[PATCH] bcache: prevent crash on changing writeback_running

[PATCH] bcache: prevent crash on changing writeback_running

[Stefan Bader]

commit a664d0f05a2ec02c8f042db536d84d15d6e19e81
    bcache: fix crash on shutdown in passthrough mode

added a safeguard in the shutdown case. At least while not being
attached it is also possible to trigger a kernel bug by writing into
writeback_running. This change  adds the same check before trying to
wake up the thread for that case.

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
---
 drivers/md/bcache/writeback.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h
index 0a9dab1..073a042 100644
--- a/drivers/md/bcache/writeback.h
+++ b/drivers/md/bcache/writeback.h
@@ -63,7 +63,8 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio,
 
 static inline void bch_writeback_queue(struct cached_dev *dc)
 {
-	wake_up_process(dc->writeback_thread);
+	if (!IS_ERR_OR_NULL(dc->writeback_thread))
+		wake_up_process(dc->writeback_thread);
 }
 
 static inline void bch_writeback_add(struct cached_dev *dc)
-- 
1.9.1

Re: [PATCH] bcache: prevent crash on changing writeback_running

[Slava Pestov]

Thanks, this is now in our development branch.

On Tue, Aug 19, 2014 at 6:01 AM, Stefan Bader
<stefan.bader@canonical.com> wrote:
> commit a664d0f05a2ec02c8f042db536d84d15d6e19e81
>     bcache: fix crash on shutdown in passthrough mode
>
> added a safeguard in the shutdown case. At least while not being
> attached it is also possible to trigger a kernel bug by writing into
> writeback_running. This change  adds the same check before trying to
> wake up the thread for that case.
>
> Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
> ---
>  drivers/md/bcache/writeback.h | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h
> index 0a9dab1..073a042 100644
> --- a/drivers/md/bcache/writeback.h
> +++ b/drivers/md/bcache/writeback.h
> @@ -63,7 +63,8 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio,
>
>  static inline void bch_writeback_queue(struct cached_dev *dc)
>  {
> -       wake_up_process(dc->writeback_thread);
> +       if (!IS_ERR_OR_NULL(dc->writeback_thread))
> +               wake_up_process(dc->writeback_thread);
>  }
>
>  static inline void bch_writeback_add(struct cached_dev *dc)
> --
> 1.9.1
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] bcache: prevent crash on changing writeback_running

[Stefan Bader]

[-- Attachment #1: Type: text/plain, Size: 2404 bytes --]

On 19.08.2014 20:03, Slava Pestov wrote:
> Thanks, this is now in our development branch.

Thanks. :) This kind of brings up another question: how are stable patches
handled for bcache. Someone handling them centrally like for the network stack
or tagging for stable and if that missed, individual proposals?

I am asking because it looks like for all kernels between 3.13..3.16 (inclusive)
the set of

a664d0f05a2ec02c8f042db536d84d15d6e19e81
  bcache: fix crash on shutdown in passthrough mode
9e5c353510b26500bd6b8309823ac9ef2837b761
  bcache: fix uninterruptible sleep in writeback thread

plus the one I did would be useful. Right now I only tested with 3.13 with those
and at least avoid the lockdep warnings and crashes I saw before. I only see an
unexplainable load of 1 when idle but having one bcache device configured in a
VM (Xen or KVM). Since writeback should be fixed, maybe gc...

-Stefan
> 
> On Tue, Aug 19, 2014 at 6:01 AM, Stefan Bader
> <stefan.bader@canonical.com> wrote:
>> commit a664d0f05a2ec02c8f042db536d84d15d6e19e81
>>     bcache: fix crash on shutdown in passthrough mode
>>
>> added a safeguard in the shutdown case. At least while not being
>> attached it is also possible to trigger a kernel bug by writing into
>> writeback_running. This change  adds the same check before trying to
>> wake up the thread for that case.
>>
>> Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
>> ---
>>  drivers/md/bcache/writeback.h | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h
>> index 0a9dab1..073a042 100644
>> --- a/drivers/md/bcache/writeback.h
>> +++ b/drivers/md/bcache/writeback.h
>> @@ -63,7 +63,8 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio,
>>
>>  static inline void bch_writeback_queue(struct cached_dev *dc)
>>  {
>> -       wake_up_process(dc->writeback_thread);
>> +       if (!IS_ERR_OR_NULL(dc->writeback_thread))
>> +               wake_up_process(dc->writeback_thread);
>>  }
>>
>>  static inline void bch_writeback_add(struct cached_dev *dc)
>> --
>> 1.9.1
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH] bcache: prevent crash on changing writeback_running

[Stefan Bader]

[-- Attachment #1: Type: text/plain, Size: 1809 bytes --]

On 19.08.2014 20:03, Slava Pestov wrote:
> Thanks, this is now in our development branch.

It has been a while but I just went back there and checked linux and linux-next
but cannot find this change anywhere. I think it is a rather simple bugfix. So I
start to wonder where it went and when it would get pushed.

-Stefan

> 
> On Tue, Aug 19, 2014 at 6:01 AM, Stefan Bader
> <stefan.bader@canonical.com> wrote:
>> commit a664d0f05a2ec02c8f042db536d84d15d6e19e81
>>     bcache: fix crash on shutdown in passthrough mode
>>
>> added a safeguard in the shutdown case. At least while not being
>> attached it is also possible to trigger a kernel bug by writing into
>> writeback_running. This change  adds the same check before trying to
>> wake up the thread for that case.
>>
>> Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
>> ---
>>  drivers/md/bcache/writeback.h | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h
>> index 0a9dab1..073a042 100644
>> --- a/drivers/md/bcache/writeback.h
>> +++ b/drivers/md/bcache/writeback.h
>> @@ -63,7 +63,8 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio,
>>
>>  static inline void bch_writeback_queue(struct cached_dev *dc)
>>  {
>> -       wake_up_process(dc->writeback_thread);
>> +       if (!IS_ERR_OR_NULL(dc->writeback_thread))
>> +               wake_up_process(dc->writeback_thread);
>>  }
>>
>>  static inline void bch_writeback_add(struct cached_dev *dc)
>> --
>> 1.9.1
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]