diff mbox series

[v1] f2fs: Fix discard bug on zoned block devices with 2MiB zone size

Message ID 20230313094825epcms2p71e6cb549251dc55e8d202dd64b9913a6@epcms2p7
State New
Headers
Series [v1] f2fs: Fix discard bug on zoned block devices with 2MiB zone size |

Commit Message

Yonggil Song March 13, 2023, 9:48 a.m. UTC 
  When using f2fs on a zoned block device with 2MiB zone size, IO errors
occurs because f2fs tries to write data to a zone that has not been reset.

The cause is that f2fs tries to discard multiple zones at once. This is
caused by a condition in f2fs_clear_prefree_segments that does not check
for zoned block devices when setting the discard range. This leads to
invalid reset commands and write pointer mismatches.

This patch fixes the zoned block device with 2MiB zone size to reset one
zone at a time.

Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
---
 fs/f2fs/segment.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

patchwork-bot+f2fs@kernel.org March 21, 2023, 4:40 p.m. UTC | #1 
Hello:

This patch was applied to jaegeuk/f2fs.git (dev)
by Jaegeuk Kim <jaegeuk@kernel.org>:

On Mon, 13 Mar 2023 18:48:25 +0900 you wrote:
> When using f2fs on a zoned block device with 2MiB zone size, IO errors
> occurs because f2fs tries to write data to a zone that has not been reset.
> 
> The cause is that f2fs tries to discard multiple zones at once. This is
> caused by a condition in f2fs_clear_prefree_segments that does not check
> for zoned block devices when setting the discard range. This leads to
> invalid reset commands and write pointer mismatches.
> 
> [...]

Here is the summary with links:
  - [f2fs-dev,v1] f2fs: Fix discard bug on zoned block devices with 2MiB zone size
    https://git.kernel.org/jaegeuk/f2fs/c/5108a09b11d7

You are awesome, thank you!
Chao Yu March 23, 2023, 3:17 p.m. UTC | #2 
On 2023/3/13 17:48, Yonggil Song wrote:
> When using f2fs on a zoned block device with 2MiB zone size, IO errors
> occurs because f2fs tries to write data to a zone that has not been reset.
> 
> The cause is that f2fs tries to discard multiple zones at once. This is
> caused by a condition in f2fs_clear_prefree_segments that does not check
> for zoned block devices when setting the discard range. This leads to
> invalid reset commands and write pointer mismatches.
> 
> This patch fixes the zoned block device with 2MiB zone size to reset one
> zone at a time.
> 
> Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
> ---
>   fs/f2fs/segment.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index acf3d3fa4363..2b6cb6df623b 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
>   					(end - 1) <= cpc->trim_end)
>   				continue;
>   
> -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
> +		if (!f2fs_sb_has_blkzoned(sbi) &&

Could you please add one line comment here for this change?

Otherwise it looks good to me.

Thanks,

> +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
>   			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
>   				(end - start) << sbi->log_blocks_per_seg);
>   			continue;
Jaegeuk Kim March 23, 2023, 10:03 p.m. UTC | #3 
On 03/23, Chao Yu wrote:
> On 2023/3/13 17:48, Yonggil Song wrote:
> > When using f2fs on a zoned block device with 2MiB zone size, IO errors
> > occurs because f2fs tries to write data to a zone that has not been reset.
> > 
> > The cause is that f2fs tries to discard multiple zones at once. This is
> > caused by a condition in f2fs_clear_prefree_segments that does not check
> > for zoned block devices when setting the discard range. This leads to
> > invalid reset commands and write pointer mismatches.
> > 
> > This patch fixes the zoned block device with 2MiB zone size to reset one
> > zone at a time.
> > 
> > Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
> > ---
> >   fs/f2fs/segment.c | 3 ++-
> >   1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> > index acf3d3fa4363..2b6cb6df623b 100644
> > --- a/fs/f2fs/segment.c
> > +++ b/fs/f2fs/segment.c
> > @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
> >   					(end - 1) <= cpc->trim_end)
> >   				continue;
> > -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
> > +		if (!f2fs_sb_has_blkzoned(sbi) &&
> 
> Could you please add one line comment here for this change?

This was merged in -dev a while ago. I don't think this would be critical
to rebase it again.

> 
> Otherwise it looks good to me.
> 
> Thanks,
> 
> > +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
> >   			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
> >   				(end - start) << sbi->log_blocks_per_seg);
> >   			continue;
Chao Yu March 24, 2023, 1:52 a.m. UTC | #4 
On 2023/3/24 6:03, Jaegeuk Kim wrote:
> On 03/23, Chao Yu wrote:
>> On 2023/3/13 17:48, Yonggil Song wrote:
>>> When using f2fs on a zoned block device with 2MiB zone size, IO errors
>>> occurs because f2fs tries to write data to a zone that has not been reset.
>>>
>>> The cause is that f2fs tries to discard multiple zones at once. This is
>>> caused by a condition in f2fs_clear_prefree_segments that does not check
>>> for zoned block devices when setting the discard range. This leads to
>>> invalid reset commands and write pointer mismatches.
>>>
>>> This patch fixes the zoned block device with 2MiB zone size to reset one
>>> zone at a time.
>>>
>>> Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
>>> ---
>>>    fs/f2fs/segment.c | 3 ++-
>>>    1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>>> index acf3d3fa4363..2b6cb6df623b 100644
>>> --- a/fs/f2fs/segment.c
>>> +++ b/fs/f2fs/segment.c
>>> @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
>>>    					(end - 1) <= cpc->trim_end)
>>>    				continue;
>>> -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
>>> +		if (!f2fs_sb_has_blkzoned(sbi) &&
>>
>> Could you please add one line comment here for this change?
> 
> This was merged in -dev a while ago. I don't think this would be critical
> to rebase it again.

Yes, it's not critical, fine to me.

Thanks,

> 
>>
>> Otherwise it looks good to me.
>>
>> Thanks,
>>
>>> +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
>>>    			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
>>>    				(end - start) << sbi->log_blocks_per_seg);
>>>    			continue;
Jaegeuk Kim March 24, 2023, 4:54 p.m. UTC | #5 
On 03/24, Chao Yu wrote:
> On 2023/3/24 6:03, Jaegeuk Kim wrote:
> > On 03/23, Chao Yu wrote:
> > > On 2023/3/13 17:48, Yonggil Song wrote:
> > > > When using f2fs on a zoned block device with 2MiB zone size, IO errors
> > > > occurs because f2fs tries to write data to a zone that has not been reset.
> > > > 
> > > > The cause is that f2fs tries to discard multiple zones at once. This is
> > > > caused by a condition in f2fs_clear_prefree_segments that does not check
> > > > for zoned block devices when setting the discard range. This leads to
> > > > invalid reset commands and write pointer mismatches.
> > > > 
> > > > This patch fixes the zoned block device with 2MiB zone size to reset one
> > > > zone at a time.
> > > > 
> > > > Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
> > > > ---
> > > >    fs/f2fs/segment.c | 3 ++-
> > > >    1 file changed, 2 insertions(+), 1 deletion(-)
> > > > 
> > > > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> > > > index acf3d3fa4363..2b6cb6df623b 100644
> > > > --- a/fs/f2fs/segment.c
> > > > +++ b/fs/f2fs/segment.c
> > > > @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
> > > >    					(end - 1) <= cpc->trim_end)
> > > >    				continue;
> > > > -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
> > > > +		if (!f2fs_sb_has_blkzoned(sbi) &&
> > > 
> > > Could you please add one line comment here for this change?
> > 
> > This was merged in -dev a while ago. I don't think this would be critical
> > to rebase it again.
> 
> Yes, it's not critical, fine to me.

Added:

/* Should cover 2MB zoned device for zone-based reset */

So lucky since I had to rebase to fix other patch. :(

"f2fs: factor out discard_cmd usage from general rb_tree use"


> 
> Thanks,
> 
> > 
> > > 
> > > Otherwise it looks good to me.
> > > 
> > > Thanks,
> > > 
> > > > +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
> > > >    			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
> > > >    				(end - start) << sbi->log_blocks_per_seg);
> > > >    			continue;
Chao Yu March 25, 2023, 7:25 a.m. UTC | #6 
On 2023/3/25 0:54, Jaegeuk Kim wrote:
> On 03/24, Chao Yu wrote:
>> On 2023/3/24 6:03, Jaegeuk Kim wrote:
>>> On 03/23, Chao Yu wrote:
>>>> On 2023/3/13 17:48, Yonggil Song wrote:
>>>>> When using f2fs on a zoned block device with 2MiB zone size, IO errors
>>>>> occurs because f2fs tries to write data to a zone that has not been reset.
>>>>>
>>>>> The cause is that f2fs tries to discard multiple zones at once. This is
>>>>> caused by a condition in f2fs_clear_prefree_segments that does not check
>>>>> for zoned block devices when setting the discard range. This leads to
>>>>> invalid reset commands and write pointer mismatches.
>>>>>
>>>>> This patch fixes the zoned block device with 2MiB zone size to reset one
>>>>> zone at a time.
>>>>>
>>>>> Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
>>>>> ---
>>>>>     fs/f2fs/segment.c | 3 ++-
>>>>>     1 file changed, 2 insertions(+), 1 deletion(-)
>>>>>
>>>>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>>>>> index acf3d3fa4363..2b6cb6df623b 100644
>>>>> --- a/fs/f2fs/segment.c
>>>>> +++ b/fs/f2fs/segment.c
>>>>> @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
>>>>>     					(end - 1) <= cpc->trim_end)
>>>>>     				continue;
>>>>> -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
>>>>> +		if (!f2fs_sb_has_blkzoned(sbi) &&
>>>>
>>>> Could you please add one line comment here for this change?
>>>
>>> This was merged in -dev a while ago. I don't think this would be critical
>>> to rebase it again.
>>
>> Yes, it's not critical, fine to me.
> 
> Added:
> 
> /* Should cover 2MB zoned device for zone-based reset */

Thanks a lot for the change. :)

Thanks,

> 
> So lucky since I had to rebase to fix other patch. :(
> 
> "f2fs: factor out discard_cmd usage from general rb_tree use"
> 
> 
>>
>> Thanks,
>>
>>>
>>>>
>>>> Otherwise it looks good to me.
>>>>
>>>> Thanks,
>>>>
>>>>> +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
>>>>>     			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
>>>>>     				(end - start) << sbi->log_blocks_per_seg);
>>>>>     			continue;
Chao Yu March 26, 2023, 3:49 a.m. UTC | #7 
On 2023/3/25 0:54, Jaegeuk Kim wrote:
> On 03/24, Chao Yu wrote:
>> On 2023/3/24 6:03, Jaegeuk Kim wrote:
>>> On 03/23, Chao Yu wrote:
>>>> On 2023/3/13 17:48, Yonggil Song wrote:
>>>>> When using f2fs on a zoned block device with 2MiB zone size, IO errors
>>>>> occurs because f2fs tries to write data to a zone that has not been reset.
>>>>>
>>>>> The cause is that f2fs tries to discard multiple zones at once. This is
>>>>> caused by a condition in f2fs_clear_prefree_segments that does not check
>>>>> for zoned block devices when setting the discard range. This leads to
>>>>> invalid reset commands and write pointer mismatches.
>>>>>
>>>>> This patch fixes the zoned block device with 2MiB zone size to reset one
>>>>> zone at a time.
>>>>>
>>>>> Signed-off-by: Yonggil Song <yonggil.song@samsung.com>
>>>>> ---
>>>>>     fs/f2fs/segment.c | 3 ++-
>>>>>     1 file changed, 2 insertions(+), 1 deletion(-)
>>>>>
>>>>> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
>>>>> index acf3d3fa4363..2b6cb6df623b 100644
>>>>> --- a/fs/f2fs/segment.c
>>>>> +++ b/fs/f2fs/segment.c
>>>>> @@ -1953,7 +1953,8 @@ void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
>>>>>     					(end - 1) <= cpc->trim_end)
>>>>>     				continue;
>>>>> -		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
>>>>> +		if (!f2fs_sb_has_blkzoned(sbi) &&
>>>>
>>>> Could you please add one line comment here for this change?
>>>
>>> This was merged in -dev a while ago. I don't think this would be critical
>>> to rebase it again.
>>
>> Yes, it's not critical, fine to me.
> 
> Added:
> 
> /* Should cover 2MB zoned device for zone-based reset */

Reviewed-by: Chao Yu <chao@kernel.org>

Thanks,

> 
> So lucky since I had to rebase to fix other patch. :(
> 
> "f2fs: factor out discard_cmd usage from general rb_tree use"
> 
> 
>>
>> Thanks,
>>
>>>
>>>>
>>>> Otherwise it looks good to me.
>>>>
>>>> Thanks,
>>>>
>>>>> +		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
>>>>>     			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
>>>>>     				(end - start) << sbi->log_blocks_per_seg);
>>>>>     			continue;
diff mbox series

Patch

diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index acf3d3fa4363..2b6cb6df623b 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -1953,7 +1953,8 @@  void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
 					(end - 1) <= cpc->trim_end)
 				continue;
 
-		if (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi)) {
+		if (!f2fs_sb_has_blkzoned(sbi) &&
+		    (!f2fs_lfs_mode(sbi) || !__is_large_section(sbi))) {
 			f2fs_issue_discard(sbi, START_BLOCK(sbi, start),
 				(end - start) << sbi->log_blocks_per_seg);
 			continue;