* [PATCH] mm/hugetlb: skip report_hugepages() output when no hugepages configured
@ 2025-09-10 9:03 lirongqing
2025-09-10 9:09 ` David Hildenbrand
0 siblings, 1 reply; 2+ messages in thread
From: lirongqing @ 2025-09-10 9:03 UTC (permalink / raw)
To: muchun.song, osalvador, david, akpm, linux-mm, linux-kernel; +Cc: Li RongQing
From: Li RongQing <lirongqing@baidu.com>
Avoid unnecessary report_hugepages() output and processing when
max_huge_pages is 0. This eliminates redundant log messages for
unconfigured hugepage sizes and saves CPU cycles during boot.
Before this change, the kernel would always print registration messages
even for hugepage sizes with zero pre-allocated pages:
[ 4.118953] HugeTLB: registered 2.00 MiB page size, pre-allocated 0 pages
[ 4.122920] HugeTLB: 28 KiB vmemmap can be freed for a 2.00 MiB page
[ 4.126920] HugeTLB: registered 1.00 GiB page size, pre-allocated 0 pages
[ 4.130920] HugeTLB: 16380 KiB vmemmap can be freed for a 1.00 GiB page
Now these messages are suppressed when no pages are configured.
Signed-off-by: Li RongQing <lirongqing@baidu.com>
---
mm/hugetlb.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index af7c36f..9fb9311 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3733,6 +3733,9 @@ static void __init report_hugepages(void)
for_each_hstate(h) {
char buf[32];
+ if (!h->max_huge_pages)
+ continue;
+
nrinvalid = hstate_boot_nrinvalid[hstate_index(h)];
h->max_huge_pages -= nrinvalid;
--
2.9.4
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] mm/hugetlb: skip report_hugepages() output when no hugepages configured
2025-09-10 9:03 [PATCH] mm/hugetlb: skip report_hugepages() output when no hugepages configured lirongqing
@ 2025-09-10 9:09 ` David Hildenbrand
0 siblings, 0 replies; 2+ messages in thread
From: David Hildenbrand @ 2025-09-10 9:09 UTC (permalink / raw)
To: lirongqing, muchun.song, osalvador, akpm, linux-mm, linux-kernel
On 10.09.25 11:03, lirongqing wrote:
> From: Li RongQing <lirongqing@baidu.com>
>
> Avoid unnecessary report_hugepages() output and processing when
> max_huge_pages is 0. This eliminates redundant log messages for
> unconfigured hugepage sizes and saves CPU cycles during boot.
>
> Before this change, the kernel would always print registration messages
> even for hugepage sizes with zero pre-allocated pages:
>
> [ 4.118953] HugeTLB: registered 2.00 MiB page size, pre-allocated 0 pages
> [ 4.122920] HugeTLB: 28 KiB vmemmap can be freed for a 2.00 MiB page
> [ 4.126920] HugeTLB: registered 1.00 GiB page size, pre-allocated 0 pages
> [ 4.130920] HugeTLB: 16380 KiB vmemmap can be freed for a 1.00 GiB page
>
> Now these messages are suppressed when no pages are configured.
Well, there is valuable information in there, especially given that we
can allocate these hugetlb pages later.
So I think we want to keep printing that information.
--
Cheers
David / dhildenb
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-09-10 9:09 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-09-10 9:03 [PATCH] mm/hugetlb: skip report_hugepages() output when no hugepages configured lirongqing
2025-09-10 9:09 ` David Hildenbrand
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox