Skip to contentSkip to conversation

venus: core: add shutdown callback for venus

PATCH Superseded Mansur Alisha Shaik 3 participants 5 messages Jun 13, 2020 – Aug 6, 2020

Subsystems (4)

v1 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 venus: core: add shutdown callback for venus

After the SMMU translation is disabled in the
arm-smmu shutdown callback during reboot, if
any subsystem are still alive then IOVAs they
are using will become PAs on bus, which may
lead to crash.

Below are the consumers of smmu from venus
arm-smmu: consumer: aa00000.video-codec supplier=15000000.iommu
arm-smmu: consumer: video-firmware.0 supplier=15000000.iommu

So implemented shutdown callback, which detach iommu maps.

Change-Id I0f0f331056e0b84b92f1d86f66618d4b1caaa24a
---
 drivers/media/platform/qcom/venus/core.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/media/platform/qcom/venus/core.c +9 -0
index 30d4b9e..acf798c 100644
@@ -371,6 +371,14 @@ static int venus_remove(struct platform_device *pdev)
 	return ret;
 }
 
+static void venus_core_shutdown(struct platform_device *pdev)
+{
+	int ret;
+
+	ret = venus_remove(pdev);
+	WARN_ON(ret < 0);
+}
+
 static __maybe_unused int venus_runtime_suspend(struct device *dev)
 {
 	struct venus_core *core = dev_get_drvdata(dev);
@@ -628,6 +636,7 @@ static struct platform_driver qcom_venus_driver = {
 		.of_match_table = venus_dt_match,
 		.pm = &venus_pm_ops,
 	},
+	.shutdown = venus_core_shutdown,
 };
 module_platform_driver(qcom_venus_driver);
 
Signature · 1 line
2.7.4
#1 lore raw

Hi Mansur,

Thanks for the patch!

How you test this? Is it enough to start playback and issue reboot (did you test with reboot -f) ?

Quoted 46 lines · Mansur Alisha Shaik

After the SMMU translation is disabled in the
arm-smmu shutdown callback during reboot, if
any subsystem are still alive then IOVAs they
are using will become PAs on bus, which may
lead to crash.

Below are the consumers of smmu from venus
arm-smmu: consumer: aa00000.video-codec supplier=15000000.iommu
arm-smmu: consumer: video-firmware.0 supplier=15000000.iommu

So implemented shutdown callback, which detach iommu maps.

Change-Id I0f0f331056e0b84b92f1d86f66618d4b1caaa24a
---
 drivers/media/platform/qcom/venus/core.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/media/platform/qcom/venus/core.c +9 -0
index 30d4b9e..acf798c 100644
@@ -371,6 +371,14 @@ static int venus_remove(struct platform_device *pdev)
 	return ret;
 }
 
+static void venus_core_shutdown(struct platform_device *pdev)
+{
+	int ret;
+
+	ret = venus_remove(pdev);
+	WARN_ON(ret < 0);
+}
+
 static __maybe_unused int venus_runtime_suspend(struct device *dev)
 {
 	struct venus_core *core = dev_get_drvdata(dev);
@@ -628,6 +636,7 @@ static struct platform_driver qcom_venus_driver = {
 		.of_match_table = venus_dt_match,
 		.pm = &venus_pm_ops,
 	},
+	.shutdown = venus_core_shutdown,
 };
 module_platform_driver(qcom_venus_driver);
 
Signature · 2 lines
regards,
Stan
#2 lore raw

Hi Stan,

Quoted 7 lines · Stanimir Varbanov
Quoted 6 lines

Hi Mansur,

Thanks for the patch!

How you test this? Is it enough to start playback and issue reboot (did you test with reboot -f) ?

Yes, I have tested it with "reboot -f" and started video playback (YouTube browser and local video) and issue reboot command.

Quoted 41 lines
Quoted 41 lines · Mansur Alisha Shaik
Quoted 40 lines

After the SMMU translation is disabled in the
arm-smmu shutdown callback during reboot, if
any subsystem are still alive then IOVAs they
are using will become PAs on bus, which may
lead to crash.

Below are the consumers of smmu from venus
arm-smmu: consumer: aa00000.video-codec supplier=15000000.iommu
arm-smmu: consumer: video-firmware.0 supplier=15000000.iommu

So implemented shutdown callback, which detach iommu maps.

Change-Id I0f0f331056e0b84b92f1d86f66618d4b1caaa24a
---
 drivers/media/platform/qcom/venus/core.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/media/platform/qcom/venus/core.c +8 -0
index 30d4b9e..acf798c 100644
@@ -371,6 +371,14 @@ static int venus_remove(struct platform_device 
 *pdev)
 	return ret;
 }
 
+static void venus_core_shutdown(struct platform_device *pdev)
+{
+	int ret;
+
+	ret = venus_remove(pdev);
+	WARN_ON(ret < 0);
+}
+
 static __maybe_unused int venus_runtime_suspend(struct device *dev)
 {
drivers/media/platform/qcom/venus/core.c +0 -0
 	struct venus_core *core = dev_get_drvdata(dev);
@@ -628,6 +636,7 @@ static struct platform_driver qcom_venus_driver = 
 {
Quoted 8 lines
Quoted 8 lines

drivers/media/platform/qcom/venus/core.c static struct platform_driver qcom_venus_driver = { · line 641

Quoted 8 lines
drivers/media/platform/qcom/venus/core.c +1 -0
 		.of_match_table = venus_dt_match,
 		.pm = &venus_pm_ops,
 	},
+	.shutdown = venus_core_shutdown,
 };
 module_platform_driver(qcom_venus_driver);

Thank,
Mansur

#3 lore raw
In reply to Mansur Alisha Shaik

Hi Mansur,

Quoted 24 lines · Mansur Alisha Shaik

After the SMMU translation is disabled in the
arm-smmu shutdown callback during reboot, if
any subsystem are still alive then IOVAs they
are using will become PAs on bus, which may
lead to crash.

Below are the consumers of smmu from venus
arm-smmu: consumer: aa00000.video-codec supplier=15000000.iommu
arm-smmu: consumer: video-firmware.0 supplier=15000000.iommu

So implemented shutdown callback, which detach iommu maps.

Change-Id I0f0f331056e0b84b92f1d86f66618d4b1caaa24a
---
 drivers/media/platform/qcom/venus/core.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/media/platform/qcom/venus/core.c +0 -0
index 30d4b9e..acf798c 100644

drivers/media/platform/qcom/venus/core.c static int venus_remove(struct platform_device · line 380

Quoted 10 lines
drivers/media/platform/qcom/venus/core.c +6 -0
@@ -371,6 +371,14 @@ static int venus_remove(struct platform_device 
 *pdev)
 	return ret;
 }
 
+static void venus_core_shutdown(struct platform_device *pdev)
+{
+	int ret;
+
+	ret = venus_remove(pdev);
+	WARN_ON(ret < 0);

I don't think you should warn here, its shutdown path and you can't do anything with this WARN unlike remove callback where you have to be sure to cleanup properly so that you are able to reload module. But if you still want a hint about this failure, then just add a dev_err() to indicate the failure instead of a big stack trace spamming kernel log.

Thanks,
Sai

Signature · 2 lines
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member
of Code Aurora Forum, hosted by The Linux Foundation
#4 lore raw

Hi Sai,

Quoted 32 lines · Sai Prakash Ranjan
Quoted 31 lines

Hi Mansur,

Quoted 28 lines · Mansur Alisha Shaik

After the SMMU translation is disabled in the
arm-smmu shutdown callback during reboot, if
any subsystem are still alive then IOVAs they
are using will become PAs on bus, which may
lead to crash.

Below are the consumers of smmu from venus
arm-smmu: consumer: aa00000.video-codec supplier=15000000.iommu
arm-smmu: consumer: video-firmware.0 supplier=15000000.iommu

So implemented shutdown callback, which detach iommu maps.

Change-Id I0f0f331056e0b84b92f1d86f66618d4b1caaa24a
---
 drivers/media/platform/qcom/venus/core.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/media/platform/qcom/venus/core.c +0 -0
index 30d4b9e..acf798c 100644
@@ -371,6 +371,14 @@ static int venus_remove(struct platform_device 
 *pdev)
 	return ret;
 }
 
Quoted 13 lines
Quoted 4 lines
Quoted 4 lines
drivers/media/platform/qcom/venus/core.c +4 -0
+static void venus_core_shutdown(struct platform_device *pdev)
+{
+	int ret;
+
Quoted 9 lines

drivers/media/platform/qcom/venus/core.c static int venus_remove(struct platform_device *pdev) · line 379

Quoted 2 lines
drivers/media/platform/qcom/venus/core.c +2 -0
+	ret = venus_remove(pdev);
+	WARN_ON(ret < 0);

I don't think you should warn here, its shutdown path and you can't do anything with this WARN unlike remove callback where you have to be sure to cleanup properly so that you are able to reload module. But if you still want a hint about this failure, then just add a dev_err() to indicate the failure instead of a big stack trace spamming kernel log.

posted V2 version by adding dev_warn during shutdown failure instead of WARN_ON. V2 version : https://lore.kernel.org/patchwork/patch/1284693/

Quoted 2 lines
Quoted 2 lines

Thanks,
Sai

---
Thanks,
Mansur