* [PATCH] ALSA: wavefront: Fix use-after-free in MIDI operations
@ 2025-11-04 14:11 moonafterrain
2025-11-05 8:41 ` Takashi Iwai
0 siblings, 1 reply; 2+ messages in thread
From: moonafterrain @ 2025-11-04 14:11 UTC (permalink / raw)
To: Jaroslav Kysela, Takashi Iwai
Cc: linux-sound, linux-kernel, stable, Yuhao Jiang, Junrui Luo
From: Junrui Luo <moonafterrain@outlook.com>
Clear substream pointers in close functions to prevent use-after-free
when timer callbacks or interrupt handlers access them after close.
Reported-by: Yuhao Jiang <danisjiang@gmail.com>
Reported-by: Junrui Luo <moonafterrain@outlook.com>
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Cc: stable@vger.kernel.org
Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
---
sound/isa/wavefront/wavefront_midi.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/sound/isa/wavefront/wavefront_midi.c b/sound/isa/wavefront/wavefront_midi.c
index 1250ecba659a..69d87c4cafae 100644
--- a/sound/isa/wavefront/wavefront_midi.c
+++ b/sound/isa/wavefront/wavefront_midi.c
@@ -278,6 +278,7 @@ static int snd_wavefront_midi_input_close(struct snd_rawmidi_substream *substrea
return -EIO;
guard(spinlock_irqsave)(&midi->open);
+ midi->substream_input[mpu] = NULL;
midi->mode[mpu] &= ~MPU401_MODE_INPUT;
return 0;
@@ -300,6 +301,7 @@ static int snd_wavefront_midi_output_close(struct snd_rawmidi_substream *substre
return -EIO;
guard(spinlock_irqsave)(&midi->open);
+ midi->substream_output[mpu] = NULL;
midi->mode[mpu] &= ~MPU401_MODE_OUTPUT;
return 0;
}
--
2.51.1.dirty
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] ALSA: wavefront: Fix use-after-free in MIDI operations
2025-11-04 14:11 [PATCH] ALSA: wavefront: Fix use-after-free in MIDI operations moonafterrain
@ 2025-11-05 8:41 ` Takashi Iwai
0 siblings, 0 replies; 2+ messages in thread
From: Takashi Iwai @ 2025-11-05 8:41 UTC (permalink / raw)
To: moonafterrain
Cc: Jaroslav Kysela, Takashi Iwai, linux-sound, linux-kernel, stable,
Yuhao Jiang
On Tue, 04 Nov 2025 15:11:55 +0100,
moonafterrain@outlook.com wrote:
>
> From: Junrui Luo <moonafterrain@outlook.com>
>
> Clear substream pointers in close functions to prevent use-after-free
> when timer callbacks or interrupt handlers access them after close.
There can be no actual access done because MPU401_MODE_INPUT_TRIGGER
is guaranteed to be off before closing the stream. That is, the
variable stream is assigned to an old pointer, but it's not accessed.
So, this is not really a use-after-free bug, per se.
Other than that, the fix itself looks good. Please resubmit after
rephrasing the patch description.
thanks,
Takashi
> Reported-by: Yuhao Jiang <danisjiang@gmail.com>
> Reported-by: Junrui Luo <moonafterrain@outlook.com>
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Cc: stable@vger.kernel.org
> Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
> ---
> sound/isa/wavefront/wavefront_midi.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/sound/isa/wavefront/wavefront_midi.c b/sound/isa/wavefront/wavefront_midi.c
> index 1250ecba659a..69d87c4cafae 100644
> --- a/sound/isa/wavefront/wavefront_midi.c
> +++ b/sound/isa/wavefront/wavefront_midi.c
> @@ -278,6 +278,7 @@ static int snd_wavefront_midi_input_close(struct snd_rawmidi_substream *substrea
> return -EIO;
>
> guard(spinlock_irqsave)(&midi->open);
> + midi->substream_input[mpu] = NULL;
> midi->mode[mpu] &= ~MPU401_MODE_INPUT;
>
> return 0;
> @@ -300,6 +301,7 @@ static int snd_wavefront_midi_output_close(struct snd_rawmidi_substream *substre
> return -EIO;
>
> guard(spinlock_irqsave)(&midi->open);
> + midi->substream_output[mpu] = NULL;
> midi->mode[mpu] &= ~MPU401_MODE_OUTPUT;
> return 0;
> }
> --
> 2.51.1.dirty
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-11-05 8:41 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-11-04 14:11 [PATCH] ALSA: wavefront: Fix use-after-free in MIDI operations moonafterrain
2025-11-05 8:41 ` Takashi Iwai
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox