[PATCH] btrfs: not a disk error if the bio_add_page fails

[PATCH] btrfs: not a disk error if the bio_add_page fails

[Anand Jain]

bio_add_page() can fail for logical reasons as from the bio_add_page()
comments:- 'This will only fail if either
bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
write error statistics for this. And set -EINVAL instead of -EIO.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
 fs/btrfs/scrub.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index d766c73eb29a..8c1508a41f90 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -4618,15 +4618,15 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
 	ret = bio_add_page(bio, page, PAGE_SIZE, 0);
 	if (ret != PAGE_SIZE) {
-leave_with_eio:
 		bio_put(bio);
+		return -EINVAL;
+	}
+
+	if (btrfsic_submit_bio_wait(bio)) {
 		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
 		return -EIO;
 	}
 
-	if (btrfsic_submit_bio_wait(bio))
-		goto leave_with_eio;
-
 	bio_put(bio);
 	return 0;
 }
-- 
2.15.0

Re: [PATCH] btrfs: not a disk error if the bio_add_page fails

[Filipe Manana]

On Wed, Jan 3, 2018 at 8:07 AM, Anand Jain <anand.jain@oracle.com> wrote:
> bio_add_page() can fail for logical reasons as from the bio_add_page()
> comments:- 'This will only fail if either
> bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
> write error statistics for this. And set -EINVAL instead of -EIO.
>
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
>  fs/btrfs/scrub.c | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
> index d766c73eb29a..8c1508a41f90 100644
> --- a/fs/btrfs/scrub.c
> +++ b/fs/btrfs/scrub.c
> @@ -4618,15 +4618,15 @@ static int write_page_nocow(struct scrub_ctx *sctx,
>         bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
>         ret = bio_add_page(bio, page, PAGE_SIZE, 0);
>         if (ret != PAGE_SIZE) {
> -leave_with_eio:
>                 bio_put(bio);
> +               return -EINVAL;
> +       }
> +
> +       if (btrfsic_submit_bio_wait(bio)) {
>                 btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
>                 return -EIO;
>         }

You are now introducing a leak of the bio. Missing bio_put(bio) before
returning.

>
> -       if (btrfsic_submit_bio_wait(bio))
> -               goto leave_with_eio;
> -
>         bio_put(bio);
>         return 0;
>  }
> --
> 2.15.0
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html



-- 
Filipe David Manana,

“Whether you think you can, or you think you can't — you're right.”

[PATCH v2] btrfs: not a disk error if the bio_add_page fails

[Anand Jain]

bio_add_page() can fail for logical reasons as from the bio_add_page()
comments:- 'This will only fail if either
bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
write error statistics for this. And set -EINVAL instead of -EIO.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
---
v1->v2: Add missing bio_put(). Thanks Filipe.

 fs/btrfs/scrub.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
index ecfe3118d9dd..57ac39a3f046 100644
--- a/fs/btrfs/scrub.c
+++ b/fs/btrfs/scrub.c
@@ -4617,15 +4617,16 @@ static int write_page_nocow(struct scrub_ctx *sctx,
 	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
 	ret = bio_add_page(bio, page, PAGE_SIZE, 0);
 	if (ret != PAGE_SIZE) {
-leave_with_eio:
+		bio_put(bio);
+		return -EINVAL;
+	}
+
+	if (btrfsic_submit_bio_wait(bio)) {
 		bio_put(bio);
 		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
 		return -EIO;
 	}
 
-	if (btrfsic_submit_bio_wait(bio))
-		goto leave_with_eio;
-
 	bio_put(bio);
 	return 0;
 }
-- 
2.15.0

Re: [PATCH] btrfs: not a disk error if the bio_add_page fails

[Anand Jain]

On 01/03/2018 07:34 PM, Filipe Manana wrote:
> On Wed, Jan 3, 2018 at 8:07 AM, Anand Jain <anand.jain@oracle.com> wrote:
>> bio_add_page() can fail for logical reasons as from the bio_add_page()
>> comments:- 'This will only fail if either
>> bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
>> write error statistics for this. And set -EINVAL instead of -EIO.
>>
>> Signed-off-by: Anand Jain <anand.jain@oracle.com>
>> ---
>>   fs/btrfs/scrub.c | 8 ++++----
>>   1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
>> index d766c73eb29a..8c1508a41f90 100644
>> --- a/fs/btrfs/scrub.c
>> +++ b/fs/btrfs/scrub.c
>> @@ -4618,15 +4618,15 @@ static int write_page_nocow(struct scrub_ctx *sctx,
>>          bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
>>          ret = bio_add_page(bio, page, PAGE_SIZE, 0);
>>          if (ret != PAGE_SIZE) {
>> -leave_with_eio:
>>                  bio_put(bio);
>> +               return -EINVAL;
>> +       }
>> +
>> +       if (btrfsic_submit_bio_wait(bio)) {
>>                  btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
>>                  return -EIO;
>>          }
> 
> You are now introducing a leak of the bio. Missing bio_put(bio) before
> returning.

  Oops. My bad. Thanks. Now fixed in v2.

- Anand

>>
>> -       if (btrfsic_submit_bio_wait(bio))
>> -               goto leave_with_eio;
>> -
>>          bio_put(bio);
>>          return 0;
>>   }
>> --
>> 2.15.0
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 
> 

Re: [PATCH v2] btrfs: not a disk error if the bio_add_page fails

[Liu Bo]

On Wed, Jan 03, 2018 at 09:52:43PM +0800, Anand Jain wrote:
> bio_add_page() can fail for logical reasons as from the bio_add_page()
> comments:- 'This will only fail if either
> bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
> write error statistics for this. And set -EINVAL instead of -EIO.
>

It's correct to skip increasing counter, but why -EINVAL?
(its caller only cares about if it's non-zero.)

thanks,

-liubo
> Signed-off-by: Anand Jain <anand.jain@oracle.com>
> ---
> v1->v2: Add missing bio_put(). Thanks Filipe.
> 
>  fs/btrfs/scrub.c | 9 +++++----
>  1 file changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
> index ecfe3118d9dd..57ac39a3f046 100644
> --- a/fs/btrfs/scrub.c
> +++ b/fs/btrfs/scrub.c
> @@ -4617,15 +4617,16 @@ static int write_page_nocow(struct scrub_ctx *sctx,
>  	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
>  	ret = bio_add_page(bio, page, PAGE_SIZE, 0);
>  	if (ret != PAGE_SIZE) {
> -leave_with_eio:
> +		bio_put(bio);
> +		return -EINVAL;
> +	}
> +
> +	if (btrfsic_submit_bio_wait(bio)) {
>  		bio_put(bio);
>  		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
>  		return -EIO;
>  	}
>  
> -	if (btrfsic_submit_bio_wait(bio))
> -		goto leave_with_eio;
> -
>  	bio_put(bio);
>  	return 0;
>  }
> -- 
> 2.15.0
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] btrfs: not a disk error if the bio_add_page fails

[Liu Bo]

On Thu, Jan 04, 2018 at 09:47:56AM -0800, Liu Bo wrote:
> On Wed, Jan 03, 2018 at 09:52:43PM +0800, Anand Jain wrote:
> > bio_add_page() can fail for logical reasons as from the bio_add_page()
> > comments:- 'This will only fail if either
> > bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
> > write error statistics for this. And set -EINVAL instead of -EIO.
> >
> 
> It's correct to skip increasing counter, but why -EINVAL?
> (its caller only cares about if it's non-zero.)

Well...there is no chance that bio_add_page() could return
non-PAGE_SIZE as this bio has been just created, we can remove the
error handling after bio_add_page().

thanks,
-liubo
> 
> thanks,
> 
> -liubo
> > Signed-off-by: Anand Jain <anand.jain@oracle.com>
> > ---
> > v1->v2: Add missing bio_put(). Thanks Filipe.
> > 
> >  fs/btrfs/scrub.c | 9 +++++----
> >  1 file changed, 5 insertions(+), 4 deletions(-)
> > 
> > diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
> > index ecfe3118d9dd..57ac39a3f046 100644
> > --- a/fs/btrfs/scrub.c
> > +++ b/fs/btrfs/scrub.c
> > @@ -4617,15 +4617,16 @@ static int write_page_nocow(struct scrub_ctx *sctx,
> >  	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
> >  	ret = bio_add_page(bio, page, PAGE_SIZE, 0);
> >  	if (ret != PAGE_SIZE) {
> > -leave_with_eio:
> > +		bio_put(bio);
> > +		return -EINVAL;
> > +	}
> > +
> > +	if (btrfsic_submit_bio_wait(bio)) {
> >  		bio_put(bio);
> >  		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
> >  		return -EIO;
> >  	}
> >  
> > -	if (btrfsic_submit_bio_wait(bio))
> > -		goto leave_with_eio;
> > -
> >  	bio_put(bio);
> >  	return 0;
> >  }
> > -- 
> > 2.15.0
> > 
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] btrfs: not a disk error if the bio_add_page fails

[Anand Jain]

On 01/05/2018 02:50 AM, Liu Bo wrote:
> On Thu, Jan 04, 2018 at 09:47:56AM -0800, Liu Bo wrote:
>> On Wed, Jan 03, 2018 at 09:52:43PM +0800, Anand Jain wrote:
>>> bio_add_page() can fail for logical reasons as from the bio_add_page()
>>> comments:- 'This will only fail if either
>>> bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.' Don't inc the
>>> write error statistics for this. And set -EINVAL instead of -EIO.
>>>
>>
>> It's correct to skip increasing counter, but why -EINVAL?
>> (its caller only cares about if it's non-zero.)
> 
> Well...there is no chance that bio_add_page() could return
> non-PAGE_SIZE as this bio has been just created, we can remove the
> error handling after bio_add_page().

  Ah. right, bio_add_page() can't fail in this thread, its not
  a bio clone, and its the first call to bio_add_page(). Will
  remove error handling part.

Thanks, Anand

> thanks,
> -liubo
>>
>> thanks,
>>
>> -liubo
>>> Signed-off-by: Anand Jain <anand.jain@oracle.com>
>>> ---
>>> v1->v2: Add missing bio_put(). Thanks Filipe.
>>>
>>>   fs/btrfs/scrub.c | 9 +++++----
>>>   1 file changed, 5 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c
>>> index ecfe3118d9dd..57ac39a3f046 100644
>>> --- a/fs/btrfs/scrub.c
>>> +++ b/fs/btrfs/scrub.c
>>> @@ -4617,15 +4617,16 @@ static int write_page_nocow(struct scrub_ctx *sctx,
>>>   	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC;
>>>   	ret = bio_add_page(bio, page, PAGE_SIZE, 0);
>>>   	if (ret != PAGE_SIZE) {
>>> -leave_with_eio:
>>> +		bio_put(bio);
>>> +		return -EINVAL;
>>> +	}
>>> +
>>> +	if (btrfsic_submit_bio_wait(bio)) {
>>>   		bio_put(bio);
>>>   		btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS);
>>>   		return -EIO;
>>>   	}
>>>   
>>> -	if (btrfsic_submit_bio_wait(bio))
>>> -		goto leave_with_eio;
>>> -
>>>   	bio_put(bio);
>>>   	return 0;
>>>   }
>>> -- 
>>> 2.15.0
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>