Skip to contentSkip to conversation

Possible bug in sdio.c?

DISCUSSION Prashanth Bhat 3 participants 4 messages May 29, 2011 MMC, SD and SDIO

lore
Share & tools
lore
#0 lore raw

Hi Michał,

In mmc/core/sdio.c, routine mmc_enable_4bit_bus():

I think:

    if (card->type == MMC_TYPE_SDIO)
              return sdio_enable_wide(card);

should really be:

    if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO )
               return sdio_enable_wide(card);

Otherwise, the SDIO component of the SD Combo card will not operate in wide mode. What do you think?

Perhaps a similar change is needed in the sdio_enable_hs() routine too.

Thanks,
Prashanth

ps: I am sorry to be emailing you directly with questions/bug reports. What is the formal process of reporting suggestions for code change?

#1 lore raw
Quoted 10 lines · Prashanth Bhat

Hi Michał,
In mmc/core/sdio.c, routine mmc_enable_4bit_bus():

I think:

    if (card->type == MMC_TYPE_SDIO)
              return sdio_enable_wide(card);

should really be:

Quoted 7 lines
    if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO )
               return sdio_enable_wide(card);

Otherwise, the SDIO component of the SD Combo card will not operate in wide mode. What do you think?

Perhaps a similar change is needed in the sdio_enable_hs() routine too.

Hi,

You sent similar problem report some time ago. I assumed you will follow on with a patch fixing mmc_card_sd() and friends (the functions you point out in this mail should use that).

Best Regards,
Michał Mirosław

#2 lore raw
Quoted 20 lines · Michał Mirosław
Quoted 10 lines · Prashanth Bhat

Hi Michał,
In mmc/core/sdio.c, routine mmc_enable_4bit_bus():

I think:

     if (card->type == MMC_TYPE_SDIO)
               return sdio_enable_wide(card);

should really be:

Quoted 8 lines
     if (card->type == MMC_TYPE_SDIO || card->type == MMC_TYPE_SD_COMBO )
                return sdio_enable_wide(card);

Otherwise, the SDIO component of the SD Combo card will not operate in wide mode. What do you think?

Perhaps a similar change is needed in the sdio_enable_hs() routine too.

Quoted 11 lines

Hi,

You sent similar problem report some time ago. I assumed you will follow on with a patch fixing mmc_card_sd() and friends (the functions you point out in this mail should use that).

Best Regards,
Michał Mirosław

Thanks for the reply. Are there other places where this new function must be used? Is it only these 2 places?

Could you please clarify the formal process for including these changes into the build? I have made the previous changes in my private code tree only so far.

Thanks,
Prashanth

#3 lore raw

Hi,

Quoted 3 lines · Prashanth Bhat

Could you please clarify the formal process for including these changes into the build? I have made the previous changes in my private code tree only so far.

See Documentation/SubmittingPatches, and send the patch you generate to the same list of e-mail addresses you're using here.

If you have a git commit corresponding to the patch, you can use git send-email or git format-patch to generate the patch file.

Thanks,

- Chris.

Signature · 2 lines
Chris Ball   <[email protected]>   <http://printf.net/>
One Laptop Per Child