* [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios
@ 2026-07-14 2:16 Song Hu
2026-09-02 0:37 ` Andrew Morton
0 siblings, 1 reply; 4+ messages in thread
From: Song Hu @ 2026-07-14 2:16 UTC (permalink / raw)
To: linux-mm
Cc: akpm, muchun.song, osalvador, david, ljs, liam, vbabka, rppt,
surenb, mhocko, shuah, linux-kselftest, linux-kernel, Song Hu
A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made
the "more than reservation", "more than cgroup limit", and "multiple
cgroups" scenarios unreachable, so they never ran.
Remove it so the first two run again. The third,
run_multiple_cgroup_test(), is separately broken -- a backslash
line-continuation breaks its local declarations and its hugetlb
reservation accounting does not charge as expected -- and has never run
in CI; skip it explicitly with a comment until fixed. Also correct two
`[[ "$method" == 2"" ]]` typos (harmless empty-string concatenation)
to plain `2`.
Signed-off-by: Song Hu <husong@kylinos.cn>
---
.../selftests/mm/charge_reserved_hugetlb.sh | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh b/tools/testing/selftests/mm/charge_reserved_hugetlb.sh
index a1cfd3a349db..18b2f2060c61 100755
--- a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh
+++ b/tools/testing/selftests/mm/charge_reserved_hugetlb.sh
@@ -439,11 +439,11 @@ for populate in "" "-o"; do
fi
# Skip populated shmem tests. Doesn't seem to be supported.
- if [[ "$method" == 2"" ]] && [[ "$populate" == "-o" ]]; then
+ if [[ "$method" == 2 ]] && [[ "$populate" == "-o" ]]; then
continue
fi
- if [[ "$method" == 2"" ]] && [[ "$reserve" == "-n" ]]; then
+ if [[ "$method" == 2 ]] && [[ "$reserve" == "-n" ]]; then
continue
fi
@@ -496,7 +496,6 @@ for populate in "" "-o"; do
echo 'PASS'
cleanup
- continue
echo
echo
echo
@@ -530,6 +529,15 @@ for populate in "" "-o"; do
cleanup
+ # The "multiple cgroups" scenario (run_multiple_cgroup_test) is
+ # multiply broken and has never run in CI: a backslash
+ # line-continuation breaks its local declarations, it passes
+ # reservation sizes in bytes rather than pages (unlike
+ # run_test), and its assertions compare page counts to byte
+ # values. Skip it until properly rewritten; the two scenarios
+ # above are restored by this change.
+ continue
+
echo
echo
echo
--
2.43.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios
2026-07-14 2:16 [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios Song Hu
@ 2026-09-02 0:37 ` Andrew Morton
2026-09-04 8:30 ` Mike Rapoport
0 siblings, 1 reply; 4+ messages in thread
From: Andrew Morton @ 2026-09-02 0:37 UTC (permalink / raw)
To: Song Hu
Cc: linux-mm, muchun.song, osalvador, david, ljs, liam, vbabka, rppt,
surenb, mhocko, shuah, linux-kselftest, linux-kernel,
Mina Almasry
On Tue, 14 Jul 2026 10:16:30 +0800 Song Hu <husong@kylinos.cn> wrote:
> A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made
> the "more than reservation", "more than cgroup limit", and "multiple
> cgroups" scenarios unreachable, so they never ran.
>
> Remove it so the first two run again.
OK.
> The third,
> run_multiple_cgroup_test(), is separately broken -- a backslash
> line-continuation breaks its local declarations and its hugetlb
> reservation accounting does not charge as expected -- and has never run
> in CI; skip it explicitly with a comment until fixed.
That's reasonable. This code was added by Mina six years ago. But
he's still around so let's bug him with a cc ;)
> Also correct two
> `[[ "$method" == 2"" ]]` typos (harmless empty-string concatenation)
> to plain `2`.
OK.
From: Song Hu <husong@kylinos.cn>
Subject: selftests/mm: restore hidden hugetlb reservation scenarios
Date: Tue, 14 Jul 2026 10:16:30 +0800
A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made the
"more than reservation", "more than cgroup limit", and "multiple cgroups"
scenarios unreachable, so they never ran.
Remove it so the first two run again. The third,
run_multiple_cgroup_test(), is separately broken -- a backslash
line-continuation breaks its local declarations and its hugetlb
reservation accounting does not charge as expected -- and has never run in
CI; skip it explicitly with a comment until fixed. Also correct two `[[
"$method" == 2"" ]]` typos (harmless empty-string concatenation) to plain
`2`.
Link: https://lore.kernel.org/20260714021630.1063823-1-husong@kylinos.cn
Signed-off-by: Song Hu <husong@kylinos.cn>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: David Hildenbrand <david@kernel.org>
Cc: Liam R. Howlett <liam@infradead.org>
Cc: Lorenzo Stoakes <ljs@kernel.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Muchun Song <muchun.song@linux.dev>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Vlastimil Babka <vbabka@kernel.org>
Cc: Mina Almasry <almasrymina@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
tools/testing/selftests/mm/charge_reserved_hugetlb.sh | 14 +++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
--- a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh~selftests-mm-restore-hidden-hugetlb-reservation-scenarios
+++ a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh
@@ -439,11 +439,11 @@ for populate in "" "-o"; do
fi
# Skip populated shmem tests. Doesn't seem to be supported.
- if [[ "$method" == 2"" ]] && [[ "$populate" == "-o" ]]; then
+ if [[ "$method" == 2 ]] && [[ "$populate" == "-o" ]]; then
continue
fi
- if [[ "$method" == 2"" ]] && [[ "$reserve" == "-n" ]]; then
+ if [[ "$method" == 2 ]] && [[ "$reserve" == "-n" ]]; then
continue
fi
@@ -496,7 +496,6 @@ for populate in "" "-o"; do
echo 'PASS'
cleanup
- continue
echo
echo
echo
@@ -530,6 +529,15 @@ for populate in "" "-o"; do
cleanup
+ # The "multiple cgroups" scenario (run_multiple_cgroup_test) is
+ # multiply broken and has never run in CI: a backslash
+ # line-continuation breaks its local declarations, it passes
+ # reservation sizes in bytes rather than pages (unlike
+ # run_test), and its assertions compare page counts to byte
+ # values. Skip it until properly rewritten; the two scenarios
+ # above are restored by this change.
+ continue
+
echo
echo
echo
_
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios
2026-09-02 0:37 ` Andrew Morton
@ 2026-09-04 8:30 ` Mike Rapoport
2026-09-04 17:48 ` Andrew Morton
0 siblings, 1 reply; 4+ messages in thread
From: Mike Rapoport @ 2026-09-04 8:30 UTC (permalink / raw)
To: Andrew Morton
Cc: Song Hu, linux-mm, muchun.song, osalvador, david, ljs, liam,
vbabka, surenb, mhocko, shuah, linux-kselftest, linux-kernel,
Mina Almasry
On Tue, Sep 01, 2026 at 05:37:48PM -0700, Andrew Morton wrote:
> On Tue, 14 Jul 2026 10:16:30 +0800 Song Hu <husong@kylinos.cn> wrote:
>
> > A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made
> > the "more than reservation", "more than cgroup limit", and "multiple
> > cgroups" scenarios unreachable, so they never ran.
> >
> > Remove it so the first two run again.
>
> OK.
>
> > The third,
> > run_multiple_cgroup_test(), is separately broken -- a backslash
> > line-continuation breaks its local declarations and its hugetlb
> > reservation accounting does not charge as expected -- and has never run
> > in CI; skip it explicitly with a comment until fixed.
>
> That's reasonable. This code was added by Mina six years ago. But
> he's still around so let's bug him with a cc ;)
And this code breaks our CI e.g:
https://github.com/linux-mm/linux-mm/actions/runs/33040981840/job/98414290236
> > Also correct two
> > `[[ "$method" == 2"" ]]` typos (harmless empty-string concatenation)
> > to plain `2`.
>
> OK.
>
>
>
> From: Song Hu <husong@kylinos.cn>
> Subject: selftests/mm: restore hidden hugetlb reservation scenarios
> Date: Tue, 14 Jul 2026 10:16:30 +0800
>
> A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made the
> "more than reservation", "more than cgroup limit", and "multiple cgroups"
> scenarios unreachable, so they never ran.
>
> Remove it so the first two run again. The third,
> run_multiple_cgroup_test(), is separately broken -- a backslash
> line-continuation breaks its local declarations and its hugetlb
> reservation accounting does not charge as expected -- and has never run in
> CI; skip it explicitly with a comment until fixed. Also correct two `[[
> "$method" == 2"" ]]` typos (harmless empty-string concatenation) to plain
> `2`.
>
> Link: https://lore.kernel.org/20260714021630.1063823-1-husong@kylinos.cn
> Signed-off-by: Song Hu <husong@kylinos.cn>
> Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
> Cc: David Hildenbrand <david@kernel.org>
> Cc: Liam R. Howlett <liam@infradead.org>
> Cc: Lorenzo Stoakes <ljs@kernel.org>
> Cc: Michal Hocko <mhocko@suse.com>
> Cc: Mike Rapoport <rppt@kernel.org>
> Cc: Muchun Song <muchun.song@linux.dev>
> Cc: Oscar Salvador <osalvador@suse.de>
> Cc: Shuah Khan <shuah@kernel.org>
> Cc: Suren Baghdasaryan <surenb@google.com>
> Cc: Vlastimil Babka <vbabka@kernel.org>
> Cc: Mina Almasry <almasrymina@google.com>
> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
> ---
>
> tools/testing/selftests/mm/charge_reserved_hugetlb.sh | 14 +++++++---
> 1 file changed, 11 insertions(+), 3 deletions(-)
>
> --- a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh~selftests-mm-restore-hidden-hugetlb-reservation-scenarios
> +++ a/tools/testing/selftests/mm/charge_reserved_hugetlb.sh
> @@ -439,11 +439,11 @@ for populate in "" "-o"; do
> fi
>
> # Skip populated shmem tests. Doesn't seem to be supported.
> - if [[ "$method" == 2"" ]] && [[ "$populate" == "-o" ]]; then
> + if [[ "$method" == 2 ]] && [[ "$populate" == "-o" ]]; then
> continue
> fi
>
> - if [[ "$method" == 2"" ]] && [[ "$reserve" == "-n" ]]; then
> + if [[ "$method" == 2 ]] && [[ "$reserve" == "-n" ]]; then
> continue
> fi
>
> @@ -496,7 +496,6 @@ for populate in "" "-o"; do
> echo 'PASS'
>
> cleanup
> - continue
> echo
> echo
> echo
> @@ -530,6 +529,15 @@ for populate in "" "-o"; do
>
> cleanup
>
> + # The "multiple cgroups" scenario (run_multiple_cgroup_test) is
> + # multiply broken and has never run in CI: a backslash
> + # line-continuation breaks its local declarations, it passes
> + # reservation sizes in bytes rather than pages (unlike
> + # run_test), and its assertions compare page counts to byte
> + # values. Skip it until properly rewritten; the two scenarios
> + # above are restored by this change.
> + continue
> +
> echo
> echo
> echo
> _
>
>
--
Sincerely yours,
Mike.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios
2026-09-04 8:30 ` Mike Rapoport
@ 2026-09-04 17:48 ` Andrew Morton
0 siblings, 0 replies; 4+ messages in thread
From: Andrew Morton @ 2026-09-04 17:48 UTC (permalink / raw)
To: Mike Rapoport
Cc: Song Hu, linux-mm, muchun.song, osalvador, david, ljs, liam,
vbabka, surenb, mhocko, shuah, linux-kselftest, linux-kernel,
Mina Almasry
On Fri, 4 Sep 2026 11:30:23 +0300 Mike Rapoport <rppt@kernel.org> wrote:
> On Tue, Sep 01, 2026 at 05:37:48PM -0700, Andrew Morton wrote:
> > On Tue, 14 Jul 2026 10:16:30 +0800 Song Hu <husong@kylinos.cn> wrote:
> >
> > > A stray 'continue' in the main loop of charge_reserved_hugetlb.sh made
> > > the "more than reservation", "more than cgroup limit", and "multiple
> > > cgroups" scenarios unreachable, so they never ran.
> > >
> > > Remove it so the first two run again.
> >
> > OK.
> >
> > > The third,
> > > run_multiple_cgroup_test(), is separately broken -- a backslash
> > > line-continuation breaks its local declarations and its hugetlb
> > > reservation accounting does not charge as expected -- and has never run
> > > in CI; skip it explicitly with a comment until fixed.
> >
> > That's reasonable. This code was added by Mina six years ago. But
> > he's still around so let's bug him with a cc ;)
>
> And this code breaks our CI e.g:
>
> https://github.com/linux-mm/linux-mm/actions/runs/33040981840/job/98414290236
What do we need to do to view the error output there?
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-09-04 17:48 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-07-14 2:16 [PATCH 2/2] selftests/mm: restore hidden hugetlb reservation scenarios Song Hu
2026-09-02 0:37 ` Andrew Morton
2026-09-04 8:30 ` Mike Rapoport
2026-09-04 17:48 ` Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox