* [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region @ 2014-11-09 21:12 Alexey Charkov 2014-11-09 21:55 ` Fabio Estevam 0 siblings, 1 reply; 8+ messages in thread From: Alexey Charkov @ 2014-11-09 21:12 UTC (permalink / raw) To: Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel, linux-mmc, linux-kernel Cc: Helmut Stengele, Alexey Charkov From: Helmut Stengele <lautriv@coldplug.net> Current code calls release_mem_region upon module unload without requesting the region first. This patch fixes it by moving to a devres-based ioremap implementation, which handles requesting and releasing memory region internally. This also makes the error/unload paths a bit simpler. Signed-off-by: Helmut Stengele <lautriv@coldplug.net> (edited commit message) Signed-off-by: Alexey Charkov <alchark@gmail.com> --- Dear Chris, Ulf, This is a pretty trivial fix to a bug in wmt-sdmmc.c that is only visible upon module unload. Is there any chance to get it merged to the current fixes tree? I think I'd also like to convert other resource requests here to devres, but that would probably be a separate patch to go into -next. That's why we didn't rename the failX: labels, for example (as most of them could be dropped altogether if the respective requesting functions are converted to devres infrastructure). Thanks a lot, Alexey drivers/mmc/host/wmt-sdmmc.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c index 54181b4..02b5e6b 100644 --- a/drivers/mmc/host/wmt-sdmmc.c +++ b/drivers/mmc/host/wmt-sdmmc.c @@ -759,6 +759,7 @@ static int wmt_mci_probe(struct platform_device *pdev) const struct wmt_mci_caps *wmt_caps; int ret; int regular_irq, dma_irq; + struct resource *res; if (!of_id || !of_id->data) { dev_err(&pdev->dev, "Controller capabilities data missing\n"); @@ -781,6 +782,11 @@ static int wmt_mci_probe(struct platform_device *pdev) goto fail1; } + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) { + ret = -ENODEV; + goto fail1; + } mmc = mmc_alloc_host(sizeof(struct wmt_mci_priv), &pdev->dev); if (!mmc) { dev_err(&pdev->dev, "Failed to allocate mmc_host\n"); @@ -813,7 +819,7 @@ static int wmt_mci_probe(struct platform_device *pdev) if (of_get_property(np, "cd-inverted", NULL)) priv->cd_inverted = 1; - priv->sdmmc_base = of_iomap(np, 0); + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); if (!priv->sdmmc_base) { dev_err(&pdev->dev, "Failed to map IO space\n"); ret = -ENOMEM; @@ -826,7 +832,7 @@ static int wmt_mci_probe(struct platform_device *pdev) ret = request_irq(regular_irq, wmt_mci_regular_isr, 0, "sdmmc", priv); if (ret) { dev_err(&pdev->dev, "Register regular IRQ fail\n"); - goto fail3; + goto fail2; } ret = request_irq(dma_irq, wmt_mci_dma_isr, 0, "sdmmc", priv); @@ -869,8 +875,6 @@ fail5: free_irq(dma_irq, priv); fail4: free_irq(regular_irq, priv); -fail3: - iounmap(priv->sdmmc_base); fail2: mmc_free_host(mmc); fail1: @@ -881,7 +885,6 @@ static int wmt_mci_remove(struct platform_device *pdev) { struct mmc_host *mmc; struct wmt_mci_priv *priv; - struct resource *res; u32 reg_tmp; mmc = platform_get_drvdata(pdev); @@ -904,14 +907,9 @@ static int wmt_mci_remove(struct platform_device *pdev) free_irq(priv->irq_regular, priv); free_irq(priv->irq_dma, priv); - iounmap(priv->sdmmc_base); - clk_disable_unprepare(priv->clk_sdmmc); clk_put(priv->clk_sdmmc); - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - release_mem_region(res->start, resource_size(res)); - mmc_free_host(mmc); dev_info(&pdev->dev, "WMT MCI device removed\n"); -- 2.1.0 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-09 21:12 [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region Alexey Charkov @ 2014-11-09 21:55 ` Fabio Estevam 2014-11-09 22:18 ` lautriv 0 siblings, 1 reply; 8+ messages in thread From: Fabio Estevam @ 2014-11-09 21:55 UTC (permalink / raw) To: Alexey Charkov Cc: Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel, Helmut Stengele On Sun, Nov 9, 2014 at 7:12 PM, Alexey Charkov <alchark@gmail.com> wrote: > + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); > + if (!res) { > + ret = -ENODEV; > + goto fail1; You could save this NULL check... > + } > mmc = mmc_alloc_host(sizeof(struct wmt_mci_priv), &pdev->dev); > if (!mmc) { > dev_err(&pdev->dev, "Failed to allocate mmc_host\n"); > @@ -813,7 +819,7 @@ static int wmt_mci_probe(struct platform_device *pdev) > if (of_get_property(np, "cd-inverted", NULL)) > priv->cd_inverted = 1; > > - priv->sdmmc_base = of_iomap(np, 0); If you move ' res = platform_get_resource' right here as devm_ioremap_resource() already does the NULL check. > + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); > if (!priv->sdmmc_base) { > dev_err(&pdev->dev, "Failed to map IO space\n"); > ret = -ENOMEM; ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-09 21:55 ` Fabio Estevam @ 2014-11-09 22:18 ` lautriv 2014-11-10 5:15 ` Alexey Charkov 0 siblings, 1 reply; 8+ messages in thread From: lautriv @ 2014-11-09 22:18 UTC (permalink / raw) To: Fabio Estevam, Alexey Charkov Cc: Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel On 11/09/2014 10:55 PM, Fabio Estevam wrote: > On Sun, Nov 9, 2014 at 7:12 PM, Alexey Charkov <alchark@gmail.com> wrote: > >> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); >> + if (!res) { >> + ret = -ENODEV; >> + goto fail1; > You could save this NULL check... > >> + } >> mmc = mmc_alloc_host(sizeof(struct wmt_mci_priv), &pdev->dev); >> if (!mmc) { >> dev_err(&pdev->dev, "Failed to allocate mmc_host\n"); >> @@ -813,7 +819,7 @@ static int wmt_mci_probe(struct platform_device *pdev) >> if (of_get_property(np, "cd-inverted", NULL)) >> priv->cd_inverted = 1; >> >> - priv->sdmmc_base = of_iomap(np, 0); > If you move ' res = platform_get_resource' right here as > devm_ioremap_resource() already does the NULL check. > >> + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); >> if (!priv->sdmmc_base) { >> dev_err(&pdev->dev, "Failed to map IO space\n"); >> ret = -ENOMEM; This was the original intention but it would fall between the failX jumps, i discussed that with Alexey and we decided to change it with the cleanup to hold this patch small. ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-09 22:18 ` lautriv @ 2014-11-10 5:15 ` Alexey Charkov 2014-11-10 20:17 ` lautriv 2014-11-20 14:45 ` lautriv 0 siblings, 2 replies; 8+ messages in thread From: Alexey Charkov @ 2014-11-10 5:15 UTC (permalink / raw) To: lautriv Cc: Fabio Estevam, Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel 2014-11-10 1:18 GMT+03:00 lautriv <lautriv@coldplug.net>: > On 11/09/2014 10:55 PM, Fabio Estevam wrote: >> >> On Sun, Nov 9, 2014 at 7:12 PM, Alexey Charkov <alchark@gmail.com> wrote: >> >>> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); >>> + if (!res) { >>> + ret = -ENODEV; >>> + goto fail1; >> >> You could save this NULL check... >> >>> + } >>> mmc = mmc_alloc_host(sizeof(struct wmt_mci_priv), &pdev->dev); >>> if (!mmc) { >>> dev_err(&pdev->dev, "Failed to allocate mmc_host\n"); >>> @@ -813,7 +819,7 @@ static int wmt_mci_probe(struct platform_device >>> *pdev) >>> if (of_get_property(np, "cd-inverted", NULL)) >>> priv->cd_inverted = 1; >>> >>> - priv->sdmmc_base = of_iomap(np, 0); >> >> If you move ' res = platform_get_resource' right here as >> devm_ioremap_resource() already does the NULL check. >> >>> + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); >>> if (!priv->sdmmc_base) { >>> dev_err(&pdev->dev, "Failed to map IO space\n"); >>> ret = -ENOMEM; > > This was the original intention but it would fall between the failX jumps, i > discussed that with Alexey and we decided to change it with the cleanup to > hold this patch small. I think Fabio means something slightly different than what we discussed. We haven't realized that the first thing devm_ioremap_resource does is exactly that (!res) check, so we can skip ours altogether. Looking at the function code a bit closer, we should also make the check look like "if (IS_ERR(priv->sdmmc_base)", kill the dev_err call (as it's also done internally), and then just do "ret = PTR_ERR(priv->sdmmc_base)" and jump to the label. Thanks a lot, Alexey ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-10 5:15 ` Alexey Charkov @ 2014-11-10 20:17 ` lautriv 2014-11-20 14:45 ` lautriv 1 sibling, 0 replies; 8+ messages in thread From: lautriv @ 2014-11-10 20:17 UTC (permalink / raw) To: Alexey Charkov Cc: Fabio Estevam, Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel [-- Attachment #1: Type: text/plain, Size: 2059 bytes --] On 11/10/2014 06:15 AM, Alexey Charkov wrote: > 2014-11-10 1:18 GMT+03:00 lautriv <lautriv@coldplug.net>: >> On 11/09/2014 10:55 PM, Fabio Estevam wrote: >>> On Sun, Nov 9, 2014 at 7:12 PM, Alexey Charkov <alchark@gmail.com> wrote: >>> >>>> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); >>>> + if (!res) { >>>> + ret = -ENODEV; >>>> + goto fail1; >>> You could save this NULL check... >>> >>>> + } >>>> mmc = mmc_alloc_host(sizeof(struct wmt_mci_priv), &pdev->dev); >>>> if (!mmc) { >>>> dev_err(&pdev->dev, "Failed to allocate mmc_host\n"); >>>> @@ -813,7 +819,7 @@ static int wmt_mci_probe(struct platform_device >>>> *pdev) >>>> if (of_get_property(np, "cd-inverted", NULL)) >>>> priv->cd_inverted = 1; >>>> >>>> - priv->sdmmc_base = of_iomap(np, 0); >>> If you move ' res = platform_get_resource' right here as >>> devm_ioremap_resource() already does the NULL check. >>> >>>> + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); >>>> if (!priv->sdmmc_base) { >>>> dev_err(&pdev->dev, "Failed to map IO space\n"); >>>> ret = -ENOMEM; >> This was the original intention but it would fall between the failX jumps, i >> discussed that with Alexey and we decided to change it with the cleanup to >> hold this patch small. > I think Fabio means something slightly different than what we > discussed. We haven't realized that the first thing > devm_ioremap_resource does is exactly that (!res) check, so we can > skip ours altogether. > > Looking at the function code a bit closer, we should also make the > check look like "if (IS_ERR(priv->sdmmc_base)", kill the dev_err call > (as it's also done internally), and then just do "ret = > PTR_ERR(priv->sdmmc_base)" and jump to the label. > > Thanks a lot, > Alexey > Sorry for the delay, 8880's are used to crash a lot since we have not enought information ;) Attached a changed patch which will hopefully meet all needs. [-- Attachment #2: 0001-mmc-vt8500-change-resource-mapping-to-devm_.patch --] [-- Type: text/x-patch, Size: 2730 bytes --] >From 7c8ebb55fde8786cd16d95659ff73675bedc287c Mon Sep 17 00:00:00 2001 From: Helmut Stengele <lautriv@coldplug.net> Date: Mon, 10 Nov 2014 11:26:51 +0100 Subject: [PATCH] mmc: vt8500: change resource-mapping to devm_* Current code calls release_mem_region upon module unload without requesting the region first. This patch fixes it by moving to a devres-based ioremap implementation, which handles requesting and releasing memory region internally. This also makes the error/unload paths a bit simpler. Signed-off-by: Helmut Stengele <lautriv@coldplug.net> --- drivers/mmc/host/wmt-sdmmc.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c index 54181b4..f874d36 100644 --- a/drivers/mmc/host/wmt-sdmmc.c +++ b/drivers/mmc/host/wmt-sdmmc.c @@ -759,6 +759,7 @@ static int wmt_mci_probe(struct platform_device *pdev) const struct wmt_mci_caps *wmt_caps; int ret; int regular_irq, dma_irq; + struct resource *res; if (!of_id || !of_id->data) { dev_err(&pdev->dev, "Controller capabilities data missing\n"); @@ -813,10 +814,11 @@ static int wmt_mci_probe(struct platform_device *pdev) if (of_get_property(np, "cd-inverted", NULL)) priv->cd_inverted = 1; - priv->sdmmc_base = of_iomap(np, 0); + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); if (!priv->sdmmc_base) { dev_err(&pdev->dev, "Failed to map IO space\n"); - ret = -ENOMEM; + ret = PTR_ERR(priv->sdmmc_base); goto fail2; } @@ -826,7 +828,7 @@ static int wmt_mci_probe(struct platform_device *pdev) ret = request_irq(regular_irq, wmt_mci_regular_isr, 0, "sdmmc", priv); if (ret) { dev_err(&pdev->dev, "Register regular IRQ fail\n"); - goto fail3; + goto fail2; } ret = request_irq(dma_irq, wmt_mci_dma_isr, 0, "sdmmc", priv); @@ -869,8 +871,6 @@ fail5: free_irq(dma_irq, priv); fail4: free_irq(regular_irq, priv); -fail3: - iounmap(priv->sdmmc_base); fail2: mmc_free_host(mmc); fail1: @@ -881,7 +881,6 @@ static int wmt_mci_remove(struct platform_device *pdev) { struct mmc_host *mmc; struct wmt_mci_priv *priv; - struct resource *res; u32 reg_tmp; mmc = platform_get_drvdata(pdev); @@ -904,14 +903,9 @@ static int wmt_mci_remove(struct platform_device *pdev) free_irq(priv->irq_regular, priv); free_irq(priv->irq_dma, priv); - iounmap(priv->sdmmc_base); - clk_disable_unprepare(priv->clk_sdmmc); clk_put(priv->clk_sdmmc); - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - release_mem_region(res->start, resource_size(res)); - mmc_free_host(mmc); dev_info(&pdev->dev, "WMT MCI device removed\n"); -- 1.7.10.4 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-10 5:15 ` Alexey Charkov 2014-11-10 20:17 ` lautriv @ 2014-11-20 14:45 ` lautriv 2014-11-24 8:34 ` Ulf Hansson 1 sibling, 1 reply; 8+ messages in thread From: lautriv @ 2014-11-20 14:45 UTC (permalink / raw) To: Alexey Charkov Cc: Fabio Estevam, Tony Prisk, Chris Ball, Ulf Hansson, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel From b0509e27e33326e6dccd67d8ebe67e2bdb0cfdde Mon Sep 17 00:00:00 2001 From: Helmut Stengele <lautriv@coldplug.net> Date: Thu, 20 Nov 2014 15:27:40 +0100 Subject: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region Current code calls release_mem_region upon module unload without requesting the region first. This patch fixes it by moving to a devres-based ioremap implementation, which handles requesting and releasing memory region internally. This also makes the error/unload paths a bit simpler. Signed-off-by: Helmut Stengele <lautriv@coldplug.net> --- This is a corrected patch based on the one Alexey sent for me. I sent it already but since i'm not used to send something upstream it was an attached file and obviously rejected. Patch is tested on bare-metal ( wm8850 ) and works like suspected. drivers/mmc/host/wmt-sdmmc.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c index 54181b4..bcaa0cc 100644 --- a/drivers/mmc/host/wmt-sdmmc.c +++ b/drivers/mmc/host/wmt-sdmmc.c @@ -759,6 +759,7 @@ static int wmt_mci_probe(struct platform_device *pdev) const struct wmt_mci_caps *wmt_caps; int ret; int regular_irq, dma_irq; + struct resource *res; if (!of_id || !of_id->data) { dev_err(&pdev->dev, "Controller capabilities data missing\n"); @@ -813,10 +814,11 @@ static int wmt_mci_probe(struct platform_device *pdev) if (of_get_property(np, "cd-inverted", NULL)) priv->cd_inverted = 1; - priv->sdmmc_base = of_iomap(np, 0); + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); if (!priv->sdmmc_base) { dev_err(&pdev->dev, "Failed to map IO space\n"); - ret = -ENOMEM; + ret = PTR_ERR(priv->sdmmc_base); goto fail2; } @@ -826,7 +828,7 @@ static int wmt_mci_probe(struct platform_device *pdev) ret = request_irq(regular_irq, wmt_mci_regular_isr, 0, "sdmmc", priv); if (ret) { dev_err(&pdev->dev, "Register regular IRQ fail\n"); - goto fail3; + goto fail2; } ret = request_irq(dma_irq, wmt_mci_dma_isr, 0, "sdmmc", priv); @@ -869,8 +871,6 @@ fail5: free_irq(dma_irq, priv); fail4: free_irq(regular_irq, priv); -fail3: - iounmap(priv->sdmmc_base); fail2: mmc_free_host(mmc); fail1: @@ -881,7 +881,6 @@ static int wmt_mci_remove(struct platform_device *pdev) { struct mmc_host *mmc; struct wmt_mci_priv *priv; - struct resource *res; u32 reg_tmp; mmc = platform_get_drvdata(pdev); @@ -909,9 +908,6 @@ static int wmt_mci_remove(struct platform_device *pdev) clk_disable_unprepare(priv->clk_sdmmc); clk_put(priv->clk_sdmmc); - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - release_mem_region(res->start, resource_size(res)); - mmc_free_host(mmc); dev_info(&pdev->dev, "WMT MCI device removed\n"); -- 2.1.3 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-20 14:45 ` lautriv @ 2014-11-24 8:34 ` Ulf Hansson 2014-11-25 22:34 ` lautriv 0 siblings, 1 reply; 8+ messages in thread From: Ulf Hansson @ 2014-11-24 8:34 UTC (permalink / raw) To: lautriv Cc: Alexey Charkov, Fabio Estevam, Tony Prisk, Chris Ball, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel On 20 November 2014 at 15:45, lautriv <lautriv@coldplug.net> wrote: > From b0509e27e33326e6dccd67d8ebe67e2bdb0cfdde Mon Sep 17 00:00:00 2001 > From: Helmut Stengele <lautriv@coldplug.net> > Date: Thu, 20 Nov 2014 15:27:40 +0100 > Subject: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region > > Current code calls release_mem_region upon module unload > without requesting the region first. This patch fixes it > by moving to a devres-based ioremap implementation, which > handles requesting and releasing memory region internally. > > This also makes the error/unload paths a bit simpler. > > Signed-off-by: Helmut Stengele <lautriv@coldplug.net> > --- > > This is a corrected patch based on the one Alexey sent for me. > I sent it already but since i'm not used to send something upstream > it was an attached file and obviously rejected. > Patch is tested on bare-metal ( wm8850 ) and works like suspected. > > drivers/mmc/host/wmt-sdmmc.c | 14 +++++--------- > 1 file changed, 5 insertions(+), 9 deletions(-) > > diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c > index 54181b4..bcaa0cc 100644 > --- a/drivers/mmc/host/wmt-sdmmc.c > +++ b/drivers/mmc/host/wmt-sdmmc.c > @@ -759,6 +759,7 @@ static int wmt_mci_probe(struct platform_device *pdev) > const struct wmt_mci_caps *wmt_caps; > int ret; > int regular_irq, dma_irq; > + struct resource *res; > > if (!of_id || !of_id->data) { > dev_err(&pdev->dev, "Controller capabilities data missing\n"); > @@ -813,10 +814,11 @@ static int wmt_mci_probe(struct platform_device *pdev) > if (of_get_property(np, "cd-inverted", NULL)) > priv->cd_inverted = 1; > > - priv->sdmmc_base = of_iomap(np, 0); > + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); > + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); You should check the return value from devm_ioremap_resource() with IS_ERR(). > if (!priv->sdmmc_base) { > dev_err(&pdev->dev, "Failed to map IO space\n"); This print can be removed, devm_ioremap_resource() handles that. > - ret = -ENOMEM; > + ret = PTR_ERR(priv->sdmmc_base); > goto fail2; > } > > @@ -826,7 +828,7 @@ static int wmt_mci_probe(struct platform_device *pdev) > ret = request_irq(regular_irq, wmt_mci_regular_isr, 0, "sdmmc", priv); > if (ret) { > dev_err(&pdev->dev, "Register regular IRQ fail\n"); > - goto fail3; > + goto fail2; > } > > ret = request_irq(dma_irq, wmt_mci_dma_isr, 0, "sdmmc", priv); > @@ -869,8 +871,6 @@ fail5: > free_irq(dma_irq, priv); > fail4: > free_irq(regular_irq, priv); > -fail3: > - iounmap(priv->sdmmc_base); > fail2: > mmc_free_host(mmc); > fail1: > @@ -881,7 +881,6 @@ static int wmt_mci_remove(struct platform_device *pdev) > { > struct mmc_host *mmc; > struct wmt_mci_priv *priv; > - struct resource *res; > u32 reg_tmp; > > mmc = platform_get_drvdata(pdev); > @@ -909,9 +908,6 @@ static int wmt_mci_remove(struct platform_device *pdev) > clk_disable_unprepare(priv->clk_sdmmc); > clk_put(priv->clk_sdmmc); > > - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); > - release_mem_region(res->start, resource_size(res)); > - > mmc_free_host(mmc); > > dev_info(&pdev->dev, "WMT MCI device removed\n"); > -- > 2.1.3 > > Kind regards Uffe ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region 2014-11-24 8:34 ` Ulf Hansson @ 2014-11-25 22:34 ` lautriv 0 siblings, 0 replies; 8+ messages in thread From: lautriv @ 2014-11-25 22:34 UTC (permalink / raw) To: Ulf Hansson Cc: Alexey Charkov, Fabio Estevam, Tony Prisk, Chris Ball, linux-arm-kernel@lists.infradead.org, linux-mmc@vger.kernel.org, linux-kernel On 11/24/2014 09:34 AM, Ulf Hansson wrote: > On 20 November 2014 at 15:45, lautriv <lautriv@coldplug.net> wrote: >> From b0509e27e33326e6dccd67d8ebe67e2bdb0cfdde Mon Sep 17 00:00:00 2001 >> From: Helmut Stengele <lautriv@coldplug.net> >> Date: Thu, 20 Nov 2014 15:27:40 +0100 >> Subject: [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region >> >> Current code calls release_mem_region upon module unload >> without requesting the region first. This patch fixes it >> by moving to a devres-based ioremap implementation, which >> handles requesting and releasing memory region internally. >> >> This also makes the error/unload paths a bit simpler. >> >> Signed-off-by: Helmut Stengele <lautriv@coldplug.net> >> --- >> >> This is a corrected patch based on the one Alexey sent for me. >> I sent it already but since i'm not used to send something upstream >> it was an attached file and obviously rejected. >> Patch is tested on bare-metal ( wm8850 ) and works like suspected. >> >> drivers/mmc/host/wmt-sdmmc.c | 14 +++++--------- >> 1 file changed, 5 insertions(+), 9 deletions(-) >> >> diff --git a/drivers/mmc/host/wmt-sdmmc.c b/drivers/mmc/host/wmt-sdmmc.c >> index 54181b4..bcaa0cc 100644 >> --- a/drivers/mmc/host/wmt-sdmmc.c >> +++ b/drivers/mmc/host/wmt-sdmmc.c >> @@ -759,6 +759,7 @@ static int wmt_mci_probe(struct platform_device *pdev) >> const struct wmt_mci_caps *wmt_caps; >> int ret; >> int regular_irq, dma_irq; >> + struct resource *res; >> >> if (!of_id || !of_id->data) { >> dev_err(&pdev->dev, "Controller capabilities data missing\n"); >> @@ -813,10 +814,11 @@ static int wmt_mci_probe(struct platform_device *pdev) >> if (of_get_property(np, "cd-inverted", NULL)) >> priv->cd_inverted = 1; >> >> - priv->sdmmc_base = of_iomap(np, 0); >> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); >> + priv->sdmmc_base = devm_ioremap_resource(&pdev->dev, res); > You should check the return value from devm_ioremap_resource() with IS_ERR(). > >> if (!priv->sdmmc_base) { >> dev_err(&pdev->dev, "Failed to map IO space\n"); > This print can be removed, devm_ioremap_resource() handles that. > >> - ret = -ENOMEM; >> + ret = PTR_ERR(priv->sdmmc_base); >> goto fail2; >> } >> >> @@ -826,7 +828,7 @@ static int wmt_mci_probe(struct platform_device *pdev) >> ret = request_irq(regular_irq, wmt_mci_regular_isr, 0, "sdmmc", priv); >> if (ret) { >> dev_err(&pdev->dev, "Register regular IRQ fail\n"); >> - goto fail3; >> + goto fail2; >> } >> >> ret = request_irq(dma_irq, wmt_mci_dma_isr, 0, "sdmmc", priv); >> @@ -869,8 +871,6 @@ fail5: >> free_irq(dma_irq, priv); >> fail4: >> free_irq(regular_irq, priv); >> -fail3: >> - iounmap(priv->sdmmc_base); >> fail2: >> mmc_free_host(mmc); >> fail1: >> @@ -881,7 +881,6 @@ static int wmt_mci_remove(struct platform_device *pdev) >> { >> struct mmc_host *mmc; >> struct wmt_mci_priv *priv; >> - struct resource *res; >> u32 reg_tmp; >> >> mmc = platform_get_drvdata(pdev); >> @@ -909,9 +908,6 @@ static int wmt_mci_remove(struct platform_device *pdev) >> clk_disable_unprepare(priv->clk_sdmmc); >> clk_put(priv->clk_sdmmc); >> >> - res = platform_get_resource(pdev, IORESOURCE_MEM, 0); >> - release_mem_region(res->start, resource_size(res)); >> - >> mmc_free_host(mmc); >> >> dev_info(&pdev->dev, "WMT MCI device removed\n"); >> -- >> 2.1.3 >> >> > Kind regards > Uffe > Thanks for the hint, will do so. ( Sent the wrong patch by accident anyway ) ^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2014-11-25 22:34 UTC | newest] Thread overview: 8+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2014-11-09 21:12 [PATCH] mmc: wmt-sdmmc: fix unmatched release_mem_region Alexey Charkov 2014-11-09 21:55 ` Fabio Estevam 2014-11-09 22:18 ` lautriv 2014-11-10 5:15 ` Alexey Charkov 2014-11-10 20:17 ` lautriv 2014-11-20 14:45 ` lautriv 2014-11-24 8:34 ` Ulf Hansson 2014-11-25 22:34 ` lautriv
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).