| Message ID | 1393555699-12752-1-git-send-email-Li.Xiubo@freescale.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 931f27c6e892fdfe98896055e0df7962e21969d9 |
| Headers | show |
On 02/28/2014 03:48 AM, Xiubo Li wrote: > For the snd_soc_cache_init(), the reg_size maybe zero and then the value > of codec->reg_cache, which is alloced via kzalloc, maybe equal to > ZERO_SIZE_PTR. If the reg parameter of snd_soc_cache_write() is large enough, > the cache[idx] = val maybe cause the kernel crash... > There are actually no users of snd_soc_cache_{read,write}() left. Since all drivers using snd_soc_set_cache_io() are now using native regmap the path in hw_{read,write} that does call snd_soc_cache_{read,write}() is never hit. If you want to avoid this theoretical issue just remove the functions. > So this patch fix this via doing the zero pionter check of it. > > Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com> > --- > sound/soc/soc-cache.c | 13 +++++++------ > 1 file changed, 7 insertions(+), 6 deletions(-) > > diff --git a/sound/soc/soc-cache.c b/sound/soc/soc-cache.c > index 375dc6d..bfed3e4 100644 > --- a/sound/soc/soc-cache.c > +++ b/sound/soc/soc-cache.c > @@ -96,8 +96,7 @@ int snd_soc_cache_exit(struct snd_soc_codec *codec) > { > dev_dbg(codec->dev, "ASoC: Destroying cache for %s codec\n", > codec->name); > - if (!codec->reg_cache) > - return 0; > + > kfree(codec->reg_cache); > codec->reg_cache = NULL; > return 0; > @@ -117,8 +116,9 @@ int snd_soc_cache_read(struct snd_soc_codec *codec, > return -EINVAL; > > mutex_lock(&codec->cache_rw_mutex); > - *value = snd_soc_get_cache_val(codec->reg_cache, reg, > - codec->driver->reg_word_size); > + if (!ZERO_OR_NULL_PTR(codec->reg_cache)) > + *value = snd_soc_get_cache_val(codec->reg_cache, reg, > + codec->driver->reg_word_size); > mutex_unlock(&codec->cache_rw_mutex); > > return 0; > @@ -136,8 +136,9 @@ int snd_soc_cache_write(struct snd_soc_codec *codec, > unsigned int reg, unsigned int value) > { > mutex_lock(&codec->cache_rw_mutex); > - snd_soc_set_cache_val(codec->reg_cache, reg, value, > - codec->driver->reg_word_size); > + if (!ZERO_OR_NULL_PTR(codec->reg_cache)) > + snd_soc_set_cache_val(codec->reg_cache, reg, value, > + codec->driver->reg_word_size); > mutex_unlock(&codec->cache_rw_mutex); > > return 0; >
On Fri, Feb 28, 2014 at 10:48:19AM +0800, Xiubo Li wrote: > For the snd_soc_cache_init(), the reg_size maybe zero and then the value > of codec->reg_cache, which is alloced via kzalloc, maybe equal to > ZERO_SIZE_PTR. If the reg parameter of snd_soc_cache_write() is large enough, > the cache[idx] = val maybe cause the kernel crash... Applied, thanks.
diff --git a/sound/soc/soc-cache.c b/sound/soc/soc-cache.c index 375dc6d..bfed3e4 100644 --- a/sound/soc/soc-cache.c +++ b/sound/soc/soc-cache.c @@ -96,8 +96,7 @@ int snd_soc_cache_exit(struct snd_soc_codec *codec) { dev_dbg(codec->dev, "ASoC: Destroying cache for %s codec\n", codec->name); - if (!codec->reg_cache) - return 0; + kfree(codec->reg_cache); codec->reg_cache = NULL; return 0; @@ -117,8 +116,9 @@ int snd_soc_cache_read(struct snd_soc_codec *codec, return -EINVAL; mutex_lock(&codec->cache_rw_mutex); - *value = snd_soc_get_cache_val(codec->reg_cache, reg, - codec->driver->reg_word_size); + if (!ZERO_OR_NULL_PTR(codec->reg_cache)) + *value = snd_soc_get_cache_val(codec->reg_cache, reg, + codec->driver->reg_word_size); mutex_unlock(&codec->cache_rw_mutex); return 0; @@ -136,8 +136,9 @@ int snd_soc_cache_write(struct snd_soc_codec *codec, unsigned int reg, unsigned int value) { mutex_lock(&codec->cache_rw_mutex); - snd_soc_set_cache_val(codec->reg_cache, reg, value, - codec->driver->reg_word_size); + if (!ZERO_OR_NULL_PTR(codec->reg_cache)) + snd_soc_set_cache_val(codec->reg_cache, reg, value, + codec->driver->reg_word_size); mutex_unlock(&codec->cache_rw_mutex); return 0;
For the snd_soc_cache_init(), the reg_size maybe zero and then the value of codec->reg_cache, which is alloced via kzalloc, maybe equal to ZERO_SIZE_PTR. If the reg parameter of snd_soc_cache_write() is large enough, the cache[idx] = val maybe cause the kernel crash... So this patch fix this via doing the zero pionter check of it. Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com> --- sound/soc/soc-cache.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)