linux-kernel - Re: [PATCH] mm/page_alloc: simplify lowmem

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <78E63F8C-7A6F-4B6B-AD92-25477CF59E4E@nvidia.com>
Date: Thu, 14 Aug 2025 21:59:22 -0400
From: Zi Yan <ziy@...dia.com>
To: Ye Liu <ye.liu@...ux.dev>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
 Vlastimil Babka <vbabka@...e.cz>, Ye Liu <liuye@...inos.cn>,
 Suren Baghdasaryan <surenb@...gle.com>, Michal Hocko <mhocko@...e.com>,
 Brendan Jackman <jackmanb@...gle.com>, Johannes Weiner <hannes@...xchg.org>,
 linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm/page_alloc: simplify lowmem_reserve max calculation

On 14 Aug 2025, at 21:38, Ye Liu wrote:

> 在 2025/8/14 22:47, Zi Yan 写道:
>> On 14 Aug 2025, at 5:00, Ye Liu wrote:
>>
>>> From: Ye Liu <liuye@...inos.cn>
>>>
>>> Use max() macro to simplify the calculation of maximum lowmem_reserve
>>> value in calculate_totalreserve_pages(), instead of open-coding the
>>> comparison. The functionality remains identical.
>>>
>>> Signed-off-by: Ye Liu <liuye@...inos.cn>
>>> ---
>>>  mm/page_alloc.c | 3 +--
>>>  1 file changed, 1 insertion(+), 2 deletions(-)
>>>
>>> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
>>> index 64872214bc7d..8a55a4951d19 100644
>>> --- a/mm/page_alloc.c
>>> +++ b/mm/page_alloc.c
>>> @@ -6236,8 +6236,7 @@ static void calculate_totalreserve_pages(void)
>>>
>>>  			/* Find valid and maximum lowmem_reserve in the zone */
>>>  			for (j = i; j < MAX_NR_ZONES; j++) {
>>> -				if (zone->lowmem_reserve[j] > max)
>>> -					max = zone->lowmem_reserve[j];
>>> +				max = max(max, zone->lowmem_reserve[j]);
>>>  			}
>>
>> There is a “if (max > managed_pages)” below. Maybe convert that as well?
>
> I should use min() here, but I noticed the two variables have different types:
> one is 'long' and the other is 'unsigned long'. So, I should use min_t().
> But then again, why is lowmem_reserve of type 'long'?

Based on commit 3484b2de9499 ("mm: rearrange zone fields into read-only, page
alloc, statistics and page reclaim lines”), lowmem_reserve type is changed
due to the watermark calculation. Looking at current __zone_watermark_ok(),
the comment above free_pages -= ... says free_pages may go negative and
later free_pages is compared with lowmem_reserve.

> It should be a non-negative number, right?
> Is it possible to change the type of lowmem_reserve to 'unsigned long' and
> change all uses of it at the same time? Is it necessary?

In calculate_totalreserve_pages(), max cannot be negative, so you
probably can use min_t(unsigned long, max, managed_pages) and without
changing lowmem_reserve type.


>
>>
>> Feel free to add Acked-by: Zi Yan <ziy@...dia.com>.
>>
>> Best Regards,
>> Yan, Zi
>
> -- 
> Thanks,
> Ye Liu


Best Regards,
Yan, Zi