At Wed, 17 Oct 2012 12:43:44 +0200, David Henningsson wrote:
Even when CONFIG_SND_DEBUG is not enabled, we don't want to return an arbitrary memory location when the channel count is larger than we expected.
Cc: stable@kernel.org Signed-off-by: David Henningsson david.henningsson@canonical.com
Notes: This was found while debugging an OOPS on kernel 3.2. See https://bugs.launchpad.net/ubuntu/+source/alsa-driver/+bug/1067269 for details. The badness calculation was improved in kernel 3.4; so not sure if the same could happen in later kernels, but at least apply to kernels <= 3.3 would make sense. Also adding to current trees (3.7, 3.8) would make sense IMO, just as extra safety.
Applied, thanks.
Takashi
sound/pci/hda/patch_realtek.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 2d2bb66..651822c 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -2599,8 +2599,10 @@ static const char *alc_get_line_out_pfx(struct alc_spec *spec, int ch, return "PCM"; break; }
- if (snd_BUG_ON(ch >= ARRAY_SIZE(channel_name)))
if (ch >= ARRAY_SIZE(channel_name)) {
snd_BUG();return "PCM";
}
return channel_name[ch];
}
1.7.9.5