| Message ID | 20180202123223.GA4410@embeddedor.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Fri, Feb 02, 2018 at 06:32:23AM -0600, Gustavo A. R. Silva wrote: > _minor_ is being dereferenced before it is null checked, hence there > is a potential null pointer dereference. Fix this by moving the pointer > dereference after _minor_ has been null checked. > > Fixes: 024ad8df763f ("drm/msm: add a5xx specific debugfs") > Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> > --- > > I wonder if a better solution for this would be to WARN_ON in case _minor_ > happens to be NULL and return -EINVAL, instead of just returning zero. > > Something like: > > struct drm_device *dev; > > if (WARN_ON(!minor) > return -EINVAL; > > dev = minor->dev; > > What do you think? In my opinion everything in debugfs is optional. I'm not sure if it is even possible for dev->primary, dev->render or dev->control to be NULL from the DRM core but if so I think the failure should be silent. Jordan > > drivers/gpu/drm/msm/adreno/a5xx_debugfs.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c > index 6b27941..059ec7d 100644 > --- a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c > +++ b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c > @@ -159,13 +159,15 @@ DEFINE_SIMPLE_ATTRIBUTE(reset_fops, NULL, reset_set, "%llx\n"); > > int a5xx_debugfs_init(struct msm_gpu *gpu, struct drm_minor *minor) > { > - struct drm_device *dev = minor->dev; > + struct drm_device *dev; > struct dentry *ent; > int ret; > > if (!minor) > return 0; > > + dev = minor->dev; > + > ret = drm_debugfs_create_files(a5xx_debugfs_list, > ARRAY_SIZE(a5xx_debugfs_list), > minor->debugfs_root, minor); > -- > 2.7.4 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-arm-msm" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Feb 2, 2018 at 11:30 AM, Jordan Crouse <jcrouse@codeaurora.org> wrote: > On Fri, Feb 02, 2018 at 06:32:23AM -0600, Gustavo A. R. Silva wrote: >> _minor_ is being dereferenced before it is null checked, hence there >> is a potential null pointer dereference. Fix this by moving the pointer >> dereference after _minor_ has been null checked. >> >> Fixes: 024ad8df763f ("drm/msm: add a5xx specific debugfs") >> Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> >> --- >> >> I wonder if a better solution for this would be to WARN_ON in case _minor_ >> happens to be NULL and return -EINVAL, instead of just returning zero. >> >> Something like: >> >> struct drm_device *dev; >> >> if (WARN_ON(!minor) >> return -EINVAL; >> >> dev = minor->dev; >> >> What do you think? > > In my opinion everything in debugfs is optional. I'm not sure if it is even > possible for dev->primary, dev->render or dev->control to be NULL from the DRM > core but if so I think the failure should be silent. > Don't have code in front of me atm, but I think this is one of those things you can hit both before and after we have the minor, depending on whether filesystem and fw are present when the driver loads or not.. so I don't think it should be a WARN_ON(). BR, -R > Jordan >> >> drivers/gpu/drm/msm/adreno/a5xx_debugfs.c | 4 +++- >> 1 file changed, 3 insertions(+), 1 deletion(-) >> >> diff --git a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >> index 6b27941..059ec7d 100644 >> --- a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >> +++ b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >> @@ -159,13 +159,15 @@ DEFINE_SIMPLE_ATTRIBUTE(reset_fops, NULL, reset_set, "%llx\n"); >> >> int a5xx_debugfs_init(struct msm_gpu *gpu, struct drm_minor *minor) >> { >> - struct drm_device *dev = minor->dev; >> + struct drm_device *dev; >> struct dentry *ent; >> int ret; >> >> if (!minor) >> return 0; >> >> + dev = minor->dev; >> + >> ret = drm_debugfs_create_files(a5xx_debugfs_list, >> ARRAY_SIZE(a5xx_debugfs_list), >> minor->debugfs_root, minor); >> -- >> 2.7.4 >> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-arm-msm" in >> the body of a message to majordomo@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html > > -- > The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, > a Linux Foundation Collaborative Project
Quoting Rob Clark <robdclark@gmail.com>: > On Fri, Feb 2, 2018 at 11:30 AM, Jordan Crouse > <jcrouse@codeaurora.org> wrote: >> On Fri, Feb 02, 2018 at 06:32:23AM -0600, Gustavo A. R. Silva wrote: >>> _minor_ is being dereferenced before it is null checked, hence there >>> is a potential null pointer dereference. Fix this by moving the pointer >>> dereference after _minor_ has been null checked. >>> >>> Fixes: 024ad8df763f ("drm/msm: add a5xx specific debugfs") >>> Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> >>> --- >>> >>> I wonder if a better solution for this would be to WARN_ON in case _minor_ >>> happens to be NULL and return -EINVAL, instead of just returning zero. >>> >>> Something like: >>> >>> struct drm_device *dev; >>> >>> if (WARN_ON(!minor) >>> return -EINVAL; >>> >>> dev = minor->dev; >>> >>> What do you think? >> >> In my opinion everything in debugfs is optional. I'm not sure if it is even >> possible for dev->primary, dev->render or dev->control to be NULL >> from the DRM >> core but if so I think the failure should be silent. >> > > Don't have code in front of me atm, but I think this is one of those > things you can hit both before and after we have the minor, depending > on whether filesystem and fw are present when the driver loads or > not.. so I don't think it should be a WARN_ON(). > I get it. Then in this case this patch should work just fine. Thank you both Jordan and Rob for the feedback. -- Gustavo > BR, > -R > > >> Jordan >>> >>> drivers/gpu/drm/msm/adreno/a5xx_debugfs.c | 4 +++- >>> 1 file changed, 3 insertions(+), 1 deletion(-) >>> >>> diff --git a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >>> b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >>> index 6b27941..059ec7d 100644 >>> --- a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >>> +++ b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c >>> @@ -159,13 +159,15 @@ DEFINE_SIMPLE_ATTRIBUTE(reset_fops, NULL, >>> reset_set, "%llx\n"); >>> >>> int a5xx_debugfs_init(struct msm_gpu *gpu, struct drm_minor *minor) >>> { >>> - struct drm_device *dev = minor->dev; >>> + struct drm_device *dev; >>> struct dentry *ent; >>> int ret; >>> >>> if (!minor) >>> return 0; >>> >>> + dev = minor->dev; >>> + >>> ret = drm_debugfs_create_files(a5xx_debugfs_list, >>> ARRAY_SIZE(a5xx_debugfs_list), >>> minor->debugfs_root, minor); >>> -- >>> 2.7.4 >>> >>> -- >>> To unsubscribe from this list: send the line "unsubscribe linux-arm-msm" in >>> the body of a message to majordomo@vger.kernel.org >>> More majordomo info at http://vger.kernel.org/majordomo-info.html >> >> -- >> The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, >> a Linux Foundation Collaborative Project
diff --git a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c index 6b27941..059ec7d 100644 --- a/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c +++ b/drivers/gpu/drm/msm/adreno/a5xx_debugfs.c @@ -159,13 +159,15 @@ DEFINE_SIMPLE_ATTRIBUTE(reset_fops, NULL, reset_set, "%llx\n"); int a5xx_debugfs_init(struct msm_gpu *gpu, struct drm_minor *minor) { - struct drm_device *dev = minor->dev; + struct drm_device *dev; struct dentry *ent; int ret; if (!minor) return 0; + dev = minor->dev; + ret = drm_debugfs_create_files(a5xx_debugfs_list, ARRAY_SIZE(a5xx_debugfs_list), minor->debugfs_root, minor);
_minor_ is being dereferenced before it is null checked, hence there is a potential null pointer dereference. Fix this by moving the pointer dereference after _minor_ has been null checked. Fixes: 024ad8df763f ("drm/msm: add a5xx specific debugfs") Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com> --- I wonder if a better solution for this would be to WARN_ON in case _minor_ happens to be NULL and return -EINVAL, instead of just returning zero. Something like: struct drm_device *dev; if (WARN_ON(!minor) return -EINVAL; dev = minor->dev; What do you think? Thanks drivers/gpu/drm/msm/adreno/a5xx_debugfs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)