[PATCH] mmc: sdhci: Ignore unexpected CARD_INT interrupts

[PATCH] mmc: sdhci: Ignore unexpected CARD_INT interrupts

[Gabriel Krisman Bertazi]

Hi Greg,

The following patch prevents an illegal command from getting sent to SD
cards, which in turn avoids a lock-up during boot.  Can you please apply
to the 4.9 and 4.4 stable trees?

The patch was generated against the 4.9 and there will be a trivial
context conflict when applying to the 4.4.y branch.  If you prefer, I
can resend another version against that branch.

Thanks!

-- >8 --

commit 5b19196de18e1522630326c80a40a64181c78a1c upstream.

One of our kernelCI boxes hanged at boot because a faulty eSDHC device
was triggering spurious CARD_INT interrupts for SD cards, causing CMD52
reads, which are not allowed for SD devices.  This adds a sanity check
to the interruption path, preventing that illegal command from getting
sent if the CARD_INT interruption should be disabled.

This quirk allows that particular machine to resume boot despite the
faulty hardware, instead of getting hung dealing with thousands of
mishandled interrupts.

Suggested-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Gabriel Krisman Bertazi <krisman@collabora.co.uk>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
(cherry picked from commit 5b19196de18e1522630326c80a40a64181c78a1c)
---
 drivers/mmc/host/sdhci.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c
index e1e274a0a34f..ba637ff8aa7e 100644
--- a/drivers/mmc/host/sdhci.c
+++ b/drivers/mmc/host/sdhci.c
@@ -2719,7 +2719,8 @@ static irqreturn_t sdhci_irq(int irq, void *dev_id)
 		if (intmask & SDHCI_INT_RETUNE)
 			mmc_retune_needed(host->mmc);
 
-		if (intmask & SDHCI_INT_CARD_INT) {
+		if ((intmask & SDHCI_INT_CARD_INT) &&
+		    (host->ier & SDHCI_INT_CARD_INT)) {
 			sdhci_enable_sdio_irq_nolock(host, false);
 			host->thread_isr |= SDHCI_INT_CARD_INT;
 			result = IRQ_WAKE_THREAD;
-- 
2.11.0

Re: [PATCH] mmc: sdhci: Ignore unexpected CARD_INT interrupts

[Greg KH]

On Wed, Jan 25, 2017 at 11:23:35PM -0200, Gabriel Krisman Bertazi wrote:
> Hi Greg,
> 
> The following patch prevents an illegal command from getting sent to SD
> cards, which in turn avoids a lock-up during boot.  Can you please apply
> to the 4.9 and 4.4 stable trees?
> 
> The patch was generated against the 4.9 and there will be a trivial
> context conflict when applying to the 4.4.y branch.  If you prefer, I
> can resend another version against that branch.
> 
> Thanks!
> 
> -- >8 --
> 
> commit 5b19196de18e1522630326c80a40a64181c78a1c upstream.

I don't see this commit in Linus's tree, are you sure you got it
correct?

> One of our kernelCI boxes hanged at boot because a faulty eSDHC device
> was triggering spurious CARD_INT interrupts for SD cards, causing CMD52
> reads, which are not allowed for SD devices.  This adds a sanity check
> to the interruption path, preventing that illegal command from getting
> sent if the CARD_INT interruption should be disabled.
> 
> This quirk allows that particular machine to resume boot despite the
> faulty hardware, instead of getting hung dealing with thousands of
> mishandled interrupts.
> 
> Suggested-by: Adrian Hunter <adrian.hunter@intel.com>
> Signed-off-by: Gabriel Krisman Bertazi <krisman@collabora.co.uk>
> Acked-by: Adrian Hunter <adrian.hunter@intel.com>
> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
> (cherry picked from commit 5b19196de18e1522630326c80a40a64181c78a1c)
> ---
>  drivers/mmc/host/sdhci.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)

Are you sure this is in Linus's tree?  I can't find it anywhere.

You did read Documentation/stable_kernel_rules.txt, right?

thanks,

greg k-h

Re: [PATCH] mmc: sdhci: Ignore unexpected CARD_INT interrupts

[Gabriel Krisman Bertazi]

Greg KH <gregkh@linuxfoundation.org> writes:

>> commit 5b19196de18e1522630326c80a40a64181c78a1c upstream.
>
> I don't see this commit in Linus's tree, are you sure you got it
> correct?

Hi Greg,

Sorry, I got confused.  The patch is not in Linus tree yet, it was only
accepted in Ulf Hansson's tree for the next merge window, which goes
against stable rules.

I'll send you a ping with new version after the merge window.

-- 
Gabriel Krisman Bertazi