mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* Re: [PATCH v3] watchdog/perf: one digit too short in the raw event config copy
  2026-10-03 11:02 [PATCH v3] watchdog/perf: one digit too short in the raw event config copy Bradley Morgan
@ 2026-10-03 10:41 ` Bradley Morgan
  0 siblings, 0 replies; 2+ messages in thread
From: Bradley Morgan @ 2026-10-03 10:41 UTC (permalink / raw)
  To: akpm; +Cc: blum, tglx, dianders, linux-kernel

On 3 October 2026 12:02:00 BST, Bradley Morgan <brads@mainlining.org>
wrote:
>Commit 6164be01f179 ("watchdog/perf: optimize bytes copied and remove
>manual NUL-termination") swapped strscpy(buf, str, sizeof(buf)) plus a
>manual buf[len] = 0 for strscpy(buf, str, len), and that count is one
>short. strscpy() keeps the last byte of the destination for the NUL,
>so the config loses its final digit, nmi_watchdog=r300,panic for
>example ends up with buf = "30" and arms the raw event with the wrong
>config.
>
>The empty case falls over too, strscpy() with a zero count writes
>nothing at all, so nmi_watchdog=r,1 leaves buf uninitialized and
>kstrtoull() reads whatever stack garbage is sitting there.
>
>The old code was safe on both by accident, it filled the whole buffer
>first and buf[len] = 0 then stomped the comma position, so the worst
>you got was a truncated parse failure.
>
>Pass len + 1 so the copy includes the character the NUL replaces, and
>reject len >= sizeof(buf) like it was before the optimization, which
>also turns an empty config back into a clean parse failure.
>
>Fixes: 6164be01f179 ("watchdog/perf: optimize bytes copied and remove manual NUL-termination")
>Signed-off-by: Bradley Morgan <brads@mainlining.org>
>Changes since v2:
>
>- rebased on pristine mainline, v2 was cut on top of v1 so the diff
>  carried v1's hunks
>- the intro note moved below the cut, it ended up in the commit
>  message in v2
>- dropped the Suggested-by, review feedback doesn't warrant one

For god sake, not again, since I'm lazy, just ommit the changes since in
the description Andrew. Imagine it's just below the ---

>
>---
> kernel/watchdog_perf.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
>diff --git a/kernel/watchdog_perf.c b/kernel/watchdog_perf.c
>index cf05775a96d3..a4f677c16b20 100644
>--- a/kernel/watchdog_perf.c
>+++ b/kernel/watchdog_perf.c
>@@ -301,10 +301,10 @@ void __init hardlockup_config_perf_event(const char *str)
> 	} else {
> 		unsigned int len = comma - str;
> 
>-		if (len > sizeof(buf))
>+		if (len >= sizeof(buf))
> 			return;
> 
>-		strscpy(buf, str, len);
>+		strscpy(buf, str, len + 1);
> 		if (kstrtoull(buf, 16, &config))
> 			return;
> 	}
>

--- Thanks!
"I'm not a very positive person" - Linus torvalds

^ permalink raw reply	[flat|nested] 2+ messages in thread

* [PATCH v3] watchdog/perf: one digit too short in the raw event config copy
@ 2026-10-03 11:02 Bradley Morgan
  2026-10-03 10:41 ` Bradley Morgan
  0 siblings, 1 reply; 2+ messages in thread
From: Bradley Morgan @ 2026-10-03 11:02 UTC (permalink / raw)
  To: akpm; +Cc: blum, tglx, dianders, linux-kernel, brads

Commit 6164be01f179 ("watchdog/perf: optimize bytes copied and remove
manual NUL-termination") swapped strscpy(buf, str, sizeof(buf)) plus a
manual buf[len] = 0 for strscpy(buf, str, len), and that count is one
short. strscpy() keeps the last byte of the destination for the NUL,
so the config loses its final digit, nmi_watchdog=r300,panic for
example ends up with buf = "30" and arms the raw event with the wrong
config.

The empty case falls over too, strscpy() with a zero count writes
nothing at all, so nmi_watchdog=r,1 leaves buf uninitialized and
kstrtoull() reads whatever stack garbage is sitting there.

The old code was safe on both by accident, it filled the whole buffer
first and buf[len] = 0 then stomped the comma position, so the worst
you got was a truncated parse failure.

Pass len + 1 so the copy includes the character the NUL replaces, and
reject len >= sizeof(buf) like it was before the optimization, which
also turns an empty config back into a clean parse failure.

Fixes: 6164be01f179 ("watchdog/perf: optimize bytes copied and remove manual NUL-termination")
Signed-off-by: Bradley Morgan <brads@mainlining.org>
Changes since v2:

- rebased on pristine mainline, v2 was cut on top of v1 so the diff
  carried v1's hunks
- the intro note moved below the cut, it ended up in the commit
  message in v2
- dropped the Suggested-by, review feedback doesn't warrant one

---
 kernel/watchdog_perf.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/watchdog_perf.c b/kernel/watchdog_perf.c
index cf05775a96d3..a4f677c16b20 100644
--- a/kernel/watchdog_perf.c
+++ b/kernel/watchdog_perf.c
@@ -301,10 +301,10 @@ void __init hardlockup_config_perf_event(const char *str)
 	} else {
 		unsigned int len = comma - str;
 
-		if (len > sizeof(buf))
+		if (len >= sizeof(buf))
 			return;
 
-		strscpy(buf, str, len);
+		strscpy(buf, str, len + 1);
 		if (kstrtoull(buf, 16, &config))
 			return;
 	}
-- 
2.53.0


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-10-03 10:41 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-10-03 11:02 [PATCH v3] watchdog/perf: one digit too short in the raw event config copy Bradley Morgan
2026-10-03 10:41 ` Bradley Morgan

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®