* [PATCH] drm/i810: Prevent underflow in ioctl
@ 2019-10-04 10:22 Dan Carpenter
2019-10-04 14:08 ` Chris Wilson
0 siblings, 1 reply; 3+ messages in thread
From: Dan Carpenter @ 2019-10-04 10:22 UTC (permalink / raw)
To: David Airlie
Cc: Daniel Vetter, Sam Ravnborg, Emil Velikov, dri-devel,
linux-kernel, kernel-janitors
The "used" variables here come from the user in the ioctl and it can be
negative. It could result in an out of bounds write.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
---
drivers/gpu/drm/i810/i810_dma.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 2a77823b8e9a..e66c38332df4 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -728,7 +728,7 @@ static void i810_dma_dispatch_vertex(struct drm_device *dev,
if (nbox > I810_NR_SAREA_CLIPRECTS)
nbox = I810_NR_SAREA_CLIPRECTS;
- if (used > 4 * 1024)
+ if (used < 0 || used > 4 * 1024)
used = 0;
if (sarea_priv->dirty)
@@ -1048,7 +1048,7 @@ static void i810_dma_dispatch_mc(struct drm_device *dev, struct drm_buf *buf, in
if (u != I810_BUF_CLIENT)
DRM_DEBUG("MC found buffer that isn't mine!\n");
- if (used > 4 * 1024)
+ if (used < 0 || used > 4 * 1024)
used = 0;
sarea_priv->dirty = 0x7f;
--
2.20.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] drm/i810: Prevent underflow in ioctl
2019-10-04 10:22 [PATCH] drm/i810: Prevent underflow in ioctl Dan Carpenter
@ 2019-10-04 14:08 ` Chris Wilson
2019-10-04 16:04 ` Chris Wilson
0 siblings, 1 reply; 3+ messages in thread
From: Chris Wilson @ 2019-10-04 14:08 UTC (permalink / raw)
To: Dan Carpenter, David Airlie
Cc: kernel-janitors, linux-kernel, dri-devel, Sam Ravnborg,
Emil Velikov
Quoting Dan Carpenter (2019-10-04 11:22:51)
> The "used" variables here come from the user in the ioctl and it can be
> negative. It could result in an out of bounds write.
>
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---
> drivers/gpu/drm/i810/i810_dma.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
> index 2a77823b8e9a..e66c38332df4 100644
> --- a/drivers/gpu/drm/i810/i810_dma.c
> +++ b/drivers/gpu/drm/i810/i810_dma.c
> @@ -728,7 +728,7 @@ static void i810_dma_dispatch_vertex(struct drm_device *dev,
> if (nbox > I810_NR_SAREA_CLIPRECTS)
> nbox = I810_NR_SAREA_CLIPRECTS;
>
> - if (used > 4 * 1024)
> + if (used < 0 || used > 4 * 1024)
> used = 0;
Yes, as passed to the GPU instruction, negative used is invalid.
Then it is used as an offset into a memblock, where a negative offset
would be very bad.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
-Chris
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] drm/i810: Prevent underflow in ioctl
2019-10-04 14:08 ` Chris Wilson
@ 2019-10-04 16:04 ` Chris Wilson
0 siblings, 0 replies; 3+ messages in thread
From: Chris Wilson @ 2019-10-04 16:04 UTC (permalink / raw)
To: Dan Carpenter, David Airlie
Cc: kernel-janitors, linux-kernel, dri-devel, Sam Ravnborg,
Emil Velikov
Quoting Chris Wilson (2019-10-04 15:08:57)
> Quoting Dan Carpenter (2019-10-04 11:22:51)
> > The "used" variables here come from the user in the ioctl and it can be
> > negative. It could result in an out of bounds write.
> >
> > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> > ---
> > drivers/gpu/drm/i810/i810_dma.c | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
> > index 2a77823b8e9a..e66c38332df4 100644
> > --- a/drivers/gpu/drm/i810/i810_dma.c
> > +++ b/drivers/gpu/drm/i810/i810_dma.c
> > @@ -728,7 +728,7 @@ static void i810_dma_dispatch_vertex(struct drm_device *dev,
> > if (nbox > I810_NR_SAREA_CLIPRECTS)
> > nbox = I810_NR_SAREA_CLIPRECTS;
> >
> > - if (used > 4 * 1024)
> > + if (used < 0 || used > 4 * 1024)
> > used = 0;
>
> Yes, as passed to the GPU instruction, negative used is invalid.
>
> Then it is used as an offset into a memblock, where a negative offset
> would be very bad.
>
> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Applied to drm-misc-next with cc'ed stable.
-Chris
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-10-04 16:05 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-10-04 10:22 [PATCH] drm/i810: Prevent underflow in ioctl Dan Carpenter
2019-10-04 14:08 ` Chris Wilson
2019-10-04 16:04 ` Chris Wilson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox