[PATCH] mmc: sdhci: Avoid swiotlb buffer being full

[PATCH] mmc: sdhci: Avoid swiotlb buffer being full

[Ulf Hansson]

The commit de3ee99b097d ("mmc: Delete bounce buffer handling") deletes the
bounce buffer handling, but also causes the max_req_size for sdhci to be
increased, in case when max_segs == 1. This causes errors for sdhci-pci
Ricoh variant, about the swiotlb buffer to become full.

Fix the issue, by taking IO_TLB_SEGSIZE and IO_TLB_SHIFT into account when
deciding the max_req_size for sdhci.

Reported-by: Jiri Slaby <jslaby@suse.cz>
Fixes: de3ee99b097d ("mmc: Delete bounce buffer handling")
Cc: <stable@vger.kernel.org> # v4.14+
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Tested-by: Jiri Slaby <jslaby@suse.cz>
---
 drivers/mmc/host/sdhci.c | 28 ++++++++++++++++++----------
 1 file changed, 18 insertions(+), 10 deletions(-)

diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index 2f14334..e9290a3 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -21,6 +21,7 @@
 #include <linux/dma-mapping.h>
 #include <linux/slab.h>
 #include <linux/scatterlist.h>
+#include <linux/swiotlb.h>
 #include <linux/regulator/consumer.h>
 #include <linux/pm_runtime.h>
 #include <linux/of.h>
@@ -3651,22 +3652,29 @@ int sdhci_setup_host(struct sdhci_host *host)
 	spin_lock_init(&host->lock);
 
 	/*
+	 * Maximum number of sectors in one transfer. Limited by SDMA boundary
+	 * size (512KiB). Note some tuning modes impose a 4MiB limit, but this
+	 * is less anyway.
+	 */
+	mmc->max_req_size = 524288;
+
+	/*
 	 * Maximum number of segments. Depends on if the hardware
 	 * can do scatter/gather or not.
 	 */
-	if (host->flags & SDHCI_USE_ADMA)
+	if (host->flags & SDHCI_USE_ADMA) {
 		mmc->max_segs = SDHCI_MAX_SEGS;
-	else if (host->flags & SDHCI_USE_SDMA)
+	} else if (host->flags & SDHCI_USE_SDMA) {
 		mmc->max_segs = 1;
-	else /* PIO */
+		if (swiotlb_max_segment()) {
+			unsigned int max_req_size = (1 << IO_TLB_SHIFT) *
+						IO_TLB_SEGSIZE;
+			mmc->max_req_size = min(mmc->max_req_size,
+						max_req_size);
+		}
+	} else { /* PIO */
 		mmc->max_segs = SDHCI_MAX_SEGS;
-
-	/*
-	 * Maximum number of sectors in one transfer. Limited by SDMA boundary
-	 * size (512KiB). Note some tuning modes impose a 4MiB limit, but this
-	 * is less anyway.
-	 */
-	mmc->max_req_size = 524288;
+	}
 
 	/*
 	 * Maximum segment size. Could be one segment with the maximum number
-- 
2.7.4

Re: [PATCH] mmc: sdhci: Avoid swiotlb buffer being full

[Adrian Hunter]

On 27/11/17 12:28, Ulf Hansson wrote:
> The commit de3ee99b097d ("mmc: Delete bounce buffer handling") deletes the
> bounce buffer handling, but also causes the max_req_size for sdhci to be
> increased, in case when max_segs == 1. This causes errors for sdhci-pci
> Ricoh variant, about the swiotlb buffer to become full.
> 
> Fix the issue, by taking IO_TLB_SEGSIZE and IO_TLB_SHIFT into account when
> deciding the max_req_size for sdhci.
> 
> Reported-by: Jiri Slaby <jslaby@suse.cz>
> Fixes: de3ee99b097d ("mmc: Delete bounce buffer handling")
> Cc: <stable@vger.kernel.org> # v4.14+
> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
> Tested-by: Jiri Slaby <jslaby@suse.cz>

Acked-by: Adrian Hunter <adrian.hunter@intel.com>

> ---
>  drivers/mmc/host/sdhci.c | 28 ++++++++++++++++++----------
>  1 file changed, 18 insertions(+), 10 deletions(-)
> 
> diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
> index 2f14334..e9290a3 100644
> --- a/drivers/mmc/host/sdhci.c
> +++ b/drivers/mmc/host/sdhci.c
> @@ -21,6 +21,7 @@
>  #include <linux/dma-mapping.h>
>  #include <linux/slab.h>
>  #include <linux/scatterlist.h>
> +#include <linux/swiotlb.h>
>  #include <linux/regulator/consumer.h>
>  #include <linux/pm_runtime.h>
>  #include <linux/of.h>
> @@ -3651,22 +3652,29 @@ int sdhci_setup_host(struct sdhci_host *host)
>  	spin_lock_init(&host->lock);
>  
>  	/*
> +	 * Maximum number of sectors in one transfer. Limited by SDMA boundary
> +	 * size (512KiB). Note some tuning modes impose a 4MiB limit, but this
> +	 * is less anyway.
> +	 */
> +	mmc->max_req_size = 524288;
> +
> +	/*
>  	 * Maximum number of segments. Depends on if the hardware
>  	 * can do scatter/gather or not.
>  	 */
> -	if (host->flags & SDHCI_USE_ADMA)
> +	if (host->flags & SDHCI_USE_ADMA) {
>  		mmc->max_segs = SDHCI_MAX_SEGS;
> -	else if (host->flags & SDHCI_USE_SDMA)
> +	} else if (host->flags & SDHCI_USE_SDMA) {
>  		mmc->max_segs = 1;
> -	else /* PIO */
> +		if (swiotlb_max_segment()) {
> +			unsigned int max_req_size = (1 << IO_TLB_SHIFT) *
> +						IO_TLB_SEGSIZE;
> +			mmc->max_req_size = min(mmc->max_req_size,
> +						max_req_size);
> +		}
> +	} else { /* PIO */
>  		mmc->max_segs = SDHCI_MAX_SEGS;
> -
> -	/*
> -	 * Maximum number of sectors in one transfer. Limited by SDMA boundary
> -	 * size (512KiB). Note some tuning modes impose a 4MiB limit, but this
> -	 * is less anyway.
> -	 */
> -	mmc->max_req_size = 524288;
> +	}
>  
>  	/*
>  	 * Maximum segment size. Could be one segment with the maximum number
> 

Re: [PATCH] mmc: sdhci: Avoid swiotlb buffer being full

[Linus Walleij]

On Mon, Nov 27, 2017 at 11:28 AM, Ulf Hansson <ulf.hansson@linaro.org> wrote:

> The commit de3ee99b097d ("mmc: Delete bounce buffer handling") deletes the
> bounce buffer handling, but also causes the max_req_size for sdhci to be
> increased, in case when max_segs == 1. This causes errors for sdhci-pci
> Ricoh variant, about the swiotlb buffer to become full.
>
> Fix the issue, by taking IO_TLB_SEGSIZE and IO_TLB_SHIFT into account when
> deciding the max_req_size for sdhci.
>
> Reported-by: Jiri Slaby <jslaby@suse.cz>
> Fixes: de3ee99b097d ("mmc: Delete bounce buffer handling")
> Cc: <stable@vger.kernel.org> # v4.14+
> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
> Tested-by: Jiri Slaby <jslaby@suse.cz>

Reviewed-by: Linus Walleij <linus.walleij@linaro.org>

Thank you for fixing my screwups :/

Yours,
Linus Walleij

Re: [PATCH] mmc: sdhci: Avoid swiotlb buffer being full

[Ulf Hansson]

On 1 December 2017 at 11:01, Linus Walleij <linus.walleij@linaro.org> wrote:
> On Mon, Nov 27, 2017 at 11:28 AM, Ulf Hansson <ulf.hansson@linaro.org> wrote:
>
>> The commit de3ee99b097d ("mmc: Delete bounce buffer handling") deletes the
>> bounce buffer handling, but also causes the max_req_size for sdhci to be
>> increased, in case when max_segs == 1. This causes errors for sdhci-pci
>> Ricoh variant, about the swiotlb buffer to become full.
>>
>> Fix the issue, by taking IO_TLB_SEGSIZE and IO_TLB_SHIFT into account when
>> deciding the max_req_size for sdhci.
>>
>> Reported-by: Jiri Slaby <jslaby@suse.cz>
>> Fixes: de3ee99b097d ("mmc: Delete bounce buffer handling")
>> Cc: <stable@vger.kernel.org> # v4.14+
>> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
>> Tested-by: Jiri Slaby <jslaby@suse.cz>
>
> Reviewed-by: Linus Walleij <linus.walleij@linaro.org>

Sorry, but this was too late, I just sent the PR to Linus.

Thanks anyway!

[...]

Kind regards
Uffe