3.16.60-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai tiwai@suse.de
commit f526afcd8f71945c23ce581d7864ace93de8a4f7 upstream.
As recently Smatch suggested, one place in RME9652 driver may expand the array directly from the user-space value with speculation: sound/pci/rme9652/rme9652.c:2074 snd_rme9652_channel_info() warn: potential spectre issue 'rme9652->channel_map' (local cap)
This patch puts array_index_nospec() for hardening against it.
BugLink: https://marc.info/?l=linux-kernel&m=152411496503418&w=2 Reported-by: Dan Carpenter dan.carpenter@oracle.com Signed-off-by: Takashi Iwai tiwai@suse.de [bwh: Backported to 3.16: adjust context] Signed-off-by: Ben Hutchings ben@decadent.org.uk --- sound/pci/rme9652/rme9652.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
--- a/sound/pci/rme9652/rme9652.c +++ b/sound/pci/rme9652/rme9652.c @@ -35,6 +35,7 @@
#include <asm/current.h> #include <asm/io.h> +#include <linux/nospec.h>
static int index[SNDRV_CARDS] = SNDRV_DEFAULT_IDX; /* Index 0-MAX */ static char *id[SNDRV_CARDS] = SNDRV_DEFAULT_STR; /* ID for this card */ @@ -2059,9 +2060,10 @@ static int snd_rme9652_channel_info(stru if (snd_BUG_ON(info->channel >= RME9652_NCHANNELS)) return -EINVAL;
- if ((chn = rme9652->channel_map[info->channel]) < 0) { + chn = rme9652->channel_map[array_index_nospec(info->channel, + RME9652_NCHANNELS)]; + if (chn < 0) return -EINVAL; - }
info->offset = chn * RME9652_CHANNEL_BUFFER_BYTES; info->first = 0;