diff mbox series

[04/21] ext4: get correct ext4_group_info in ext4_mb_prefetch_fini

Message ID 20230209194825.511043-5-shikemeng@huaweicloud.com
State New
Headers
Series Some bugfix and cleanup to mballoc |

Commit Message

Kemeng Shi Feb. 9, 2023, 7:48 p.m. UTC 
  We always get ext4_group_desc with group + 1 and ext4_group_info with
group to check if we need do initialize ext4_group_info for the group.
Just get ext4_group_desc with group for ext4_group_info initialization
check.

Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
---
 fs/ext4/mballoc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

Comments

Ojaswin Mujoo Feb. 13, 2023, 7:03 a.m. UTC | #1 
On Fri, Feb 10, 2023 at 03:48:08AM +0800, Kemeng Shi wrote:
> We always get ext4_group_desc with group + 1 and ext4_group_info with
> group to check if we need do initialize ext4_group_info for the group.
> Just get ext4_group_desc with group for ext4_group_info initialization
> check.
> 
> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
> ---
>  fs/ext4/mballoc.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index 352ac9139fee..f24f80ecf318 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
>  			   unsigned int nr)
>  {
>  	while (nr-- > 0) {
> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
> -								  NULL);
> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
> +		struct ext4_group_desc *gdp;
> +		struct ext4_group_info *grp;
>  
>  		if (!group)
>  			group = ext4_get_groups_count(sb);
>  		group--;
> +		gdp = ext4_get_group_desc(sb, group, NULL);
>  		grp = ext4_get_group_info(sb, group);
>  
>  		if (EXT4_MB_GRP_NEED_INIT(grp) &&
> -- 
> 2.30.0
> 

This is a duplicate of [1] :)

But I'm okay with this going in as that patchseries might take some time
to get merged. Feel free to add:

Reviewed-by: Ojaswin Mujoo <ojaswin@linux.ibm.com> 

[1] https://lore.kernel.org/r/85bbcb3774e38de65b737ef0000241ddbdda73aa.1674822311.git.ojaswin@linux.ibm.com
Kemeng Shi Feb. 13, 2023, 12:27 p.m. UTC | #2 
on 2/13/2023 3:03 PM, Ojaswin Mujoo wrote:
> On Fri, Feb 10, 2023 at 03:48:08AM +0800, Kemeng Shi wrote:
>> We always get ext4_group_desc with group + 1 and ext4_group_info with
>> group to check if we need do initialize ext4_group_info for the group.
>> Just get ext4_group_desc with group for ext4_group_info initialization
>> check.
>>
>> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
>> ---
>>  fs/ext4/mballoc.c | 6 +++---
>>  1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
>> index 352ac9139fee..f24f80ecf318 100644
>> --- a/fs/ext4/mballoc.c
>> +++ b/fs/ext4/mballoc.c
>> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
>>  			   unsigned int nr)
>>  {
>>  	while (nr-- > 0) {
>> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
>> -								  NULL);
>> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
>> +		struct ext4_group_desc *gdp;
>> +		struct ext4_group_info *grp;
>>  
>>  		if (!group)
>>  			group = ext4_get_groups_count(sb);
>>  		group--;
>> +		gdp = ext4_get_group_desc(sb, group, NULL);
>>  		grp = ext4_get_group_info(sb, group);
>>  
>>  		if (EXT4_MB_GRP_NEED_INIT(grp) &&
>> -- 
>> 2.30.0
>>
> 
> This is a duplicate of [1] :)
> 
> But I'm okay with this going in as that patchseries might take some time
> to get merged. Feel free to add:
> 
> Reviewed-by: Ojaswin Mujoo <ojaswin@linux.ibm.com> 
> 
> [1] https://lore.kernel.org/r/85bbcb3774e38de65b737ef0000241ddbdda73aa.1674822311.git.ojaswin@linux.ibm.com
> 
Hi Ojaswin, Thank you so much to review my code. I 'm sorry that I didn't
notice this patch is duplicated and I really appreciate that you allow this
one to get merged first. I will add your sign to this patch in next version.
Thanks!
Ojaswin Mujoo Feb. 13, 2023, 8:14 p.m. UTC | #3 
On Mon, Feb 13, 2023 at 08:27:32PM +0800, Kemeng Shi wrote:
> 
> 
> on 2/13/2023 3:03 PM, Ojaswin Mujoo wrote:
> > On Fri, Feb 10, 2023 at 03:48:08AM +0800, Kemeng Shi wrote:
> >> We always get ext4_group_desc with group + 1 and ext4_group_info with
> >> group to check if we need do initialize ext4_group_info for the group.
> >> Just get ext4_group_desc with group for ext4_group_info initialization
> >> check.
> >>
> >> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
> >> ---
> >>  fs/ext4/mballoc.c | 6 +++---
> >>  1 file changed, 3 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> >> index 352ac9139fee..f24f80ecf318 100644
> >> --- a/fs/ext4/mballoc.c
> >> +++ b/fs/ext4/mballoc.c
> >> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
> >>  			   unsigned int nr)
> >>  {
> >>  	while (nr-- > 0) {
> >> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
> >> -								  NULL);
> >> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
> >> +		struct ext4_group_desc *gdp;
> >> +		struct ext4_group_info *grp;
> >>  
> >>  		if (!group)
> >>  			group = ext4_get_groups_count(sb);
> >>  		group--;
> >> +		gdp = ext4_get_group_desc(sb, group, NULL);
> >>  		grp = ext4_get_group_info(sb, group);
> >>  
> >>  		if (EXT4_MB_GRP_NEED_INIT(grp) &&
> >> -- 
> >> 2.30.0
> >>
> > 
> > This is a duplicate of [1] :)
> > 
> > But I'm okay with this going in as that patchseries might take some time
> > to get merged. Feel free to add:
> > 
> > Reviewed-by: Ojaswin Mujoo <ojaswin@linux.ibm.com> 
> > 
> > [1] https://lore.kernel.org/r/85bbcb3774e38de65b737ef0000241ddbdda73aa.1674822311.git.ojaswin@linux.ibm.com
> > 
> Hi Ojaswin, Thank you so much to review my code. I 'm sorry that I didn't
> notice this patch is duplicated and I really appreciate that you allow this
> one to get merged first. I will add your sign to this patch in next version.
> Thanks!
Hi Kemeng,

So I'm not sure what the norm is when dealing with such duplicate
patches, but if you do plan to add the Signed-off-by then I'd just like
to point out that the patch I linked is mainly from Ritesh Harjani, so
you can pick his Signed-off-by rather than mine.

Thanks,
ojaswin
> 
> -- 
> Best wishes
> Kemeng Shi
>
Kemeng Shi Feb. 14, 2023, 1:12 a.m. UTC | #4 
on 2/14/2023 4:14 AM, Ojaswin Mujoo wrote:
> On Mon, Feb 13, 2023 at 08:27:32PM +0800, Kemeng Shi wrote:
>>
>>
>> on 2/13/2023 3:03 PM, Ojaswin Mujoo wrote:
>>> On Fri, Feb 10, 2023 at 03:48:08AM +0800, Kemeng Shi wrote:
>>>> We always get ext4_group_desc with group + 1 and ext4_group_info with
>>>> group to check if we need do initialize ext4_group_info for the group.
>>>> Just get ext4_group_desc with group for ext4_group_info initialization
>>>> check.
>>>>
>>>> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
>>>> ---
>>>>  fs/ext4/mballoc.c | 6 +++---
>>>>  1 file changed, 3 insertions(+), 3 deletions(-)
>>>>
>>>> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
>>>> index 352ac9139fee..f24f80ecf318 100644
>>>> --- a/fs/ext4/mballoc.c
>>>> +++ b/fs/ext4/mballoc.c
>>>> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
>>>>  			   unsigned int nr)
>>>>  {
>>>>  	while (nr-- > 0) {
>>>> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
>>>> -								  NULL);
>>>> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
>>>> +		struct ext4_group_desc *gdp;
>>>> +		struct ext4_group_info *grp;
>>>>  
>>>>  		if (!group)
>>>>  			group = ext4_get_groups_count(sb);
>>>>  		group--;
>>>> +		gdp = ext4_get_group_desc(sb, group, NULL);
>>>>  		grp = ext4_get_group_info(sb, group);
>>>>  
>>>>  		if (EXT4_MB_GRP_NEED_INIT(grp) &&
>>>> -- 
>>>> 2.30.0
>>>>
>>>
>>> This is a duplicate of [1] :)
>>>
>>> But I'm okay with this going in as that patchseries might take some time
>>> to get merged. Feel free to add:
>>>
>>> Reviewed-by: Ojaswin Mujoo <ojaswin@linux.ibm.com> 
>>>
>>> [1] https://lore.kernel.org/r/85bbcb3774e38de65b737ef0000241ddbdda73aa.1674822311.git.ojaswin@linux.ibm.com
>>>
>> Hi Ojaswin, Thank you so much to review my code. I 'm sorry that I didn't
>> notice this patch is duplicated and I really appreciate that you allow this
>> one to get merged first. I will add your sign to this patch in next version.
>> Thanks!
> Hi Kemeng,
> 
> So I'm not sure what the norm is when dealing with such duplicate
> patches, but if you do plan to add the Signed-off-by then I'd just like
> to point out that the patch I linked is mainly from Ritesh Harjani, so
> you can pick his Signed-off-by rather than mine.
> 
Sorry that I miss that there are two Signed-off-bys in patch you have already
sent. I will collect both signs to this patch. Thanks!
Ritesh Harjani (IBM) Feb. 17, 2023, 6:46 a.m. UTC | #5 
Kemeng Shi <shikemeng@huaweicloud.com> writes:

> We always get ext4_group_desc with group + 1 and ext4_group_info with
> group to check if we need do initialize ext4_group_info for the group.
> Just get ext4_group_desc with group for ext4_group_info initialization
> check.
>
> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
> ---
>  fs/ext4/mballoc.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index 352ac9139fee..f24f80ecf318 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
>  			   unsigned int nr)
>  {
>  	while (nr-- > 0) {
> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
> -								  NULL);
> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
> +		struct ext4_group_desc *gdp;
> +		struct ext4_group_info *grp;

We can even declare these variables at the begining of the function like
in [1]. Also I would advise to rearrange any "fixes" patches at the
begining of the patch series and "cleanup" patches at the end.
e.g. this looks like a fix to me.

That way it is sometimes easier for people to cherry-pick any fixes if
required in their older kernel trees. ;)

[1]: https://lore.kernel.org/all/85bbcb3774e38de65b737ef0000241ddbdda73aa.1674822311.git.ojaswin@linux.ibm.com/

-ritesh

>
>  		if (!group)
>  			group = ext4_get_groups_count(sb);
>  		group--;
> +		gdp = ext4_get_group_desc(sb, group, NULL);
>  		grp = ext4_get_group_info(sb, group);
>
>  		if (EXT4_MB_GRP_NEED_INIT(grp) &&
> --
> 2.30.0
Kemeng Shi Feb. 17, 2023, 7:19 a.m. UTC | #6 
on 2/17/2023 2:46 PM, Ritesh Harjani (IBM) wrote:
> Kemeng Shi <shikemeng@huaweicloud.com> writes:
> 
>> We always get ext4_group_desc with group + 1 and ext4_group_info with
>> group to check if we need do initialize ext4_group_info for the group.
>> Just get ext4_group_desc with group for ext4_group_info initialization
>> check.
>>
>> Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
>> ---
>>  fs/ext4/mballoc.c | 6 +++---
>>  1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
>> index 352ac9139fee..f24f80ecf318 100644
>> --- a/fs/ext4/mballoc.c
>> +++ b/fs/ext4/mballoc.c
>> @@ -2570,13 +2570,13 @@ void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
>>  			   unsigned int nr)
>>  {
>>  	while (nr-- > 0) {
>> -		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
>> -								  NULL);
>> -		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
>> +		struct ext4_group_desc *gdp;
>> +		struct ext4_group_info *grp;
> 
> We can even declare these variables at the begining of the function like
> in [1]. Also I would advise to rearrange any "fixes" patches at the
> begining of the patch series and "cleanup" patches at the end.
> e.g. this looks like a fix to me.
> 
> That way it is sometimes easier for people to cherry-pick any fixes if
> required in their older kernel trees. ;)
> 
Hi Ritesh, Thanks for feedback. I declare these variables at the begining
of the function in next version.
I agree that we should keep bugfix patches at the beginning. Actually,
patch 1-8 are all fix patches from my view. So I think current patch sort
is fine.
Thanks.
diff mbox series

Patch

diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 352ac9139fee..f24f80ecf318 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -2570,13 +2570,13 @@  void ext4_mb_prefetch_fini(struct super_block *sb, ext4_group_t group,
 			   unsigned int nr)
 {
 	while (nr-- > 0) {
-		struct ext4_group_desc *gdp = ext4_get_group_desc(sb, group,
-								  NULL);
-		struct ext4_group_info *grp = ext4_get_group_info(sb, group);
+		struct ext4_group_desc *gdp;
+		struct ext4_group_info *grp;
 
 		if (!group)
 			group = ext4_get_groups_count(sb);
 		group--;
+		gdp = ext4_get_group_desc(sb, group, NULL);
 		grp = ext4_get_group_info(sb, group);
 
 		if (EXT4_MB_GRP_NEED_INIT(grp) &&