[PATCH] Btrfs: fix oops while writing data to SSD partitions

[PATCH] Btrfs: fix oops while writing data to SSD partitions

[liubo]

Here I have a two SSD-partitions btrfs, and they are defaultly set to
"data=raid0, metadata=raid1", then I try to fill my btrfs partition
till "No space left on device", via "dd if=/dev/zero of=/mnt/btrfs/tmp".

I get an oops panic from kernel BUG at fs/btrfs/extent-tree.c:5199!, which
refers to find_free_extent's
BUG_ON(index != get_block_group_index(block_group));

In SSD mode, in order to find enough space to alloc, we may check the
block_group cache which has been checked sometime before, but the index is not
updated, where it hits the BUG_ON.

Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
---
 fs/btrfs/extent-tree.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 71cd456..3213c39 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -5065,7 +5065,9 @@ have_block_group:
 			 * group is does point to and try again
 			 */
 			if (!last_ptr_loop && last_ptr->block_group &&
-			    last_ptr->block_group != block_group) {
+			    last_ptr->block_group != block_group &&
+			    index <=
+				 get_block_group_index(last_ptr->block_group)) {
 
 				btrfs_put_block_group(block_group);
 				block_group = last_ptr->block_group;
-- 
1.6.5.2

Re: [PATCH] Btrfs: fix oops while writing data to SSD partitions

[Josef Bacik]

On 07/27/2011 05:49 AM, liubo wrote:
> Here I have a two SSD-partitions btrfs, and they are defaultly set to
> "data=raid0, metadata=raid1", then I try to fill my btrfs partition
> till "No space left on device", via "dd if=/dev/zero of=/mnt/btrfs/tmp".
> 
> I get an oops panic from kernel BUG at fs/btrfs/extent-tree.c:5199!, which
> refers to find_free_extent's
> BUG_ON(index != get_block_group_index(block_group));
> 
> In SSD mode, in order to find enough space to alloc, we may check the
> block_group cache which has been checked sometime before, but the index is not
> updated, where it hits the BUG_ON.
> 
> Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
> ---
>  fs/btrfs/extent-tree.c |    4 +++-
>  1 files changed, 3 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
> index 71cd456..3213c39 100644
> --- a/fs/btrfs/extent-tree.c
> +++ b/fs/btrfs/extent-tree.c
> @@ -5065,7 +5065,9 @@ have_block_group:
>  			 * group is does point to and try again
>  			 */
>  			if (!last_ptr_loop && last_ptr->block_group &&
> -			    last_ptr->block_group != block_group) {
> +			    last_ptr->block_group != block_group &&
> +			    index <=
> +				 get_block_group_index(last_ptr->block_group)) {
>  
>  				btrfs_put_block_group(block_group);
>  				block_group = last_ptr->block_group;

Ugh I wish we could make this prettier but

Acked-by: Josef Bacik <josef@redhat.com>

Thanks,

Josef

Re: [PATCH] Btrfs: fix oops while writing data to SSD partitions

[liubo]

On 07/27/2011 10:26 PM, Josef Bacik wrote:
> On 07/27/2011 05:49 AM, liubo wrote:
>> Here I have a two SSD-partitions btrfs, and they are defaultly set to
>> "data=raid0, metadata=raid1", then I try to fill my btrfs partition
>> till "No space left on device", via "dd if=/dev/zero of=/mnt/btrfs/tmp".
>>
>> I get an oops panic from kernel BUG at fs/btrfs/extent-tree.c:5199!, which
>> refers to find_free_extent's
>> BUG_ON(index != get_block_group_index(block_group));
>>
>> In SSD mode, in order to find enough space to alloc, we may check the
>> block_group cache which has been checked sometime before, but the index is not
>> updated, where it hits the BUG_ON.
>>
>> Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
>> ---
>>  fs/btrfs/extent-tree.c |    4 +++-
>>  1 files changed, 3 insertions(+), 1 deletions(-)
>>
>> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
>> index 71cd456..3213c39 100644
>> --- a/fs/btrfs/extent-tree.c
>> +++ b/fs/btrfs/extent-tree.c
>> @@ -5065,7 +5065,9 @@ have_block_group:
>>  			 * group is does point to and try again
>>  			 */
>>  			if (!last_ptr_loop && last_ptr->block_group &&
>> -			    last_ptr->block_group != block_group) {
>> +			    last_ptr->block_group != block_group &&
>> +			    index <=
>> +				 get_block_group_index(last_ptr->block_group)) {
>>  
>>  				btrfs_put_block_group(block_group);
>>  				block_group = last_ptr->block_group;
> 
> Ugh I wish we could make this prettier but
> 
> Acked-by: Josef Bacik <josef@redhat.com>
> 

I wish that, too. ;)

But lots of goto really make life harder...we can somehow improve it later.

thanks,
liubo

> Thanks,
> 
> Josef
>