From patchwork Thu Nov 2 07:12:22 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858315 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=b6f4yQG8; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLb9c6KYGz1yQt for ; Thu, 2 Nov 2023 18:29:40 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8B-0003D3-CD; Thu, 02 Nov 2023 03:28:43 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS89-0003Bo-8P; Thu, 02 Nov 2023 03:28:41 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS85-0003H2-B7; Thu, 02 Nov 2023 03:28:41 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910117; x=1730446117; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=WcudhB5q7MGJ5qgLSh9aRR0eBPwPGEKs8NT4Lk0tffQ=; b=b6f4yQG8/9Z52l5Qwdg5ud9mBzIFFRB+VGZaAlpdrQnqFqQ2vAYs/YlM DzCK494WgAdV8aaERpWpiFTmliI7prARBFUhXaRT8mYAtpex+bFUsRIJC 8cnay6odcxhtn4hMg7YFLkzXDfT1XW6DFlz/hWhOwiE4yhGpEIETA6BI5 rW7MFnGhvyQF6rIindcbkyo34CNHURiW1VOGpgrtSc7Lagysq/MEhoK4Q GX6C8ZZ1tOGHSYQCC6A6WVr0S4NcfQeiL/lm02a8/JD+e4tn76hjzBlSG M35cx6xcfsJUiXFX6m98aGBTJ5k11bsqr3tzGOMB92inqZ0nyywWIUCVC g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452952956" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452952956" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:34 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055087" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055087" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:19 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Eric Farman , Nicholas Piggin , Daniel Henrique Barboza , =?utf-8?q?C=C3=A9dric_Le_G?= =?utf-8?q?oater?= , David Gibson , Harsh Prateek Bora , Tony Krowiak , Halil Pasic , Jason Herne , Thomas Huth , Matthew Rosato , qemu-ppc@nongnu.org (open list:sPAPR (pseries)), qemu-s390x@nongnu.org (open list:vfio-ap) Subject: [PATCH v4 01/41] vfio/container: Move IBM EEH related functions into spapr_pci_vfio.c Date: Thu, 2 Nov 2023 15:12:22 +0800 Message-Id: <20231102071302.1818071-2-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org With vfio_eeh_as_ok/vfio_eeh_as_op moved and made static, vfio.h becomes empty and is deleted. No functional changes intended. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan Acked-by: Eric Farman Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio.h | 7 --- hw/ppc/spapr_pci_vfio.c | 100 +++++++++++++++++++++++++++++++++++++++- hw/vfio/ap.c | 1 - hw/vfio/ccw.c | 1 - hw/vfio/common.c | 1 - hw/vfio/container.c | 98 --------------------------------------- hw/vfio/helpers.c | 1 - 7 files changed, 99 insertions(+), 110 deletions(-) delete mode 100644 include/hw/vfio/vfio.h diff --git a/include/hw/vfio/vfio.h b/include/hw/vfio/vfio.h deleted file mode 100644 index 86248f5436..0000000000 --- a/include/hw/vfio/vfio.h +++ /dev/null @@ -1,7 +0,0 @@ -#ifndef HW_VFIO_H -#define HW_VFIO_H - -bool vfio_eeh_as_ok(AddressSpace *as); -int vfio_eeh_as_op(AddressSpace *as, uint32_t op); - -#endif diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index 9016720547..f283f7e38d 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -18,14 +18,112 @@ */ #include "qemu/osdep.h" +#include #include #include "hw/ppc/spapr.h" #include "hw/pci-host/spapr.h" #include "hw/pci/msix.h" #include "hw/pci/pci_device.h" -#include "hw/vfio/vfio.h" +#include "hw/vfio/vfio-common.h" #include "qemu/error-report.h" +/* + * Interfaces for IBM EEH (Enhanced Error Handling) + */ +static bool vfio_eeh_container_ok(VFIOContainer *container) +{ + /* + * As of 2016-03-04 (linux-4.5) the host kernel EEH/VFIO + * implementation is broken if there are multiple groups in a + * container. The hardware works in units of Partitionable + * Endpoints (== IOMMU groups) and the EEH operations naively + * iterate across all groups in the container, without any logic + * to make sure the groups have their state synchronized. For + * certain operations (ENABLE) that might be ok, until an error + * occurs, but for others (GET_STATE) it's clearly broken. + */ + + /* + * XXX Once fixed kernels exist, test for them here + */ + + if (QLIST_EMPTY(&container->group_list)) { + return false; + } + + if (QLIST_NEXT(QLIST_FIRST(&container->group_list), container_next)) { + return false; + } + + return true; +} + +static int vfio_eeh_container_op(VFIOContainer *container, uint32_t op) +{ + struct vfio_eeh_pe_op pe_op = { + .argsz = sizeof(pe_op), + .op = op, + }; + int ret; + + if (!vfio_eeh_container_ok(container)) { + error_report("vfio/eeh: EEH_PE_OP 0x%x: " + "kernel requires a container with exactly one group", op); + return -EPERM; + } + + ret = ioctl(container->fd, VFIO_EEH_PE_OP, &pe_op); + if (ret < 0) { + error_report("vfio/eeh: EEH_PE_OP 0x%x failed: %m", op); + return -errno; + } + + return ret; +} + +static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) +{ + VFIOAddressSpace *space = vfio_get_address_space(as); + VFIOContainer *container = NULL; + + if (QLIST_EMPTY(&space->containers)) { + /* No containers to act on */ + goto out; + } + + container = QLIST_FIRST(&space->containers); + + if (QLIST_NEXT(container, next)) { + /* + * We don't yet have logic to synchronize EEH state across + * multiple containers + */ + container = NULL; + goto out; + } + +out: + vfio_put_address_space(space); + return container; +} + +static bool vfio_eeh_as_ok(AddressSpace *as) +{ + VFIOContainer *container = vfio_eeh_as_container(as); + + return (container != NULL) && vfio_eeh_container_ok(container); +} + +static int vfio_eeh_as_op(AddressSpace *as, uint32_t op) +{ + VFIOContainer *container = vfio_eeh_as_container(as); + + if (!container) { + return -ENODEV; + } + return vfio_eeh_container_op(container, op); +} + bool spapr_phb_eeh_available(SpaprPhbState *sphb) { return vfio_eeh_as_ok(&sphb->iommu_as); diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index 5f257bffb9..bbf69ff55a 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -14,7 +14,6 @@ #include #include #include "qapi/error.h" -#include "hw/vfio/vfio.h" #include "hw/vfio/vfio-common.h" #include "hw/s390x/ap-device.h" #include "qemu/error-report.h" diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index 6623ae237b..d857bb8d0f 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -20,7 +20,6 @@ #include #include "qapi/error.h" -#include "hw/vfio/vfio.h" #include "hw/vfio/vfio-common.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/vfio-ccw.h" diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 9c5c6433f2..e72055e752 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -26,7 +26,6 @@ #include #include "hw/vfio/vfio-common.h" -#include "hw/vfio/vfio.h" #include "hw/vfio/pci.h" #include "exec/address-spaces.h" #include "exec/memory.h" diff --git a/hw/vfio/container.c b/hw/vfio/container.c index fc88222377..83c0f05bba 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -26,7 +26,6 @@ #include #include "hw/vfio/vfio-common.h" -#include "hw/vfio/vfio.h" #include "exec/address-spaces.h" #include "exec/memory.h" #include "exec/ram_addr.h" @@ -1011,103 +1010,6 @@ static void vfio_put_base_device(VFIODevice *vbasedev) close(vbasedev->fd); } -/* - * Interfaces for IBM EEH (Enhanced Error Handling) - */ -static bool vfio_eeh_container_ok(VFIOContainer *container) -{ - /* - * As of 2016-03-04 (linux-4.5) the host kernel EEH/VFIO - * implementation is broken if there are multiple groups in a - * container. The hardware works in units of Partitionable - * Endpoints (== IOMMU groups) and the EEH operations naively - * iterate across all groups in the container, without any logic - * to make sure the groups have their state synchronized. For - * certain operations (ENABLE) that might be ok, until an error - * occurs, but for others (GET_STATE) it's clearly broken. - */ - - /* - * XXX Once fixed kernels exist, test for them here - */ - - if (QLIST_EMPTY(&container->group_list)) { - return false; - } - - if (QLIST_NEXT(QLIST_FIRST(&container->group_list), container_next)) { - return false; - } - - return true; -} - -static int vfio_eeh_container_op(VFIOContainer *container, uint32_t op) -{ - struct vfio_eeh_pe_op pe_op = { - .argsz = sizeof(pe_op), - .op = op, - }; - int ret; - - if (!vfio_eeh_container_ok(container)) { - error_report("vfio/eeh: EEH_PE_OP 0x%x: " - "kernel requires a container with exactly one group", op); - return -EPERM; - } - - ret = ioctl(container->fd, VFIO_EEH_PE_OP, &pe_op); - if (ret < 0) { - error_report("vfio/eeh: EEH_PE_OP 0x%x failed: %m", op); - return -errno; - } - - return ret; -} - -static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) -{ - VFIOAddressSpace *space = vfio_get_address_space(as); - VFIOContainer *container = NULL; - - if (QLIST_EMPTY(&space->containers)) { - /* No containers to act on */ - goto out; - } - - container = QLIST_FIRST(&space->containers); - - if (QLIST_NEXT(container, next)) { - /* - * We don't yet have logic to synchronize EEH state across - * multiple containers - */ - container = NULL; - goto out; - } - -out: - vfio_put_address_space(space); - return container; -} - -bool vfio_eeh_as_ok(AddressSpace *as) -{ - VFIOContainer *container = vfio_eeh_as_container(as); - - return (container != NULL) && vfio_eeh_container_ok(container); -} - -int vfio_eeh_as_op(AddressSpace *as, uint32_t op) -{ - VFIOContainer *container = vfio_eeh_as_container(as); - - if (!container) { - return -ENODEV; - } - return vfio_eeh_container_op(container, op); -} - static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) { char *tmp, group_path[PATH_MAX], *group_name; diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index 7e5da21b31..168847e7c5 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -23,7 +23,6 @@ #include #include "hw/vfio/vfio-common.h" -#include "hw/vfio/vfio.h" #include "hw/hw.h" #include "trace.h" #include "qapi/error.h" From patchwork Thu Nov 2 07:12:23 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858329 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=mdWDItKD; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbCX2LzBz1yQ4 for ; Thu, 2 Nov 2023 18:31:20 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8C-0003E6-4p; Thu, 02 Nov 2023 03:28:44 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8A-0003C9-3Q; Thu, 02 Nov 2023 03:28:42 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS87-0003HZ-T8; Thu, 02 Nov 2023 03:28:41 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910119; x=1730446119; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=qjbDkT4Dazzn57crwRpv6ay+fQ5la+aq92kPhp4CtRg=; b=mdWDItKDXnr/r/lvvi2ZjhIMOmw3FgpNGMgzjKhGU6ttz4xXiYiJ0DYb jYfLybARqOzgH4vhSKo1qQz0Z/MO7/bsG9dvoerC3LONWn6r7Q7kZ97NL bRx3kqwgZLdNKeMDdwwy3kvbWJxzw6SmWPzVMaauCKRleTMfxLEADfYCy AqJwi8qOEPKB89Vq1/Q7afnPjfeQTgYWqpwS9Y1/pVxrV+1cu46wLZtcB AwIUhlcbGIE9NtpuIvA6n9MWvTdMSoxOvxaAroNDblDblwgNBeJ0/V0LC HbQZMZQwuR0oC8fcaZ4EwY0r8SdRKTpjt+59PFvaL0FOlNDN7Vv1LDZz7 Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452952967" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452952967" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:35 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055107" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055107" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:26 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 02/41] vfio/container: Move vfio_container_add/del_section_window into spapr.c Date: Thu, 2 Nov 2023 15:12:23 +0800 Message-Id: <20231102071302.1818071-3-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org vfio_container_add/del_section_window are spapr specific functions, so move them into spapr.c to make container.c cleaner. No functional changes intended. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- hw/vfio/container.c | 90 --------------------------------------------- hw/vfio/spapr.c | 90 +++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 90 insertions(+), 90 deletions(-) diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 83c0f05bba..7a3f005d1b 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -20,9 +20,6 @@ #include "qemu/osdep.h" #include -#ifdef CONFIG_KVM -#include -#endif #include #include "hw/vfio/vfio-common.h" @@ -32,7 +29,6 @@ #include "hw/hw.h" #include "qemu/error-report.h" #include "qemu/range.h" -#include "sysemu/kvm.h" #include "sysemu/reset.h" #include "trace.h" #include "qapi/error.h" @@ -204,92 +200,6 @@ int vfio_dma_map(VFIOContainer *container, hwaddr iova, return -errno; } -int vfio_container_add_section_window(VFIOContainer *container, - MemoryRegionSection *section, - Error **errp) -{ - VFIOHostDMAWindow *hostwin; - hwaddr pgsize = 0; - int ret; - - if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { - return 0; - } - - /* For now intersections are not allowed, we may relax this later */ - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { - if (ranges_overlap(hostwin->min_iova, - hostwin->max_iova - hostwin->min_iova + 1, - section->offset_within_address_space, - int128_get64(section->size))) { - error_setg(errp, - "region [0x%"PRIx64",0x%"PRIx64"] overlaps with existing" - "host DMA window [0x%"PRIx64",0x%"PRIx64"]", - section->offset_within_address_space, - section->offset_within_address_space + - int128_get64(section->size) - 1, - hostwin->min_iova, hostwin->max_iova); - return -EINVAL; - } - } - - ret = vfio_spapr_create_window(container, section, &pgsize); - if (ret) { - error_setg_errno(errp, -ret, "Failed to create SPAPR window"); - return ret; - } - - vfio_host_win_add(container, section->offset_within_address_space, - section->offset_within_address_space + - int128_get64(section->size) - 1, pgsize); -#ifdef CONFIG_KVM - if (kvm_enabled()) { - VFIOGroup *group; - IOMMUMemoryRegion *iommu_mr = IOMMU_MEMORY_REGION(section->mr); - struct kvm_vfio_spapr_tce param; - struct kvm_device_attr attr = { - .group = KVM_DEV_VFIO_GROUP, - .attr = KVM_DEV_VFIO_GROUP_SET_SPAPR_TCE, - .addr = (uint64_t)(unsigned long)¶m, - }; - - if (!memory_region_iommu_get_attr(iommu_mr, IOMMU_ATTR_SPAPR_TCE_FD, - ¶m.tablefd)) { - QLIST_FOREACH(group, &container->group_list, container_next) { - param.groupfd = group->fd; - if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { - error_setg_errno(errp, errno, - "vfio: failed GROUP_SET_SPAPR_TCE for " - "KVM VFIO device %d and group fd %d", - param.tablefd, param.groupfd); - return -errno; - } - trace_vfio_spapr_group_attach(param.groupfd, param.tablefd); - } - } - } -#endif - return 0; -} - -void vfio_container_del_section_window(VFIOContainer *container, - MemoryRegionSection *section) -{ - if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { - return; - } - - vfio_spapr_remove_window(container, - section->offset_within_address_space); - if (vfio_host_win_del(container, - section->offset_within_address_space, - section->offset_within_address_space + - int128_get64(section->size) - 1) < 0) { - hw_error("%s: Cannot delete missing window at %"HWADDR_PRIx, - __func__, section->offset_within_address_space); - } -} - int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start) { int ret; diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 9ec1e95f6d..9a7517c042 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -11,6 +11,10 @@ #include "qemu/osdep.h" #include #include +#ifdef CONFIG_KVM +#include +#endif +#include "sysemu/kvm.h" #include "hw/vfio/vfio-common.h" #include "hw/hw.h" @@ -253,3 +257,89 @@ int vfio_spapr_remove_window(VFIOContainer *container, return 0; } + +int vfio_container_add_section_window(VFIOContainer *container, + MemoryRegionSection *section, + Error **errp) +{ + VFIOHostDMAWindow *hostwin; + hwaddr pgsize = 0; + int ret; + + if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { + return 0; + } + + /* For now intersections are not allowed, we may relax this later */ + QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + if (ranges_overlap(hostwin->min_iova, + hostwin->max_iova - hostwin->min_iova + 1, + section->offset_within_address_space, + int128_get64(section->size))) { + error_setg(errp, + "region [0x%"PRIx64",0x%"PRIx64"] overlaps with existing" + "host DMA window [0x%"PRIx64",0x%"PRIx64"]", + section->offset_within_address_space, + section->offset_within_address_space + + int128_get64(section->size) - 1, + hostwin->min_iova, hostwin->max_iova); + return -EINVAL; + } + } + + ret = vfio_spapr_create_window(container, section, &pgsize); + if (ret) { + error_setg_errno(errp, -ret, "Failed to create SPAPR window"); + return ret; + } + + vfio_host_win_add(container, section->offset_within_address_space, + section->offset_within_address_space + + int128_get64(section->size) - 1, pgsize); +#ifdef CONFIG_KVM + if (kvm_enabled()) { + VFIOGroup *group; + IOMMUMemoryRegion *iommu_mr = IOMMU_MEMORY_REGION(section->mr); + struct kvm_vfio_spapr_tce param; + struct kvm_device_attr attr = { + .group = KVM_DEV_VFIO_GROUP, + .attr = KVM_DEV_VFIO_GROUP_SET_SPAPR_TCE, + .addr = (uint64_t)(unsigned long)¶m, + }; + + if (!memory_region_iommu_get_attr(iommu_mr, IOMMU_ATTR_SPAPR_TCE_FD, + ¶m.tablefd)) { + QLIST_FOREACH(group, &container->group_list, container_next) { + param.groupfd = group->fd; + if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { + error_setg_errno(errp, errno, + "vfio: failed GROUP_SET_SPAPR_TCE for " + "KVM VFIO device %d and group fd %d", + param.tablefd, param.groupfd); + return -errno; + } + trace_vfio_spapr_group_attach(param.groupfd, param.tablefd); + } + } + } +#endif + return 0; +} + +void vfio_container_del_section_window(VFIOContainer *container, + MemoryRegionSection *section) +{ + if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { + return; + } + + vfio_spapr_remove_window(container, + section->offset_within_address_space); + if (vfio_host_win_del(container, + section->offset_within_address_space, + section->offset_within_address_space + + int128_get64(section->size) - 1) < 0) { + hw_error("%s: Cannot delete missing window at %"HWADDR_PRIx, + __func__, section->offset_within_address_space); + } +} From patchwork Thu Nov 2 07:12:24 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858314 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=ETvMvDyA; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLb9c3srCz1yQs for ; Thu, 2 Nov 2023 18:29:40 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8E-0003FJ-Pl; Thu, 02 Nov 2023 03:28:46 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8D-0003Ep-CG; Thu, 02 Nov 2023 03:28:45 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8A-0003H2-46; Thu, 02 Nov 2023 03:28:45 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910122; x=1730446122; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=fa1hVBkiJF9ggE12WPPLphRmMrq/8tCZaj5CDa4UGGk=; b=ETvMvDyAoNWwMldyXwAOlAcp6becLuOoCSfvESQtNYl0yYHxwaWag4lC 0v9cib9vh+6Nib1y9IWmmWKmKkTo0gw6bKgBvNuQI4Pq+Kt7ilUONnBcx fOfoE8unzQVX9XVHI5OpOXwJKZiKI0/5p/cU1o+skskLzJYolF+QRbcQM WyGX3NHxO6G9b4XCg2DEiGosJ0t7BljpE/nASh05jrhksGY5ojO+VBitl M+3eaEkmhfGiIvOB4yy6iuyDAh8TIZA6afIrPNLBAZ4eXybMxWj8sUCsy hieZlTWWj5DCGiOEa3jGEJo6pKp0wDBGuFo4Icux2uHt2hykEgJyr3CCY Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452952983" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452952983" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:37 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055122" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055122" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:33 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 03/41] vfio/container: Move spapr specific init/deinit into spapr.c Date: Thu, 2 Nov 2023 15:12:24 +0800 Message-Id: <20231102071302.1818071-4-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Move spapr specific init/deinit code into spapr.c and wrap them with vfio_spapr_container_init/deinit, this way footprint of spapr is further reduced, vfio_prereg_listener could also be made static. vfio_listener_release is unnecessary when prereg_listener is moved out, so have it removed. No functional changes intended. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 4 +- hw/vfio/container.c | 82 +++++------------------------------ hw/vfio/spapr.c | 81 +++++++++++++++++++++++++++++++++- 3 files changed, 95 insertions(+), 72 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 0c3d390e8b..ed5a8e4754 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -225,11 +225,14 @@ int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start); int vfio_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); +/* SPAPR specific */ int vfio_container_add_section_window(VFIOContainer *container, MemoryRegionSection *section, Error **errp); void vfio_container_del_section_window(VFIOContainer *container, MemoryRegionSection *section); +int vfio_spapr_container_init(VFIOContainer *container, Error **errp); +void vfio_spapr_container_deinit(VFIOContainer *container); void vfio_disable_irqindex(VFIODevice *vbasedev, int index); void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); @@ -289,7 +292,6 @@ vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); #endif -extern const MemoryListener vfio_prereg_listener; int vfio_spapr_create_window(VFIOContainer *container, MemoryRegionSection *section, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 7a3f005d1b..204b244b11 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -264,14 +264,6 @@ int vfio_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, return ret; } -static void vfio_listener_release(VFIOContainer *container) -{ - memory_listener_unregister(&container->listener); - if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU) { - memory_listener_unregister(&container->prereg_listener); - } -} - static struct vfio_info_cap_header * vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) { @@ -612,69 +604,11 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, case VFIO_SPAPR_TCE_v2_IOMMU: case VFIO_SPAPR_TCE_IOMMU: { - struct vfio_iommu_spapr_tce_info info; - bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; - - /* - * The host kernel code implementing VFIO_IOMMU_DISABLE is called - * when container fd is closed so we do not call it explicitly - * in this file. - */ - if (!v2) { - ret = ioctl(fd, VFIO_IOMMU_ENABLE); - if (ret) { - error_setg_errno(errp, errno, "failed to enable container"); - ret = -errno; - goto enable_discards_exit; - } - } else { - container->prereg_listener = vfio_prereg_listener; - - memory_listener_register(&container->prereg_listener, - &address_space_memory); - if (container->error) { - memory_listener_unregister(&container->prereg_listener); - ret = -1; - error_propagate_prepend(errp, container->error, - "RAM memory listener initialization failed: "); - goto enable_discards_exit; - } - } - - info.argsz = sizeof(info); - ret = ioctl(fd, VFIO_IOMMU_SPAPR_TCE_GET_INFO, &info); + ret = vfio_spapr_container_init(container, errp); if (ret) { - error_setg_errno(errp, errno, - "VFIO_IOMMU_SPAPR_TCE_GET_INFO failed"); - ret = -errno; - if (v2) { - memory_listener_unregister(&container->prereg_listener); - } goto enable_discards_exit; } - - if (v2) { - container->pgsizes = info.ddw.pgsizes; - /* - * There is a default window in just created container. - * To make region_add/del simpler, we better remove this - * window now and let those iommu_listener callbacks - * create/remove them when needed. - */ - ret = vfio_spapr_remove_window(container, info.dma32_window_start); - if (ret) { - error_setg_errno(errp, -ret, - "failed to remove existing window"); - goto enable_discards_exit; - } - } else { - /* The default table uses 4K pages */ - container->pgsizes = 0x1000; - vfio_host_win_add(container, info.dma32_window_start, - info.dma32_window_start + - info.dma32_window_size - 1, - 0x1000); - } + break; } } @@ -704,7 +638,11 @@ listener_release_exit: QLIST_REMOVE(group, container_next); QLIST_REMOVE(container, next); vfio_kvm_device_del_group(group); - vfio_listener_release(container); + memory_listener_unregister(&container->listener); + if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU || + container->iommu_type == VFIO_SPAPR_TCE_IOMMU) { + vfio_spapr_container_deinit(container); + } enable_discards_exit: vfio_ram_block_discard_disable(container, false); @@ -734,7 +672,11 @@ static void vfio_disconnect_container(VFIOGroup *group) * group. */ if (QLIST_EMPTY(&container->group_list)) { - vfio_listener_release(container); + memory_listener_unregister(&container->listener); + if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU || + container->iommu_type == VFIO_SPAPR_TCE_IOMMU) { + vfio_spapr_container_deinit(container); + } } if (ioctl(group->fd, VFIO_GROUP_UNSET_CONTAINER, &container->fd)) { diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 9a7517c042..00dbd7af11 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -15,6 +15,7 @@ #include #endif #include "sysemu/kvm.h" +#include "exec/address-spaces.h" #include "hw/vfio/vfio-common.h" #include "hw/hw.h" @@ -139,7 +140,7 @@ static void vfio_prereg_listener_region_del(MemoryListener *listener, trace_vfio_prereg_unregister(reg.vaddr, reg.size, ret ? -errno : 0); } -const MemoryListener vfio_prereg_listener = { +static const MemoryListener vfio_prereg_listener = { .name = "vfio-pre-reg", .region_add = vfio_prereg_listener_region_add, .region_del = vfio_prereg_listener_region_del, @@ -343,3 +344,81 @@ void vfio_container_del_section_window(VFIOContainer *container, __func__, section->offset_within_address_space); } } + +int vfio_spapr_container_init(VFIOContainer *container, Error **errp) +{ + struct vfio_iommu_spapr_tce_info info; + bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; + int ret, fd = container->fd; + + /* + * The host kernel code implementing VFIO_IOMMU_DISABLE is called + * when container fd is closed so we do not call it explicitly + * in this file. + */ + if (!v2) { + ret = ioctl(fd, VFIO_IOMMU_ENABLE); + if (ret) { + error_setg_errno(errp, errno, "failed to enable container"); + return -errno; + } + } else { + container->prereg_listener = vfio_prereg_listener; + + memory_listener_register(&container->prereg_listener, + &address_space_memory); + if (container->error) { + ret = -1; + error_propagate_prepend(errp, container->error, + "RAM memory listener initialization failed: "); + goto listener_unregister_exit; + } + } + + info.argsz = sizeof(info); + ret = ioctl(fd, VFIO_IOMMU_SPAPR_TCE_GET_INFO, &info); + if (ret) { + error_setg_errno(errp, errno, + "VFIO_IOMMU_SPAPR_TCE_GET_INFO failed"); + ret = -errno; + goto listener_unregister_exit; + } + + if (v2) { + container->pgsizes = info.ddw.pgsizes; + /* + * There is a default window in just created container. + * To make region_add/del simpler, we better remove this + * window now and let those iommu_listener callbacks + * create/remove them when needed. + */ + ret = vfio_spapr_remove_window(container, info.dma32_window_start); + if (ret) { + error_setg_errno(errp, -ret, + "failed to remove existing window"); + goto listener_unregister_exit; + } + } else { + /* The default table uses 4K pages */ + container->pgsizes = 0x1000; + vfio_host_win_add(container, info.dma32_window_start, + info.dma32_window_start + + info.dma32_window_size - 1, + 0x1000); + } + + return 0; + +listener_unregister_exit: + if (v2) { + memory_listener_unregister(&container->prereg_listener); + } + return ret; +} + +void vfio_spapr_container_deinit(VFIOContainer *container) +{ + if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU) { + memory_listener_unregister(&container->prereg_listener); + } +} From patchwork Thu Nov 2 07:12:25 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858357 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=SagubfKy; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKw5zVVz1yQq for ; Thu, 2 Nov 2023 18:36:52 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8H-0003Hf-BG; Thu, 02 Nov 2023 03:28:49 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8F-0003GI-QJ; Thu, 02 Nov 2023 03:28:47 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8E-0003IK-6t; Thu, 02 Nov 2023 03:28:47 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910126; x=1730446126; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=QwLk7ZuVKFYm0XZNyQpHNFnmiD1IGAidoZKA/z51gVc=; b=SagubfKyJYhpbIp6fC6Ly3nbDT29QxuUAf3iwwFngRiNDqGwAnp8FGCL Fus6+1vqYuRTz20wG+q52ADPhKSWTyaxCw0qVsS45+zyc1O716CE1cs/n ZVOavPx3ZXy+5s0QQbX6Cq0lPZZY3Ya/5hiMQoSxi9yIYJdQyRppwNQTL KdgXCjgyDR31abQbG6MhTTQ/2v7cK0DB5gGtM/6v41PnqBTUq0/+Iaqqb +o0ojDyzLYbHJ7t79VI9dc+PkuVlP1yP4k76A5z4JE2w3pjigiptMwG4T rP9UYWe3vPc7+Rh9bD9IT1ID7cz0gd7OPnmrzxifLA9cZwHKNRuSdcm+n A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953034" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953034" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:43 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055150" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055150" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:37 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , =?utf-8?q?C=C3=A9dric_Le_G?= =?utf-8?q?oater?= , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 04/41] vfio/spapr: Make vfio_spapr_create/remove_window static Date: Thu, 2 Nov 2023 15:12:25 +0800 Message-Id: <20231102071302.1818071-5-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org vfio_spapr_create_window calls vfio_spapr_remove_window, With reoder of definition of the two, we can make vfio_spapr_create/remove_window static. No functional changes intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 6 ----- hw/vfio/spapr.c | 48 +++++++++++++++++------------------ 2 files changed, 24 insertions(+), 30 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index ed5a8e4754..87848982bd 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -293,12 +293,6 @@ struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); #endif -int vfio_spapr_create_window(VFIOContainer *container, - MemoryRegionSection *section, - hwaddr *pgsize); -int vfio_spapr_remove_window(VFIOContainer *container, - hwaddr offset_within_address_space); - bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); void vfio_migration_exit(VFIODevice *vbasedev); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 00dbd7af11..4428990c28 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -146,9 +146,30 @@ static const MemoryListener vfio_prereg_listener = { .region_del = vfio_prereg_listener_region_del, }; -int vfio_spapr_create_window(VFIOContainer *container, - MemoryRegionSection *section, - hwaddr *pgsize) +static int vfio_spapr_remove_window(VFIOContainer *container, + hwaddr offset_within_address_space) +{ + struct vfio_iommu_spapr_tce_remove remove = { + .argsz = sizeof(remove), + .start_addr = offset_within_address_space, + }; + int ret; + + ret = ioctl(container->fd, VFIO_IOMMU_SPAPR_TCE_REMOVE, &remove); + if (ret) { + error_report("Failed to remove window at %"PRIx64, + (uint64_t)remove.start_addr); + return -errno; + } + + trace_vfio_spapr_remove_window(offset_within_address_space); + + return 0; +} + +static int vfio_spapr_create_window(VFIOContainer *container, + MemoryRegionSection *section, + hwaddr *pgsize) { int ret = 0; IOMMUMemoryRegion *iommu_mr = IOMMU_MEMORY_REGION(section->mr); @@ -238,27 +259,6 @@ int vfio_spapr_create_window(VFIOContainer *container, return 0; } -int vfio_spapr_remove_window(VFIOContainer *container, - hwaddr offset_within_address_space) -{ - struct vfio_iommu_spapr_tce_remove remove = { - .argsz = sizeof(remove), - .start_addr = offset_within_address_space, - }; - int ret; - - ret = ioctl(container->fd, VFIO_IOMMU_SPAPR_TCE_REMOVE, &remove); - if (ret) { - error_report("Failed to remove window at %"PRIx64, - (uint64_t)remove.start_addr); - return -errno; - } - - trace_vfio_spapr_remove_window(offset_within_address_space); - - return 0; -} - int vfio_container_add_section_window(VFIOContainer *container, MemoryRegionSection *section, Error **errp) From patchwork Thu Nov 2 07:12:26 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858328 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=POMA85ah; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbCQ4H6pz1yQ4 for ; Thu, 2 Nov 2023 18:31:14 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8N-0003Ir-FU; Thu, 02 Nov 2023 03:28:55 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8K-0003IN-0x; Thu, 02 Nov 2023 03:28:52 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8H-0003Iu-TK; Thu, 02 Nov 2023 03:28:51 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910129; x=1730446129; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=7PD2tQefFsLc6jUwf7HpZ+5DjIGGtQEznVqGXW4nFtU=; b=POMA85ahe5F1/S/AtUZ0Bcfsxd2IcyG46oclveo7ljgwmVYZhtbxsjpr P8pYk9SR/UlaPFUS2s6Upcvo2vVf1CLbdAWBCbcdTWAjdwaMEpmsY+we6 cSHxlT3xtLUL1oyoOMHCLXxMJHTo1a8wYQfs970hOGhD7CVT/9hQTQSl0 avteHVvwpnLKKkEdN+wcXtK4iKJpUdJmBpTa5RJWnUsCZaHYaEiHB4Upy JXILi6y2fzKFc9DVOVvdCC9qyneO9hhM9eiRk4+p430fU/+OYRCKDYtOk Ejla+SsDwlDNosIoWBxqQH8t9DT9/3jUoNCUE05cPj1ZLaXP49e+KqZrM g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953076" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953076" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:47 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055185" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055185" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:43 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 05/41] vfio/common: Move vfio_host_win_add/del into spapr.c Date: Thu, 2 Nov 2023 15:12:26 +0800 Message-Id: <20231102071302.1818071-6-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Only spapr supports a customed host window list, other vfio driver assume 64bit host window. So remove the check in listener callback and move vfio_host_win_add/del into spapr.c and make it static. With the check removed, we still need to do the same check for VFIO_SPAPR_TCE_IOMMU which allows a single host window range [dma32_window_start, dma32_window_size). Move vfio_find_hostwin into spapr.c and do same check in vfio_container_add_section_window instead. When mapping a ram device section, if it's unaligned with hostwin->iova_pgsizes, this mapping is bypassed. With hostwin moved into spapr, we changed to check container->pgsizes. Suggested-by: Alex Williamson Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: add vfio_find_hostwin back for VFIO_SPAPR_TCE_IOMMU include/hw/vfio/vfio-common.h | 5 --- hw/vfio/common.c | 70 +---------------------------- hw/vfio/container.c | 16 ------- hw/vfio/spapr.c | 83 +++++++++++++++++++++++++++++++++++ 4 files changed, 85 insertions(+), 89 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 87848982bd..a4a22accb9 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -207,11 +207,6 @@ typedef struct { hwaddr pages; } VFIOBitmap; -void vfio_host_win_add(VFIOContainer *container, - hwaddr min_iova, hwaddr max_iova, - uint64_t iova_pgsizes); -int vfio_host_win_del(VFIOContainer *container, hwaddr min_iova, - hwaddr max_iova); VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); bool vfio_devices_all_running_and_saving(VFIOContainer *container); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index e72055e752..e70fdf5e0c 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -245,44 +245,6 @@ bool vfio_devices_all_running_and_mig_active(VFIOContainer *container) return true; } -void vfio_host_win_add(VFIOContainer *container, hwaddr min_iova, - hwaddr max_iova, uint64_t iova_pgsizes) -{ - VFIOHostDMAWindow *hostwin; - - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { - if (ranges_overlap(hostwin->min_iova, - hostwin->max_iova - hostwin->min_iova + 1, - min_iova, - max_iova - min_iova + 1)) { - hw_error("%s: Overlapped IOMMU are not enabled", __func__); - } - } - - hostwin = g_malloc0(sizeof(*hostwin)); - - hostwin->min_iova = min_iova; - hostwin->max_iova = max_iova; - hostwin->iova_pgsizes = iova_pgsizes; - QLIST_INSERT_HEAD(&container->hostwin_list, hostwin, hostwin_next); -} - -int vfio_host_win_del(VFIOContainer *container, - hwaddr min_iova, hwaddr max_iova) -{ - VFIOHostDMAWindow *hostwin; - - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { - if (hostwin->min_iova == min_iova && hostwin->max_iova == max_iova) { - QLIST_REMOVE(hostwin, hostwin_next); - g_free(hostwin); - return 0; - } - } - - return -1; -} - static bool vfio_listener_skipped_section(MemoryRegionSection *section) { return (!memory_region_is_ram(section->mr) && @@ -531,22 +493,6 @@ static void vfio_unregister_ram_discard_listener(VFIOContainer *container, g_free(vrdl); } -static VFIOHostDMAWindow *vfio_find_hostwin(VFIOContainer *container, - hwaddr iova, hwaddr end) -{ - VFIOHostDMAWindow *hostwin; - bool hostwin_found = false; - - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { - if (hostwin->min_iova <= iova && end <= hostwin->max_iova) { - hostwin_found = true; - break; - } - } - - return hostwin_found ? hostwin : NULL; -} - static bool vfio_known_safe_misalignment(MemoryRegionSection *section) { MemoryRegion *mr = section->mr; @@ -625,7 +571,6 @@ static void vfio_listener_region_add(MemoryListener *listener, Int128 llend, llsize; void *vaddr; int ret; - VFIOHostDMAWindow *hostwin; Error *err = NULL; if (!vfio_listener_valid_section(section, "region_add")) { @@ -647,13 +592,6 @@ static void vfio_listener_region_add(MemoryListener *listener, goto fail; } - hostwin = vfio_find_hostwin(container, iova, end); - if (!hostwin) { - error_setg(&err, "Container %p can't map guest IOVA region" - " 0x%"HWADDR_PRIx"..0x%"HWADDR_PRIx, container, iova, end); - goto fail; - } - memory_region_ref(section->mr); if (memory_region_is_iommu(section->mr)) { @@ -734,7 +672,7 @@ static void vfio_listener_region_add(MemoryListener *listener, llsize = int128_sub(llend, int128_make64(iova)); if (memory_region_is_ram_device(section->mr)) { - hwaddr pgmask = (1ULL << ctz64(hostwin->iova_pgsizes)) - 1; + hwaddr pgmask = (1ULL << ctz64(container->pgsizes)) - 1; if ((iova & pgmask) || (int128_get64(llsize) & pgmask)) { trace_vfio_listener_region_add_no_dma_map( @@ -833,12 +771,8 @@ static void vfio_listener_region_del(MemoryListener *listener, if (memory_region_is_ram_device(section->mr)) { hwaddr pgmask; - VFIOHostDMAWindow *hostwin; - - hostwin = vfio_find_hostwin(container, iova, end); - assert(hostwin); /* or region_add() would have failed */ - pgmask = (1ULL << ctz64(hostwin->iova_pgsizes)) - 1; + pgmask = (1ULL << ctz64(container->pgsizes)) - 1; try_unmap = !((iova & pgmask) || (int128_get64(llsize) & pgmask)); } else if (memory_region_has_ram_discard_manager(section->mr)) { vfio_unregister_ram_discard_listener(container, section); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 204b244b11..242010036a 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -551,7 +551,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container->dma_max_mappings = 0; container->iova_ranges = NULL; QLIST_INIT(&container->giommu_list); - QLIST_INIT(&container->hostwin_list); QLIST_INIT(&container->vrdl_list); ret = vfio_init_container(container, group->fd, errp); @@ -591,14 +590,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, vfio_get_iommu_info_migration(container, info); g_free(info); - - /* - * FIXME: We should parse VFIO_IOMMU_TYPE1_INFO_CAP_IOVA_RANGE - * information to get the actual window extent rather than assume - * a 64-bit IOVA address space. - */ - vfio_host_win_add(container, 0, (hwaddr)-1, container->pgsizes); - break; } case VFIO_SPAPR_TCE_v2_IOMMU: @@ -687,7 +678,6 @@ static void vfio_disconnect_container(VFIOGroup *group) if (QLIST_EMPTY(&container->group_list)) { VFIOAddressSpace *space = container->space; VFIOGuestIOMMU *giommu, *tmp; - VFIOHostDMAWindow *hostwin, *next; QLIST_REMOVE(container, next); @@ -698,12 +688,6 @@ static void vfio_disconnect_container(VFIOGroup *group) g_free(giommu); } - QLIST_FOREACH_SAFE(hostwin, &container->hostwin_list, hostwin_next, - next) { - QLIST_REMOVE(hostwin, hostwin_next); - g_free(hostwin); - } - trace_vfio_disconnect_container(container->fd); close(container->fd); vfio_free_container(container); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 4428990c28..83da2f7ec2 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -146,6 +146,60 @@ static const MemoryListener vfio_prereg_listener = { .region_del = vfio_prereg_listener_region_del, }; +static void vfio_host_win_add(VFIOContainer *container, hwaddr min_iova, + hwaddr max_iova, uint64_t iova_pgsizes) +{ + VFIOHostDMAWindow *hostwin; + + QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + if (ranges_overlap(hostwin->min_iova, + hostwin->max_iova - hostwin->min_iova + 1, + min_iova, + max_iova - min_iova + 1)) { + hw_error("%s: Overlapped IOMMU are not enabled", __func__); + } + } + + hostwin = g_malloc0(sizeof(*hostwin)); + + hostwin->min_iova = min_iova; + hostwin->max_iova = max_iova; + hostwin->iova_pgsizes = iova_pgsizes; + QLIST_INSERT_HEAD(&container->hostwin_list, hostwin, hostwin_next); +} + +static int vfio_host_win_del(VFIOContainer *container, + hwaddr min_iova, hwaddr max_iova) +{ + VFIOHostDMAWindow *hostwin; + + QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + if (hostwin->min_iova == min_iova && hostwin->max_iova == max_iova) { + QLIST_REMOVE(hostwin, hostwin_next); + g_free(hostwin); + return 0; + } + } + + return -1; +} + +static VFIOHostDMAWindow *vfio_find_hostwin(VFIOContainer *container, + hwaddr iova, hwaddr end) +{ + VFIOHostDMAWindow *hostwin; + bool hostwin_found = false; + + QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + if (hostwin->min_iova <= iova && end <= hostwin->max_iova) { + hostwin_found = true; + break; + } + } + + return hostwin_found ? hostwin : NULL; +} + static int vfio_spapr_remove_window(VFIOContainer *container, hwaddr offset_within_address_space) { @@ -267,6 +321,26 @@ int vfio_container_add_section_window(VFIOContainer *container, hwaddr pgsize = 0; int ret; + /* + * VFIO_SPAPR_TCE_IOMMU supports a single host window between + * [dma32_window_start, dma32_window_size), we need to ensure + * the section fall in this range. + */ + if (container->iommu_type == VFIO_SPAPR_TCE_IOMMU) { + hwaddr iova, end; + + iova = section->offset_within_address_space; + end = iova + int128_get64(section->size) - 1; + + if (!vfio_find_hostwin(container, iova, end)) { + error_setg(errp, "Container %p can't map guest IOVA region" + " 0x%"HWADDR_PRIx"..0x%"HWADDR_PRIx, container, + iova, end); + return -EINVAL; + } + return 0; + } + if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { return 0; } @@ -351,6 +425,8 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; int ret, fd = container->fd; + QLIST_INIT(&container->hostwin_list); + /* * The host kernel code implementing VFIO_IOMMU_DISABLE is called * when container fd is closed so we do not call it explicitly @@ -418,7 +494,14 @@ listener_unregister_exit: void vfio_spapr_container_deinit(VFIOContainer *container) { + VFIOHostDMAWindow *hostwin, *next; + if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU) { memory_listener_unregister(&container->prereg_listener); } + QLIST_FOREACH_SAFE(hostwin, &container->hostwin_list, hostwin_next, + next) { + QLIST_REMOVE(hostwin, hostwin_next); + g_free(hostwin); + } } From patchwork Thu Nov 2 07:12:27 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858340 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=BihX+ofB; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbFb2RrNz1yQ4 for ; Thu, 2 Nov 2023 18:33:07 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8Q-0003Jf-3c; Thu, 02 Nov 2023 03:28:58 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8M-0003Ij-MV for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:28:54 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8K-0003Iu-VR for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:28:54 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910133; x=1730446133; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=VW5pFPVjDwvzR62I3j8pzdNJf0iC/PM3XBdsI01VM1Q=; b=BihX+ofBWkQL2FSRMPi+vCXrSeUGObcYBbyO5V3ODZj9G6xPFC6Jb4xs gPOqRt1sFYJSvYDhjbkHMmptMRz8ZenKvVur6rMcliEYMHc+Kng/rSxt7 1ZdK1mgIknhcuP2mYgxNuangWKUX1CZ4u0113qIlpIwwB0h4wxSGTFPBD saX5Bnwz0XFoCuH1+Wr/89z+V0X8+OfGUuDWrMq1jvLnsZG4hlpv3R3pu lX6JF3HnlFiSEnXhPTTKfPLYK1Bfe+CYGgvfVVK+tNebWMwr2dT3jFbyZ OBukkjVjS8JAth6j5ff9wU1LXLSp9rv6G7y+mBPS3liNMzi1m6e+F8Wn6 w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953095" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953095" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:52 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055209" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055209" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:47 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Yi Sun Subject: [PATCH v4 06/41] vfio: Introduce base object for VFIOContainer and targeted interface Date: Thu, 2 Nov 2023 15:12:27 +0800 Message-Id: <20231102071302.1818071-7-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Introduce a dumb VFIOContainerBase object and its targeted interface. This is willingly not a QOM object because we don't want it to be visible from the user interface. The VFIOContainerBase will be smoothly populated in subsequent patches as well as interfaces. No fucntional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: use SPDX identifier, use const char *name parameter, HW_VFIO_VFIO_CONTAINER_BASE_H include/hw/vfio/vfio-common.h | 8 ++--- include/hw/vfio/vfio-container-base.h | 50 +++++++++++++++++++++++++++ 2 files changed, 52 insertions(+), 6 deletions(-) create mode 100644 include/hw/vfio/vfio-container-base.h diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index a4a22accb9..586d153c12 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -30,6 +30,7 @@ #include #endif #include "sysemu/sysemu.h" +#include "hw/vfio/vfio-container-base.h" #define VFIO_MSG_PREFIX "vfio %s: " @@ -81,6 +82,7 @@ typedef struct VFIOAddressSpace { struct VFIOGroup; typedef struct VFIOContainer { + VFIOContainerBase bcontainer; VFIOAddressSpace *space; int fd; /* /dev/vfio/vfio, empowered by the attached groups */ MemoryListener listener; @@ -201,12 +203,6 @@ typedef struct VFIODisplay { } dmabuf; } VFIODisplay; -typedef struct { - unsigned long *bitmap; - hwaddr size; - hwaddr pages; -} VFIOBitmap; - VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); bool vfio_devices_all_running_and_saving(VFIOContainer *container); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h new file mode 100644 index 0000000000..1d6daaea5d --- /dev/null +++ b/include/hw/vfio/vfio-container-base.h @@ -0,0 +1,50 @@ +/* + * VFIO BASE CONTAINER + * + * Copyright (C) 2023 Intel Corporation. + * Copyright Red Hat, Inc. 2023 + * + * Authors: Yi Liu + * Eric Auger + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_CONTAINER_BASE_H +#define HW_VFIO_VFIO_CONTAINER_BASE_H + +#include "exec/memory.h" + +typedef struct VFIODevice VFIODevice; +typedef struct VFIOIOMMUOps VFIOIOMMUOps; + +typedef struct { + unsigned long *bitmap; + hwaddr size; + hwaddr pages; +} VFIOBitmap; + +/* + * This is the base object for vfio container backends + */ +typedef struct VFIOContainerBase { + const VFIOIOMMUOps *ops; +} VFIOContainerBase; + +struct VFIOIOMMUOps { + /* basic feature */ + int (*dma_map)(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + void *vaddr, bool readonly); + int (*dma_unmap)(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + IOMMUTLBEntry *iotlb); + int (*attach_device)(const char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp); + void (*detach_device)(VFIODevice *vbasedev); + /* migration feature */ + int (*set_dirty_page_tracking)(VFIOContainerBase *bcontainer, bool start); + int (*query_dirty_bitmap)(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, + hwaddr iova, hwaddr size); +}; +#endif /* HW_VFIO_VFIO_CONTAINER_BASE_H */ From patchwork Thu Nov 2 07:12:28 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858324 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=Sh6rDlBf; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBs2y80z1yQ4 for ; Thu, 2 Nov 2023 18:30:45 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8R-0003KG-O5; Thu, 02 Nov 2023 03:28:59 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8Q-0003K1-Dl for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:28:58 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8O-0003Iu-QE for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:28:58 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910136; x=1730446136; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=aqP0d+Qvoe4ddCUwjKIo+Uc/4pcxFrWytxjiZV15aVY=; b=Sh6rDlBfpYKw7Z3m/ZDz2myoj7MMw2QNwvUT1JfuadRaX9V6xiSbRoC/ 2c3eZcyH/HPF7JvLEEILe0jHvmc1zzDDPD4iAbie4yjA76ISlP81fnJIB 9zg/Dgzq+b7G8H+5hPLagwRoVJwUbem0OMp/QrsxEPQOotxChj6cPWzaj poOHXrhA41OaNPRANd98i9m2lgAIMKXYzuKXGQYznTlJFxHPVkxSX7PYc x1F1NEN4Ue/hETw/9rhUBHVDix5XlogmQGKD66OSz8dTx4ge+YID1CcNd xTQep2SXPAnFcUrn06HpSdXAe5DfmbqNFFiB+WZBSQOvcvbm1uBUF04EK g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953108" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953108" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:56 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055232" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055232" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:52 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 07/41] vfio/container: Introduce a empty VFIOIOMMUOps Date: Thu, 2 Nov 2023 15:12:28 +0800 Message-Id: <20231102071302.1818071-8-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This empty VFIOIOMMUOps named vfio_legacy_ops will hold all general IOMMU ops of legacy container. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 2 +- hw/vfio/container.c | 5 +++++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 586d153c12..678161f207 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -255,7 +255,7 @@ typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIOGroupList vfio_group_list; extern VFIODeviceList vfio_device_list; - +extern const VFIOIOMMUOps vfio_legacy_ops; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 242010036a..4bc43ddfa4 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -472,6 +472,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, Error **errp) { VFIOContainer *container; + VFIOContainerBase *bcontainer; int ret, fd; VFIOAddressSpace *space; @@ -552,6 +553,8 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container->iova_ranges = NULL; QLIST_INIT(&container->giommu_list); QLIST_INIT(&container->vrdl_list); + bcontainer = &container->bcontainer; + bcontainer->ops = &vfio_legacy_ops; ret = vfio_init_container(container, group->fd, errp); if (ret) { @@ -933,3 +936,5 @@ void vfio_detach_device(VFIODevice *vbasedev) vfio_put_base_device(vbasedev); vfio_put_group(group); } + +const VFIOIOMMUOps vfio_legacy_ops; From patchwork Thu Nov 2 07:12:29 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858342 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=TRSPeo/f; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbGg2Wlwz1yQ4 for ; Thu, 2 Nov 2023 18:34:03 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8X-0003Ky-BP; Thu, 02 Nov 2023 03:29:05 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8V-0003Ke-Em for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:03 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8T-0003Iu-62 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:03 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910141; x=1730446141; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=yHD71MDi4qkLHgDSxl+cotqqPedL+szvPH1FioSrrDE=; b=TRSPeo/fzm/qDV4F8tmp4qqBVwrZTIs3H0VCKWedKWdk9WTTTIIKs/St 7EjRFozNqYeEj411mdnLb0wHRTStfM6B2ON8L6512v6pFwO7PaEPyE6ej v7duEE9xekO3HZxWORASakkncBiPn0Z/WrLYIZ/ZwitohknC56zIi5zOm nSmsLRJt6KBc3xa5AN9drOo4VdLktcr86i8V9dsEzZqCFrAT/663q/RnE 7jB4pqmbcZ4g9mazGDmyCUmM5S5Lg9cqoH+m/4EyorlJH/3ihd9TOK7Ys /QxIYSDQnc9ur/zJdBHR19HRgj0kA6KJcRs0VdMrhlIWsWtQ0jZKjkHUY A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953113" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953113" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:00 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055256" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055256" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:28:56 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan Subject: [PATCH v4 08/41] vfio/container: Switch to dma_map|unmap API Date: Thu, 2 Nov 2023 15:12:29 +0800 Message-Id: <20231102071302.1818071-9-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger No fucntional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: use SPDX identifier, use assert include/hw/vfio/vfio-common.h | 4 --- include/hw/vfio/vfio-container-base.h | 7 +++++ hw/vfio/common.c | 45 +++++++++++++++------------ hw/vfio/container-base.c | 32 +++++++++++++++++++ hw/vfio/container.c | 22 ++++++++----- hw/vfio/meson.build | 1 + hw/vfio/trace-events | 2 +- 7 files changed, 81 insertions(+), 32 deletions(-) create mode 100644 hw/vfio/container-base.c diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 678161f207..24a26345e5 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -208,10 +208,6 @@ void vfio_put_address_space(VFIOAddressSpace *space); bool vfio_devices_all_running_and_saving(VFIOContainer *container); /* container->fd */ -int vfio_dma_unmap(VFIOContainer *container, hwaddr iova, - ram_addr_t size, IOMMUTLBEntry *iotlb); -int vfio_dma_map(VFIOContainer *container, hwaddr iova, - ram_addr_t size, void *vaddr, bool readonly); int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start); int vfio_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 1d6daaea5d..56b033f59f 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -31,6 +31,13 @@ typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; } VFIOContainerBase; +int vfio_container_dma_map(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + void *vaddr, bool readonly); +int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + IOMMUTLBEntry *iotlb); + struct VFIOIOMMUOps { /* basic feature */ int (*dma_map)(VFIOContainerBase *bcontainer, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index e70fdf5e0c..e610771888 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -292,7 +292,7 @@ static bool vfio_get_xlat_addr(IOMMUTLBEntry *iotlb, void **vaddr, static void vfio_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) { VFIOGuestIOMMU *giommu = container_of(n, VFIOGuestIOMMU, n); - VFIOContainer *container = giommu->container; + VFIOContainerBase *bcontainer = &giommu->container->bcontainer; hwaddr iova = iotlb->iova + giommu->iommu_offset; void *vaddr; int ret; @@ -322,21 +322,22 @@ static void vfio_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) * of vaddr will always be there, even if the memory object is * destroyed and its backing memory munmap-ed. */ - ret = vfio_dma_map(container, iova, - iotlb->addr_mask + 1, vaddr, - read_only); + ret = vfio_container_dma_map(bcontainer, iova, + iotlb->addr_mask + 1, vaddr, + read_only); if (ret) { - error_report("vfio_dma_map(%p, 0x%"HWADDR_PRIx", " + error_report("vfio_container_dma_map(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx", %p) = %d (%s)", - container, iova, + bcontainer, iova, iotlb->addr_mask + 1, vaddr, ret, strerror(-ret)); } } else { - ret = vfio_dma_unmap(container, iova, iotlb->addr_mask + 1, iotlb); + ret = vfio_container_dma_unmap(bcontainer, iova, + iotlb->addr_mask + 1, iotlb); if (ret) { - error_report("vfio_dma_unmap(%p, 0x%"HWADDR_PRIx", " + error_report("vfio_container_dma_unmap(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", - container, iova, + bcontainer, iova, iotlb->addr_mask + 1, ret, strerror(-ret)); vfio_set_migration_error(ret); } @@ -355,9 +356,10 @@ static void vfio_ram_discard_notify_discard(RamDiscardListener *rdl, int ret; /* Unmap with a single call. */ - ret = vfio_dma_unmap(vrdl->container, iova, size , NULL); + ret = vfio_container_dma_unmap(&vrdl->container->bcontainer, + iova, size , NULL); if (ret) { - error_report("%s: vfio_dma_unmap() failed: %s", __func__, + error_report("%s: vfio_container_dma_unmap() failed: %s", __func__, strerror(-ret)); } } @@ -385,8 +387,8 @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, section->offset_within_address_space; vaddr = memory_region_get_ram_ptr(section->mr) + start; - ret = vfio_dma_map(vrdl->container, iova, next - start, - vaddr, section->readonly); + ret = vfio_container_dma_map(&vrdl->container->bcontainer, iova, + next - start, vaddr, section->readonly); if (ret) { /* Rollback */ vfio_ram_discard_notify_discard(rdl, section); @@ -684,10 +686,11 @@ static void vfio_listener_region_add(MemoryListener *listener, } } - ret = vfio_dma_map(container, iova, int128_get64(llsize), - vaddr, section->readonly); + ret = vfio_container_dma_map(&container->bcontainer, + iova, int128_get64(llsize), vaddr, + section->readonly); if (ret) { - error_setg(&err, "vfio_dma_map(%p, 0x%"HWADDR_PRIx", " + error_setg(&err, "vfio_container_dma_map(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx", %p) = %d (%s)", container, iova, int128_get64(llsize), vaddr, ret, strerror(-ret)); @@ -784,18 +787,20 @@ static void vfio_listener_region_del(MemoryListener *listener, if (int128_eq(llsize, int128_2_64())) { /* The unmap ioctl doesn't accept a full 64-bit span. */ llsize = int128_rshift(llsize, 1); - ret = vfio_dma_unmap(container, iova, int128_get64(llsize), NULL); + ret = vfio_container_dma_unmap(&container->bcontainer, iova, + int128_get64(llsize), NULL); if (ret) { - error_report("vfio_dma_unmap(%p, 0x%"HWADDR_PRIx", " + error_report("vfio_container_dma_unmap(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", container, iova, int128_get64(llsize), ret, strerror(-ret)); } iova += int128_get64(llsize); } - ret = vfio_dma_unmap(container, iova, int128_get64(llsize), NULL); + ret = vfio_container_dma_unmap(&container->bcontainer, iova, + int128_get64(llsize), NULL); if (ret) { - error_report("vfio_dma_unmap(%p, 0x%"HWADDR_PRIx", " + error_report("vfio_container_dma_unmap(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", container, iova, int128_get64(llsize), ret, strerror(-ret)); diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c new file mode 100644 index 0000000000..55d3a35fa4 --- /dev/null +++ b/hw/vfio/container-base.c @@ -0,0 +1,32 @@ +/* + * VFIO BASE CONTAINER + * + * Copyright (C) 2023 Intel Corporation. + * Copyright Red Hat, Inc. 2023 + * + * Authors: Yi Liu + * Eric Auger + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "hw/vfio/vfio-container-base.h" + +int vfio_container_dma_map(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + void *vaddr, bool readonly) +{ + g_assert(bcontainer->ops->dma_map); + return bcontainer->ops->dma_map(bcontainer, iova, size, vaddr, readonly); +} + +int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + IOMMUTLBEntry *iotlb) +{ + g_assert(bcontainer->ops->dma_unmap); + return bcontainer->ops->dma_unmap(bcontainer, iova, size, iotlb); +} diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 4bc43ddfa4..c04df26323 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -115,9 +115,11 @@ unmap_exit: /* * DMA - Mapping and unmapping for the "type1" IOMMU interface used on x86 */ -int vfio_dma_unmap(VFIOContainer *container, hwaddr iova, - ram_addr_t size, IOMMUTLBEntry *iotlb) +static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, + ram_addr_t size, IOMMUTLBEntry *iotlb) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dma_unmap unmap = { .argsz = sizeof(unmap), .flags = 0, @@ -151,7 +153,7 @@ int vfio_dma_unmap(VFIOContainer *container, hwaddr iova, */ if (errno == EINVAL && unmap.size && !(unmap.iova + unmap.size) && container->iommu_type == VFIO_TYPE1v2_IOMMU) { - trace_vfio_dma_unmap_overflow_workaround(); + trace_vfio_legacy_dma_unmap_overflow_workaround(); unmap.size -= 1ULL << ctz64(container->pgsizes); continue; } @@ -170,9 +172,11 @@ int vfio_dma_unmap(VFIOContainer *container, hwaddr iova, return 0; } -int vfio_dma_map(VFIOContainer *container, hwaddr iova, - ram_addr_t size, void *vaddr, bool readonly) +static int vfio_legacy_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, + ram_addr_t size, void *vaddr, bool readonly) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dma_map map = { .argsz = sizeof(map), .flags = VFIO_DMA_MAP_FLAG_READ, @@ -191,7 +195,8 @@ int vfio_dma_map(VFIOContainer *container, hwaddr iova, * the VGA ROM space. */ if (ioctl(container->fd, VFIO_IOMMU_MAP_DMA, &map) == 0 || - (errno == EBUSY && vfio_dma_unmap(container, iova, size, NULL) == 0 && + (errno == EBUSY && + vfio_legacy_dma_unmap(bcontainer, iova, size, NULL) == 0 && ioctl(container->fd, VFIO_IOMMU_MAP_DMA, &map) == 0)) { return 0; } @@ -937,4 +942,7 @@ void vfio_detach_device(VFIODevice *vbasedev) vfio_put_group(group); } -const VFIOIOMMUOps vfio_legacy_ops; +const VFIOIOMMUOps vfio_legacy_ops = { + .dma_map = vfio_legacy_dma_map, + .dma_unmap = vfio_legacy_dma_unmap, +}; diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index 2a6912c940..eb6ce6229d 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -2,6 +2,7 @@ vfio_ss = ss.source_set() vfio_ss.add(files( 'helpers.c', 'common.c', + 'container-base.c', 'container.c', 'spapr.c', 'migration.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index 0eb2387cf2..9f7fedee98 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -116,7 +116,7 @@ vfio_region_unmap(const char *name, unsigned long offset, unsigned long end) "Re vfio_region_sparse_mmap_header(const char *name, int index, int nr_areas) "Device %s region %d: %d sparse mmap entries" vfio_region_sparse_mmap_entry(int i, unsigned long start, unsigned long end) "sparse entry %d [0x%lx - 0x%lx]" vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -vfio_dma_unmap_overflow_workaround(void) "" +vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(int fd, uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "container fd=%d, iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 From patchwork Thu Nov 2 07:12:30 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858344 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=DaBpSiZD; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbGl72mnz1yQs for ; Thu, 2 Nov 2023 18:34:07 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8b-0003LZ-7v; Thu, 02 Nov 2023 03:29:09 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8Z-0003LC-Es for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:07 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8X-0003Iu-P4 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:07 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910145; x=1730446145; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=fUm8JL0SeQgChdWgcZIV4ZbFjiGIcOSVo1Keh6d/NJA=; b=DaBpSiZD3/rM7XdVn5wKaP5zS1GXyW2X8ZOssAK5nE7jvFpylMCCb6Dz cWbFnG2vEj0a0jByxRrVkFk7JwcmkJjgTWMfNapZVIVIicsCN1KZ2wMdG w2Ttiii0K5fFVJFjDycxcD72pwOD1cST21F90X/znwhUe5eqeNxjiVvLF /4Rp4SxIP7xhatPX3FCALqvbUhLGDkTTuXwyewcWJpuLsZZ8oXuxVHLuc BdmrE79EXj2uONokmHop+ILAoGRKYRKsCoEL4fcJjzSh00cmyl4De56aN cJM/aUbFwdOJmYTohP/kY72J8QizB9yth/5zbi9L2zSeeAMi0O/b8H/Ux g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953118" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953118" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:05 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055282" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055282" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:00 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 09/41] vfio/common: Introduce vfio_container_init/destroy helper Date: Thu, 2 Nov 2023 15:12:30 +0800 Message-Id: <20231102071302.1818071-10-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This adds two helper functions vfio_container_init/destroy which will be used by both legacy and iommufd containers to do base container specific initialization and release. No fucntional change intended. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-container-base.h | 4 ++++ hw/vfio/container-base.c | 9 +++++++++ hw/vfio/container.c | 4 +++- 3 files changed, 16 insertions(+), 1 deletion(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 56b033f59f..577f52ccbc 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -38,6 +38,10 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb); +void vfio_container_init(VFIOContainerBase *bcontainer, + const VFIOIOMMUOps *ops); +void vfio_container_destroy(VFIOContainerBase *bcontainer); + struct VFIOIOMMUOps { /* basic feature */ int (*dma_map)(VFIOContainerBase *bcontainer, diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 55d3a35fa4..e929435751 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -30,3 +30,12 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, g_assert(bcontainer->ops->dma_unmap); return bcontainer->ops->dma_unmap(bcontainer, iova, size, iotlb); } + +void vfio_container_init(VFIOContainerBase *bcontainer, const VFIOIOMMUOps *ops) +{ + bcontainer->ops = ops; +} + +void vfio_container_destroy(VFIOContainerBase *bcontainer) +{ +} diff --git a/hw/vfio/container.c b/hw/vfio/container.c index c04df26323..32a0251dd1 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -559,7 +559,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, QLIST_INIT(&container->giommu_list); QLIST_INIT(&container->vrdl_list); bcontainer = &container->bcontainer; - bcontainer->ops = &vfio_legacy_ops; + vfio_container_init(bcontainer, &vfio_legacy_ops); ret = vfio_init_container(container, group->fd, errp); if (ret) { @@ -661,6 +661,7 @@ put_space_exit: static void vfio_disconnect_container(VFIOGroup *group) { VFIOContainer *container = group->container; + VFIOContainerBase *bcontainer = &container->bcontainer; QLIST_REMOVE(group, container_next); group->container = NULL; @@ -695,6 +696,7 @@ static void vfio_disconnect_container(VFIOGroup *group) QLIST_REMOVE(giommu, giommu_next); g_free(giommu); } + vfio_container_destroy(bcontainer); trace_vfio_disconnect_container(container->fd); close(container->fd); From patchwork Thu Nov 2 07:12:31 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858313 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=UucAgDNm; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLb9c3YKbz1yQ6 for ; Thu, 2 Nov 2023 18:29:40 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8f-0003MA-TA; Thu, 02 Nov 2023 03:29:13 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8d-0003Lj-Un for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:11 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8a-0003Iu-VY for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:10 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910149; x=1730446149; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=wBFdYQwS8ayEuu44wagwcT+o6v+ysmUkaKsw6smiBH0=; b=UucAgDNmR0dfXKxXUM/QRe7aUdTN8guJw63ZVyUyysU5fUAAKtZBr/Rc 5kthv2lotz9Yy3ZD2RCp8l4FDFXmLA5hgpACMFSbg3XqRCy7Idu0t7396 rPoon3gi8k24VQLCvL0XeyrKV9VQLg6N3a5HMo7Lju4/w1LRY7WrTD+BH P8F3T2qj+LRcfeFSPznyrJkwCkIDg26fGkXncwZtFOqCiukpSF2DKQdWM YVPV1tbdH1105yHZCyWLuJzSlzrevyCj8e1qvc03Gzr6L4LO+iokErSuz g/pgwi8SSmDqlwIPhHgbqLxfxa/xjitcsnvFo2WY5SzTV3W2W1P1Zq9io Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953124" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953124" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:08 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055300" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055300" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:04 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan Subject: [PATCH v4 10/41] vfio/common: Move giommu_list in base container Date: Thu, 2 Nov 2023 15:12:31 +0800 Message-Id: <20231102071302.1818071-11-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger Move the giommu_list field in the base container and store the base container in the VFIOGuestIOMMU. No functional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 9 --------- include/hw/vfio/vfio-container-base.h | 9 +++++++++ hw/vfio/common.c | 17 +++++++++++------ hw/vfio/container-base.c | 9 +++++++++ hw/vfio/container.c | 8 -------- 5 files changed, 29 insertions(+), 23 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 24a26345e5..6be082b8f2 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -95,7 +95,6 @@ typedef struct VFIOContainer { uint64_t max_dirty_bitmap_size; unsigned long pgsizes; unsigned int dma_max_mappings; - QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; @@ -104,14 +103,6 @@ typedef struct VFIOContainer { GList *iova_ranges; } VFIOContainer; -typedef struct VFIOGuestIOMMU { - VFIOContainer *container; - IOMMUMemoryRegion *iommu_mr; - hwaddr iommu_offset; - IOMMUNotifier n; - QLIST_ENTRY(VFIOGuestIOMMU) giommu_next; -} VFIOGuestIOMMU; - typedef struct VFIORamDiscardListener { VFIOContainer *container; MemoryRegion *mr; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 577f52ccbc..a11aec5755 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -29,8 +29,17 @@ typedef struct { */ typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; + QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; } VFIOContainerBase; +typedef struct VFIOGuestIOMMU { + VFIOContainerBase *bcontainer; + IOMMUMemoryRegion *iommu_mr; + hwaddr iommu_offset; + IOMMUNotifier n; + QLIST_ENTRY(VFIOGuestIOMMU) giommu_next; +} VFIOGuestIOMMU; + int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index e610771888..43580bcc43 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -292,7 +292,7 @@ static bool vfio_get_xlat_addr(IOMMUTLBEntry *iotlb, void **vaddr, static void vfio_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) { VFIOGuestIOMMU *giommu = container_of(n, VFIOGuestIOMMU, n); - VFIOContainerBase *bcontainer = &giommu->container->bcontainer; + VFIOContainerBase *bcontainer = giommu->bcontainer; hwaddr iova = iotlb->iova + giommu->iommu_offset; void *vaddr; int ret; @@ -569,6 +569,7 @@ static void vfio_listener_region_add(MemoryListener *listener, MemoryRegionSection *section) { VFIOContainer *container = container_of(listener, VFIOContainer, listener); + VFIOContainerBase *bcontainer = &container->bcontainer; hwaddr iova, end; Int128 llend, llsize; void *vaddr; @@ -612,7 +613,7 @@ static void vfio_listener_region_add(MemoryListener *listener, giommu->iommu_mr = iommu_mr; giommu->iommu_offset = section->offset_within_address_space - section->offset_within_region; - giommu->container = container; + giommu->bcontainer = bcontainer; llend = int128_add(int128_make64(section->offset_within_region), section->size); llend = int128_sub(llend, int128_one()); @@ -647,7 +648,7 @@ static void vfio_listener_region_add(MemoryListener *listener, g_free(giommu); goto fail; } - QLIST_INSERT_HEAD(&container->giommu_list, giommu, giommu_next); + QLIST_INSERT_HEAD(&bcontainer->giommu_list, giommu, giommu_next); memory_region_iommu_replay(giommu->iommu_mr, &giommu->n); return; @@ -732,6 +733,7 @@ static void vfio_listener_region_del(MemoryListener *listener, MemoryRegionSection *section) { VFIOContainer *container = container_of(listener, VFIOContainer, listener); + VFIOContainerBase *bcontainer = &container->bcontainer; hwaddr iova, end; Int128 llend, llsize; int ret; @@ -744,7 +746,7 @@ static void vfio_listener_region_del(MemoryListener *listener, if (memory_region_is_iommu(section->mr)) { VFIOGuestIOMMU *giommu; - QLIST_FOREACH(giommu, &container->giommu_list, giommu_next) { + QLIST_FOREACH(giommu, &bcontainer->giommu_list, giommu_next) { if (MEMORY_REGION(giommu->iommu_mr) == section->mr && giommu->n.start == section->offset_within_region) { memory_region_unregister_iommu_notifier(section->mr, @@ -1206,7 +1208,9 @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) vfio_giommu_dirty_notifier *gdn = container_of(n, vfio_giommu_dirty_notifier, n); VFIOGuestIOMMU *giommu = gdn->giommu; - VFIOContainer *container = giommu->container; + VFIOContainerBase *bcontainer = giommu->bcontainer; + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); hwaddr iova = iotlb->iova + giommu->iommu_offset; ram_addr_t translated_addr; int ret = -EINVAL; @@ -1284,12 +1288,13 @@ static int vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainer *container, static int vfio_sync_dirty_bitmap(VFIOContainer *container, MemoryRegionSection *section) { + VFIOContainerBase *bcontainer = &container->bcontainer; ram_addr_t ram_addr; if (memory_region_is_iommu(section->mr)) { VFIOGuestIOMMU *giommu; - QLIST_FOREACH(giommu, &container->giommu_list, giommu_next) { + QLIST_FOREACH(giommu, &bcontainer->giommu_list, giommu_next) { if (MEMORY_REGION(giommu->iommu_mr) == section->mr && giommu->n.start == section->offset_within_region) { Int128 llend; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index e929435751..20bcb9669a 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -34,8 +34,17 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, void vfio_container_init(VFIOContainerBase *bcontainer, const VFIOIOMMUOps *ops) { bcontainer->ops = ops; + QLIST_INIT(&bcontainer->giommu_list); } void vfio_container_destroy(VFIOContainerBase *bcontainer) { + VFIOGuestIOMMU *giommu, *tmp; + + QLIST_FOREACH_SAFE(giommu, &bcontainer->giommu_list, giommu_next, tmp) { + memory_region_unregister_iommu_notifier( + MEMORY_REGION(giommu->iommu_mr), &giommu->n); + QLIST_REMOVE(giommu, giommu_next); + g_free(giommu); + } } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 32a0251dd1..133d3c8f5c 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -556,7 +556,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container->dirty_pages_supported = false; container->dma_max_mappings = 0; container->iova_ranges = NULL; - QLIST_INIT(&container->giommu_list); QLIST_INIT(&container->vrdl_list); bcontainer = &container->bcontainer; vfio_container_init(bcontainer, &vfio_legacy_ops); @@ -686,16 +685,9 @@ static void vfio_disconnect_container(VFIOGroup *group) if (QLIST_EMPTY(&container->group_list)) { VFIOAddressSpace *space = container->space; - VFIOGuestIOMMU *giommu, *tmp; QLIST_REMOVE(container, next); - QLIST_FOREACH_SAFE(giommu, &container->giommu_list, giommu_next, tmp) { - memory_region_unregister_iommu_notifier( - MEMORY_REGION(giommu->iommu_mr), &giommu->n); - QLIST_REMOVE(giommu, giommu_next); - g_free(giommu); - } vfio_container_destroy(bcontainer); trace_vfio_disconnect_container(container->fd); From patchwork Thu Nov 2 07:12:32 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858331 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=aXynzagv; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbCg55Pnz1yQs for ; Thu, 2 Nov 2023 18:31:27 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8k-0003Q1-TL; Thu, 02 Nov 2023 03:29:18 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8j-0003PI-4S; Thu, 02 Nov 2023 03:29:17 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8h-0003Iu-0h; Thu, 02 Nov 2023 03:29:16 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910155; x=1730446155; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=XOCOsjJC353dGK656kRHn7Xra3X3VewuBrMK5UsfgY8=; b=aXynzagvSpfNOMCaA7ZvvDRLtlj/8cICsNBVxJjOx83gckX3X+Wh9+hV 9cG8Zy6NvEB7As53Q+htDKPKdABLCGTyp1kPIW4laNtzCC3828Q6CRp6g Kr4lAz0gEKk72VFlH9zwo9cr7nPOjFNmZxxd1uH87h2wNDCd4xwqxk2vM NQRdw/Iv4SFP/iucXeZxtkcfeIt0bt4cjRQh8LgVcB7lQdKtxiymrCWqh oQnPxA4fhCL9489ftgzQsqcJ8WphRe29jGxZSsD7ViLW7S50h+jnggiIW srah2CD2soF05dSWBp+Yu1JXyE5s2Y0Jve1au6yx4wn9Ra5mxvkpUg30h Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953129" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953129" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:13 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055323" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055323" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:08 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , =?utf-8?q?C=C3=A9dric_Le_G?= =?utf-8?q?oater?= , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 11/41] vfio/container: Move space field to base container Date: Thu, 2 Nov 2023 15:12:32 +0800 Message-Id: <20231102071302.1818071-12-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger Move the space field to the base object. Also the VFIOAddressSpace now contains a list of base containers. No fucntional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: use bcontainer->space->as instead of container->bcontainer.space->as include/hw/vfio/vfio-common.h | 8 -------- include/hw/vfio/vfio-container-base.h | 9 +++++++++ hw/ppc/spapr_pci_vfio.c | 10 +++++----- hw/vfio/common.c | 4 ++-- hw/vfio/container-base.c | 6 +++++- hw/vfio/container.c | 18 ++++++++---------- 6 files changed, 29 insertions(+), 26 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 6be082b8f2..bd4de6cb3a 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -73,17 +73,10 @@ typedef struct VFIOMigration { bool initial_data_sent; } VFIOMigration; -typedef struct VFIOAddressSpace { - AddressSpace *as; - QLIST_HEAD(, VFIOContainer) containers; - QLIST_ENTRY(VFIOAddressSpace) list; -} VFIOAddressSpace; - struct VFIOGroup; typedef struct VFIOContainer { VFIOContainerBase bcontainer; - VFIOAddressSpace *space; int fd; /* /dev/vfio/vfio, empowered by the attached groups */ MemoryListener listener; MemoryListener prereg_listener; @@ -98,7 +91,6 @@ typedef struct VFIOContainer { QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; - QLIST_ENTRY(VFIOContainer) next; QLIST_HEAD(, VFIODevice) device_list; GList *iova_ranges; } VFIOContainer; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index a11aec5755..c7cc6ec9c5 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -24,12 +24,20 @@ typedef struct { hwaddr pages; } VFIOBitmap; +typedef struct VFIOAddressSpace { + AddressSpace *as; + QLIST_HEAD(, VFIOContainerBase) containers; + QLIST_ENTRY(VFIOAddressSpace) list; +} VFIOAddressSpace; + /* * This is the base object for vfio container backends */ typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; + VFIOAddressSpace *space; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; + QLIST_ENTRY(VFIOContainerBase) next; } VFIOContainerBase; typedef struct VFIOGuestIOMMU { @@ -48,6 +56,7 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, IOMMUTLBEntry *iotlb); void vfio_container_init(VFIOContainerBase *bcontainer, + VFIOAddressSpace *space, const VFIOIOMMUOps *ops); void vfio_container_destroy(VFIOContainerBase *bcontainer); diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index f283f7e38d..d1d07bec46 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -84,27 +84,27 @@ static int vfio_eeh_container_op(VFIOContainer *container, uint32_t op) static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) { VFIOAddressSpace *space = vfio_get_address_space(as); - VFIOContainer *container = NULL; + VFIOContainerBase *bcontainer = NULL; if (QLIST_EMPTY(&space->containers)) { /* No containers to act on */ goto out; } - container = QLIST_FIRST(&space->containers); + bcontainer = QLIST_FIRST(&space->containers); - if (QLIST_NEXT(container, next)) { + if (QLIST_NEXT(bcontainer, next)) { /* * We don't yet have logic to synchronize EEH state across * multiple containers */ - container = NULL; + bcontainer = NULL; goto out; } out: vfio_put_address_space(space); - return container; + return container_of(bcontainer, VFIOContainer, bcontainer); } static bool vfio_eeh_as_ok(AddressSpace *as) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 43580bcc43..1d8202537e 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -145,7 +145,7 @@ void vfio_unblock_multiple_devices_migration(void) bool vfio_viommu_preset(VFIODevice *vbasedev) { - return vbasedev->container->space->as != &address_space_memory; + return vbasedev->container->bcontainer.space->as != &address_space_memory; } static void vfio_set_migration_error(int err) @@ -922,7 +922,7 @@ static void vfio_dirty_tracking_init(VFIOContainer *container, dirty.container = container; memory_listener_register(&dirty.listener, - container->space->as); + container->bcontainer.space->as); *ranges = dirty.ranges; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 20bcb9669a..3933391e0d 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -31,9 +31,11 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, return bcontainer->ops->dma_unmap(bcontainer, iova, size, iotlb); } -void vfio_container_init(VFIOContainerBase *bcontainer, const VFIOIOMMUOps *ops) +void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, + const VFIOIOMMUOps *ops) { bcontainer->ops = ops; + bcontainer->space = space; QLIST_INIT(&bcontainer->giommu_list); } @@ -41,6 +43,8 @@ void vfio_container_destroy(VFIOContainerBase *bcontainer) { VFIOGuestIOMMU *giommu, *tmp; + QLIST_REMOVE(bcontainer, next); + QLIST_FOREACH_SAFE(giommu, &bcontainer->giommu_list, giommu_next, tmp) { memory_region_unregister_iommu_notifier( MEMORY_REGION(giommu->iommu_mr), &giommu->n); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 133d3c8f5c..f12fcb6fe1 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -514,7 +514,8 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, * details once we know which type of IOMMU we are using. */ - QLIST_FOREACH(container, &space->containers, next) { + QLIST_FOREACH(bcontainer, &space->containers, next) { + container = container_of(bcontainer, VFIOContainer, bcontainer); if (!ioctl(group->fd, VFIO_GROUP_SET_CONTAINER, &container->fd)) { ret = vfio_ram_block_discard_disable(container, true); if (ret) { @@ -550,7 +551,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, } container = g_malloc0(sizeof(*container)); - container->space = space; container->fd = fd; container->error = NULL; container->dirty_pages_supported = false; @@ -558,7 +558,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container->iova_ranges = NULL; QLIST_INIT(&container->vrdl_list); bcontainer = &container->bcontainer; - vfio_container_init(bcontainer, &vfio_legacy_ops); + vfio_container_init(bcontainer, space, &vfio_legacy_ops); ret = vfio_init_container(container, group->fd, errp); if (ret) { @@ -613,14 +613,14 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, vfio_kvm_device_add_group(group); QLIST_INIT(&container->group_list); - QLIST_INSERT_HEAD(&space->containers, container, next); + QLIST_INSERT_HEAD(&space->containers, bcontainer, next); group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); container->listener = vfio_memory_listener; - memory_listener_register(&container->listener, container->space->as); + memory_listener_register(&container->listener, bcontainer->space->as); if (container->error) { ret = -1; @@ -634,7 +634,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, return 0; listener_release_exit: QLIST_REMOVE(group, container_next); - QLIST_REMOVE(container, next); + QLIST_REMOVE(bcontainer, next); vfio_kvm_device_del_group(group); memory_listener_unregister(&container->listener); if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU || @@ -684,9 +684,7 @@ static void vfio_disconnect_container(VFIOGroup *group) } if (QLIST_EMPTY(&container->group_list)) { - VFIOAddressSpace *space = container->space; - - QLIST_REMOVE(container, next); + VFIOAddressSpace *space = bcontainer->space; vfio_container_destroy(bcontainer); @@ -707,7 +705,7 @@ static VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp) QLIST_FOREACH(group, &vfio_group_list, next) { if (group->groupid == groupid) { /* Found it. Now is it already in the right context? */ - if (group->container->space->as == as) { + if (group->container->bcontainer.space->as == as) { return group; } else { error_setg(errp, "group %d used in multiple address spaces", From patchwork Thu Nov 2 07:12:33 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858338 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=ZYCveTV/; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbDz3D9dz1yQ4 for ; Thu, 2 Nov 2023 18:32:35 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8o-0003Qh-Eo; Thu, 02 Nov 2023 03:29:22 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8m-0003QW-Fh for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:20 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8k-0003Iu-Az for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:20 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910158; x=1730446158; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=ctq67vD+99Xopap7tH2d0vKzlrZ5Ssejmx6yfYitmAk=; b=ZYCveTV/ZGPmpM7LZCjMk4XjE5XzCzwP3Wo8jsk0fRSmTHnQHRI11gLJ U1j8bWodAb39aSYjzOdoJfQyfDqaxHV1KrlOjvfx0dWXUPJoAp/tLsyAL JzbtieXaWP1uywDLqksFCujsMCnr/Hzyjrw/pQvPmFdAj2lmQ6GeZYEeZ L0wSQ7fn27NMBKjbXx+1E7x0uD+r6H+tP/ONnRmODw/3gBaBqbLR3P5EI 2XRj4micSMBFyrvIe07tS292EhLUtyaEuBRhxqV2jlYpARpSPl8+bxQBg L7cWq/f6wmG7N088dBa7bTpzmuPu/L6FXfVhFTQVWG7INAkcg/Jx9aPX7 Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953144" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953144" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:17 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055339" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055339" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:13 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan Subject: [PATCH v4 12/41] vfio/container: Switch to IOMMU BE set_dirty_page_tracking/query_dirty_bitmap API Date: Thu, 2 Nov 2023 15:12:33 +0800 Message-Id: <20231102071302.1818071-13-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger dirty_pages_supported field is also moved to the base container No fucntional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: use assert include/hw/vfio/vfio-common.h | 6 ------ include/hw/vfio/vfio-container-base.h | 6 ++++++ hw/vfio/common.c | 12 ++++++++---- hw/vfio/container-base.c | 16 ++++++++++++++++ hw/vfio/container.c | 21 ++++++++++++++------- 5 files changed, 44 insertions(+), 17 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index bd4de6cb3a..60f2785fe0 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -83,7 +83,6 @@ typedef struct VFIOContainer { unsigned iommu_type; Error *error; bool initialized; - bool dirty_pages_supported; uint64_t dirty_pgsizes; uint64_t max_dirty_bitmap_size; unsigned long pgsizes; @@ -190,11 +189,6 @@ VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); bool vfio_devices_all_running_and_saving(VFIOContainer *container); -/* container->fd */ -int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start); -int vfio_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, - hwaddr iova, hwaddr size); - /* SPAPR specific */ int vfio_container_add_section_window(VFIOContainer *container, MemoryRegionSection *section, diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index c7cc6ec9c5..f244f003d0 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -36,6 +36,7 @@ typedef struct VFIOAddressSpace { typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; VFIOAddressSpace *space; + bool dirty_pages_supported; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; QLIST_ENTRY(VFIOContainerBase) next; } VFIOContainerBase; @@ -54,6 +55,11 @@ int vfio_container_dma_map(VFIOContainerBase *bcontainer, int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb); +int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, + bool start); +int vfio_container_query_dirty_bitmap(VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, + hwaddr iova, hwaddr size); void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 1d8202537e..b1a875ca93 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -1079,7 +1079,8 @@ static void vfio_listener_log_global_start(MemoryListener *listener) if (vfio_devices_all_device_dirty_tracking(container)) { ret = vfio_devices_dma_logging_start(container); } else { - ret = vfio_set_dirty_page_tracking(container, true); + ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, + true); } if (ret) { @@ -1097,7 +1098,8 @@ static void vfio_listener_log_global_stop(MemoryListener *listener) if (vfio_devices_all_device_dirty_tracking(container)) { vfio_devices_dma_logging_stop(container); } else { - ret = vfio_set_dirty_page_tracking(container, false); + ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, + false); } if (ret) { @@ -1165,7 +1167,8 @@ int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, VFIOBitmap vbmap; int ret; - if (!container->dirty_pages_supported && !all_device_dirty_tracking) { + if (!container->bcontainer.dirty_pages_supported && + !all_device_dirty_tracking) { cpu_physical_memory_set_dirty_range(ram_addr, size, tcg_enabled() ? DIRTY_CLIENTS_ALL : DIRTY_CLIENTS_NOCODE); @@ -1180,7 +1183,8 @@ int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, if (all_device_dirty_tracking) { ret = vfio_devices_query_dirty_bitmap(container, &vbmap, iova, size); } else { - ret = vfio_query_dirty_bitmap(container, &vbmap, iova, size); + ret = vfio_container_query_dirty_bitmap(&container->bcontainer, &vbmap, + iova, size); } if (ret) { diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 3933391e0d..5d654ae172 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -31,11 +31,27 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, return bcontainer->ops->dma_unmap(bcontainer, iova, size, iotlb); } +int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, + bool start) +{ + g_assert(bcontainer->ops->set_dirty_page_tracking); + return bcontainer->ops->set_dirty_page_tracking(bcontainer, start); +} + +int vfio_container_query_dirty_bitmap(VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, + hwaddr iova, hwaddr size) +{ + g_assert(bcontainer->ops->query_dirty_bitmap); + return bcontainer->ops->query_dirty_bitmap(bcontainer, vbmap, iova, size); +} + void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, const VFIOIOMMUOps *ops) { bcontainer->ops = ops; bcontainer->space = space; + bcontainer->dirty_pages_supported = false; QLIST_INIT(&bcontainer->giommu_list); } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index f12fcb6fe1..3ab74e2615 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -131,7 +131,7 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, if (iotlb && vfio_devices_all_running_and_mig_active(container)) { if (!vfio_devices_all_device_dirty_tracking(container) && - container->dirty_pages_supported) { + container->bcontainer.dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); } @@ -205,14 +205,17 @@ static int vfio_legacy_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, return -errno; } -int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start) +static int vfio_legacy_set_dirty_page_tracking(VFIOContainerBase *bcontainer, + bool start) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); int ret; struct vfio_iommu_type1_dirty_bitmap dirty = { .argsz = sizeof(dirty), }; - if (!container->dirty_pages_supported) { + if (!bcontainer->dirty_pages_supported) { return 0; } @@ -232,9 +235,12 @@ int vfio_set_dirty_page_tracking(VFIOContainer *container, bool start) return ret; } -int vfio_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, - hwaddr iova, hwaddr size) +static int vfio_legacy_query_dirty_bitmap(VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, + hwaddr iova, hwaddr size) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dirty_bitmap *dbitmap; struct vfio_iommu_type1_dirty_bitmap_get *range; int ret; @@ -461,7 +467,7 @@ static void vfio_get_iommu_info_migration(VFIOContainer *container, * qemu_real_host_page_size to mark those dirty. */ if (cap_mig->pgsize_bitmap & qemu_real_host_page_size()) { - container->dirty_pages_supported = true; + container->bcontainer.dirty_pages_supported = true; container->max_dirty_bitmap_size = cap_mig->max_dirty_bitmap_size; container->dirty_pgsizes = cap_mig->pgsize_bitmap; } @@ -553,7 +559,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container = g_malloc0(sizeof(*container)); container->fd = fd; container->error = NULL; - container->dirty_pages_supported = false; container->dma_max_mappings = 0; container->iova_ranges = NULL; QLIST_INIT(&container->vrdl_list); @@ -937,4 +942,6 @@ void vfio_detach_device(VFIODevice *vbasedev) const VFIOIOMMUOps vfio_legacy_ops = { .dma_map = vfio_legacy_dma_map, .dma_unmap = vfio_legacy_dma_unmap, + .set_dirty_page_tracking = vfio_legacy_set_dirty_page_tracking, + .query_dirty_bitmap = vfio_legacy_query_dirty_bitmap, }; From patchwork Thu Nov 2 07:12:34 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858319 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=SgjeZBJf; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBC4pYHz1yQ4 for ; Thu, 2 Nov 2023 18:30:11 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8r-0003RP-2r; Thu, 02 Nov 2023 03:29:25 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8q-0003RD-AP for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:24 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8o-0003Iu-CX for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:24 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910162; x=1730446162; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=/k6nU6bkA/1Q7aVeLCfHu1LES7excWL5i7FhEWgQWBQ=; b=SgjeZBJfmC4+l01hpbr+7QkSyjizrSX9KibpXM/XtoebcDBsfnKVwS9R l6rwsvhvl+y4CWpzzcoL8O4Xp561acmirqrR1rPfhhjLHk3qicmAPyTT+ 2rAj1wmPX5HjYlETPTpVu/1ITRADFYfijFwco0tl0OE+8cXD26/EI7mK+ 4G1btpBQUWxf4WXzqgYr7aiNkCNhVo4PwHcIpgeiO2iSunLdlRzovwDpA r7HH6XxCbfFa/a6mZD+OFFow1ElJ8K2Rh9Iqq+/ZjpPrlOa8BPuOLxl1e 2qLSBlnr6LUBeO46BIuEOelRrkEhXBpodaGYh4ApfdZpo7DE+XIkOJbUk g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953153" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953153" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:21 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055364" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055364" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:17 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 13/41] vfio/container: Move per container device list in base container Date: Thu, 2 Nov 2023 15:12:34 +0800 Message-Id: <20231102071302.1818071-14-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org VFIO Device is also changed to point to base container instead of legacy container. No fucntional change intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 3 +-- include/hw/vfio/vfio-container-base.h | 1 + hw/vfio/common.c | 23 +++++++++++++++-------- hw/vfio/container.c | 12 ++++++------ 4 files changed, 23 insertions(+), 16 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 60f2785fe0..9740cf9fbc 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -90,7 +90,6 @@ typedef struct VFIOContainer { QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; - QLIST_HEAD(, VFIODevice) device_list; GList *iova_ranges; } VFIOContainer; @@ -118,7 +117,7 @@ typedef struct VFIODevice { QLIST_ENTRY(VFIODevice) container_next; QLIST_ENTRY(VFIODevice) global_next; struct VFIOGroup *group; - VFIOContainer *container; + VFIOContainerBase *bcontainer; char *sysfsdev; char *name; DeviceState *dev; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index f244f003d0..7090962496 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -39,6 +39,7 @@ typedef struct VFIOContainerBase { bool dirty_pages_supported; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; QLIST_ENTRY(VFIOContainerBase) next; + QLIST_HEAD(, VFIODevice) device_list; } VFIOContainerBase; typedef struct VFIOGuestIOMMU { diff --git a/hw/vfio/common.c b/hw/vfio/common.c index b1a875ca93..9415395ed9 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -145,7 +145,7 @@ void vfio_unblock_multiple_devices_migration(void) bool vfio_viommu_preset(VFIODevice *vbasedev) { - return vbasedev->container->bcontainer.space->as != &address_space_memory; + return vbasedev->bcontainer->space->as != &address_space_memory; } static void vfio_set_migration_error(int err) @@ -179,6 +179,7 @@ bool vfio_device_state_is_precopy(VFIODevice *vbasedev) static bool vfio_devices_all_dirty_tracking(VFIOContainer *container) { + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; MigrationState *ms = migrate_get_current(); @@ -187,7 +188,7 @@ static bool vfio_devices_all_dirty_tracking(VFIOContainer *container) return false; } - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { VFIOMigration *migration = vbasedev->migration; if (!migration) { @@ -205,9 +206,10 @@ static bool vfio_devices_all_dirty_tracking(VFIOContainer *container) bool vfio_devices_all_device_dirty_tracking(VFIOContainer *container) { + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { if (!vbasedev->dirty_pages_supported) { return false; } @@ -222,13 +224,14 @@ bool vfio_devices_all_device_dirty_tracking(VFIOContainer *container) */ bool vfio_devices_all_running_and_mig_active(VFIOContainer *container) { + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; if (!migration_is_active(migrate_get_current())) { return false; } - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { VFIOMigration *migration = vbasedev->migration; if (!migration) { @@ -833,12 +836,13 @@ static bool vfio_section_is_vfio_pci(MemoryRegionSection *section, VFIOContainer *container) { VFIOPCIDevice *pcidev; + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; Object *owner; owner = memory_region_owner(section->mr); - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { if (vbasedev->type != VFIO_DEVICE_TYPE_PCI) { continue; } @@ -939,13 +943,14 @@ static void vfio_devices_dma_logging_stop(VFIOContainer *container) uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature), sizeof(uint64_t))] = {}; struct vfio_device_feature *feature = (struct vfio_device_feature *)buf; + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; feature->argsz = sizeof(buf); feature->flags = VFIO_DEVICE_FEATURE_SET | VFIO_DEVICE_FEATURE_DMA_LOGGING_STOP; - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { if (!vbasedev->dirty_tracking) { continue; } @@ -1036,6 +1041,7 @@ static int vfio_devices_dma_logging_start(VFIOContainer *container) { struct vfio_device_feature *feature; VFIODirtyRanges ranges; + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; int ret = 0; @@ -1046,7 +1052,7 @@ static int vfio_devices_dma_logging_start(VFIOContainer *container) return -errno; } - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { if (vbasedev->dirty_tracking) { continue; } @@ -1139,10 +1145,11 @@ int vfio_devices_query_dirty_bitmap(VFIOContainer *container, VFIOBitmap *vbmap, hwaddr iova, hwaddr size) { + VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; int ret; - QLIST_FOREACH(vbasedev, &container->device_list, container_next) { + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { ret = vfio_device_dma_logging_report(vbasedev, iova, size, vbmap->bitmap); if (ret) { diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 3ab74e2615..63a906de93 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -888,7 +888,7 @@ int vfio_attach_device(char *name, VFIODevice *vbasedev, int groupid = vfio_device_groupid(vbasedev, errp); VFIODevice *vbasedev_iter; VFIOGroup *group; - VFIOContainer *container; + VFIOContainerBase *bcontainer; int ret; if (groupid < 0) { @@ -915,9 +915,9 @@ int vfio_attach_device(char *name, VFIODevice *vbasedev, return ret; } - container = group->container; - vbasedev->container = container; - QLIST_INSERT_HEAD(&container->device_list, vbasedev, container_next); + bcontainer = &group->container->bcontainer; + vbasedev->bcontainer = bcontainer; + QLIST_INSERT_HEAD(&bcontainer->device_list, vbasedev, container_next); QLIST_INSERT_HEAD(&vfio_device_list, vbasedev, global_next); return ret; @@ -927,13 +927,13 @@ void vfio_detach_device(VFIODevice *vbasedev) { VFIOGroup *group = vbasedev->group; - if (!vbasedev->container) { + if (!vbasedev->bcontainer) { return; } QLIST_REMOVE(vbasedev, global_next); QLIST_REMOVE(vbasedev, container_next); - vbasedev->container = NULL; + vbasedev->bcontainer = NULL; trace_vfio_detach_device(vbasedev->name, group->groupid); vfio_put_base_device(vbasedev); vfio_put_group(group); From patchwork Thu Nov 2 07:12:35 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858345 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=XlvtIN/X; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbHB1Xktz1yQ4 for ; Thu, 2 Nov 2023 18:34:30 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS8w-0003Sc-W1; Thu, 02 Nov 2023 03:29:31 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8v-0003SF-C6 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:29 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8s-0003Iu-6q for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:29 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910166; x=1730446166; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=O25HrN7bi2D82W+06OemKsqjZDfD6BFl3siJe3Uotns=; b=XlvtIN/XHNbQ77eldu3SNAgGs4D6xKCneP3ClYJf1rDdjfU8YFVxdbTL bWhRbaGOjp3Fd3JUCSR/eJre4JXI8QkuFnKW3LlpfmtTohBxp8NKk37Vw +EvPqefGju8WV4WsAPa2ie1UfMGIISibvbS6Pm+zMdAoxWkVqh++vr9tC OBuuF/urw5zgQCs+tfshhvME0bVIiRCSrQ7bF9zlILhTDhKD2AmCUQcHh R/rzonAC5JrDJ0vvhHzlDJYdvZdt2AAfL5SOk3+5zkjOCV9bySnUs1q3n B/v0fCAmtviZmf4fLSCQP1d6pkNZXVnRkAgKLD65IgZSLoM3MyxKOxw35 A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953163" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953163" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:25 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055370" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055370" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:21 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 14/41] vfio/container: Convert functions to base container Date: Thu, 2 Nov 2023 15:12:35 +0800 Message-Id: <20231102071302.1818071-15-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger In the prospect to get rid of VFIOContainer refs in common.c lets convert misc functions to use the base container object instead: vfio_devices_all_dirty_tracking vfio_devices_all_device_dirty_tracking vfio_devices_all_running_and_mig_active vfio_devices_query_dirty_bitmap vfio_get_dirty_bitmap Signed-off-by: Eric Auger Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 9 ++++---- hw/vfio/common.c | 42 +++++++++++++++-------------------- hw/vfio/container.c | 6 ++--- hw/vfio/trace-events | 2 +- 4 files changed, 26 insertions(+), 33 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 9740cf9fbc..bc67e1316c 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -186,7 +186,6 @@ typedef struct VFIODisplay { VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); -bool vfio_devices_all_running_and_saving(VFIOContainer *container); /* SPAPR specific */ int vfio_container_add_section_window(VFIOContainer *container, @@ -260,11 +259,11 @@ bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); void vfio_migration_exit(VFIODevice *vbasedev); int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); -bool vfio_devices_all_running_and_mig_active(VFIOContainer *container); -bool vfio_devices_all_device_dirty_tracking(VFIOContainer *container); -int vfio_devices_query_dirty_bitmap(VFIOContainer *container, +bool vfio_devices_all_running_and_mig_active(VFIOContainerBase *bcontainer); +bool vfio_devices_all_device_dirty_tracking(VFIOContainerBase *bcontainer); +int vfio_devices_query_dirty_bitmap(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); -int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, +int vfio_get_dirty_bitmap(VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr); #endif /* HW_VFIO_VFIO_COMMON_H */ diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 9415395ed9..cf6618f6ed 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -177,9 +177,8 @@ bool vfio_device_state_is_precopy(VFIODevice *vbasedev) migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; } -static bool vfio_devices_all_dirty_tracking(VFIOContainer *container) +static bool vfio_devices_all_dirty_tracking(VFIOContainerBase *bcontainer) { - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; MigrationState *ms = migrate_get_current(); @@ -204,9 +203,8 @@ static bool vfio_devices_all_dirty_tracking(VFIOContainer *container) return true; } -bool vfio_devices_all_device_dirty_tracking(VFIOContainer *container) +bool vfio_devices_all_device_dirty_tracking(VFIOContainerBase *bcontainer) { - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { @@ -222,9 +220,8 @@ bool vfio_devices_all_device_dirty_tracking(VFIOContainer *container) * Check if all VFIO devices are running and migration is active, which is * essentially equivalent to the migration being in pre-copy phase. */ -bool vfio_devices_all_running_and_mig_active(VFIOContainer *container) +bool vfio_devices_all_running_and_mig_active(VFIOContainerBase *bcontainer) { - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; if (!migration_is_active(migrate_get_current())) { @@ -1082,7 +1079,7 @@ static void vfio_listener_log_global_start(MemoryListener *listener) VFIOContainer *container = container_of(listener, VFIOContainer, listener); int ret; - if (vfio_devices_all_device_dirty_tracking(container)) { + if (vfio_devices_all_device_dirty_tracking(&container->bcontainer)) { ret = vfio_devices_dma_logging_start(container); } else { ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, @@ -1101,7 +1098,7 @@ static void vfio_listener_log_global_stop(MemoryListener *listener) VFIOContainer *container = container_of(listener, VFIOContainer, listener); int ret = 0; - if (vfio_devices_all_device_dirty_tracking(container)) { + if (vfio_devices_all_device_dirty_tracking(&container->bcontainer)) { vfio_devices_dma_logging_stop(container); } else { ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, @@ -1141,11 +1138,10 @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } -int vfio_devices_query_dirty_bitmap(VFIOContainer *container, +int vfio_devices_query_dirty_bitmap(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size) { - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; int ret; @@ -1165,17 +1161,16 @@ int vfio_devices_query_dirty_bitmap(VFIOContainer *container, return 0; } -int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, +int vfio_get_dirty_bitmap(VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr) { bool all_device_dirty_tracking = - vfio_devices_all_device_dirty_tracking(container); + vfio_devices_all_device_dirty_tracking(bcontainer); uint64_t dirty_pages; VFIOBitmap vbmap; int ret; - if (!container->bcontainer.dirty_pages_supported && - !all_device_dirty_tracking) { + if (!bcontainer->dirty_pages_supported && !all_device_dirty_tracking) { cpu_physical_memory_set_dirty_range(ram_addr, size, tcg_enabled() ? DIRTY_CLIENTS_ALL : DIRTY_CLIENTS_NOCODE); @@ -1188,10 +1183,9 @@ int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, } if (all_device_dirty_tracking) { - ret = vfio_devices_query_dirty_bitmap(container, &vbmap, iova, size); + ret = vfio_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size); } else { - ret = vfio_container_query_dirty_bitmap(&container->bcontainer, &vbmap, - iova, size); + ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size); } if (ret) { @@ -1201,8 +1195,7 @@ int vfio_get_dirty_bitmap(VFIOContainer *container, uint64_t iova, dirty_pages = cpu_physical_memory_set_dirty_lebitmap(vbmap.bitmap, ram_addr, vbmap.pages); - trace_vfio_get_dirty_bitmap(container->fd, iova, size, vbmap.size, - ram_addr, dirty_pages); + trace_vfio_get_dirty_bitmap(iova, size, vbmap.size, ram_addr, dirty_pages); out: g_free(vbmap.bitmap); @@ -1236,8 +1229,8 @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) rcu_read_lock(); if (vfio_get_xlat_addr(iotlb, NULL, &translated_addr, NULL)) { - ret = vfio_get_dirty_bitmap(container, iova, iotlb->addr_mask + 1, - translated_addr); + ret = vfio_get_dirty_bitmap(&container->bcontainer, iova, + iotlb->addr_mask + 1, translated_addr); if (ret) { error_report("vfio_iommu_map_dirty_notify(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", @@ -1266,7 +1259,8 @@ static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, * Sync the whole mapped region (spanning multiple individual mappings) * in one go. */ - return vfio_get_dirty_bitmap(vrdl->container, iova, size, ram_addr); + return vfio_get_dirty_bitmap(&vrdl->container->bcontainer, iova, size, + ram_addr); } static int vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainer *container, @@ -1335,7 +1329,7 @@ static int vfio_sync_dirty_bitmap(VFIOContainer *container, ram_addr = memory_region_get_ram_addr(section->mr) + section->offset_within_region; - return vfio_get_dirty_bitmap(container, + return vfio_get_dirty_bitmap(&container->bcontainer, REAL_HOST_PAGE_ALIGN(section->offset_within_address_space), int128_get64(section->size), ram_addr); } @@ -1350,7 +1344,7 @@ static void vfio_listener_log_sync(MemoryListener *listener, return; } - if (vfio_devices_all_dirty_tracking(container)) { + if (vfio_devices_all_dirty_tracking(&container->bcontainer)) { ret = vfio_sync_dirty_bitmap(container, section); if (ret) { error_report("vfio: Failed to sync dirty bitmap, err: %d (%s)", ret, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 63a906de93..7bd81eab09 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -129,8 +129,8 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, bool need_dirty_sync = false; int ret; - if (iotlb && vfio_devices_all_running_and_mig_active(container)) { - if (!vfio_devices_all_device_dirty_tracking(container) && + if (iotlb && vfio_devices_all_running_and_mig_active(bcontainer)) { + if (!vfio_devices_all_device_dirty_tracking(bcontainer) && container->bcontainer.dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); } @@ -162,7 +162,7 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, } if (need_dirty_sync) { - ret = vfio_get_dirty_bitmap(container, iova, size, + ret = vfio_get_dirty_bitmap(bcontainer, iova, size, iotlb->translated_addr); if (ret) { return ret; diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index 9f7fedee98..08a1f9dfa4 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -117,7 +117,7 @@ vfio_region_sparse_mmap_header(const char *name, int index, int nr_areas) "Devic vfio_region_sparse_mmap_entry(int i, unsigned long start, unsigned long end) "sparse entry %d [0x%lx - 0x%lx]" vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" vfio_legacy_dma_unmap_overflow_workaround(void) "" -vfio_get_dirty_bitmap(int fd, uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "container fd=%d, iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 # platform.c From patchwork Thu Nov 2 07:12:36 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858341 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=QKcC/bsc; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbG33RP5z1yQ4 for ; Thu, 2 Nov 2023 18:33:31 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS92-0003Tz-Hi; Thu, 02 Nov 2023 03:29:36 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS91-0003TV-9D; Thu, 02 Nov 2023 03:29:35 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS8z-0003Iu-3p; Thu, 02 Nov 2023 03:29:35 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910173; x=1730446173; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=isyJZX6GqnJB+enC/OfCc7VnmeD1C9hXkNCedvWoDLE=; b=QKcC/bsc0HABy8VJ3N3uzqJ9cEchK8biZ7q0OvI9gCeFNlVVDBrzdW1O k0wVmNwJHIOGdgSrE29KH1HN7b0XDIAZ/On/Ld72ai0DLoQd/n5cv2Zsq wJAZLyuViEG7oCqGMV8Oi7CkpeGZLXdt/inw3dH6/BC7sfEA14q96qtK0 nysv/upo/0Tb5V02SmgWLzk7qremdWJyYBN5i2OCmKhis4RvbKXlccX00 jTchnITpv0YO0XNNj7VplNqTczOfxlmzNgMOyKdOWnZPXRdSAkcqrdrow zUc18jQXMyKpRHYh+jP3nMMjy4dghxN95afZUOsEgvF2RQXNj45zXHngo A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953177" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953177" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:30 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055385" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055385" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:25 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 15/41] vfio/container: Move pgsizes and dma_max_mappings to base container Date: Thu, 2 Nov 2023 15:12:36 +0800 Message-Id: <20231102071302.1818071-16-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger No functional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: Split vrdl_list change out in a seperate patch include/hw/vfio/vfio-common.h | 2 -- include/hw/vfio/vfio-container-base.h | 2 ++ hw/vfio/common.c | 17 +++++++++-------- hw/vfio/container-base.c | 1 + hw/vfio/container.c | 11 +++++------ hw/vfio/spapr.c | 10 ++++++---- 6 files changed, 23 insertions(+), 20 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index bc67e1316c..d3dc2f9dcb 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -85,8 +85,6 @@ typedef struct VFIOContainer { bool initialized; uint64_t dirty_pgsizes; uint64_t max_dirty_bitmap_size; - unsigned long pgsizes; - unsigned int dma_max_mappings; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 7090962496..85ec7e1a56 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -36,6 +36,8 @@ typedef struct VFIOAddressSpace { typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; VFIOAddressSpace *space; + unsigned long pgsizes; + unsigned int dma_max_mappings; bool dirty_pages_supported; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; QLIST_ENTRY(VFIOContainerBase) next; diff --git a/hw/vfio/common.c b/hw/vfio/common.c index cf6618f6ed..1cb53d369e 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -401,6 +401,7 @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, static void vfio_register_ram_discard_listener(VFIOContainer *container, MemoryRegionSection *section) { + VFIOContainerBase *bcontainer = &container->bcontainer; RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); VFIORamDiscardListener *vrdl; @@ -419,8 +420,8 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, section->mr); g_assert(vrdl->granularity && is_power_of_2(vrdl->granularity)); - g_assert(container->pgsizes && - vrdl->granularity >= 1ULL << ctz64(container->pgsizes)); + g_assert(bcontainer->pgsizes && + vrdl->granularity >= 1ULL << ctz64(bcontainer->pgsizes)); ram_discard_listener_init(&vrdl->listener, vfio_ram_discard_notify_populate, @@ -441,7 +442,7 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, * number of sections in the address space we could have over time, * also consuming DMA mappings. */ - if (container->dma_max_mappings) { + if (bcontainer->dma_max_mappings) { unsigned int vrdl_count = 0, vrdl_mappings = 0, max_memslots = 512; #ifdef CONFIG_KVM @@ -462,11 +463,11 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, } if (vrdl_mappings + max_memslots - vrdl_count > - container->dma_max_mappings) { + bcontainer->dma_max_mappings) { warn_report("%s: possibly running out of DMA mappings. E.g., try" " increasing the 'block-size' of virtio-mem devies." " Maximum possible DMA mappings: %d, Maximum possible" - " memslots: %d", __func__, container->dma_max_mappings, + " memslots: %d", __func__, bcontainer->dma_max_mappings, max_memslots); } } @@ -626,7 +627,7 @@ static void vfio_listener_region_add(MemoryListener *listener, iommu_idx); ret = memory_region_iommu_set_page_size_mask(giommu->iommu_mr, - container->pgsizes, + bcontainer->pgsizes, &err); if (ret) { g_free(giommu); @@ -675,7 +676,7 @@ static void vfio_listener_region_add(MemoryListener *listener, llsize = int128_sub(llend, int128_make64(iova)); if (memory_region_is_ram_device(section->mr)) { - hwaddr pgmask = (1ULL << ctz64(container->pgsizes)) - 1; + hwaddr pgmask = (1ULL << ctz64(bcontainer->pgsizes)) - 1; if ((iova & pgmask) || (int128_get64(llsize) & pgmask)) { trace_vfio_listener_region_add_no_dma_map( @@ -777,7 +778,7 @@ static void vfio_listener_region_del(MemoryListener *listener, if (memory_region_is_ram_device(section->mr)) { hwaddr pgmask; - pgmask = (1ULL << ctz64(container->pgsizes)) - 1; + pgmask = (1ULL << ctz64(bcontainer->pgsizes)) - 1; try_unmap = !((iova & pgmask) || (int128_get64(llsize) & pgmask)); } else if (memory_region_has_ram_discard_manager(section->mr)) { vfio_unregister_ram_discard_listener(container, section); diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 5d654ae172..dcce111349 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -52,6 +52,7 @@ void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, bcontainer->ops = ops; bcontainer->space = space; bcontainer->dirty_pages_supported = false; + bcontainer->dma_max_mappings = 0; QLIST_INIT(&bcontainer->giommu_list); } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 7bd81eab09..c5a6262882 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -154,7 +154,7 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, if (errno == EINVAL && unmap.size && !(unmap.iova + unmap.size) && container->iommu_type == VFIO_TYPE1v2_IOMMU) { trace_vfio_legacy_dma_unmap_overflow_workaround(); - unmap.size -= 1ULL << ctz64(container->pgsizes); + unmap.size -= 1ULL << ctz64(bcontainer->pgsizes); continue; } error_report("VFIO_UNMAP_DMA failed: %s", strerror(errno)); @@ -559,7 +559,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container = g_malloc0(sizeof(*container)); container->fd = fd; container->error = NULL; - container->dma_max_mappings = 0; container->iova_ranges = NULL; QLIST_INIT(&container->vrdl_list); bcontainer = &container->bcontainer; @@ -589,13 +588,13 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, } if (info->flags & VFIO_IOMMU_INFO_PGSIZES) { - container->pgsizes = info->iova_pgsizes; + bcontainer->pgsizes = info->iova_pgsizes; } else { - container->pgsizes = qemu_real_host_page_size(); + bcontainer->pgsizes = qemu_real_host_page_size(); } - if (!vfio_get_info_dma_avail(info, &container->dma_max_mappings)) { - container->dma_max_mappings = 65535; + if (!vfio_get_info_dma_avail(info, &bcontainer->dma_max_mappings)) { + bcontainer->dma_max_mappings = 65535; } vfio_get_info_iova_range(info, container); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 83da2f7ec2..4f76bdd3ca 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -226,6 +226,7 @@ static int vfio_spapr_create_window(VFIOContainer *container, hwaddr *pgsize) { int ret = 0; + VFIOContainerBase *bcontainer = &container->bcontainer; IOMMUMemoryRegion *iommu_mr = IOMMU_MEMORY_REGION(section->mr); uint64_t pagesize = memory_region_iommu_get_min_page_size(iommu_mr), pgmask; unsigned entries, bits_total, bits_per_level, max_levels; @@ -239,13 +240,13 @@ static int vfio_spapr_create_window(VFIOContainer *container, if (pagesize > rampagesize) { pagesize = rampagesize; } - pgmask = container->pgsizes & (pagesize | (pagesize - 1)); + pgmask = bcontainer->pgsizes & (pagesize | (pagesize - 1)); pagesize = pgmask ? (1ULL << (63 - clz64(pgmask))) : 0; if (!pagesize) { error_report("Host doesn't support page size 0x%"PRIx64 ", the supported mask is 0x%lx", memory_region_iommu_get_min_page_size(iommu_mr), - container->pgsizes); + bcontainer->pgsizes); return -EINVAL; } @@ -421,6 +422,7 @@ void vfio_container_del_section_window(VFIOContainer *container, int vfio_spapr_container_init(VFIOContainer *container, Error **errp) { + VFIOContainerBase *bcontainer = &container->bcontainer; struct vfio_iommu_spapr_tce_info info; bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; int ret, fd = container->fd; @@ -461,7 +463,7 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) } if (v2) { - container->pgsizes = info.ddw.pgsizes; + bcontainer->pgsizes = info.ddw.pgsizes; /* * There is a default window in just created container. * To make region_add/del simpler, we better remove this @@ -476,7 +478,7 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) } } else { /* The default table uses 4K pages */ - container->pgsizes = 0x1000; + bcontainer->pgsizes = 0x1000; vfio_host_win_add(container, info.dma32_window_start, info.dma32_window_start + info.dma32_window_size - 1, From patchwork Thu Nov 2 07:12:37 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858321 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=NsW6qx8W; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBg5c2mz1yQ4 for ; Thu, 2 Nov 2023 18:30:35 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS98-0003nn-RK; Thu, 02 Nov 2023 03:29:42 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS93-0003UW-RR for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:38 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS91-0003OA-Uh for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:37 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910175; x=1730446175; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=9p/nY/JHKHl9Dombgfsh3VRNf+QMpQ4oRhV9AFjEKEA=; b=NsW6qx8WKeLxkDySi/nPHB4VgfMpainGGnjRLwOiw/BwTg5kd1ctNDJs e+C6u3sM8a8IAd6nEfoPfFFUODmyPCEVWxQJDRGZBrB9nfr4VPAQVXiCw jighwSYGgJB8PG3rdi/7iQGKGxe4W4mxRyY7dEsqNdQBprcKimDH6oI7G A1IlGs6Q5o/hlRAqcJzr3WCtNscWg6fICS6LIRPC1yYk48mUJV8mukHTx 9+1b3/Qs/oD+HIhGfbSC/vJ0a1tF6FQMr6P8g7ABiDdHhMJWcCGszqM71 1sNFYR+CY8e5z+DuL6to8mhJEiCz5Pi0ZWP3FnbrKdOutnlxWj1oylwqZ w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953185" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953185" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:34 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055402" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055402" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:30 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 16/41] vfio/container: Move vrdl_list to base container Date: Thu, 2 Nov 2023 15:12:37 +0800 Message-Id: <20231102071302.1818071-17-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org No functional change intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 11 -------- include/hw/vfio/vfio-container-base.h | 11 ++++++++ hw/vfio/common.c | 38 +++++++++++++-------------- hw/vfio/container-base.c | 1 + hw/vfio/container.c | 1 - 5 files changed, 31 insertions(+), 31 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index d3dc2f9dcb..8a607a4c17 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -87,20 +87,9 @@ typedef struct VFIOContainer { uint64_t max_dirty_bitmap_size; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; - QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; GList *iova_ranges; } VFIOContainer; -typedef struct VFIORamDiscardListener { - VFIOContainer *container; - MemoryRegion *mr; - hwaddr offset_within_address_space; - hwaddr size; - uint64_t granularity; - RamDiscardListener listener; - QLIST_ENTRY(VFIORamDiscardListener) next; -} VFIORamDiscardListener; - typedef struct VFIOHostDMAWindow { hwaddr min_iova; hwaddr max_iova; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 85ec7e1a56..8e05b5ac5a 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -40,6 +40,7 @@ typedef struct VFIOContainerBase { unsigned int dma_max_mappings; bool dirty_pages_supported; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list; + QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; QLIST_ENTRY(VFIOContainerBase) next; QLIST_HEAD(, VFIODevice) device_list; } VFIOContainerBase; @@ -52,6 +53,16 @@ typedef struct VFIOGuestIOMMU { QLIST_ENTRY(VFIOGuestIOMMU) giommu_next; } VFIOGuestIOMMU; +typedef struct VFIORamDiscardListener { + VFIOContainerBase *bcontainer; + MemoryRegion *mr; + hwaddr offset_within_address_space; + hwaddr size; + uint64_t granularity; + RamDiscardListener listener; + QLIST_ENTRY(VFIORamDiscardListener) next; +} VFIORamDiscardListener; + int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 1cb53d369e..f15665789f 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -351,13 +351,13 @@ static void vfio_ram_discard_notify_discard(RamDiscardListener *rdl, { VFIORamDiscardListener *vrdl = container_of(rdl, VFIORamDiscardListener, listener); + VFIOContainerBase *bcontainer = vrdl->bcontainer; const hwaddr size = int128_get64(section->size); const hwaddr iova = section->offset_within_address_space; int ret; /* Unmap with a single call. */ - ret = vfio_container_dma_unmap(&vrdl->container->bcontainer, - iova, size , NULL); + ret = vfio_container_dma_unmap(bcontainer, iova, size , NULL); if (ret) { error_report("%s: vfio_container_dma_unmap() failed: %s", __func__, strerror(-ret)); @@ -369,6 +369,7 @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, { VFIORamDiscardListener *vrdl = container_of(rdl, VFIORamDiscardListener, listener); + VFIOContainerBase *bcontainer = vrdl->bcontainer; const hwaddr end = section->offset_within_region + int128_get64(section->size); hwaddr start, next, iova; @@ -387,8 +388,8 @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, section->offset_within_address_space; vaddr = memory_region_get_ram_ptr(section->mr) + start; - ret = vfio_container_dma_map(&vrdl->container->bcontainer, iova, - next - start, vaddr, section->readonly); + ret = vfio_container_dma_map(bcontainer, iova, next - start, + vaddr, section->readonly); if (ret) { /* Rollback */ vfio_ram_discard_notify_discard(rdl, section); @@ -398,10 +399,9 @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, return 0; } -static void vfio_register_ram_discard_listener(VFIOContainer *container, +static void vfio_register_ram_discard_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { - VFIOContainerBase *bcontainer = &container->bcontainer; RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); VFIORamDiscardListener *vrdl; @@ -412,7 +412,7 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, g_assert(QEMU_IS_ALIGNED(int128_get64(section->size), TARGET_PAGE_SIZE)); vrdl = g_new0(VFIORamDiscardListener, 1); - vrdl->container = container; + vrdl->bcontainer = bcontainer; vrdl->mr = section->mr; vrdl->offset_within_address_space = section->offset_within_address_space; vrdl->size = int128_get64(section->size); @@ -427,7 +427,7 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, vfio_ram_discard_notify_populate, vfio_ram_discard_notify_discard, true); ram_discard_manager_register_listener(rdm, &vrdl->listener, section); - QLIST_INSERT_HEAD(&container->vrdl_list, vrdl, next); + QLIST_INSERT_HEAD(&bcontainer->vrdl_list, vrdl, next); /* * Sanity-check if we have a theoretically problematic setup where we could @@ -451,7 +451,7 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, } #endif - QLIST_FOREACH(vrdl, &container->vrdl_list, next) { + QLIST_FOREACH(vrdl, &bcontainer->vrdl_list, next) { hwaddr start, end; start = QEMU_ALIGN_DOWN(vrdl->offset_within_address_space, @@ -473,13 +473,13 @@ static void vfio_register_ram_discard_listener(VFIOContainer *container, } } -static void vfio_unregister_ram_discard_listener(VFIOContainer *container, +static void vfio_unregister_ram_discard_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); VFIORamDiscardListener *vrdl = NULL; - QLIST_FOREACH(vrdl, &container->vrdl_list, next) { + QLIST_FOREACH(vrdl, &bcontainer->vrdl_list, next) { if (vrdl->mr == section->mr && vrdl->offset_within_address_space == section->offset_within_address_space) { @@ -663,7 +663,7 @@ static void vfio_listener_region_add(MemoryListener *listener, * about changes. */ if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_register_ram_discard_listener(container, section); + vfio_register_ram_discard_listener(bcontainer, section); return; } @@ -781,7 +781,7 @@ static void vfio_listener_region_del(MemoryListener *listener, pgmask = (1ULL << ctz64(bcontainer->pgsizes)) - 1; try_unmap = !((iova & pgmask) || (int128_get64(llsize) & pgmask)); } else if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_unregister_ram_discard_listener(container, section); + vfio_unregister_ram_discard_listener(bcontainer, section); /* Unregistering will trigger an unmap. */ try_unmap = false; } @@ -1260,17 +1260,17 @@ static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, * Sync the whole mapped region (spanning multiple individual mappings) * in one go. */ - return vfio_get_dirty_bitmap(&vrdl->container->bcontainer, iova, size, - ram_addr); + return vfio_get_dirty_bitmap(vrdl->bcontainer, iova, size, ram_addr); } -static int vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainer *container, - MemoryRegionSection *section) +static int +vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainerBase *bcontainer, + MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); VFIORamDiscardListener *vrdl = NULL; - QLIST_FOREACH(vrdl, &container->vrdl_list, next) { + QLIST_FOREACH(vrdl, &bcontainer->vrdl_list, next) { if (vrdl->mr == section->mr && vrdl->offset_within_address_space == section->offset_within_address_space) { @@ -1324,7 +1324,7 @@ static int vfio_sync_dirty_bitmap(VFIOContainer *container, } return 0; } else if (memory_region_has_ram_discard_manager(section->mr)) { - return vfio_sync_ram_discard_listener_dirty_bitmap(container, section); + return vfio_sync_ram_discard_listener_dirty_bitmap(bcontainer, section); } ram_addr = memory_region_get_ram_addr(section->mr) + diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index dcce111349..584eee4ba1 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -54,6 +54,7 @@ void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, bcontainer->dirty_pages_supported = false; bcontainer->dma_max_mappings = 0; QLIST_INIT(&bcontainer->giommu_list); + QLIST_INIT(&bcontainer->vrdl_list); } void vfio_container_destroy(VFIOContainerBase *bcontainer) diff --git a/hw/vfio/container.c b/hw/vfio/container.c index c5a6262882..6ba2e2f8c4 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -560,7 +560,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container->fd = fd; container->error = NULL; container->iova_ranges = NULL; - QLIST_INIT(&container->vrdl_list); bcontainer = &container->bcontainer; vfio_container_init(bcontainer, space, &vfio_legacy_ops); From patchwork Thu Nov 2 07:12:38 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858322 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=QeCvap9k; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBh6j3Yz1yQs for ; Thu, 2 Nov 2023 18:30:36 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9Y-0005SN-4R; Thu, 02 Nov 2023 03:30:08 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9U-0005K8-LL; Thu, 02 Nov 2023 03:30:04 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9R-0003OZ-L8; Thu, 02 Nov 2023 03:30:04 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910201; x=1730446201; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=i0Ht7Khj/fK6A4Q7Pt3T45CjblgzXnOkJiF1CLSwfkE=; b=QeCvap9k9c5ThmQgSOK3Tl6RfmRn3ytC9avHxJG/nlfKWYl6RvX0+cef i251gh8DJZY6vidD8yZbalIIbTAZkeOgGGrl/7TNRsZ8c8ZEQOJtRG8eE 0HDU5h8UZrLOVQ8zlOTa8oiE+a1/DYxDc+nfJtlq4vka9O9LmmiMe8ceL VKWSWthaGzMY2WFx+qwx1NqhiVdeWIAVN8f/CWS6381I8fMDmVv2lNZe8 oEtn06efCtQfHdqsfUkM4Cg8nm5t0OYiyBF4V+6lJ3oCV4OisdDsGyf/y pw0T38htlHgewfYdtM+ZOfl403NahPoQ6KelIu3bFyhpWMGJiN5YiAkMN w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953200" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953200" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:39 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055411" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055411" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:34 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 17/41] vfio/container: Move listener to base container Date: Thu, 2 Nov 2023 15:12:38 +0800 Message-Id: <20231102071302.1818071-18-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger Move listener to base container. Also error and initialized fields are moved at the same time. No functional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 3 - include/hw/vfio/vfio-container-base.h | 3 + hw/vfio/common.c | 110 +++++++++++++------------- hw/vfio/container-base.c | 1 + hw/vfio/container.c | 19 +++-- hw/vfio/spapr.c | 11 +-- 6 files changed, 74 insertions(+), 73 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 8a607a4c17..922022cbc6 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -78,11 +78,8 @@ struct VFIOGroup; typedef struct VFIOContainer { VFIOContainerBase bcontainer; int fd; /* /dev/vfio/vfio, empowered by the attached groups */ - MemoryListener listener; MemoryListener prereg_listener; unsigned iommu_type; - Error *error; - bool initialized; uint64_t dirty_pgsizes; uint64_t max_dirty_bitmap_size; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 8e05b5ac5a..95f8d319e0 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -36,6 +36,9 @@ typedef struct VFIOAddressSpace { typedef struct VFIOContainerBase { const VFIOIOMMUOps *ops; VFIOAddressSpace *space; + MemoryListener listener; + Error *error; + bool initialized; unsigned long pgsizes; unsigned int dma_max_mappings; bool dirty_pages_supported; diff --git a/hw/vfio/common.c b/hw/vfio/common.c index f15665789f..be623e544b 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -541,7 +541,7 @@ static bool vfio_listener_valid_section(MemoryRegionSection *section, return true; } -static bool vfio_get_section_iova_range(VFIOContainer *container, +static bool vfio_get_section_iova_range(VFIOContainerBase *bcontainer, MemoryRegionSection *section, hwaddr *out_iova, hwaddr *out_end, Int128 *out_llend) @@ -569,8 +569,10 @@ static bool vfio_get_section_iova_range(VFIOContainer *container, static void vfio_listener_region_add(MemoryListener *listener, MemoryRegionSection *section) { - VFIOContainer *container = container_of(listener, VFIOContainer, listener); - VFIOContainerBase *bcontainer = &container->bcontainer; + VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, + listener); + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); hwaddr iova, end; Int128 llend, llsize; void *vaddr; @@ -581,7 +583,8 @@ static void vfio_listener_region_add(MemoryListener *listener, return; } - if (!vfio_get_section_iova_range(container, section, &iova, &end, &llend)) { + if (!vfio_get_section_iova_range(bcontainer, section, &iova, &end, + &llend)) { if (memory_region_is_ram_device(section->mr)) { trace_vfio_listener_region_add_no_dma_map( memory_region_name(section->mr), @@ -688,13 +691,12 @@ static void vfio_listener_region_add(MemoryListener *listener, } } - ret = vfio_container_dma_map(&container->bcontainer, - iova, int128_get64(llsize), vaddr, - section->readonly); + ret = vfio_container_dma_map(bcontainer, iova, int128_get64(llsize), + vaddr, section->readonly); if (ret) { error_setg(&err, "vfio_container_dma_map(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx", %p) = %d (%s)", - container, iova, int128_get64(llsize), vaddr, ret, + bcontainer, iova, int128_get64(llsize), vaddr, ret, strerror(-ret)); if (memory_region_is_ram_device(section->mr)) { /* Allow unexpected mappings not to be fatal for RAM devices */ @@ -716,9 +718,9 @@ fail: * can gracefully fail. Runtime, there's not much we can do other * than throw a hardware error. */ - if (!container->initialized) { - if (!container->error) { - error_propagate_prepend(&container->error, err, + if (!bcontainer->initialized) { + if (!bcontainer->error) { + error_propagate_prepend(&bcontainer->error, err, "Region %s: ", memory_region_name(section->mr)); } else { @@ -733,8 +735,10 @@ fail: static void vfio_listener_region_del(MemoryListener *listener, MemoryRegionSection *section) { - VFIOContainer *container = container_of(listener, VFIOContainer, listener); - VFIOContainerBase *bcontainer = &container->bcontainer; + VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, + listener); + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); hwaddr iova, end; Int128 llend, llsize; int ret; @@ -767,7 +771,8 @@ static void vfio_listener_region_del(MemoryListener *listener, */ } - if (!vfio_get_section_iova_range(container, section, &iova, &end, &llend)) { + if (!vfio_get_section_iova_range(bcontainer, section, &iova, &end, + &llend)) { return; } @@ -790,22 +795,22 @@ static void vfio_listener_region_del(MemoryListener *listener, if (int128_eq(llsize, int128_2_64())) { /* The unmap ioctl doesn't accept a full 64-bit span. */ llsize = int128_rshift(llsize, 1); - ret = vfio_container_dma_unmap(&container->bcontainer, iova, + ret = vfio_container_dma_unmap(bcontainer, iova, int128_get64(llsize), NULL); if (ret) { error_report("vfio_container_dma_unmap(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", - container, iova, int128_get64(llsize), ret, + bcontainer, iova, int128_get64(llsize), ret, strerror(-ret)); } iova += int128_get64(llsize); } - ret = vfio_container_dma_unmap(&container->bcontainer, iova, + ret = vfio_container_dma_unmap(bcontainer, iova, int128_get64(llsize), NULL); if (ret) { error_report("vfio_container_dma_unmap(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", - container, iova, int128_get64(llsize), ret, + bcontainer, iova, int128_get64(llsize), ret, strerror(-ret)); } } @@ -825,16 +830,15 @@ typedef struct VFIODirtyRanges { } VFIODirtyRanges; typedef struct VFIODirtyRangesListener { - VFIOContainer *container; + VFIOContainerBase *bcontainer; VFIODirtyRanges ranges; MemoryListener listener; } VFIODirtyRangesListener; static bool vfio_section_is_vfio_pci(MemoryRegionSection *section, - VFIOContainer *container) + VFIOContainerBase *bcontainer) { VFIOPCIDevice *pcidev; - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; Object *owner; @@ -863,7 +867,7 @@ static void vfio_dirty_tracking_update(MemoryListener *listener, hwaddr iova, end, *min, *max; if (!vfio_listener_valid_section(section, "tracking_update") || - !vfio_get_section_iova_range(dirty->container, section, + !vfio_get_section_iova_range(dirty->bcontainer, section, &iova, &end, NULL)) { return; } @@ -887,7 +891,7 @@ static void vfio_dirty_tracking_update(MemoryListener *listener, * The alternative would be an IOVATree but that has a much bigger runtime * overhead and unnecessary complexity. */ - if (vfio_section_is_vfio_pci(section, dirty->container) && + if (vfio_section_is_vfio_pci(section, dirty->bcontainer) && iova >= UINT32_MAX) { min = &range->minpci64; max = &range->maxpci64; @@ -911,7 +915,7 @@ static const MemoryListener vfio_dirty_tracking_listener = { .region_add = vfio_dirty_tracking_update, }; -static void vfio_dirty_tracking_init(VFIOContainer *container, +static void vfio_dirty_tracking_init(VFIOContainerBase *bcontainer, VFIODirtyRanges *ranges) { VFIODirtyRangesListener dirty; @@ -921,10 +925,10 @@ static void vfio_dirty_tracking_init(VFIOContainer *container, dirty.ranges.min64 = UINT64_MAX; dirty.ranges.minpci64 = UINT64_MAX; dirty.listener = vfio_dirty_tracking_listener; - dirty.container = container; + dirty.bcontainer = bcontainer; memory_listener_register(&dirty.listener, - container->bcontainer.space->as); + bcontainer->space->as); *ranges = dirty.ranges; @@ -936,12 +940,11 @@ static void vfio_dirty_tracking_init(VFIOContainer *container, memory_listener_unregister(&dirty.listener); } -static void vfio_devices_dma_logging_stop(VFIOContainer *container) +static void vfio_devices_dma_logging_stop(VFIOContainerBase *bcontainer) { uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature), sizeof(uint64_t))] = {}; struct vfio_device_feature *feature = (struct vfio_device_feature *)buf; - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; feature->argsz = sizeof(buf); @@ -962,7 +965,7 @@ static void vfio_devices_dma_logging_stop(VFIOContainer *container) } static struct vfio_device_feature * -vfio_device_feature_dma_logging_start_create(VFIOContainer *container, +vfio_device_feature_dma_logging_start_create(VFIOContainerBase *bcontainer, VFIODirtyRanges *tracking) { struct vfio_device_feature *feature; @@ -1035,16 +1038,15 @@ static void vfio_device_feature_dma_logging_start_destroy( g_free(feature); } -static int vfio_devices_dma_logging_start(VFIOContainer *container) +static int vfio_devices_dma_logging_start(VFIOContainerBase *bcontainer) { struct vfio_device_feature *feature; VFIODirtyRanges ranges; - VFIOContainerBase *bcontainer = &container->bcontainer; VFIODevice *vbasedev; int ret = 0; - vfio_dirty_tracking_init(container, &ranges); - feature = vfio_device_feature_dma_logging_start_create(container, + vfio_dirty_tracking_init(bcontainer, &ranges); + feature = vfio_device_feature_dma_logging_start_create(bcontainer, &ranges); if (!feature) { return -errno; @@ -1067,7 +1069,7 @@ static int vfio_devices_dma_logging_start(VFIOContainer *container) out: if (ret) { - vfio_devices_dma_logging_stop(container); + vfio_devices_dma_logging_stop(bcontainer); } vfio_device_feature_dma_logging_start_destroy(feature); @@ -1077,14 +1079,14 @@ out: static void vfio_listener_log_global_start(MemoryListener *listener) { - VFIOContainer *container = container_of(listener, VFIOContainer, listener); + VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, + listener); int ret; - if (vfio_devices_all_device_dirty_tracking(&container->bcontainer)) { - ret = vfio_devices_dma_logging_start(container); + if (vfio_devices_all_device_dirty_tracking(bcontainer)) { + ret = vfio_devices_dma_logging_start(bcontainer); } else { - ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, - true); + ret = vfio_container_set_dirty_page_tracking(bcontainer, true); } if (ret) { @@ -1096,14 +1098,14 @@ static void vfio_listener_log_global_start(MemoryListener *listener) static void vfio_listener_log_global_stop(MemoryListener *listener) { - VFIOContainer *container = container_of(listener, VFIOContainer, listener); + VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, + listener); int ret = 0; - if (vfio_devices_all_device_dirty_tracking(&container->bcontainer)) { - vfio_devices_dma_logging_stop(container); + if (vfio_devices_all_device_dirty_tracking(bcontainer)) { + vfio_devices_dma_logging_stop(bcontainer); } else { - ret = vfio_container_set_dirty_page_tracking(&container->bcontainer, - false); + ret = vfio_container_set_dirty_page_tracking(bcontainer, false); } if (ret) { @@ -1214,8 +1216,6 @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) vfio_giommu_dirty_notifier, n); VFIOGuestIOMMU *giommu = gdn->giommu; VFIOContainerBase *bcontainer = giommu->bcontainer; - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); hwaddr iova = iotlb->iova + giommu->iommu_offset; ram_addr_t translated_addr; int ret = -EINVAL; @@ -1230,12 +1230,12 @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) rcu_read_lock(); if (vfio_get_xlat_addr(iotlb, NULL, &translated_addr, NULL)) { - ret = vfio_get_dirty_bitmap(&container->bcontainer, iova, - iotlb->addr_mask + 1, translated_addr); + ret = vfio_get_dirty_bitmap(bcontainer, iova, iotlb->addr_mask + 1, + translated_addr); if (ret) { error_report("vfio_iommu_map_dirty_notify(%p, 0x%"HWADDR_PRIx", " "0x%"HWADDR_PRIx") = %d (%s)", - container, iova, iotlb->addr_mask + 1, ret, + bcontainer, iova, iotlb->addr_mask + 1, ret, strerror(-ret)); } } @@ -1291,10 +1291,9 @@ vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainerBase *bcontainer, &vrdl); } -static int vfio_sync_dirty_bitmap(VFIOContainer *container, +static int vfio_sync_dirty_bitmap(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { - VFIOContainerBase *bcontainer = &container->bcontainer; ram_addr_t ram_addr; if (memory_region_is_iommu(section->mr)) { @@ -1330,7 +1329,7 @@ static int vfio_sync_dirty_bitmap(VFIOContainer *container, ram_addr = memory_region_get_ram_addr(section->mr) + section->offset_within_region; - return vfio_get_dirty_bitmap(&container->bcontainer, + return vfio_get_dirty_bitmap(bcontainer, REAL_HOST_PAGE_ALIGN(section->offset_within_address_space), int128_get64(section->size), ram_addr); } @@ -1338,15 +1337,16 @@ static int vfio_sync_dirty_bitmap(VFIOContainer *container, static void vfio_listener_log_sync(MemoryListener *listener, MemoryRegionSection *section) { - VFIOContainer *container = container_of(listener, VFIOContainer, listener); + VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, + listener); int ret; if (vfio_listener_skipped_section(section)) { return; } - if (vfio_devices_all_dirty_tracking(&container->bcontainer)) { - ret = vfio_sync_dirty_bitmap(container, section); + if (vfio_devices_all_dirty_tracking(bcontainer)) { + ret = vfio_sync_dirty_bitmap(bcontainer, section); if (ret) { error_report("vfio: Failed to sync dirty bitmap, err: %d (%s)", ret, strerror(-ret)); diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 584eee4ba1..7f508669f5 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -51,6 +51,7 @@ void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, { bcontainer->ops = ops; bcontainer->space = space; + bcontainer->error = NULL; bcontainer->dirty_pages_supported = false; bcontainer->dma_max_mappings = 0; QLIST_INIT(&bcontainer->giommu_list); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 6ba2e2f8c4..5c1dee8c9f 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -453,6 +453,7 @@ static void vfio_get_iommu_info_migration(VFIOContainer *container, { struct vfio_info_cap_header *hdr; struct vfio_iommu_type1_info_cap_migration *cap_mig; + VFIOContainerBase *bcontainer = &container->bcontainer; hdr = vfio_get_iommu_info_cap(info, VFIO_IOMMU_TYPE1_INFO_CAP_MIGRATION); if (!hdr) { @@ -467,7 +468,7 @@ static void vfio_get_iommu_info_migration(VFIOContainer *container, * qemu_real_host_page_size to mark those dirty. */ if (cap_mig->pgsize_bitmap & qemu_real_host_page_size()) { - container->bcontainer.dirty_pages_supported = true; + bcontainer->dirty_pages_supported = true; container->max_dirty_bitmap_size = cap_mig->max_dirty_bitmap_size; container->dirty_pgsizes = cap_mig->pgsize_bitmap; } @@ -558,7 +559,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container = g_malloc0(sizeof(*container)); container->fd = fd; - container->error = NULL; container->iova_ranges = NULL; bcontainer = &container->bcontainer; vfio_container_init(bcontainer, space, &vfio_legacy_ops); @@ -621,25 +621,24 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); - container->listener = vfio_memory_listener; - - memory_listener_register(&container->listener, bcontainer->space->as); + bcontainer->listener = vfio_memory_listener; + memory_listener_register(&bcontainer->listener, bcontainer->space->as); - if (container->error) { + if (bcontainer->error) { ret = -1; - error_propagate_prepend(errp, container->error, + error_propagate_prepend(errp, bcontainer->error, "memory listener initialization failed: "); goto listener_release_exit; } - container->initialized = true; + bcontainer->initialized = true; return 0; listener_release_exit: QLIST_REMOVE(group, container_next); QLIST_REMOVE(bcontainer, next); vfio_kvm_device_del_group(group); - memory_listener_unregister(&container->listener); + memory_listener_unregister(&bcontainer->listener); if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU || container->iommu_type == VFIO_SPAPR_TCE_IOMMU) { vfio_spapr_container_deinit(container); @@ -674,7 +673,7 @@ static void vfio_disconnect_container(VFIOGroup *group) * group. */ if (QLIST_EMPTY(&container->group_list)) { - memory_listener_unregister(&container->listener); + memory_listener_unregister(&bcontainer->listener); if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU || container->iommu_type == VFIO_SPAPR_TCE_IOMMU) { vfio_spapr_container_deinit(container); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 4f76bdd3ca..7a50975f25 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -46,6 +46,7 @@ static void vfio_prereg_listener_region_add(MemoryListener *listener, { VFIOContainer *container = container_of(listener, VFIOContainer, prereg_listener); + VFIOContainerBase *bcontainer = &container->bcontainer; const hwaddr gpa = section->offset_within_address_space; hwaddr end; int ret; @@ -88,9 +89,9 @@ static void vfio_prereg_listener_region_add(MemoryListener *listener, * can gracefully fail. Runtime, there's not much we can do other * than throw a hardware error. */ - if (!container->initialized) { - if (!container->error) { - error_setg_errno(&container->error, -ret, + if (!bcontainer->initialized) { + if (!bcontainer->error) { + error_setg_errno(&bcontainer->error, -ret, "Memory registering failed"); } } else { @@ -445,9 +446,9 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) memory_listener_register(&container->prereg_listener, &address_space_memory); - if (container->error) { + if (bcontainer->error) { ret = -1; - error_propagate_prepend(errp, container->error, + error_propagate_prepend(errp, bcontainer->error, "RAM memory listener initialization failed: "); goto listener_unregister_exit; } From patchwork Thu Nov 2 07:12:39 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858323 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=khj44zbN; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBk52tTz1yQ4 for ; Thu, 2 Nov 2023 18:30:38 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9O-0004qG-PE; Thu, 02 Nov 2023 03:29:58 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9M-0004hB-Nw for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:56 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9L-0003Ox-2u for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:29:56 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910195; x=1730446195; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=31JYkYEjp0/hjcg8A2CZRlpGcKKCzVxH23KgUl4EDzE=; b=khj44zbNj5V2aPYa2UTVcGknYZfNZw9X/B0efmoy8nCZrgGxYk77WH3u +UgaSXKPoVtFzIkFcgktkrhBtPBrTWtQa/vrR4I44C/ewJ1HsKT1VrwIy MR3Ovo3wW0E6ZhTat+AE6G0BJCMiQQW8Ma1qpbsga7KV8vsOJeO4kgKSO MMXFBx4VgS0SHhSgxjP1ekzMDuVAL6ZPkB0t01Ezzno3VNhMf50MKEIDc uyAdRCXV0y+Xs79L3K/VBUtmg0o2KyVhZsn0fo2uKCjxMz67RES1mTaMi fRJIqPLc7hgUqQqA8Of7/BKOzFjsxNuHNReXhtFwpIvfMxj9Mel95ahfU g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953208" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953208" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:43 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055421" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055421" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:39 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan Subject: [PATCH v4 18/41] vfio/container: Move dirty_pgsizes and max_dirty_bitmap_size to base container Date: Thu, 2 Nov 2023 15:12:39 +0800 Message-Id: <20231102071302.1818071-19-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger No functional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 2 -- include/hw/vfio/vfio-container-base.h | 2 ++ hw/vfio/container.c | 9 +++++---- 3 files changed, 7 insertions(+), 6 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 922022cbc6..b1c9fe711b 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -80,8 +80,6 @@ typedef struct VFIOContainer { int fd; /* /dev/vfio/vfio, empowered by the attached groups */ MemoryListener prereg_listener; unsigned iommu_type; - uint64_t dirty_pgsizes; - uint64_t max_dirty_bitmap_size; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; GList *iova_ranges; diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 95f8d319e0..80e4a993c5 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -39,6 +39,8 @@ typedef struct VFIOContainerBase { MemoryListener listener; Error *error; bool initialized; + uint64_t dirty_pgsizes; + uint64_t max_dirty_bitmap_size; unsigned long pgsizes; unsigned int dma_max_mappings; bool dirty_pages_supported; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 5c1dee8c9f..c8088a8174 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -64,6 +64,7 @@ static int vfio_dma_unmap_bitmap(VFIOContainer *container, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb) { + VFIOContainerBase *bcontainer = &container->bcontainer; struct vfio_iommu_type1_dma_unmap *unmap; struct vfio_bitmap *bitmap; VFIOBitmap vbmap; @@ -91,7 +92,7 @@ static int vfio_dma_unmap_bitmap(VFIOContainer *container, bitmap->size = vbmap.size; bitmap->data = (__u64 *)vbmap.bitmap; - if (vbmap.size > container->max_dirty_bitmap_size) { + if (vbmap.size > bcontainer->max_dirty_bitmap_size) { error_report("UNMAP: Size of bitmap too big 0x%"PRIx64, vbmap.size); ret = -E2BIG; goto unmap_exit; @@ -131,7 +132,7 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, if (iotlb && vfio_devices_all_running_and_mig_active(bcontainer)) { if (!vfio_devices_all_device_dirty_tracking(bcontainer) && - container->bcontainer.dirty_pages_supported) { + bcontainer->dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); } @@ -469,8 +470,8 @@ static void vfio_get_iommu_info_migration(VFIOContainer *container, */ if (cap_mig->pgsize_bitmap & qemu_real_host_page_size()) { bcontainer->dirty_pages_supported = true; - container->max_dirty_bitmap_size = cap_mig->max_dirty_bitmap_size; - container->dirty_pgsizes = cap_mig->pgsize_bitmap; + bcontainer->max_dirty_bitmap_size = cap_mig->max_dirty_bitmap_size; + bcontainer->dirty_pgsizes = cap_mig->pgsize_bitmap; } } From patchwork Thu Nov 2 07:12:40 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858325 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=LKGX8sHe; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBw73v2z1yQ4 for ; Thu, 2 Nov 2023 18:30:48 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9S-0005DN-Vx; Thu, 02 Nov 2023 03:30:03 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9R-00053L-5o for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:01 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9P-0003PH-8Q for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:00 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910199; x=1730446199; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=mTN336HkMtcIJyI1oPV+52daJ6m8f9QoN514J7P+dWs=; b=LKGX8sHeTsTrSh6gcB/pzKvQ3SnJPC2ALBrkW1tskCkEs7SMMBVkD2T2 1w4Rk634kPpWD6S6d4lGStE+dAg8wKLEBCUVST7wJ89KUivl7mMSulT8g oRwT2AXbzThhlaIGX+bztcNbsMy/Z0OFg6ZTW3e3sTLZjmhdTwP/XTtzP TNe2wz+CqIzSOL2HOuPB00mx/3cL9jUCKom5/NWyjXL8/+5K1Vz7IaOYT 09BFAZYwlnxH/inixUjAf4s0U7+zsx6IkLDJlPjQaLeIlpIA+AqHnEt9a Gucj9FARY11FyvQsr2Ekc9mM3MilxvVq8Fy+qe3Yc7xUZD/nlcfbvLWzz g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953218" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953218" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:47 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055436" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055436" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:43 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 19/41] vfio/container: Move iova_ranges to base container Date: Thu, 2 Nov 2023 15:12:40 +0800 Message-Id: <20231102071302.1818071-20-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Meanwhile remove the helper function vfio_free_container as it only calls g_free now. No functional change intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 1 - include/hw/vfio/vfio-container-base.h | 1 + hw/vfio/common.c | 5 +++-- hw/vfio/container-base.c | 3 +++ hw/vfio/container.c | 19 ++++++------------- 5 files changed, 13 insertions(+), 16 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index b1c9fe711b..b9e5a0e64b 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -82,7 +82,6 @@ typedef struct VFIOContainer { unsigned iommu_type; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; - GList *iova_ranges; } VFIOContainer; typedef struct VFIOHostDMAWindow { diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 80e4a993c5..9658ffb526 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -48,6 +48,7 @@ typedef struct VFIOContainerBase { QLIST_HEAD(, VFIORamDiscardListener) vrdl_list; QLIST_ENTRY(VFIOContainerBase) next; QLIST_HEAD(, VFIODevice) device_list; + GList *iova_ranges; } VFIOContainerBase; typedef struct VFIOGuestIOMMU { diff --git a/hw/vfio/common.c b/hw/vfio/common.c index be623e544b..8ef2e7967d 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -637,9 +637,10 @@ static void vfio_listener_region_add(MemoryListener *listener, goto fail; } - if (container->iova_ranges) { + if (bcontainer->iova_ranges) { ret = memory_region_iommu_set_iova_ranges(giommu->iommu_mr, - container->iova_ranges, &err); + bcontainer->iova_ranges, + &err); if (ret) { g_free(giommu); goto fail; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 7f508669f5..0177f43741 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -54,6 +54,7 @@ void vfio_container_init(VFIOContainerBase *bcontainer, VFIOAddressSpace *space, bcontainer->error = NULL; bcontainer->dirty_pages_supported = false; bcontainer->dma_max_mappings = 0; + bcontainer->iova_ranges = NULL; QLIST_INIT(&bcontainer->giommu_list); QLIST_INIT(&bcontainer->vrdl_list); } @@ -70,4 +71,6 @@ void vfio_container_destroy(VFIOContainerBase *bcontainer) QLIST_REMOVE(giommu, giommu_next); g_free(giommu); } + + g_list_free_full(bcontainer->iova_ranges, g_free); } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index c8088a8174..721c0d7375 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -308,7 +308,7 @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, } static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, - VFIOContainer *container) + VFIOContainerBase *bcontainer) { struct vfio_info_cap_header *hdr; struct vfio_iommu_type1_info_cap_iova_range *cap; @@ -326,8 +326,8 @@ static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, range_set_bounds(range, cap->iova_ranges[i].start, cap->iova_ranges[i].end); - container->iova_ranges = - range_list_insert(container->iova_ranges, range); + bcontainer->iova_ranges = + range_list_insert(bcontainer->iova_ranges, range); } return true; @@ -475,12 +475,6 @@ static void vfio_get_iommu_info_migration(VFIOContainer *container, } } -static void vfio_free_container(VFIOContainer *container) -{ - g_list_free_full(container->iova_ranges, g_free); - g_free(container); -} - static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, Error **errp) { @@ -560,7 +554,6 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container = g_malloc0(sizeof(*container)); container->fd = fd; - container->iova_ranges = NULL; bcontainer = &container->bcontainer; vfio_container_init(bcontainer, space, &vfio_legacy_ops); @@ -597,7 +590,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, bcontainer->dma_max_mappings = 65535; } - vfio_get_info_iova_range(info, container); + vfio_get_info_iova_range(info, bcontainer); vfio_get_iommu_info_migration(container, info); g_free(info); @@ -649,7 +642,7 @@ enable_discards_exit: vfio_ram_block_discard_disable(container, false); free_container_exit: - vfio_free_container(container); + g_free(container); close_fd_exit: close(fd); @@ -693,7 +686,7 @@ static void vfio_disconnect_container(VFIOGroup *group) trace_vfio_disconnect_container(container->fd); close(container->fd); - vfio_free_container(container); + g_free(container); vfio_put_address_space(space); } From patchwork Thu Nov 2 07:12:41 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858361 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=Snil1fu8; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbLb5Z0Fz1yQs for ; Thu, 2 Nov 2023 18:37:27 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9b-0005hO-OW; Thu, 02 Nov 2023 03:30:11 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9Z-0005Yf-Ja for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:09 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9X-0003Ox-72 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:09 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910207; x=1730446207; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=xP0P3YLnFVXz6lbotxMQPU3y5ur7D1Y6XVxWm0xG5cA=; b=Snil1fu8Xq5BKq4/VKtKz3f6VFH1P+4TgOpQ6dEfj9uIKddqkGm5UKyb vPqe7TE9f94Rh7HS9HFyYn6XcD9wBv2ZH4kjPa4xGDUjDB0oURkNBhyhc RLMJUVT/1mT/ysvKoePXGruIZY+Ag6mF0ogIXjEYOU/oWewFuoE2/bn8I 9DPq4i6NLxQYo7ANwsSnvFFgGjsa15E5Hz5NiXI1tUp2Jeazd33YkNjsc u4+/eeiv7HWkJ1ju7RLRK3HFP2isY9kiLzXtNlsjBDLerAgpc39KqEBc1 j8qbs9LMpaNxZKbh0bpbuOHSeXLRmb+CRpTL/tYGLH4vgCL2wk/X/tER1 w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953226" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953226" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:51 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055448" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055448" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:47 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Yi Sun , Zhenzhong Duan Subject: [PATCH v4 20/41] vfio/container: Implement attach/detach_device Date: Thu, 2 Nov 2023 15:12:41 +0800 Message-Id: <20231102071302.1818071-21-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger No fucntional change intended. Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Yi Sun Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- hw/vfio/common.c | 16 ++++++++++++++++ hw/vfio/container.c | 12 +++++------- 2 files changed, 21 insertions(+), 7 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 8ef2e7967d..483ba82089 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -1498,3 +1498,19 @@ retry: return info; } + +int vfio_attach_device(char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp) +{ + const VFIOIOMMUOps *ops = &vfio_legacy_ops; + + return ops->attach_device(name, vbasedev, as, errp); +} + +void vfio_detach_device(VFIODevice *vbasedev) +{ + if (!vbasedev->bcontainer) { + return; + } + vbasedev->bcontainer->ops->detach_device(vbasedev); +} diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 721c0d7375..6bacf38222 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -873,8 +873,8 @@ static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) * @name and @vbasedev->name are likely to be different depending * on the type of the device, hence the need for passing @name */ -int vfio_attach_device(char *name, VFIODevice *vbasedev, - AddressSpace *as, Error **errp) +static int vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp) { int groupid = vfio_device_groupid(vbasedev, errp); VFIODevice *vbasedev_iter; @@ -914,14 +914,10 @@ int vfio_attach_device(char *name, VFIODevice *vbasedev, return ret; } -void vfio_detach_device(VFIODevice *vbasedev) +static void vfio_legacy_detach_device(VFIODevice *vbasedev) { VFIOGroup *group = vbasedev->group; - if (!vbasedev->bcontainer) { - return; - } - QLIST_REMOVE(vbasedev, global_next); QLIST_REMOVE(vbasedev, container_next); vbasedev->bcontainer = NULL; @@ -933,6 +929,8 @@ void vfio_detach_device(VFIODevice *vbasedev) const VFIOIOMMUOps vfio_legacy_ops = { .dma_map = vfio_legacy_dma_map, .dma_unmap = vfio_legacy_dma_unmap, + .attach_device = vfio_legacy_attach_device, + .detach_device = vfio_legacy_detach_device, .set_dirty_page_tracking = vfio_legacy_set_dirty_page_tracking, .query_dirty_bitmap = vfio_legacy_query_dirty_bitmap, }; From patchwork Thu Nov 2 07:12:42 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858333 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=Illy0yr2; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbCr3k79z1yQ4 for ; Thu, 2 Nov 2023 18:31:36 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9p-0006QY-GM; Thu, 02 Nov 2023 03:30:25 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9n-0006Do-Mr; Thu, 02 Nov 2023 03:30:23 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9m-0003PH-3b; Thu, 02 Nov 2023 03:30:23 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910222; x=1730446222; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=l3YZkNwIq5jBPCpzck0bFALAvay/0LAv+uWO0ePCl98=; b=Illy0yr2KmN/KKP5fgo8D1mJ4rlb05gG+B1t38RTB7fICZBRVi4LxsnZ sMUlinilLekScagEQFyu1pLgx3r3Cq/WAsjAj4zE2ycrQcEoBXxpqOkpP 8NkWcV14lEib+JsYa2veN+P00sHmDFGIOu77shP2NWl0hBIUW7CklZCh+ hyaS0JB4p6X0Lb3IQsIALakKez3pqRfUt+sZ5AAqUUvsQfOep7nv/jek6 GARRNlDE/Sqeo0OgRWaAQcz+6YANMIHHMO90kPL7dM/9ja04Tm2R2K7FJ AomZTL91N7jgW3+apw/K0J7ule41oww9oDQfsJI8vr92jzRxN+/EVhWEi Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953232" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953232" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:56 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055459" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055459" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:51 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , =?utf-8?q?C=C3=A9dric_Le_G?= =?utf-8?q?oater?= , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 21/41] vfio/spapr: Introduce spapr backend and target interface Date: Thu, 2 Nov 2023 15:12:42 +0800 Message-Id: <20231102071302.1818071-22-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Introduce an empty spapr backend which will hold spapr specific content, currently only prereg_listener and hostwin_list. Also introduce two spapr specific callbacks add/del_window into VFIOIOMMUOps. Instantiate a spapr ops with a helper setup_spapr_ops and assign it to bcontainer->ops. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- v4: remove VFIOIOMMUSpaprOps include/hw/vfio/vfio-container-base.h | 6 ++++++ hw/vfio/spapr.c | 14 ++++++++++++++ 2 files changed, 20 insertions(+) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 9658ffb526..f62a14ac73 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -101,5 +101,11 @@ struct VFIOIOMMUOps { int (*set_dirty_page_tracking)(VFIOContainerBase *bcontainer, bool start); int (*query_dirty_bitmap)(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); + /* SPAPR specific */ + int (*add_window)(VFIOContainerBase *bcontainer, + MemoryRegionSection *section, + Error **errp); + void (*del_window)(VFIOContainerBase *bcontainer, + MemoryRegionSection *section); }; #endif /* HW_VFIO_VFIO_CONTAINER_BASE_H */ diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 7a50975f25..e1a6b35563 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -24,6 +24,10 @@ #include "qapi/error.h" #include "trace.h" +typedef struct VFIOSpaprContainer { + VFIOContainer container; +} VFIOSpaprContainer; + static bool vfio_prereg_listener_skipped_section(MemoryRegionSection *section) { if (memory_region_is_iommu(section->mr)) { @@ -421,6 +425,14 @@ void vfio_container_del_section_window(VFIOContainer *container, } } +static VFIOIOMMUOps vfio_iommu_spapr_ops; + +static void setup_spapr_ops(VFIOContainerBase *bcontainer) +{ + vfio_iommu_spapr_ops = *bcontainer->ops; + bcontainer->ops = &vfio_iommu_spapr_ops; +} + int vfio_spapr_container_init(VFIOContainer *container, Error **errp) { VFIOContainerBase *bcontainer = &container->bcontainer; @@ -486,6 +498,8 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) 0x1000); } + setup_spapr_ops(bcontainer); + return 0; listener_unregister_exit: From patchwork Thu Nov 2 07:12:43 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858352 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=aHwOXMOq; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKj2wLCz1yQs for ; Thu, 2 Nov 2023 18:36:41 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9u-000727-PB; Thu, 02 Nov 2023 03:30:30 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9s-0006mI-IT; Thu, 02 Nov 2023 03:30:28 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9q-0003OZ-LS; Thu, 02 Nov 2023 03:30:28 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910226; x=1730446226; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=1jY3uHhxMwZwdNDtdWIik+YcNcyh+TkP1T3KvqzilT8=; b=aHwOXMOqWisCwbDhlmEU6BJ6+SVqEwXbQJBhXwHrrbwMloEJcHOyqD09 xzBUCIofyM8vBDa/bER36jDnviXWiV3fcAxM0WMBGD9ImQzNoV6cRuEAx JxITeUN8/C3PUxqkh3zv2sJa8+0QeShSwgXCt87AW6O6ESYAUJElsScCF v/DfnIs23IoLoe5/C6HCZg3Wu+gR46pfqnNUkBicPziPiNzxX4iKTF52q 9ylWHL8G0RJPYAbORfAX/rYQpP3R9PCfRJaLtJes6PN2DUa7J/e06beny M6ckAuqgeE9M8fNRevSQ6ohsgLfj+f3OR5HH7SOn71VCHnG2UuRMke4g0 g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953245" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953245" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:01 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055473" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055473" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:29:56 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 22/41] vfio/spapr: switch to spapr IOMMU BE add/del_section_window Date: Thu, 2 Nov 2023 15:12:43 +0800 Message-Id: <20231102071302.1818071-23-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org No fucntional change intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 5 ----- include/hw/vfio/vfio-container-base.h | 5 +++++ hw/vfio/common.c | 8 ++------ hw/vfio/container-base.c | 21 +++++++++++++++++++++ hw/vfio/spapr.c | 19 ++++++++++++++----- 5 files changed, 42 insertions(+), 16 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index b9e5a0e64b..055f679363 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -169,11 +169,6 @@ VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); /* SPAPR specific */ -int vfio_container_add_section_window(VFIOContainer *container, - MemoryRegionSection *section, - Error **errp); -void vfio_container_del_section_window(VFIOContainer *container, - MemoryRegionSection *section); int vfio_spapr_container_init(VFIOContainer *container, Error **errp); void vfio_spapr_container_deinit(VFIOContainer *container); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index f62a14ac73..4b6f017c6f 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -75,6 +75,11 @@ int vfio_container_dma_map(VFIOContainerBase *bcontainer, int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb); +int vfio_container_add_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section, + Error **errp); +void vfio_container_del_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section); int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start); int vfio_container_query_dirty_bitmap(VFIOContainerBase *bcontainer, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 483ba82089..572ae7c934 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -571,8 +571,6 @@ static void vfio_listener_region_add(MemoryListener *listener, { VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, listener); - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); hwaddr iova, end; Int128 llend, llsize; void *vaddr; @@ -595,7 +593,7 @@ static void vfio_listener_region_add(MemoryListener *listener, return; } - if (vfio_container_add_section_window(container, section, &err)) { + if (vfio_container_add_section_window(bcontainer, section, &err)) { goto fail; } @@ -738,8 +736,6 @@ static void vfio_listener_region_del(MemoryListener *listener, { VFIOContainerBase *bcontainer = container_of(listener, VFIOContainerBase, listener); - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); hwaddr iova, end; Int128 llend, llsize; int ret; @@ -818,7 +814,7 @@ static void vfio_listener_region_del(MemoryListener *listener, memory_region_unref(section->mr); - vfio_container_del_section_window(container, section); + vfio_container_del_section_window(bcontainer, section); } typedef struct VFIODirtyRanges { diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 0177f43741..71f7274973 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -31,6 +31,27 @@ int vfio_container_dma_unmap(VFIOContainerBase *bcontainer, return bcontainer->ops->dma_unmap(bcontainer, iova, size, iotlb); } +int vfio_container_add_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section, + Error **errp) +{ + if (!bcontainer->ops->add_window) { + return 0; + } + + return bcontainer->ops->add_window(bcontainer, section, errp); +} + +void vfio_container_del_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section) +{ + if (!bcontainer->ops->del_window) { + return; + } + + return bcontainer->ops->del_window(bcontainer, section); +} + int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start) { diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index e1a6b35563..5be1911aad 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -319,10 +319,13 @@ static int vfio_spapr_create_window(VFIOContainer *container, return 0; } -int vfio_container_add_section_window(VFIOContainer *container, - MemoryRegionSection *section, - Error **errp) +static int +vfio_spapr_container_add_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section, + Error **errp) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); VFIOHostDMAWindow *hostwin; hwaddr pgsize = 0; int ret; @@ -407,9 +410,13 @@ int vfio_container_add_section_window(VFIOContainer *container, return 0; } -void vfio_container_del_section_window(VFIOContainer *container, - MemoryRegionSection *section) +static void +vfio_spapr_container_del_section_window(VFIOContainerBase *bcontainer, + MemoryRegionSection *section) { + VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); + if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { return; } @@ -430,6 +437,8 @@ static VFIOIOMMUOps vfio_iommu_spapr_ops; static void setup_spapr_ops(VFIOContainerBase *bcontainer) { vfio_iommu_spapr_ops = *bcontainer->ops; + vfio_iommu_spapr_ops.add_window = vfio_spapr_container_add_section_window; + vfio_iommu_spapr_ops.del_window = vfio_spapr_container_del_section_window; bcontainer->ops = &vfio_iommu_spapr_ops; } From patchwork Thu Nov 2 07:12:44 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858347 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=Ov0/1EF0; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbHr0kDdz1yQ4 for ; Thu, 2 Nov 2023 18:35:04 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9y-0007SF-Ep; Thu, 02 Nov 2023 03:30:34 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9w-0007E7-Cp; Thu, 02 Nov 2023 03:30:32 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9u-0003Ox-BN; Thu, 02 Nov 2023 03:30:32 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910230; x=1730446230; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=KXQqgqbLJd95sOJC11Mdq5aN1v8GXo0opX/2P4N8Aps=; b=Ov0/1EF0DCxpCOoONxbtSnqCvl+/2cd346zmNNULIzvLw8DTdeQ25YhR sPdLMcyKLk0QHSUUDe0FtVbwVN86opWoFmyaLdCRz8LceC5rSAFwBXLSa LPEJBrwR9oC+sDkbpaugpO1Bz5PA6V1rqXXm5+e0g/BpVy/96MuTYz1TJ zmu0YYdMxA0KYmYIDN7ikEFzsZGtmzAYEMWYFV3LQr6gZHOlp7nqoeU4C yjWBmb3y64ggvaR6AV2kOHSJtVw/TY32/JtvahZyDKOv4kwOKmSBlXty1 MMVTJm4OK4U4JfJHmgtBJ91xF8bFRzq9d1VmVS2ipk/37rbHYvNax820L A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953255" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953255" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:06 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055510" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055510" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:01 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 23/41] vfio/spapr: Move prereg_listener into spapr container Date: Thu, 2 Nov 2023 15:12:44 +0800 Message-Id: <20231102071302.1818071-24-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org No functional changes intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 1 - hw/vfio/spapr.c | 24 ++++++++++++++++-------- 2 files changed, 16 insertions(+), 9 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 055f679363..ed6148c058 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -78,7 +78,6 @@ struct VFIOGroup; typedef struct VFIOContainer { VFIOContainerBase bcontainer; int fd; /* /dev/vfio/vfio, empowered by the attached groups */ - MemoryListener prereg_listener; unsigned iommu_type; QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 5be1911aad..68c3dd6c75 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -26,6 +26,7 @@ typedef struct VFIOSpaprContainer { VFIOContainer container; + MemoryListener prereg_listener; } VFIOSpaprContainer; static bool vfio_prereg_listener_skipped_section(MemoryRegionSection *section) @@ -48,8 +49,9 @@ static void *vfio_prereg_gpa_to_vaddr(MemoryRegionSection *section, hwaddr gpa) static void vfio_prereg_listener_region_add(MemoryListener *listener, MemoryRegionSection *section) { - VFIOContainer *container = container_of(listener, VFIOContainer, - prereg_listener); + VFIOSpaprContainer *scontainer = container_of(listener, VFIOSpaprContainer, + prereg_listener); + VFIOContainer *container = &scontainer->container; VFIOContainerBase *bcontainer = &container->bcontainer; const hwaddr gpa = section->offset_within_address_space; hwaddr end; @@ -107,8 +109,9 @@ static void vfio_prereg_listener_region_add(MemoryListener *listener, static void vfio_prereg_listener_region_del(MemoryListener *listener, MemoryRegionSection *section) { - VFIOContainer *container = container_of(listener, VFIOContainer, - prereg_listener); + VFIOSpaprContainer *scontainer = container_of(listener, VFIOSpaprContainer, + prereg_listener); + VFIOContainer *container = &scontainer->container; const hwaddr gpa = section->offset_within_address_space; hwaddr end; int ret; @@ -445,6 +448,8 @@ static void setup_spapr_ops(VFIOContainerBase *bcontainer) int vfio_spapr_container_init(VFIOContainer *container, Error **errp) { VFIOContainerBase *bcontainer = &container->bcontainer; + VFIOSpaprContainer *scontainer = container_of(container, VFIOSpaprContainer, + container); struct vfio_iommu_spapr_tce_info info; bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; int ret, fd = container->fd; @@ -463,9 +468,9 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) return -errno; } } else { - container->prereg_listener = vfio_prereg_listener; + scontainer->prereg_listener = vfio_prereg_listener; - memory_listener_register(&container->prereg_listener, + memory_listener_register(&scontainer->prereg_listener, &address_space_memory); if (bcontainer->error) { ret = -1; @@ -513,7 +518,7 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) listener_unregister_exit: if (v2) { - memory_listener_unregister(&container->prereg_listener); + memory_listener_unregister(&scontainer->prereg_listener); } return ret; } @@ -523,7 +528,10 @@ void vfio_spapr_container_deinit(VFIOContainer *container) VFIOHostDMAWindow *hostwin, *next; if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU) { - memory_listener_unregister(&container->prereg_listener); + VFIOSpaprContainer *scontainer = container_of(container, + VFIOSpaprContainer, + container); + memory_listener_unregister(&scontainer->prereg_listener); } QLIST_FOREACH_SAFE(hostwin, &container->hostwin_list, hostwin_next, next) { From patchwork Thu Nov 2 07:12:45 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858320 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=BRIz+9NB; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbBc4VJmz1yQ4 for ; Thu, 2 Nov 2023 18:30:32 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9s-0006mK-OZ; Thu, 02 Nov 2023 03:30:28 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9r-0006XJ-1t; Thu, 02 Nov 2023 03:30:27 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9o-0003PH-HY; Thu, 02 Nov 2023 03:30:26 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910224; x=1730446224; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=/KZAB1sGU6lU1/7WA1ztyPgXR60ROtf6O0wC7GYMFpQ=; b=BRIz+9NBsQNptwPwPd04xL4kBRabSWURC4/IYKUxNagg67XRMbD2a+0C AJ90dAMxFgE/aegFcJqHUmk1P6Jh54jmprEJ2MCeJT6Bu2vxzbeTCMkIB u7rrGJBzJB/dbl/hlzNlcmikpFTbKCZGSzIM9pGDvcBAQ4NI7VxjzAwPY c8MwYYi8rPBlKP+hoxNKWIw6uKf6Any36QlT+hbdvbtxeQ1z6bM6Nh3Kl JHYy7dt1iB5Bl1yLcKFnMRp0wPJ4Ap8+E2zjTtdxgUjWFQRrzpt7cg5k8 nXIVUcq2zCTX2M8Ga8bPLDO8FwWPOrpcEj57X7TQFP6Bz/lzuFjJRh1r7 g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953276" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953276" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:11 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055544" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055544" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:06 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Nicholas Piggin , Daniel Henrique Barboza , =?utf-8?q?C=C3=A9dric_Le_G?= =?utf-8?q?oater?= , David Gibson , Harsh Prateek Bora , qemu-ppc@nongnu.org (open list:sPAPR (pseries)) Subject: [PATCH v4 24/41] vfio/spapr: Move hostwin_list into spapr container Date: Thu, 2 Nov 2023 15:12:45 +0800 Message-Id: <20231102071302.1818071-25-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org No functional changes intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- include/hw/vfio/vfio-common.h | 1 - hw/vfio/spapr.c | 36 +++++++++++++++++++---------------- 2 files changed, 20 insertions(+), 17 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index ed6148c058..24ecc0e7ee 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -79,7 +79,6 @@ typedef struct VFIOContainer { VFIOContainerBase bcontainer; int fd; /* /dev/vfio/vfio, empowered by the attached groups */ unsigned iommu_type; - QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; QLIST_HEAD(, VFIOGroup) group_list; } VFIOContainer; diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index 68c3dd6c75..5c6426e697 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -27,6 +27,7 @@ typedef struct VFIOSpaprContainer { VFIOContainer container; MemoryListener prereg_listener; + QLIST_HEAD(, VFIOHostDMAWindow) hostwin_list; } VFIOSpaprContainer; static bool vfio_prereg_listener_skipped_section(MemoryRegionSection *section) @@ -154,12 +155,12 @@ static const MemoryListener vfio_prereg_listener = { .region_del = vfio_prereg_listener_region_del, }; -static void vfio_host_win_add(VFIOContainer *container, hwaddr min_iova, +static void vfio_host_win_add(VFIOSpaprContainer *scontainer, hwaddr min_iova, hwaddr max_iova, uint64_t iova_pgsizes) { VFIOHostDMAWindow *hostwin; - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + QLIST_FOREACH(hostwin, &scontainer->hostwin_list, hostwin_next) { if (ranges_overlap(hostwin->min_iova, hostwin->max_iova - hostwin->min_iova + 1, min_iova, @@ -173,15 +174,15 @@ static void vfio_host_win_add(VFIOContainer *container, hwaddr min_iova, hostwin->min_iova = min_iova; hostwin->max_iova = max_iova; hostwin->iova_pgsizes = iova_pgsizes; - QLIST_INSERT_HEAD(&container->hostwin_list, hostwin, hostwin_next); + QLIST_INSERT_HEAD(&scontainer->hostwin_list, hostwin, hostwin_next); } -static int vfio_host_win_del(VFIOContainer *container, +static int vfio_host_win_del(VFIOSpaprContainer *scontainer, hwaddr min_iova, hwaddr max_iova) { VFIOHostDMAWindow *hostwin; - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + QLIST_FOREACH(hostwin, &scontainer->hostwin_list, hostwin_next) { if (hostwin->min_iova == min_iova && hostwin->max_iova == max_iova) { QLIST_REMOVE(hostwin, hostwin_next); g_free(hostwin); @@ -192,7 +193,7 @@ static int vfio_host_win_del(VFIOContainer *container, return -1; } -static VFIOHostDMAWindow *vfio_find_hostwin(VFIOContainer *container, +static VFIOHostDMAWindow *vfio_find_hostwin(VFIOSpaprContainer *container, hwaddr iova, hwaddr end) { VFIOHostDMAWindow *hostwin; @@ -329,6 +330,8 @@ vfio_spapr_container_add_section_window(VFIOContainerBase *bcontainer, { VFIOContainer *container = container_of(bcontainer, VFIOContainer, bcontainer); + VFIOSpaprContainer *scontainer = container_of(container, VFIOSpaprContainer, + container); VFIOHostDMAWindow *hostwin; hwaddr pgsize = 0; int ret; @@ -344,7 +347,7 @@ vfio_spapr_container_add_section_window(VFIOContainerBase *bcontainer, iova = section->offset_within_address_space; end = iova + int128_get64(section->size) - 1; - if (!vfio_find_hostwin(container, iova, end)) { + if (!vfio_find_hostwin(scontainer, iova, end)) { error_setg(errp, "Container %p can't map guest IOVA region" " 0x%"HWADDR_PRIx"..0x%"HWADDR_PRIx, container, iova, end); @@ -358,7 +361,7 @@ vfio_spapr_container_add_section_window(VFIOContainerBase *bcontainer, } /* For now intersections are not allowed, we may relax this later */ - QLIST_FOREACH(hostwin, &container->hostwin_list, hostwin_next) { + QLIST_FOREACH(hostwin, &scontainer->hostwin_list, hostwin_next) { if (ranges_overlap(hostwin->min_iova, hostwin->max_iova - hostwin->min_iova + 1, section->offset_within_address_space, @@ -380,7 +383,7 @@ vfio_spapr_container_add_section_window(VFIOContainerBase *bcontainer, return ret; } - vfio_host_win_add(container, section->offset_within_address_space, + vfio_host_win_add(scontainer, section->offset_within_address_space, section->offset_within_address_space + int128_get64(section->size) - 1, pgsize); #ifdef CONFIG_KVM @@ -419,6 +422,8 @@ vfio_spapr_container_del_section_window(VFIOContainerBase *bcontainer, { VFIOContainer *container = container_of(bcontainer, VFIOContainer, bcontainer); + VFIOSpaprContainer *scontainer = container_of(container, VFIOSpaprContainer, + container); if (container->iommu_type != VFIO_SPAPR_TCE_v2_IOMMU) { return; @@ -426,7 +431,7 @@ vfio_spapr_container_del_section_window(VFIOContainerBase *bcontainer, vfio_spapr_remove_window(container, section->offset_within_address_space); - if (vfio_host_win_del(container, + if (vfio_host_win_del(scontainer, section->offset_within_address_space, section->offset_within_address_space + int128_get64(section->size) - 1) < 0) { @@ -454,7 +459,7 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) bool v2 = container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU; int ret, fd = container->fd; - QLIST_INIT(&container->hostwin_list); + QLIST_INIT(&scontainer->hostwin_list); /* * The host kernel code implementing VFIO_IOMMU_DISABLE is called @@ -506,7 +511,7 @@ int vfio_spapr_container_init(VFIOContainer *container, Error **errp) } else { /* The default table uses 4K pages */ bcontainer->pgsizes = 0x1000; - vfio_host_win_add(container, info.dma32_window_start, + vfio_host_win_add(scontainer, info.dma32_window_start, info.dma32_window_start + info.dma32_window_size - 1, 0x1000); @@ -525,15 +530,14 @@ listener_unregister_exit: void vfio_spapr_container_deinit(VFIOContainer *container) { + VFIOSpaprContainer *scontainer = container_of(container, VFIOSpaprContainer, + container); VFIOHostDMAWindow *hostwin, *next; if (container->iommu_type == VFIO_SPAPR_TCE_v2_IOMMU) { - VFIOSpaprContainer *scontainer = container_of(container, - VFIOSpaprContainer, - container); memory_listener_unregister(&scontainer->prereg_listener); } - QLIST_FOREACH_SAFE(hostwin, &container->hostwin_list, hostwin_next, + QLIST_FOREACH_SAFE(hostwin, &scontainer->hostwin_list, hostwin_next, next) { QLIST_REMOVE(hostwin, hostwin_next); g_free(hostwin); From patchwork Thu Nov 2 07:12:46 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858343 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=CRUfU71e; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbGl0SmQz1yQ4 for ; Thu, 2 Nov 2023 18:34:07 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qyS9w-00075F-5z; Thu, 02 Nov 2023 03:30:32 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9t-0006qY-4h for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:29 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qyS9r-0003PH-39 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:28 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910227; x=1730446227; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=VD1ocG9/fRSfYw95BHHJrAKocj8tx/oJLWrA9fiPlz4=; b=CRUfU71eQID7TlNq2O9ChE2dFsW3JLmIMYj+Xs/lzM/aQovPFSrRjwgM GMQObZ4FjBQDFEi1EHWqq1rBXeYICzp/m5I0LFcARjL1MIKj+vXjmf24N pj+DkLtkpRu3hrMBuDX+9a37W8WhnEJHjnPlb4W+f3VJM84Hohn42TaFa cVj+v19wJAn7a+nZQ07aGJWW5jgS9yohGiQZd4DaeIAmfLRjbey593LKD SBSaEfvBHa29TE3seXRW+WC/vVype9ygf7xBvaj3WMP57fvBWGccSQxC5 0qnzwkxnJwi7Fbp9t1fHnGi587xcJw5KvE3gzq/iwTxzkheyQsSQXzhbX Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953291" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953291" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:16 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055567" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055567" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:11 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Paolo Bonzini , =?utf-8?q?Marc-Andr=C3=A9_Lureau?= , =?utf-8?q?Daniel_P=2E_Berrang=C3=A9?= , Thomas Huth , =?utf-8?q?Philippe_Mathieu-Daud=C3=A9?= Subject: [PATCH v4 25/41] Add iommufd configure option Date: Thu, 2 Nov 2023 15:12:46 +0800 Message-Id: <20231102071302.1818071-26-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This adds "--enable-iommufd/--disable-iommufd" to enable or disable iommufd support, enabled by default. Signed-off-by: Zhenzhong Duan --- meson.build | 6 ++++++ meson_options.txt | 2 ++ scripts/meson-buildoptions.sh | 3 +++ 3 files changed, 11 insertions(+) diff --git a/meson.build b/meson.build index dcef8b1e79..72a57288a0 100644 --- a/meson.build +++ b/meson.build @@ -560,6 +560,10 @@ have_tpm = get_option('tpm') \ .require(targetos != 'windows', error_message: 'TPM emulation only available on POSIX systems') \ .allowed() +have_iommufd = get_option('iommufd') \ + .require(targetos == 'linux', error_message: 'iommufd is supported only on Linux') \ + .allowed() + # vhost have_vhost_user = get_option('vhost_user') \ .disable_auto_if(targetos != 'linux') \ @@ -2133,6 +2137,7 @@ if get_option('tcg').allowed() endif config_host_data.set('CONFIG_TPM', have_tpm) config_host_data.set('CONFIG_TSAN', get_option('tsan')) +config_host_data.set('CONFIG_IOMMUFD', have_iommufd) config_host_data.set('CONFIG_USB_LIBUSB', libusb.found()) config_host_data.set('CONFIG_VDE', vde.found()) config_host_data.set('CONFIG_VHOST', have_vhost) @@ -4075,6 +4080,7 @@ summary_info += {'vhost-user-crypto support': have_vhost_user_crypto} summary_info += {'vhost-user-blk server support': have_vhost_user_blk_server} summary_info += {'vhost-vdpa support': have_vhost_vdpa} summary_info += {'build guest agent': have_ga} +summary_info += {'iommufd support': have_iommufd} summary(summary_info, bool_yn: true, section: 'Configurable features') # Compilation information diff --git a/meson_options.txt b/meson_options.txt index 3c7398f3c6..91bb958cae 100644 --- a/meson_options.txt +++ b/meson_options.txt @@ -109,6 +109,8 @@ option('dbus_display', type: 'feature', value: 'auto', description: '-display dbus support') option('tpm', type : 'feature', value : 'auto', description: 'TPM support') +option('iommufd', type : 'feature', value : 'auto', + description: 'iommufd support') # Do not enable it by default even for Mingw32, because it doesn't # work on Wine. diff --git a/scripts/meson-buildoptions.sh b/scripts/meson-buildoptions.sh index 7ca4b77eae..1effc46f7d 100644 --- a/scripts/meson-buildoptions.sh +++ b/scripts/meson-buildoptions.sh @@ -125,6 +125,7 @@ meson_options_help() { printf "%s\n" ' guest-agent-msi Build MSI package for the QEMU Guest Agent' printf "%s\n" ' hvf HVF acceleration support' printf "%s\n" ' iconv Font glyph conversion support' + printf "%s\n" ' iommufd iommufd support' printf "%s\n" ' jack JACK sound support' printf "%s\n" ' keyring Linux keyring support' printf "%s\n" ' kvm KVM acceleration support' @@ -342,6 +343,8 @@ _meson_option_parse() { --enable-install-blobs) printf "%s" -Dinstall_blobs=true ;; --disable-install-blobs) printf "%s" -Dinstall_blobs=false ;; --interp-prefix=*) quote_sh "-Dinterp_prefix=$2" ;; + --enable-iommufd) printf "%s" -Diommufd=enabled ;; + --disable-iommufd) printf "%s" -Diommufd=disabled ;; --enable-jack) printf "%s" -Djack=enabled ;; --disable-jack) printf "%s" -Djack=disabled ;; --enable-keyring) printf "%s" -Dkeyring=enabled ;; From patchwork Thu Nov 2 07:12:47 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858364 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=QVES2LhT; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbLg628wz1yQs for ; Thu, 2 Nov 2023 18:37:31 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySA7-00085E-5u; Thu, 02 Nov 2023 03:30:43 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySA5-0007up-M8 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:42 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySA3-0003OZ-0B for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:41 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910239; x=1730446239; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=ZRnUM1UmhA0Av1npfuEYFt+Mmoe6CHlwX/e7PIpFUAQ=; b=QVES2LhTZ/dkJ+QfomgZ9uRv90j5e+nAa4fqYQCPLdHp8NBkDWG5qKIk fppWx4j26LhkXlPcbuL+bmZwejbpP932twHeIReEaalxZl8WSvwF8KH1A HbkFbj5x47jyC0ntCsg49EC8m3zKcPyVXzgtMolnyByOsxWXUWOlHBeWC QRvB5NJznXCMkZM7qnWttueVaOrg4ajr5DuAPOfpPydRHuE/ss+gKrBPG dfoQpjbsQgoKl5HlTkNwdq+04+Fawf63OFeaDafh9WWcmdkfkVhYecVDu 0gCsL+ja06ZqLg+u6qPKPvZXZ0pso+BIPkVkHBxCFojl+YqyGpMs1fRB2 g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953301" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953301" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:21 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055613" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055613" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:16 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Paolo Bonzini , Eric Blake , Markus Armbruster , =?utf-8?q?Daniel_P=2E_Berrang=C3=A9?= , Eduardo Habkost Subject: [PATCH v4 26/41] backends/iommufd: Introduce the iommufd object Date: Thu, 2 Nov 2023 15:12:47 +0800 Message-Id: <20231102071302.1818071-27-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger Introduce an iommufd object which allows the interaction with the host /dev/iommu device. The /dev/iommu can have been already pre-opened outside of qemu, in which case the fd can be passed directly along with the iommufd object: This allows the iommufd object to be shared accross several subsystems (VFIO, VDPA, ...). For example, libvirt would open the /dev/iommu once. If no fd is passed along with the iommufd object, the /dev/iommu is opened by the qemu code. The CONFIG_IOMMUFD option must be set to compile this new object. Suggested-by: Alex Williamson Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Zhenzhong Duan --- v4: add CONFIG_IOMMUFD check, document default case MAINTAINERS | 7 ++ qapi/qom.json | 22 ++++ include/sysemu/iommufd.h | 46 +++++++ backends/iommufd-stub.c | 59 +++++++++ backends/iommufd.c | 257 +++++++++++++++++++++++++++++++++++++++ backends/Kconfig | 4 + backends/meson.build | 5 + backends/trace-events | 12 ++ qemu-options.hx | 13 ++ 9 files changed, 425 insertions(+) create mode 100644 include/sysemu/iommufd.h create mode 100644 backends/iommufd-stub.c create mode 100644 backends/iommufd.c diff --git a/MAINTAINERS b/MAINTAINERS index cd8d6b140f..6f35159255 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2135,6 +2135,13 @@ F: hw/vfio/ap.c F: docs/system/s390x/vfio-ap.rst L: qemu-s390x@nongnu.org +iommufd +M: Yi Liu +M: Eric Auger +S: Supported +F: backends/iommufd.c +F: include/sysemu/iommufd.h + vhost M: Michael S. Tsirkin S: Supported diff --git a/qapi/qom.json b/qapi/qom.json index c53ef978ff..27300add48 100644 --- a/qapi/qom.json +++ b/qapi/qom.json @@ -794,6 +794,24 @@ { 'struct': 'VfioUserServerProperties', 'data': { 'socket': 'SocketAddress', 'device': 'str' } } +## +# @IOMMUFDProperties: +# +# Properties for iommufd objects. +# +# @fd: file descriptor name previously passed via 'getfd' command, +# which represents a pre-opened /dev/iommu. This allows the +# iommufd object to be shared accross several subsystems +# (VFIO, VDPA, ...), and the file descriptor to be shared +# with other process, e.g. DPDK. (default: QEMU opens +# /dev/iommu by itself) +# +# Since: 8.2 +## +{ 'struct': 'IOMMUFDProperties', + 'data': { '*fd': 'str' }, + 'if': 'CONFIG_IOMMUFD' } + ## # @RngProperties: # @@ -934,6 +952,8 @@ 'input-barrier', { 'name': 'input-linux', 'if': 'CONFIG_LINUX' }, + { 'name': 'iommufd', + 'if': 'CONFIG_IOMMUFD' }, 'iothread', 'main-loop', { 'name': 'memory-backend-epc', @@ -1003,6 +1023,8 @@ 'input-barrier': 'InputBarrierProperties', 'input-linux': { 'type': 'InputLinuxProperties', 'if': 'CONFIG_LINUX' }, + 'iommufd': { 'type': 'IOMMUFDProperties', + 'if': 'CONFIG_IOMMUFD' }, 'iothread': 'IothreadProperties', 'main-loop': 'MainLoopProperties', 'memory-backend-epc': { 'type': 'MemoryBackendEpcProperties', diff --git a/include/sysemu/iommufd.h b/include/sysemu/iommufd.h new file mode 100644 index 0000000000..f0e5c7eeb8 --- /dev/null +++ b/include/sysemu/iommufd.h @@ -0,0 +1,46 @@ +#ifndef SYSEMU_IOMMUFD_H +#define SYSEMU_IOMMUFD_H + +#include "qom/object.h" +#include "qemu/thread.h" +#include "exec/hwaddr.h" +#include "exec/cpu-common.h" + +#define TYPE_IOMMUFD_BACKEND "iommufd" +OBJECT_DECLARE_TYPE(IOMMUFDBackend, IOMMUFDBackendClass, + IOMMUFD_BACKEND) +#define IOMMUFD_BACKEND(obj) \ + OBJECT_CHECK(IOMMUFDBackend, (obj), TYPE_IOMMUFD_BACKEND) +#define IOMMUFD_BACKEND_GET_CLASS(obj) \ + OBJECT_GET_CLASS(IOMMUFDBackendClass, (obj), TYPE_IOMMUFD_BACKEND) +#define IOMMUFD_BACKEND_CLASS(klass) \ + OBJECT_CLASS_CHECK(IOMMUFDBackendClass, (klass), TYPE_IOMMUFD_BACKEND) +struct IOMMUFDBackendClass { + ObjectClass parent_class; +}; + +struct IOMMUFDBackend { + Object parent; + + /*< protected >*/ + int fd; /* /dev/iommu file descriptor */ + bool owned; /* is the /dev/iommu opened internally */ + QemuMutex lock; + uint32_t users; + + /*< public >*/ +}; + +int iommufd_backend_connect(IOMMUFDBackend *be, Error **errp); +void iommufd_backend_disconnect(IOMMUFDBackend *be); + +int iommufd_backend_get_ioas(IOMMUFDBackend *be, uint32_t *ioas_id); +void iommufd_backend_put_ioas(IOMMUFDBackend *be, uint32_t ioas_id); +void iommufd_backend_free_id(int fd, uint32_t id); +int iommufd_backend_map_dma(IOMMUFDBackend *be, uint32_t ioas_id, hwaddr iova, + ram_addr_t size, void *vaddr, bool readonly); +int iommufd_backend_unmap_dma(IOMMUFDBackend *be, uint32_t ioas_id, + hwaddr iova, ram_addr_t size); +int iommufd_backend_alloc_hwpt(int iommufd, uint32_t dev_id, + uint32_t pt_id, uint32_t *out_hwpt); +#endif diff --git a/backends/iommufd-stub.c b/backends/iommufd-stub.c new file mode 100644 index 0000000000..02ac844c17 --- /dev/null +++ b/backends/iommufd-stub.c @@ -0,0 +1,59 @@ +/* + * iommufd container backend stub + * + * Copyright (C) 2023 Intel Corporation. + * Copyright Red Hat, Inc. 2023 + * + * Authors: Yi Liu + * Eric Auger + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + + * You should have received a copy of the GNU General Public License along + * with this program; if not, see . + */ + +#include "qemu/osdep.h" +#include "sysemu/iommufd.h" +#include "qemu/error-report.h" + +int iommufd_backend_connect(IOMMUFDBackend *be, Error **errp) +{ + return 0; +} +void iommufd_backend_disconnect(IOMMUFDBackend *be) +{ +} +void iommufd_backend_free_id(int fd, uint32_t id) +{ +} +int iommufd_backend_get_ioas(IOMMUFDBackend *be, uint32_t *ioas_id) +{ + return 0; +} +void iommufd_backend_put_ioas(IOMMUFDBackend *be, uint32_t ioas_id) +{ +} +int iommufd_backend_map_dma(IOMMUFDBackend *be, uint32_t ioas_id, hwaddr iova, + ram_addr_t size, void *vaddr, bool readonly) +{ + return 0; +} +int iommufd_backend_unmap_dma(IOMMUFDBackend *be, uint32_t ioas_id, + hwaddr iova, ram_addr_t size) +{ + return 0; +} +int iommufd_backend_alloc_hwpt(int iommufd, uint32_t dev_id, + uint32_t pt_id, uint32_t *out_hwpt) +{ + return 0; +} diff --git a/backends/iommufd.c b/backends/iommufd.c new file mode 100644 index 0000000000..a526d58824 --- /dev/null +++ b/backends/iommufd.c @@ -0,0 +1,257 @@ +/* + * iommufd container backend + * + * Copyright (C) 2023 Intel Corporation. + * Copyright Red Hat, Inc. 2023 + * + * Authors: Yi Liu + * Eric Auger + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "sysemu/iommufd.h" +#include "qapi/error.h" +#include "qapi/qmp/qerror.h" +#include "qemu/module.h" +#include "qom/object_interfaces.h" +#include "qemu/error-report.h" +#include "monitor/monitor.h" +#include "trace.h" +#include +#include + +static void iommufd_backend_init(Object *obj) +{ + IOMMUFDBackend *be = IOMMUFD_BACKEND(obj); + + be->fd = -1; + be->users = 0; + be->owned = true; + qemu_mutex_init(&be->lock); +} + +static void iommufd_backend_finalize(Object *obj) +{ + IOMMUFDBackend *be = IOMMUFD_BACKEND(obj); + + if (be->owned) { + close(be->fd); + be->fd = -1; + } +} + +static void iommufd_backend_set_fd(Object *obj, const char *str, Error **errp) +{ + IOMMUFDBackend *be = IOMMUFD_BACKEND(obj); + int fd = -1; + + fd = monitor_fd_param(monitor_cur(), str, errp); + if (fd == -1) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + qemu_mutex_lock(&be->lock); + be->fd = fd; + be->owned = false; + qemu_mutex_unlock(&be->lock); + trace_iommu_backend_set_fd(be->fd); +} + +static void iommufd_backend_class_init(ObjectClass *oc, void *data) +{ + object_class_property_add_str(oc, "fd", NULL, iommufd_backend_set_fd); +} + +int iommufd_backend_connect(IOMMUFDBackend *be, Error **errp) +{ + int fd, ret = 0; + + qemu_mutex_lock(&be->lock); + if (be->users == UINT32_MAX) { + error_setg(errp, "too many connections"); + ret = -E2BIG; + goto out; + } + if (be->owned && !be->users) { + fd = qemu_open_old("/dev/iommu", O_RDWR); + if (fd < 0) { + error_setg_errno(errp, errno, "/dev/iommu opening failed"); + ret = fd; + goto out; + } + be->fd = fd; + } + be->users++; +out: + trace_iommufd_backend_connect(be->fd, be->owned, + be->users, ret); + qemu_mutex_unlock(&be->lock); + return ret; +} + +void iommufd_backend_disconnect(IOMMUFDBackend *be) +{ + qemu_mutex_lock(&be->lock); + if (!be->users) { + goto out; + } + be->users--; + if (!be->users && be->owned) { + close(be->fd); + be->fd = -1; + } +out: + trace_iommufd_backend_disconnect(be->fd, be->users); + qemu_mutex_unlock(&be->lock); +} + +static int iommufd_backend_alloc_ioas(int fd, uint32_t *ioas_id) +{ + int ret; + struct iommu_ioas_alloc alloc_data = { + .size = sizeof(alloc_data), + .flags = 0, + }; + + ret = ioctl(fd, IOMMU_IOAS_ALLOC, &alloc_data); + if (ret) { + error_report("Failed to allocate ioas %m"); + } + + *ioas_id = alloc_data.out_ioas_id; + trace_iommufd_backend_alloc_ioas(fd, *ioas_id, ret); + + return ret; +} + +void iommufd_backend_free_id(int fd, uint32_t id) +{ + int ret; + struct iommu_destroy des = { + .size = sizeof(des), + .id = id, + }; + + ret = ioctl(fd, IOMMU_DESTROY, &des); + trace_iommufd_backend_free_id(fd, id, ret); + if (ret) { + error_report("Failed to free id: %u %m", id); + } +} + +int iommufd_backend_get_ioas(IOMMUFDBackend *be, uint32_t *ioas_id) +{ + int ret; + + ret = iommufd_backend_alloc_ioas(be->fd, ioas_id); + trace_iommufd_backend_get_ioas(be->fd, *ioas_id, ret); + return ret; +} + +void iommufd_backend_put_ioas(IOMMUFDBackend *be, uint32_t ioas_id) +{ + iommufd_backend_free_id(be->fd, ioas_id); + trace_iommufd_backend_put_ioas(be->fd, ioas_id); +} + +int iommufd_backend_map_dma(IOMMUFDBackend *be, uint32_t ioas_id, hwaddr iova, + ram_addr_t size, void *vaddr, bool readonly) +{ + int ret; + struct iommu_ioas_map map = { + .size = sizeof(map), + .flags = IOMMU_IOAS_MAP_READABLE | + IOMMU_IOAS_MAP_FIXED_IOVA, + .ioas_id = ioas_id, + .__reserved = 0, + .user_va = (uintptr_t)vaddr, + .iova = iova, + .length = size, + }; + + if (!readonly) { + map.flags |= IOMMU_IOAS_MAP_WRITEABLE; + } + + ret = ioctl(be->fd, IOMMU_IOAS_MAP, &map); + trace_iommufd_backend_map_dma(be->fd, ioas_id, iova, size, + vaddr, readonly, ret); + if (ret) { + error_report("IOMMU_IOAS_MAP failed: %m"); + } + return !ret ? 0 : -errno; +} + +int iommufd_backend_unmap_dma(IOMMUFDBackend *be, uint32_t ioas_id, + hwaddr iova, ram_addr_t size) +{ + int ret; + struct iommu_ioas_unmap unmap = { + .size = sizeof(unmap), + .ioas_id = ioas_id, + .iova = iova, + .length = size, + }; + + ret = ioctl(be->fd, IOMMU_IOAS_UNMAP, &unmap); + trace_iommufd_backend_unmap_dma(be->fd, ioas_id, iova, size, ret); + /* + * TODO: IOMMUFD doesn't support mapping PCI BARs for now. + * It's not a problem if there is no p2p dma, relax it here + * and avoid many noisy trigger from vIOMMU side. + */ + if (ret && errno == ENOENT) { + ret = 0; + } + if (ret) { + error_report("IOMMU_IOAS_UNMAP failed: %m"); + } + return !ret ? 0 : -errno; +} + +int iommufd_backend_alloc_hwpt(int iommufd, uint32_t dev_id, + uint32_t pt_id, uint32_t *out_hwpt) +{ + int ret; + struct iommu_hwpt_alloc alloc_hwpt = { + .size = sizeof(struct iommu_hwpt_alloc), + .flags = 0, + .dev_id = dev_id, + .pt_id = pt_id, + .__reserved = 0, + }; + + ret = ioctl(iommufd, IOMMU_HWPT_ALLOC, &alloc_hwpt); + trace_iommufd_backend_alloc_hwpt(iommufd, dev_id, pt_id, + alloc_hwpt.out_hwpt_id, ret); + + if (ret) { + error_report("IOMMU_HWPT_ALLOC failed: %m"); + } else { + *out_hwpt = alloc_hwpt.out_hwpt_id; + } + return !ret ? 0 : -errno; +} + +static const TypeInfo iommufd_backend_info = { + .name = TYPE_IOMMUFD_BACKEND, + .parent = TYPE_OBJECT, + .instance_size = sizeof(IOMMUFDBackend), + .instance_init = iommufd_backend_init, + .instance_finalize = iommufd_backend_finalize, + .class_size = sizeof(IOMMUFDBackendClass), + .class_init = iommufd_backend_class_init, + .interfaces = (InterfaceInfo[]) { + { TYPE_USER_CREATABLE }, + { } + } +}; + +static void register_types(void) +{ + type_register_static(&iommufd_backend_info); +} + +type_init(register_types); diff --git a/backends/Kconfig b/backends/Kconfig index f35abc1609..2cb23f62fa 100644 --- a/backends/Kconfig +++ b/backends/Kconfig @@ -1 +1,5 @@ source tpm/Kconfig + +config IOMMUFD + bool + depends on VFIO diff --git a/backends/meson.build b/backends/meson.build index 914c7c4afb..05ac57ff15 100644 --- a/backends/meson.build +++ b/backends/meson.build @@ -20,6 +20,11 @@ if have_vhost_user system_ss.add(when: 'CONFIG_VIRTIO', if_true: files('vhost-user.c')) endif system_ss.add(when: 'CONFIG_VIRTIO_CRYPTO', if_true: files('cryptodev-vhost.c')) +if have_iommufd + system_ss.add(files('iommufd.c')) +else + system_ss.add(files('iommufd-stub.c')) +endif if have_vhost_user_crypto system_ss.add(when: 'CONFIG_VIRTIO_CRYPTO', if_true: files('cryptodev-vhost-user.c')) endif diff --git a/backends/trace-events b/backends/trace-events index 652eb76a57..e5f828bca2 100644 --- a/backends/trace-events +++ b/backends/trace-events @@ -5,3 +5,15 @@ dbus_vmstate_pre_save(void) dbus_vmstate_post_load(int version_id) "version_id: %d" dbus_vmstate_loading(const char *id) "id: %s" dbus_vmstate_saving(const char *id) "id: %s" + +# iommufd.c +iommufd_backend_connect(int fd, bool owned, uint32_t users, int ret) "fd=%d owned=%d users=%d (%d)" +iommufd_backend_disconnect(int fd, uint32_t users) "fd=%d users=%d" +iommu_backend_set_fd(int fd) "pre-opened /dev/iommu fd=%d" +iommufd_backend_get_ioas(int iommufd, uint32_t ioas, int ret) " iommufd=%d ioas=%d (%d)" +iommufd_backend_put_ioas(int iommufd, uint32_t ioas) " iommufd=%d ioas=%d" +iommufd_backend_map_dma(int iommufd, uint32_t ioas, uint64_t iova, uint64_t size, void *vaddr, bool readonly, int ret) " iommufd=%d ioas=%d iova=0x%"PRIx64" size=0x%"PRIx64" addr=%p readonly=%d (%d)" +iommufd_backend_unmap_dma(int iommufd, uint32_t ioas, uint64_t iova, uint64_t size, int ret) " iommufd=%d ioas=%d iova=0x%"PRIx64" size=0x%"PRIx64" (%d)" +iommufd_backend_alloc_ioas(int iommufd, uint32_t ioas, int ret) " iommufd=%d ioas=%d (%d)" +iommufd_backend_free_id(int iommufd, uint32_t id, int ret) " iommufd=%d id=%d (%d)" +iommufd_backend_alloc_hwpt(int iommufd, uint32_t dev_id, uint32_t pt_id, uint32_t out_hwpt_id, int ret) " iommufd=%d dev_id=%u pt_id=%u out_hwpt=%u (%d)" diff --git a/qemu-options.hx b/qemu-options.hx index e26230bac5..ddfaddf8ce 100644 --- a/qemu-options.hx +++ b/qemu-options.hx @@ -5210,6 +5210,19 @@ SRST The ``share`` boolean option is on by default with memfd. +#ifdef CONFIG_IOMMUFD + ``-object iommufd,id=id[,fd=fd]`` + Creates an iommufd backend which allows control of DMA mapping + through the /dev/iommu device. + + The ``id`` parameter is a unique ID which frontends (such as + vfio-pci of vdpa) will use to connect with the iommufd backend. + + The ``fd`` parameter is an optional pre-opened file descriptor + resulting from /dev/iommu opening. Usually the iommufd is shared + across all subsystems, bringing the benefit of centralized + reference counting. +#endif ``-object rng-builtin,id=id`` Creates a random number generator backend which obtains entropy from QEMU builtin functions. The ``id`` parameter is a unique ID From patchwork Thu Nov 2 07:12:48 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858335 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=ZxGvv18S; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbDV5rQSz1yQ4 for ; Thu, 2 Nov 2023 18:32:10 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySA8-0008Bq-7i; Thu, 02 Nov 2023 03:30:44 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySA5-0007uo-GH for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:42 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySA3-0003PH-El for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:41 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910239; x=1730446239; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=HEgrWyu8R1RdkYN1ejrLTS512+fFMo6yzI9EdpNqkGU=; b=ZxGvv18SSrCjY45u8u+kiO2gURPdf10QgXYrgEPVXlkLGXp9V2kG28l7 CwhCvpbcW7ctTthQhGWeXKbSyqalKLAOlT5THk+TSngT9r+jXl1xYDc24 uo2y8i5VVhbkdaBPRTtwOP5XSWPgQypC8XZtp5Y9f9m3RdQszxsbxA8UK 2xcpljvQg8ZgWfnJPWzMNVo+TrdCj2SerFmUzTxWe2mKkTBDusYKzmiIF jipl2LF7qGBnRXcdtTnEn9YabNy74GBlme6Az8lnSTCs6I+UVIvsIieU0 uADysZBtZh4V9Sew7naP0/2YtGGYUEr/w6QnyqvpkUSrs5S3oEPaiVTUc w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953311" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953311" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:25 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055634" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055634" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:21 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 27/41] util/char_dev: Add open_cdev() Date: Thu, 2 Nov 2023 15:12:48 +0800 Message-Id: <20231102071302.1818071-28-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Yi Liu /dev/vfio/devices/vfioX may not exist. In that case it is still possible to open /dev/char/$major:$minor instead. Add helper function to abstract the cdev open. Suggested-by: Jason Gunthorpe Signed-off-by: Yi Liu Signed-off-by: Zhenzhong Duan --- MAINTAINERS | 6 +++ include/qemu/chardev_open.h | 16 ++++++++ util/chardev_open.c | 81 +++++++++++++++++++++++++++++++++++++ util/meson.build | 1 + 4 files changed, 104 insertions(+) create mode 100644 include/qemu/chardev_open.h create mode 100644 util/chardev_open.c diff --git a/MAINTAINERS b/MAINTAINERS index 6f35159255..eada773975 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3473,6 +3473,12 @@ S: Maintained F: include/qemu/iova-tree.h F: util/iova-tree.c +cdev Open +M: Yi Liu +S: Maintained +F: include/qemu/chardev_open.h +F: util/chardev_open.c + elf2dmp M: Viktor Prutyanov S: Maintained diff --git a/include/qemu/chardev_open.h b/include/qemu/chardev_open.h new file mode 100644 index 0000000000..64e8fcfdcb --- /dev/null +++ b/include/qemu/chardev_open.h @@ -0,0 +1,16 @@ +/* + * QEMU Chardev Helper + * + * Copyright (C) 2023 Intel Corporation. + * + * Authors: Yi Liu + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + */ + +#ifndef QEMU_CHARDEV_OPEN_H +#define QEMU_CHARDEV_OPEN_H + +int open_cdev(const char *devpath, dev_t cdev); +#endif diff --git a/util/chardev_open.c b/util/chardev_open.c new file mode 100644 index 0000000000..f776429788 --- /dev/null +++ b/util/chardev_open.c @@ -0,0 +1,81 @@ +/* + * Copyright (c) 2019, Mellanox Technologies. All rights reserved. + * Copyright (C) 2023 Intel Corporation. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + * + * Authors: Yi Liu + * + * Copied from + * https://github.com/linux-rdma/rdma-core/blob/master/util/open_cdev.c + * + */ + +#include "qemu/osdep.h" +#include "qemu/chardev_open.h" + +static int open_cdev_internal(const char *path, dev_t cdev) +{ + struct stat st; + int fd; + + fd = qemu_open_old(path, O_RDWR); + if (fd == -1) { + return -1; + } + if (fstat(fd, &st) || !S_ISCHR(st.st_mode) || + (cdev != 0 && st.st_rdev != cdev)) { + close(fd); + return -1; + } + return fd; +} + +static int open_cdev_robust(dev_t cdev) +{ + g_autofree char *devpath = NULL; + + /* + * This assumes that udev is being used and is creating the /dev/char/ + * symlinks. + */ + devpath = g_strdup_printf("/dev/char/%u:%u", major(cdev), minor(cdev)); + return open_cdev_internal(devpath, cdev); +} + +int open_cdev(const char *devpath, dev_t cdev) +{ + int fd; + + fd = open_cdev_internal(devpath, cdev); + if (fd == -1 && cdev != 0) { + return open_cdev_robust(cdev); + } + return fd; +} diff --git a/util/meson.build b/util/meson.build index eb677b40c2..eda0b06062 100644 --- a/util/meson.build +++ b/util/meson.build @@ -107,6 +107,7 @@ if have_block util_ss.add(files('filemonitor-stub.c')) endif util_ss.add(when: 'CONFIG_LINUX', if_true: files('vfio-helpers.c')) + util_ss.add(when: 'CONFIG_LINUX', if_true: files('chardev_open.c')) endif if cpu == 'aarch64' From patchwork Thu Nov 2 07:12:49 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858349 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=QRL0xD0B; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbJC5T5Tz1yQq for ; Thu, 2 Nov 2023 18:35:23 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAQ-0000QN-5v; Thu, 02 Nov 2023 03:31:02 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAB-0008K0-6R for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:50 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySA7-0003Ox-Ak for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:45 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910243; x=1730446243; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=lR0aRuT655G39EX168JFPH2V0AaPW6r49Pj5N7e+Cu8=; b=QRL0xD0BFnkGoDFwzIs8cIyNgyNXvfn7yfziKGJRjJsPfJJpHn5EYTVp n8QDViwziZUVR9X8DmYubPguvoBMyncNK+lffP7vpdkgqaAPAYkfEvI8X Fmy9Wzc+sTGWUj43Mo23/18BL2xmblCL/i6NjhQE3Qz6VXl7/eFsHgKk6 pKHzhr7bl+Tc1OGBYxGdN3UlX88nQYpeI0KwpjGltIQ0N8sIu7tBVDS6H twtWff6mrpXJ1anmuR7IrVBQA4g3ICFFfyrVzsiKqZ8alzTeYQ8R0DbsR bFUAPPUPDSBPGHpd0K4i534ibNFoGdUPic2WdfILwvAzKwmI6Ls5La+aF A==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953327" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953327" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:29 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055643" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055643" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:25 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 28/41] vfio/iommufd: Implement the iommufd backend Date: Thu, 2 Nov 2023 15:12:49 +0800 Message-Id: <20231102071302.1818071-29-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Yi Liu Add the iommufd backend. The IOMMUFD container class is implemented based on the new /dev/iommu user API. This backend obviously depends on CONFIG_IOMMUFD. So far, the iommufd backend doesn't support dirty page sync yet due to missing support in the host kernel. Co-authored-by: Eric Auger Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Zhenzhong Duan --- v4: use SPDX identifier, use iommufd_cdev_* prefix, merge with manual alloc patch include/hw/vfio/vfio-common.h | 23 ++ hw/vfio/common.c | 19 +- hw/vfio/iommufd.c | 504 ++++++++++++++++++++++++++++++++++ hw/vfio/meson.build | 3 + hw/vfio/trace-events | 13 + 5 files changed, 558 insertions(+), 4 deletions(-) create mode 100644 hw/vfio/iommufd.c diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 24ecc0e7ee..3f1a39a991 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -89,6 +89,23 @@ typedef struct VFIOHostDMAWindow { QLIST_ENTRY(VFIOHostDMAWindow) hostwin_next; } VFIOHostDMAWindow; +#ifdef CONFIG_IOMMUFD +typedef struct VFIOIOASHwpt { + uint32_t hwpt_id; + QLIST_HEAD(, VFIODevice) device_list; + QLIST_ENTRY(VFIOIOASHwpt) next; +} VFIOIOASHwpt; + +typedef struct IOMMUFDBackend IOMMUFDBackend; + +typedef struct VFIOIOMMUFDContainer { + VFIOContainerBase bcontainer; + IOMMUFDBackend *be; + uint32_t ioas_id; + QLIST_HEAD(, VFIOIOASHwpt) hwpt_list; +} VFIOIOMMUFDContainer; +#endif + typedef struct VFIODeviceOps VFIODeviceOps; typedef struct VFIODevice { @@ -116,6 +133,11 @@ typedef struct VFIODevice { OnOffAuto pre_copy_dirty_page_tracking; bool dirty_pages_supported; bool dirty_tracking; +#ifdef CONFIG_IOMMUFD + int devid; + VFIOIOASHwpt *hwpt; + IOMMUFDBackend *iommufd; +#endif } VFIODevice; struct VFIODeviceOps { @@ -201,6 +223,7 @@ typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIOGroupList vfio_group_list; extern VFIODeviceList vfio_device_list; extern const VFIOIOMMUOps vfio_legacy_ops; +extern const VFIOIOMMUOps vfio_iommufd_ops; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 572ae7c934..a61dce2845 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -1462,10 +1462,13 @@ VFIOAddressSpace *vfio_get_address_space(AddressSpace *as) void vfio_put_address_space(VFIOAddressSpace *space) { - if (QLIST_EMPTY(&space->containers)) { - QLIST_REMOVE(space, list); - g_free(space); + if (!QLIST_EMPTY(&space->containers)) { + return; } + + QLIST_REMOVE(space, list); + g_free(space); + if (QLIST_EMPTY(&vfio_address_spaces)) { qemu_unregister_reset(vfio_reset_handler, NULL); } @@ -1498,8 +1501,16 @@ retry: int vfio_attach_device(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { - const VFIOIOMMUOps *ops = &vfio_legacy_ops; + const VFIOIOMMUOps *ops; +#ifdef CONFIG_IOMMUFD + if (vbasedev->iommufd) { + ops = &vfio_iommufd_ops; + } else +#endif + { + ops = &vfio_legacy_ops; + } return ops->attach_device(name, vbasedev, as, errp); } diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c new file mode 100644 index 0000000000..1bb55ca2c4 --- /dev/null +++ b/hw/vfio/iommufd.c @@ -0,0 +1,504 @@ +/* + * iommufd container backend + * + * Copyright (C) 2023 Intel Corporation. + * Copyright Red Hat, Inc. 2023 + * + * Authors: Yi Liu + * Eric Auger + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include +#include +#include + +#include "hw/vfio/vfio-common.h" +#include "qemu/error-report.h" +#include "trace.h" +#include "qapi/error.h" +#include "sysemu/iommufd.h" +#include "hw/qdev-core.h" +#include "sysemu/reset.h" +#include "qemu/cutils.h" +#include "qemu/chardev_open.h" + +static int iommufd_map(VFIOContainerBase *bcontainer, hwaddr iova, + ram_addr_t size, void *vaddr, bool readonly) +{ + VFIOIOMMUFDContainer *container = + container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); + + return iommufd_backend_map_dma(container->be, + container->ioas_id, + iova, size, vaddr, readonly); +} + +static int iommufd_unmap(VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + IOMMUTLBEntry *iotlb) +{ + VFIOIOMMUFDContainer *container = + container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); + + /* TODO: Handle dma_unmap_bitmap with iotlb args (migration) */ + return iommufd_backend_unmap_dma(container->be, + container->ioas_id, iova, size); +} + +static void iommufd_cdev_kvm_device_add(VFIODevice *vbasedev) +{ + Error *err = NULL; + + if (vfio_kvm_device_add_fd(vbasedev->fd, &err)) { + error_report_err(err); + } +} + +static void iommufd_cdev_kvm_device_del(VFIODevice *vbasedev) +{ + Error *err = NULL; + + if (vfio_kvm_device_del_fd(vbasedev->fd, &err)) { + error_report_err(err); + } +} + +static int iommufd_connect_and_bind(VFIODevice *vbasedev, Error **errp) +{ + IOMMUFDBackend *iommufd = vbasedev->iommufd; + struct vfio_device_bind_iommufd bind = { + .argsz = sizeof(bind), + .flags = 0, + }; + int ret; + + ret = iommufd_backend_connect(iommufd, errp); + if (ret) { + return ret; + } + + /* + * Add device to kvm-vfio to be prepared for the tracking + * in KVM. Especially for some emulated devices, it requires + * to have kvm information in the device open. + */ + iommufd_cdev_kvm_device_add(vbasedev); + + /* Bind device to iommufd */ + bind.iommufd = iommufd->fd; + ret = ioctl(vbasedev->fd, VFIO_DEVICE_BIND_IOMMUFD, &bind); + if (ret) { + error_setg_errno(errp, errno, "error bind device fd=%d to iommufd=%d", + vbasedev->fd, bind.iommufd); + goto err_bind; + } + + vbasedev->devid = bind.out_devid; + trace_iommufd_connect_and_bind(bind.iommufd, vbasedev->name, vbasedev->fd, + vbasedev->devid); + return ret; +err_bind: + iommufd_cdev_kvm_device_del(vbasedev); + iommufd_backend_disconnect(iommufd); + return ret; +} + +static void iommufd_unbind_and_disconnect(VFIODevice *vbasedev) +{ + /* Unbind is automatically conducted when device fd is closed */ + iommufd_cdev_kvm_device_del(vbasedev); + iommufd_backend_disconnect(vbasedev->iommufd); +} + +static int iommufd_cdev_getfd(const char *sysfs_path, Error **errp) +{ + long int ret = -ENOTTY; + char *path, *vfio_dev_path = NULL, *vfio_path = NULL; + DIR *dir = NULL; + struct dirent *dent; + gchar *contents; + struct stat st; + gsize length; + int major, minor; + dev_t vfio_devt; + + path = g_strdup_printf("%s/vfio-dev", sysfs_path); + if (stat(path, &st) < 0) { + error_setg_errno(errp, errno, "no such host device"); + goto out_free_path; + } + + dir = opendir(path); + if (!dir) { + error_setg_errno(errp, errno, "couldn't open dirrectory %s", path); + goto out_free_path; + } + + while ((dent = readdir(dir))) { + if (!strncmp(dent->d_name, "vfio", 4)) { + vfio_dev_path = g_strdup_printf("%s/%s/dev", path, dent->d_name); + break; + } + } + + if (!vfio_dev_path) { + error_setg(errp, "failed to find vfio-dev/vfioX/dev"); + goto out_close_dir; + } + + if (!g_file_get_contents(vfio_dev_path, &contents, &length, NULL)) { + error_setg(errp, "failed to load \"%s\"", vfio_dev_path); + goto out_free_dev_path; + } + + if (sscanf(contents, "%d:%d", &major, &minor) != 2) { + error_setg(errp, "failed to get major:minor for \"%s\"", vfio_dev_path); + goto out_free_dev_path; + } + g_free(contents); + vfio_devt = makedev(major, minor); + + vfio_path = g_strdup_printf("/dev/vfio/devices/%s", dent->d_name); + ret = open_cdev(vfio_path, vfio_devt); + if (ret < 0) { + error_setg(errp, "Failed to open %s", vfio_path); + } + + trace_iommufd_cdev_getfd(vfio_path, ret); + g_free(vfio_path); + +out_free_dev_path: + g_free(vfio_dev_path); +out_close_dir: + closedir(dir); +out_free_path: + if (*errp) { + error_prepend(errp, VFIO_MSG_PREFIX, path); + } + g_free(path); + + return ret; +} + +static VFIOIOASHwpt *iommufd_container_get_hwpt(VFIOIOMMUFDContainer *container, + uint32_t hwpt_id) +{ + VFIOIOASHwpt *hwpt; + + QLIST_FOREACH(hwpt, &container->hwpt_list, next) { + if (hwpt->hwpt_id == hwpt_id) { + return hwpt; + } + } + + hwpt = g_malloc0(sizeof(*hwpt)); + + hwpt->hwpt_id = hwpt_id; + QLIST_INIT(&hwpt->device_list); + QLIST_INSERT_HEAD(&container->hwpt_list, hwpt, next); + + return hwpt; +} + +static void iommufd_container_put_hwpt(IOMMUFDBackend *be, VFIOIOASHwpt *hwpt) +{ + QLIST_REMOVE(hwpt, next); + iommufd_backend_free_id(be->fd, hwpt->hwpt_id); + g_free(hwpt); +} + +static int iommufd_cdev_attach_hwpt(VFIODevice *vbasedev, uint32_t hwpt_id, + Error **errp) +{ + int ret, iommufd = vbasedev->iommufd->fd; + struct vfio_device_attach_iommufd_pt attach_data = { + .argsz = sizeof(attach_data), + .flags = 0, + .pt_id = hwpt_id, + }; + + /* Attach device to an hwpt within iommufd */ + ret = ioctl(vbasedev->fd, VFIO_DEVICE_ATTACH_IOMMUFD_PT, &attach_data); + if (ret) { + error_setg_errno(errp, errno, + "[iommufd=%d] error attach %s (%d) to hwpt_id=%d", + iommufd, vbasedev->name, vbasedev->fd, hwpt_id); + } + trace_iommufd_cdev_attach_hwpt(iommufd, vbasedev->name, vbasedev->fd, + hwpt_id); + return ret; +} + +static int iommufd_cdev_detach_hwpt(VFIODevice *vbasedev, Error **errp) +{ + int ret, iommufd = vbasedev->iommufd->fd; + struct vfio_device_detach_iommufd_pt detach_data = { + .argsz = sizeof(detach_data), + .flags = 0, + }; + + ret = ioctl(vbasedev->fd, VFIO_DEVICE_DETACH_IOMMUFD_PT, &detach_data); + if (ret) { + error_setg_errno(errp, errno, "detach %s from ioas failed", + vbasedev->name); + } + trace_iommufd_cdev_detach_hwpt(iommufd, vbasedev->name, + vbasedev->hwpt->hwpt_id); + return ret; +} + +static int iommufd_cdev_attach_container(VFIODevice *vbasedev, + VFIOIOMMUFDContainer *container, + Error **errp) +{ + int ret, iommufd = vbasedev->iommufd->fd; + VFIOIOASHwpt *hwpt; + uint32_t hwpt_id; + Error *err = NULL; + + /* try to attach to an existing hwpt in this container */ + QLIST_FOREACH(hwpt, &container->hwpt_list, next) { + ret = iommufd_cdev_attach_hwpt(vbasedev, hwpt->hwpt_id, &err); + if (ret) { + const char *msg = error_get_pretty(err); + + trace_iommufd_cdev_fail_attach_existing_hwpt(msg); + error_free(err); + err = NULL; + } else { + goto found_hwpt; + } + } + + ret = iommufd_backend_alloc_hwpt(iommufd, vbasedev->devid, + container->ioas_id, &hwpt_id); + + if (ret) { + error_setg_errno(errp, errno, "error alloc shadow hwpt"); + return ret; + } + + /* Attach cdev to a new allocated hwpt within iommufd */ + ret = iommufd_cdev_attach_hwpt(vbasedev, hwpt_id, errp); + if (ret) { + iommufd_backend_free_id(iommufd, hwpt_id); + return ret; + } + + hwpt = iommufd_container_get_hwpt(container, hwpt_id); +found_hwpt: + QLIST_INSERT_HEAD(&hwpt->device_list, vbasedev, next); + vbasedev->hwpt = hwpt; + + trace_iommufd_cdev_attach_container(iommufd, vbasedev->name, vbasedev->fd, + container->ioas_id, hwpt->hwpt_id); + return ret; +} + +static void iommufd_cdev_detach_container(VFIODevice *vbasedev, + VFIOIOMMUFDContainer *container) +{ + VFIOIOASHwpt *hwpt = vbasedev->hwpt; + Error *err = NULL; + int ret; + + ret = iommufd_cdev_detach_hwpt(vbasedev, &err); + if (ret) { + error_report_err(err); + } + + QLIST_REMOVE(vbasedev, next); + vbasedev->hwpt = NULL; + if (QLIST_EMPTY(&hwpt->device_list)) { + iommufd_container_put_hwpt(vbasedev->iommufd, hwpt); + } + + trace_iommufd_cdev_detach_container(container->be->fd, vbasedev->name, + container->ioas_id); +} + +static void iommufd_container_destroy(VFIOIOMMUFDContainer *container) +{ + VFIOContainerBase *bcontainer = &container->bcontainer; + + if (!QLIST_EMPTY(&container->hwpt_list)) { + return; + } + memory_listener_unregister(&bcontainer->listener); + vfio_container_destroy(bcontainer); + iommufd_backend_put_ioas(container->be, container->ioas_id); + g_free(container); +} + +static int iommufd_ram_block_discard_disable(bool state) +{ + /* + * We support coordinated discarding of RAM via the RamDiscardManager. + */ + return ram_block_uncoordinated_discard_disable(state); +} + +static int iommufd_attach_device(const char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp) +{ + VFIOContainerBase *bcontainer; + VFIOIOMMUFDContainer *container; + VFIOAddressSpace *space; + struct vfio_device_info dev_info = { .argsz = sizeof(dev_info) }; + int ret, devfd; + uint32_t ioas_id; + Error *err = NULL; + + devfd = iommufd_cdev_getfd(vbasedev->sysfsdev, errp); + if (devfd < 0) { + return devfd; + } + vbasedev->fd = devfd; + + ret = iommufd_connect_and_bind(vbasedev, errp); + if (ret) { + goto err_connect_bind; + } + + space = vfio_get_address_space(as); + + /* try to attach to an existing container in this space */ + QLIST_FOREACH(bcontainer, &space->containers, next) { + container = container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); + if (bcontainer->ops != &vfio_iommufd_ops || + vbasedev->iommufd != container->be) { + continue; + } + if (iommufd_cdev_attach_container(vbasedev, container, &err)) { + const char *msg = error_get_pretty(err); + + trace_iommufd_cdev_fail_attach_existing_container(msg); + error_free(err); + err = NULL; + } else { + ret = iommufd_ram_block_discard_disable(true); + if (ret) { + error_setg(errp, + "Cannot set discarding of RAM broken (%d)", ret); + goto err_discard_disable; + } + goto found_container; + } + } + + /* Need to allocate a new dedicated container */ + ret = iommufd_backend_get_ioas(vbasedev->iommufd, &ioas_id); + if (ret < 0) { + error_setg_errno(errp, errno, "Failed to alloc ioas"); + goto err_get_ioas; + } + + trace_iommufd_cdev_alloc_ioas(vbasedev->iommufd->fd, ioas_id); + + container = g_malloc0(sizeof(*container)); + container->be = vbasedev->iommufd; + container->ioas_id = ioas_id; + QLIST_INIT(&container->hwpt_list); + + bcontainer = &container->bcontainer; + vfio_container_init(bcontainer, space, &vfio_iommufd_ops); + QLIST_INSERT_HEAD(&space->containers, bcontainer, next); + + ret = iommufd_cdev_attach_container(vbasedev, container, errp); + if (ret) { + goto err_attach_container; + } + + ret = iommufd_ram_block_discard_disable(true); + if (ret) { + goto err_discard_disable; + } + + bcontainer->pgsizes = qemu_real_host_page_size(); + + bcontainer->listener = vfio_memory_listener; + memory_listener_register(&bcontainer->listener, bcontainer->space->as); + + if (bcontainer->error) { + ret = -1; + error_propagate_prepend(errp, bcontainer->error, + "memory listener initialization failed: "); + goto err_listener_register; + } + + bcontainer->initialized = true; + +found_container: + ret = ioctl(devfd, VFIO_DEVICE_GET_INFO, &dev_info); + if (ret) { + error_setg_errno(errp, errno, "error getting device info"); + goto err_listener_register; + } + + /* + * TODO: examine RAM_BLOCK_DISCARD stuff, should we do group level + * for discarding incompatibility check as well? + */ + if (vbasedev->ram_block_discard_allowed) { + iommufd_ram_block_discard_disable(false); + } + + vbasedev->group = 0; + vbasedev->num_irqs = dev_info.num_irqs; + vbasedev->num_regions = dev_info.num_regions; + vbasedev->flags = dev_info.flags; + vbasedev->reset_works = !!(dev_info.flags & VFIO_DEVICE_FLAGS_RESET); + vbasedev->bcontainer = bcontainer; + QLIST_INSERT_HEAD(&bcontainer->device_list, vbasedev, container_next); + QLIST_INSERT_HEAD(&vfio_device_list, vbasedev, global_next); + + trace_iommufd_cdev_device_info(vbasedev->name, devfd, vbasedev->num_irqs, + vbasedev->num_regions, vbasedev->flags); + return 0; + +err_listener_register: + iommufd_ram_block_discard_disable(false); +err_discard_disable: + iommufd_cdev_detach_container(vbasedev, container); +err_attach_container: + iommufd_container_destroy(container); +err_get_ioas: + vfio_put_address_space(space); + iommufd_unbind_and_disconnect(vbasedev); +err_connect_bind: + close(vbasedev->fd); + return ret; +} + +static void iommufd_detach_device(VFIODevice *vbasedev) +{ + VFIOContainerBase *bcontainer = vbasedev->bcontainer; + VFIOIOMMUFDContainer *container; + VFIOAddressSpace *space = bcontainer->space; + + QLIST_REMOVE(vbasedev, global_next); + QLIST_REMOVE(vbasedev, container_next); + vbasedev->bcontainer = NULL; + + if (!vbasedev->ram_block_discard_allowed) { + iommufd_ram_block_discard_disable(false); + } + + container = container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); + iommufd_cdev_detach_container(vbasedev, container); + iommufd_container_destroy(container); + vfio_put_address_space(space); + + iommufd_unbind_and_disconnect(vbasedev); + close(vbasedev->fd); +} + +const VFIOIOMMUOps vfio_iommufd_ops = { + .dma_map = iommufd_map, + .dma_unmap = iommufd_unmap, + .attach_device = iommufd_attach_device, + .detach_device = iommufd_detach_device, +}; diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index eb6ce6229d..9cae2c9e21 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -7,6 +7,9 @@ vfio_ss.add(files( 'spapr.c', 'migration.c', )) +if have_iommufd + vfio_ss.add(files('iommufd.c')) +endif vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files( 'display.c', 'pci-quirks.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index 08a1f9dfa4..d85342b65f 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -164,3 +164,16 @@ vfio_state_pending_estimate(const char *name, uint64_t precopy, uint64_t postcop vfio_state_pending_exact(const char *name, uint64_t precopy, uint64_t postcopy, uint64_t stopcopy_size, uint64_t precopy_init_size, uint64_t precopy_dirty_size) " (%s) precopy 0x%"PRIx64" postcopy 0x%"PRIx64" stopcopy size 0x%"PRIx64" precopy initial size 0x%"PRIx64" precopy dirty size 0x%"PRIx64 vfio_vmstate_change(const char *name, int running, const char *reason, const char *dev_state) " (%s) running %d reason %s device state %s" vfio_vmstate_change_prepare(const char *name, int running, const char *reason, const char *dev_state) " (%s) running %d reason %s device state %s" + +#iommufd.c + +iommufd_connect_and_bind(int iommufd, const char *name, int devfd, int devid) " [iommufd=%d] Successfully bound device %s (fd=%d): output devid=%d" +iommufd_cdev_getfd(const char *dev, int devfd) " %s (fd=%d)" +iommufd_cdev_attach_hwpt(int iommufd, const char *name, int devfd, int hwptid) " [iommufd=%d] Successfully attached device %s (%d) to hwptd=%d" +iommufd_cdev_detach_hwpt(int iommufd, const char *name, int hwptid) " [iommufd=%d] Detached %s from hwpt=%d" +iommufd_cdev_fail_attach_existing_hwpt(const char *msg) " %s" +iommufd_cdev_attach_container(int iommufd, const char *name, int devfd, int ioasid, int hwptid) " [iommufd=%d] Successfully attached device %s (%d) to ioasid=%d: output hwptd=%d" +iommufd_cdev_detach_container(int iommufd, const char *name, int ioasid) " [iommufd=%d] Detached %s from ioasid=%d" +iommufd_cdev_fail_attach_existing_container(const char *msg) " %s" +iommufd_cdev_alloc_ioas(int iommufd, int ioas_id) " [iommufd=%d] new IOMMUFD container with ioasid=%d" +iommufd_cdev_device_info(char *name, int devfd, int num_irqs, int num_regions, int flags) " %s (%d) num_irqs=%d num_regions=%d flags=%d" From patchwork Thu Nov 2 07:12:50 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858358 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=lLutUUvK; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbL34WWnz1yQq for ; Thu, 2 Nov 2023 18:36:59 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAZ-0000ur-7j; Thu, 02 Nov 2023 03:31:11 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAH-0000Hs-Oc for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:58 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAG-0003PH-02 for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:53 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910252; x=1730446252; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=5ejhTNqCwupSTAJs1yfEG2tXWeKhQlFW3fCS+eeO2Bc=; b=lLutUUvKh+Zp4kwmjGoqY7ufVJE3HF+cdgTTQAqzGXIuL+ZF1Y1/DuXX wOoDYnnMzUJ02UskkXYaIGvR+lmDfP4RGe7hgQqaQzTrS9kRIQeM7BlJO AVEVVtHUTNdC12SKU6KLM7OnyPeTmOJsiaI6oJdSWaqIrt6hKUNWOt06z P8jWBUMcBFY55q3B3jjmpWnaFmfz0cLK0Grs82XmYMwRcttcl11g/E7LQ 5xCdz6C0unoEiodoDTXd37LvQoJ4PmBr4mPX1PtrzZ6lOhTMmFR+VTwbj lMiOCnJCjmYcHpqX6/SA7HmRHGxu9yK2dTRuVZHOZwEsYGkukikSx1P9b g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953334" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953334" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:32 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055665" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055665" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:29 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 29/41] vfio/iommufd: Relax assert check for iommufd backend Date: Thu, 2 Nov 2023 15:12:50 +0800 Message-Id: <20231102071302.1818071-30-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Currently iommufd doesn't support dirty page sync yet, but it will not block us doing live migration if VFIO migration is force enabled. So in this case we allow set_dirty_page_tracking to be NULL. Note we don't need same change for query_dirty_bitmap because when dirty page sync isn't supported, query_dirty_bitmap will never be called. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan --- hw/vfio/container-base.c | 4 ++++ hw/vfio/container.c | 4 ---- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index 71f7274973..eee2dcfe76 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -55,6 +55,10 @@ void vfio_container_del_section_window(VFIOContainerBase *bcontainer, int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start) { + if (!bcontainer->dirty_pages_supported) { + return 0; + } + g_assert(bcontainer->ops->set_dirty_page_tracking); return bcontainer->ops->set_dirty_page_tracking(bcontainer, start); } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index 6bacf38222..ed2d721b2b 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -216,10 +216,6 @@ static int vfio_legacy_set_dirty_page_tracking(VFIOContainerBase *bcontainer, .argsz = sizeof(dirty), }; - if (!bcontainer->dirty_pages_supported) { - return 0; - } - if (start) { dirty.flags = VFIO_IOMMU_DIRTY_PAGES_FLAG_START; } else { From patchwork Thu Nov 2 07:12:51 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858330 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=G+jvX6hi; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbCd5KY7z1yQ4 for ; Thu, 2 Nov 2023 18:31:25 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAW-0000eZ-IR; Thu, 02 Nov 2023 03:31:08 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAJ-0000LU-3G for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:58 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAG-0003OZ-0Y for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:54 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910252; x=1730446252; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=r/VMm/1Z4Z01iP7zIyomy6106isBuemKDquVKatVN7I=; b=G+jvX6hiKAeJOxQcQx8h3+i5mdssYIIlFvRouEq4wSUmDhiwqlFdtGsr 1cfOXzxOElHgVZGafSv7+R2RxqCc9sUwxvhof+kU8HIpWUj2nfw2HUzxO FfBoOZoKso0JNVUM/7ahwRv7Tk+d4YaiTWz6T3dS49evfBy81XX4eAn5y z8ori6ZM/Dmke6+DiWCeeFcZg0hVDQmmsAWHBedcnjbBw8ZXKv4z2hxIl G5ts5D1dW7ZMIjXbwGfYqeDMWeD51rtI3zaJkX6/w000GUaPyo75A0PqN ehpfNFky8V2ubCODFp/ObJ9sM+D/XR9ZLKVV3YXvXvLvBpLSZEWGgzd08 w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953340" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953340" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:36 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055678" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055678" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:33 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 30/41] vfio/iommufd: Add support for iova_ranges Date: Thu, 2 Nov 2023 15:12:51 +0800 Message-Id: <20231102071302.1818071-31-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Some vIOMMU such as virtio-iommu use iova ranges from host side to setup reserved ranges for passthrough device, so that guest will not use an iova range beyond host support. Use an uAPI of IOMMUFD to get iova ranges of host side and pass to vIOMMU just like the legacy backend. Signed-off-by: Zhenzhong Duan --- v4: fix build error in 32bit fedora hw/vfio/iommufd.c | 47 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 47 insertions(+) diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index 1bb55ca2c4..22f02f92a9 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -341,6 +341,52 @@ static int iommufd_ram_block_discard_disable(bool state) return ram_block_uncoordinated_discard_disable(state); } +static int vfio_get_info_iova_range(VFIOIOMMUFDContainer *container, + uint32_t ioas_id) +{ + VFIOContainerBase *bcontainer = &container->bcontainer; + struct iommu_ioas_iova_ranges *info; + struct iommu_iova_range *iova_ranges; + int ret, sz, fd = container->be->fd; + + info = g_malloc0(sizeof(*info)); + info->size = sizeof(*info); + info->ioas_id = ioas_id; + + ret = ioctl(fd, IOMMU_IOAS_IOVA_RANGES, info); + if (ret && errno != EMSGSIZE) { + goto error; + } + + sz = info->num_iovas * sizeof(struct iommu_iova_range); + info = g_realloc(info, sizeof(*info) + sz); + info->allowed_iovas = (uintptr_t)(info + 1); + + ret = ioctl(fd, IOMMU_IOAS_IOVA_RANGES, info); + if (ret) { + goto error; + } + + iova_ranges = (struct iommu_iova_range *)info->allowed_iovas; + + for (int i = 0; i < info->num_iovas; i++) { + Range *range = g_new(Range, 1); + + range_set_bounds(range, iova_ranges[i].start, iova_ranges[i].last); + bcontainer->iova_ranges = + range_list_insert(bcontainer->iova_ranges, range); + } + + g_free(info); + return 0; + +error: + ret = -errno; + g_free(info); + error_report("vfio/iommufd: Cannot get iova ranges: %m"); + return ret; +} + static int iommufd_attach_device(const char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { @@ -418,6 +464,7 @@ static int iommufd_attach_device(const char *name, VFIODevice *vbasedev, } bcontainer->pgsizes = qemu_real_host_page_size(); + vfio_get_info_iova_range(container, ioas_id); bcontainer->listener = vfio_memory_listener; memory_listener_register(&bcontainer->listener, bcontainer->space->as); From patchwork Thu Nov 2 07:12:52 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858350 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=nAlM3qyT; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKb1qGYz1yQq for ; Thu, 2 Nov 2023 18:36:35 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAY-0000re-RI; Thu, 02 Nov 2023 03:31:10 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAM-0000N9-7o for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:58 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAK-0003Ox-Ip for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:30:58 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910256; x=1730446256; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=mgfP+9nkXo1nd7QBOZYlI1vQZnttEZ8bVbqSddms4hg=; b=nAlM3qyT7SdzRTTvdquUr3EirWGyxQF4W0ZYrr4cgKKF18EVEs3r7qlx Qq2ox0Lm63h0xPJ9QuL7CNkxoaaiJS9TdefrtRnTEGyWQcVK72s6wa5D/ 8u1uiaobcxtTHQOUttsHEcvgJoGc5tdACfrHgWufAmx3GgvEzdqhzRtXN GG5wtpI5lLlf7uYViOTSHo+tz+cDMHMQmkAgDP65jRLNGLaPvPbj4i1P3 A8SFUJbG6MO2lJDUhPC0IHE4Hmbahli6B4fIZEx380pscKaeev4n0KuGa ycVp9rIogGCuG6wFAuLXpTKMAUMxvggND7gOGsH1EFygHKzKESA/PVWUT Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953345" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953345" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:40 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055685" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055685" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:36 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 31/41] vfio/pci: Extract out a helper vfio_pci_get_pci_hot_reset_info Date: Thu, 2 Nov 2023 15:12:52 +0800 Message-Id: <20231102071302.1818071-32-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This helper will be used by both legacy and iommufd backends. No functional changes intended. Signed-off-by: Zhenzhong Duan Reviewed-by: Cédric Le Goater --- hw/vfio/pci.h | 3 +++ hw/vfio/pci.c | 54 +++++++++++++++++++++++++++++++++++---------------- 2 files changed, 40 insertions(+), 17 deletions(-) diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index fba8737ab2..1006061afb 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -218,6 +218,9 @@ void vfio_probe_igd_bar4_quirk(VFIOPCIDevice *vdev, int nr); extern const PropertyInfo qdev_prop_nv_gpudirect_clique; +int vfio_pci_get_pci_hot_reset_info(VFIOPCIDevice *vdev, + struct vfio_pci_hot_reset_info **info_p); + int vfio_populate_vga(VFIOPCIDevice *vdev, Error **errp); int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev, diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index c62c02f7b6..eb55e8ae88 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -2445,22 +2445,13 @@ static bool vfio_pci_host_match(PCIHostDeviceAddress *addr, const char *name) return (strcmp(tmp, name) == 0); } -static int vfio_pci_hot_reset(VFIOPCIDevice *vdev, bool single) +int vfio_pci_get_pci_hot_reset_info(VFIOPCIDevice *vdev, + struct vfio_pci_hot_reset_info **info_p) { - VFIOGroup *group; struct vfio_pci_hot_reset_info *info; - struct vfio_pci_dependent_device *devices; - struct vfio_pci_hot_reset *reset; - int32_t *fds; - int ret, i, count; - bool multi = false; + int ret, count; - trace_vfio_pci_hot_reset(vdev->vbasedev.name, single ? "one" : "multi"); - - if (!single) { - vfio_pci_pre_reset(vdev); - } - vdev->vbasedev.needs_reset = false; + assert(info_p && !*info_p); info = g_malloc0(sizeof(*info)); info->argsz = sizeof(*info); @@ -2468,24 +2459,53 @@ static int vfio_pci_hot_reset(VFIOPCIDevice *vdev, bool single) ret = ioctl(vdev->vbasedev.fd, VFIO_DEVICE_GET_PCI_HOT_RESET_INFO, info); if (ret && errno != ENOSPC) { ret = -errno; + g_free(info); if (!vdev->has_pm_reset) { error_report("vfio: Cannot reset device %s, " "no available reset mechanism.", vdev->vbasedev.name); } - goto out_single; + return ret; } count = info->count; - info = g_realloc(info, sizeof(*info) + (count * sizeof(*devices))); - info->argsz = sizeof(*info) + (count * sizeof(*devices)); - devices = &info->devices[0]; + info = g_realloc(info, sizeof(*info) + (count * sizeof(info->devices[0]))); + info->argsz = sizeof(*info) + (count * sizeof(info->devices[0])); ret = ioctl(vdev->vbasedev.fd, VFIO_DEVICE_GET_PCI_HOT_RESET_INFO, info); if (ret) { ret = -errno; + g_free(info); error_report("vfio: hot reset info failed: %m"); + return ret; + } + + *info_p = info; + return 0; +} + +static int vfio_pci_hot_reset(VFIOPCIDevice *vdev, bool single) +{ + VFIOGroup *group; + struct vfio_pci_hot_reset_info *info = NULL; + struct vfio_pci_dependent_device *devices; + struct vfio_pci_hot_reset *reset; + int32_t *fds; + int ret, i, count; + bool multi = false; + + trace_vfio_pci_hot_reset(vdev->vbasedev.name, single ? "one" : "multi"); + + if (!single) { + vfio_pci_pre_reset(vdev); + } + vdev->vbasedev.needs_reset = false; + + ret = vfio_pci_get_pci_hot_reset_info(vdev, &info); + + if (ret) { goto out_single; } + devices = &info->devices[0]; trace_vfio_pci_hot_reset_has_dep_devices(vdev->vbasedev.name); From patchwork Thu Nov 2 07:12:53 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858360 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=Ym+0Wtyk; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbLH66wpz1yQs for ; Thu, 2 Nov 2023 18:37:11 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAa-000155-4y; Thu, 02 Nov 2023 03:31:12 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAU-0000ag-DG for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:06 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAS-0003PH-0Q for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:05 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910264; x=1730446264; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=kSokffqzS+mxdyV6rsa1kPj97RxOw80RU6XfZVz/QCU=; b=Ym+0WtykXRX5/Ma0hoa7nxkpWkEOtzSDS7iebGpKeUug9rhPtInu64uk 3ou563QGS1M/OTE/W48Q56AKuH2zTuBdihWeAIQ3g9ZLW+m2tGNoWdC47 ZqfGQNB773J44UaTOe0Njj7tQacQN+Yc6D4yUa3yQMPio5s3tKNzBXd8j T5UJXNSx/ti5xjNNfyrWzekHSnYuDAY3KgnDRLAFjG/4gyVSbuwHprrmo y9PxAV/gvgLi25/aMx633PzB3iEM3Ne1AKmxlTvT1vQTL1BW14HSfz6ez UC4qoM5vgc12ET9YVDTpRnOm4EED0v1EvbDdp8P0IhcHisKIoApvhrgo+ Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953359" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953359" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:45 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055709" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055709" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:40 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 32/41] vfio/pci: Introduce a vfio pci hot reset interface Date: Thu, 2 Nov 2023 15:12:53 +0800 Message-Id: <20231102071302.1818071-33-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Legacy vfio pci and iommufd cdev have different process to hot reset vfio device, expand current code to abstract out pci_hot_reset callback for legacy vfio, this same interface will also be used by iommufd cdev vfio device. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan --- hw/vfio/pci.h | 1 + include/hw/vfio/vfio-container-base.h | 3 +++ hw/vfio/container.c | 2 ++ hw/vfio/pci.c | 11 ++++++++++- 4 files changed, 16 insertions(+), 1 deletion(-) diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index 1006061afb..12cc765821 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -220,6 +220,7 @@ extern const PropertyInfo qdev_prop_nv_gpudirect_clique; int vfio_pci_get_pci_hot_reset_info(VFIOPCIDevice *vdev, struct vfio_pci_hot_reset_info **info_p); +int vfio_legacy_pci_hot_reset(VFIODevice *vbasedev, bool single); int vfio_populate_vga(VFIOPCIDevice *vdev, Error **errp); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 4b6f017c6f..45bb19c767 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -106,6 +106,9 @@ struct VFIOIOMMUOps { int (*set_dirty_page_tracking)(VFIOContainerBase *bcontainer, bool start); int (*query_dirty_bitmap)(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); + /* PCI specific */ + int (*pci_hot_reset)(VFIODevice *vbasedev, bool single); + /* SPAPR specific */ int (*add_window)(VFIOContainerBase *bcontainer, MemoryRegionSection *section, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index ed2d721b2b..f27cc15d09 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -33,6 +33,7 @@ #include "trace.h" #include "qapi/error.h" #include "migration/migration.h" +#include "pci.h" VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); @@ -929,4 +930,5 @@ const VFIOIOMMUOps vfio_legacy_ops = { .detach_device = vfio_legacy_detach_device, .set_dirty_page_tracking = vfio_legacy_set_dirty_page_tracking, .query_dirty_bitmap = vfio_legacy_query_dirty_bitmap, + .pci_hot_reset = vfio_legacy_pci_hot_reset, }; diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index eb55e8ae88..a6194b7bfe 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -2483,8 +2483,9 @@ int vfio_pci_get_pci_hot_reset_info(VFIOPCIDevice *vdev, return 0; } -static int vfio_pci_hot_reset(VFIOPCIDevice *vdev, bool single) +int vfio_legacy_pci_hot_reset(VFIODevice *vbasedev, bool single) { + VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev); VFIOGroup *group; struct vfio_pci_hot_reset_info *info = NULL; struct vfio_pci_dependent_device *devices; @@ -2647,6 +2648,14 @@ out_single: return ret; } +static int vfio_pci_hot_reset(VFIOPCIDevice *vdev, bool single) +{ + VFIODevice *vbasedev = &vdev->vbasedev; + const VFIOIOMMUOps *ops = vbasedev->bcontainer->ops; + + return ops->pci_hot_reset(vbasedev, single); +} + /* * We want to differentiate hot reset of multiple in-use devices vs hot reset * of a single in-use device. VFIO_DEVICE_RESET will already handle the case From patchwork Thu Nov 2 07:12:54 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858362 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=hFM+boJi; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbLb4B2Hz1yQq for ; Thu, 2 Nov 2023 18:37:27 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAa-00017z-Dv; Thu, 02 Nov 2023 03:31:12 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAW-0000f3-Bh for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:08 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAU-0003OZ-8M for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:08 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910266; x=1730446266; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=qYbjFJsZTdqafFCCXBwd0B8hhpm9nk9rj4ooJuAp0t4=; b=hFM+boJicjju8MKSXX0eFrWk60z+q0xcAnMsXo1ZZf8iQTSpHlZbmoe7 Ca6w83hscW5tG/l1C9zftk3Fkoyj/Wr9MEpG2GTC0H+Z0AIWzZdBnHY9o hpO/i6JYW6OA1/iP7EJazDThUsvcnkNPiHazJWkz/+xTEVoNm8Mc/a2Nj vT7Q7raJgAObf6JobPYzz2P8IaX/BDLlmtHp7Tx6Fzb0SlJWtwx3XqA2p EjYoeIyugk7bPihJpn0NBga9By6+SlbbArtH8XzSWePWVvjESSs0yuRp3 pV1UsgP/iosAetz1kqkzvfcRD1xcyamGyJ1QiOyRG/pRbIgqVacBzNsOA w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953387" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953387" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:49 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055723" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055723" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:45 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 33/41] vfio/iommufd: Enable pci hot reset through iommufd cdev interface Date: Thu, 2 Nov 2023 15:12:54 +0800 Message-Id: <20231102071302.1818071-34-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Add a new callback iommufd_pci_hot_reset to do iommufd specific check and reset operation. Signed-off-by: Zhenzhong Duan --- hw/vfio/pci.h | 2 + hw/vfio/iommufd.c | 142 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 4 +- hw/vfio/trace-events | 1 + 4 files changed, 147 insertions(+), 2 deletions(-) diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index 12cc765821..ec4a03aecd 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -218,6 +218,8 @@ void vfio_probe_igd_bar4_quirk(VFIOPCIDevice *vdev, int nr); extern const PropertyInfo qdev_prop_nv_gpudirect_clique; +void vfio_pci_pre_reset(VFIOPCIDevice *vdev); +void vfio_pci_post_reset(VFIOPCIDevice *vdev); int vfio_pci_get_pci_hot_reset_info(VFIOPCIDevice *vdev, struct vfio_pci_hot_reset_info **info_p); int vfio_legacy_pci_hot_reset(VFIODevice *vbasedev, bool single); diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index 22f02f92a9..aedfe31c3c 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -24,6 +24,7 @@ #include "sysemu/reset.h" #include "qemu/cutils.h" #include "qemu/chardev_open.h" +#include "pci.h" static int iommufd_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) @@ -543,9 +544,150 @@ static void iommufd_detach_device(VFIODevice *vbasedev) close(vbasedev->fd); } +static VFIODevice *vfio_pci_find_by_iommufd_devid(__u32 devid) +{ + VFIODevice *vbasedev_iter; + + QLIST_FOREACH(vbasedev_iter, &vfio_device_list, global_next) { + if (vbasedev_iter->bcontainer->ops != &vfio_iommufd_ops) { + continue; + } + if (devid == vbasedev_iter->devid) { + return vbasedev_iter; + } + } + return NULL; +} + +static int iommufd_pci_hot_reset(VFIODevice *vbasedev, bool single) +{ + VFIOPCIDevice *vdev = container_of(vbasedev, VFIOPCIDevice, vbasedev); + struct vfio_pci_hot_reset_info *info = NULL; + struct vfio_pci_dependent_device *devices; + struct vfio_pci_hot_reset *reset; + int ret, i; + bool multi = false; + + trace_vfio_pci_hot_reset(vdev->vbasedev.name, single ? "one" : "multi"); + + if (!single) { + vfio_pci_pre_reset(vdev); + } + vdev->vbasedev.needs_reset = false; + + ret = vfio_pci_get_pci_hot_reset_info(vdev, &info); + + if (ret) { + goto out_single; + } + + assert(info->flags & VFIO_PCI_HOT_RESET_FLAG_DEV_ID); + + devices = &info->devices[0]; + + if (!(info->flags & VFIO_PCI_HOT_RESET_FLAG_DEV_ID_OWNED)) { + if (!vdev->has_pm_reset) { + for (i = 0; i < info->count; i++) { + if (devices[i].devid == VFIO_PCI_DEVID_NOT_OWNED) { + error_report("vfio: Cannot reset device %s, " + "depends on device %04x:%02x:%02x.%x " + "which is not owned.", + vdev->vbasedev.name, devices[i].segment, + devices[i].bus, PCI_SLOT(devices[i].devfn), + PCI_FUNC(devices[i].devfn)); + } + } + } + ret = -EPERM; + goto out_single; + } + + trace_vfio_pci_hot_reset_has_dep_devices(vdev->vbasedev.name); + + for (i = 0; i < info->count; i++) { + VFIOPCIDevice *tmp; + VFIODevice *vbasedev_iter; + + trace_vfio_pci_hot_reset_dep_devices_iommufd(devices[i].segment, + devices[i].bus, + PCI_SLOT(devices[i].devfn), + PCI_FUNC(devices[i].devfn), + devices[i].devid); + + /* + * If a VFIO cdev device is resettable, all the dependent devices + * are either bound to same iommufd or within same iommu_groups as + * one of the iommufd bound devices. + */ + assert(devices[i].devid != VFIO_PCI_DEVID_NOT_OWNED); + + if (devices[i].devid == vdev->vbasedev.devid || + devices[i].devid == VFIO_PCI_DEVID_OWNED) { + continue; + } + + vbasedev_iter = vfio_pci_find_by_iommufd_devid(devices[i].devid); + if (!vbasedev_iter || !vbasedev_iter->dev->realized || + vbasedev_iter->type != VFIO_DEVICE_TYPE_PCI) { + continue; + } + tmp = container_of(vbasedev_iter, VFIOPCIDevice, vbasedev); + if (single) { + ret = -EINVAL; + goto out_single; + } + vfio_pci_pre_reset(tmp); + tmp->vbasedev.needs_reset = false; + multi = true; + } + + if (!single && !multi) { + ret = -EINVAL; + goto out_single; + } + + /* Use zero length array for hot reset with iommufd backend */ + reset = g_malloc0(sizeof(*reset)); + reset->argsz = sizeof(*reset); + + /* Bus reset! */ + ret = ioctl(vdev->vbasedev.fd, VFIO_DEVICE_PCI_HOT_RESET, reset); + g_free(reset); + + trace_vfio_pci_hot_reset_result(vdev->vbasedev.name, + ret ? strerror(errno) : "Success"); + + /* Re-enable INTx on affected devices */ + for (i = 0; i < info->count; i++) { + VFIOPCIDevice *tmp; + VFIODevice *vbasedev_iter; + + if (devices[i].devid == vdev->vbasedev.devid || + devices[i].devid == VFIO_PCI_DEVID_OWNED) { + continue; + } + + vbasedev_iter = vfio_pci_find_by_iommufd_devid(devices[i].devid); + if (!vbasedev_iter || !vbasedev_iter->dev->realized || + vbasedev_iter->type != VFIO_DEVICE_TYPE_PCI) { + continue; + } + tmp = container_of(vbasedev_iter, VFIOPCIDevice, vbasedev); + vfio_pci_post_reset(tmp); + } +out_single: + if (!single) { + vfio_pci_post_reset(vdev); + } + g_free(info); + + return ret; +} + const VFIOIOMMUOps vfio_iommufd_ops = { .dma_map = iommufd_map, .dma_unmap = iommufd_unmap, .attach_device = iommufd_attach_device, .detach_device = iommufd_detach_device, + .pci_hot_reset = iommufd_pci_hot_reset, }; diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index a6194b7bfe..eb662fd086 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -2374,7 +2374,7 @@ static int vfio_add_capabilities(VFIOPCIDevice *vdev, Error **errp) return 0; } -static void vfio_pci_pre_reset(VFIOPCIDevice *vdev) +void vfio_pci_pre_reset(VFIOPCIDevice *vdev) { PCIDevice *pdev = &vdev->pdev; uint16_t cmd; @@ -2411,7 +2411,7 @@ static void vfio_pci_pre_reset(VFIOPCIDevice *vdev) vfio_pci_write_config(pdev, PCI_COMMAND, cmd, 2); } -static void vfio_pci_post_reset(VFIOPCIDevice *vdev) +void vfio_pci_post_reset(VFIOPCIDevice *vdev) { Error *err = NULL; int nr; diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index d85342b65f..e88a7d5ccc 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -34,6 +34,7 @@ vfio_check_af_flr(const char *name) "%s Supports FLR via AF cap" vfio_pci_hot_reset(const char *name, const char *type) " (%s) %s" vfio_pci_hot_reset_has_dep_devices(const char *name) "%s: hot reset dependent devices:" vfio_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int group_id) "\t%04x:%02x:%02x.%x group %d" +vfio_pci_hot_reset_dep_devices_iommufd(int domain, int bus, int slot, int function, int dev_id) "\t%04x:%02x:%02x.%x devid %d" vfio_pci_hot_reset_result(const char *name, const char *result) "%s hot reset: %s" vfio_populate_device_config(const char *name, unsigned long size, unsigned long offset, unsigned long flags) "Device %s config:\n size: 0x%lx, offset: 0x%lx, flags: 0x%lx" vfio_populate_device_get_irq_info_failure(const char *errstr) "VFIO_DEVICE_GET_IRQ_INFO failure: %s" From patchwork Thu Nov 2 07:12:55 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858353 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=KFr0IIES; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKr0mKKz1yQq for ; Thu, 2 Nov 2023 18:36:48 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAb-0001GJ-EU; Thu, 02 Nov 2023 03:31:13 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAY-0000px-Eo for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:10 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAW-0003Ox-MA for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:10 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910268; x=1730446268; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=N177sKSc3z0WYF39Q+c1NtUnCGcyPBi7xR3+aLU6nSY=; b=KFr0IIESs9dk+6auAm/0yoJjqnjNuH9urrNqrXDDt7WgT1TLKQx0FVB9 kJj2KB1jLf8LpqtiljxpvQFDQtG1cB9WLBZUsXaSSLPrIHSWZXnCSxY05 4Cv8ACXodpG95jvv/PzvbuuhX+v/cS0JZIEfhSb7xJW7DCf+JFH1pm6sh pmKzslH4RIVgX6Vh6o7v3ckA0SaftsC5l1HK1YGMe4U7s/W9zyAub/+jZ kv7wESdgCpCD0tTMbbiipw2bL4qFdrvX7SZBX3XzekJMN0WzMqjhdTrxY 34I9sV1HIeiWx7AbxILUuF2KD4jstiyh/+JdtF7mILzw6+2n4uhsb6Sjv Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953407" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953407" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:53 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055740" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055740" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:49 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 34/41] vfio/pci: Allow the selection of a given iommu backend Date: Thu, 2 Nov 2023 15:12:55 +0800 Message-Id: <20231102071302.1818071-35-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org From: Eric Auger Now we support two types of iommu backends, let's add the capability to select one of them. This depends on whether an iommufd object has been linked with the vfio-pci device: if the user wants to use the legacy backend, it shall not link the vfio-pci device with any iommufd object: -device vfio-pci,host=0000:02:00.0 This is called the legacy mode/backend. If the user wants to use the iommufd backend (/dev/iommu) it shall pass an iommufd object id in the vfio-pci device options: -object iommufd,id=iommufd0 -device vfio-pci,host=0000:02:00.0,iommufd=iommufd0 Suggested-by: Alex Williamson Signed-off-by: Eric Auger Signed-off-by: Yi Liu Signed-off-by: Zhenzhong Duan --- hw/vfio/pci.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index eb662fd086..7a6696ca55 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -42,6 +42,7 @@ #include "qapi/error.h" #include "migration/blocker.h" #include "migration/qemu-file.h" +#include "sysemu/iommufd.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" @@ -3551,6 +3552,10 @@ static Property vfio_pci_dev_properties[] = { * DEFINE_PROP_STRING("vfiofd", VFIOPCIDevice, vfiofd_name), * DEFINE_PROP_STRING("vfiogroupfd, VFIOPCIDevice, vfiogroupfd_name), */ +#ifdef CONFIG_IOMMUFD + DEFINE_PROP_LINK("iommufd", VFIOPCIDevice, vbasedev.iommufd, + TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), +#endif DEFINE_PROP_END_OF_LIST(), }; From patchwork Thu Nov 2 07:12:56 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858337 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=HzLkB8LY; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbDv4Yngz1yQ4 for ; Thu, 2 Nov 2023 18:32:31 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAj-0002Md-Co; Thu, 02 Nov 2023 03:31:21 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAh-00022S-5N for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:19 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAe-0003PH-Rm for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:18 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910276; x=1730446276; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=hnwDTX235fTol+1nkm3A5OU5kCT/KmaYpp2TcnmHZzU=; b=HzLkB8LYD0mQU2ppJ7hgL/PARbFDpP/aD2R6jO2RJLmNelYG+af+SPnq hDqrbon6W4YR4f25U2esd0YR0gBr4ei93tQpAiu6XRfr+XyxZBYQPqsVQ wDXa+0ym0dR+736ArLGkCZQWesv9dpys6JuIjExqptNj0OXpcFXoxAhaK i1yD0tHO1WxtZ0NEUMPZh3INzabJ37ARDpGTS3zM9SdvRTveN3mxLk9Nb pdp07Mag6MWGN1nmmVlFZYsw+3dSHuKUsQFLct11/qqMVPtFUkzfNICO3 OFW3Hx6A/Wy63aIgHqtmbwcD7vqxhaTS1GAaPeoAbNZoJDStXQweiUAjD g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953430" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953430" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:57 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055756" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055756" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:53 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 35/41] vfio/pci: Make vfio cdev pre-openable by passing a file handle Date: Thu, 2 Nov 2023 15:12:56 +0800 Message-Id: <20231102071302.1818071-36-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This gives management tools like libvirt a chance to open the vfio cdev with privilege and pass FD to qemu. This way qemu never needs to have privilege to open a VFIO or iommu cdev node. Together with the earlier support of pre-opening /dev/iommu device, now we have full support of passing a vfio device to unprivileged qemu by management tool. This mode is no more considered for the legacy backend. So let's remove the "TODO" comment. Add a helper function vfio_device_get_name() to check fd and get device name, it will also be used by other vfio devices. There is no easy way to check if a device is mdev with FD passing, so fail the x-balloon-allowed check unconditionally in this case. There is also no easy way to get BDF as name with FD passing, so we fake a name by VFIO_FD[fd]. Signed-off-by: Zhenzhong Duan --- include/hw/vfio/vfio-common.h | 1 + hw/vfio/helpers.c | 33 +++++++++++++++++++++++++++++ hw/vfio/iommufd.c | 12 +++++++---- hw/vfio/pci.c | 40 ++++++++++++++++++++++++----------- 4 files changed, 70 insertions(+), 16 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 3f1a39a991..854c32e4ce 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -250,6 +250,7 @@ struct vfio_info_cap_header * vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); +int vfio_device_get_name(VFIODevice *vbasedev, Error **errp); #endif bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index 168847e7c5..044dbbc501 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -609,3 +609,36 @@ bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) return ret; } + +int vfio_device_get_name(VFIODevice *vbasedev, Error **errp) +{ + struct stat st; + + if (vbasedev->fd < 0) { + if (stat(vbasedev->sysfsdev, &st) < 0) { + error_setg_errno(errp, errno, "no such host device"); + error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); + return -errno; + } + /* User may specify a name, e.g: VFIO platform device */ + if (!vbasedev->name) { + vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); + } + } +#ifdef CONFIG_IOMMUFD + else { + if (!vbasedev->iommufd) { + error_setg(errp, "Use FD passing only with iommufd backend"); + return -EINVAL; + } + /* + * Give a name with fd so any function printing out vbasedev->name + * will not break. + */ + if (!vbasedev->name) { + vbasedev->name = g_strdup_printf("VFIO_FD%d", vbasedev->fd); + } + } +#endif + return 0; +} diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index aedfe31c3c..1fb1c7e853 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -399,11 +399,15 @@ static int iommufd_attach_device(const char *name, VFIODevice *vbasedev, uint32_t ioas_id; Error *err = NULL; - devfd = iommufd_cdev_getfd(vbasedev->sysfsdev, errp); - if (devfd < 0) { - return devfd; + if (vbasedev->fd < 0) { + devfd = iommufd_cdev_getfd(vbasedev->sysfsdev, errp); + if (devfd < 0) { + return devfd; + } + vbasedev->fd = devfd; + } else { + devfd = vbasedev->fd; } - vbasedev->fd = devfd; ret = iommufd_connect_and_bind(vbasedev, errp); if (ret) { diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index 7a6696ca55..d8f658ea47 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -43,6 +43,7 @@ #include "migration/blocker.h" #include "migration/qemu-file.h" #include "sysemu/iommufd.h" +#include "monitor/monitor.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" @@ -3108,18 +3109,23 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) VFIODevice *vbasedev = &vdev->vbasedev; char *tmp, *subsys; Error *err = NULL; - struct stat st; int i, ret; bool is_mdev; char uuid[UUID_STR_LEN]; char *name; - if (!vbasedev->sysfsdev) { + if (vbasedev->fd < 0 && !vbasedev->sysfsdev) { if (!(~vdev->host.domain || ~vdev->host.bus || ~vdev->host.slot || ~vdev->host.function)) { error_setg(errp, "No provided host device"); +#ifdef CONFIG_IOMMUFD + error_append_hint(errp, "Use -device vfio-pci,host=DDDD:BB:DD.F, " + "-device vfio-pci,sysfsdev=PATH_TO_DEVICE " + "or -device vfio-pci,fd=DEVICE_FD\n"); +#else error_append_hint(errp, "Use -device vfio-pci,host=DDDD:BB:DD.F " "or -device vfio-pci,sysfsdev=PATH_TO_DEVICE\n"); +#endif return; } vbasedev->sysfsdev = @@ -3128,13 +3134,9 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) vdev->host.slot, vdev->host.function); } - if (stat(vbasedev->sysfsdev, &st) < 0) { - error_setg_errno(errp, errno, "no such host device"); - error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); + if (vfio_device_get_name(vbasedev, errp)) { return; } - - vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); vbasedev->ops = &vfio_pci_ops; vbasedev->type = VFIO_DEVICE_TYPE_PCI; vbasedev->dev = DEVICE(vdev); @@ -3494,6 +3496,7 @@ static void vfio_instance_init(Object *obj) vdev->host.bus = ~0U; vdev->host.slot = ~0U; vdev->host.function = ~0U; + vdev->vbasedev.fd = -1; vdev->nv_gpudirect_clique = 0xFF; @@ -3547,11 +3550,6 @@ static Property vfio_pci_dev_properties[] = { qdev_prop_nv_gpudirect_clique, uint8_t), DEFINE_PROP_OFF_AUTO_PCIBAR("x-msix-relocation", VFIOPCIDevice, msix_relo, OFF_AUTOPCIBAR_OFF), - /* - * TODO - support passed fds... is this necessary? - * DEFINE_PROP_STRING("vfiofd", VFIOPCIDevice, vfiofd_name), - * DEFINE_PROP_STRING("vfiogroupfd, VFIOPCIDevice, vfiogroupfd_name), - */ #ifdef CONFIG_IOMMUFD DEFINE_PROP_LINK("iommufd", VFIOPCIDevice, vbasedev.iommufd, TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), @@ -3559,6 +3557,21 @@ static Property vfio_pci_dev_properties[] = { DEFINE_PROP_END_OF_LIST(), }; +#ifdef CONFIG_IOMMUFD +static void vfio_pci_set_fd(Object *obj, const char *str, Error **errp) +{ + VFIOPCIDevice *vdev = VFIO_PCI(obj); + int fd = -1; + + fd = monitor_fd_param(monitor_cur(), str, errp); + if (fd == -1) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vdev->vbasedev.fd = fd; +} +#endif + static void vfio_pci_dev_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); @@ -3566,6 +3579,9 @@ static void vfio_pci_dev_class_init(ObjectClass *klass, void *data) dc->reset = vfio_pci_reset; device_class_set_props(dc, vfio_pci_dev_properties); +#ifdef CONFIG_IOMMUFD + object_class_property_add_str(klass, "fd", NULL, vfio_pci_set_fd); +#endif dc->desc = "VFIO-based PCI device assignment"; set_bit(DEVICE_CATEGORY_MISC, dc->categories); pdc->realize = vfio_realize; From patchwork Thu Nov 2 07:12:57 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858354 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=h9kFJjx4; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKs25CTz1yQq for ; Thu, 2 Nov 2023 18:36:49 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAv-0003R8-P4; Thu, 02 Nov 2023 03:31:33 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAt-0003Q1-Jh; Thu, 02 Nov 2023 03:31:31 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAr-0003OZ-R4; Thu, 02 Nov 2023 03:31:31 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910289; x=1730446289; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=mzAi6vxpNOGWPYVzQLzPA/i470ZxYWBSSEc9XR7tpwI=; b=h9kFJjx4TVLtuNy874xT7ULzWRaZTDQT5YToFZiY4XFJJn7RT8CSgaEv 0QYI1xmCtOT7qYPx0R8DYHgw39ddmc+uhESq/2BiglaGchz6culy0f3eD cK+8i5GeYHlx6a2k3uZAKQCbieKgbMRtzT74BlbvP56u9RICAeeE06kYm hNJ5UE/LdmxnWrl4UEDujIPY2YQmSXCNzUCH7AYuIr4leR2DjFNF9CzHy yY/adFA/AWjXlaFo6XKBnKmJ5Q9qlteDgkvwJaMQn8UfoZzCIjfTeMYfg Wr4uN2oHJnyr51FJj/BHvgOfE4t0qAbBxr4Ul39vZm0sajhCMfqAdIdnP g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953439" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953439" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:02 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055769" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055769" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:30:57 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Thomas Huth , Tony Krowiak , Halil Pasic , Jason Herne , Eric Farman , Matthew Rosato , qemu-s390x@nongnu.org (open list:S390 general arch...) Subject: [PATCH v4 36/41] vfio: Allow the selection of a given iommu backend for platform ap and ccw Date: Thu, 2 Nov 2023 15:12:57 +0800 Message-Id: <20231102071302.1818071-37-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Previously we added support to select iommu backend for vfio pci device. Now we added others, E.g: platform, ap and ccw. Signed-off-by: Zhenzhong Duan --- include/hw/vfio/vfio-platform.h | 1 + hw/vfio/ap.c | 5 +++++ hw/vfio/ccw.c | 5 +++++ hw/vfio/platform.c | 4 ++++ 4 files changed, 15 insertions(+) diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h index c414c3dffc..f57f4276f2 100644 --- a/include/hw/vfio/vfio-platform.h +++ b/include/hw/vfio/vfio-platform.h @@ -18,6 +18,7 @@ #include "hw/sysbus.h" #include "hw/vfio/vfio-common.h" +#include "sysemu/iommufd.h" #include "qemu/event_notifier.h" #include "qemu/queue.h" #include "qom/object.h" diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index bbf69ff55a..6a4186ccd3 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -15,6 +15,7 @@ #include #include "qapi/error.h" #include "hw/vfio/vfio-common.h" +#include "sysemu/iommufd.h" #include "hw/s390x/ap-device.h" #include "qemu/error-report.h" #include "qemu/event_notifier.h" @@ -204,6 +205,10 @@ static void vfio_ap_unrealize(DeviceState *dev) static Property vfio_ap_properties[] = { DEFINE_PROP_STRING("sysfsdev", VFIOAPDevice, vdev.sysfsdev), +#ifdef CONFIG_IOMMUFD + DEFINE_PROP_LINK("iommufd", VFIOAPDevice, vdev.iommufd, + TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), +#endif DEFINE_PROP_END_OF_LIST(), }; diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index d857bb8d0f..7695ede0fc 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -21,6 +21,7 @@ #include "qapi/error.h" #include "hw/vfio/vfio-common.h" +#include "sysemu/iommufd.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/vfio-ccw.h" #include "hw/qdev-properties.h" @@ -677,6 +678,10 @@ static void vfio_ccw_unrealize(DeviceState *dev) static Property vfio_ccw_properties[] = { DEFINE_PROP_STRING("sysfsdev", VFIOCCWDevice, vdev.sysfsdev), DEFINE_PROP_BOOL("force-orb-pfch", VFIOCCWDevice, force_orb_pfch, false), +#ifdef CONFIG_IOMMUFD + DEFINE_PROP_LINK("iommufd", VFIOCCWDevice, vdev.iommufd, + TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), +#endif DEFINE_PROP_END_OF_LIST(), }; diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index 8e3d4ac458..a1c25e0337 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -649,6 +649,10 @@ static Property vfio_platform_dev_properties[] = { DEFINE_PROP_UINT32("mmap-timeout-ms", VFIOPlatformDevice, mmap_timeout, 1100), DEFINE_PROP_BOOL("x-irqfd", VFIOPlatformDevice, irqfd_allowed, true), +#ifdef CONFIG_IOMMUFD + DEFINE_PROP_LINK("iommufd", VFIOPlatformDevice, vbasedev.iommufd, + TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), +#endif DEFINE_PROP_END_OF_LIST(), }; From patchwork Thu Nov 2 07:12:58 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858334 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=dEu59uyO; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbD64Q60z1yQ4 for ; Thu, 2 Nov 2023 18:31:50 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySAn-00034q-NZ; Thu, 02 Nov 2023 03:31:26 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAl-0002iB-9m for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:23 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySAj-0003Ox-5h for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:23 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910281; x=1730446281; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=LmRhWqNN5MpdYLIWPBzUIr7um0fmtv6LZuOxukM3jx8=; b=dEu59uyOkLjh7FV5o9HnWgvC0VqVDd9ummADkYGFoNxiYjny1wX9rwxO VB/pZ37x1snx1cxzx5dvx9dZvt5guXOg1xO1Ff9HbyvQJrkoRBNjTDMYm B4CV8hDq4CeZzS+16lFaeApCACn6HohFXJvJmYnI2Xcq8FAYVgyhXpLvk lBxoXr6jUJjjdR+w94WeFweApYYF2wx1OyOorCiP9DjabO5JNsGw0hvAV lQKqsDohK8JcpB6qwKhuC0/oP4AFgTopgU7keKlBkHuBGl0GTX7vZm51z 1CMVdZXblhVA3w6TIMixKglvyiLTRDAJZtHXX4MBSQQpwfNHIUUXsm1A/ w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953504" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953504" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:06 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055787" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055787" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:02 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 37/41] vfio/platform: Make vfio cdev pre-openable by passing a file handle Date: Thu, 2 Nov 2023 15:12:58 +0800 Message-Id: <20231102071302.1818071-38-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This gives management tools like libvirt a chance to open the vfio cdev with privilege and pass FD to qemu. This way qemu never needs to have privilege to open a VFIO or iommu cdev node. Signed-off-by: Zhenzhong Duan --- hw/vfio/platform.c | 41 +++++++++++++++++++++++++++++++++-------- 1 file changed, 33 insertions(+), 8 deletions(-) diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index a1c25e0337..aa0b2b9583 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -35,6 +35,7 @@ #include "hw/platform-bus.h" #include "hw/qdev-properties.h" #include "sysemu/kvm.h" +#include "monitor/monitor.h" /* * Functions used whatever the injection method @@ -529,14 +530,13 @@ static VFIODeviceOps vfio_platform_ops = { */ static int vfio_base_device_init(VFIODevice *vbasedev, Error **errp) { - struct stat st; int ret; - /* @sysfsdev takes precedence over @host */ - if (vbasedev->sysfsdev) { + /* @fd takes precedence over @sysfsdev which takes precedence over @host */ + if (vbasedev->fd < 0 && vbasedev->sysfsdev) { g_free(vbasedev->name); vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); - } else { + } else if (vbasedev->fd < 0) { if (!vbasedev->name || strchr(vbasedev->name, '/')) { error_setg(errp, "wrong host device name"); return -EINVAL; @@ -546,10 +546,9 @@ static int vfio_base_device_init(VFIODevice *vbasedev, Error **errp) vbasedev->name); } - if (stat(vbasedev->sysfsdev, &st) < 0) { - error_setg_errno(errp, errno, - "failed to get the sysfs host device file status"); - return -errno; + ret = vfio_device_get_name(vbasedev, errp); + if (ret) { + return ret; } ret = vfio_attach_device(vbasedev->name, vbasedev, @@ -656,6 +655,28 @@ static Property vfio_platform_dev_properties[] = { DEFINE_PROP_END_OF_LIST(), }; +static void vfio_platform_instance_init(Object *obj) +{ + VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(obj); + + vdev->vbasedev.fd = -1; +} + +#ifdef CONFIG_IOMMUFD +static void vfio_platform_set_fd(Object *obj, const char *str, Error **errp) +{ + VFIOPlatformDevice *vdev = VFIO_PLATFORM_DEVICE(obj); + int fd = -1; + + fd = monitor_fd_param(monitor_cur(), str, errp); + if (fd == -1) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vdev->vbasedev.fd = fd; +} +#endif + static void vfio_platform_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); @@ -663,6 +684,9 @@ static void vfio_platform_class_init(ObjectClass *klass, void *data) dc->realize = vfio_platform_realize; device_class_set_props(dc, vfio_platform_dev_properties); +#ifdef CONFIG_IOMMUFD + object_class_property_add_str(klass, "fd", NULL, vfio_platform_set_fd); +#endif dc->vmsd = &vfio_platform_vmstate; dc->desc = "VFIO-based platform device assignment"; sbc->connect_irq_notifier = vfio_start_irqfd_injection; @@ -675,6 +699,7 @@ static const TypeInfo vfio_platform_dev_info = { .name = TYPE_VFIO_PLATFORM, .parent = TYPE_SYS_BUS_DEVICE, .instance_size = sizeof(VFIOPlatformDevice), + .instance_init = vfio_platform_instance_init, .class_init = vfio_platform_class_init, .class_size = sizeof(VFIOPlatformDeviceClass), }; From patchwork Thu Nov 2 07:12:59 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858351 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=ZJEqaUCC; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKj0ydHz1yQq for ; Thu, 2 Nov 2023 18:36:41 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySBE-0004mz-6h; Thu, 02 Nov 2023 03:31:52 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB5-0004Xg-VC; Thu, 02 Nov 2023 03:31:43 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB2-0003PH-88; Thu, 02 Nov 2023 03:31:43 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910300; x=1730446300; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=hg846DZ/r6i0oF59a4jSiOSK6VqIvw8JEveZ3UP+CpQ=; b=ZJEqaUCCZBEzIaLonJi0s60qWLkWGCFtYdx5Tk4vcPHvcigfGznUANwn 8cCDZEKrIPsbtvK71dnv+zfVIzv/rU3zjnUpBR3wqdaG6MlmmhJShSroV hdIH21E9Qr7dD/bVFXBmMX/9N8RucLloD/FTUQeGmO4yf0BpOn7CMpYl9 cAgjBIoLFhDw6w687Pit9DXsWEawNR1nGFmn/wAgk8pXycCaV7U7dR+8g hrSejRRLwlUEkOEvZxZ8/chVt3G6XwxWq4DBE/sWkAvp+yCrvfCqcDuXw LzGtQXDNittEvXMcPM6TM3ATiW5SStAC3RxKN/ZrvtIxnv0Tg+2v3a/D5 Q==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953537" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953537" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:11 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055802" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055802" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:06 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Thomas Huth , Tony Krowiak , Halil Pasic , Jason Herne , qemu-s390x@nongnu.org (open list:S390 general arch...) Subject: [PATCH v4 38/41] vfio/ap: Make vfio cdev pre-openable by passing a file handle Date: Thu, 2 Nov 2023 15:12:59 +0800 Message-Id: <20231102071302.1818071-39-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This gives management tools like libvirt a chance to open the vfio cdev with privilege and pass FD to qemu. This way qemu never needs to have privilege to open a VFIO or iommu cdev node. Opportunisticly, remove some unnecessory double-cast. Signed-off-by: Zhenzhong Duan --- hw/vfio/ap.c | 32 +++++++++++++++++++++++++++++++- 1 file changed, 31 insertions(+), 1 deletion(-) diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index 6a4186ccd3..0a810f8b88 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -29,6 +29,7 @@ #include "hw/s390x/ap-bridge.h" #include "exec/address-spaces.h" #include "qom/object.h" +#include "monitor/monitor.h" #define TYPE_VFIO_AP_DEVICE "vfio-ap" @@ -159,7 +160,10 @@ static void vfio_ap_realize(DeviceState *dev, Error **errp) VFIOAPDevice *vapdev = VFIO_AP_DEVICE(dev); VFIODevice *vbasedev = &vapdev->vdev; - vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); + if (vfio_device_get_name(vbasedev, errp)) { + return; + } + vbasedev->ops = &vfio_ap_ops; vbasedev->type = VFIO_DEVICE_TYPE_AP; vbasedev->dev = dev; @@ -229,11 +233,36 @@ static const VMStateDescription vfio_ap_vmstate = { .unmigratable = 1, }; +static void vfio_ap_instance_init(Object *obj) +{ + VFIOAPDevice *vapdev = VFIO_AP_DEVICE(obj); + + vapdev->vdev.fd = -1; +} + +#ifdef CONFIG_IOMMUFD +static void vfio_ap_set_fd(Object *obj, const char *str, Error **errp) +{ + VFIOAPDevice *vapdev = VFIO_AP_DEVICE(obj); + int fd = -1; + + fd = monitor_fd_param(monitor_cur(), str, errp); + if (fd == -1) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vapdev->vdev.fd = fd; +} +#endif + static void vfio_ap_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); device_class_set_props(dc, vfio_ap_properties); +#ifdef CONFIG_IOMMUFD + object_class_property_add_str(klass, "fd", NULL, vfio_ap_set_fd); +#endif dc->vmsd = &vfio_ap_vmstate; dc->desc = "VFIO-based AP device assignment"; set_bit(DEVICE_CATEGORY_MISC, dc->categories); @@ -248,6 +277,7 @@ static const TypeInfo vfio_ap_info = { .name = TYPE_VFIO_AP_DEVICE, .parent = TYPE_AP_DEVICE, .instance_size = sizeof(VFIOAPDevice), + .instance_init = vfio_ap_instance_init, .class_init = vfio_ap_class_init, }; From patchwork Thu Nov 2 07:13:00 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858356 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=bTOhYTNY; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbKt5pGmz1yQq for ; Thu, 2 Nov 2023 18:36:50 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySBM-0005II-Hk; Thu, 02 Nov 2023 03:32:00 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB9-0004n6-I3; Thu, 02 Nov 2023 03:31:50 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB6-0003Ox-AX; Thu, 02 Nov 2023 03:31:46 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910304; x=1730446304; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=jfdiZ9n2PNbWWNnBELCHU65fEvg1lmnCoImg81PrcFc=; b=bTOhYTNYNEf18oS2OCkGM9BDzPmRvK+UjfEEUXlaCV5mJkEkz1M5wq36 RHXUPqDBvA3nSfDNSHC9q1AdhETdp2f8hTKCNCNRRPUXf1NhtmVWO2rnp p5uoxQcwRU1s9BF9/v+6Fg3ZaZNCd70znRRlKDKmKeN2bHOizJqs5GGT4 gG5/D42YsNqp9Elzu260IhZug8z8hG8XgcinxV/OU4qhK7oAYAT9aSE/I kBwAi1w24U812laiy66eSk/1pI3V+MEUA9Ne9CfPh4j27dF5kvuggiDfU BYSwwFSW3Cn+b9WIK3CdbDIcus462mppTpDZ4p8nVOfF0sLNKTFaO6MIZ g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953550" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953550" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:15 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055820" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055820" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:11 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan , Eric Farman , Matthew Rosato , Thomas Huth , qemu-s390x@nongnu.org (open list:vfio-ccw) Subject: [PATCH v4 39/41] vfio/ccw: Make vfio cdev pre-openable by passing a file handle Date: Thu, 2 Nov 2023 15:13:00 +0800 Message-Id: <20231102071302.1818071-40-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org This gives management tools like libvirt a chance to open the vfio cdev with privilege and pass FD to qemu. This way qemu never needs to have privilege to open a VFIO or iommu cdev node. Opportunisticly, remove a redundant definition of TYPE_VFIO_CCW. Signed-off-by: Zhenzhong Duan --- hw/vfio/ccw.c | 34 +++++++++++++++++++++++++++++++--- 1 file changed, 31 insertions(+), 3 deletions(-) diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index 7695ede0fc..a674bd8d6d 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -30,6 +30,7 @@ #include "qemu/error-report.h" #include "qemu/main-loop.h" #include "qemu/module.h" +#include "monitor/monitor.h" struct VFIOCCWDevice { S390CCWDevice cdev; @@ -589,11 +590,12 @@ static void vfio_ccw_realize(DeviceState *dev, Error **errp) } } + if (vfio_device_get_name(vbasedev, errp)) { + return; + } + vbasedev->ops = &vfio_ccw_ops; vbasedev->type = VFIO_DEVICE_TYPE_CCW; - vbasedev->name = g_strdup_printf("%x.%x.%04x", vcdev->cdev.hostid.cssid, - vcdev->cdev.hostid.ssid, - vcdev->cdev.hostid.devid); vbasedev->dev = dev; /* @@ -690,12 +692,37 @@ static const VMStateDescription vfio_ccw_vmstate = { .unmigratable = 1, }; +static void vfio_ccw_instance_init(Object *obj) +{ + VFIOCCWDevice *vcdev = VFIO_CCW(obj); + + vcdev->vdev.fd = -1; +} + +#ifdef CONFIG_IOMMUFD +static void vfio_ccw_set_fd(Object *obj, const char *str, Error **errp) +{ + VFIOCCWDevice *vcdev = VFIO_CCW(obj); + int fd = -1; + + fd = monitor_fd_param(monitor_cur(), str, errp); + if (fd == -1) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vcdev->vdev.fd = fd; +} +#endif + static void vfio_ccw_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); S390CCWDeviceClass *cdc = S390_CCW_DEVICE_CLASS(klass); device_class_set_props(dc, vfio_ccw_properties); +#ifdef CONFIG_IOMMUFD + object_class_property_add_str(klass, "fd", NULL, vfio_ccw_set_fd); +#endif dc->vmsd = &vfio_ccw_vmstate; dc->desc = "VFIO-based subchannel assignment"; set_bit(DEVICE_CATEGORY_MISC, dc->categories); @@ -713,6 +740,7 @@ static const TypeInfo vfio_ccw_info = { .name = TYPE_VFIO_CCW, .parent = TYPE_S390_CCW, .instance_size = sizeof(VFIOCCWDevice), + .instance_init = vfio_ccw_instance_init, .class_init = vfio_ccw_class_init, }; From patchwork Thu Nov 2 07:13:01 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858363 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=ICVsx2Aj; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbLf4QH0z1yQq for ; Thu, 2 Nov 2023 18:37:30 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySBM-0005N7-QS; Thu, 02 Nov 2023 03:32:00 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB9-0004n5-Hw for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:50 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySB4-0003OZ-1g for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:45 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910302; x=1730446302; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=V79B63jbxqsm8f4qWarh0zT/m5obcDJzksGaxcQ4Q3s=; b=ICVsx2AjqY5kBHfxyHmdTh19n0G9lm1NN+CrkcT5x2mU85Aw02jHQNj9 Fowj1pzwgi3MJyIVV/+rv22mY9bXghdOajaYAh/7jvER2ZoyxHuMYmZNm QeEOnLglcvb2mpbmPRvVQWRu7UiDNIT+zYmlSJsekHYW3QhbIBWw2E6tW fpyQRwjttTnTrTO2UTysk6wHS8izxfkfSPW65XcTm5/S9oz59Sa3Cx4EF acFvjt+bBHVdv/4pu46m+XXwvyjFqwIDg32kTf0os+tXvesw+BeJwwvge uxhQKtMQhcDcG02SDvrYAi127bwy4ialE7Ea0tdBPSktkB1oz9S5YzYNu g==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953560" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953560" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:19 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055830" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055830" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:15 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 40/41] vfio: Make VFIOContainerBase poiner parameter const in VFIOIOMMUOps callbacks Date: Thu, 2 Nov 2023 15:13:01 +0800 Message-Id: <20231102071302.1818071-41-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Some of the callbacks in VFIOIOMMUOps pass VFIOContainerBase poiner, those callbacks only need read access to the sub object of VFIOContainerBase. So make VFIOContainerBase, VFIOContainer and VFIOIOMMUFDContainer as const in these callbacks. Local functions called by those callbacks also need same changes to avoid build error. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan --- include/hw/vfio/vfio-common.h | 12 ++++++---- include/hw/vfio/vfio-container-base.h | 12 ++++++---- hw/vfio/common.c | 9 +++---- hw/vfio/container-base.c | 2 +- hw/vfio/container.c | 34 ++++++++++++++------------- hw/vfio/iommufd.c | 8 +++---- 6 files changed, 42 insertions(+), 35 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index 854c32e4ce..2fc976c7f0 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -257,11 +257,13 @@ bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); void vfio_migration_exit(VFIODevice *vbasedev); int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); -bool vfio_devices_all_running_and_mig_active(VFIOContainerBase *bcontainer); -bool vfio_devices_all_device_dirty_tracking(VFIOContainerBase *bcontainer); -int vfio_devices_query_dirty_bitmap(VFIOContainerBase *bcontainer, +bool +vfio_devices_all_running_and_mig_active(const VFIOContainerBase *bcontainer); +bool +vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); +int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); -int vfio_get_dirty_bitmap(VFIOContainerBase *bcontainer, uint64_t iova, - uint64_t size, ram_addr_t ram_addr); +int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, + uint64_t size, ram_addr_t ram_addr); #endif /* HW_VFIO_VFIO_COMMON_H */ diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index 45bb19c767..2ae297ccda 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -82,7 +82,7 @@ void vfio_container_del_section_window(VFIOContainerBase *bcontainer, MemoryRegionSection *section); int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start); -int vfio_container_query_dirty_bitmap(VFIOContainerBase *bcontainer, +int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size); @@ -93,18 +93,20 @@ void vfio_container_destroy(VFIOContainerBase *bcontainer); struct VFIOIOMMUOps { /* basic feature */ - int (*dma_map)(VFIOContainerBase *bcontainer, + int (*dma_map)(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly); - int (*dma_unmap)(VFIOContainerBase *bcontainer, + int (*dma_unmap)(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb); int (*attach_device)(const char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp); void (*detach_device)(VFIODevice *vbasedev); /* migration feature */ - int (*set_dirty_page_tracking)(VFIOContainerBase *bcontainer, bool start); - int (*query_dirty_bitmap)(VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, + int (*set_dirty_page_tracking)(const VFIOContainerBase *bcontainer, + bool start); + int (*query_dirty_bitmap)(const VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, hwaddr iova, hwaddr size); /* PCI specific */ int (*pci_hot_reset)(VFIODevice *vbasedev, bool single); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index a61dce2845..1c9203183d 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -203,7 +203,7 @@ static bool vfio_devices_all_dirty_tracking(VFIOContainerBase *bcontainer) return true; } -bool vfio_devices_all_device_dirty_tracking(VFIOContainerBase *bcontainer) +bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -220,7 +220,8 @@ bool vfio_devices_all_device_dirty_tracking(VFIOContainerBase *bcontainer) * Check if all VFIO devices are running and migration is active, which is * essentially equivalent to the migration being in pre-copy phase. */ -bool vfio_devices_all_running_and_mig_active(VFIOContainerBase *bcontainer) +bool +vfio_devices_all_running_and_mig_active(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -1138,7 +1139,7 @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } -int vfio_devices_query_dirty_bitmap(VFIOContainerBase *bcontainer, +int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size) { @@ -1161,7 +1162,7 @@ int vfio_devices_query_dirty_bitmap(VFIOContainerBase *bcontainer, return 0; } -int vfio_get_dirty_bitmap(VFIOContainerBase *bcontainer, uint64_t iova, +int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr) { bool all_device_dirty_tracking = diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index eee2dcfe76..1ffd25bbfa 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -63,7 +63,7 @@ int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, return bcontainer->ops->set_dirty_page_tracking(bcontainer, start); } -int vfio_container_query_dirty_bitmap(VFIOContainerBase *bcontainer, +int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size) { diff --git a/hw/vfio/container.c b/hw/vfio/container.c index f27cc15d09..31681db52b 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -61,11 +61,11 @@ static int vfio_ram_block_discard_disable(VFIOContainer *container, bool state) } } -static int vfio_dma_unmap_bitmap(VFIOContainer *container, +static int vfio_dma_unmap_bitmap(const VFIOContainer *container, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb) { - VFIOContainerBase *bcontainer = &container->bcontainer; + const VFIOContainerBase *bcontainer = &container->bcontainer; struct vfio_iommu_type1_dma_unmap *unmap; struct vfio_bitmap *bitmap; VFIOBitmap vbmap; @@ -117,11 +117,12 @@ unmap_exit: /* * DMA - Mapping and unmapping for the "type1" IOMMU interface used on x86 */ -static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, - ram_addr_t size, IOMMUTLBEntry *iotlb) +static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, + hwaddr iova, ram_addr_t size, + IOMMUTLBEntry *iotlb) { - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); + const VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dma_unmap unmap = { .argsz = sizeof(unmap), .flags = 0, @@ -174,11 +175,11 @@ static int vfio_legacy_dma_unmap(VFIOContainerBase *bcontainer, hwaddr iova, return 0; } -static int vfio_legacy_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, +static int vfio_legacy_dma_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) { - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); + const VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dma_map map = { .argsz = sizeof(map), .flags = VFIO_DMA_MAP_FLAG_READ, @@ -207,11 +208,12 @@ static int vfio_legacy_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, return -errno; } -static int vfio_legacy_set_dirty_page_tracking(VFIOContainerBase *bcontainer, - bool start) +static int +vfio_legacy_set_dirty_page_tracking(const VFIOContainerBase *bcontainer, + bool start) { - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); + const VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); int ret; struct vfio_iommu_type1_dirty_bitmap dirty = { .argsz = sizeof(dirty), @@ -233,12 +235,12 @@ static int vfio_legacy_set_dirty_page_tracking(VFIOContainerBase *bcontainer, return ret; } -static int vfio_legacy_query_dirty_bitmap(VFIOContainerBase *bcontainer, +static int vfio_legacy_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size) { - VFIOContainer *container = container_of(bcontainer, VFIOContainer, - bcontainer); + const VFIOContainer *container = container_of(bcontainer, VFIOContainer, + bcontainer); struct vfio_iommu_type1_dirty_bitmap *dbitmap; struct vfio_iommu_type1_dirty_bitmap_get *range; int ret; diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index 1fb1c7e853..ea33e64933 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -26,10 +26,10 @@ #include "qemu/chardev_open.h" #include "pci.h" -static int iommufd_map(VFIOContainerBase *bcontainer, hwaddr iova, +static int iommufd_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) { - VFIOIOMMUFDContainer *container = + const VFIOIOMMUFDContainer *container = container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); return iommufd_backend_map_dma(container->be, @@ -37,11 +37,11 @@ static int iommufd_map(VFIOContainerBase *bcontainer, hwaddr iova, iova, size, vaddr, readonly); } -static int iommufd_unmap(VFIOContainerBase *bcontainer, +static int iommufd_unmap(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, IOMMUTLBEntry *iotlb) { - VFIOIOMMUFDContainer *container = + const VFIOIOMMUFDContainer *container = container_of(bcontainer, VFIOIOMMUFDContainer, bcontainer); /* TODO: Handle dma_unmap_bitmap with iotlb args (migration) */ From patchwork Thu Nov 2 07:13:02 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: "Duan, Zhenzhong" X-Patchwork-Id: 1858336 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@legolas.ozlabs.org Authentication-Results: legolas.ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=intel.com header.i=@intel.com header.a=rsa-sha256 header.s=Intel header.b=AUhix5dM; dkim-atps=neutral Authentication-Results: legolas.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org (client-ip=209.51.188.17; helo=lists.gnu.org; envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org; receiver=patchwork.ozlabs.org) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by legolas.ozlabs.org (Postfix) with ESMTPS id 4SLbDs318Lz1yQ4 for ; Thu, 2 Nov 2023 18:32:29 +1100 (AEDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1qySBP-0005mT-6G; Thu, 02 Nov 2023 03:32:03 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySBJ-0005EC-Ag for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:58 -0400 Received: from mgamail.intel.com ([134.134.136.31]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1qySBH-0003PH-JZ for qemu-devel@nongnu.org; Thu, 02 Nov 2023 03:31:57 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1698910315; x=1730446315; h=from:to:cc:subject:date:message-id:in-reply-to: references:mime-version:content-transfer-encoding; bh=T+1+8hQGhVkQiItYHEuZThhC59tnU8gRzCkf2uADcnc=; b=AUhix5dMTbvcEV8EJjD8dRkyB2GnT7t2C0gisUrjQPYMDYC6CTP3srYK KsFP9+X3jUPlAezSWsCGm/JoBUvttsqEP9zsg4AgIlUgLGrqlMDLf82vW fIF963KBAa5z41q9VLngllCHql4bRWL9OyJUEZi1yjHa/t10M2UpA4bcJ 014fU9F2QwC/rIKbp5KCWXuLaLAWH9qUYu5V9FWG4ffc9lUyPxyb9JHS9 xnU3s+KcAztNbav5dzHFjpH4CSTjBW17256ofetBA/fKXBg1th7N4Wci/ AcYUfkRi5pjcEAikUXSdyRnoeQNHZ5+ntJ8xB6oBJ+GYFULTn2LHk0tMw w==; X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="452953576" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="452953576" Received: from orsmga003.jf.intel.com ([10.7.209.27]) by orsmga104.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:23 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10881"; a="711055859" X-IronPort-AV: E=Sophos;i="6.03,270,1694761200"; d="scan'208";a="711055859" Received: from duan-server-s2600bt.bj.intel.com ([10.240.192.147]) by orsmga003-auth.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 02 Nov 2023 00:31:19 -0700 From: Zhenzhong Duan To: qemu-devel@nongnu.org Cc: alex.williamson@redhat.com, clg@redhat.com, jgg@nvidia.com, nicolinc@nvidia.com, joao.m.martins@oracle.com, eric.auger@redhat.com, peterx@redhat.com, jasowang@redhat.com, kevin.tian@intel.com, yi.l.liu@intel.com, yi.y.sun@intel.com, chao.p.peng@intel.com, Zhenzhong Duan Subject: [PATCH v4 41/41] vfio: Compile out iommufd for PPC target Date: Thu, 2 Nov 2023 15:13:02 +0800 Message-Id: <20231102071302.1818071-42-zhenzhong.duan@intel.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20231102071302.1818071-1-zhenzhong.duan@intel.com> References: <20231102071302.1818071-1-zhenzhong.duan@intel.com> MIME-Version: 1.0 Received-SPF: pass client-ip=134.134.136.31; envelope-from=zhenzhong.duan@intel.com; helo=mgamail.intel.com X-Spam_score_int: -47 X-Spam_score: -4.8 X-Spam_bar: ---- X-Spam_report: (-4.8 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.393, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, T_SCC_BODY_TEXT_LINE=-0.01 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Sender: qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org Since PPC doesn't support IOMMUFD, make iommufd related code compiled out. Suggested-by: Cédric Le Goater Signed-off-by: Zhenzhong Duan --- hw/vfio/common.c | 2 +- hw/vfio/pci.c | 2 +- hw/vfio/platform.c | 2 +- backends/meson.build | 4 ++-- hw/vfio/meson.build | 2 +- 5 files changed, 6 insertions(+), 6 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index 1c9203183d..000717cef3 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -1504,7 +1504,7 @@ int vfio_attach_device(char *name, VFIODevice *vbasedev, { const VFIOIOMMUOps *ops; -#ifdef CONFIG_IOMMUFD +#if defined(CONFIG_IOMMUFD) && !defined(TARGET_PPC) if (vbasedev->iommufd) { ops = &vfio_iommufd_ops; } else diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index d8f658ea47..2287e45119 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -3550,7 +3550,7 @@ static Property vfio_pci_dev_properties[] = { qdev_prop_nv_gpudirect_clique, uint8_t), DEFINE_PROP_OFF_AUTO_PCIBAR("x-msix-relocation", VFIOPCIDevice, msix_relo, OFF_AUTOPCIBAR_OFF), -#ifdef CONFIG_IOMMUFD +#if defined(CONFIG_IOMMUFD) && !defined(TARGET_PPC) DEFINE_PROP_LINK("iommufd", VFIOPCIDevice, vbasedev.iommufd, TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), #endif diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index aa0b2b9583..c8f4ae5a06 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -648,7 +648,7 @@ static Property vfio_platform_dev_properties[] = { DEFINE_PROP_UINT32("mmap-timeout-ms", VFIOPlatformDevice, mmap_timeout, 1100), DEFINE_PROP_BOOL("x-irqfd", VFIOPlatformDevice, irqfd_allowed, true), -#ifdef CONFIG_IOMMUFD +#if defined(CONFIG_IOMMUFD) && !defined(TARGET_PPC) DEFINE_PROP_LINK("iommufd", VFIOPlatformDevice, vbasedev.iommufd, TYPE_IOMMUFD_BACKEND, IOMMUFDBackend *), #endif diff --git a/backends/meson.build b/backends/meson.build index 05ac57ff15..9dbdfa87f7 100644 --- a/backends/meson.build +++ b/backends/meson.build @@ -21,9 +21,9 @@ if have_vhost_user endif system_ss.add(when: 'CONFIG_VIRTIO_CRYPTO', if_true: files('cryptodev-vhost.c')) if have_iommufd - system_ss.add(files('iommufd.c')) + system_ss.add(when: 'TARGET_PPC', if_false: files('iommufd.c')) else - system_ss.add(files('iommufd-stub.c')) + system_ss.add(when: 'TARGET_PPC', if_false: files('iommufd-stub.c')) endif if have_vhost_user_crypto system_ss.add(when: 'CONFIG_VIRTIO_CRYPTO', if_true: files('cryptodev-vhost-user.c')) diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index 9cae2c9e21..4423bb3cd4 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -8,7 +8,7 @@ vfio_ss.add(files( 'migration.c', )) if have_iommufd - vfio_ss.add(files('iommufd.c')) + vfio_ss.add(when: 'TARGET_PPC', if_false: files('iommufd.c')) endif vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files( 'display.c',