[PATCH] mmc: core: fix au_size according to sd spec3.0

[PATCH] mmc: core: fix au_size according to sd spec3.0

[hyeonsu.kim]

From: Hyeonsu Kim <hyeonsu.kim@samsung.com>

This "if phrase" is illogical.
and SD Spec 3.0 supports au_size from 0 to 0xF.

Signed-off-by: Hyeonsu Kim <hyeonsu.kim@samsung.com>
---
 drivers/mmc/core/sd.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
index c63ad03..cf8bf0e 100644
--- a/drivers/mmc/core/sd.c
+++ b/drivers/mmc/core/sd.c
@@ -244,7 +244,7 @@ static int mmc_read_ssr(struct mmc_card *card)
 	 * bitfield positions accordingly.
 	 */
 	au = UNSTUFF_BITS(ssr, 428 - 384, 4);
-	if (au > 0 || au <= 9) {
+	if (au >= 0 && au < 16) {
 		card->ssr.au = 1 << (au + 4);
 		es = UNSTUFF_BITS(ssr, 408 - 384, 16);
 		et = UNSTUFF_BITS(ssr, 402 - 384, 6);
-- 
1.7.5.4

Re: [PATCH] mmc: core: fix au_size according to sd spec3.0

[Jaehoon Chung]

Hi.
On 02/17/2012 07:14 PM, hyeonsu.kim@samsung.com wrote:

> From: Hyeonsu Kim <hyeonsu.kim@samsung.com>
> 
> This "if phrase" is illogical.
> and SD Spec 3.0 supports au_size from 0 to 0xF.
> 
> Signed-off-by: Hyeonsu Kim <hyeonsu.kim@samsung.com>
> ---
>  drivers/mmc/core/sd.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
> index c63ad03..cf8bf0e 100644
> --- a/drivers/mmc/core/sd.c
> +++ b/drivers/mmc/core/sd.c
> @@ -244,7 +244,7 @@ static int mmc_read_ssr(struct mmc_card *card)
>  	 * bitfield positions accordingly.
>  	 */
>  	au = UNSTUFF_BITS(ssr, 428 - 384, 4);
> -	if (au > 0 || au <= 9) {
> +	if (au >= 0 && au < 16) {

I think "if (au > 0 && au <= 0xf)" is more readable and 
need to check the sd-spec.
Because in SD2.0, didn't define from 0xa to 0xf.

>  		card->ssr.au = 1 << (au + 4);

Is it right???
card->ssr.au = 1 << (au + 4) is something wrong.
If au is 1, card->ssr.au is 32KB. BUt in spec, defined 16KB.
I think the wrong bit operation. this is more problem.

If my point is right, will send the patch.

Best Regards,
Jaehoon Chung

>  		es = UNSTUFF_BITS(ssr, 408 - 384, 16);
>  		et = UNSTUFF_BITS(ssr, 402 - 384, 6);

Re: [PATCH] mmc: core: fix au_size according to sd spec3.0

[Adrian Hunter]

On 17/02/12 12:46, Jaehoon Chung wrote:
> Hi.
> On 02/17/2012 07:14 PM, hyeonsu.kim@samsung.com wrote:
> 
>> From: Hyeonsu Kim <hyeonsu.kim@samsung.com>
>>
>> This "if phrase" is illogical.
>> and SD Spec 3.0 supports au_size from 0 to 0xF.
>>
>> Signed-off-by: Hyeonsu Kim <hyeonsu.kim@samsung.com>
>> ---
>>  drivers/mmc/core/sd.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c
>> index c63ad03..cf8bf0e 100644
>> --- a/drivers/mmc/core/sd.c
>> +++ b/drivers/mmc/core/sd.c
>> @@ -244,7 +244,7 @@ static int mmc_read_ssr(struct mmc_card *card)
>>  	 * bitfield positions accordingly.
>>  	 */
>>  	au = UNSTUFF_BITS(ssr, 428 - 384, 4);
>> -	if (au > 0 || au <= 9) {
>> +	if (au >= 0 && au < 16) {
> 
> I think "if (au > 0 && au <= 0xf)" is more readable and 
> need to check the sd-spec.
> Because in SD2.0, didn't define from 0xa to 0xf.
> 
>>  		card->ssr.au = 1 << (au + 4);
> 
> Is it right???
> card->ssr.au = 1 << (au + 4) is something wrong.
> If au is 1, card->ssr.au is 32KB. BUt in spec, defined 16KB.
> I think the wrong bit operation. this is more problem.

No it is 16KiB because "au" is in sectors.

RE: [PATCH] mmc: core: fix au_size according to sd spec3.0

[Subhash Jadavani]

> -----Original Message-----
> From: linux-mmc-owner@vger.kernel.org [mailto:linux-mmc-
> owner@vger.kernel.org] On Behalf Of hyeonsu.kim@samsung.com
> Sent: Friday, February 17, 2012 3:45 PM
> To: linux-mmc@vger.kernel.org; cjb@laptop.org
> Cc: kyungmin.park@samsung.com; jh80.chung@samsung.com; Hyeonsu Kim
> Subject: [PATCH] mmc: core: fix au_size according to sd spec3.0
> 
> From: Hyeonsu Kim <hyeonsu.kim@samsung.com>
> 
> This "if phrase" is illogical.
> and SD Spec 3.0 supports au_size from 0 to 0xF.
> 
> Signed-off-by: Hyeonsu Kim <hyeonsu.kim@samsung.com>
> ---
>  drivers/mmc/core/sd.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/mmc/core/sd.c b/drivers/mmc/core/sd.c index
> c63ad03..cf8bf0e 100644
> --- a/drivers/mmc/core/sd.c
> +++ b/drivers/mmc/core/sd.c
> @@ -244,7 +244,7 @@ static int mmc_read_ssr(struct mmc_card *card)
>  	 * bitfield positions accordingly.
>  	 */
>  	au = UNSTUFF_BITS(ssr, 428 - 384, 4);
> -	if (au > 0 || au <= 9) {
> +	if (au >= 0 && au < 16) {

au = 0 is still invalid. And as "Jaehoon Chung" mentioned, we would need to
check the SD spec version here as au value from 10 to 15 is invalide for
cards till SD2.0 spec.

>  		card->ssr.au = 1 << (au + 4);
>  		es = UNSTUFF_BITS(ssr, 408 - 384, 16);
>  		et = UNSTUFF_BITS(ssr, 402 - 384, 6);
> --
> 1.7.5.4
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in
the
> body of a message to majordomo@vger.kernel.org More majordomo info at
> http://vger.kernel.org/majordomo-info.html