media: use pm_runtime_resume_and_get() on non-i2c drivers

PATCH In review Mauro Carvalho Chehab 6 participants 38 messages May 6, 2021 – May 18, 2021

Subsystems (24)

Media (V4L/DVB) Arm Qualcomm mailing list Arm i.MX and Layerscape Staging +20 more subsystems

lore

Share & tools

lore

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #0 lore raw

v5 cover media: use pm_runtime_resume_and_get() on non-i2c drivers

Dealing with PM runtime (RPM) is different than dealing with other kAPIs used on media, as most pm_runtime_get_*() functions won't return to the the state before the call if an error rises. They, instead, increment an usage_count.

Due to that, there were several bugs on media. Just on this review, we found 24 such errors.

So, let's use pm_runtime_resume_and_get() whenever possible, as it has two advantages over :

1. On errors, it decrements the usage count;
2. It always return zero on success or an error code. This prevents a common error pattern of checking if ret is not zero to identify for errors.

There are however a few places where calls to pm_runtime_get_sync() are kept. On several of those, a comment was added, in order to help preventing trivial patches that could try to change them.

PS.: This series was submitted already together with the fix patches at:

https://lore.kernel.org/linux-media/[email protected]/

I opted to break it on 3 parts, in order to make easier to review.

This is the third (and final) part.

Mauro Carvalho Chehab (25):
staging: media: imx7-mipi-csis: use pm_runtime_resume_and_get()
staging: media: atomisp: use pm_runtime_resume_and_get()
staging: media: ipu3: use pm_runtime_resume_and_get()
staging: media: cedrus_video: use pm_runtime_resume_and_get()
staging: media: tegra-vde: use pm_runtime_resume_and_get()
staging: media: tegra-video: use pm_runtime_resume_and_get()
media: rockchip/rga: use pm_runtime_resume_and_get()
media: sti/hva: use pm_runtime_resume_and_get()
media: ipu3: use pm_runtime_resume_and_get()
media: coda: use pm_runtime_resume_and_get()
media: mtk-jpeg: use pm_runtime_resume_and_get()
media: camss: use pm_runtime_resume_and_get()
media: venus: core: use pm_runtime_resume_and_get()
media: venus: vdec: use pm_runtime_resume_and_get()
media: venus: venc: use pm_runtime_resume_and_get()
media: rcar-fcp: use pm_runtime_resume_and_get()
media: rkisp1: use pm_runtime_resume_and_get()
media: s3c-camif: use pm_runtime_resume_and_get()
media: s5p-mfc: use pm_runtime_resume_and_get()
media: stm32: use pm_runtime_resume_and_get()
media: sunxi: use pm_runtime_resume_and_get()
media: ti-vpe: use pm_runtime_resume_and_get()
media: vsp1: use pm_runtime_resume_and_get()
media: rcar-vin: use pm_runtime_resume_and_get()
media: hantro: use pm_runtime_resume_and_get()

 drivers/media/pci/intel/ipu3/ipu3-cio2-main.c |  3 +--
 drivers/media/platform/coda/coda-common.c     |  7 ++++---
 .../media/platform/mtk-jpeg/mtk_jpeg_core.c   |  4 ++--
 .../media/platform/qcom/camss/camss-csid.c    |  6 ++----
 .../media/platform/qcom/camss/camss-csiphy.c  |  6 ++----
 .../media/platform/qcom/camss/camss-ispif.c   |  6 ++----
 drivers/media/platform/qcom/camss/camss-vfe.c |  5 +++--
 .../media/platform/qcom/venus/pm_helpers.c    |  3 +--
 drivers/media/platform/qcom/venus/vdec.c      |  6 +++---
 drivers/media/platform/qcom/venus/venc.c      |  5 +++--
 drivers/media/platform/rcar-fcp.c             |  8 +------
 drivers/media/platform/rcar-vin/rcar-csi2.c   | 15 ++++++++++---
 drivers/media/platform/rcar-vin/rcar-dma.c    |  6 ++----
 drivers/media/platform/rcar-vin/rcar-v4l2.c   |  6 ++----
 drivers/media/platform/rockchip/rga/rga-buf.c |  3 +--
 drivers/media/platform/rockchip/rga/rga.c     |  4 +++-
 .../platform/rockchip/rkisp1/rkisp1-capture.c |  3 +--
 .../media/platform/s3c-camif/camif-capture.c  |  2 +-
 drivers/media/platform/s3c-camif/camif-core.c |  5 +++--
 drivers/media/platform/s5p-mfc/s5p_mfc_pm.c   |  6 ++----
 drivers/media/platform/sti/hva/hva-hw.c       | 17 ++++++++-------
 drivers/media/platform/stm32/stm32-dcmi.c     |  5 +++--
 .../platform/sunxi/sun4i-csi/sun4i_v4l2.c     |  6 ++++--
 drivers/media/platform/ti-vpe/cal-video.c     |  4 +++-
 drivers/media/platform/ti-vpe/cal.c           |  8 ++++---
 drivers/media/platform/ti-vpe/vpe.c           |  8 +++----
 drivers/media/platform/vsp1/vsp1_drv.c        | 10 +--------
 .../staging/media/atomisp/pci/atomisp_fops.c  |  6 +++---
 drivers/staging/media/hantro/hantro_drv.c     |  5 ++---
 drivers/staging/media/imx/imx7-mipi-csis.c    |  7 +++----
 drivers/staging/media/ipu3/ipu3.c             |  3 +--
 .../staging/media/sunxi/cedrus/cedrus_video.c |  6 ++----
 drivers/staging/media/tegra-vde/vde.c         | 21 ++++++++++++++++---
 drivers/staging/media/tegra-video/csi.c       |  3 +--
 drivers/staging/media/tegra-video/vi.c        |  3 +--
 35 files changed, 110 insertions(+), 111 deletions(-)

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #1 lore raw

PATCH v5 1/25 staging: media: imx7-mipi-csis: use pm_runtime_resume_and_get()

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.

Use the new API, in order to cleanup the error check logic.

Acked-by Rui Miguel Silva <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/imx/imx7-mipi-csis.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

drivers/staging/media/imx/imx7-mipi-csis.c +3 -4

index 025fdc488bd6..1dc680d94a46 100644
@@ -695,11 +695,10 @@ static int mipi_csis_s_stream(struct v4l2_subdev *mipi_sd, int enable)
 
 		mipi_csis_clear_counters(state);
 
-		ret = pm_runtime_get_sync(&state->pdev->dev);
-		if (ret < 0) {
-			pm_runtime_put_noidle(&state->pdev->dev);
+		ret = pm_runtime_resume_and_get(&state->pdev->dev);
+		if (ret < 0)
 			return ret;
-		}
+
 		ret = v4l2_subdev_call(state->src_sd, core, s_power, 1);
 		if (ret < 0 && ret != -ENOIOCTLCMD)
 			goto done;

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #2 lore raw

PATCH v5 2/25 staging: media: atomisp: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Besides that, the de-init order in case of css error was wrong. This change should also fix that.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/atomisp/pci/atomisp_fops.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

drivers/staging/media/atomisp/pci/atomisp_fops.c +3 -3

index f1e6b2597853..26d05474a035 100644
@@ -837,7 +837,7 @@ static int atomisp_open(struct file *file)
 	}
 
 	/* runtime power management, turn on ISP */
-	ret = pm_runtime_get_sync(vdev->v4l2_dev->dev);
+	ret = pm_runtime_resume_and_get(vdev->v4l2_dev->dev);
 	if (ret < 0) {
 		dev_err(isp->dev, "Failed to power on device\n");
 		goto error;

@@ -881,9 +881,9 @@ static int atomisp_open(struct file *file)
 
 css_error:
 	atomisp_css_uninit(isp);
-error:
-	hmm_pool_unregister(HMM_POOL_TYPE_DYNAMIC);
 	pm_runtime_put(vdev->v4l2_dev->dev);
+error:
+	hmm_pool_unregister(HMM_POOL_TYPE_DYNAMIC);
 	rt_mutex_unlock(&isp->mutex);
 	return ret;
 }

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #3 lore raw

PATCH v5 3/25 staging: media: ipu3: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Jonathan Cameron <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/ipu3/ipu3.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/staging/media/ipu3/ipu3.c +1 -2

index ee1bba6bdcac..8e1e9e46e604 100644
@@ -392,10 +392,9 @@ int imgu_s_stream(struct imgu_device *imgu, int enable)
 	}
 
 	/* Set Power */
-	r = pm_runtime_get_sync(dev);
+	r = pm_runtime_resume_and_get(dev);
 	if (r < 0) {
 		dev_err(dev, "failed to set imgu power\n");
-		pm_runtime_put(dev);
 		return r;
 	}

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #4 lore raw

PATCH v5 4/25 staging: media: cedrus_video: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Acked-by Jonathan Cameron <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/sunxi/cedrus/cedrus_video.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

drivers/staging/media/sunxi/cedrus/cedrus_video.c +2 -4

index b62eb8e84057..9ddd789d0b1f 100644
@@ -490,11 +490,9 @@ static int cedrus_start_streaming(struct vb2_queue *vq, unsigned int count)
 	}
 
 	if (V4L2_TYPE_IS_OUTPUT(vq->type)) {
-		ret = pm_runtime_get_sync(dev->dev);
-		if (ret < 0) {
-			pm_runtime_put_noidle(dev->dev);
+		ret = pm_runtime_resume_and_get(dev->dev);
+		if (ret < 0)
 			goto err_cleanup;
-		}
 
 		if (dev->dec_ops[ctx->current_codec]->start) {
 			ret = dev->dec_ops[ctx->current_codec]->start(ctx);

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #5 lore raw

PATCH v5 5/25 staging: media: tegra-vde: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/tegra-vde/vde.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

drivers/staging/media/tegra-vde/vde.c +18 -3

index 28845b5bafaf..e025b69776f2 100644
@@ -775,9 +775,9 @@ static int tegra_vde_ioctl_decode_h264(struct tegra_vde *vde,
 	if (ret)
 		goto release_dpb_frames;
 
-	ret = pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
 	if (ret < 0)
-		goto put_runtime_pm;
+		goto unlock;
 
 	/*
 	 * We rely on the VDE registers reset value, otherwise VDE

@@ -843,6 +843,8 @@ static int tegra_vde_ioctl_decode_h264(struct tegra_vde *vde,
 put_runtime_pm:
 	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);
+
+unlock:
 	mutex_unlock(&vde->lock);
 
 release_dpb_frames:

@@ -1069,11 +1071,19 @@ static int tegra_vde_probe(struct platform_device *pdev)
 	 * power-cycle it in order to put hardware into a predictable lower
 	 * power state.
 	 */
-	pm_runtime_get_sync(dev);
+	if (pm_runtime_resume_and_get(dev) < 0)
+		goto err_pm_runtime;
+
 	pm_runtime_put(dev);
 
 	return 0;
 
+err_pm_runtime:
+	misc_deregister(&vde->miscdev);
+
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+
 err_deinit_iommu:
 	tegra_vde_iommu_deinit(vde);

@@ -1089,7 +1099,12 @@ static int tegra_vde_remove(struct platform_device *pdev)
 	struct tegra_vde *vde = platform_get_drvdata(pdev);
 	struct device *dev = &pdev->dev;
 
+	/*
+	 * As it increments RPM usage_count even on errors, we don't need to
+	 * check the returned code here.
+	 */
 	pm_runtime_get_sync(dev);
+
 	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);

Signature · 1 line

2.30.2

Dan Carpenter

2021-05-07 05:14 UTC #6 lore raw

Re: [PATCH v5 05/25] staging: media: tegra-vde: use pm_runtime_resume_and_get()

drivers/staging/media/tegra-vde/vde.c static int tegra_vde_probe(struct platform_device *pdev) · line 1075

Quoted 7 lines · Mauro Carvalho Chehab

drivers/staging/media/tegra-vde/vde.c +2 -1

@@ -1069,11 +1071,19 @@ static int tegra_vde_probe(struct platform_device *pdev)
 	 * power-cycle it in order to put hardware into a predictable lower
 	 * power state.
 	 */
-	pm_runtime_get_sync(dev);
+	if (pm_runtime_resume_and_get(dev) < 0)
+		goto err_pm_runtime;

Needs an error code on this path. These days the kbuild bot will send a warning for this eventually.

Quoted 4 lines

drivers/staging/media/tegra-vde/vde.c +1 -0
+
 	pm_runtime_put(dev);
 
 	return 0;

drivers/staging/media/tegra-vde/vde.c static int tegra_vde_probe(struct platform_device *pdev) · line 1089

Quoted 10 lines

drivers/staging/media/tegra-vde/vde.c +6 -0

 
+err_pm_runtime:
+	misc_deregister(&vde->miscdev);
+
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+
 err_deinit_iommu:
 	tegra_vde_iommu_deinit(vde);

regards,
dan carpenter

Dmitry Osipenko

2021-05-10 12:26 UTC #7 lore raw

Re: [PATCH v5 05/25] staging: media: tegra-vde: use pm_runtime_resume_and_get()

07.05.2021 08:13, Dan Carpenter пишет:

Quoted 11 lines · Mauro Carvalho Chehab

drivers/staging/media/tegra-vde/vde.c static int tegra_vde_probe(struct platform_device *pdev) · line 1075
Quoted 7 lines
drivers/staging/media/tegra-vde/vde.c +2 -1
@@ -1069,11 +1071,19 @@ static int tegra_vde_probe(struct platform_device *pdev)
 	 * power-cycle it in order to put hardware into a predictable lower
 	 * power state.
 	 */
-	pm_runtime_get_sync(dev);
+	if (pm_runtime_resume_and_get(dev) < 0)
+		goto err_pm_runtime;
Needs an error code on this path. These days the kbuild bot will send a warning for this eventually.

Good catch, thank you.

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #8 lore raw

PATCH v5 6/25 staging: media: tegra-video: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Jonathan Cameron <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/tegra-video/csi.c | 3 +--
 drivers/staging/media/tegra-video/vi.c  | 3 +--
 2 files changed, 2 insertions(+), 4 deletions(-)

drivers/staging/media/tegra-video/csi.c +1 -2

index 033a6935c26d..e938bf4c48b6 100644
@@ -298,10 +298,9 @@ static int tegra_csi_enable_stream(struct v4l2_subdev *subdev)
 	struct tegra_csi *csi = csi_chan->csi;
 	int ret, err;
 
-	ret = pm_runtime_get_sync(csi->dev);
+	ret = pm_runtime_resume_and_get(csi->dev);
 	if (ret < 0) {
 		dev_err(csi->dev, "failed to get runtime PM: %d\n", ret);
-		pm_runtime_put_noidle(csi->dev);
 		return ret;
 	}

drivers/staging/media/tegra-video/vi.c +1 -2

index 7a09061cda57..1298740a9c6c 100644
@@ -297,10 +297,9 @@ static int tegra_channel_start_streaming(struct vb2_queue *vq, u32 count)
 	struct tegra_vi_channel *chan = vb2_get_drv_priv(vq);
 	int ret;
 
-	ret = pm_runtime_get_sync(chan->vi->dev);
+	ret = pm_runtime_resume_and_get(chan->vi->dev);
 	if (ret < 0) {
 		dev_err(chan->vi->dev, "failed to get runtime PM: %d\n", ret);
-		pm_runtime_put_noidle(chan->vi->dev);
 		return ret;
 	}

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #9 lore raw

PATCH v5 7/25 media: rockchip/rga: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Ezequiel Garcia <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/rockchip/rga/rga-buf.c | 3 +--
 drivers/media/platform/rockchip/rga/rga.c     | 4 +++-
 2 files changed, 4 insertions(+), 3 deletions(-)

drivers/media/platform/rockchip/rga/rga-buf.c +1 -2

index bf9a75b75083..81508ed5abf3 100644
@@ -79,9 +79,8 @@ static int rga_buf_start_streaming(struct vb2_queue *q, unsigned int count)
 	struct rockchip_rga *rga = ctx->rga;
 	int ret;
 
-	ret = pm_runtime_get_sync(rga->dev);
+	ret = pm_runtime_resume_and_get(rga->dev);
 	if (ret < 0) {
-		pm_runtime_put_noidle(rga->dev);
 		rga_buf_return_buffers(q, VB2_BUF_STATE_QUEUED);
 		return ret;
 	}

drivers/media/platform/rockchip/rga/rga.c +3 -1

index 9d122429706e..bf3fd71ec3af 100644
@@ -866,7 +866,9 @@ static int rga_probe(struct platform_device *pdev)
 		goto unreg_video_dev;
 	}
 
-	pm_runtime_get_sync(rga->dev);
+	ret = pm_runtime_resume_and_get(rga->dev);
+	if (ret < 0)
+		goto unreg_video_dev;
 
 	rga->version.major = (rga_read(rga, RGA_VERSION_INFO) >> 24) & 0xFF;
 	rga->version.minor = (rga_read(rga, RGA_VERSION_INFO) >> 20) & 0x0F;

Signature · 1 line

2.30.2

Heiko Stuebner

maintainer Reviewed-by on v5 7/25

2021-05-13 22:38 UTC #10 lore raw

Re: [PATCH v5 07/25] media: rockchip/rga: use pm_runtime_resume_and_get()

Am Donnerstag, 6. Mai 2021, 17:25:45 CEST schrieb Mauro Carvalho Chehab:

Quoted 18 lines

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Reviewed-by Ezequiel Garcia <[email protected]>
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/rockchip/rga/rga-buf.c | 3 +--
 drivers/media/platform/rockchip/rga/rga.c     | 4 +++-
 2 files changed, 4 insertions(+), 3 deletions(-)
drivers/media/platform/rockchip/rga/rga-buf.c +0 -0
index bf9a75b75083..81508ed5abf3 100644

drivers/media/platform/rockchip/rga/rga-buf.c static int rga_buf_start_streaming(struct vb2_queue *q, unsigned int count) · line 86

Quoted 10 lines

drivers/media/platform/rockchip/rga/rga-buf.c +1 -2

@@ -79,9 +79,8 @@ static int rga_buf_start_streaming(struct vb2_queue *q, unsigned int count)
 	struct rockchip_rga *rga = ctx->rga;
 	int ret;
 
-	ret = pm_runtime_get_sync(rga->dev);
+	ret = pm_runtime_resume_and_get(rga->dev);
 	if (ret < 0) {
-		pm_runtime_put_noidle(rga->dev);
 		rga_buf_return_buffers(q, VB2_BUF_STATE_QUEUED);
 		return ret;
 	}

looks ok

Quoted 5 lines

drivers/media/platform/rockchip/rga/rga.c +0 -0
index 9d122429706e..bf3fd71ec3af 100644

drivers/media/platform/rockchip/rga/rga.c static int rga_probe(struct platform_device *pdev) · line 874

Quoted 10 lines

drivers/media/platform/rockchip/rga/rga.c +3 -1

@@ -866,7 +866,9 @@ static int rga_probe(struct platform_device *pdev)
 		goto unreg_video_dev;
 	}
 
-	pm_runtime_get_sync(rga->dev);
+	ret = pm_runtime_resume_and_get(rga->dev);
+	if (ret < 0)
+		goto unreg_video_dev;
 
 	rga->version.major = (rga_read(rga, RGA_VERSION_INFO) >> 24) & 0xFF;
 	rga->version.minor = (rga_read(rga, RGA_VERSION_INFO) >> 20) & 0x0F;

hmm, the whole error handling in rga_probe looks fishy to my untrained eyes.

But I guess the easiest will be to apply your patch first and then investigate and clean up the non-matching alloc - releae calls, so

Reviewed-by Heiko Stuebner <[email protected]>

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #11 lore raw

PATCH v5 8/25 media: sti/hva: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

While the hva driver does it right, there are lots of errors on other drivers due to its misusage. So, let's change this driver to also use pm_runtime_resume_and_get(), as we're doing similar changes all over the media subsystem.

Reviewed-by Jonathan Cameron <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/sti/hva/hva-hw.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

drivers/media/platform/sti/hva/hva-hw.c +9 -8

index f59811e27f51..77b8bfa5e0c5 100644
@@ -270,9 +270,8 @@ static unsigned long int hva_hw_get_ip_version(struct hva_dev *hva)
 	struct device *dev = hva_to_dev(hva);
 	unsigned long int version;
 
-	if (pm_runtime_get_sync(dev) < 0) {
+	if (pm_runtime_resume_and_get(dev) < 0) {
 		dev_err(dev, "%s     failed to get pm_runtime\n", HVA_PREFIX);
-		pm_runtime_put_noidle(dev);
 		mutex_unlock(&hva->protect_mutex);
 		return -EFAULT;
 	}

@@ -386,10 +385,10 @@ int hva_hw_probe(struct platform_device *pdev, struct hva_dev *hva)
 	pm_runtime_set_suspended(dev);
 	pm_runtime_enable(dev);
 
-	ret = pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
 	if (ret < 0) {
 		dev_err(dev, "%s     failed to set PM\n", HVA_PREFIX);
-		goto err_pm;
+		goto err_clk;
 	}
 
 	/* check IP hardware version */

@@ -462,6 +461,7 @@ int hva_hw_execute_task(struct hva_ctx *ctx, enum hva_hw_cmd_type cmd,
 	u8 client_id = ctx->id;
 	int ret;
 	u32 reg = 0;
+	bool got_pm = false;
 
 	mutex_lock(&hva->protect_mutex);

@@ -469,12 +469,13 @@ int hva_hw_execute_task(struct hva_ctx *ctx, enum hva_hw_cmd_type cmd,
 	enable_irq(hva->irq_its);
 	enable_irq(hva->irq_err);
 
-	if (pm_runtime_get_sync(dev) < 0) {
+	if (pm_runtime_resume_and_get(dev) < 0) {
 		dev_err(dev, "%s     failed to get pm_runtime\n", ctx->name);
 		ctx->sys_errors++;
 		ret = -EFAULT;
 		goto out;
 	}
+	got_pm = true;
 
 	reg = readl_relaxed(hva->regs + HVA_HIF_REG_CLK_GATING);
 	switch (cmd) {

@@ -537,7 +538,8 @@ int hva_hw_execute_task(struct hva_ctx *ctx, enum hva_hw_cmd_type cmd,
 		dev_dbg(dev, "%s     unknown command 0x%x\n", ctx->name, cmd);
 	}
 
-	pm_runtime_put_autosuspend(dev);
+	if (got_pm)
+		pm_runtime_put_autosuspend(dev);
 	mutex_unlock(&hva->protect_mutex);
 
 	return ret;

@@ -553,9 +555,8 @@ void hva_hw_dump_regs(struct hva_dev *hva, struct seq_file *s)
 
 	mutex_lock(&hva->protect_mutex);
 
-	if (pm_runtime_get_sync(dev) < 0) {
+	if (pm_runtime_resume_and_get(dev) < 0) {
 		seq_puts(s, "Cannot wake up IP\n");
-		pm_runtime_put_noidle(dev);
 		mutex_unlock(&hva->protect_mutex);
 		return;
 	}

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #12 lore raw

PATCH v5 9/25 media: ipu3: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/pci/intel/ipu3/ipu3-cio2-main.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/media/pci/intel/ipu3/ipu3-cio2-main.c +1 -2

index fecef85bd62e..ca8040d1a725 100644
@@ -975,10 +975,9 @@ static int cio2_vb2_start_streaming(struct vb2_queue *vq, unsigned int count)
 	cio2->cur_queue = q;
 	atomic_set(&q->frame_sequence, 0);
 
-	r = pm_runtime_get_sync(&cio2->pci_dev->dev);
+	r = pm_runtime_resume_and_get(&cio2->pci_dev->dev);
 	if (r < 0) {
 		dev_info(&cio2->pci_dev->dev, "failed to set power %d\n", r);
-		pm_runtime_put_noidle(&cio2->pci_dev->dev);
 		return r;
 	}

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #13 lore raw

PATCH v5 10/25 media: coda: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

While here, as noted by Phillip, the labels at the coda_open() function are currently named after what operation failed, instead of what they do in response. So, change the name of the error label that it is called when clk_enable fails, in order to be consistent.

Reviewed-by Philipp Zabel <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/coda/coda-common.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

drivers/media/platform/coda/coda-common.c +4 -3

index bd666c858fa1..2017de85713e 100644
@@ -2660,7 +2660,7 @@ static int coda_open(struct file *file)
 	ctx->use_vdoa = false;
 
 	/* Power up and upload firmware if necessary */
-	ret = pm_runtime_get_sync(dev->dev);
+	ret = pm_runtime_resume_and_get(dev->dev);
 	if (ret < 0) {
 		v4l2_err(&dev->v4l2_dev, "failed to power up: %d\n", ret);
 		goto err_pm_get;

@@ -2668,7 +2668,7 @@ static int coda_open(struct file *file)
 
 	ret = clk_prepare_enable(dev->clk_per);
 	if (ret)
-		goto err_pm_get;
+		goto err_clk_enable;
 
 	ret = clk_prepare_enable(dev->clk_ahb);
 	if (ret)

@@ -2707,8 +2707,9 @@ static int coda_open(struct file *file)
 	clk_disable_unprepare(dev->clk_ahb);
 err_clk_ahb:
 	clk_disable_unprepare(dev->clk_per);
+err_clk_enable:
+	pm_runtime_put_sync(dev->dev);
 err_pm_get:
-	pm_runtime_put_sync(dev->dev);
 	v4l2_fh_del(&ctx->fh);
 	v4l2_fh_exit(&ctx->fh);
 err_coda_name_init:

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #14 lore raw

PATCH v5 11/25 media: mtk-jpeg: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/mtk-jpeg/mtk_jpeg_core.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

drivers/media/platform/mtk-jpeg/mtk_jpeg_core.c +2 -2

index 88a23bce569d..a89c7b206eef 100644
@@ -920,7 +920,7 @@ static void mtk_jpeg_enc_device_run(void *priv)
 	src_buf = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
 	dst_buf = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
 
-	ret = pm_runtime_get_sync(jpeg->dev);
+	ret = pm_runtime_resume_and_get(jpeg->dev);
 	if (ret < 0)
 		goto enc_end;

@@ -973,7 +973,7 @@ static void mtk_jpeg_dec_device_run(void *priv)
 		return;
 	}
 
-	ret = pm_runtime_get_sync(jpeg->dev);
+	ret = pm_runtime_resume_and_get(jpeg->dev);
 	if (ret < 0)
 		goto dec_end;

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #15 lore raw

PATCH v5 12/25 media: camss: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Robert Foss <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/qcom/camss/camss-csid.c   | 6 ++----
 drivers/media/platform/qcom/camss/camss-csiphy.c | 6 ++----
 drivers/media/platform/qcom/camss/camss-ispif.c  | 6 ++----
 drivers/media/platform/qcom/camss/camss-vfe.c    | 5 +++--
 4 files changed, 9 insertions(+), 14 deletions(-)

drivers/media/platform/qcom/camss/camss-csid.c +2 -4

index cc11fbfdae13..d2a7f2a64f26 100644
@@ -156,11 +156,9 @@ static int csid_set_power(struct v4l2_subdev *sd, int on)
 	int ret;
 
 	if (on) {
-		ret = pm_runtime_get_sync(dev);
-		if (ret < 0) {
-			pm_runtime_put_sync(dev);
+		ret = pm_runtime_resume_and_get(dev);
+		if (ret < 0)
 			return ret;
-		}
 
 		ret = regulator_enable(csid->vdda);
 		if (ret < 0) {

drivers/media/platform/qcom/camss/camss-csiphy.c +2 -4

index b3c3bf19e522..8e18b8e668cf 100644
@@ -197,11 +197,9 @@ static int csiphy_set_power(struct v4l2_subdev *sd, int on)
 	if (on) {
 		int ret;
 
-		ret = pm_runtime_get_sync(dev);
-		if (ret < 0) {
-			pm_runtime_put_sync(dev);
+		ret = pm_runtime_resume_and_get(dev);
+		if (ret < 0)
 			return ret;
-		}
 
 		ret = csiphy_set_clock_rates(csiphy);
 		if (ret < 0) {

drivers/media/platform/qcom/camss/camss-ispif.c +2 -4

index 37611c8861da..d9907742ba79 100644
@@ -372,11 +372,9 @@ static int ispif_set_power(struct v4l2_subdev *sd, int on)
 			goto exit;
 		}
 
-		ret = pm_runtime_get_sync(dev);
-		if (ret < 0) {
-			pm_runtime_put_sync(dev);
+		ret = pm_runtime_resume_and_get(dev);
+		if (ret < 0)
 			goto exit;
-		}
 
 		ret = camss_enable_clocks(ispif->nclocks, ispif->clock, dev);
 		if (ret < 0) {

drivers/media/platform/qcom/camss/camss-vfe.c +3 -2

index 15695fd466c4..cf743e61f798 100644
@@ -584,9 +584,9 @@ static int vfe_get(struct vfe_device *vfe)
 		if (ret < 0)
 			goto error_pm_domain;
 
-		ret = pm_runtime_get_sync(vfe->camss->dev);
+		ret = pm_runtime_resume_and_get(vfe->camss->dev);
 		if (ret < 0)
-			goto error_pm_runtime_get;
+			goto error_domain_off;
 
 		ret = vfe_set_clock_rates(vfe);
 		if (ret < 0)

@@ -620,6 +620,7 @@ static int vfe_get(struct vfe_device *vfe)
 
 error_pm_runtime_get:
 	pm_runtime_put_sync(vfe->camss->dev);
+error_domain_off:
 	vfe->ops->pm_domain_off(vfe);
 
 error_pm_domain:

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #16 lore raw

PATCH v5 13/25 media: venus: core: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/qcom/venus/pm_helpers.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/media/platform/qcom/venus/pm_helpers.c +1 -2

index c7e1ebec47ee..d0fddf5e9a69 100644
@@ -990,9 +990,8 @@ static int core_power_v4(struct venus_core *core, int on)
 
 	if (on == POWER_ON) {
 		if (pmctrl) {
-			ret = pm_runtime_get_sync(pmctrl);
+			ret = pm_runtime_resume_and_get(pmctrl);
 			if (ret < 0) {
-				pm_runtime_put_noidle(pmctrl);
 				return ret;
 			}
 		}

Signature · 1 line

2.30.2

Stanimir Varbanov

Tested-by Acked-by on v5 13/25

2021-05-10 14:19 UTC #17 lore raw

Re: [PATCH v5 13/25] media: venus: core: use pm_runtime_resume_and_get()

Quoted 10 lines · Mauro Carvalho Chehab

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/pm_helpers.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

Tested-by Stanimir Varbanov <[email protected]>

Acked-by Stanimir Varbanov <[email protected]>

Quoted 17 lines

drivers/media/platform/qcom/venus/pm_helpers.c +1 -2

index c7e1ebec47ee..d0fddf5e9a69 100644
@@ -990,9 +990,8 @@ static int core_power_v4(struct venus_core *core, int on)
 
 	if (on == POWER_ON) {
 		if (pmctrl) {
-			ret = pm_runtime_get_sync(pmctrl);
+			ret = pm_runtime_resume_and_get(pmctrl);
 			if (ret < 0) {
-				pm_runtime_put_noidle(pmctrl);
 				return ret;
 			}
 		}

Signature · 2 lines

regards,
Stan

Stanimir Varbanov

2021-05-17 16:33 UTC #18 lore raw

Re: [PATCH v5 13/25] media: venus: core: use pm_runtime_resume_and_get()

Hi Mauro,

Quoted 9 lines · Stanimir Varbanov

Quoted 6 lines · Mauro Carvalho Chehab
Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.

Quoted 8 lines

Quoted 4 lines
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/pm_helpers.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
Tested-by Stanimir Varbanov <[email protected]>
Acked-by Stanimir Varbanov <[email protected]>

What is the plan for these venus patches. Do I need to take them through my Venus pull request for v5.14 or you will take them directly?

Quoted 18 lines

Quoted 17 lines

drivers/media/platform/qcom/venus/pm_helpers.c +1 -2

index c7e1ebec47ee..d0fddf5e9a69 100644
@@ -990,9 +990,8 @@ static int core_power_v4(struct venus_core *core, int on)
 
 	if (on == POWER_ON) {
 		if (pmctrl) {
-			ret = pm_runtime_get_sync(pmctrl);
+			ret = pm_runtime_resume_and_get(pmctrl);
 			if (ret < 0) {
-				pm_runtime_put_noidle(pmctrl);
 				return ret;
 			}
 		}

Signature · 2 lines

regards,
Stan

Mauro Carvalho Chehab

author

2021-05-18 15:20 UTC #19 lore raw

Re: [PATCH v5 13/25] media: venus: core: use pm_runtime_resume_and_get()

Em Mon, 17 May 2021 18:26:14 +0300
Stanimir Varbanov <[email protected]> escreveu:

Quoted 17 lines

Hi Mauro,
Quoted 14 lines · Stanimir Varbanov
Quoted 10 lines · Mauro Carvalho Chehab
Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/pm_helpers.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)  

Quoted 6 lines

Quoted 3 lines
Tested-by Stanimir Varbanov <[email protected]>
Acked-by Stanimir Varbanov <[email protected]>
What is the plan for these venus patches. Do I need to take them through my Venus pull request for v5.14 or you will take them directly?

Whatever works best for you. In principle, I should apply them on my tree probably tomorrow, if ok for you.

Regards,
Mauro

Quoted 8 lines

Quoted 8 lines
Quoted 8 lines
drivers/media/platform/qcom/venus/pm_helpers.c +0 -0
index c7e1ebec47ee..d0fddf5e9a69 100644
@@ -990,9 +990,8 @@ static int core_power_v4(struct venus_core *core, int on)
 

Quoted 12 lines

Quoted 11 lines
drivers/media/platform/qcom/venus/pm_helpers.c static int core_power_v4(struct venus_core *core, int on) · line 997
Quoted 10 lines
drivers/media/platform/qcom/venus/pm_helpers.c +1 -2
 	if (on == POWER_ON) {
 		if (pmctrl) {
-			ret = pm_runtime_get_sync(pmctrl);
+			ret = pm_runtime_resume_and_get(pmctrl);
 			if (ret < 0) {
-				pm_runtime_put_noidle(pmctrl);
 				return ret;
 			}
 		}

Thanks,
Mauro

Stanimir Varbanov

2021-05-18 15:49 UTC #20 lore raw

Re: [PATCH v5 13/25] media: venus: core: use pm_runtime_resume_and_get()

Hi Mauro,

Quoted 24 lines · Mauro Carvalho Chehab

Em Mon, 17 May 2021 18:26:14 +0300
Stanimir Varbanov <[email protected]> escreveu:
Quoted 21 lines
Hi Mauro,
Quoted 17 lines · Stanimir Varbanov
Quoted 10 lines · Mauro Carvalho Chehab
Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/pm_helpers.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)  
Tested-by Stanimir Varbanov <[email protected]>
Acked-by Stanimir Varbanov <[email protected]>

Quoted 6 lines

Quoted 2 lines
What is the plan for these venus patches. Do I need to take them through my Venus pull request for v5.14 or you will take them directly?
Whatever works best for you. In principle, I should apply them on my tree probably tomorrow, if ok for you.

It is perfectly fine, please apply them. Thank you!

Signature · 2 lines

regards,
Stan

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #21 lore raw

PATCH v5 14/25 media: venus: vdec: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

As a bonus, there's no need to check if ret == 1, as pm_runtime_resume_and_get() always return 0 on success.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/qcom/venus/vdec.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

drivers/media/platform/qcom/venus/vdec.c +3 -3

index ddb7cd39424e..198e47eb63f4 100644
@@ -568,10 +568,10 @@ static int vdec_pm_get(struct venus_inst *inst)
 	int ret;
 
 	mutex_lock(&core->pm_lock);
-	ret = pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
 	mutex_unlock(&core->pm_lock);
 
-	return ret < 0 ? ret : 0;
+	return ret;
 }
 
 static int vdec_pm_put(struct venus_inst *inst, bool autosuspend)

@@ -601,7 +601,7 @@ static int vdec_pm_get_put(struct venus_inst *inst)
 	mutex_lock(&core->pm_lock);
 
 	if (pm_runtime_suspended(dev)) {
-		ret = pm_runtime_get_sync(dev);
+		ret = pm_runtime_resume_and_get(dev);
 		if (ret < 0)
 			goto error;

Signature · 1 line

2.30.2

Stanimir Varbanov

Tested-by Acked-by on v5 14/25

2021-05-10 14:19 UTC #22 lore raw

Re: [PATCH v5 14/25] media: venus: vdec: use pm_runtime_resume_and_get()

Quoted 6 lines · Mauro Carvalho Chehab

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.

Quoted 7 lines

As a bonus, there's no need to check if ret == 1, as pm_runtime_resume_and_get() always return 0 on success.
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/vdec.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

Tested-by Stanimir Varbanov <[email protected]>

Acked-by Stanimir Varbanov <[email protected]>

Signature · 2 lines

regards,
Stan

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #23 lore raw

PATCH v5 15/25 media: venus: venc: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/qcom/venus/venc.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

drivers/media/platform/qcom/venus/venc.c +3 -2

index 4a7291f934b6..8dd49d4f124c 100644
@@ -1205,9 +1205,9 @@ static int venc_open(struct file *file)
 
 	venus_helper_init_instance(inst);
 
-	ret = pm_runtime_get_sync(core->dev_enc);
+	ret = pm_runtime_resume_and_get(core->dev_enc);
 	if (ret < 0)
-		goto err_put_sync;
+		goto err_free;
 
 	ret = venc_ctrl_init(inst);
 	if (ret)

@@ -1252,6 +1252,7 @@ static int venc_open(struct file *file)
 	venc_ctrl_deinit(inst);
 err_put_sync:
 	pm_runtime_put_sync(core->dev_enc);
+err_free:
 	kfree(inst);
 	return ret;
 }

Signature · 1 line

2.30.2

Stanimir Varbanov

Tested-by Acked-by on v5 15/25

2021-05-10 14:21 UTC #24 lore raw

Re: [PATCH v5 15/25] media: venus: venc: use pm_runtime_resume_and_get()

Quoted 10 lines · Mauro Carvalho Chehab

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Signed-off-by Mauro Carvalho Chehab <[email protected]>
---
 drivers/media/platform/qcom/venus/venc.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

Tested-by Stanimir Varbanov <[email protected]>

Acked-by Stanimir Varbanov <[email protected]>

Signature · 2 lines

regards,
Stan

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #25 lore raw

PATCH v5 16/25 media: rcar-fcp: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

As a bonus, pm_runtime_resume_and_get() always return 0 on success. So, the code can be simplified.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/rcar-fcp.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

drivers/media/platform/rcar-fcp.c +1 -7

index 5c03318ae07b..a3a7afc03d7b 100644
@@ -101,13 +101,7 @@ int rcar_fcp_enable(struct rcar_fcp_device *fcp)
 	if (!fcp)
 		return 0;
 
-	ret = pm_runtime_get_sync(fcp->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(fcp->dev);
-		return ret;
-	}
-
-	return 0;
+	return pm_runtime_resume_and_get(fcp->dev);
 }
 EXPORT_SYMBOL_GPL(rcar_fcp_enable);

Signature · 1 line

2.30.2

Laurent Pinchart

maintainer Reviewed-by on v5 16/25

2021-05-06 16:05 UTC #26 lore raw

Re: [PATCH v5 16/25] media: rcar-fcp: use pm_runtime_resume_and_get()

Hi Mauro,

Thank you for the patch.

Quoted 10 lines · Mauro Carvalho Chehab

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
As a bonus, pm_runtime_resume_and_get() always return 0 on success. So, the code can be simplified.
Signed-off-by Mauro Carvalho Chehab <[email protected]>

Reviewed-by Laurent Pinchart <[email protected]>

Quoted 23 lines

---
 drivers/media/platform/rcar-fcp.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

drivers/media/platform/rcar-fcp.c +1 -7

index 5c03318ae07b..a3a7afc03d7b 100644
@@ -101,13 +101,7 @@ int rcar_fcp_enable(struct rcar_fcp_device *fcp)
 	if (!fcp)
 		return 0;
 
-	ret = pm_runtime_get_sync(fcp->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(fcp->dev);
-		return ret;
-	}
-
-	return 0;
+	return pm_runtime_resume_and_get(fcp->dev);
 }
 EXPORT_SYMBOL_GPL(rcar_fcp_enable);

Signature · 3 lines

Regards,

Laurent Pinchart

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #27 lore raw

PATCH v5 17/25 media: rkisp1: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

drivers/media/platform/rockchip/rkisp1/rkisp1-capture.c +1 -2

index 5f6c9d1623e4..3730376897d9 100644
@@ -1003,9 +1003,8 @@ rkisp1_vb2_start_streaming(struct vb2_queue *queue, unsigned int count)
 	if (ret)
 		goto err_pipeline_stop;
 
-	ret = pm_runtime_get_sync(cap->rkisp1->dev);
+	ret = pm_runtime_resume_and_get(cap->rkisp1->dev);
 	if (ret < 0) {
-		pm_runtime_put_noidle(cap->rkisp1->dev);
 		dev_err(cap->rkisp1->dev, "power up failed %d\n", ret);
 		goto err_destroy_dummy;
 	}

Signature · 1 line

2.30.2

Heiko Stuebner

maintainer Reviewed-by on v5 17/25

2021-05-13 22:27 UTC #28 lore raw

Re: [PATCH v5 17/25] media: rkisp1: use pm_runtime_resume_and_get()

Am Donnerstag, 6. Mai 2021, 17:25:55 CEST schrieb Mauro Carvalho Chehab:

Quoted 7 lines

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
Signed-off-by Mauro Carvalho Chehab <[email protected]>

Reviewed-by Heiko Stuebner <[email protected]>

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #29 lore raw

PATCH v5 18/25 media: s3c-camif: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Sylwester Nawrocki <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/s3c-camif/camif-capture.c | 2 +-
 drivers/media/platform/s3c-camif/camif-core.c    | 5 +++--
 2 files changed, 4 insertions(+), 3 deletions(-)

drivers/media/platform/s3c-camif/camif-capture.c +1 -1

index 9ca49af29542..62241ec3b978 100644
@@ -547,7 +547,7 @@ static int s3c_camif_open(struct file *file)
 	if (ret < 0)
 		goto unlock;
 
-	ret = pm_runtime_get_sync(camif->dev);
+	ret = pm_runtime_resume_and_get(camif->dev);
 	if (ret < 0)
 		goto err_pm;

drivers/media/platform/s3c-camif/camif-core.c +3 -2

index 4c3c00d59c92..e1d51fd3e700 100644
@@ -460,9 +460,9 @@ static int s3c_camif_probe(struct platform_device *pdev)
 
 	pm_runtime_enable(dev);
 
-	ret = pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
 	if (ret < 0)
-		goto err_pm;
+		goto err_disable;
 
 	ret = camif_media_dev_init(camif);
 	if (ret < 0)

@@ -502,6 +502,7 @@ static int s3c_camif_probe(struct platform_device *pdev)
 	camif_unregister_media_entities(camif);
 err_pm:
 	pm_runtime_put(dev);
+err_disable:
 	pm_runtime_disable(dev);
 	camif_clk_put(camif);
 err_clk:

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #30 lore raw

PATCH v5 19/25 media: s5p-mfc: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Sylwester Nawrocki <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/s5p-mfc/s5p_mfc_pm.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

drivers/media/platform/s5p-mfc/s5p_mfc_pm.c +2 -4

index 62d2320a7218..88b7d33c9197 100644
@@ -78,11 +78,9 @@ int s5p_mfc_power_on(void)
 {
 	int i, ret = 0;
 
-	ret = pm_runtime_get_sync(pm->device);
-	if (ret < 0) {
-		pm_runtime_put_noidle(pm->device);
+	ret = pm_runtime_resume_and_get(pm->device);
+	if (ret < 0)
 		return ret;
-	}
 
 	/* clock control */
 	for (i = 0; i < pm->num_clocks; i++) {

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #31 lore raw

PATCH v5 20/25 media: stm32: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/stm32/stm32-dcmi.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

drivers/media/platform/stm32/stm32-dcmi.c +3 -2

index d9b4ad0abf0c..b33c6e7ae0a1 100644
@@ -723,11 +723,11 @@ static int dcmi_start_streaming(struct vb2_queue *vq, unsigned int count)
 	u32 val = 0;
 	int ret;
 
-	ret = pm_runtime_get_sync(dcmi->dev);
+	ret = pm_runtime_resume_and_get(dcmi->dev);
 	if (ret < 0) {
 		dev_err(dcmi->dev, "%s: Failed to start streaming, cannot get sync (%d)\n",
 			__func__, ret);
-		goto err_pm_put;
+		goto err_unlocked;
 	}
 
 	ret = media_pipeline_start(&dcmi->vdev->entity, &dcmi->pipeline);

@@ -848,6 +848,7 @@ static int dcmi_start_streaming(struct vb2_queue *vq, unsigned int count)
 
 err_pm_put:
 	pm_runtime_put(dcmi->dev);
+err_unlocked:
 	spin_lock_irq(&dcmi->irqlock);
 	/*
 	 * Return all buffers to vb2 in QUEUED state.

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #32 lore raw

PATCH v5 21/25 media: sunxi: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/sunxi/sun4i-csi/sun4i_v4l2.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

drivers/media/platform/sunxi/sun4i-csi/sun4i_v4l2.c +4 -2

index 4785faddf630..54b909987caa 100644
@@ -206,9 +206,9 @@ static int sun4i_csi_open(struct file *file)
 	if (ret)
 		return ret;
 
-	ret = pm_runtime_get_sync(csi->dev);
+	ret = pm_runtime_resume_and_get(csi->dev);
 	if (ret < 0)
-		goto err_pm_put;
+		goto err_unlock;
 
 	ret = v4l2_pipeline_pm_get(&csi->vdev.entity);
 	if (ret)

@@ -227,6 +227,8 @@ static int sun4i_csi_open(struct file *file)
 
 err_pm_put:
 	pm_runtime_put(csi->dev);
+
+err_unlock:
 	mutex_unlock(&csi->lock);
 
 	return ret;

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #33 lore raw

PATCH v5 22/25 media: ti-vpe: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/ti-vpe/cal-video.c | 4 +++-
 drivers/media/platform/ti-vpe/cal.c       | 8 +++++---
 drivers/media/platform/ti-vpe/vpe.c       | 8 +++-----
 3 files changed, 11 insertions(+), 9 deletions(-)

drivers/media/platform/ti-vpe/cal-video.c +3 -1

index 7b7436a355ee..15fb5360cf13 100644
@@ -700,7 +700,9 @@ static int cal_start_streaming(struct vb2_queue *vq, unsigned int count)
 
 	addr = vb2_dma_contig_plane_dma_addr(&buf->vb.vb2_buf, 0);
 
-	pm_runtime_get_sync(ctx->cal->dev);
+	ret = pm_runtime_resume_and_get(ctx->cal->dev);
+	if (ret < 0)
+		goto error_pipeline;
 
 	cal_ctx_set_dma_addr(ctx, addr);
 	cal_ctx_start(ctx);

drivers/media/platform/ti-vpe/cal.c +5 -3

index 2e2bef91b2b0..76fe7a8b33f6 100644
@@ -1024,7 +1024,7 @@ static int cal_probe(struct platform_device *pdev)
 
 	/* Read the revision and hardware info to verify hardware access. */
 	pm_runtime_enable(&pdev->dev);
-	ret = pm_runtime_get_sync(&pdev->dev);
+	ret = pm_runtime_resume_and_get(&pdev->dev);
 	if (ret)
 		goto error_pm_runtime;

@@ -1098,10 +1098,11 @@ static int cal_remove(struct platform_device *pdev)
 {
 	struct cal_dev *cal = platform_get_drvdata(pdev);
 	unsigned int i;
+	int ret;
 
 	cal_dbg(1, cal, "Removing %s\n", CAL_MODULE_NAME);
 
-	pm_runtime_get_sync(&pdev->dev);
+	ret = pm_runtime_resume_and_get(&pdev->dev);
 
 	cal_media_unregister(cal);

@@ -1115,7 +1116,8 @@ static int cal_remove(struct platform_device *pdev)
 	for (i = 0; i < cal->data->num_csi2_phy; i++)
 		cal_camerarx_destroy(cal->phy[i]);
 
-	pm_runtime_put_sync(&pdev->dev);
+	if (ret >= 0)
+		pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
 	return 0;

drivers/media/platform/ti-vpe/vpe.c +3 -5

index 10251b787674..5b1c5d96a407 100644
@@ -2471,11 +2471,9 @@ static int vpe_runtime_get(struct platform_device *pdev)
 
 	dev_dbg(&pdev->dev, "vpe_runtime_get\n");
 
-	r = pm_runtime_get_sync(&pdev->dev);
+	r = pm_runtime_resume_and_get(&pdev->dev);
 	WARN_ON(r < 0);
-	if (r)
-		pm_runtime_put_noidle(&pdev->dev);
-	return r < 0 ? r : 0;
+	return r;
 }
 
 static void vpe_runtime_put(struct platform_device *pdev)

@@ -2580,7 +2578,7 @@ static int vpe_probe(struct platform_device *pdev)
 	pm_runtime_enable(&pdev->dev);
 
 	ret = vpe_runtime_get(pdev);
-	if (ret)
+	if (ret < 0)
 		goto rel_m2m;
 
 	/* Perform clk enable followed by reset */

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #34 lore raw

PATCH v5 23/25 media: vsp1: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

As a bonus, pm_runtime_resume_and_get() always return 0 on success. So, the code can be simplified.

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/vsp1/vsp1_drv.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

drivers/media/platform/vsp1/vsp1_drv.c +1 -9

index aa66e4f5f3f3..de442d6c9926 100644
@@ -559,15 +559,7 @@ static int vsp1_device_init(struct vsp1_device *vsp1)
  */
 int vsp1_device_get(struct vsp1_device *vsp1)
 {
-	int ret;
-
-	ret = pm_runtime_get_sync(vsp1->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(vsp1->dev);
-		return ret;
-	}
-
-	return 0;
+	return pm_runtime_resume_and_get(vsp1->dev);
 }
 
 /*

Signature · 1 line

2.30.2

Laurent Pinchart

maintainer Reviewed-by on v5 23/25

2021-05-06 16:06 UTC #35 lore raw

Re: [PATCH v5 23/25] media: vsp1: use pm_runtime_resume_and_get()

Hi Mauro,

Thank you for the patch.

Quoted 10 lines · Mauro Carvalho Chehab

Commit dd8088d5a896 ("PM: runtime: Add pm_runtime_resume_and_get to deal with usage counter") added pm_runtime_resume_and_get() in order to automatically handle dev->power.usage_count decrement on errors.
Use the new API, in order to cleanup the error check logic.
As a bonus, pm_runtime_resume_and_get() always return 0 on success. So, the code can be simplified.
Signed-off-by Mauro Carvalho Chehab <[email protected]>

Reviewed-by Laurent Pinchart <[email protected]>

Quoted 25 lines

---
 drivers/media/platform/vsp1/vsp1_drv.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

drivers/media/platform/vsp1/vsp1_drv.c +1 -9

index aa66e4f5f3f3..de442d6c9926 100644
@@ -559,15 +559,7 @@ static int vsp1_device_init(struct vsp1_device *vsp1)
  */
 int vsp1_device_get(struct vsp1_device *vsp1)
 {
-	int ret;
-
-	ret = pm_runtime_get_sync(vsp1->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(vsp1->dev);
-		return ret;
-	}
-
-	return 0;
+	return pm_runtime_resume_and_get(vsp1->dev);
 }
 
 /*

Signature · 3 lines

Regards,

Laurent Pinchart

Mauro Carvalho Chehab

author

2021-05-06 15:27 UTC #36 lore raw

PATCH v5 24/25 media: rcar-vin: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

Use the new API, in order to cleanup the error check logic.

Reviewed-by Niklas Söderlund <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/media/platform/rcar-vin/rcar-csi2.c | 15 ++++++++++++---
 drivers/media/platform/rcar-vin/rcar-dma.c  |  6 ++----
 drivers/media/platform/rcar-vin/rcar-v4l2.c |  6 ++----
 3 files changed, 16 insertions(+), 11 deletions(-)

drivers/media/platform/rcar-vin/rcar-csi2.c +12 -3

index e06cd512aba2..99bf814eb2a7 100644
@@ -406,10 +406,17 @@ static void rcsi2_enter_standby(struct rcar_csi2 *priv)
 	pm_runtime_put(priv->dev);
 }
 
-static void rcsi2_exit_standby(struct rcar_csi2 *priv)
+static int rcsi2_exit_standby(struct rcar_csi2 *priv)
 {
-	pm_runtime_get_sync(priv->dev);
+	int ret;
+
+	ret = pm_runtime_resume_and_get(priv->dev);
+	if (ret < 0)
+		return ret;
+
 	reset_control_deassert(priv->rstc);
+
+	return 0;
 }
 
 static int rcsi2_wait_phy_start(struct rcar_csi2 *priv,

@@ -657,7 +664,9 @@ static int rcsi2_start(struct rcar_csi2 *priv)
 {
 	int ret;
 
-	rcsi2_exit_standby(priv);
+	ret = rcsi2_exit_standby(priv);
+	if (ret < 0)
+		return ret;
 
 	ret = rcsi2_start_receiver(priv);
 	if (ret) {

drivers/media/platform/rcar-vin/rcar-dma.c +2 -4

index f30dafbdf61c..f5f722ab1d4e 100644
@@ -1458,11 +1458,9 @@ int rvin_set_channel_routing(struct rvin_dev *vin, u8 chsel)
 	u32 vnmc;
 	int ret;
 
-	ret = pm_runtime_get_sync(vin->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(vin->dev);
+	ret = pm_runtime_resume_and_get(vin->dev);
+	if (ret < 0)
 		return ret;
-	}
 
 	/* Make register writes take effect immediately. */
 	vnmc = rvin_read(vin, VNMC_REG);

drivers/media/platform/rcar-vin/rcar-v4l2.c +2 -4

index 457a65bf6b66..b1e9f86caa5c 100644
@@ -870,11 +870,9 @@ static int rvin_open(struct file *file)
 	struct rvin_dev *vin = video_drvdata(file);
 	int ret;
 
-	ret = pm_runtime_get_sync(vin->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(vin->dev);
+	ret = pm_runtime_resume_and_get(vin->dev);
+	if (ret < 0)
 		return ret;
-	}
 
 	ret = mutex_lock_interruptible(&vin->lock);
 	if (ret)

Signature · 1 line

2.30.2

Mauro Carvalho Chehab

author

2021-05-06 15:26 UTC #37 lore raw

PATCH v5 25/25 media: hantro: use pm_runtime_resume_and_get()

In reply to Mauro Carvalho Chehab

While there's nothing wrong with the current usage on this driver, as we're getting rid of the pm_runtime_get_sync() call all over the media subsystem, let's remove the last occurrence on this driver.

Reviewed-by Ezequiel Garcia <[email protected]>

Signed-off-by Mauro Carvalho Chehab <[email protected]>

---
 drivers/staging/media/hantro/hantro_drv.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

drivers/staging/media/hantro/hantro_drv.c +2 -3

index eea2009fa17b..74d6545395f9 100644
@@ -160,9 +160,8 @@ static void device_run(void *priv)
 	src = hantro_get_src_buf(ctx);
 	dst = hantro_get_dst_buf(ctx);
 
-	ret = pm_runtime_get_sync(ctx->dev->dev);
-	if (ret < 0) {
-		pm_runtime_put_noidle(ctx->dev->dev);
+	ret = pm_runtime_resume_and_get(ctx->dev->dev);
+	if (ret < 0)
 		goto err_cancel_job;
 	}

Signature · 1 line

2.30.2