forked from Minki/linux
76c4718322
The master and next_conj of rcs_ops are used for iterating the resource list entries, and currently those are supposed to return the current value. The problem is that next_conf may go over the last entry before the loop abort condition is evaluated, and it may return the "current" value that is beyond the array size. It was caught recently as a GPF, for example. Those return values are, however, never actually evaluated, hence basically we don't have to consider the current value as the return at all. By dropping those return values, the potential out-of-range access above is also fixed automatically. This patch changes the return type of master and next_conj callbacks to void and drop the superfluous code accordingly. BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=214985 Cc: <stable@vger.kernel.org> Link: https://lore.kernel.org/r/20211118215729.26257-1-tiwai@suse.de Signed-off-by: Takashi Iwai <tiwai@suse.de> |
||
|---|---|---|
| .. | ||
| ct20k1reg.h | ||
| ct20k2reg.h | ||
| ctamixer.c | ||
| ctamixer.h | ||
| ctatc.c | ||
| ctatc.h | ||
| ctdaio.c | ||
| ctdaio.h | ||
| cthardware.c | ||
| cthardware.h | ||
| cthw20k1.c | ||
| cthw20k1.h | ||
| cthw20k2.c | ||
| cthw20k2.h | ||
| ctimap.c | ||
| ctimap.h | ||
| ctmixer.c | ||
| ctmixer.h | ||
| ctpcm.c | ||
| ctpcm.h | ||
| ctresource.c | ||
| ctresource.h | ||
| ctsrc.c | ||
| ctsrc.h | ||
| cttimer.c | ||
| cttimer.h | ||
| ctvmem.c | ||
| ctvmem.h | ||
| Makefile | ||
| xfi.c | ||