2019-03-19 02:59:03

by Bo YU

Subject: [PATCH] drm: etnaviv: fix strncpy sizeof argument

Calling strncpy with a maximum size argument of 64 bytes on destination
array "domain->name" of size 64 bytes might leave the destination string
unterminated.

Detected by CoverityScan, CID# 1443992: Memory - illegal accesses
(BUFFER_SIZE_WARNING)

Fixes: 9e2c2e2730126 (drm/etnaviv: add infrastructure to query perf counter)
Signed-off-by: Bo YU <[email protected]>
---
drivers/gpu/drm/etnaviv/etnaviv_perfmon.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
index 4227a4006c34..08ca3c44be48 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
@@ -414,7 +414,7 @@ int etnaviv_pm_query_dom(struct etnaviv_gpu *gpu,

domain->id = domain->iter;
domain->nr_signals = dom->nr_signals;
- strncpy(domain->name, dom->name, sizeof(domain->name));
+ strncpy(domain->name, dom->name, sizeof(dom->name));

domain->iter++;
if (domain->iter == meta->nr_domains)
--
2.11.0

2019-03-19 10:24:57

by Russell King (Oracle)

Subject: Re: [PATCH] drm: etnaviv: fix strncpy sizeof argument

On Mon, Mar 18, 2019 at 10:57:55PM -0400, Bo YU wrote:
> Calling strncpy with a maximum size argument of 64 bytes on destination
> array "domain->name" of size 64 bytes might leave the destination string
> unterminated.
>
> Detected by CoverityScan, CID# 1443992: Memory - illegal accesses
> (BUFFER_SIZE_WARNING)
>
> Fixes: 9e2c2e2730126 (drm/etnaviv: add infrastructure to query perf counter)
> Signed-off-by: Bo YU <[email protected]>
> ---
> drivers/gpu/drm/etnaviv/etnaviv_perfmon.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> index 4227a4006c34..08ca3c44be48 100644
> --- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> +++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> @@ -414,7 +414,7 @@ int etnaviv_pm_query_dom(struct etnaviv_gpu *gpu,
>
> domain->id = domain->iter;
> domain->nr_signals = dom->nr_signals;
> - strncpy(domain->name, dom->name, sizeof(domain->name));
> + strncpy(domain->name, dom->name, sizeof(dom->name));

This introduces an overflow bug if sizeof(dom->name) >
sizeof(domain->name). If both sizes are the same, then there is no
effect.

strlcpy() would be a better replacement, it guarantees that the
destination will be correctly terminated.

--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
According to speedtest.net: 11.9Mbps down 500kbps up

2019-03-19 11:10:47

by Bo YU

Subject: Re: [PATCH] drm: etnaviv: fix strncpy sizeof argument

On Tue, Mar 19, 2019 at 6:22 PM Russell King - ARM Linux admin
<[email protected]> wrote:
>
> On Mon, Mar 18, 2019 at 10:57:55PM -0400, Bo YU wrote:
> > Calling strncpy with a maximum size argument of 64 bytes on destination
> > array "domain->name" of size 64 bytes might leave the destination string
> > unterminated.
> >
> > Detected by CoverityScan, CID# 1443992: Memory - illegal accesses
> > (BUFFER_SIZE_WARNING)
> >
> > Fixes: 9e2c2e2730126 (drm/etnaviv: add infrastructure to query perf counter)
> > Signed-off-by: Bo YU <[email protected]>
> > ---
> > drivers/gpu/drm/etnaviv/etnaviv_perfmon.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> > index 4227a4006c34..08ca3c44be48 100644
> > --- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> > +++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
> > @@ -414,7 +414,7 @@ int etnaviv_pm_query_dom(struct etnaviv_gpu *gpu,
> >
> > domain->id = domain->iter;
> > domain->nr_signals = dom->nr_signals;
> > - strncpy(domain->name, dom->name, sizeof(domain->name));
> > + strncpy(domain->name, dom->name, sizeof(dom->name));
>
> This introduces an overflow bug if sizeof(dom->name) >
> sizeof(domain->name). If both sizes are the same, then there is no
> effect.
Oops,It seems more worse than original code.
>
> strlcpy() would be a better replacement, it guarantees that the
> destination will be correctly terminated.
But there are too many strcpy like usage in kernel, Does it matter?
>
> --
> RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
> FTTC broadband for 0.8mile line in suburbia: sync at 12.1Mbps down 622kbps up
> According to speedtest.net: 11.9Mbps down 500kbps up