* [PATCH] drm/loongson: clean up KMS polling on probe failure
@ 2026-05-11 17:01 Myeonghun Pak
2026-05-11 17:22 ` Icenowy Zheng
0 siblings, 1 reply; 3+ messages in thread
From: Myeonghun Pak @ 2026-05-11 17:01 UTC (permalink / raw)
To: dri-devel
Cc: Myeonghun Pak, Sui Jingfeng, Jianmin Lv, Qianhai Wu, Huacai Chen,
Mingcong Bai, Xi Ruoyao, Icenowy Zheng, Maarten Lankhorst,
Maxime Ripard, Thomas Zimmermann, David Airlie, Simona Vetter,
linux-kernel, stable, Ijae Kim
lsdc_pci_probe() initializes KMS polling before setting up vblank support,
requesting the IRQ and registering the DRM device. If any of those later
steps fails, probe returns without finalizing polling. The remove path has
the same lifetime gap when tearing down a successfully registered device.
Route those probe failures through a poll cleanup label. Also finalize
polling from remove before unregistering the DRM device.
This issue was identified during our ongoing static-analysis research while
reviewing kernel code.
Fixes: f39db26c5428 ("drm: Add kms driver for loongson display controller")
Cc: stable@vger.kernel.org
Co-developed-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
---
drivers/gpu/drm/loongson/lsdc_drv.c | 11 ++++++++---
1 file changed, 8 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/loongson/lsdc_drv.c b/drivers/gpu/drm/loongson/lsdc_drv.c
index abf5bf68ee..3db1f8690a 100644
--- a/drivers/gpu/drm/loongson/lsdc_drv.c
+++ b/drivers/gpu/drm/loongson/lsdc_drv.c
@@ -297,7 +297,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
if (loongson_vblank) {
ret = drm_vblank_init(ddev, descp->num_of_crtc);
if (ret)
- return ret;
+ goto err_poll_fini;
ret = devm_request_irq(&pdev->dev, pdev->irq,
descp->funcs->irq_handler,
@@ -305,7 +305,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
dev_name(&pdev->dev), ddev);
if (ret) {
drm_err(ddev, "Failed to register interrupt: %d\n", ret);
- return ret;
+ goto err_poll_fini;
}
drm_info(ddev, "registered irq: %u\n", pdev->irq);
@@ -313,17 +313,22 @@ static int lsdc_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
ret = drm_dev_register(ddev, 0);
if (ret)
- return ret;
+ goto err_poll_fini;
drm_client_setup(ddev, NULL);
return 0;
+
+err_poll_fini:
+ drm_kms_helper_poll_fini(ddev);
+ return ret;
}
static void lsdc_pci_remove(struct pci_dev *pdev)
{
struct drm_device *ddev = pci_get_drvdata(pdev);
+ drm_kms_helper_poll_fini(ddev);
drm_dev_unregister(ddev);
drm_atomic_helper_shutdown(ddev);
}
--
2.47.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] drm/loongson: clean up KMS polling on probe failure
2026-05-11 17:01 [PATCH] drm/loongson: clean up KMS polling on probe failure Myeonghun Pak
@ 2026-05-11 17:22 ` Icenowy Zheng
2026-05-12 6:26 ` Thomas Zimmermann
0 siblings, 1 reply; 3+ messages in thread
From: Icenowy Zheng @ 2026-05-11 17:22 UTC (permalink / raw)
To: Myeonghun Pak, dri-devel
Cc: Sui Jingfeng, Jianmin Lv, Qianhai Wu, Huacai Chen, Mingcong Bai,
Xi Ruoyao, Maarten Lankhorst, Maxime Ripard, Thomas Zimmermann,
David Airlie, Simona Vetter, linux-kernel, stable, Ijae Kim
在 2026-05-12二的 02:01 +0900，Myeonghun Pak写道：
> lsdc_pci_probe() initializes KMS polling before setting up vblank
> support,
> requesting the IRQ and registering the DRM device. If any of those
> later
> steps fails, probe returns without finalizing polling. The remove
> path has
> the same lifetime gap when tearing down a successfully registered
> device.
>
> Route those probe failures through a poll cleanup label. Also
> finalize
> polling from remove before unregistering the DRM device.
Interesting, but it looks like a `drmm_kms_helper_poll_init` function
exists (while rarely used).
Maybe it's better to switch to this? Or maybe there's some reason not
to use this?
Thanks,
Icenowy
>
> This issue was identified during our ongoing static-analysis research
> while
> reviewing kernel code.
>
> Fixes: f39db26c5428 ("drm: Add kms driver for loongson display
> controller")
> Cc: stable@vger.kernel.org
> Co-developed-by: Ijae Kim <ae878000@gmail.com>
> Signed-off-by: Ijae Kim <ae878000@gmail.com>
> Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
> ---
> drivers/gpu/drm/loongson/lsdc_drv.c | 11 ++++++++---
> 1 file changed, 8 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/gpu/drm/loongson/lsdc_drv.c
> b/drivers/gpu/drm/loongson/lsdc_drv.c
> index abf5bf68ee..3db1f8690a 100644
> --- a/drivers/gpu/drm/loongson/lsdc_drv.c
> +++ b/drivers/gpu/drm/loongson/lsdc_drv.c
> @@ -297,7 +297,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
> const struct pci_device_id *ent)
> if (loongson_vblank) {
> ret = drm_vblank_init(ddev, descp->num_of_crtc);
> if (ret)
> - return ret;
> + goto err_poll_fini;
>
> ret = devm_request_irq(&pdev->dev, pdev->irq,
> descp->funcs->irq_handler,
> @@ -305,7 +305,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
> const struct pci_device_id *ent)
> dev_name(&pdev->dev), ddev);
> if (ret) {
> drm_err(ddev, "Failed to register interrupt:
> %d\n", ret);
> - return ret;
> + goto err_poll_fini;
> }
>
> drm_info(ddev, "registered irq: %u\n", pdev->irq);
> @@ -313,17 +313,22 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
> const struct pci_device_id *ent)
>
> ret = drm_dev_register(ddev, 0);
> if (ret)
> - return ret;
> + goto err_poll_fini;
>
> drm_client_setup(ddev, NULL);
>
> return 0;
> +
> +err_poll_fini:
> + drm_kms_helper_poll_fini(ddev);
> + return ret;
> }
>
> static void lsdc_pci_remove(struct pci_dev *pdev)
> {
> struct drm_device *ddev = pci_get_drvdata(pdev);
>
> + drm_kms_helper_poll_fini(ddev);
> drm_dev_unregister(ddev);
> drm_atomic_helper_shutdown(ddev);
> }
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] drm/loongson: clean up KMS polling on probe failure
2026-05-11 17:22 ` Icenowy Zheng
@ 2026-05-12 6:26 ` Thomas Zimmermann
0 siblings, 0 replies; 3+ messages in thread
From: Thomas Zimmermann @ 2026-05-12 6:26 UTC (permalink / raw)
To: Icenowy Zheng, Myeonghun Pak, dri-devel
Cc: Sui Jingfeng, Jianmin Lv, Qianhai Wu, Huacai Chen, Mingcong Bai,
Xi Ruoyao, Maarten Lankhorst, Maxime Ripard, David Airlie,
Simona Vetter, linux-kernel, stable, Ijae Kim
Hi
Am 11.05.26 um 19:22 schrieb Icenowy Zheng:
> 在 2026-05-12二的 02:01 +0900，Myeonghun Pak写道：
>> lsdc_pci_probe() initializes KMS polling before setting up vblank
>> support,
>> requesting the IRQ and registering the DRM device. If any of those
>> later
>> steps fails, probe returns without finalizing polling. The remove
>> path has
>> the same lifetime gap when tearing down a successfully registered
>> device.
>>
>> Route those probe failures through a poll cleanup label. Also
>> finalize
>> polling from remove before unregistering the DRM device.
> Interesting, but it looks like a `drmm_kms_helper_poll_init` function
> exists (while rarely used).
>
> Maybe it's better to switch to this? Or maybe there's some reason not
> to use this?
Agreed. DRM drivers are advised to use managed cleanup for their data
structures. So rather switch to drmm_kms_helper_poll_init(). That also
resolves the problem that loongson never calls poll_fini on regular
removals.
Best regards
Thomas
>
> Thanks,
> Icenowy
>
>> This issue was identified during our ongoing static-analysis research
>> while
>> reviewing kernel code.
>>
>> Fixes: f39db26c5428 ("drm: Add kms driver for loongson display
>> controller")
>> Cc: stable@vger.kernel.org
>> Co-developed-by: Ijae Kim <ae878000@gmail.com>
>> Signed-off-by: Ijae Kim <ae878000@gmail.com>
>> Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
>> ---
>> drivers/gpu/drm/loongson/lsdc_drv.c | 11 ++++++++---
>> 1 file changed, 8 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/loongson/lsdc_drv.c
>> b/drivers/gpu/drm/loongson/lsdc_drv.c
>> index abf5bf68ee..3db1f8690a 100644
>> --- a/drivers/gpu/drm/loongson/lsdc_drv.c
>> +++ b/drivers/gpu/drm/loongson/lsdc_drv.c
>> @@ -297,7 +297,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
>> const struct pci_device_id *ent)
>> if (loongson_vblank) {
>> ret = drm_vblank_init(ddev, descp->num_of_crtc);
>> if (ret)
>> - return ret;
>> + goto err_poll_fini;
>>
>> ret = devm_request_irq(&pdev->dev, pdev->irq,
>> descp->funcs->irq_handler,
>> @@ -305,7 +305,7 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
>> const struct pci_device_id *ent)
>> dev_name(&pdev->dev), ddev);
>> if (ret) {
>> drm_err(ddev, "Failed to register interrupt:
>> %d\n", ret);
>> - return ret;
>> + goto err_poll_fini;
>> }
>>
>> drm_info(ddev, "registered irq: %u\n", pdev->irq);
>> @@ -313,17 +313,22 @@ static int lsdc_pci_probe(struct pci_dev *pdev,
>> const struct pci_device_id *ent)
>>
>> ret = drm_dev_register(ddev, 0);
>> if (ret)
>> - return ret;
>> + goto err_poll_fini;
>>
>> drm_client_setup(ddev, NULL);
>>
>> return 0;
>> +
>> +err_poll_fini:
>> + drm_kms_helper_poll_fini(ddev);
>> + return ret;
>> }
>>
>> static void lsdc_pci_remove(struct pci_dev *pdev)
>> {
>> struct drm_device *ddev = pci_get_drvdata(pdev);
>>
>> + drm_kms_helper_poll_fini(ddev);
>> drm_dev_unregister(ddev);
>> drm_atomic_helper_shutdown(ddev);
>> }
--
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Frankenstr. 146, 90461 Nürnberg, Germany, www.suse.com
GF: Jochen Jaser, Andrew McDonald, Werner Knoblich, (HRB 36809, AG Nürnberg)
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-05-12 6:26 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-05-11 17:01 [PATCH] drm/loongson: clean up KMS polling on probe failure Myeonghun Pak
2026-05-11 17:22 ` Icenowy Zheng
2026-05-12 6:26 ` Thomas Zimmermann
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox