[PATCH 0/2] Two small perfcnt fixes for Panfrost

[PATCH 0/2] Two small perfcnt fixes for Panfrost

[Adrián Larumbe]

To: Boris Brezillon <boris.brezillon@collabora.com>
To: Rob Herring <robh@kernel.org>
To: Steven Price <steven.price@arm.com>
To: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
To: Maxime Ripard <mripard@kernel.org>
To: Thomas Zimmermann <tzimmermann@suse.de>
To: David Airlie <airlied@gmail.com>
To: Simona Vetter <simona@ffwll.ch>
Cc: dri-devel@lists.freedesktop.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Adrián Larumbe <adrian.larumbe@collabora.com>

First fix makes sure cycle counter is enabled when starting a new
session. This is required by some profiling frameworks like
Perfetto. Second fix deals with a recently-introduced race condition.

---
Adrián Larumbe (1):
      drm/panfrost: Fix race when restoring reset state in perfcnt

Boris Brezillon (1):
      drm/panfrost: Enable cycle counting when perfcnt is enabled

 drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)
---
base-commit: 9944dd07f3cf7b6c19d65c2585147887ca06677c
change-id: 20261014-panfrost-reset-races-fixes-6f5f2459dc9b

Best regards,
--  
Adrián Larumbe <adrian.larumbe@collabora.com>

[PATCH 1/2] drm/panfrost: Enable cycle counting when perfcnt is enabled

[Adrián Larumbe]

From: Boris Brezillon <boris.brezillon@collabora.com>

If we don't do that, we end up with invalid timestamps in the block
headers.

Fixes: 7786fd108777 ("drm/panfrost: Expose performance counters through unstable ioctls")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
---
 drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
index a9def452d5ef..b26da057a1a0 100644
--- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
+++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
@@ -17,6 +17,7 @@
 #include "panfrost_device.h"
 #include "panfrost_features.h"
 #include "panfrost_gem.h"
+#include "panfrost_gpu.h"
 #include "panfrost_issues.h"
 #include "panfrost_job.h"
 #include "panfrost_mmu.h"
@@ -52,6 +53,7 @@ static void panfrost_perfcnt_hw_disable(struct panfrost_device *pfdev)
 	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
 
 	if (perfcnt->owns_as_ref) {
+		panfrost_cycle_counter_put(pfdev);
 		panfrost_mmu_as_put(pfdev, perfcnt->mapping->mmu);
 		perfcnt->owns_as_ref = false;
 	}
@@ -84,6 +86,7 @@ static int panfrost_perfcnt_hw_enable(struct panfrost_device *pfdev)
 		return ret;
 
 	perfcnt->owns_as_ref = true;
+	panfrost_cycle_counter_get(pfdev);
 
 	as = ret;
 	cfg = GPU_PERFCNT_CFG_AS(as) |

-- 
2.55.0

[PATCH 2/2] drm/panfrost: Fix race when restoring reset state in perfcnt

[Adrián Larumbe]

struct panfrost_perfcnt::state should always be updated atomically, both
in the dump ioctl and the reset sequence.

Fixes: 9944dd07f3cf ("drm/panfrost: Fix races between perfcnt and reset sequence")
Signed-off-by: Adrián Larumbe <adrian.larumbe@collabora.com>
---
 drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
index b26da057a1a0..d50b6cc2ecfa 100644
--- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
+++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
@@ -148,18 +148,17 @@ static int panfrost_perfcnt_dump_locked(struct panfrost_device *pfdev, u32 *stat
 	 */
 	ret = wait_for_completion_interruptible_timeout(&pfdev->perfcnt->dump_comp,
 							msecs_to_jiffies(1000));
-
 	if (!ret)
 		ret = -ETIMEDOUT;
 
-	/* We need to restore perfcnt state so the next successful dump can relay it to UM */
-	if (ret < 0) {
-		perfcnt->state |= *state;
-		*state = 0;
-		return ret;
-	}
-
 	scoped_guard(rwsem_read, &pfdev->reset.lock) {
+		/* Restore perfcnt state so the next successful dump can relay it to UM */
+		if (ret < 0) {
+			perfcnt->state |= *state;
+			*state = 0;
+			return ret;
+		}
+
 		*state |= perfcnt->state;
 
 		/* UM must re-enable their session before requesting new dumps. */

-- 
2.55.0

Re: [PATCH 1/2] drm/panfrost: Enable cycle counting when perfcnt is enabled

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] drm/panfrost: Spurious drm_WARN_ON during GPU reset with active perfcnt
--

commit d9ae50cec3ff06c2a614193efb813114ed5ac709
Author: Boris Brezillon <boris.brezillon@collabora.com>
Date:   Thu Oct 15 18:22:31 2026 +0000

    drm/panfrost: Enable cycle counting when perfcnt is enabled

    If we don't do that, we end up with invalid timestamps in the block
    headers.

> diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> index a9def452d5ef9..b26da057a1a01 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c

[ ... ]

> @@ -84,6 +86,7 @@ static int panfrost_perfcnt_hw_enable(struct panfrost_device *pfdev)
>  		return ret;
>  
>  	perfcnt->owns_as_ref = true;
> +	panfrost_cycle_counter_get(pfdev);

[Severity: Medium]
Will this cause a spurious drm_WARN_ON() during a GPU reset?

When a GPU reset is triggered (e.g. by a job timeout), panfrost_device_reset()
is called. This function calls panfrost_gpu_soft_reset() before it handles
the perfcnt state via panfrost_perfcnt_reset():

drivers/gpu/drm/panfrost/panfrost_device.c:panfrost_device_reset() {
    ...
	panfrost_gpu_soft_reset(pfdev);
    ...
	panfrost_perfcnt_reset(pfdev);
}

Inside panfrost_gpu_soft_reset(), it expects all cycle counter references
to be released:

drivers/gpu/drm/panfrost/panfrost_gpu.c:panfrost_gpu_soft_reset() {
    ...
	/*
	 * All in-flight jobs should have released their cycle
	 * counter references upon reset, but let us make sure
	 */
	if (drm_WARN_ON(&pfdev->base, atomic_read(&pfdev->cycle_counter.use_count) != 0))
		atomic_set(&pfdev->cycle_counter.use_count, 0);
    ...
}

Because the perfcnt still holds the reference when panfrost_gpu_soft_reset()
is called, won't this trigger the drm_WARN_ON() and potentially panic the kernel
if panic_on_warn is enabled?

>  
>  	as = ret;
>  	cfg = GPU_PERFCNT_CFG_AS(as) |

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261015-panfrost-reset-races-fixes-v1-0-bb74c1786a0b@collabora.com?part=1

Re: [PATCH 1/2] drm/panfrost: Enable cycle counting when perfcnt is enabled

[Boris Brezillon]

On Thu, 15 Oct 2026 19:22:31 +0100
Adrián Larumbe <adrian.larumbe@collabora.com> wrote:

> From: Boris Brezillon <boris.brezillon@collabora.com>
> 
> If we don't do that, we end up with invalid timestamps in the block
> headers.
> 
> Fixes: 7786fd108777 ("drm/panfrost: Expose performance counters through unstable ioctls")
> Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>

Let's make sure this is merged in the same release the perfcnt uAPI
changes are in, so userspace can check the driver version to know if
the fix is present or not (we have a workaround in mesa for that which
we can drop when the fix is in).

> ---
>  drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> index a9def452d5ef..b26da057a1a0 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> @@ -17,6 +17,7 @@
>  #include "panfrost_device.h"
>  #include "panfrost_features.h"
>  #include "panfrost_gem.h"
> +#include "panfrost_gpu.h"
>  #include "panfrost_issues.h"
>  #include "panfrost_job.h"
>  #include "panfrost_mmu.h"
> @@ -52,6 +53,7 @@ static void panfrost_perfcnt_hw_disable(struct panfrost_device *pfdev)
>  	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
>  
>  	if (perfcnt->owns_as_ref) {
> +		panfrost_cycle_counter_put(pfdev);
>  		panfrost_mmu_as_put(pfdev, perfcnt->mapping->mmu);
>  		perfcnt->owns_as_ref = false;
>  	}
> @@ -84,6 +86,7 @@ static int panfrost_perfcnt_hw_enable(struct panfrost_device *pfdev)
>  		return ret;
>  
>  	perfcnt->owns_as_ref = true;
> +	panfrost_cycle_counter_get(pfdev);
>  
>  	as = ret;
>  	cfg = GPU_PERFCNT_CFG_AS(as) |
> 

Re: [PATCH 1/2] drm/panfrost: Enable cycle counting when perfcnt is enabled

[Boris Brezillon]

On Thu, 15 Oct 2026 19:22:31 +0100
Adrián Larumbe <adrian.larumbe@collabora.com> wrote:

> From: Boris Brezillon <boris.brezillon@collabora.com>
> 
> If we don't do that, we end up with invalid timestamps in the block
> headers.
> 
> Fixes: 7786fd108777 ("drm/panfrost: Expose performance counters through unstable ioctls")
> Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>

Also, you need to add your SoB here.

> ---
>  drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> index a9def452d5ef..b26da057a1a0 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> @@ -17,6 +17,7 @@
>  #include "panfrost_device.h"
>  #include "panfrost_features.h"
>  #include "panfrost_gem.h"
> +#include "panfrost_gpu.h"
>  #include "panfrost_issues.h"
>  #include "panfrost_job.h"
>  #include "panfrost_mmu.h"
> @@ -52,6 +53,7 @@ static void panfrost_perfcnt_hw_disable(struct panfrost_device *pfdev)
>  	gpu_write(pfdev, GPU_PRFCNT_TILER_EN, 0);
>  
>  	if (perfcnt->owns_as_ref) {
> +		panfrost_cycle_counter_put(pfdev);
>  		panfrost_mmu_as_put(pfdev, perfcnt->mapping->mmu);
>  		perfcnt->owns_as_ref = false;
>  	}
> @@ -84,6 +86,7 @@ static int panfrost_perfcnt_hw_enable(struct panfrost_device *pfdev)
>  		return ret;
>  
>  	perfcnt->owns_as_ref = true;
> +	panfrost_cycle_counter_get(pfdev);
>  
>  	as = ret;
>  	cfg = GPU_PERFCNT_CFG_AS(as) |
> 

Re: [PATCH 2/2] drm/panfrost: Fix race when restoring reset state in perfcnt

[Boris Brezillon]

On Thu, 15 Oct 2026 19:22:32 +0100
Adrián Larumbe <adrian.larumbe@collabora.com> wrote:

> struct panfrost_perfcnt::state should always be updated atomically, both
> in the dump ioctl and the reset sequence.
> 
> Fixes: 9944dd07f3cf ("drm/panfrost: Fix races between perfcnt and reset sequence")
> Signed-off-by: Adrián Larumbe <adrian.larumbe@collabora.com>

Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com>

> ---
>  drivers/gpu/drm/panfrost/panfrost_perfcnt.c | 15 +++++++--------
>  1 file changed, 7 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> index b26da057a1a0..d50b6cc2ecfa 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_perfcnt.c
> @@ -148,18 +148,17 @@ static int panfrost_perfcnt_dump_locked(struct panfrost_device *pfdev, u32 *stat
>  	 */
>  	ret = wait_for_completion_interruptible_timeout(&pfdev->perfcnt->dump_comp,
>  							msecs_to_jiffies(1000));
> -
>  	if (!ret)
>  		ret = -ETIMEDOUT;
>  
> -	/* We need to restore perfcnt state so the next successful dump can relay it to UM */
> -	if (ret < 0) {
> -		perfcnt->state |= *state;
> -		*state = 0;
> -		return ret;
> -	}
> -
>  	scoped_guard(rwsem_read, &pfdev->reset.lock) {
> +		/* Restore perfcnt state so the next successful dump can relay it to UM */
> +		if (ret < 0) {
> +			perfcnt->state |= *state;
> +			*state = 0;
> +			return ret;
> +		}
> +
>  		*state |= perfcnt->state;
>  
>  		/* UM must re-enable their session before requesting new dumps. */
>