[PATCH] sdio: fix suspend/resume regression

[ohad@wizery.com (Ohad Ben-Cohen)]

On Sat, Oct 23, 2010 at 4:18 PM, Maxim Levitsky <maximlevitsky@gmail.com> wrote:
> On Sat, 2010-10-23 at 12:09 +0200, Ohad Ben-Cohen wrote:
...
>> >> diff --git a/drivers/mmc/core/core.c b/drivers/mmc/core/core.c
>> >> index c94565d..515ff39 100644
>> >> --- a/drivers/mmc/core/core.c
>> >> +++ b/drivers/mmc/core/core.c
>> >> @@ -1682,6 +1682,19 @@ int mmc_suspend_host(struct mmc_host *host)
>> >> ? ? ? if (host->bus_ops && !host->bus_dead) {
>> >> ? ? ? ? ? ? ? if (host->bus_ops->suspend)
>> >> ? ? ? ? ? ? ? ? ? ? ? err = host->bus_ops->suspend(host);
>> >> + ? ? ? ? ? ? if (err == -ENOSYS || !host->bus_ops->resume) {

>> > This reintroduces the bug I fixed.
...
>> In this case, mmc_suspend_host() should skip the code you are concerned about.

> You are right here.
> There is a very unlikely case that suspend of sd/mmc card fails (with
> CONFIG_MMC_UNSAFE_RESUME set, and that will trigger the hang, but that
> currently isn't possible because both mmc_sd_suspend and mmc_suspend
> just return 0 unconditionally).

Please note that the card is removed only on -ENOSYS, which is not a
random error that can happen. In this context, it's a deliberate
request (from the underlying bus) to remove the card.

So even if mmc_{sd_}suspend will be changed to propagate an error code
one day, it is unlikely that they will trigger this bug even if they
fail.

> Btw, two above functions are identical, so some refactoring won't hurt.

Feel free to send a patch ;)

Regards,
Ohad.