Re: [PATCH 4.19 03/16] mm: kasan: do not panic if both panic_on_warn and kasan_multishot set

[SeongJae Park]

On Wed, 1 Feb 2023 21:25:51 -0800 Eric Biggers <ebiggers@xxxxxxxxxx> wrote:

> From: David Gow <davidgow@xxxxxxxxxx>
> 
> commit be4f1ae978ffe98cc95ec49ceb95386fb4474974 upstream.
> 
> KASAN errors will currently trigger a panic when panic_on_warn is set.
> This renders kasan_multishot useless, as further KASAN errors won't be
> reported if the kernel has already paniced.  By making kasan_multishot
> disable this behaviour for KASAN errors, we can still have the benefits of
> panic_on_warn for non-KASAN warnings, yet be able to use kasan_multishot.
> 
> This is particularly important when running KASAN tests, which need to
> trigger multiple KASAN errors: previously these would panic the system if
> panic_on_warn was set, now they can run (and will panic the system should
> non-KASAN warnings show up).
> 
> Signed-off-by: David Gow <davidgow@xxxxxxxxxx>
> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Tested-by: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
> Reviewed-by: Andrey Konovalov <andreyknvl@xxxxxxxxxx>
> Reviewed-by: Brendan Higgins <brendanhiggins@xxxxxxxxxx>
> Cc: Andrey Ryabinin <aryabinin@xxxxxxxxxxxxx>
> Cc: Dmitry Vyukov <dvyukov@xxxxxxxxxx>
> Cc: Ingo Molnar <mingo@xxxxxxxxxx>
> Cc: Juri Lelli <juri.lelli@xxxxxxxxxx>
> Cc: Patricia Alfonso <trishalfonso@xxxxxxxxxx>
> Cc: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
> Cc: Shuah Khan <shuah@xxxxxxxxxx>
> Cc: Vincent Guittot <vincent.guittot@xxxxxxxxxx>
> Link: https://lkml.kernel.org/r/20200915035828.570483-6-davidgow@xxxxxxxxxx
> Link: https://lkml.kernel.org/r/20200910070331.3358048-6-davidgow@xxxxxxxxxx
> Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> Signed-off-by: Eric Biggers <ebiggers@xxxxxxxxxx>
> ---
>  mm/kasan/report.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/kasan/report.c b/mm/kasan/report.c
> index 5c169aa688fde..90fdb261a5e2d 100644
> --- a/mm/kasan/report.c
> +++ b/mm/kasan/report.c
> @@ -176,7 +176,7 @@ static void kasan_end_report(unsigned long *flags)
>  	pr_err("==================================================================\n");
>  	add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE);
>  	spin_unlock_irqrestore(&report_lock, *flags);
> -	if (panic_on_warn)
> +	if (panic_on_warn && !test_bit(KASAN_BIT_MULTI_SHOT, &kasan_flags))
>  		panic("panic_on_warn set ...\n");
>  	kasan_enable_current();


Seems this introduced a build failure when CONFIG_KASAN is enabled, as also
reported by Sasha[1].

    mm/kasan/report.c: In function ‘kasan_end_report’:
    mm/kasan/report.c:179:16: error: ‘KASAN_BIT_MULTI_SHOT’ undeclared (first use in this function)
      179 |  if (!test_bit(KASAN_BIT_MULTI_SHOT, &kasan_flags))
          |                ^~~~~~~~~~~~~~~~~~~~
    arch/x86/include/asm/bitops.h:342:25: note: in definition of macro ‘test_bit’
      342 |  (__builtin_constant_p((nr))  \
          |                         ^~
    mm/kasan/report.c:179:16: note: each undeclared identifier is reported only once for each function it appears in
      179 |  if (!test_bit(KASAN_BIT_MULTI_SHOT, &kasan_flags))
          |                ^~~~~~~~~~~~~~~~~~~~
    arch/x86/include/asm/bitops.h:342:25: note: in definition of macro ‘test_bit’
      342 |  (__builtin_constant_p((nr))  \
          |                         ^~
    mm/kasan/report.c:179:39: error: ‘kasan_flags’ undeclared (first use in this function)
      179 |  if (!test_bit(KASAN_BIT_MULTI_SHOT, &kasan_flags))
          |                                       ^~~~~~~~~~~
    arch/x86/include/asm/bitops.h:343:30: note: in definition of macro ‘test_bit’
      343 |   ? constant_test_bit((nr), (addr)) \
          |                              ^~~~

I confirmed dropping this patch fixes the build failure.  It causes a conflict
to a following patch[1], but seems it's not that difficult to resolve.  I
updated kernel/panic.c part of the patch like attached below to resolve the
conflict.

[1] https://lore.kernel.org/stable/Y9v3G6UantaCo29G@sashalap/
[2] https://lore.kernel.org/stable/20230202052604.179184-13-ebiggers@xxxxxxxxxx/


Thanks,
SJ


================================ >8 ===========================================

diff --git a/kernel/panic.c b/kernel/panic.c
index a078d413042f..08b8adc55b2b 100644
--- a/kernel/panic.c
+++ b/kernel/panic.c
@@ -125,6 +125,12 @@ void nmi_panic(struct pt_regs *regs, const char *msg)
 }
 EXPORT_SYMBOL(nmi_panic);

+void check_panic_on_warn(const char *origin)
+{
+       if (panic_on_warn)
+               panic("%s: panic_on_warn set ...\n", origin);
+}
+
 /**
  *     panic - halt the system
  *     @fmt: The text string to print
@@ -540,8 +546,7 @@ void __warn(const char *file, int line, void *caller, unsigned taint,
        if (args)
                vprintk(args->fmt, args->args);

-       if (panic_on_warn)
-               panic("panic_on_warn set ...\n");
+       check_panic_on_warn("kernel");

        print_modules();