From patchwork Thu Mar 6 03:52:28 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Alistair Popple X-Patchwork-Id: 327274 Return-Path: X-Original-To: incoming-dt@patchwork.ozlabs.org Delivered-To: patchwork-incoming-dt@bilbo.ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 07C2B2C03F1 for ; Thu, 6 Mar 2014 14:55:58 +1100 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755531AbaCFDyt (ORCPT ); Wed, 5 Mar 2014 22:54:49 -0500 Received: from ipmail04.adl6.internode.on.net ([150.101.137.141]:13911 "EHLO ipmail04.adl6.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751573AbaCFDy1 (ORCPT ); Wed, 5 Mar 2014 22:54:27 -0500 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AiAEAKzwF1Ol5H4JgWdsb2JhbABahkxSuh2Dc4ExDgEBFiYogiYBBSMECwFGEAgdAiYCAlcQCYd5rhagaxeBKY0oB4JvgUkEiF6FaoYKg2uVbCg Received: from ibmaus65.lnk.telstra.net (HELO localhost) ([165.228.126.9]) by ipmail04.adl6.internode.on.net with ESMTP; 06 Mar 2014 14:24:25 +1030 From: Alistair Popple To: benh@kernel.crashing.org Cc: Alistair Popple , linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org Subject: [PATCH 5/5] powerpc: Added PCI MSI support using the HSTA module Date: Thu, 6 Mar 2014 14:52:28 +1100 Message-Id: <1394077948-8395-6-git-send-email-alistair@popple.id.au> X-Mailer: git-send-email 1.8.3.2 In-Reply-To: <1394077948-8395-1-git-send-email-alistair@popple.id.au> References: <1394077948-8395-1-git-send-email-alistair@popple.id.au> MIME-Version: 1.0 Sender: devicetree-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: devicetree@vger.kernel.org The PPC476GTR SoC supports message signalled interrupts (MSI) by writing to special addresses within the High Speed Transfer Assist (HSTA) module. This patch adds support for PCI MSI with a new system device. The DMA window is also updated to allow access to the entire 42-bit address range to allow PCI devices write access to the HSTA module. Signed-off-by: Alistair Popple --- .../devicetree/bindings/powerpc/4xx/hsta.txt | 19 ++ arch/powerpc/boot/dts/akebono.dts | 46 ++++- arch/powerpc/boot/treeboot-akebono.c | 15 -- arch/powerpc/platforms/44x/Kconfig | 2 + arch/powerpc/sysdev/Kconfig | 6 + arch/powerpc/sysdev/Makefile | 1 + arch/powerpc/sysdev/ppc4xx_hsta_msi.c | 215 +++++++++++++++++++++ arch/powerpc/sysdev/ppc4xx_pci.c | 8 +- 8 files changed, 287 insertions(+), 25 deletions(-) create mode 100644 Documentation/devicetree/bindings/powerpc/4xx/hsta.txt create mode 100644 arch/powerpc/sysdev/ppc4xx_hsta_msi.c diff --git a/Documentation/devicetree/bindings/powerpc/4xx/hsta.txt b/Documentation/devicetree/bindings/powerpc/4xx/hsta.txt new file mode 100644 index 0000000..c737c83 --- /dev/null +++ b/Documentation/devicetree/bindings/powerpc/4xx/hsta.txt @@ -0,0 +1,19 @@ + +ppc476gtr High Speed Serial Assist (HSTA) node +============================================== + +The 476gtr SoC contains a high speed serial assist module attached +between the plb4 and plb6 system buses to provide high speed data +transfer between memory and system peripherals as well as support for +PCI message signalled interrupts. + +Currently only the MSI support is used by Linux using the following +device tree entries: + +Require properties: +- compatible : "ibm,476gtr-hsta-msi", "ibm,hsta-msi" +- reg : register mapping for the HSTA MSI space +- interrupt-parent : parent controller for mapping interrupts +- interrupts : ordered interrupt mapping for each MSI in the register + space. The first interrupt should be associated with a + register offset of 0x00, the second to 0x10, etc. diff --git a/arch/powerpc/boot/dts/akebono.dts b/arch/powerpc/boot/dts/akebono.dts index 96ac13b..f92ecfe 100644 --- a/arch/powerpc/boot/dts/akebono.dts +++ b/arch/powerpc/boot/dts/akebono.dts @@ -82,6 +82,28 @@ ranges; clock-frequency = <200000000>; // 200Mhz + HSTA0: hsta@310000e0000 { + compatible = "ibm,476gtr-hsta-msi", "ibm,hsta-msi"; + reg = <0x310 0x000e0000 0x0 0xf0>; + interrupt-parent = <&MPIC>; + interrupts = <108 0 + 109 0 + 110 0 + 111 0 + 112 0 + 113 0 + 114 0 + 115 0 + 116 0 + 117 0 + 118 0 + 119 0 + 120 0 + 121 0 + 122 0 + 123 0>; + }; + MAL0: mcmal { compatible = "ibm,mcmal-476gtr", "ibm,mcmal2"; dcr-reg = <0xc0000000 0x062>; @@ -242,8 +264,10 @@ ranges = <0x02000000 0x00000000 0x80000000 0x00000110 0x80000000 0x0 0x80000000 0x01000000 0x0 0x0 0x00000140 0x0 0x0 0x00010000>; - /* Inbound starting at 0 to memsize filled in by zImage */ - dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x0 0x0>; + /* Inbound starting at 0x0 to 0x40000000000. In order to use MSI + * PCI devices must be able to write to the HSTA module. + */ + dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x400 0x0>; /* This drives busses 0 to 0xf */ bus-range = <0x0 0xf>; @@ -280,8 +304,10 @@ ranges = <0x02000000 0x00000000 0x80000000 0x00000210 0x80000000 0x0 0x80000000 0x01000000 0x0 0x0 0x00000240 0x0 0x0 0x00010000>; - /* Inbound starting at 0 to memsize filled in by zImage */ - dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x0 0x0>; + /* Inbound starting at 0x0 to 0x40000000000. In order to use MSI + * PCI devices must be able to write to the HSTA module. + */ + dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x400 0x0>; /* This drives busses 0 to 0xf */ bus-range = <0x0 0xf>; @@ -318,8 +344,10 @@ ranges = <0x02000000 0x00000000 0x80000000 0x00000190 0x80000000 0x0 0x80000000 0x01000000 0x0 0x0 0x000001c0 0x0 0x0 0x00010000>; - /* Inbound starting at 0 to memsize filled in by zImage */ - dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x0 0x0>; + /* Inbound starting at 0x0 to 0x40000000000. In order to use MSI + * PCI devices must be able to write to the HSTA module. + */ + dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x400 0x0>; /* This drives busses 0 to 0xf */ bus-range = <0x0 0xf>; @@ -356,8 +384,10 @@ ranges = <0x02000000 0x00000000 0x80000000 0x00000290 0x80000000 0x0 0x80000000 0x01000000 0x0 0x0 0x000002c0 0x0 0x0 0x00010000>; - /* Inbound starting at 0 to memsize filled in by zImage */ - dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x0 0x0>; + /* Inbound starting at 0x0 to 0x40000000000. In order to use MSI + * PCI devices must be able to write to the HSTA module. + */ + dma-ranges = <0x42000000 0x0 0x0 0x0 0x0 0x400 0x0>; /* This drives busses 0 to 0xf */ bus-range = <0x0 0xf>; diff --git a/arch/powerpc/boot/treeboot-akebono.c b/arch/powerpc/boot/treeboot-akebono.c index 070a20f..b73174c 100644 --- a/arch/powerpc/boot/treeboot-akebono.c +++ b/arch/powerpc/boot/treeboot-akebono.c @@ -75,24 +75,9 @@ static void ibm_akebono_fixups(void) { void *emac; u32 reg; - void *devp = finddevice("/"); - u32 dma_ranges[7]; dt_fixup_memory(0x0ULL, ibm_akebono_memsize); - while ((devp = find_node_by_devtype(devp, "pci"))) { - if (getprop(devp, "dma-ranges", dma_ranges, - sizeof(dma_ranges)) < 0) { - printf("%s: Failed to get dma-ranges\r\n", __func__); - continue; - } - - dma_ranges[5] = ibm_akebono_memsize >> 32; - dma_ranges[6] = ibm_akebono_memsize & 0xffffffffUL; - - setprop(devp, "dma-ranges", dma_ranges, sizeof(dma_ranges)); - } - /* Fixup the SD timeout frequency */ mtdcrx(CCTL0_MCO4, 0x1); diff --git a/arch/powerpc/platforms/44x/Kconfig b/arch/powerpc/platforms/44x/Kconfig index b0202d8..8beec7d 100644 --- a/arch/powerpc/platforms/44x/Kconfig +++ b/arch/powerpc/platforms/44x/Kconfig @@ -206,6 +206,8 @@ config AKEBONO select SWIOTLB select 476FPE select PPC4xx_PCI_EXPRESS + select PCI_MSI + select PPC4xx_HSTA_MSI select I2C select I2C_IBM_IIC select NETDEVICES diff --git a/arch/powerpc/sysdev/Kconfig b/arch/powerpc/sysdev/Kconfig index 7baa70d..a19332a 100644 --- a/arch/powerpc/sysdev/Kconfig +++ b/arch/powerpc/sysdev/Kconfig @@ -7,6 +7,12 @@ config PPC4xx_PCI_EXPRESS depends on PCI && 4xx default n +config PPC4xx_HSTA_MSI + bool + depends on PCI_MSI + depends on PCI && 4xx + default n + config PPC4xx_MSI bool depends on PCI_MSI diff --git a/arch/powerpc/sysdev/Makefile b/arch/powerpc/sysdev/Makefile index afbcc37..f7cb2a1 100644 --- a/arch/powerpc/sysdev/Makefile +++ b/arch/powerpc/sysdev/Makefile @@ -45,6 +45,7 @@ obj-$(CONFIG_OF_RTC) += of_rtc.o ifeq ($(CONFIG_PCI),y) obj-$(CONFIG_4xx) += ppc4xx_pci.o endif +obj-$(CONFIG_PPC4xx_HSTA_MSI) += ppc4xx_hsta_msi.o obj-$(CONFIG_PPC4xx_MSI) += ppc4xx_msi.o obj-$(CONFIG_PPC4xx_CPM) += ppc4xx_cpm.o obj-$(CONFIG_PPC4xx_GPIO) += ppc4xx_gpio.o diff --git a/arch/powerpc/sysdev/ppc4xx_hsta_msi.c b/arch/powerpc/sysdev/ppc4xx_hsta_msi.c new file mode 100644 index 0000000..11c8884 --- /dev/null +++ b/arch/powerpc/sysdev/ppc4xx_hsta_msi.c @@ -0,0 +1,215 @@ +/* + * MSI support for PPC4xx SoCs using High Speed Transfer Assist (HSTA) for + * generation of the interrupt. + * + * Copyright © 2013 Alistair Popple IBM Corporation + * + * 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. + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +struct ppc4xx_hsta_msi { + struct device *dev; + + /* The ioremapped HSTA MSI IO space */ + u32 __iomem *data; + + /* Physical address of HSTA MSI IO space */ + u64 address; + struct msi_bitmap bmp; + + /* An array mapping offsets to hardware IRQs */ + int *irq_map; + + /* Number of hwirqs supported */ + int irq_count; +}; +static struct ppc4xx_hsta_msi ppc4xx_hsta_msi; + +static int hsta_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) +{ + struct msi_msg msg; + struct msi_desc *entry; + int irq, hwirq; + u64 addr; + + list_for_each_entry(entry, &dev->msi_list, list) { + irq = msi_bitmap_alloc_hwirqs(&ppc4xx_hsta_msi.bmp, 1); + if (irq < 0) { + pr_debug("%s: Failed to allocate msi interrupt\n", + __func__); + return irq; + } + + hwirq = ppc4xx_hsta_msi.irq_map[irq]; + if (hwirq == NO_IRQ) { + pr_err("%s: Failed mapping irq %d\n", __func__, irq); + return -EINVAL; + } + + /* + * HSTA generates interrupts on writes to 128-bit aligned + * addresses. + */ + addr = ppc4xx_hsta_msi.address + irq*0x10; + msg.address_hi = upper_32_bits(addr); + msg.address_lo = lower_32_bits(addr); + + /* Data is not used by the HSTA. */ + msg.data = 0; + + pr_debug("%s: Setup irq %d (0x%0llx)\n", __func__, hwirq, + (((u64) msg.address_hi) << 32) | msg.address_lo); + + if (irq_set_msi_desc(hwirq, entry)) { + pr_err( + "%s: Invalid hwirq %d specified in device tree\n", + __func__, hwirq); + msi_bitmap_free_hwirqs(&ppc4xx_hsta_msi.bmp, irq, 1); + return -EINVAL; + } + write_msi_msg(hwirq, &msg); + } + + return 0; +} + +static int hsta_find_hwirq_offset(int hwirq) +{ + int irq; + + /* Find the offset given the hwirq */ + for (irq = 0; irq < ppc4xx_hsta_msi.irq_count; irq++) + if (ppc4xx_hsta_msi.irq_map[irq] == hwirq) + return irq; + + return -EINVAL; +} + +static void hsta_teardown_msi_irqs(struct pci_dev *dev) +{ + struct msi_desc *entry; + int irq; + + list_for_each_entry(entry, &dev->msi_list, list) { + if (entry->irq == NO_IRQ) + continue; + + irq = hsta_find_hwirq_offset(entry->irq); + + /* entry->irq should always be in irq_map */ + BUG_ON(irq < 0); + irq_set_msi_desc(entry->irq, NULL); + msi_bitmap_free_hwirqs(&ppc4xx_hsta_msi.bmp, irq, 1); + pr_debug("%s: Teardown IRQ %u (index %u)\n", __func__, + entry->irq, irq); + } +} + +static int hsta_msi_check_device(struct pci_dev *pdev, int nvec, int type) +{ + /* We don't support MSI-X */ + if (type == PCI_CAP_ID_MSIX) { + pr_debug("%s: MSI-X not supported.\n", __func__); + return -EINVAL; + } + + return 0; +} + +static int hsta_msi_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct resource *mem; + int irq, ret, irq_count; + + mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (IS_ERR(mem)) { + dev_err(dev, "Unable to get mmio space\n"); + return -EINVAL; + } + + irq_count = of_irq_count(dev->of_node); + if (!irq_count) { + dev_err(dev, "Unable to find IRQ range\n"); + return -EINVAL; + } + + ppc4xx_hsta_msi.dev = dev; + ppc4xx_hsta_msi.address = mem->start; + ppc4xx_hsta_msi.data = ioremap(mem->start, resource_size(mem)); + ppc4xx_hsta_msi.irq_count = irq_count; + if (IS_ERR(ppc4xx_hsta_msi.data)) { + dev_err(dev, "Unable to map memory\n"); + return -ENOMEM; + } + + ret = msi_bitmap_alloc(&ppc4xx_hsta_msi.bmp, irq_count, dev->of_node); + if (ret) + goto out; + + ppc4xx_hsta_msi.irq_map = kmalloc(sizeof(int) * irq_count, GFP_KERNEL); + if (IS_ERR(ppc4xx_hsta_msi.irq_map)) { + ret = -ENOMEM; + goto out1; + } + + /* Setup a mapping from irq offsets to hardware irq numbers */ + for (irq = 0; irq < irq_count; irq++) { + ppc4xx_hsta_msi.irq_map[irq] = + irq_of_parse_and_map(dev->of_node, irq); + if (ppc4xx_hsta_msi.irq_map[irq] == NO_IRQ) { + dev_err(dev, "Unable to map IRQ\n"); + ret = -EINVAL; + goto out2; + } + } + + ppc_md.setup_msi_irqs = hsta_setup_msi_irqs; + ppc_md.teardown_msi_irqs = hsta_teardown_msi_irqs; + ppc_md.msi_check_device = hsta_msi_check_device; + return 0; + +out2: + kfree(ppc4xx_hsta_msi.irq_map); + +out1: + msi_bitmap_free(&ppc4xx_hsta_msi.bmp); + +out: + iounmap(ppc4xx_hsta_msi.data); + return ret; +} + +static const struct of_device_id hsta_msi_ids[] = { + { + .compatible = "ibm,hsta-msi", + }, + {} +}; + +static struct platform_driver hsta_msi_driver = { + .probe = hsta_msi_probe, + .driver = { + .name = "hsta-msi", + .owner = THIS_MODULE, + .of_match_table = hsta_msi_ids, + }, +}; + +static int hsta_msi_init(void) +{ + return platform_driver_register(&hsta_msi_driver); +} +subsys_initcall(hsta_msi_init); diff --git a/arch/powerpc/sysdev/ppc4xx_pci.c b/arch/powerpc/sysdev/ppc4xx_pci.c index 54ec1d5..7cc3acc 100644 --- a/arch/powerpc/sysdev/ppc4xx_pci.c +++ b/arch/powerpc/sysdev/ppc4xx_pci.c @@ -176,8 +176,12 @@ static int __init ppc4xx_parse_dma_ranges(struct pci_controller *hose, return -ENXIO; } - /* Check that we are fully contained within 32 bits space */ - if (res->end > 0xffffffff) { + /* Check that we are fully contained within 32 bits space if we are not + * running on a 460sx or 476fpe which have 64 bit bus addresses. + */ + if (res->end > 0xffffffff && + !(of_device_is_compatible(hose->dn, "ibm,plb-pciex-460sx") + || of_device_is_compatible(hose->dn, "ibm,plb-pciex-476fpe"))) { printk(KERN_ERR "%s: dma-ranges outside of 32 bits space\n", hose->dn->full_name); return -ENXIO;