[PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Brendan Jackman]

Currently order is signed in one version of the function and unsigned in
the other. Tidy that up.

In page_alloc.c, order is unsigned in the vast majority of cases. But,
there is a cluster of exceptions in compaction-related code (probably
stemming from the fact that compact_control.order is signed). So, prefer
local consistency and make this one signed too.

Signed-off-by: Brendan Jackman <jackmanb@google.com>
---
 mm/page_alloc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index d1d037f97c5fc76f8a7739e8515d7593e0ad44f9..8faa0ad9f461fbe151ec347e331d83c2fdc8cad2 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -4182,7 +4182,7 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
 }
 
 static inline bool
-should_compact_retry(struct alloc_context *ac, unsigned int order, int alloc_flags,
+should_compact_retry(struct alloc_context *ac, int order, int alloc_flags,
 		     enum compact_result compact_result,
 		     enum compact_priority *compact_priority,
 		     int *compaction_retries)

---
base-commit: fab1beda7597fac1cecc01707d55eadb6bbe773c
change-id: 20250826-cleanup-should_compact_retry-9c6b5cdf8d27

Best regards,
-- 
Brendan Jackman <jackmanb@google.com>

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Zi Yan]

On 26 Aug 2025, at 10:06, Brendan Jackman wrote:

> Currently order is signed in one version of the function and unsigned in
> the other. Tidy that up.
>
> In page_alloc.c, order is unsigned in the vast majority of cases. But,
> there is a cluster of exceptions in compaction-related code (probably
> stemming from the fact that compact_control.order is signed). So, prefer
> local consistency and make this one signed too.
>
> Signed-off-by: Brendan Jackman <jackmanb@google.com>
> ---
>  mm/page_alloc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index d1d037f97c5fc76f8a7739e8515d7593e0ad44f9..8faa0ad9f461fbe151ec347e331d83c2fdc8cad2 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4182,7 +4182,7 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
>  }
>
>  static inline bool
> -should_compact_retry(struct alloc_context *ac, unsigned int order, int alloc_flags,
> +should_compact_retry(struct alloc_context *ac, int order, int alloc_flags,
>  		     enum compact_result compact_result,
>  		     enum compact_priority *compact_priority,
>  		     int *compaction_retries)
>
> ---
> base-commit: fab1beda7597fac1cecc01707d55eadb6bbe773c
> change-id: 20250826-cleanup-should_compact_retry-9c6b5cdf8d27
>

LGTM. Reviewed-by: Zi Yan <ziy@nvidia.com>

--
Best Regards,
Yan, Zi

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Andrew Morton]

On Tue, 26 Aug 2025 14:06:54 +0000 Brendan Jackman <jackmanb@google.com> wrote:

> Currently order is signed in one version of the function and unsigned in
> the other. Tidy that up.
> 
> In page_alloc.c, order is unsigned in the vast majority of cases. But,
> there is a cluster of exceptions in compaction-related code (probably
> stemming from the fact that compact_control.order is signed). So, prefer
> local consistency and make this one signed too.
> 

grumble, pet peeve.  Negative orders make no sense.  Can we make
cc->order unsigned in order (heh) to make everything nice?

> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4182,7 +4182,7 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
>  }
>  
>  static inline bool
> -should_compact_retry(struct alloc_context *ac, unsigned int order, int alloc_flags,
> +should_compact_retry(struct alloc_context *ac, int order, int alloc_flags,
>  		     enum compact_result compact_result,
>  		     enum compact_priority *compact_priority,
>  		     int *compaction_retries)
> 

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Zi Yan]

On 26 Aug 2025, at 22:13, Andrew Morton wrote:

> On Tue, 26 Aug 2025 14:06:54 +0000 Brendan Jackman <jackmanb@google.com> wrote:
>
>> Currently order is signed in one version of the function and unsigned in
>> the other. Tidy that up.
>>
>> In page_alloc.c, order is unsigned in the vast majority of cases. But,
>> there is a cluster of exceptions in compaction-related code (probably
>> stemming from the fact that compact_control.order is signed). So, prefer
>> local consistency and make this one signed too.
>>
>
> grumble, pet peeve.  Negative orders make no sense.  Can we make
> cc->order unsigned in order (heh) to make everything nice?

Unless we do not do order--, where order can go negative. See next_search_order()
in mm/compaction.c as an example.

>
>> --- a/mm/page_alloc.c
>> +++ b/mm/page_alloc.c
>> @@ -4182,7 +4182,7 @@ __alloc_pages_direct_compact(gfp_t gfp_mask, unsigned int order,
>>  }
>>
>>  static inline bool
>> -should_compact_retry(struct alloc_context *ac, unsigned int order, int alloc_flags,
>> +should_compact_retry(struct alloc_context *ac, int order, int alloc_flags,
>>  		     enum compact_result compact_result,
>>  		     enum compact_priority *compact_priority,
>>  		     int *compaction_retries)
>>


--
Best Regards,
Yan, Zi

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Brendan Jackman]

On Wed Aug 27, 2025 at 2:13 AM UTC, Andrew Morton wrote:
> On Tue, 26 Aug 2025 14:06:54 +0000 Brendan Jackman <jackmanb@google.com> wrote:
>
>> Currently order is signed in one version of the function and unsigned in
>> the other. Tidy that up.
>> 
>> In page_alloc.c, order is unsigned in the vast majority of cases. But,
>> there is a cluster of exceptions in compaction-related code (probably
>> stemming from the fact that compact_control.order is signed). So, prefer
>> local consistency and make this one signed too.
>> 
>
> grumble, pet peeve.  Negative orders make no sense.  Can we make
> cc->order unsigned in order (heh) to make everything nice?

I think we can't "just" do that:

/*
 * order == -1 is expected when compacting proactively via
 * 1. /proc/sys/vm/compact_memory
 * 2. /sys/devices/system/node/nodex/compact
 * 3. /proc/sys/vm/compaction_proactiveness
 */
static inline bool is_via_compact_memory(int order)
{
	return order == -1;
}

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Vlastimil Babka]

On 8/27/25 17:30, Brendan Jackman wrote:
> On Wed Aug 27, 2025 at 2:13 AM UTC, Andrew Morton wrote:
>> On Tue, 26 Aug 2025 14:06:54 +0000 Brendan Jackman <jackmanb@google.com> wrote:
>>
>>> Currently order is signed in one version of the function and unsigned in
>>> the other. Tidy that up.
>>> 
>>> In page_alloc.c, order is unsigned in the vast majority of cases. But,
>>> there is a cluster of exceptions in compaction-related code (probably
>>> stemming from the fact that compact_control.order is signed). So, prefer
>>> local consistency and make this one signed too.
>>> 
>>
>> grumble, pet peeve.  Negative orders make no sense.  Can we make
>> cc->order unsigned in order (heh) to make everything nice?
> 
> I think we can't "just" do that:

That part should be easy to convert to a compact_control flag.
Zi's point about going negative seems like more prone to overlook some case.
But worth trying and the cleanups I'd say.
> /*
>  * order == -1 is expected when compacting proactively via
>  * 1. /proc/sys/vm/compact_memory
>  * 2. /sys/devices/system/node/nodex/compact
>  * 3. /proc/sys/vm/compaction_proactiveness
>  */
> static inline bool is_via_compact_memory(int order)
> {
> 	return order == -1;
> }

Re: [PATCH] mm/page_alloc: Harmonize should_compact_retry() type

[Brendan Jackman]

On Thu Aug 28, 2025 at 6:31 PM UTC, Vlastimil Babka wrote:
> On 8/27/25 17:30, Brendan Jackman wrote:
>> On Wed Aug 27, 2025 at 2:13 AM UTC, Andrew Morton wrote:
>>> On Tue, 26 Aug 2025 14:06:54 +0000 Brendan Jackman <jackmanb@google.com> wrote:
>>>
>>>> Currently order is signed in one version of the function and unsigned in
>>>> the other. Tidy that up.
>>>> 
>>>> In page_alloc.c, order is unsigned in the vast majority of cases. But,
>>>> there is a cluster of exceptions in compaction-related code (probably
>>>> stemming from the fact that compact_control.order is signed). So, prefer
>>>> local consistency and make this one signed too.
>>>> 
>>>
>>> grumble, pet peeve.  Negative orders make no sense.  Can we make
>>> cc->order unsigned in order (heh) to make everything nice?
>> 
>> I think we can't "just" do that:
>
> That part should be easy to convert to a compact_control flag.
> Zi's point about going negative seems like more prone to overlook some case.
> But worth trying and the cleanups I'd say.

OK, I can take a look next week. From a quick glance it does seem to be
worth having a sniff, there could be bugs in there where code is
expecting non-negative and doing stuff like using it as a shift index,
in cases where it can actually be negative.