gpu: host1x: fix uninitialized variable use
PATCH 1/1 gpu: host1x: fix uninitialized variable use
From: Arnd Bergmann <[email protected]>
The error handling for platform_get_irq() failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]
if (syncpt_irq < 0)
^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/gpu/host1x/dev.c +2 -3
@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );
@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);
Signature · 1 line
2.39.0
Quoted 7 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]
Quoted 8 lines
if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>
I had the same diff pending but civic duty called today :)
Quoted 30 lines
--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0
Mikko Perttunen
Reviewed-byQuoted 46 lines · Nathan Chancellor
Quoted 7 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]Quoted 8 lines
if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>I had the same diff pending but civic duty called today :)
Reviewed-by Nathan Chancellor <[email protected]>Quoted 25 lines
--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;
Quoted 5 lines
Quoted 5 lines
drivers/gpu/host1x/dev.c +0
-0mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0
Thanks both for the fix :)
FWIW,
Quoted 7 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]
Quoted 8 lines
if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>
Thanks Arnd, I saw some reports from kernelci about this, too.
https://lore.kernel.org/linux-next/?q=warning%3A+variable+%27syncpt_irq%27+is+uninitialized+when+used+here
Quoted 31 lines
--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0
Hi Thierry, Daniel, and David,
Quoted 45 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0
Apologies if this has been reported already or has a solution in progress but mainline is now broken because this change got separated from the change it is fixing:
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/4249931209/jobs/7391912774
https://storage.tuxsuite.com/public/clangbuiltlinux/continuous-integration2/builds/2M7y9HpiXB13qiC2mkHMyeZOcLW/build.log
I see this change sitting in the drm-tegra tree [1], which is getting merged into -next, so it is fixed there, which is why we did not notice any issues until the drm-next tree was merged into mainline. Can this be fast tracked to Linus to unbreak clang builds with -Werror?
[1]: https://gitlab.freedesktop.org/drm/tegra/-/commit/b9930311641cf2ed905a84aabe27e8f3868aee4a
Cheers,
Nathan
Ping? This warning is now in 6.3-rc1.
Quoted 62 lines · Nathan Chancellor
Hi Thierry, Daniel, and David,
Quoted 45 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0Apologies if this has been reported already or has a solution in progress but mainline is now broken because this change got separated from the change it is fixing:
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/4249931209/jobs/7391912774
https://storage.tuxsuite.com/public/clangbuiltlinux/continuous-integration2/builds/2M7y9HpiXB13qiC2mkHMyeZOcLW/build.logI see this change sitting in the drm-tegra tree [1], which is getting merged into -next, so it is fixed there, which is why we did not notice any issues until the drm-next tree was merged into mainline. Can this be fast tracked to Linus to unbreak clang builds with -Werror?
[1]: https://gitlab.freedesktop.org/drm/tegra/-/commit/b9930311641cf2ed905a84aabe27e8f3868aee4a
Quoted 1 line · Nathan Chancellor
Ping? This warning is now in 6.3-rc1.
Thierry is away at the moment.
David, Daniel, do you want to pick this up directly in the meantime as a fix for 6.3? Mikko has already reviewed and FWIW ...
Thanks
Jon
Quoted 63 lines · Nathan Chancellor
Quoted 62 lines
Hi Thierry, Daniel, and David,
Quoted 45 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0Apologies if this has been reported already or has a solution in progress but mainline is now broken because this change got separated from the change it is fixing:
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/4249931209/jobs/7391912774
https://storage.tuxsuite.com/public/clangbuiltlinux/continuous-integration2/builds/2M7y9HpiXB13qiC2mkHMyeZOcLW/build.logI see this change sitting in the drm-tegra tree [1], which is getting merged into -next, so it is fixed there, which is why we did not notice any issues until the drm-next tree was merged into mainline. Can this be fast tracked to Linus to unbreak clang builds with -Werror?
[1]: https://gitlab.freedesktop.org/drm/tegra/-/commit/b9930311641cf2ed905a84aabe27e8f3868aee4a
Signature · 1 line
nvpublic
Quoted 9 lines · Jon Hunter
Quoted 1 line · Nathan Chancellor
Ping? This warning is now in 6.3-rc1.
Thierry is away at the moment.
David, Daniel, do you want to pick this up directly in the meantime as a fix for 6.3? Mikko has already reviewed and FWIW ...
Generally first fallback should be drm-misc maintainers (or anyone else with commit rights), but since this fell through cracks for weeks it seems I'll pick it up to drm-fixes now directly.
-Daniel
Quoted 70 lines
Reviewed-by Jon Hunter <[email protected]>Thanks
JonQuoted 60 lines · Nathan Chancellor
Quoted 59 lines
Hi Thierry, Daniel, and David,
Quoted 42 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-3@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices);Signature · 1 line
2.39.0Apologies if this has been reported already or has a solution in progress but mainline is now broken because this change got separated from the change it is fixing:
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/4249931209/jobs/7391912774
https://storage.tuxsuite.com/public/clangbuiltlinux/continuous-integration2/builds/2M7y9HpiXB13qiC2mkHMyeZOcLW/build.logI see this change sitting in the drm-tegra tree [1], which is getting merged into -next, so it is fixed there, which is why we did not notice any issues until the drm-next tree was merged into mainline. Can this be fast tracked to Linus to unbreak clang builds with -Werror?
[1]: https://gitlab.freedesktop.org/drm/tegra/-/commit/b9930311641cf2ed905a84aabe27e8f3868aee4a
Signature · 1 line
nvpublic
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
Quoted 7 lines · Daniel Vetter
Quoted 5 lines · Jon Hunter
Quoted 1 line · Nathan Chancellor
Ping? This warning is now in 6.3-rc1.
Quoted 8 lines
Quoted 4 lines
Thierry is away at the moment.
David, Daniel, do you want to pick this up directly in the meantime as a fix for 6.3? Mikko has already reviewed and FWIW ...
Generally first fallback should be drm-misc maintainers (or anyone else with commit rights), but since this fell through cracks for weeks it seems I'll pick it up to drm-fixes now directly.
Ok I think I found out why this fell through cracks, MAINTAINERS isn't updated that host1x&tegra is maintained in drm-misc.
John, since Thierry is out, can you pls create the MAINTAINERS patch to
- point at drm-misc git repo everywhere needed
- add any missing host1x paths to the drm-misc entry so that get_maintainers.pl adds the right people for this patch (currently it doesn't)
Also should we have at least a 2nd person for tegra stuff (or well maybe nvidia stuff in general) for drm-misc? Currently it's just Thierry, and I don't think that's enough. Whomever gets volunteered please follow
https://drm.pages.freedesktop.org/maintainer-tools/commit-access.html#drm-misc
Thanks, Daniel
Quoted 76 lines
Quoted 67 lines
Reviewed-by Jon Hunter <[email protected]>Thanks
JonQuoted 59 lines · Nathan Chancellor
Quoted 58 lines
Hi Thierry, Daniel, and David,
Quoted 40 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The error handling for
platform_get_irq()failing no longer works after a recent change, clang now points this out with a warning:
drivers/gpu/host1x/dev.c:520:6: error: variable 'syncpt_irq' is uninitialized when used here [-Werror,-Wuninitialized]if (syncpt_irq < 0) ^~~~~~~~~~Fix this by removing the variable and checking the correct error status.
Fixes 625d4ffb438c ("gpu: host1x: Rewrite syncpoint interrupt handling")Signed-off-by Arnd Bergmann <[email protected]>--- drivers/gpu/host1x/dev.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)drivers/gpu/host1x/dev.c +2
-4@@ -487,7 +487,6 @@ static int host1x_get_resets(struct host1x *host)static int host1x_probe (struct platform_device *pdev) {struct host1x *host; -int syncpt_irq;int err; host =devm_kzalloc (&pdev->dev,sizeof (*host),GFP_KERNEL );@@ -517,8 +516,8 @@ static int host1x_probe(struct platform_device *pdev) } host->syncpt_irq =platform_get_irq (pdev,0 ); -if (syncpt_irq <0 ) -return syncpt_irq; +if (host->syncpt_irq <0 ) +return host->syncpt_irq;mutex_init (&host->devices_lock);INIT_LIST_HEAD (&host->devices); --Quoted 2 lines
2.39.0
Apologies if this has been reported already or has a solution in progress but mainline is now broken because this change got separated from the change it is fixing:
https://github.com/ClangBuiltLinux/continuous-integration2/actions/runs/4249931209/jobs/7391912774
https://storage.tuxsuite.com/public/clangbuiltlinux/continuous-integration2/builds/2M7y9HpiXB13qiC2mkHMyeZOcLW/build.logI see this change sitting in the drm-tegra tree [1], which is getting merged into -next, so it is fixed there, which is why we did not notice any issues until the drm-next tree was merged into mainline. Can this be fast tracked to Linus to unbreak clang builds with -Werror?
Quoted 4 lines
Quoted 1 line
--
nvpublic--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch
Quoted 26 lines · Daniel Vetter
Quoted 7 lines · Daniel Vetter
Quoted 5 lines · Jon Hunter
Quoted 1 line · Nathan Chancellor
Ping? This warning is now in 6.3-rc1.
Quoted 8 lines
Quoted 4 lines
Thierry is away at the moment.
David, Daniel, do you want to pick this up directly in the meantime as a fix for 6.3? Mikko has already reviewed and FWIW ...
Generally first fallback should be drm-misc maintainers (or anyone else with commit rights), but since this fell through cracks for weeks it seems I'll pick it up to drm-fixes now directly.
Ok I think I found out why this fell through cracks,
MAINTAINERSisn't updated that host1x&tegra is maintained in drm-misc.John, since Thierry is out, can you pls create the
MAINTAINERSpatch to
- point at drm-misc git repo everywhere needed
- add any missing host1x paths to the drm-misc entry so that get_maintainers.pl adds the right people for this patch (currently it doesn't)
Quoted 6 lines
Also should we have at least a 2nd person for tegra stuff (or well maybe nvidia stuff in general) for drm-misc? Currently it's just Thierry, and I don't think that's enough. Whomever gets volunteered please follow
https://drm.pages.freedesktop.org/maintainer-tools/commit-access.html#drm-misc
We never actually moved host1x and Tegra DRM to drm-misc completely. There's still a separate tree that feeds into linux-next. This made sense a while ago because there was a lot going on, but development speed has decreased recently, so we may want to fix that. The big benefit is that drm-misc is better oiled than drm-tegra, so getting things merged would be easier for everyone.
I'm all in favor of adding a second Tegra person to take pick up the slack when necessary.
Thierry
Quoted 33 lines · Thierry Reding
Quoted 31 lines · Daniel Vetter
Quoted 7 lines · Daniel Vetter
Quoted 5 lines · Jon Hunter
Quoted 1 line · Nathan Chancellor
Ping? This warning is now in 6.3-rc1.
Quoted 8 lines
Quoted 4 lines
Thierry is away at the moment.
David, Daniel, do you want to pick this up directly in the meantime as a fix for 6.3? Mikko has already reviewed and FWIW ...
Generally first fallback should be drm-misc maintainers (or anyone else with commit rights), but since this fell through cracks for weeks it seems I'll pick it up to drm-fixes now directly.
Ok I think I found out why this fell through cracks,
MAINTAINERSisn't updated that host1x&tegra is maintained in drm-misc.John, since Thierry is out, can you pls create the
MAINTAINERSpatch to
- point at drm-misc git repo everywhere needed
- add any missing host1x paths to the drm-misc entry so that get_maintainers.pl adds the right people for this patch (currently it doesn't)Also should we have at least a 2nd person for tegra stuff (or well maybe nvidia stuff in general) for drm-misc? Currently it's just Thierry, and I don't think that's enough. Whomever gets volunteered please follow
Quoted 8 lines
We never actually moved host1x and Tegra DRM to drm-misc completely. There's still a separate tree that feeds into linux-next. This made sense a while ago because there was a lot going on, but development speed has decreased recently, so we may want to fix that. The big benefit is that drm-misc is better oiled than drm-tegra, so getting things merged would be easier for everyone.
Hm I thought you had a split where you feed -fixes through misc and features through tegra.git? Just listing both trees would cover that.
Quoted 2 lines
I'm all in favor of adding a second Tegra person to take pick up the slack when necessary.
Imo features still through tegra.git with just you is fine, there's not going to be anyone else (like ci farms) blocked if that's held up for a bit. It's more the bugfixes where a 2nd nvidia person would be great.
-Daniel
Signature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch