linux-kernel - Re: [PATCH] selftests/mm: Reduce uffd-unit-test poison test to minimum

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] [day] [month] [year] [list]

Message-ID: <aF2RSsjuEOtzXcUa@x1.local>
Date: Thu, 26 Jun 2025 14:28:26 -0400
From: Peter Xu <peterx@...hat.com>
To: Axel Rasmussen <axelrasmussen@...gle.com>
Cc: linux-kernel@...r.kernel.org, linux-mm@...ck.org,
	Andrew Morton <akpm@...ux-foundation.org>,
	Ujwal Kundur <ujwal.kundur@...il.com>,
	David Hildenbrand <david@...hat.com>, shuah@...nel.org,
	jackmanb@...gle.com
Subject: Re: [PATCH] selftests/mm: Reduce uffd-unit-test poison test to
 minimum

On Thu, Jun 26, 2025 at 10:32:20AM -0700, Axel Rasmussen wrote:
> On Fri, Jun 20, 2025 at 8:01 AM Peter Xu <peterx@...hat.com> wrote:
> >
> > The test will still generate quite some unwanted MCE error messages to
> > syslog.  There was old proposal ratelimiting the MCE messages from kernel,
> > but that has risk of hiding real useful information on production systems.
> >
> > We can at least reduce the test to minimum to not over-pollute dmesg,
> > however trying to not lose its coverage too much.
> >
> > Cc: Axel Rasmussen <axelrasmussen@...gle.com>
> > Signed-off-by: Peter Xu <peterx@...hat.com>
> 
> Besides a small nitpick you can take:
> 
> Reviewed-by: Axel Rasmussen <axelrasmussen@...gle.com>

Thanks, Axel.

> 
> Making the functional tests small makes sense to me, especially for
> poisoning. Only reason to use a huge number of pages is if we're
> trying to stress racy bugs or so, but really for that you'd want even
> more pages / more threads / run for a longer time. It makes sense to
> separate that use case out / maybe not run it by default, and leave
> the functional tests small + fast.

IIUC the major complain from others are the pollutions in the dmesg.  And
yes, it's also slow and IIUC it's because of the messages piped, at least
when I ran it in VMs with a serial console.

> 
> > ---
> >  tools/testing/selftests/mm/uffd-unit-tests.c | 20 ++++++++++++++------
> >  1 file changed, 14 insertions(+), 6 deletions(-)
> >
> > diff --git a/tools/testing/selftests/mm/uffd-unit-tests.c b/tools/testing/selftests/mm/uffd-unit-tests.c
> > index c73fd5d455c8..39b3fd1b7bf2 100644
> > --- a/tools/testing/selftests/mm/uffd-unit-tests.c
> > +++ b/tools/testing/selftests/mm/uffd-unit-tests.c
> > @@ -1027,6 +1027,9 @@ static void uffd_poison_handle_fault(
> >                 do_uffdio_poison(uffd, offset);
> >  }
> >
> > +/* Make sure to cover odd/even, and minimum duplications */
> > +#define  UFFD_POISON_TEST_NPAGES  4
> > +
> >  static void uffd_poison_test(uffd_test_args_t *targs)
> >  {
> >         pthread_t uffd_mon;
> > @@ -1034,12 +1037,17 @@ static void uffd_poison_test(uffd_test_args_t *targs)
> >         struct uffd_args args = { 0 };
> >         struct sigaction act = { 0 };
> >         unsigned long nr_sigbus = 0;
> > -       unsigned long nr;
> > +       unsigned long nr, poison_pages = UFFD_POISON_TEST_NPAGES;
> > +
> > +       if (nr_pages < poison_pages) {
> > +               uffd_test_skip("Too less pages for POISON test");
> 
> I think "Too few pages for POISON test" is more grammatically correct.

Right..  This is currently in mm-unstable.  Andrew, would you mind take
below as a fixup?  TIA!

===8<===

commit 810f5674e1b990775600ffca533dbf75505adbef (HEAD -> test-poison)
Author: Peter Xu <peterx@...hat.com>
Date:   Thu Jun 26 14:22:21 2025 -0400

    fixup! selftests/mm: reduce uffd-unit-test poison test to minimum
    
    Signed-off-by: Peter Xu <peterx@...hat.com>
---
 tools/testing/selftests/mm/uffd-unit-tests.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/mm/uffd-unit-tests.c b/tools/testing/selftests/mm/uffd-unit-tests.c
index 39b3fd1b7bf2..50501b38e34e 100644
--- a/tools/testing/selftests/mm/uffd-unit-tests.c
+++ b/tools/testing/selftests/mm/uffd-unit-tests.c
@@ -1040,7 +1040,7 @@ static void uffd_poison_test(uffd_test_args_t *targs)
        unsigned long nr, poison_pages = UFFD_POISON_TEST_NPAGES;
 
        if (nr_pages < poison_pages) {
-               uffd_test_skip("Too less pages for POISON test");
+               uffd_test_skip("Too few pages for POISON test");
                return;
        }

-- 
Peter Xu