From patchwork Thu Mar 20 17:34:37 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ben Dooks X-Patchwork-Id: 332336 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 50FDB2C007C for ; Fri, 21 Mar 2014 04:34:50 +1100 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1759557AbaCTReo (ORCPT ); Thu, 20 Mar 2014 13:34:44 -0400 Received: from access.ducie-dc1.codethink.co.uk ([185.25.241.217]:35148 "EHLO rt.codethink.co.uk" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1754990AbaCTRen (ORCPT ); Thu, 20 Mar 2014 13:34:43 -0400 Received: from rainbowdash.ducie.codethink.co.uk (localhost [127.0.0.1]) by rt.codethink.co.uk (Postfix) with ESMTPS id 61FCA1A0BDD; Thu, 20 Mar 2014 17:34:42 +0000 (GMT) Received: from ben by rainbowdash.ducie.codethink.co.uk with local (Exim 4.82) (envelope-from ) id 1WQgrZ-0003Eb-5y; Thu, 20 Mar 2014 18:34:41 +0100 From: Ben Dooks To: linux-kernel@lists.codethink.co.uk, netdev@vger.kernel.org Cc: linux-sh@vger.kernel.org, davem@davemloft.net, Ben Dooks , sergei.shtylyov@cogentembedded.com, laurent.pinchart+renesas@ideasonboard.com Subject: [PATCH v3] sh_eth: ensure pm_runtime cannot suspend the device during init Date: Thu, 20 Mar 2014 18:34:37 +0100 Message-Id: <1395336877-12402-1-git-send-email-ben.dooks@codethink.co.uk> X-Mailer: git-send-email 1.9.0 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org The pm_rumtime work queue is causing the device to be suspended during initialisation, thus the initialisation may not be able to access registers properly. As the code is called from a work queue, it is possible that this is not seen from certain configurations/builds due to the asynchronos nature of the code. Another issue has also been found where the network device registration calls back into the driver thus causing further pm_runtime calls that also caused issues with the MDIO bus code. Use pm_runtime_get_sync() and pm_runtime_put() to ensure that the pm system does not suspend it during the probe() call and remove the now unnecessary pm_runtime_resume() call. This fixes the external abort that can cause /sbin/init or other such init processed to die. Signed-off-by: Ben Dooks Tested-by: Geert Uytterhoeven --- Cc: sergei.shtylyov@cogentembedded.com Cc: laurent.pinchart+renesas@ideasonboard.com cc: netdev@vger.kernel.org Note, Laurent this should probably be applied before your current series as it may require changes to the exit sequence. Fixes from v1: - use pm_runtime_put() over pm_runtime_put_sync() as we do not need to guaranteed the device has powered off after probe. Fixed from v2: - call pm_runtime_put() in error path --- drivers/net/ethernet/renesas/sh_eth.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/renesas/sh_eth.c b/drivers/net/ethernet/renesas/sh_eth.c index 4f76b5e..610da51 100644 --- a/drivers/net/ethernet/renesas/sh_eth.c +++ b/drivers/net/ethernet/renesas/sh_eth.c @@ -2841,6 +2841,9 @@ static int sh_eth_drv_probe(struct platform_device *pdev) goto out; } + pm_runtime_enable(&pdev->dev); + pm_runtime_get_sync(&pdev->dev); + /* The sh Ether-specific entries in the device structure. */ ndev->base_addr = res->start; devno = pdev->id; @@ -2868,8 +2871,6 @@ static int sh_eth_drv_probe(struct platform_device *pdev) spin_lock_init(&mdp->lock); mdp->pdev = pdev; - pm_runtime_enable(&pdev->dev); - pm_runtime_resume(&pdev->dev); if (pdev->dev.of_node) pd = sh_eth_parse_dt(&pdev->dev); @@ -2961,6 +2962,7 @@ static int sh_eth_drv_probe(struct platform_device *pdev) pr_info("Base address at 0x%x, %pM, IRQ %d.\n", (u32)ndev->base_addr, ndev->dev_addr, ndev->irq); + pm_runtime_put(&pdev->dev); platform_set_drvdata(pdev, ndev); return ret; @@ -2976,6 +2978,7 @@ out_release: if (ndev) free_netdev(ndev); + pm_runtime_put(&pdev->dev); out: return ret; }