Skip to contentSkip to conversation

ioat: ioat_alloc_ring() failure handling.

PATCH In review Sasha Levin 6 participants 213 messages Jan 16, 2020 – Jan 23, 2020 Page 1 of 2

Subsystems (43)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/205 ioat: ioat_alloc_ring() failure handling.

From: "[email protected]" <[email protected]>

[ Upstream commit b0b5ce1010ffc50015eaec72b0028aaae3f526bb ]

If dma_alloc_coherent() returns NULL in ioat_alloc_ring(), ring allocation must not proceed.

Until now, if the first call to dma_alloc_coherent() in ioat_alloc_ring() returned NULL, the processing could proceed, failing with NULL-pointer dereferencing further down the line.

---
 drivers/dma/ioat/dma.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/dma/ioat/dma.c +2 -1
index 1a422a8b43cf..18c011e57592 100644
@@ -377,10 +377,11 @@ ioat_alloc_ring(struct dma_chan *c, int order, gfp_t flags)
 
 		descs->virt = dma_alloc_coherent(to_dev(ioat_chan),
 						 SZ_2M, &descs->hw, flags);
-		if (!descs->virt && (i > 0)) {
+		if (!descs->virt) {
 			int idx;
 
 			for (idx = 0; idx < i; idx++) {
+				descs = &ioat_chan->descs[idx];
 				dma_free_coherent(to_dev(ioat_chan), SZ_2M,
 						  descs->virt, descs->hw);
 				descs->virt = NULL;
Signature · 1 line
2.20.1
author maintainer
#1 lore raw

PATCH 2/205 drm/panfrost: Add missing check for pfdev->regulator

From: Steven Price <[email protected]>

[ Upstream commit 52282163dfa651849e905886845bcf6850dd83c2 ]

When modifying panfrost_devfreq_target() to support a device without a regulator defined I missed the check on the error path. Let's add it.

Fixes e21dd290881b ("drm/panfrost: Enable devfreq to work without regulator")
---
 drivers/gpu/drm/panfrost/panfrost_devfreq.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/gpu/drm/panfrost/panfrost_devfreq.c +4 -2
index 12ff77dacc95..c1eb8cfe6aeb 100644
@@ -53,8 +53,10 @@ static int panfrost_devfreq_target(struct device *dev, unsigned long *freq,
 	if (err) {
 		dev_err(dev, "Cannot set frequency %lu (%d)\n", target_rate,
 			err);
-		regulator_set_voltage(pfdev->regulator, pfdev->devfreq.cur_volt,
-				      pfdev->devfreq.cur_volt);
+		if (pfdev->regulator)
+			regulator_set_voltage(pfdev->regulator,
+					      pfdev->devfreq.cur_volt,
+					      pfdev->devfreq.cur_volt);
 		return err;
 	}
 
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#2 lore raw

Re: [PATCH AUTOSEL 5.4 002/205] drm/panfrost: Add missing check for pfdev->regulator

Quoted 3 lines · Sasha Levin

From: Steven Price <[email protected]>

[ Upstream commit 52282163dfa651849e905886845bcf6850dd83c2 ]

This commit is effectively already in 5.4. Confusingly there were two versions of this upstream:

52282163dfa6 ("drm/panfrost: Add missing check for pfdev->regulator") c90f30812a79 ("drm/panfrost: Add missing check for pfdev->regulator")

It got merged both through a -fixes branch and through the normal merge window. The two copies caused a bad merge in mainline and this was effectively reverted in commit 603e398a3db2 ("drm/panfrost: Remove NULL check for regulator").

c90f30812a79 is included in v5.4 so should already be in any v5.4.y release.

Steve

Quoted 34 lines

When modifying panfrost_devfreq_target() to support a device without a regulator defined I missed the check on the error path. Let's add it.

Fixes e21dd290881b ("drm/panfrost: Enable devfreq to work without regulator")
---
 drivers/gpu/drm/panfrost/panfrost_devfreq.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/gpu/drm/panfrost/panfrost_devfreq.c +4 -2
index 12ff77dacc95..c1eb8cfe6aeb 100644
@@ -53,8 +53,10 @@ static int panfrost_devfreq_target(struct device *dev, unsigned long *freq,
 	if (err) {
 		dev_err(dev, "Cannot set frequency %lu (%d)\n", target_rate,
 			err);
-		regulator_set_voltage(pfdev->regulator, pfdev->devfreq.cur_volt,
-				      pfdev->devfreq.cur_volt);
+		if (pfdev->regulator)
+			regulator_set_voltage(pfdev->regulator,
+					      pfdev->devfreq.cur_volt,
+					      pfdev->devfreq.cur_volt);
 		return err;
 	}
 
Signature · 1 line
2.20.1
#3 lore raw

Re: [PATCH AUTOSEL 5.4 002/205] drm/panfrost: Add missing check for pfdev->regulator

Quoted 11 lines · Steven Price
Quoted 3 lines · Sasha Levin

From: Steven Price <[email protected]>

[ Upstream commit 52282163dfa651849e905886845bcf6850dd83c2 ]

This commit is effectively already in 5.4. Confusingly there were two versions of this upstream:

52282163dfa6 ("drm/panfrost: Add missing check for pfdev->regulator") c90f30812a79 ("drm/panfrost: Add missing check for pfdev->regulator")

Quoted 7 lines

It got merged both through a -fixes branch and through the normal merge window. The two copies caused a bad merge in mainline and this was effectively reverted in commit 603e398a3db2 ("drm/panfrost: Remove NULL check for regulator").

c90f30812a79 is included in v5.4 so should already be in any v5.4.y release.

Have I mentioned this month just how much I hate the way the DRM tree handles stable patches like this? This kind of fallout is a pain for stable maintainers, I dred every time I see a drm patch tagged for stable.

But we've been over this all before :(

greg k-h

#4 lore raw

Re: [PATCH AUTOSEL 5.4 002/205] drm/panfrost: Add missing check for pfdev->regulator

Quoted 20 lines · Greg KH
Quoted 11 lines · Steven Price
Quoted 3 lines · Sasha Levin

From: Steven Price <[email protected]>

[ Upstream commit 52282163dfa651849e905886845bcf6850dd83c2 ]

This commit is effectively already in 5.4. Confusingly there were two versions of this upstream:

52282163dfa6 ("drm/panfrost: Add missing check for pfdev->regulator") c90f30812a79 ("drm/panfrost: Add missing check for pfdev->regulator")

Quoted 7 lines

It got merged both through a -fixes branch and through the normal merge window. The two copies caused a bad merge in mainline and this was effectively reverted in commit 603e398a3db2 ("drm/panfrost: Remove NULL check for regulator").

c90f30812a79 is included in v5.4 so should already be in any v5.4.y release.

Quoted 6 lines

Have I mentioned this month just how much I hate the way the DRM tree handles stable patches like this? This kind of fallout is a pain for stable maintainers, I dred every time I see a drm patch tagged for stable.

But we've been over this all before :(

Another example is:

29cd13cfd762 ("drm/v3d: Fix memory leak in v3d_submit_cl_ioctl") 0d352a3a8a1f ("drm/v3d: don't leak bin job if v3d_job_init fails.")

Two fixes for a memory leak were merged so now it's a double free. I sent a patch on Jan 10 but no one responded.

regards,
dan carpenter

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

#5 lore raw

Re: [PATCH AUTOSEL 5.4 002/205] drm/panfrost: Add missing check for pfdev->regulator

Quoted 28 lines · Dan Carpenter
Quoted 20 lines · Greg KH
Quoted 11 lines · Steven Price
Quoted 3 lines · Sasha Levin

From: Steven Price <[email protected]>

[ Upstream commit 52282163dfa651849e905886845bcf6850dd83c2 ]

This commit is effectively already in 5.4. Confusingly there were two versions of this upstream:

52282163dfa6 ("drm/panfrost: Add missing check for pfdev->regulator") c90f30812a79 ("drm/panfrost: Add missing check for pfdev->regulator")

Quoted 7 lines

It got merged both through a -fixes branch and through the normal merge window. The two copies caused a bad merge in mainline and this was effectively reverted in commit 603e398a3db2 ("drm/panfrost: Remove NULL check for regulator").

c90f30812a79 is included in v5.4 so should already be in any v5.4.y release.

Quoted 6 lines

Have I mentioned this month just how much I hate the way the DRM tree handles stable patches like this? This kind of fallout is a pain for stable maintainers, I dred every time I see a drm patch tagged for stable.

But we've been over this all before :(

Quoted 7 lines

Another example is:

29cd13cfd762 ("drm/v3d: Fix memory leak in v3d_submit_cl_ioctl") 0d352a3a8a1f ("drm/v3d: don't leak bin job if v3d_job_init fails.")

Two fixes for a memory leak were merged so now it's a double free. I sent a patch on Jan 10 but no one responded.

Have a link to the patch? I can't seem to find it :(

author maintainer
#6 lore raw

PATCH 3/205 drm/v3d: don't leak bin job if v3d_job_init fails.

In reply to Sasha Levin

From: Iago Toral Quiroga <[email protected]>

[ Upstream commit 0d352a3a8a1f26168d09f7073e61bb4b328e3bb9 ]

If the initialization of the job fails we need to kfree() it before returning.

---
 drivers/gpu/drm/v3d/v3d_gem.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/v3d/v3d_gem.c +1 -0
index 19c092d75266..6316bf3646af 100644
@@ -565,6 +565,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
 		ret = v3d_job_init(v3d, file_priv, &bin->base,
 				   v3d_job_free, args->in_sync_bcl);
 		if (ret) {
+			kfree(bin);
 			v3d_job_put(&render->base);
 			kfree(bin);
 			return ret;
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
#7 lore raw

Re: [PATCH AUTOSEL 5.4 003/205] drm/v3d: don't leak bin job if v3d_job_init fails.

Hi Sasha,

please notice that there were two separate patches that addressed the same issue and applying both simultaneously leads to a double free (which is what I see is happening with this patch: see the second call to kfree(bin) right below the one added here). This issue was raised previously here:

https://lists.freedesktop.org/archives/dri-devel/2019-October/241425.html

Iago

Quoted 31 lines · Sasha Levin

From: Iago Toral Quiroga <[email protected]>

[ Upstream commit 0d352a3a8a1f26168d09f7073e61bb4b328e3bb9 ]

If the initialization of the job fails we need to kfree() it before returning.

Link:
https://patchwork.freedesktop.org/patch/msgid/[email protected]

Fixes a783a09ee76d ("drm/v3d: Refactor job management.")
---
 drivers/gpu/drm/v3d/v3d_gem.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/v3d/v3d_gem.c +1 -0
index 19c092d75266..6316bf3646af 100644
@@ -565,6 +565,7 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void
 *data,
 		ret = v3d_job_init(v3d, file_priv, &bin->base,
 				   v3d_job_free, args->in_sync_bcl);
 		if (ret) {
+			kfree(bin);
 			v3d_job_put(&render->base);
 			kfree(bin);
Quoted 1 line

return ret;

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel

author maintainer
#8 lore raw

Re: [PATCH AUTOSEL 5.4 003/205] drm/v3d: don't leak bin job if v3d_job_init fails.

Quoted 10 lines · Iago Toral

Hi Sasha,

please notice that there were two separate patches that addressed the same issue and applying both simultaneously leads to a double free (which is what I see is happening with this patch: see the second call to kfree(bin) right below the one added here). This issue was raised previously here:

https://lists.freedesktop.org/archives/dri-devel/2019-October/241425.html

I'll drop this patch for now. Any idea why upstream didn't pick up the fix yet? I see the problem still exists there.

Signature · 2 lines
Thanks,
Sasha
#9 lore raw

Re: [PATCH AUTOSEL 5.4 003/205] drm/v3d: don't leak bin job if v3d_job_init fails.

Quoted 13 lines · Sasha Levin
Quoted 12 lines · Iago Toral

Hi Sasha,

please notice that there were two separate patches that addressed the
same issue and applying both simultaneously leads to a double free (which is what I see is happening with this patch: see the second call
to kfree(bin) right below the one added here). This issue was raised
previously here:

Quoted 5 lines
Quoted 1 line

https://lists.freedesktop.org/archives/dri-devel/2019-October/241425.html

I'll drop this patch for now. Any idea why upstream didn't pick up the
fix yet? I see the problem still exists there.

+Daniel

I am not sure, when this issue was found Daniel added a few people on CC as heads-up, but maybe nobody actually got to fix the merge conflict in the end?

Iago

author maintainer
#10 lore raw

PATCH 4/205 drm: panel-lvds: Potential Oops in probe error handling

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit fb2ee9bf084bcaeff1e5be100decc0eacb4af2d5 ]

The "lvds->backlight" pointer could be NULL in situations where of_parse_phandle() returns NULL. This code is cleaner if we use the managed devm_of_find_backlight() so the clean up is automatic.

---
 drivers/gpu/drm/panel/panel-lvds.c | 21 ++++-----------------
 1 file changed, 4 insertions(+), 17 deletions(-)
drivers/gpu/drm/panel/panel-lvds.c +4 -17
index ad47cc95459e..bf5fcc3e5379 100644
@@ -197,7 +197,6 @@ static int panel_lvds_parse_dt(struct panel_lvds *lvds)
 static int panel_lvds_probe(struct platform_device *pdev)
 {
 	struct panel_lvds *lvds;
-	struct device_node *np;
 	int ret;
 
 	lvds = devm_kzalloc(&pdev->dev, sizeof(*lvds), GFP_KERNEL);
@@ -243,14 +242,9 @@ static int panel_lvds_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	np = of_parse_phandle(lvds->dev->of_node, "backlight", 0);
-	if (np) {
-		lvds->backlight = of_find_backlight_by_node(np);
-		of_node_put(np);
-
-		if (!lvds->backlight)
-			return -EPROBE_DEFER;
-	}
+	lvds->backlight = devm_of_find_backlight(lvds->dev);
+	if (IS_ERR(lvds->backlight))
+		return PTR_ERR(lvds->backlight);
 
 	/*
 	 * TODO: Handle all power supplies specified in the DT node in a generic
@@ -266,14 +260,10 @@ static int panel_lvds_probe(struct platform_device *pdev)
 
 	ret = drm_panel_add(&lvds->panel);
 	if (ret < 0)
-		goto error;
+		return ret;
 
 	dev_set_drvdata(lvds->dev, lvds);
 	return 0;
-
-error:
-	put_device(&lvds->backlight->dev);
-	return ret;
 }
 
 static int panel_lvds_remove(struct platform_device *pdev)
@@ -284,9 +274,6 @@ static int panel_lvds_remove(struct platform_device *pdev)
 
 	panel_lvds_disable(&lvds->panel);
 
-	if (lvds->backlight)
-		put_device(&lvds->backlight->dev);
-
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#11 lore raw

PATCH 5/205 scsi: lpfc: Fix hdwq sgl locks and irq handling

In reply to Sasha Levin

From: James Smart <[email protected]>

[ Upstream commit a4c21acca2be6729ecbe72eda9b08092725b0a77 ]

Many of the sgl-per-hdwq paths are locking with spin_lock_irq() and spin_unlock_irq() and may unwittingly raising irq when it shouldn't. Hard deadlocks were seen around lpfc_scsi_prep_cmnd().

Fix by converting the locks to irqsave/irqrestore.

Fixes d79c9e9d4b3d ("scsi: lpfc: Support dynamic unbounded SGL lists on G7 hardware.")
---
 drivers/scsi/lpfc/lpfc_sli.c | 38 +++++++++++++++++++++---------------
 1 file changed, 22 insertions(+), 16 deletions(-)
drivers/scsi/lpfc/lpfc_sli.c +22 -16
index 8860f41af3ff..dd542bd161d2 100644
@@ -20414,8 +20414,9 @@ lpfc_get_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 	struct sli4_hybrid_sgl *allocated_sgl = NULL;
 	struct lpfc_sli4_hdw_queue *hdwq = lpfc_buf->hdwq;
 	struct list_head *buf_list = &hdwq->sgl_list;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	if (likely(!list_empty(buf_list))) {
 		/* break off 1 chunk from the sgl_list */
@@ -20427,7 +20428,7 @@ lpfc_get_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 		}
 	} else {
 		/* allocate more */
-		spin_unlock_irq(&hdwq->hdwq_lock);
+		spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 		tmp = kmalloc_node(sizeof(*tmp), GFP_ATOMIC,
 				   cpu_to_node(smp_processor_id()));
 		if (!tmp) {
@@ -20449,7 +20450,7 @@ lpfc_get_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 			return NULL;
 		}
 
-		spin_lock_irq(&hdwq->hdwq_lock);
+		spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 		list_add_tail(&tmp->list_node, &lpfc_buf->dma_sgl_xtra_list);
 	}
 
@@ -20457,7 +20458,7 @@ lpfc_get_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 					struct sli4_hybrid_sgl,
 					list_node);
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 
 	return allocated_sgl;
 }
@@ -20481,8 +20482,9 @@ lpfc_put_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 	struct sli4_hybrid_sgl *tmp = NULL;
 	struct lpfc_sli4_hdw_queue *hdwq = lpfc_buf->hdwq;
 	struct list_head *buf_list = &hdwq->sgl_list;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	if (likely(!list_empty(&lpfc_buf->dma_sgl_xtra_list))) {
 		list_for_each_entry_safe(list_entry, tmp,
@@ -20495,7 +20497,7 @@ lpfc_put_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 		rc = -EINVAL;
 	}
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 	return rc;
 }
 
@@ -20516,8 +20518,9 @@ lpfc_free_sgl_per_hdwq(struct lpfc_hba *phba,
 	struct list_head *buf_list = &hdwq->sgl_list;
 	struct sli4_hybrid_sgl *list_entry = NULL;
 	struct sli4_hybrid_sgl *tmp = NULL;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	/* Free sgl pool */
 	list_for_each_entry_safe(list_entry, tmp,
@@ -20529,7 +20532,7 @@ lpfc_free_sgl_per_hdwq(struct lpfc_hba *phba,
 		kfree(list_entry);
 	}
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 }
 
 /**
@@ -20553,8 +20556,9 @@ lpfc_get_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 	struct fcp_cmd_rsp_buf *allocated_buf = NULL;
 	struct lpfc_sli4_hdw_queue *hdwq = lpfc_buf->hdwq;
 	struct list_head *buf_list = &hdwq->cmd_rsp_buf_list;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	if (likely(!list_empty(buf_list))) {
 		/* break off 1 chunk from the list */
@@ -20567,7 +20571,7 @@ lpfc_get_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 		}
 	} else {
 		/* allocate more */
-		spin_unlock_irq(&hdwq->hdwq_lock);
+		spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 		tmp = kmalloc_node(sizeof(*tmp), GFP_ATOMIC,
 				   cpu_to_node(smp_processor_id()));
 		if (!tmp) {
@@ -20594,7 +20598,7 @@ lpfc_get_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 		tmp->fcp_rsp = (struct fcp_rsp *)((uint8_t *)tmp->fcp_cmnd +
 				sizeof(struct fcp_cmnd));
 
-		spin_lock_irq(&hdwq->hdwq_lock);
+		spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 		list_add_tail(&tmp->list_node, &lpfc_buf->dma_cmd_rsp_list);
 	}
 
@@ -20602,7 +20606,7 @@ lpfc_get_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 					struct fcp_cmd_rsp_buf,
 					list_node);
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 
 	return allocated_buf;
 }
@@ -20627,8 +20631,9 @@ lpfc_put_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 	struct fcp_cmd_rsp_buf *tmp = NULL;
 	struct lpfc_sli4_hdw_queue *hdwq = lpfc_buf->hdwq;
 	struct list_head *buf_list = &hdwq->cmd_rsp_buf_list;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	if (likely(!list_empty(&lpfc_buf->dma_cmd_rsp_list))) {
 		list_for_each_entry_safe(list_entry, tmp,
@@ -20641,7 +20646,7 @@ lpfc_put_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 		rc = -EINVAL;
 	}
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 	return rc;
 }
 
@@ -20662,8 +20667,9 @@ lpfc_free_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 	struct list_head *buf_list = &hdwq->cmd_rsp_buf_list;
 	struct fcp_cmd_rsp_buf *list_entry = NULL;
 	struct fcp_cmd_rsp_buf *tmp = NULL;
+	unsigned long iflags;
 
-	spin_lock_irq(&hdwq->hdwq_lock);
+	spin_lock_irqsave(&hdwq->hdwq_lock, iflags);
 
 	/* Free cmd_rsp buf pool */
 	list_for_each_entry_safe(list_entry, tmp,
@@ -20676,5 +20682,5 @@ lpfc_free_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 		kfree(list_entry);
 	}
 
-	spin_unlock_irq(&hdwq->hdwq_lock);
+	spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 }
Signature · 1 line
2.20.1
author maintainer
#12 lore raw

PATCH 6/205 scsi: lpfc: Fix list corruption detected in lpfc_put_sgl_per_hdwq

In reply to Sasha Levin

From: James Smart <[email protected]>

[ Upstream commit 35a635af54ce79881eb35ba20b64dcb1e81b0389 ]

In lpfc_release_io_buf, an lpfc_io_buf is returned to the 'available' pool before any associated sgl or cmd and rsp buffers are returned via their respective 'put' routines. If xri rebalancing occurs and an lpfc_io_buf structure is reused quickly, there may be a race condition between release of old and association of new resources.

Re-ordered lpfc_release_io_buf to release sgl and cmd/rsp buffer lists before releasing the lpfc_io_buf structure for re-use.

Fixes d79c9e9d4b3d ("scsi: lpfc: Support dynamic unbounded SGL lists on G7 hardware.")
---
 drivers/scsi/lpfc/lpfc_sli.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
drivers/scsi/lpfc/lpfc_sli.c +7 -7
index dd542bd161d2..9771638b64ba 100644
@@ -20108,6 +20108,13 @@ void lpfc_release_io_buf(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_ncmd,
 	lpfc_ncmd->cur_iocbq.wqe_cmpl = NULL;
 	lpfc_ncmd->cur_iocbq.iocb_cmpl = NULL;
 
+	if (phba->cfg_xpsgl && !phba->nvmet_support &&
+	    !list_empty(&lpfc_ncmd->dma_sgl_xtra_list))
+		lpfc_put_sgl_per_hdwq(phba, lpfc_ncmd);
+
+	if (!list_empty(&lpfc_ncmd->dma_cmd_rsp_list))
+		lpfc_put_cmd_rsp_buf_per_hdwq(phba, lpfc_ncmd);
+
 	if (phba->cfg_xri_rebalancing) {
 		if (lpfc_ncmd->expedite) {
 			/* Return to expedite pool */
@@ -20172,13 +20179,6 @@ void lpfc_release_io_buf(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_ncmd,
 		spin_unlock_irqrestore(&qp->io_buf_list_put_lock,
 				       iflag);
 	}
-
-	if (phba->cfg_xpsgl && !phba->nvmet_support &&
-	    !list_empty(&lpfc_ncmd->dma_sgl_xtra_list))
-		lpfc_put_sgl_per_hdwq(phba, lpfc_ncmd);
-
-	if (!list_empty(&lpfc_ncmd->dma_cmd_rsp_list))
-		lpfc_put_cmd_rsp_buf_per_hdwq(phba, lpfc_ncmd);
 }
 
 /**
Signature · 1 line
2.20.1
author maintainer
#13 lore raw

PATCH 7/205 arm64: dts: renesas: r8a77970: Fix PWM3

In reply to Sasha Levin

From: Kieran Bingham <[email protected]>

[ Upstream commit 28a1b34c00dad4be91108369ca25ef8dc8bf850d ]

The pwm3 was incorrectly added with a compatible reference to the renesas,pwm-r8a7790 (H2) due to a single characther ommision.

Fix the compatible string.

Fixes de625477c632 ("arm64: dts: renesas: r8a779{7|8}0: add PWM support")
---
 arch/arm64/boot/dts/renesas/r8a77970.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/boot/dts/renesas/r8a77970.dtsi +1 -1
index 0cd3b376635d..4952981bb6ba 100644
@@ -652,7 +652,7 @@
 		};
 
 		pwm3: pwm@e6e33000 {
-			compatible = "renesas,pwm-r8a7790", "renesas,pwm-rcar";
+			compatible = "renesas,pwm-r8a77970", "renesas,pwm-rcar";
 			reg = <0 0xe6e33000 0 8>;
 			#pwm-cells = <2>;
 			clocks = <&cpg CPG_MOD 523>;
Signature · 1 line
2.20.1
author maintainer
#14 lore raw

PATCH 8/205 mt7601u: fix bbp version check in mt7601u_wait_bbp_ready

In reply to Sasha Levin

From: Lorenzo Bianconi <[email protected]>

[ Upstream commit 15e14f76f85f4f0eab3b8146e1cd3c58ce272823 ]

Fix bbp ready check in mt7601u_wait_bbp_ready. The issue is reported by coverity with the following error:

Logical vs. bitwise operator
The expression's value does not depend on the operands; inadvertent use of the wrong operator is a likely logic error.

Addresses-Coverity-ID: 1309441 ("Logical vs. bitwise operator")

Fixes c869f77d6abb ("add mt7601u driver")
---
 drivers/net/wireless/mediatek/mt7601u/phy.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/mediatek/mt7601u/phy.c +1 -1
index 06f5702ab4bd..d863ab4a66c9 100644
@@ -213,7 +213,7 @@ int mt7601u_wait_bbp_ready(struct mt7601u_dev *dev)
 
 	do {
 		val = mt7601u_bbp_rr(dev, MT_BBP_REG_VERSION);
-		if (val && ~val)
+		if (val && val != 0xff)
 			break;
 	} while (--i);
 
Signature · 1 line
2.20.1
author maintainer
#15 lore raw

PATCH 9/205 RDMA/counter: Prevent QP counter manual binding in auto mode

In reply to Sasha Levin

From: Mark Zhang <[email protected]>

[ Upstream commit 663912a6378a34fd4f43b8d873f0c6c6322d9d0e ]

If auto mode is configured, manual counter allocation and QP bind is not allowed.

Fixes 1bd8e0a9d0fd ("RDMA/counter: Allow manual mode configuration support")
---
 drivers/infiniband/core/counters.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/infiniband/core/counters.c +11 -1
index 023478107f0e..46dd50ff7c85 100644
@@ -466,10 +466,15 @@ static struct rdma_counter *rdma_get_counter_by_id(struct ib_device *dev,
 int rdma_counter_bind_qpn(struct ib_device *dev, u8 port,
 			  u32 qp_num, u32 counter_id)
 {
+	struct rdma_port_counter *port_counter;
 	struct rdma_counter *counter;
 	struct ib_qp *qp;
 	int ret;
 
+	port_counter = &dev->port_data[port].port_counter;
+	if (port_counter->mode.mode == RDMA_COUNTER_MODE_AUTO)
+		return -EINVAL;
+
 	qp = rdma_counter_get_qp(dev, qp_num);
 	if (!qp)
 		return -ENOENT;
@@ -506,6 +511,7 @@ int rdma_counter_bind_qpn(struct ib_device *dev, u8 port,
 int rdma_counter_bind_qpn_alloc(struct ib_device *dev, u8 port,
 				u32 qp_num, u32 *counter_id)
 {
+	struct rdma_port_counter *port_counter;
 	struct rdma_counter *counter;
 	struct ib_qp *qp;
 	int ret;
@@ -513,9 +519,13 @@ int rdma_counter_bind_qpn_alloc(struct ib_device *dev, u8 port,
 	if (!rdma_is_port_valid(dev, port))
 		return -EINVAL;
 
-	if (!dev->port_data[port].port_counter.hstats)
+	port_counter = &dev->port_data[port].port_counter;
+	if (!port_counter->hstats)
 		return -EOPNOTSUPP;
 
+	if (port_counter->mode.mode == RDMA_COUNTER_MODE_AUTO)
+		return -EINVAL;
+
 	qp = rdma_counter_get_qp(dev, qp_num);
 	if (!qp)
 		return -ENOENT;
Signature · 1 line
2.20.1
author maintainer
#16 lore raw

PATCH 10/205 RDMA/siw: Fix port number endianness in a debug message

In reply to Sasha Levin

From: Bart Van Assche <[email protected]>

[ Upstream commit 050dbddf249eee3e936b5734c30b2e1b427efdc3 ]

sin_port and sin6_port are big endian member variables. Convert these port numbers into CPU endianness before printing.

---
 drivers/infiniband/sw/siw/siw_cm.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)
drivers/infiniband/sw/siw/siw_cm.c +1 -8
index 8c1931a57f4a..0454561718d9 100644
@@ -1867,14 +1867,7 @@ static int siw_listen_address(struct iw_cm_id *id, int backlog,
 	list_add_tail(&cep->listenq, (struct list_head *)id->provider_data);
 	cep->state = SIW_EPSTATE_LISTENING;
 
-	if (addr_family == AF_INET)
-		siw_dbg(id->device, "Listen at laddr %pI4 %u\n",
-			&(((struct sockaddr_in *)laddr)->sin_addr),
-			((struct sockaddr_in *)laddr)->sin_port);
-	else
-		siw_dbg(id->device, "Listen at laddr %pI6 %u\n",
-			&(((struct sockaddr_in6 *)laddr)->sin6_addr),
-			((struct sockaddr_in6 *)laddr)->sin6_port);
+	siw_dbg(id->device, "Listen at laddr %pISp\n", laddr);
 
 	return 0;
 
Signature · 1 line
2.20.1
author maintainer
#17 lore raw

PATCH 11/205 soc/tegra: pmc: Fix crashes for hierarchical interrupts

In reply to Sasha Levin

From: Thierry Reding <[email protected]>

[ Upstream commit c9e753767a9c75d2044fb7343950a6a992d34a16 ]

Interrupts that don't have an associated wake event or GPIO wake events end up with an associate IRQ chip that is NULL and which causes IRQ code to crash. This is because we don't implicitly set the parent IRQ chip by allocating the interrupt at the parent. However, there really isn't a corresponding interrupt at the parent, so we need to work around this by setting the special no_irq_chip as the IRQ chip for these interrupts.

Fixes 19906e6b1667 ("soc/tegra: pmc: Add wake event support")
---
 drivers/soc/tegra/pmc.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)
drivers/soc/tegra/pmc.c +27 -1
index 9f9c1c677cf4..0447afa970f5 100644
@@ -1899,6 +1899,20 @@ static int tegra_pmc_irq_alloc(struct irq_domain *domain, unsigned int virq,
 							    event->id,
 							    &pmc->irq, pmc);
 
+			/*
+			 * GPIOs don't have an equivalent interrupt in the
+			 * parent controller (GIC). However some code, such
+			 * as the one in irq_get_irqchip_state(), require a
+			 * valid IRQ chip to be set. Make sure that's the
+			 * case by passing NULL here, which will install a
+			 * dummy IRQ chip for the interrupt in the parent
+			 * domain.
+			 */
+			if (domain->parent)
+				irq_domain_set_hwirq_and_chip(domain->parent,
+							      virq, 0, NULL,
+							      NULL);
+
 			break;
 		}
 	}
@@ -1908,10 +1922,22 @@ static int tegra_pmc_irq_alloc(struct irq_domain *domain, unsigned int virq,
 	 * dummy hardware IRQ number. This is used in the ->irq_set_type()
 	 * and ->irq_set_wake() callbacks to return early for these IRQs.
 	 */
-	if (i == soc->num_wake_events)
+	if (i == soc->num_wake_events) {
 		err = irq_domain_set_hwirq_and_chip(domain, virq, ULONG_MAX,
 						    &pmc->irq, pmc);
 
+		/*
+		 * Interrupts without a wake event don't have a corresponding
+		 * interrupt in the parent controller (GIC). Pass NULL for the
+		 * chip here, which causes a dummy IRQ chip to be installed
+		 * for the interrupt in the parent domain, to make this
+		 * explicit.
+		 */
+		if (domain->parent)
+			irq_domain_set_hwirq_and_chip(domain->parent, virq, 0,
+						      NULL, NULL);
+	}
+
 	return err;
 }
 
Signature · 1 line
2.20.1
author maintainer
#18 lore raw

PATCH 12/205 spi: lpspi: fix memory leak in fsl_lpspi_probe

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit 057b8945f78f76d0b04eeb5c27cd9225e5e7ad86 ]

In fsl_lpspi_probe an SPI controller is allocated either via spi_alloc_slave or spi_alloc_master. In all but one error cases this controller is put by going to error handling code. This commit fixes the case when pm_runtime_get_sync fails and it should go to the error handling path.

Fixes 944c01a889d9 ("spi: lpspi: enable runtime pm for lpspi")
---
 drivers/spi/spi-fsl-lpspi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/spi/spi-fsl-lpspi.c +1 -1
index d08e9324140e..3528ed5eea9b 100644
@@ -938,7 +938,7 @@ static int fsl_lpspi_probe(struct platform_device *pdev)
 	ret = pm_runtime_get_sync(fsl_lpspi->dev);
 	if (ret < 0) {
 		dev_err(fsl_lpspi->dev, "failed to enable clock\n");
-		return ret;
+		goto out_controller_put;
 	}
 
 	temp = readl(fsl_lpspi->base + IMX7ULP_PARAM);
Signature · 1 line
2.20.1
author maintainer
#19 lore raw

PATCH 13/205 media: rcar-vin: Fix incorrect return statement in rvin_try_format()

In reply to Sasha Levin

From: Niklas Söderlund <[email protected]>

[ Upstream commit a0862a40364e2f87109317e31c51c9d7bc89e33f ]

While refactoring code the return statement became corrupted, fix it by returning the correct return code.

Fixes 897e371389e77514 ("media: rcar-vin: simplify how formats are set and reset"
---
 drivers/media/platform/rcar-vin/rcar-v4l2.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/media/platform/rcar-vin/rcar-v4l2.c +2 -1
index cbc1c07f0a96..ec2796413e26 100644
@@ -208,6 +208,7 @@ static int rvin_try_format(struct rvin_dev *vin, u32 which,
 	ret = v4l2_subdev_call(sd, pad, set_fmt, pad_cfg, &format);
 	if (ret < 0 && ret != -ENOIOCTLCMD)
 		goto done;
+	ret = 0;
 
 	v4l2_fill_pix_format(pix, &format.format);
 
@@ -242,7 +243,7 @@ static int rvin_try_format(struct rvin_dev *vin, u32 which,
 done:
 	v4l2_subdev_free_pad_config(pad_cfg);
 
-	return 0;
+	return ret;
 }
 
 static int rvin_querycap(struct file *file, void *priv,
Signature · 1 line
2.20.1
author maintainer
#20 lore raw

PATCH 14/205 media: ov6650: Fix incorrect use of JPEG colorspace

In reply to Sasha Levin

From: Janusz Krzysztofik <[email protected]>

[ Upstream commit 12500731895ef09afc5b66b86b76c0884fb9c7bf ]

Since its initial submission, the driver selects V4L2_COLORSPACE_JPEG for supported formats other than V4L2_MBUS_FMT_SBGGR8_1X8. According to v4l2-compliance test program, V4L2_COLORSPACE_JPEG applies exclusively to V4L2_PIX_FMT_JPEG. Since the sensor does not support JPEG format, fix it to always select V4L2_COLORSPACE_SRGB.

Fixes 2f6e2404799a ("[media] SoC Camera: add driver for OV6650 sensor")
---
 drivers/media/i2c/ov6650.c | 13 ++-----------
 1 file changed, 2 insertions(+), 11 deletions(-)
drivers/media/i2c/ov6650.c +2 -11
index a5b2448c0abc..a654b402ef07 100644
@@ -201,7 +201,6 @@ struct ov6650 {
 	unsigned long		pclk_max;	/* from resolution and format */
 	struct v4l2_fract	tpf;		/* as requested with s_frame_interval */
 	u32 code;
-	enum v4l2_colorspace	colorspace;
 };
 
 
@@ -517,7 +516,7 @@ static int ov6650_get_fmt(struct v4l2_subdev *sd,
 	mf->width	= priv->rect.width >> priv->half_scale;
 	mf->height	= priv->rect.height >> priv->half_scale;
 	mf->code	= priv->code;
-	mf->colorspace	= priv->colorspace;
+	mf->colorspace	= V4L2_COLORSPACE_SRGB;
 	mf->field	= V4L2_FIELD_NONE;
 
 	return 0;
@@ -624,11 +623,6 @@ static int ov6650_s_fmt(struct v4l2_subdev *sd, struct v4l2_mbus_framefmt *mf)
 		priv->pclk_max = 8000000;
 	}
 
-	if (code == MEDIA_BUS_FMT_SBGGR8_1X8)
-		priv->colorspace = V4L2_COLORSPACE_SRGB;
-	else if (code != 0)
-		priv->colorspace = V4L2_COLORSPACE_JPEG;
-
 	if (half_scale) {
 		dev_dbg(&client->dev, "max resolution: QCIF\n");
 		coma_set |= COMA_QCIF;
@@ -663,7 +657,6 @@ static int ov6650_s_fmt(struct v4l2_subdev *sd, struct v4l2_mbus_framefmt *mf)
 		priv->code = code;
 
 	if (!ret) {
-		mf->colorspace	= priv->colorspace;
 		mf->width = priv->rect.width >> half_scale;
 		mf->height = priv->rect.height >> half_scale;
 	}
@@ -686,6 +679,7 @@ static int ov6650_set_fmt(struct v4l2_subdev *sd,
 				&mf->height, 2, H_CIF, 1, 0);
 
 	mf->field = V4L2_FIELD_NONE;
+	mf->colorspace = V4L2_COLORSPACE_SRGB;
 
 	switch (mf->code) {
 	case MEDIA_BUS_FMT_Y10_1X10:
@@ -696,13 +690,11 @@ static int ov6650_set_fmt(struct v4l2_subdev *sd,
 	case MEDIA_BUS_FMT_YUYV8_2X8:
 	case MEDIA_BUS_FMT_VYUY8_2X8:
 	case MEDIA_BUS_FMT_UYVY8_2X8:
-		mf->colorspace = V4L2_COLORSPACE_JPEG;
 		break;
 	default:
 		mf->code = MEDIA_BUS_FMT_SBGGR8_1X8;
 		/* fall through */
 	case MEDIA_BUS_FMT_SBGGR8_1X8:
-		mf->colorspace = V4L2_COLORSPACE_SRGB;
 		break;
 	}
 
@@ -1008,7 +1000,6 @@ static int ov6650_probe(struct i2c_client *client,
 	priv->rect.height = H_CIF;
 	priv->half_scale  = false;
 	priv->code	  = MEDIA_BUS_FMT_YUYV8_2X8;
-	priv->colorspace  = V4L2_COLORSPACE_JPEG;
 
 	/* Hardware default frame interval */
 	priv->tpf.numerator   = GET_CLKRC_DIV(DEF_CLKRC);
Signature · 1 line
2.20.1
author maintainer
#21 lore raw

PATCH 15/205 media: ov6650: Fix some format attributes not under control

In reply to Sasha Levin

From: Janusz Krzysztofik <[email protected]>

[ Upstream commit 1c6a2b63095154bbf9e8f38d79487a728331bf65 ]

User arguments passed to .get/set_fmt() pad operation callbacks may contain unsupported values. The driver takes control over frame size and pixel code as well as colorspace and field attributes but has never cared for remainig format attributes, i.e., ycbcr_enc, quantization and xfer_func, introduced by commit 11ff030c7365 ("[media]
v4l2-mediabus: improve colorspace support"). Fix it.

Set up a static v4l2_mbus_framefmt structure with attributes initialized to reasonable defaults and use it for updating content of user provided arguments. In case of V4L2_SUBDEV_FORMAT_ACTIVE, postpone frame size update, now performed from inside ov6650_s_fmt() helper, util the user argument is first updated in ov6650_set_fmt() with default frame format content. For V4L2_SUBDEV_FORMAT_TRY, don't copy all attributes to pad config, only those handled by the driver, then fill the response with the default frame format updated with resulting pad config format code and frame size.

Fixes 11ff030c7365 ("[media] v4l2-mediabus: improve colorspace support")
---
 drivers/media/i2c/ov6650.c | 51 +++++++++++++++++++++++++++++---------
 1 file changed, 39 insertions(+), 12 deletions(-)
drivers/media/i2c/ov6650.c +39 -12
index a654b402ef07..5fa3b2088b52 100644
@@ -213,6 +213,17 @@ static u32 ov6650_codes[] = {
 	MEDIA_BUS_FMT_Y8_1X8,
 };
 
+static const struct v4l2_mbus_framefmt ov6650_def_fmt = {
+	.width		= W_CIF,
+	.height		= H_CIF,
+	.code		= MEDIA_BUS_FMT_SBGGR8_1X8,
+	.colorspace	= V4L2_COLORSPACE_SRGB,
+	.field		= V4L2_FIELD_NONE,
+	.ycbcr_enc	= V4L2_YCBCR_ENC_DEFAULT,
+	.quantization	= V4L2_QUANTIZATION_DEFAULT,
+	.xfer_func	= V4L2_XFER_FUNC_DEFAULT,
+};
+
 /* read a register */
 static int ov6650_reg_read(struct i2c_client *client, u8 reg, u8 *val)
 {
@@ -513,11 +524,13 @@ static int ov6650_get_fmt(struct v4l2_subdev *sd,
 	if (format->pad)
 		return -EINVAL;
 
+	/* initialize response with default media bus frame format */
+	*mf = ov6650_def_fmt;
+
+	/* update media bus format code and frame size */
 	mf->width	= priv->rect.width >> priv->half_scale;
 	mf->height	= priv->rect.height >> priv->half_scale;
 	mf->code	= priv->code;
-	mf->colorspace	= V4L2_COLORSPACE_SRGB;
-	mf->field	= V4L2_FIELD_NONE;
 
 	return 0;
 }
@@ -656,10 +669,6 @@ static int ov6650_s_fmt(struct v4l2_subdev *sd, struct v4l2_mbus_framefmt *mf)
 	if (!ret)
 		priv->code = code;
 
-	if (!ret) {
-		mf->width = priv->rect.width >> half_scale;
-		mf->height = priv->rect.height >> half_scale;
-	}
 	return ret;
 }
 
@@ -678,9 +687,6 @@ static int ov6650_set_fmt(struct v4l2_subdev *sd,
 		v4l_bound_align_image(&mf->width, 2, W_CIF, 1,
 				&mf->height, 2, H_CIF, 1, 0);
 
-	mf->field = V4L2_FIELD_NONE;
-	mf->colorspace = V4L2_COLORSPACE_SRGB;
-
 	switch (mf->code) {
 	case MEDIA_BUS_FMT_Y10_1X10:
 		mf->code = MEDIA_BUS_FMT_Y8_1X8;
@@ -698,10 +704,31 @@ static int ov6650_set_fmt(struct v4l2_subdev *sd,
 		break;
 	}
 
-	if (format->which == V4L2_SUBDEV_FORMAT_ACTIVE)
-		return ov6650_s_fmt(sd, mf);
-	cfg->try_fmt = *mf;
+	if (format->which == V4L2_SUBDEV_FORMAT_TRY) {
+		/* store media bus format code and frame size in pad config */
+		cfg->try_fmt.width = mf->width;
+		cfg->try_fmt.height = mf->height;
+		cfg->try_fmt.code = mf->code;
 
+		/* return default mbus frame format updated with pad config */
+		*mf = ov6650_def_fmt;
+		mf->width = cfg->try_fmt.width;
+		mf->height = cfg->try_fmt.height;
+		mf->code = cfg->try_fmt.code;
+
+	} else {
+		/* apply new media bus format code and frame size */
+		int ret = ov6650_s_fmt(sd, mf);
+
+		if (ret)
+			return ret;
+
+		/* return default format updated with active size and code */
+		*mf = ov6650_def_fmt;
+		mf->width = priv->rect.width >> priv->half_scale;
+		mf->height = priv->rect.height >> priv->half_scale;
+		mf->code = priv->code;
+	}
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#22 lore raw

PATCH 16/205 media: ov6650: Fix .get_fmt() V4L2_SUBDEV_FORMAT_TRY support

In reply to Sasha Levin

From: Janusz Krzysztofik <[email protected]>

[ Upstream commit 39034bb0c26b76a2c3abc54aa28c185f18b40c2f ]

Commit da298c6d98d5 ("[media] v4l2: replace video op g_mbus_fmt by pad op get_fmt") converted a former ov6650_g_fmt() video operation callback to an ov6650_get_fmt() pad operation callback. However, the converted function disregards a format->which flag that pad operations should obey and always returns active frame format settings.

That can be fixed by always responding to V4L2_SUBDEV_FORMAT_TRY with -EINVAL, or providing the response from a pad config argument, likely updated by a former user call to V4L2_SUBDEV_FORMAT_TRY .set_fmt(). Since implementation of the latter is trivial, go for it.

Fixes da298c6d98d5 ("[media] v4l2: replace video op g_mbus_fmt by pad op get_fmt")
---
 drivers/media/i2c/ov6650.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/media/i2c/ov6650.c +9 -3
index 5fa3b2088b52..bbf15644e989 100644
@@ -528,10 +528,16 @@ static int ov6650_get_fmt(struct v4l2_subdev *sd,
 	*mf = ov6650_def_fmt;
 
 	/* update media bus format code and frame size */
-	mf->width	= priv->rect.width >> priv->half_scale;
-	mf->height	= priv->rect.height >> priv->half_scale;
-	mf->code	= priv->code;
+	if (format->which == V4L2_SUBDEV_FORMAT_TRY) {
+		mf->width = cfg->try_fmt.width;
+		mf->height = cfg->try_fmt.height;
+		mf->code = cfg->try_fmt.code;
 
+	} else {
+		mf->width = priv->rect.width >> priv->half_scale;
+		mf->height = priv->rect.height >> priv->half_scale;
+		mf->code = priv->code;
+	}
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#23 lore raw

PATCH 17/205 media: ov6650: Fix default format not applied on device probe

In reply to Sasha Levin

From: Janusz Krzysztofik <[email protected]>

[ Upstream commit 5439fa9263cb293e41168bc03711ec18c4f11cba ]

It is not clear what pixel format is actually configured in hardware on reset. MEDIA_BUS_FMT_YUYV8_2X8, assumed on device probe since the driver was intiially submitted, is for sure not the one.

Fix it by explicitly applying a known, driver default frame format just after initial device reset.

Fixes 2f6e2404799a ("[media] SoC Camera: add driver for OV6650 sensor")
---
 drivers/media/i2c/ov6650.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/media/i2c/ov6650.c +5 -2
index bbf15644e989..af482620f94a 100644
@@ -877,6 +877,11 @@ static int ov6650_video_probe(struct v4l2_subdev *sd)
 	ret = ov6650_reset(client);
 	if (!ret)
 		ret = ov6650_prog_dflt(client);
+	if (!ret) {
+		struct v4l2_mbus_framefmt mf = ov6650_def_fmt;
+
+		ret = ov6650_s_fmt(sd, &mf);
+	}
 	if (!ret)
 		ret = v4l2_ctrl_handler_setup(&priv->hdl);
 
@@ -1031,8 +1036,6 @@ static int ov6650_probe(struct i2c_client *client,
 	priv->rect.top	  = DEF_VSTRT << 1;
 	priv->rect.width  = W_CIF;
 	priv->rect.height = H_CIF;
-	priv->half_scale  = false;
-	priv->code	  = MEDIA_BUS_FMT_YUYV8_2X8;
 
 	/* Hardware default frame interval */
 	priv->tpf.numerator   = GET_CLKRC_DIV(DEF_CLKRC);
Signature · 1 line
2.20.1
author maintainer
#24 lore raw

PATCH 18/205 cw1200: Fix a signedness bug in cw1200_load_firmware()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 4a50d454502f1401171ff061a5424583f91266db ]

The "priv->hw_type" is an enum and in this context GCC will treat it as an unsigned int so the error handling will never trigger.

Fixes a910e4a94f69 ("cw1200: add driver for the ST-E CW1100 & CW1200 WLAN chipsets")
---
 drivers/net/wireless/st/cw1200/fwio.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/wireless/st/cw1200/fwio.c +3 -3
index 6574e78e05ea..2a03dc533b6a 100644
@@ -320,12 +320,12 @@ int cw1200_load_firmware(struct cw1200_common *priv)
 		goto out;
 	}
 
-	priv->hw_type = cw1200_get_hw_type(val32, &major_revision);
-	if (priv->hw_type < 0) {
+	ret = cw1200_get_hw_type(val32, &major_revision);
+	if (ret < 0) {
 		pr_err("Can't deduce hardware type.\n");
-		ret = -ENOTSUPP;
 		goto out;
 	}
+	priv->hw_type = ret;
 
 	/* Set DPLL Reg value, and read back to confirm writes work */
 	ret = cw1200_reg_write_32(priv, ST90TDS_TSET_GEN_R_W_REG_ID,
Signature · 1 line
2.20.1
author maintainer
#25 lore raw

PATCH 19/205 drm/amdgpu: remove excess function parameter description

In reply to Sasha Levin

From: yu kuai <[email protected]>

[ Upstream commit d0580c09c65cff211f589a40e08eabc62da463fb ]

Fixes gcc warning:

drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c:431: warning: Excess function parameter 'sw' description in 'vcn_v2_5_disable_clock_gating' drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c:550: warning: Excess function parameter 'sw' description in 'vcn_v2_5_enable_clock_gating'

Fixes cbead2bdfcf1 ("drm/amdgpu: add VCN2.5 VCPU start and stop")
---
 drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 2 --
 1 file changed, 2 deletions(-)
drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +0 -2
index 395c2259f979..9d778a0b2c5e 100644
@@ -423,7 +423,6 @@ static void vcn_v2_5_mc_resume(struct amdgpu_device *adev)
  * vcn_v2_5_disable_clock_gating - disable VCN clock gating
  *
  * @adev: amdgpu_device pointer
- * @sw: enable SW clock gating
  *
  * Disable clock gating for VCN block
  */
@@ -542,7 +541,6 @@ static void vcn_v2_5_disable_clock_gating(struct amdgpu_device *adev)
  * vcn_v2_5_enable_clock_gating - enable VCN clock gating
  *
  * @adev: amdgpu_device pointer
- * @sw: enable SW clock gating
  *
  * Enable clock gating for VCN block
  */
Signature · 6 lines
2.20.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#26 lore raw

PATCH 20/205 arm64: dts: meson: axg: fix audio fifo reg size

In reply to Sasha Levin

From: Jerome Brunet <[email protected]>

[ Upstream commit 301b94d434ac3a3cd576a4bc1053cc243d6bd841 ]

The register region size initially is too small to access all the fifo registers.

Fixes f2b8f6a93357 ("arm64: dts: meson-axg: add audio fifos")
---
 arch/arm64/boot/dts/amlogic/meson-axg.dtsi | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
arch/arm64/boot/dts/amlogic/meson-axg.dtsi +6 -6
index 82919b106010..bb4a2acb9970 100644
@@ -1162,7 +1162,7 @@
 
 			toddr_a: audio-controller@100 {
 				compatible = "amlogic,axg-toddr";
-				reg = <0x0 0x100 0x0 0x1c>;
+				reg = <0x0 0x100 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "TODDR_A";
 				interrupts = <GIC_SPI 84 IRQ_TYPE_EDGE_RISING>;
@@ -1173,7 +1173,7 @@
 
 			toddr_b: audio-controller@140 {
 				compatible = "amlogic,axg-toddr";
-				reg = <0x0 0x140 0x0 0x1c>;
+				reg = <0x0 0x140 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "TODDR_B";
 				interrupts = <GIC_SPI 85 IRQ_TYPE_EDGE_RISING>;
@@ -1184,7 +1184,7 @@
 
 			toddr_c: audio-controller@180 {
 				compatible = "amlogic,axg-toddr";
-				reg = <0x0 0x180 0x0 0x1c>;
+				reg = <0x0 0x180 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "TODDR_C";
 				interrupts = <GIC_SPI 86 IRQ_TYPE_EDGE_RISING>;
@@ -1195,7 +1195,7 @@
 
 			frddr_a: audio-controller@1c0 {
 				compatible = "amlogic,axg-frddr";
-				reg = <0x0 0x1c0 0x0 0x1c>;
+				reg = <0x0 0x1c0 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "FRDDR_A";
 				interrupts = <GIC_SPI 88 IRQ_TYPE_EDGE_RISING>;
@@ -1206,7 +1206,7 @@
 
 			frddr_b: audio-controller@200 {
 				compatible = "amlogic,axg-frddr";
-				reg = <0x0 0x200 0x0 0x1c>;
+				reg = <0x0 0x200 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "FRDDR_B";
 				interrupts = <GIC_SPI 89 IRQ_TYPE_EDGE_RISING>;
@@ -1217,7 +1217,7 @@
 
 			frddr_c: audio-controller@240 {
 				compatible = "amlogic,axg-frddr";
-				reg = <0x0 0x240 0x0 0x1c>;
+				reg = <0x0 0x240 0x0 0x2c>;
 				#sound-dai-cells = <0>;
 				sound-name-prefix = "FRDDR_C";
 				interrupts = <GIC_SPI 90 IRQ_TYPE_EDGE_RISING>;
Signature · 1 line
2.20.1
author maintainer
#27 lore raw

PATCH 21/205 arm64: dts: meson: g12: fix audio fifo reg size

In reply to Sasha Levin

From: Jerome Brunet <[email protected]>

[ Upstream commit 22c4b148a0a1085e57a470e6f7dc515cf08f5a5c ]

The register region size initially is too small to access all the fifo registers.

Fixes c59b7fe5aafd ("arm64: dts: meson: g12a: add audio fifos")
---
 arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi +6 -6
index 3f39e020f74e..0ee8a369c547 100644
@@ -1509,7 +1509,7 @@
 				toddr_a: audio-controller@100 {
 					compatible = "amlogic,g12a-toddr",
 						     "amlogic,axg-toddr";
-					reg = <0x0 0x100 0x0 0x1c>;
+					reg = <0x0 0x100 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "TODDR_A";
 					interrupts = <GIC_SPI 148 IRQ_TYPE_EDGE_RISING>;
@@ -1521,7 +1521,7 @@
 				toddr_b: audio-controller@140 {
 					compatible = "amlogic,g12a-toddr",
 						     "amlogic,axg-toddr";
-					reg = <0x0 0x140 0x0 0x1c>;
+					reg = <0x0 0x140 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "TODDR_B";
 					interrupts = <GIC_SPI 149 IRQ_TYPE_EDGE_RISING>;
@@ -1533,7 +1533,7 @@
 				toddr_c: audio-controller@180 {
 					compatible = "amlogic,g12a-toddr",
 						     "amlogic,axg-toddr";
-					reg = <0x0 0x180 0x0 0x1c>;
+					reg = <0x0 0x180 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "TODDR_C";
 					interrupts = <GIC_SPI 150 IRQ_TYPE_EDGE_RISING>;
@@ -1545,7 +1545,7 @@
 				frddr_a: audio-controller@1c0 {
 					compatible = "amlogic,g12a-frddr",
 						     "amlogic,axg-frddr";
-					reg = <0x0 0x1c0 0x0 0x1c>;
+					reg = <0x0 0x1c0 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "FRDDR_A";
 					interrupts = <GIC_SPI 152 IRQ_TYPE_EDGE_RISING>;
@@ -1557,7 +1557,7 @@
 				frddr_b: audio-controller@200 {
 					compatible = "amlogic,g12a-frddr",
 						     "amlogic,axg-frddr";
-					reg = <0x0 0x200 0x0 0x1c>;
+					reg = <0x0 0x200 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "FRDDR_B";
 					interrupts = <GIC_SPI 153 IRQ_TYPE_EDGE_RISING>;
@@ -1569,7 +1569,7 @@
 				frddr_c: audio-controller@240 {
 					compatible = "amlogic,g12a-frddr",
 						     "amlogic,axg-frddr";
-					reg = <0x0 0x240 0x0 0x1c>;
+					reg = <0x0 0x240 0x0 0x2c>;
 					#sound-dai-cells = <0>;
 					sound-name-prefix = "FRDDR_C";
 					interrupts = <GIC_SPI 154 IRQ_TYPE_EDGE_RISING>;
Signature · 1 line
2.20.1
author maintainer
#28 lore raw

PATCH 22/205 rtw88: fix beaconing mode rsvd_page memory violation issue

In reply to Sasha Levin

From: Yan-Hsuan Chuang <[email protected]>

[ Upstream commit c3594559f49c601d410dee4b767c3536a5535bfd ]

When downloading the reserved page, the first page always contains a beacon for the firmware to reference. For non-beaconing modes such as station mode, also put a blank skb with length=1.

And for the beaconing modes, driver will get a real beacon with a length approximate to the page size. But as the beacon is always put at the first page, it does not need a tx_desc, because the TX path will generate one when TXing the reserved page to the hardware. So we could allocate a buffer with a size smaller than the reserved page, when using memcpy() to copy the content of reserved page to the buffer, the over-sized reserved page will violate the kernel memory.

To fix it, add the tx_desc before memcpy() the reserved packets to the buffer, then we can get SKBs with correct length when counting the pages in total. And for page 0, count the extra tx_desc_sz that the TX path will generate. This way, the first beacon that allocated without tx_desc can be counted with the extra tx_desc_sz to get actual pages it requires.

Fixes e3037485c68e ("rtw88: new Realtek 802.11ac driver")
---
 drivers/net/wireless/realtek/rtw88/fw.c | 52 ++++++++++++++++++++-----
 1 file changed, 43 insertions(+), 9 deletions(-)
drivers/net/wireless/realtek/rtw88/fw.c +43 -9
index b082e2cc95f5..35dbdb3c4f1e 100644
@@ -498,9 +498,6 @@ static void rtw_rsvd_page_list_to_buf(struct rtw_dev *rtwdev, u8 page_size,
 {
 	struct sk_buff *skb = rsvd_pkt->skb;
 
-	if (rsvd_pkt->add_txdesc)
-		rtw_fill_rsvd_page_desc(rtwdev, skb);
-
 	if (page >= 1)
 		memcpy(buf + page_margin + page_size * (page - 1),
 		       skb->data, skb->len);
@@ -625,16 +622,37 @@ static u8 *rtw_build_rsvd_page(struct rtw_dev *rtwdev,
 	list_for_each_entry(rsvd_pkt, &rtwdev->rsvd_page_list, list) {
 		iter = rtw_get_rsvd_page_skb(hw, vif, rsvd_pkt->type);
 		if (!iter) {
-			rtw_err(rtwdev, "fail to build rsvd packet\n");
+			rtw_err(rtwdev, "failed to build rsvd packet\n");
 			goto release_skb;
 		}
+
+		/* Fill the tx_desc for the rsvd pkt that requires one.
+		 * And iter->len will be added with size of tx_desc_sz.
+		 */
+		if (rsvd_pkt->add_txdesc)
+			rtw_fill_rsvd_page_desc(rtwdev, iter);
+
 		rsvd_pkt->skb = iter;
 		rsvd_pkt->page = total_page;
-		if (rsvd_pkt->add_txdesc)
+
+		/* Reserved page is downloaded via TX path, and TX path will
+		 * generate a tx_desc at the header to describe length of
+		 * the buffer. If we are not counting page numbers with the
+		 * size of tx_desc added at the first rsvd_pkt (usually a
+		 * beacon, firmware default refer to the first page as the
+		 * content of beacon), we could generate a buffer which size
+		 * is smaller than the actual size of the whole rsvd_page
+		 */
+		if (total_page == 0) {
+			if (rsvd_pkt->type != RSVD_BEACON) {
+				rtw_err(rtwdev, "first page should be a beacon\n");
+				goto release_skb;
+			}
 			total_page += rtw_len_to_page(iter->len + tx_desc_sz,
 						      page_size);
-		else
+		} else {
 			total_page += rtw_len_to_page(iter->len, page_size);
+		}
 	}
 
 	if (total_page > rtwdev->fifo.rsvd_drv_pg_num) {
@@ -647,13 +665,24 @@ static u8 *rtw_build_rsvd_page(struct rtw_dev *rtwdev,
 	if (!buf)
 		goto release_skb;
 
+	/* Copy the content of each rsvd_pkt to the buf, and they should
+	 * be aligned to the pages.
+	 *
+	 * Note that the first rsvd_pkt is a beacon no matter what vif->type.
+	 * And that rsvd_pkt does not require tx_desc because when it goes
+	 * through TX path, the TX path will generate one for it.
+	 */
 	list_for_each_entry(rsvd_pkt, &rtwdev->rsvd_page_list, list) {
 		rtw_rsvd_page_list_to_buf(rtwdev, page_size, page_margin,
 					  page, buf, rsvd_pkt);
-		page += rtw_len_to_page(rsvd_pkt->skb->len, page_size);
-	}
-	list_for_each_entry(rsvd_pkt, &rtwdev->rsvd_page_list, list)
+		if (page == 0)
+			page += rtw_len_to_page(rsvd_pkt->skb->len +
+						tx_desc_sz, page_size);
+		else
+			page += rtw_len_to_page(rsvd_pkt->skb->len, page_size);
+
 		kfree_skb(rsvd_pkt->skb);
+	}
 
 	return buf;
 
@@ -706,6 +735,11 @@ int rtw_fw_download_rsvd_page(struct rtw_dev *rtwdev, struct ieee80211_vif *vif)
 		goto free;
 	}
 
+	/* The last thing is to download the *ONLY* beacon again, because
+	 * the previous tx_desc is to describe the total rsvd page. Download
+	 * the beacon again to replace the TX desc header, and we will get
+	 * a correct tx_desc for the beacon in the rsvd page.
+	 */
 	ret = rtw_download_beacon(rtwdev, vif);
 	if (ret) {
 		rtw_err(rtwdev, "failed to download beacon\n");
Signature · 1 line
2.20.1
author maintainer
#29 lore raw

PATCH 23/205 rtw88: fix error handling when setup efuse info

In reply to Sasha Levin

From: Ping-Ke Shih <[email protected]>

[ Upstream commit f4268729eb1eefe23f6746849c1b5626d9030532 ]

Disable efuse if the efuse is enabled when we failed to setup the efuse information, otherwise the hardware will not turn off.

Fixes e3037485c68e ("rtw88: new Realtek 802.11ac driver")
---
 drivers/net/wireless/realtek/rtw88/main.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/net/wireless/realtek/rtw88/main.c +6 -5
index 7a3a4911bde2..806af37192bc 100644
@@ -1048,19 +1048,19 @@ static int rtw_chip_efuse_info_setup(struct rtw_dev *rtwdev)
 	/* power on mac to read efuse */
 	ret = rtw_chip_efuse_enable(rtwdev);
 	if (ret)
-		goto out;
+		goto out_unlock;
 
 	ret = rtw_parse_efuse_map(rtwdev);
 	if (ret)
-		goto out;
+		goto out_disable;
 
 	ret = rtw_dump_hw_feature(rtwdev);
 	if (ret)
-		goto out;
+		goto out_disable;
 
 	ret = rtw_check_supported_rfe(rtwdev);
 	if (ret)
-		goto out;
+		goto out_disable;
 
 	if (efuse->crystal_cap == 0xff)
 		efuse->crystal_cap = 0;
@@ -1087,9 +1087,10 @@ static int rtw_chip_efuse_info_setup(struct rtw_dev *rtwdev)
 	efuse->ext_pa_5g = efuse->pa_type_5g & BIT(0) ? 1 : 0;
 	efuse->ext_lna_2g = efuse->lna_type_5g & BIT(3) ? 1 : 0;
 
+out_disable:
 	rtw_chip_efuse_disable(rtwdev);
 
-out:
+out_unlock:
 	mutex_unlock(&rtwdev->mutex);
 	return ret;
 }
Signature · 1 line
2.20.1
author maintainer
#30 lore raw

PATCH 24/205 crypto: algif_skcipher - Use chunksize instead of blocksize

In reply to Sasha Levin

From: Herbert Xu <[email protected]>

[ Upstream commit 5b0fe9552336338acb52756daf65dd7a4eeca73f ]

When algif_skcipher does a partial operation it always process data that is a multiple of blocksize. However, for algorithms such as CTR this is wrong because even though it can process any number of bytes overall, the partial block must come at the very end and not in the middle.

This is exactly what chunksize is meant to describe so this patch changes blocksize to chunksize.

Fixes 8ff590903d5f ("crypto: algif_skcipher - User-space...")
---
 crypto/algif_skcipher.c            |  2 +-
 include/crypto/internal/skcipher.h | 30 ------------------------------
 include/crypto/skcipher.h          | 30 ++++++++++++++++++++++++++++++
 3 files changed, 31 insertions(+), 31 deletions(-)
crypto/algif_skcipher.c +1 -1
index c1601edd70e3..e2c8ab408bed 100644
@@ -56,7 +56,7 @@ static int _skcipher_recvmsg(struct socket *sock, struct msghdr *msg,
 	struct alg_sock *pask = alg_sk(psk);
 	struct af_alg_ctx *ctx = ask->private;
 	struct crypto_skcipher *tfm = pask->private;
-	unsigned int bs = crypto_skcipher_blocksize(tfm);
+	unsigned int bs = crypto_skcipher_chunksize(tfm);
 	struct af_alg_async_req *areq;
 	int err = 0;
 	size_t len = 0;
include/crypto/internal/skcipher.h +0 -30
index 734b6f7081b8..3175dfeaed2c 100644
@@ -205,19 +205,6 @@ static inline unsigned int crypto_skcipher_alg_max_keysize(
 	return alg->max_keysize;
 }
 
-static inline unsigned int crypto_skcipher_alg_chunksize(
-	struct skcipher_alg *alg)
-{
-	if ((alg->base.cra_flags & CRYPTO_ALG_TYPE_MASK) ==
-	    CRYPTO_ALG_TYPE_BLKCIPHER)
-		return alg->base.cra_blocksize;
-
-	if (alg->base.cra_ablkcipher.encrypt)
-		return alg->base.cra_blocksize;
-
-	return alg->chunksize;
-}
-
 static inline unsigned int crypto_skcipher_alg_walksize(
 	struct skcipher_alg *alg)
 {
@@ -231,23 +218,6 @@ static inline unsigned int crypto_skcipher_alg_walksize(
 	return alg->walksize;
 }
 
-/**
- * crypto_skcipher_chunksize() - obtain chunk size
- * @tfm: cipher handle
- *
- * The block size is set to one for ciphers such as CTR.  However,
- * you still need to provide incremental updates in multiples of
- * the underlying block size as the IV does not have sub-block
- * granularity.  This is known in this API as the chunk size.
- *
- * Return: chunk size in bytes
- */
-static inline unsigned int crypto_skcipher_chunksize(
-	struct crypto_skcipher *tfm)
-{
-	return crypto_skcipher_alg_chunksize(crypto_skcipher_alg(tfm));
-}
-
 /**
  * crypto_skcipher_walksize() - obtain walk size
  * @tfm: cipher handle
include/crypto/skcipher.h +30 -0
index 37c164234d97..aada87916918 100644
@@ -304,6 +304,36 @@ static inline unsigned int crypto_skcipher_blocksize(
 	return crypto_tfm_alg_blocksize(crypto_skcipher_tfm(tfm));
 }
 
+static inline unsigned int crypto_skcipher_alg_chunksize(
+	struct skcipher_alg *alg)
+{
+	if ((alg->base.cra_flags & CRYPTO_ALG_TYPE_MASK) ==
+	    CRYPTO_ALG_TYPE_BLKCIPHER)
+		return alg->base.cra_blocksize;
+
+	if (alg->base.cra_ablkcipher.encrypt)
+		return alg->base.cra_blocksize;
+
+	return alg->chunksize;
+}
+
+/**
+ * crypto_skcipher_chunksize() - obtain chunk size
+ * @tfm: cipher handle
+ *
+ * The block size is set to one for ciphers such as CTR.  However,
+ * you still need to provide incremental updates in multiples of
+ * the underlying block size as the IV does not have sub-block
+ * granularity.  This is known in this API as the chunk size.
+ *
+ * Return: chunk size in bytes
+ */
+static inline unsigned int crypto_skcipher_chunksize(
+	struct crypto_skcipher *tfm)
+{
+	return crypto_skcipher_alg_chunksize(crypto_skcipher_alg(tfm));
+}
+
 static inline unsigned int crypto_sync_skcipher_blocksize(
 	struct crypto_sync_skcipher *tfm)
 {
Signature · 1 line
2.20.1
author maintainer
#31 lore raw

PATCH 25/205 hwrng: omap3-rom - Fix missing clock by probing with device tree

In reply to Sasha Levin

From: Tony Lindgren <[email protected]>

[ Upstream commit 0c0ef9ea6f3f0d5979dc7b094b0a184c1a94716b ]

Commit 0ed266d7ae5e ("clk: ti: omap3: cleanup unnecessary clock aliases") removed old omap3 clock framework aliases but caused omap3-rom-rng to stop working with clock not found error.

Based on discussions on the mailing list it was requested by Tero Kristo that it would be best to fix this issue by probing omap3-rom-rng using device tree to provide a proper clk property. The other option would be to add back the missing clock alias, but that does not help moving things forward with removing old legacy platform_data.

Let's also add a proper device tree binding and keep it together with the fix.

Cc Aaro Koskinen <[email protected]>
Cc Adam Ford <[email protected]>
Cc Pali Rohár <[email protected]>
Cc Rob Herring <[email protected]>
Cc Sebastian Reichel <[email protected]>
Cc Tero Kristo <[email protected]>
Fixes 0ed266d7ae5e ("clk: ti: omap3: cleanup unnecessary clock aliases")
---
 .../devicetree/bindings/rng/omap3_rom_rng.txt | 27 +++++++++++++++++++
 arch/arm/boot/dts/omap3-n900.dts              |  6 +++++
 arch/arm/mach-omap2/pdata-quirks.c            | 12 +--------
 drivers/char/hw_random/omap3-rom-rng.c        | 17 ++++++++++--
 4 files changed, 49 insertions(+), 13 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/rng/omap3_rom_rng.txt
Documentation/devicetree/bindings/rng/omap3_rom_rng.txt new file +27 -0
new file mode 100644
index 000000000000..f315c9723bd2
@@ -0,0 +1,27 @@
+OMAP ROM RNG driver binding
+
+Secure SoCs may provide RNG via secure ROM calls like Nokia N900 does. The
+implementation can depend on the SoC secure ROM used.
+
+- compatible:
+	Usage: required
+	Value type: <string>
+	Definition: must be "nokia,n900-rom-rng"
+
+- clocks:
+	Usage: required
+	Value type: <prop-encoded-array>
+	Definition: reference to the the RNG interface clock
+
+- clock-names:
+	Usage: required
+	Value type: <stringlist>
+	Definition: must be "ick"
+
+Example:
+
+	rom_rng: rng {
+		compatible = "nokia,n900-rom-rng";
+		clocks = <&rng_ick>;
+		clock-names = "ick";
+	};
arch/arm/boot/dts/omap3-n900.dts +6 -0
index 84a5ade1e865..63659880eeb3 100644
@@ -155,6 +155,12 @@
 		pwms = <&pwm9 0 26316 0>; /* 38000 Hz */
 	};
 
+	rom_rng: rng {
+		compatible = "nokia,n900-rom-rng";
+		clocks = <&rng_ick>;
+		clock-names = "ick";
+	};
+
 	/* controlled (enabled/disabled) directly by bcm2048 and wl1251 */
 	vctcxo: vctcxo {
 		compatible = "fixed-clock";
arch/arm/mach-omap2/pdata-quirks.c +1 -11
index 1b7cf81ff035..7189535cbb32 100644
@@ -268,14 +268,6 @@ static void __init am3517_evm_legacy_init(void)
 	am35xx_emac_reset();
 }
 
-static struct platform_device omap3_rom_rng_device = {
-	.name		= "omap3-rom-rng",
-	.id		= -1,
-	.dev	= {
-		.platform_data	= rx51_secure_rng_call,
-	},
-};
-
 static void __init nokia_n900_legacy_init(void)
 {
 	hsmmc2_internal_input_clk();
@@ -291,9 +283,6 @@ static void __init nokia_n900_legacy_init(void)
 			pr_warn("RX-51: Not enabling ARM errata 430973 workaround\n");
 			pr_warn("Thumb binaries may crash randomly without this workaround\n");
 		}
-
-		pr_info("RX-51: Registering OMAP3 HWRNG device\n");
-		platform_device_register(&omap3_rom_rng_device);
 	}
 }
 
@@ -534,6 +523,7 @@ static struct of_dev_auxdata omap_auxdata_lookup[] = {
 	OF_DEV_AUXDATA("ti,davinci_mdio", 0x5c030000, "davinci_mdio.0", NULL),
 	OF_DEV_AUXDATA("ti,am3517-emac", 0x5c000000, "davinci_emac.0",
 		       &am35xx_emac_pdata),
+	OF_DEV_AUXDATA("nokia,n900-rom-rng", 0, NULL, rx51_secure_rng_call),
 	/* McBSP modules with sidetone core */
 #if IS_ENABLED(CONFIG_SND_SOC_OMAP_MCBSP)
 	OF_DEV_AUXDATA("ti,omap3-mcbsp", 0x49022000, "49022000.mcbsp", &mcbsp_pdata),
drivers/char/hw_random/omap3-rom-rng.c +15 -2
index 648e39ce6bd9..8df3cad7c97a 100644
@@ -20,6 +20,8 @@
 #include <linux/workqueue.h>
 #include <linux/clk.h>
 #include <linux/err.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 
 #define RNG_RESET			0x01
@@ -86,14 +88,18 @@ static int omap3_rom_rng_read(struct hwrng *rng, void *data, size_t max, bool w)
 
 static struct hwrng omap3_rom_rng_ops = {
 	.name		= "omap3-rom",
-	.read		= omap3_rom_rng_read,
 };
 
 static int omap3_rom_rng_probe(struct platform_device *pdev)
 {
 	int ret = 0;
 
-	pr_info("initializing\n");
+	omap3_rom_rng_ops.read = of_device_get_match_data(&pdev->dev);
+	if (!omap3_rom_rng_ops.read) {
+		dev_err(&pdev->dev, "missing rom code handler\n");
+
+		return -ENODEV;
+	}
 
 	omap3_rom_rng_call = pdev->dev.platform_data;
 	if (!omap3_rom_rng_call) {
@@ -126,9 +132,16 @@ static int omap3_rom_rng_remove(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct of_device_id omap_rom_rng_match[] = {
+	{ .compatible = "nokia,n900-rom-rng", .data = omap3_rom_rng_read, },
+	{ /* sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, omap_rom_rng_match);
+
 static struct platform_driver omap3_rom_rng_driver = {
 	.driver = {
 		.name		= "omap3-rom-rng",
+		.of_match_table = omap_rom_rng_match,
 	},
 	.probe		= omap3_rom_rng_probe,
 	.remove		= omap3_rom_rng_remove,
Signature · 1 line
2.20.1
author maintainer
#32 lore raw

PATCH 26/205 crypto: cavium/nitrox - fix firmware assignment to AE cores

In reply to Sasha Levin

From: Phani Kiran Hemadri <[email protected]>

[ Upstream commit 6a97a99db848748d582d79447f7c9c330ce1688e ]

This patch fixes assigning UCD block number of Asymmetric crypto firmware to AE cores of CNN55XX device.

Fixes a7268c4d4205 ("crypto: cavium/nitrox - Add support for loading asymmetric crypto firmware")
---
 drivers/crypto/cavium/nitrox/nitrox_main.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
drivers/crypto/cavium/nitrox/nitrox_main.c +4 -5
index bc924980e10c..c4632d84c9a1 100644
@@ -103,8 +103,7 @@ static void write_to_ucd_unit(struct nitrox_device *ndev, u32 ucode_size,
 	offset = UCD_UCODE_LOAD_BLOCK_NUM;
 	nitrox_write_csr(ndev, offset, block_num);
 
-	code_size = ucode_size;
-	code_size = roundup(code_size, 8);
+	code_size = roundup(ucode_size, 16);
 	while (code_size) {
 		data = ucode_data[i];
 		/* write 8 bytes at a time */
@@ -220,11 +219,11 @@ static int nitrox_load_fw(struct nitrox_device *ndev)
 
 	/* write block number and firmware length
 	 * bit:<2:0> block number
-	 * bit:3 is set SE uses 32KB microcode
-	 * bit:3 is clear SE uses 64KB microcode
+	 * bit:3 is set AE uses 32KB microcode
+	 * bit:3 is clear AE uses 64KB microcode
 	 */
 	core_2_eid_val.value = 0ULL;
-	core_2_eid_val.ucode_blk = 0;
+	core_2_eid_val.ucode_blk = 2;
 	if (ucode_size <= CNN55XX_UCD_BLOCK_SIZE)
 		core_2_eid_val.ucode_len = 1;
 	else
Signature · 1 line
2.20.1
author maintainer
#33 lore raw

PATCH 27/205 pinctl: ti: iodelay: fix error checking on pinctrl_count_index_with_args call

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit 5ff8aca906f3a7a7db79fad92f2a4401107ef50d ]

The call to pinctrl_count_index_with_args checks for a -EINVAL return however this function calls pinctrl_get_list_and_count and this can return -ENOENT. Rather than check for a specific error, fix this by checking for any error return to catch the -ENOENT case.

Addresses-Coverity: ("Improper use of negative")

Fixes 003910ebc83b ("pinctrl: Introduce TI IOdelay configuration driver")
---
 drivers/pinctrl/ti/pinctrl-ti-iodelay.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pinctrl/ti/pinctrl-ti-iodelay.c +1 -1
index e5e7f1f22813..b522ca010332 100644
@@ -496,7 +496,7 @@ static int ti_iodelay_dt_node_to_map(struct pinctrl_dev *pctldev,
 		return -EINVAL;
 
 	rows = pinctrl_count_index_with_args(np, name);
-	if (rows == -EINVAL)
+	if (rows < 0)
 		return rows;
 
 	*map = devm_kzalloc(iod->dev, sizeof(**map), GFP_KERNEL);
Signature · 1 line
2.20.1
author maintainer
#34 lore raw

PATCH 28/205 net: phy: broadcom: Fix RGMII delays configuration for BCM54210E

In reply to Sasha Levin

From: Florian Fainelli <[email protected]>

[ Upstream commit fea7fda7f50a6059220f83251e70709e45cc8040 ]

Commit 0fc9ae107669 ("net: phy: broadcom: add support for BCM54210E") added support for BCM54210E but also unconditionally cleared the RXC to RXD skew and the TXD to TXC skew, thus only making PHY_INTERFACE_MODE_RGMII a possible configuration. Use bcm54xx_config_clock_delay() which correctly sets the registers depending on the 4 possible PHY interface values that exist for RGMII.

Fixes 0fc9ae107669 ("net: phy: broadcom: add support for BCM54210E")
---
 drivers/net/phy/broadcom.c | 11 +++--------
 1 file changed, 3 insertions(+), 8 deletions(-)
drivers/net/phy/broadcom.c +3 -8
index 937d0059e8ac..5e956089bf52 100644
@@ -26,18 +26,13 @@ MODULE_DESCRIPTION("Broadcom PHY driver");
 MODULE_AUTHOR("Maciej W. Rozycki");
 MODULE_LICENSE("GPL");
 
+static int bcm54xx_config_clock_delay(struct phy_device *phydev);
+
 static int bcm54210e_config_init(struct phy_device *phydev)
 {
 	int val;
 
-	val = bcm54xx_auxctl_read(phydev, MII_BCM54XX_AUXCTL_SHDWSEL_MISC);
-	val &= ~MII_BCM54XX_AUXCTL_SHDWSEL_MISC_RGMII_SKEW_EN;
-	val |= MII_BCM54XX_AUXCTL_MISC_WREN;
-	bcm54xx_auxctl_write(phydev, MII_BCM54XX_AUXCTL_SHDWSEL_MISC, val);
-
-	val = bcm_phy_read_shadow(phydev, BCM54810_SHD_CLK_CTL);
-	val &= ~BCM54810_SHD_CLK_CTL_GTXCLK_EN;
-	bcm_phy_write_shadow(phydev, BCM54810_SHD_CLK_CTL, val);
+	bcm54xx_config_clock_delay(phydev);
 
 	if (phydev->dev_flags & PHY_BRCM_EN_MASTER_MODE) {
 		val = phy_read(phydev, MII_CTRL1000);
Signature · 1 line
2.20.1
author maintainer
#35 lore raw

PATCH 29/205 clk: imx: pll14xx: Fix quick switch of S/K parameter

In reply to Sasha Levin

From: Leonard Crestez <[email protected]>

[ Upstream commit 094234fcf46146339caaac8282aa15d225a5911a ]

The PLL14xx on imx8m can change the S and K parameter without requiring a reset and relock of the whole PLL.

Fix clk_pll144xx_mp_change register reading and use it for pll1443 as well since no reset+relock is required on K changes either.

Fixes 8646d4dcc7fb ("clk: imx: Add PLLs driver for imx8mm soc")
---
 drivers/clk/imx/clk-pll14xx.c | 40 +++++++----------------------------
 1 file changed, 8 insertions(+), 32 deletions(-)
drivers/clk/imx/clk-pll14xx.c +8 -32
index d43b4a3c0de8..047f1d8fe323 100644
@@ -112,43 +112,17 @@ static unsigned long clk_pll1443x_recalc_rate(struct clk_hw *hw,
 	return fvco;
 }
 
-static inline bool clk_pll1416x_mp_change(const struct imx_pll14xx_rate_table *rate,
+static inline bool clk_pll14xx_mp_change(const struct imx_pll14xx_rate_table *rate,
 					  u32 pll_div)
 {
 	u32 old_mdiv, old_pdiv;
 
-	old_mdiv = (pll_div >> MDIV_SHIFT) & MDIV_MASK;
-	old_pdiv = (pll_div >> PDIV_SHIFT) & PDIV_MASK;
+	old_mdiv = (pll_div & MDIV_MASK) >> MDIV_SHIFT;
+	old_pdiv = (pll_div & PDIV_MASK) >> PDIV_SHIFT;
 
 	return rate->mdiv != old_mdiv || rate->pdiv != old_pdiv;
 }
 
-static inline bool clk_pll1443x_mpk_change(const struct imx_pll14xx_rate_table *rate,
-					  u32 pll_div_ctl0, u32 pll_div_ctl1)
-{
-	u32 old_mdiv, old_pdiv, old_kdiv;
-
-	old_mdiv = (pll_div_ctl0 >> MDIV_SHIFT) & MDIV_MASK;
-	old_pdiv = (pll_div_ctl0 >> PDIV_SHIFT) & PDIV_MASK;
-	old_kdiv = (pll_div_ctl1 >> KDIV_SHIFT) & KDIV_MASK;
-
-	return rate->mdiv != old_mdiv || rate->pdiv != old_pdiv ||
-		rate->kdiv != old_kdiv;
-}
-
-static inline bool clk_pll1443x_mp_change(const struct imx_pll14xx_rate_table *rate,
-					  u32 pll_div_ctl0, u32 pll_div_ctl1)
-{
-	u32 old_mdiv, old_pdiv, old_kdiv;
-
-	old_mdiv = (pll_div_ctl0 >> MDIV_SHIFT) & MDIV_MASK;
-	old_pdiv = (pll_div_ctl0 >> PDIV_SHIFT) & PDIV_MASK;
-	old_kdiv = (pll_div_ctl1 >> KDIV_SHIFT) & KDIV_MASK;
-
-	return rate->mdiv != old_mdiv || rate->pdiv != old_pdiv ||
-		rate->kdiv != old_kdiv;
-}
-
 static int clk_pll14xx_wait_lock(struct clk_pll14xx *pll)
 {
 	u32 val;
@@ -174,7 +148,7 @@ static int clk_pll1416x_set_rate(struct clk_hw *hw, unsigned long drate,
 
 	tmp = readl_relaxed(pll->base + 4);
 
-	if (!clk_pll1416x_mp_change(rate, tmp)) {
+	if (!clk_pll14xx_mp_change(rate, tmp)) {
 		tmp &= ~(SDIV_MASK) << SDIV_SHIFT;
 		tmp |= rate->sdiv << SDIV_SHIFT;
 		writel_relaxed(tmp, pll->base + 4);
@@ -239,13 +213,15 @@ static int clk_pll1443x_set_rate(struct clk_hw *hw, unsigned long drate,
 	}
 
 	tmp = readl_relaxed(pll->base + 4);
-	div_val = readl_relaxed(pll->base + 8);
 
-	if (!clk_pll1443x_mpk_change(rate, tmp, div_val)) {
+	if (!clk_pll14xx_mp_change(rate, tmp)) {
 		tmp &= ~(SDIV_MASK) << SDIV_SHIFT;
 		tmp |= rate->sdiv << SDIV_SHIFT;
 		writel_relaxed(tmp, pll->base + 4);
 
+		tmp = rate->kdiv << KDIV_SHIFT;
+		writel_relaxed(tmp, pll->base + 8);
+
 		return 0;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#36 lore raw

PATCH 30/205 dpaa2-eth: Fix minor bug in ethtool stats reporting

In reply to Sasha Levin

From: Ioana Radulescu <[email protected]>

[ Upstream commit 4b177f065e7ec37399b18e18412a8c7b75f8f299 ]

Don't print error message for a successful return value.

Fixes d84c3a4ded96 ("dpaa2-eth: Add new DPNI statistics counters")
---
 drivers/net/ethernet/freescale/dpaa2/dpaa2-ethtool.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/freescale/dpaa2/dpaa2-ethtool.c +1 -1
index 0aa1c34019bb..dc9a6c36cac0 100644
@@ -216,7 +216,7 @@ static void dpaa2_eth_get_ethtool_stats(struct net_device *net_dev,
 		if (err == -EINVAL)
 			/* Older firmware versions don't support all pages */
 			memset(&dpni_stats, 0, sizeof(dpni_stats));
-		else
+		else if (err)
 			netdev_warn(net_dev, "dpni_get_stats(%d) failed\n", j);
 
 		num_cnt = dpni_stats_page_size[j] / sizeof(u64);
Signature · 1 line
2.20.1
author maintainer
#37 lore raw

PATCH 31/205 arm64: dts: meson-gxl-s905x-khadas-vim: fix gpio-keys-polled node

In reply to Sasha Levin

From: Christian Hewitt <[email protected]>

[ Upstream commit d5f6fa904ecbadbb8e9fa6302b0fc165bec0559a ]

Fix DTC warnings:

arch/arm/dts/meson-gxl-s905x-khadas-vim.dtb: Warning (avoid_unnecessary_addr_size):
/gpio-keys-polled: unnecessary #address-cells/#size-cells
without "ranges" or child "reg" property

Fixes e15d2774b8c0 ("ARM64: dts: meson-gxl: add support for the Khadas VIM board")
---
 arch/arm64/boot/dts/amlogic/meson-gxl-s905x-khadas-vim.dts | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
arch/arm64/boot/dts/amlogic/meson-gxl-s905x-khadas-vim.dts +1 -3
index 8d6f316a5c7b..440bc23c7342 100644
@@ -33,11 +33,9 @@
 
 	gpio-keys-polled {
 		compatible = "gpio-keys-polled";
-		#address-cells = <1>;
-		#size-cells = <0>;
 		poll-interval = <100>;
 
-		button@0 {
+		power-button {
 			label = "power";
 			linux,code = <KEY_POWER>;
 			gpios = <&gpio_ao GPIOAO_2 GPIO_ACTIVE_LOW>;
Signature · 1 line
2.20.1
author maintainer
#38 lore raw

PATCH 32/205 clk: meson: axg-audio: fix regmap last register

In reply to Sasha Levin

From: Jerome Brunet <[email protected]>

[ Upstream commit 255cab9d27d78703f7450d720859ee146d0ee6e1 ]

Since the addition of the g12a, the last register is AUDIO_CLK_SPDIFOUT_B_CTRL.

Fixes 075001385c66 ("clk: meson: axg-audio: add g12a support")
---
 drivers/clk/meson/axg-audio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/meson/axg-audio.c +1 -1
index 18b23cdf679c..aa2522624fd3 100644
@@ -1001,7 +1001,7 @@ static const struct regmap_config axg_audio_regmap_cfg = {
 	.reg_bits	= 32,
 	.val_bits	= 32,
 	.reg_stride	= 4,
-	.max_register	= AUDIO_CLK_PDMIN_CTRL1,
+	.max_register	= AUDIO_CLK_SPDIFOUT_B_CTRL,
 };
 
 struct audioclk_data {
Signature · 1 line
2.20.1
author maintainer
#39 lore raw

PATCH 33/205 arm64: dts: marvell: Add AP806-dual missing CPU clocks

In reply to Sasha Levin

From: Miquel Raynal <[email protected]>

[ Upstream commit e231c6d47cca4b5df51bcf72dec1af767e63feaf ]

CPU clocks have been added to AP806-quad but not to the -dual variant.

Fixes c00bc38354cf ("arm64: dts: marvell: Add cpu clock node on Armada 7K/8K")
---
 arch/arm64/boot/dts/marvell/armada-ap806-dual.dtsi | 2 ++
 1 file changed, 2 insertions(+)
arch/arm64/boot/dts/marvell/armada-ap806-dual.dtsi +2 -0
index 9024a2d9db07..62ae016ee6aa 100644
@@ -21,6 +21,7 @@
 			reg = <0x000>;
 			enable-method = "psci";
 			#cooling-cells = <2>;
+			clocks = <&cpu_clk 0>;
 		};
 		cpu1: cpu@1 {
 			device_type = "cpu";
@@ -28,6 +29,7 @@
 			reg = <0x001>;
 			enable-method = "psci";
 			#cooling-cells = <2>;
+			clocks = <&cpu_clk 0>;
 		};
 	};
 };
Signature · 1 line
2.20.1
author maintainer
#40 lore raw

PATCH 34/205 soc: qcom: llcc: Name regmaps to avoid collisions

In reply to Sasha Levin

From: Stephen Boyd <[email protected]>

[ Upstream commit 2bfd3e7651addcaf48f12d4f11ea9d8fca6c3aa8 ]

We'll end up with debugfs collisions if we don't give names to the regmaps created by this driver. Change the name of the config before registering it so we don't collide in debugfs.

Fixes 7f9c136216c7 ("soc: qcom: Add broadcast base for Last Level Cache Controller (LLCC)")
Cc Venkata Narendra Kumar Gutta <[email protected]>
---
 drivers/soc/qcom/llcc-slice.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/soc/qcom/llcc-slice.c +2 -1
index 9090ea12eaf3..4a6111635f82 100644
@@ -48,7 +48,7 @@
 
 static struct llcc_drv_data *drv_data = (void *) -EPROBE_DEFER;
 
-static const struct regmap_config llcc_regmap_config = {
+static struct regmap_config llcc_regmap_config = {
 	.reg_bits = 32,
 	.reg_stride = 4,
 	.val_bits = 32,
@@ -323,6 +323,7 @@ static struct regmap *qcom_llcc_init_mmio(struct platform_device *pdev,
 	if (IS_ERR(base))
 		return ERR_CAST(base);
 
+	llcc_regmap_config.name = name;
 	return devm_regmap_init_mmio(&pdev->dev, base, &llcc_regmap_config);
 }
 
Signature · 1 line
2.20.1
author maintainer
#41 lore raw

PATCH 35/205 drm/rockchip: Round up _before_ giving to the clock framework

In reply to Sasha Levin

From: Douglas Anderson <[email protected]>

[ Upstream commit 287422a95fe28e05c1952de0472e0dfdffa6caae ]

I'm embarassed to say that even though I've touched vop_crtc_mode_fixup() twice and I swear I tested it, there's still a stupid glaring bug in it. Specifically, on veyron_minnie (with all the latest display timings) we want to be setting our pixel clock to 66,666,666.67 Hz and we tell userspace that's what we set, but we're actually choosing 66,000,000 Hz. This is confirmed by looking at the clock tree.

The problem is that in drm_display_mode_from_videomode() we convert from Hz to kHz with:

  dmode->clock = vm->pixelclock / 1000;

...and drm_display_mode_from_videomode() is called from panel-simple when we have an "override_mode" like we do on veyron_minnie. See commit 123643e5c40a ("ARM: dts: rockchip: Specify rk3288-veyron-minnie's display timings").

...so when the device tree specifies a clock of 66666667 for the panel then DRM translates that to 66666000. The clock framework will always pick a clock that is _lower_ than the one requested, so it will refuse to pick 66666667 and we'll end up at 66000000.

While we could try to fix drm_display_mode_from_videomode() to round to the nearest kHz and it would fix our problem, it wouldn't help if the clock we actually needed was 60,000,001 Hz. We could alternatively have DRM always round up, but maybe this would break someone else who already baked in the assumption that DRM rounds down. Specifically note that clock drivers are not consistent about whether they round up or round down when you call clk_set_rate(). We know how Rockchip's clock driver works, but (for instance) you can see that on most Qualcomm clocks the default is clk_rcg2_ops which rounds up.

Let's solve this by just adding 999 Hz before calling clk_round_rate(). This should be safe and work everywhere. As discussed in more detail in comments in the commit, Rockchip's PLLs are configured in a way that there shouldn't be another PLL setting that is only a few kHz off so we won't get mixed up.

NOTE: if this is picked to stable, it's probably easiest to first pick commit 527e4ca3b6d1 ("drm/rockchip: Base adjustments of the mode based on prev adjustments") which shouldn't hurt in stable.

---
 drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 37 +++++++++++++++++++--
 1 file changed, 34 insertions(+), 3 deletions(-)
drivers/gpu/drm/rockchip/rockchip_drm_vop.c +34 -3
index 613404f86668..84e3decb17b1 100644
@@ -1040,10 +1040,41 @@ static bool vop_crtc_mode_fixup(struct drm_crtc *crtc,
 				struct drm_display_mode *adjusted_mode)
 {
 	struct vop *vop = to_vop(crtc);
+	unsigned long rate;
 
-	adjusted_mode->clock =
-		DIV_ROUND_UP(clk_round_rate(vop->dclk,
-					    adjusted_mode->clock * 1000), 1000);
+	/*
+	 * Clock craziness.
+	 *
+	 * Key points:
+	 *
+	 * - DRM works in in kHz.
+	 * - Clock framework works in Hz.
+	 * - Rockchip's clock driver picks the clock rate that is the
+	 *   same _OR LOWER_ than the one requested.
+	 *
+	 * Action plan:
+	 *
+	 * 1. When DRM gives us a mode, we should add 999 Hz to it.  That way
+	 *    if the clock we need is 60000001 Hz (~60 MHz) and DRM tells us to
+	 *    make 60000 kHz then the clock framework will actually give us
+	 *    the right clock.
+	 *
+	 *    NOTE: if the PLL (maybe through a divider) could actually make
+	 *    a clock rate 999 Hz higher instead of the one we want then this
+	 *    could be a problem.  Unfortunately there's not much we can do
+	 *    since it's baked into DRM to use kHz.  It shouldn't matter in
+	 *    practice since Rockchip PLLs are controlled by tables and
+	 *    even if there is a divider in the middle I wouldn't expect PLL
+	 *    rates in the table that are just a few kHz different.
+	 *
+	 * 2. Get the clock framework to round the rate for us to tell us
+	 *    what it will actually make.
+	 *
+	 * 3. Store the rounded up rate so that we don't need to worry about
+	 *    this in the actual clk_set_rate().
+	 */
+	rate = clk_round_rate(vop->dclk, adjusted_mode->clock * 1000 + 999);
+	adjusted_mode->clock = DIV_ROUND_UP(rate, 1000);
 
 	return true;
 }
Signature · 1 line
2.20.1
author maintainer
#42 lore raw

PATCH 36/205 media: v4l: cadence: Fix how unsued lanes are handled in 'csi2rx_start()'

In reply to Sasha Levin

From: Christophe JAILLET <[email protected]>

[ Upstream commit 2eca8e4c1df4864b937752c3aa2f7925114f4806 ]

The 2nd parameter of 'find_first_zero_bit()' is a number of bits, not of bytes. So use 'csi2rx->max_lanes' instead of 'sizeof(lanes_used)'.

Fixes 1fc3b37f34f6 ("media: v4l: cadence: Add Cadence MIPI-CSI2 RX driver")
---
 drivers/media/platform/cadence/cdns-csi2rx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/cadence/cdns-csi2rx.c +1 -1
index 31ace114eda1..be9ec59774d6 100644
@@ -129,7 +129,7 @@ static int csi2rx_start(struct csi2rx_priv *csi2rx)
 	 */
 	for (i = csi2rx->num_lanes; i < csi2rx->max_lanes; i++) {
 		unsigned int idx = find_first_zero_bit(&lanes_used,
-						       sizeof(lanes_used));
+						       csi2rx->max_lanes);
 		set_bit(idx, &lanes_used);
 		reg |= CSI2RX_STATIC_CFG_DLANE_MAP(i, i + 1);
 	}
Signature · 1 line
2.20.1
author maintainer
#43 lore raw

PATCH 37/205 software node: Get reference to parent swnode in get_parent op

In reply to Sasha Levin

From: Sakari Ailus <[email protected]>

[ Upstream commit 51c100a651a471fcb8ead1ecc1224471eb0d61b9 ]

The software_node_get_parent() returned a pointer to the parent swnode, but did not take a reference to it, leading the caller to put a reference that was not taken. Take that reference now.

Fixes 59abd83672f7 ("drivers: base: Introducing software nodes to the firmware node framework")
---
 drivers/base/swnode.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/base/swnode.c +4 -1
index a1f3f0994f9f..d5b4905e2adb 100644
@@ -520,7 +520,10 @@ software_node_get_parent(const struct fwnode_handle *fwnode)
 {
 	struct swnode *swnode = to_swnode(fwnode);
 
-	return swnode ? (swnode->parent ? &swnode->parent->fwnode : NULL) : NULL;
+	if (!swnode || !swnode->parent)
+		return NULL;
+
+	return fwnode_handle_get(&swnode->parent->fwnode);
 }
 
 static struct fwnode_handle *
Signature · 1 line
2.20.1
author maintainer
#44 lore raw

PATCH 38/205 arm64: dts: apq8096-db820c: Increase load on l21 for SDCARD

In reply to Sasha Levin

From: Loic Poulain <[email protected]>

[ Upstream commit e38161bd325ea541ef2f258d8e28281077dde524 ]

In the same way as for msm8974-hammerhead, l21 load, used for SDCARD VMMC, needs to be increased in order to prevent any voltage drop issues (due to limited current) happening with some SDCARDS or during specific operations (e.g. write).

Fixes 660a9763c6a9 (arm64: dts: qcom: db820c: Add pm8994 regulator node)
---
 arch/arm64/boot/dts/qcom/apq8096-db820c.dtsi | 2 ++
 1 file changed, 2 insertions(+)
arch/arm64/boot/dts/qcom/apq8096-db820c.dtsi +2 -0
index 04ad2fb22b9a..dba3488492f1 100644
@@ -623,6 +623,8 @@
 				l21 {
 					regulator-min-microvolt = <2950000>;
 					regulator-max-microvolt = <2950000>;
+					regulator-allow-set-load;
+					regulator-system-load = <200000>;
 				};
 				l22 {
 					regulator-min-microvolt = <3300000>;
Signature · 1 line
2.20.1
author maintainer
#45 lore raw

PATCH 39/205 dmaengine: dw: platform: Mark 'hclk' clock optional

In reply to Sasha Levin

From: Andy Shevchenko <[email protected]>

[ Upstream commit f27c22736d133baff0ab3fdc7b015d998267d817 ]

On some platforms the clock can be fixed rate, always running one and there is no need to do anything with it.

In order to support those platforms, switch to use optional clock.

Fixes f8d9ddbc2851 ("dmaengine: dw: platform: Enable iDMA 32-bit on Intel Elkhart Lake")

Depends-on: 60b8f0ddf1a9 ("clk: Add (devm_)clk_get_optional() functions")

---
 drivers/dma/dw/platform.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/dw/platform.c +1 -1
index c90c798e5ec3..0585d749d935 100644
@@ -66,7 +66,7 @@ static int dw_probe(struct platform_device *pdev)
 
 	data->chip = chip;
 
-	chip->clk = devm_clk_get(chip->dev, "hclk");
+	chip->clk = devm_clk_get_optional(chip->dev, "hclk");
 	if (IS_ERR(chip->clk))
 		return PTR_ERR(chip->clk);
 	err = clk_prepare_enable(chip->clk);
Signature · 1 line
2.20.1
author maintainer
#46 lore raw

PATCH 40/205 pinctrl: sh-pfc: Fix PINMUX_IPSR_PHYS() to set GPSR

In reply to Sasha Levin

From: Keiya Nobuta <[email protected]>

[ Upstream commit d30710b8cce3a581c170d69002e311cc18ed47d3 ]

This patch allows PINMUX_IPSR_PHYS() to set bits in GPSR. When assigning function to pin, GPSR should be set to peripheral function.
For example when using SCL3, GPSR2 bit7 (PWM1_A pin) should be set to peripheral function.

Fixes 50d1ba1764b3e00a ("pinctrl: sh-pfc: Add physical pin multiplexing helper macros")
---
 drivers/pinctrl/sh-pfc/sh_pfc.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/pinctrl/sh-pfc/sh_pfc.h +2 -2
index 835148fc0f28..cab7da130925 100644
@@ -422,12 +422,12 @@ extern const struct sh_pfc_soc_info shx3_pinmux_info;
 /*
  * Describe a pinmux configuration in which a pin is physically multiplexed
  * with other pins.
- *   - ipsr: IPSR field (unused, for documentation purposes only)
+ *   - ipsr: IPSR field
  *   - fn: Function name
  *   - psel: Physical multiplexing selector
  */
 #define PINMUX_IPSR_PHYS(ipsr, fn, psel) \
-	PINMUX_DATA(fn##_MARK, FN_##psel)
+	PINMUX_DATA(fn##_MARK, FN_##psel, FN_##ipsr)
 
 /*
  * Describe a pinmux configuration for a single-function pin with GPIO
Signature · 1 line
2.20.1
author maintainer
#47 lore raw

PATCH 41/205 PCI: Fix missing bridge dma_ranges resource list cleanup

In reply to Sasha Levin

From: Rob Herring <[email protected]>

[ Upstream commit 7608158df3ed87a5c938c4a0b91f5b11101a9be1 ]

Commit e80a91ad302b ("PCI: Add dma_ranges window list") added a dma_ranges resource list, but failed to correctly free the list when devm_pci_alloc_host_bridge() is used.

Only the iproc host bridge driver is using the dma_ranges list.

Fixes e80a91ad302b ("PCI: Add dma_ranges window list")
Cc Srinath Mannam <[email protected]>
---
 drivers/pci/probe.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pci/probe.c +1 -0
index 64ebe3e5e611..d3033873395d 100644
@@ -572,6 +572,7 @@ static void devm_pci_release_host_bridge_dev(struct device *dev)
 		bridge->release_fn(bridge);
 
 	pci_free_resource_list(&bridge->windows);
+	pci_free_resource_list(&bridge->dma_ranges);
 }
 
 static void pci_release_host_bridge_dev(struct device *dev)
Signature · 1 line
2.20.1
author maintainer
#48 lore raw

PATCH 42/205 PCI: aardvark: Use LTSSM state to build link training flag

In reply to Sasha Levin

From: Remi Pommarel <[email protected]>

[ Upstream commit 364b3f1ff8f096d45f042a9c85daf7a1fc78413e ]

Aardvark's PCI_EXP_LNKSTA_LT flag in its link status register is not implemented and does not reflect the actual link training state (the flag is always set to 0). In order to support link re-training feature this flag has to be emulated. The Link Training and Status State Machine (LTSSM) flag in Aardvark LMI config register could be used as a link training indicator. Indeed if the LTSSM is in L0 or upper state then link training has completed (see [1]).

Unfortunately because after asking a link retraining it takes a while for the LTSSM state to become less than 0x10 (due to L0s to recovery state transition delays), LTSSM can still be in L0 while link training has not finished yet. So this waits for link to be in recovery or lesser state before returning after asking for a link retrain.

[1] "PCI Express Base Specification", REV. 4.0

    PCI Express, February 19 2014, Table 4-14
Fixes 8a3ebd8de328 ("PCI: aardvark: Implement emulated root PCI bridge config space")
---
 drivers/pci/controller/pci-aardvark.c | 29 ++++++++++++++++++++++++++-
 1 file changed, 28 insertions(+), 1 deletion(-)
drivers/pci/controller/pci-aardvark.c +28 -1
index fc0fe4d4de49..fe471861f801 100644
@@ -180,6 +180,8 @@
 #define LINK_WAIT_MAX_RETRIES		10
 #define LINK_WAIT_USLEEP_MIN		90000
 #define LINK_WAIT_USLEEP_MAX		100000
+#define RETRAIN_WAIT_MAX_RETRIES	10
+#define RETRAIN_WAIT_USLEEP_US		2000
 
 #define MSI_IRQ_NUM			32
 
@@ -239,6 +241,17 @@ static int advk_pcie_wait_for_link(struct advk_pcie *pcie)
 	return -ETIMEDOUT;
 }
 
+static void advk_pcie_wait_for_retrain(struct advk_pcie *pcie)
+{
+	size_t retries;
+
+	for (retries = 0; retries < RETRAIN_WAIT_MAX_RETRIES; ++retries) {
+		if (!advk_pcie_link_up(pcie))
+			break;
+		udelay(RETRAIN_WAIT_USLEEP_US);
+	}
+}
+
 static void advk_pcie_setup_hw(struct advk_pcie *pcie)
 {
 	u32 reg;
@@ -426,11 +439,20 @@ advk_pci_bridge_emul_pcie_conf_read(struct pci_bridge_emul *bridge,
 		return PCI_BRIDGE_EMUL_HANDLED;
 	}
 
+	case PCI_EXP_LNKCTL: {
+		/* u32 contains both PCI_EXP_LNKCTL and PCI_EXP_LNKSTA */
+		u32 val = advk_readl(pcie, PCIE_CORE_PCIEXP_CAP + reg) &
+			~(PCI_EXP_LNKSTA_LT << 16);
+		if (!advk_pcie_link_up(pcie))
+			val |= (PCI_EXP_LNKSTA_LT << 16);
+		*value = val;
+		return PCI_BRIDGE_EMUL_HANDLED;
+	}
+
 	case PCI_CAP_LIST_ID:
 	case PCI_EXP_DEVCAP:
 	case PCI_EXP_DEVCTL:
 	case PCI_EXP_LNKCAP:
-	case PCI_EXP_LNKCTL:
 		*value = advk_readl(pcie, PCIE_CORE_PCIEXP_CAP + reg);
 		return PCI_BRIDGE_EMUL_HANDLED;
 	default:
@@ -447,8 +469,13 @@ advk_pci_bridge_emul_pcie_conf_write(struct pci_bridge_emul *bridge,
 
 	switch (reg) {
 	case PCI_EXP_DEVCTL:
+		advk_writel(pcie, new, PCIE_CORE_PCIEXP_CAP + reg);
+		break;
+
 	case PCI_EXP_LNKCTL:
 		advk_writel(pcie, new, PCIE_CORE_PCIEXP_CAP + reg);
+		if (new & PCI_EXP_LNKCTL_RL)
+			advk_pcie_wait_for_retrain(pcie);
 		break;
 
 	case PCI_EXP_RTCTL:
Signature · 1 line
2.20.1
author maintainer
#49 lore raw

PATCH 43/205 PCI: aardvark: Fix PCI_EXP_RTCTL register configuration

In reply to Sasha Levin

From: Remi Pommarel <[email protected]>

[ Upstream commit c0f05a6ab52535c1bf5f43272eede3e11c5701a5 ]

PCI_EXP_RTCTL is used to activate PME interrupt only, so writing into it should not modify other interrupts' mask. The ISR mask polarity was also inverted, when PCI_EXP_RTCTL_PMEIE is set PCIE_MSG_PM_PME_MASK mask bit should actually be cleared.

Fixes 8a3ebd8de328 ("PCI: aardvark: Implement emulated root PCI bridge config space")
---
 drivers/pci/controller/pci-aardvark.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
drivers/pci/controller/pci-aardvark.c +9 -4
index fe471861f801..97245e076548 100644
@@ -428,7 +428,7 @@ advk_pci_bridge_emul_pcie_conf_read(struct pci_bridge_emul *bridge,
 
 	case PCI_EXP_RTCTL: {
 		u32 val = advk_readl(pcie, PCIE_ISR0_MASK_REG);
-		*value = (val & PCIE_MSG_PM_PME_MASK) ? PCI_EXP_RTCTL_PMEIE : 0;
+		*value = (val & PCIE_MSG_PM_PME_MASK) ? 0 : PCI_EXP_RTCTL_PMEIE;
 		return PCI_BRIDGE_EMUL_HANDLED;
 	}
 
@@ -478,10 +478,15 @@ advk_pci_bridge_emul_pcie_conf_write(struct pci_bridge_emul *bridge,
 			advk_pcie_wait_for_retrain(pcie);
 		break;
 
-	case PCI_EXP_RTCTL:
-		new = (new & PCI_EXP_RTCTL_PMEIE) << 3;
-		advk_writel(pcie, new, PCIE_ISR0_MASK_REG);
+	case PCI_EXP_RTCTL: {
+		/* Only mask/unmask PME interrupt */
+		u32 val = advk_readl(pcie, PCIE_ISR0_MASK_REG) &
+			~PCIE_MSG_PM_PME_MASK;
+		if ((new & PCI_EXP_RTCTL_PMEIE) == 0)
+			val |= PCIE_MSG_PM_PME_MASK;
+		advk_writel(pcie, val, PCIE_ISR0_MASK_REG);
 		break;
+	}
 
 	case PCI_EXP_RTSTA:
 		new = (new & PCI_EXP_RTSTA_PME) >> 9;
Signature · 1 line
2.20.1
author maintainer
#50 lore raw

PATCH 44/205 memory: mtk-smi: Add PM suspend and resume ops

In reply to Sasha Levin

From: Yong Wu <[email protected]>

[ Upstream commit fb03082a54acd66c61535edfefe96b2ff88ce7e2 ]

In the commit 4f0a1a1ae351 ("memory: mtk-smi: Invoke pm runtime_callback to enable clocks"), we use pm_runtime callback to enable/disable the smi larb clocks. It will cause the larb's clock may not be disabled when suspend. That is because device_prepare will call pm_runtime_get_noresume which will keep the larb's PM runtime status still is active when suspend, then it won't enter our pm_runtime suspend callback to disable the corresponding clocks.

This patch adds suspend pm_ops to force disable the clocks, Use "LATE" to make sure it disable the larb's clocks after the multimedia devices.

Fixes 4f0a1a1ae351 ("memory: mtk-smi: Invoke pm runtime_callback to enable clocks")
---
 drivers/memory/mtk-smi.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/memory/mtk-smi.c +4 -0
index 439d7d886873..a113e811faab 100644
@@ -366,6 +366,8 @@ static int __maybe_unused mtk_smi_larb_suspend(struct device *dev)
 
 static const struct dev_pm_ops smi_larb_pm_ops = {
 	SET_RUNTIME_PM_OPS(mtk_smi_larb_suspend, mtk_smi_larb_resume, NULL)
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				     pm_runtime_force_resume)
 };
 
 static struct platform_driver mtk_smi_larb_driver = {
@@ -507,6 +509,8 @@ static int __maybe_unused mtk_smi_common_suspend(struct device *dev)
 
 static const struct dev_pm_ops smi_common_pm_ops = {
 	SET_RUNTIME_PM_OPS(mtk_smi_common_suspend, mtk_smi_common_resume, NULL)
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				     pm_runtime_force_resume)
 };
 
 static struct platform_driver mtk_smi_common_driver = {
Signature · 1 line
2.20.1
author maintainer
#51 lore raw

PATCH 45/205 PCI: mobiveil: Fix csr_read()/write() build issue

In reply to Sasha Levin

From: Kefeng Wang <[email protected]>

[ Upstream commit 4906c05b87d44c19b225935e24d62e4480ca556d ]

RISCV has csr_read()/write() macros in arch/riscv/include/asm/csr.h.

The same function naming is used in the PCI mobiveil driver thus causing build error.

Rename csr_[read,write][l,] to mobiveil_csr_read()/write() to fix it.

drivers/pci/controller/pcie-mobiveil.c:238:69: error: macro "csr_read" passed 3 arguments, but takes just 1
static u32 csr_read(struct mobiveil_pcie *pcie, u32 off, size_t size)

drivers/pci/controller/pcie-mobiveil.c:253:80: error: macro "csr_write" passed 4 arguments, but takes just 2
static void csr_write(struct mobiveil_pcie *pcie, u32 val, u32 off, size_t size)

Fixes bcbe0d9a8d93 ("PCI: mobiveil: Unify register accessors")
Cc Hou Zhiqiang <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
Cc Minghuan Lian <[email protected]>
Cc Subrahmanya Lingappa <[email protected]>
Cc Andrew Murray <[email protected]>
---
 drivers/pci/controller/pcie-mobiveil.c | 119 +++++++++++++------------
 1 file changed, 62 insertions(+), 57 deletions(-)
drivers/pci/controller/pcie-mobiveil.c +62 -57
index a45a6447b01d..32f37d08d5bc 100644
@@ -235,7 +235,7 @@ static int mobiveil_pcie_write(void __iomem *addr, int size, u32 val)
 	return PCIBIOS_SUCCESSFUL;
 }
 
-static u32 csr_read(struct mobiveil_pcie *pcie, u32 off, size_t size)
+static u32 mobiveil_csr_read(struct mobiveil_pcie *pcie, u32 off, size_t size)
 {
 	void *addr;
 	u32 val;
@@ -250,7 +250,8 @@ static u32 csr_read(struct mobiveil_pcie *pcie, u32 off, size_t size)
 	return val;
 }
 
-static void csr_write(struct mobiveil_pcie *pcie, u32 val, u32 off, size_t size)
+static void mobiveil_csr_write(struct mobiveil_pcie *pcie, u32 val, u32 off,
+			       size_t size)
 {
 	void *addr;
 	int ret;
@@ -262,19 +263,19 @@ static void csr_write(struct mobiveil_pcie *pcie, u32 val, u32 off, size_t size)
 		dev_err(&pcie->pdev->dev, "write CSR address failed\n");
 }
 
-static u32 csr_readl(struct mobiveil_pcie *pcie, u32 off)
+static u32 mobiveil_csr_readl(struct mobiveil_pcie *pcie, u32 off)
 {
-	return csr_read(pcie, off, 0x4);
+	return mobiveil_csr_read(pcie, off, 0x4);
 }
 
-static void csr_writel(struct mobiveil_pcie *pcie, u32 val, u32 off)
+static void mobiveil_csr_writel(struct mobiveil_pcie *pcie, u32 val, u32 off)
 {
-	csr_write(pcie, val, off, 0x4);
+	mobiveil_csr_write(pcie, val, off, 0x4);
 }
 
 static bool mobiveil_pcie_link_up(struct mobiveil_pcie *pcie)
 {
-	return (csr_readl(pcie, LTSSM_STATUS) &
+	return (mobiveil_csr_readl(pcie, LTSSM_STATUS) &
 		LTSSM_STATUS_L0_MASK) == LTSSM_STATUS_L0;
 }
 
@@ -323,7 +324,7 @@ static void __iomem *mobiveil_pcie_map_bus(struct pci_bus *bus,
 		PCI_SLOT(devfn) << PAB_DEVICE_SHIFT |
 		PCI_FUNC(devfn) << PAB_FUNCTION_SHIFT;
 
-	csr_writel(pcie, value, PAB_AXI_AMAP_PEX_WIN_L(WIN_NUM_0));
+	mobiveil_csr_writel(pcie, value, PAB_AXI_AMAP_PEX_WIN_L(WIN_NUM_0));
 
 	return pcie->config_axi_slave_base + where;
 }
@@ -353,13 +354,14 @@ static void mobiveil_pcie_isr(struct irq_desc *desc)
 	chained_irq_enter(chip, desc);
 
 	/* read INTx status */
-	val = csr_readl(pcie, PAB_INTP_AMBA_MISC_STAT);
-	mask = csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
+	val = mobiveil_csr_readl(pcie, PAB_INTP_AMBA_MISC_STAT);
+	mask = mobiveil_csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
 	intr_status = val & mask;
 
 	/* Handle INTx */
 	if (intr_status & PAB_INTP_INTX_MASK) {
-		shifted_status = csr_readl(pcie, PAB_INTP_AMBA_MISC_STAT);
+		shifted_status = mobiveil_csr_readl(pcie,
+						    PAB_INTP_AMBA_MISC_STAT);
 		shifted_status &= PAB_INTP_INTX_MASK;
 		shifted_status >>= PAB_INTX_START;
 		do {
@@ -373,12 +375,13 @@ static void mobiveil_pcie_isr(struct irq_desc *desc)
 							    bit);
 
 				/* clear interrupt handled */
-				csr_writel(pcie, 1 << (PAB_INTX_START + bit),
-					   PAB_INTP_AMBA_MISC_STAT);
+				mobiveil_csr_writel(pcie,
+						    1 << (PAB_INTX_START + bit),
+						    PAB_INTP_AMBA_MISC_STAT);
 			}
 
-			shifted_status = csr_readl(pcie,
-						   PAB_INTP_AMBA_MISC_STAT);
+			shifted_status = mobiveil_csr_readl(pcie,
+							    PAB_INTP_AMBA_MISC_STAT);
 			shifted_status &= PAB_INTP_INTX_MASK;
 			shifted_status >>= PAB_INTX_START;
 		} while (shifted_status != 0);
@@ -413,7 +416,7 @@ static void mobiveil_pcie_isr(struct irq_desc *desc)
 	}
 
 	/* Clear the interrupt status */
-	csr_writel(pcie, intr_status, PAB_INTP_AMBA_MISC_STAT);
+	mobiveil_csr_writel(pcie, intr_status, PAB_INTP_AMBA_MISC_STAT);
 	chained_irq_exit(chip, desc);
 }
 
@@ -474,24 +477,24 @@ static void program_ib_windows(struct mobiveil_pcie *pcie, int win_num,
 		return;
 	}
 
-	value = csr_readl(pcie, PAB_PEX_AMAP_CTRL(win_num));
+	value = mobiveil_csr_readl(pcie, PAB_PEX_AMAP_CTRL(win_num));
 	value &= ~(AMAP_CTRL_TYPE_MASK << AMAP_CTRL_TYPE_SHIFT | WIN_SIZE_MASK);
 	value |= type << AMAP_CTRL_TYPE_SHIFT | 1 << AMAP_CTRL_EN_SHIFT |
 		 (lower_32_bits(size64) & WIN_SIZE_MASK);
-	csr_writel(pcie, value, PAB_PEX_AMAP_CTRL(win_num));
+	mobiveil_csr_writel(pcie, value, PAB_PEX_AMAP_CTRL(win_num));
 
-	csr_writel(pcie, upper_32_bits(size64),
-		   PAB_EXT_PEX_AMAP_SIZEN(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(size64),
+			    PAB_EXT_PEX_AMAP_SIZEN(win_num));
 
-	csr_writel(pcie, lower_32_bits(cpu_addr),
-		   PAB_PEX_AMAP_AXI_WIN(win_num));
-	csr_writel(pcie, upper_32_bits(cpu_addr),
-		   PAB_EXT_PEX_AMAP_AXI_WIN(win_num));
+	mobiveil_csr_writel(pcie, lower_32_bits(cpu_addr),
+			    PAB_PEX_AMAP_AXI_WIN(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(cpu_addr),
+			    PAB_EXT_PEX_AMAP_AXI_WIN(win_num));
 
-	csr_writel(pcie, lower_32_bits(pci_addr),
-		   PAB_PEX_AMAP_PEX_WIN_L(win_num));
-	csr_writel(pcie, upper_32_bits(pci_addr),
-		   PAB_PEX_AMAP_PEX_WIN_H(win_num));
+	mobiveil_csr_writel(pcie, lower_32_bits(pci_addr),
+			    PAB_PEX_AMAP_PEX_WIN_L(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(pci_addr),
+			    PAB_PEX_AMAP_PEX_WIN_H(win_num));
 
 	pcie->ib_wins_configured++;
 }
@@ -515,27 +518,29 @@ static void program_ob_windows(struct mobiveil_pcie *pcie, int win_num,
 	 * program Enable Bit to 1, Type Bit to (00) base 2, AXI Window Size Bit
 	 * to 4 KB in PAB_AXI_AMAP_CTRL register
 	 */
-	value = csr_readl(pcie, PAB_AXI_AMAP_CTRL(win_num));
+	value = mobiveil_csr_readl(pcie, PAB_AXI_AMAP_CTRL(win_num));
 	value &= ~(WIN_TYPE_MASK << WIN_TYPE_SHIFT | WIN_SIZE_MASK);
 	value |= 1 << WIN_ENABLE_SHIFT | type << WIN_TYPE_SHIFT |
 		 (lower_32_bits(size64) & WIN_SIZE_MASK);
-	csr_writel(pcie, value, PAB_AXI_AMAP_CTRL(win_num));
+	mobiveil_csr_writel(pcie, value, PAB_AXI_AMAP_CTRL(win_num));
 
-	csr_writel(pcie, upper_32_bits(size64), PAB_EXT_AXI_AMAP_SIZE(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(size64),
+			    PAB_EXT_AXI_AMAP_SIZE(win_num));
 
 	/*
 	 * program AXI window base with appropriate value in
 	 * PAB_AXI_AMAP_AXI_WIN0 register
 	 */
-	csr_writel(pcie, lower_32_bits(cpu_addr) & (~AXI_WINDOW_ALIGN_MASK),
-		   PAB_AXI_AMAP_AXI_WIN(win_num));
-	csr_writel(pcie, upper_32_bits(cpu_addr),
-		   PAB_EXT_AXI_AMAP_AXI_WIN(win_num));
+	mobiveil_csr_writel(pcie,
+			    lower_32_bits(cpu_addr) & (~AXI_WINDOW_ALIGN_MASK),
+			    PAB_AXI_AMAP_AXI_WIN(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(cpu_addr),
+			    PAB_EXT_AXI_AMAP_AXI_WIN(win_num));
 
-	csr_writel(pcie, lower_32_bits(pci_addr),
-		   PAB_AXI_AMAP_PEX_WIN_L(win_num));
-	csr_writel(pcie, upper_32_bits(pci_addr),
-		   PAB_AXI_AMAP_PEX_WIN_H(win_num));
+	mobiveil_csr_writel(pcie, lower_32_bits(pci_addr),
+			    PAB_AXI_AMAP_PEX_WIN_L(win_num));
+	mobiveil_csr_writel(pcie, upper_32_bits(pci_addr),
+			    PAB_AXI_AMAP_PEX_WIN_H(win_num));
 
 	pcie->ob_wins_configured++;
 }
@@ -579,42 +584,42 @@ static int mobiveil_host_init(struct mobiveil_pcie *pcie)
 	struct resource_entry *win;
 
 	/* setup bus numbers */
-	value = csr_readl(pcie, PCI_PRIMARY_BUS);
+	value = mobiveil_csr_readl(pcie, PCI_PRIMARY_BUS);
 	value &= 0xff000000;
 	value |= 0x00ff0100;
-	csr_writel(pcie, value, PCI_PRIMARY_BUS);
+	mobiveil_csr_writel(pcie, value, PCI_PRIMARY_BUS);
 
 	/*
 	 * program Bus Master Enable Bit in Command Register in PAB Config
 	 * Space
 	 */
-	value = csr_readl(pcie, PCI_COMMAND);
+	value = mobiveil_csr_readl(pcie, PCI_COMMAND);
 	value |= PCI_COMMAND_IO | PCI_COMMAND_MEMORY | PCI_COMMAND_MASTER;
-	csr_writel(pcie, value, PCI_COMMAND);
+	mobiveil_csr_writel(pcie, value, PCI_COMMAND);
 
 	/*
 	 * program PIO Enable Bit to 1 (and PEX PIO Enable to 1) in PAB_CTRL
 	 * register
 	 */
-	pab_ctrl = csr_readl(pcie, PAB_CTRL);
+	pab_ctrl = mobiveil_csr_readl(pcie, PAB_CTRL);
 	pab_ctrl |= (1 << AMBA_PIO_ENABLE_SHIFT) | (1 << PEX_PIO_ENABLE_SHIFT);
-	csr_writel(pcie, pab_ctrl, PAB_CTRL);
+	mobiveil_csr_writel(pcie, pab_ctrl, PAB_CTRL);
 
-	csr_writel(pcie, (PAB_INTP_INTX_MASK | PAB_INTP_MSI_MASK),
-		   PAB_INTP_AMBA_MISC_ENB);
+	mobiveil_csr_writel(pcie, (PAB_INTP_INTX_MASK | PAB_INTP_MSI_MASK),
+			    PAB_INTP_AMBA_MISC_ENB);
 
 	/*
 	 * program PIO Enable Bit to 1 and Config Window Enable Bit to 1 in
 	 * PAB_AXI_PIO_CTRL Register
 	 */
-	value = csr_readl(pcie, PAB_AXI_PIO_CTRL);
+	value = mobiveil_csr_readl(pcie, PAB_AXI_PIO_CTRL);
 	value |= APIO_EN_MASK;
-	csr_writel(pcie, value, PAB_AXI_PIO_CTRL);
+	mobiveil_csr_writel(pcie, value, PAB_AXI_PIO_CTRL);
 
 	/* Enable PCIe PIO master */
-	value = csr_readl(pcie, PAB_PEX_PIO_CTRL);
+	value = mobiveil_csr_readl(pcie, PAB_PEX_PIO_CTRL);
 	value |= 1 << PIO_ENABLE_SHIFT;
-	csr_writel(pcie, value, PAB_PEX_PIO_CTRL);
+	mobiveil_csr_writel(pcie, value, PAB_PEX_PIO_CTRL);
 
 	/*
 	 * we'll program one outbound window for config reads and
@@ -647,10 +652,10 @@ static int mobiveil_host_init(struct mobiveil_pcie *pcie)
 	}
 
 	/* fixup for PCIe class register */
-	value = csr_readl(pcie, PAB_INTP_AXI_PIO_CLASS);
+	value = mobiveil_csr_readl(pcie, PAB_INTP_AXI_PIO_CLASS);
 	value &= 0xff;
 	value |= (PCI_CLASS_BRIDGE_PCI << 16);
-	csr_writel(pcie, value, PAB_INTP_AXI_PIO_CLASS);
+	mobiveil_csr_writel(pcie, value, PAB_INTP_AXI_PIO_CLASS);
 
 	/* setup MSI hardware registers */
 	mobiveil_pcie_enable_msi(pcie);
@@ -668,9 +673,9 @@ static void mobiveil_mask_intx_irq(struct irq_data *data)
 	pcie = irq_desc_get_chip_data(desc);
 	mask = 1 << ((data->hwirq + PAB_INTX_START) - 1);
 	raw_spin_lock_irqsave(&pcie->intx_mask_lock, flags);
-	shifted_val = csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
+	shifted_val = mobiveil_csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
 	shifted_val &= ~mask;
-	csr_writel(pcie, shifted_val, PAB_INTP_AMBA_MISC_ENB);
+	mobiveil_csr_writel(pcie, shifted_val, PAB_INTP_AMBA_MISC_ENB);
 	raw_spin_unlock_irqrestore(&pcie->intx_mask_lock, flags);
 }
 
@@ -684,9 +689,9 @@ static void mobiveil_unmask_intx_irq(struct irq_data *data)
 	pcie = irq_desc_get_chip_data(desc);
 	mask = 1 << ((data->hwirq + PAB_INTX_START) - 1);
 	raw_spin_lock_irqsave(&pcie->intx_mask_lock, flags);
-	shifted_val = csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
+	shifted_val = mobiveil_csr_readl(pcie, PAB_INTP_AMBA_MISC_ENB);
 	shifted_val |= mask;
-	csr_writel(pcie, shifted_val, PAB_INTP_AMBA_MISC_ENB);
+	mobiveil_csr_writel(pcie, shifted_val, PAB_INTP_AMBA_MISC_ENB);
 	raw_spin_unlock_irqrestore(&pcie->intx_mask_lock, flags);
 }
 
Signature · 1 line
2.20.1
author maintainer
#52 lore raw

PATCH 46/205 PCI: dwc: Fix find_next_bit() usage

In reply to Sasha Levin

From: Niklas Cassel <[email protected]>

[ Upstream commit 1137e61dcb99f7f8b54e77ed83f68b5b485a3e34 ]

find_next_bit() takes a parameter of size long, and performs arithmetic that assumes that the argument is of size long.

Therefore we cannot pass a u32, since this will cause find_next_bit() to read outside the stack buffer and will produce the following print:

BUG: KASAN: stack-out-of-bounds in find_next_bit+0x38/0xb0
Fixes 1b497e6493c4 ("PCI: dwc: Fix uninitialized variable in dw_handle_msi_irq()")
---
 drivers/pci/controller/dwc/pcie-designware-host.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/pci/controller/dwc/pcie-designware-host.c +6 -5
index 0f36a926059a..8615f1548882 100644
@@ -78,7 +78,8 @@ static struct msi_domain_info dw_pcie_msi_domain_info = {
 irqreturn_t dw_handle_msi_irq(struct pcie_port *pp)
 {
 	int i, pos, irq;
-	u32 val, num_ctrls;
+	unsigned long val;
+	u32 status, num_ctrls;
 	irqreturn_t ret = IRQ_NONE;
 
 	num_ctrls = pp->num_vectors / MAX_MSI_IRQS_PER_CTRL;
@@ -86,14 +87,14 @@ irqreturn_t dw_handle_msi_irq(struct pcie_port *pp)
 	for (i = 0; i < num_ctrls; i++) {
 		dw_pcie_rd_own_conf(pp, PCIE_MSI_INTR0_STATUS +
 					(i * MSI_REG_CTRL_BLOCK_SIZE),
-				    4, &val);
-		if (!val)
+				    4, &status);
+		if (!status)
 			continue;
 
 		ret = IRQ_HANDLED;
+		val = status;
 		pos = 0;
-		while ((pos = find_next_bit((unsigned long *) &val,
-					    MAX_MSI_IRQS_PER_CTRL,
+		while ((pos = find_next_bit(&val, MAX_MSI_IRQS_PER_CTRL,
 					    pos)) != MAX_MSI_IRQS_PER_CTRL) {
 			irq = irq_find_mapping(pp->irq_domain,
 					       (i * MAX_MSI_IRQS_PER_CTRL) +
Signature · 1 line
2.20.1
author maintainer
#53 lore raw

PATCH 47/205 iio: imu: st_lsm6dsx: fix gyro gain definitions for LSM9DS1

In reply to Sasha Levin

From: Lorenzo Bianconi <[email protected]>

[ Upstream commit 1b3751017e09f0857bc38f9b1be08dce38f3d92c ]

Fix typos in gyro gain definitions for LSM9DS1 sensor

Fixes 52f4b1f19679 ("iio: imu: st_lsm6dsx: add support for accel/gyro unit of lsm9ds1")
---
 drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/iio/imu/st_lsm6dsx/st_lsm6dsx_core.c +4 -3
index 28e011b35f21..3e0528793d95 100644
@@ -152,9 +152,10 @@ static const struct st_lsm6dsx_settings st_lsm6dsx_sensor_settings[] = {
 					.addr = 0x10,
 					.mask = GENMASK(4, 3),
 				},
-				.fs_avl[0] = {  IIO_DEGREE_TO_RAD(245), 0x0 },
-				.fs_avl[1] = {  IIO_DEGREE_TO_RAD(500), 0x1 },
-				.fs_avl[2] = { IIO_DEGREE_TO_RAD(2000), 0x3 },
+
+				.fs_avl[0] = {  IIO_DEGREE_TO_RAD(8750), 0x0 },
+				.fs_avl[1] = { IIO_DEGREE_TO_RAD(17500), 0x1 },
+				.fs_avl[2] = { IIO_DEGREE_TO_RAD(70000), 0x3 },
 				.fs_len = 3,
 			},
 		},
Signature · 1 line
2.20.1
author maintainer
#54 lore raw

PATCH 48/205 drm: rcar_lvds: Fix color mismatches on R-Car H2 ES2.0 and later

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 3986457110a054466bf02f9c4a85aa2bba96177b ]

Commit 5cca30ebe089be23 ("drm/rcar-du: Add LVDS_LANES quirk") states that LVDS lanes 1 and 3 are inverted on R-Car H2 ES1 only, and that the problem has been fixed in newer revisions.

However, the code didn't take into account the actual hardware revision, thus applying the quirk also on newer hardware revisions, causing green color reversals.

Fix this by applying the quirk when running on R-Car H2 ES1.x only.

---
 drivers/gpu/drm/rcar-du/rcar_lvds.c | 28 +++++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)
drivers/gpu/drm/rcar-du/rcar_lvds.c +21 -7
index 3fc7e6899cab..50c11a7f0467 100644
@@ -16,6 +16,7 @@
 #include <linux/of_graph.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/sys_soc.h>
 
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
@@ -842,8 +843,23 @@ static int rcar_lvds_get_clocks(struct rcar_lvds *lvds)
 	return 0;
 }
 
+static const struct rcar_lvds_device_info rcar_lvds_r8a7790es1_info = {
+	.gen = 2,
+	.quirks = RCAR_LVDS_QUIRK_LANES,
+	.pll_setup = rcar_lvds_pll_setup_gen2,
+};
+
+static const struct soc_device_attribute lvds_quirk_matches[] = {
+	{
+		.soc_id = "r8a7790", .revision = "ES1.*",
+		.data = &rcar_lvds_r8a7790es1_info,
+	},
+	{ /* sentinel */ }
+};
+
 static int rcar_lvds_probe(struct platform_device *pdev)
 {
+	const struct soc_device_attribute *attr;
 	struct rcar_lvds *lvds;
 	struct resource *mem;
 	int ret;
@@ -857,6 +873,10 @@ static int rcar_lvds_probe(struct platform_device *pdev)
 	lvds->dev = &pdev->dev;
 	lvds->info = of_device_get_match_data(&pdev->dev);
 
+	attr = soc_device_match(lvds_quirk_matches);
+	if (attr)
+		lvds->info = attr->data;
+
 	ret = rcar_lvds_parse_dt(lvds);
 	if (ret < 0)
 		return ret;
@@ -893,12 +913,6 @@ static const struct rcar_lvds_device_info rcar_lvds_gen2_info = {
 	.pll_setup = rcar_lvds_pll_setup_gen2,
 };
 
-static const struct rcar_lvds_device_info rcar_lvds_r8a7790_info = {
-	.gen = 2,
-	.quirks = RCAR_LVDS_QUIRK_LANES,
-	.pll_setup = rcar_lvds_pll_setup_gen2,
-};
-
 static const struct rcar_lvds_device_info rcar_lvds_gen3_info = {
 	.gen = 3,
 	.quirks = RCAR_LVDS_QUIRK_PWD,
@@ -930,7 +944,7 @@ static const struct of_device_id rcar_lvds_of_table[] = {
 	{ .compatible = "renesas,r8a7744-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a774a1-lvds", .data = &rcar_lvds_gen3_info },
 	{ .compatible = "renesas,r8a774c0-lvds", .data = &rcar_lvds_r8a77990_info },
-	{ .compatible = "renesas,r8a7790-lvds", .data = &rcar_lvds_r8a7790_info },
+	{ .compatible = "renesas,r8a7790-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7791-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7793-lvds", .data = &rcar_lvds_gen2_info },
 	{ .compatible = "renesas,r8a7795-lvds", .data = &rcar_lvds_gen3_info },
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#55 lore raw

PATCH 49/205 net: netsec: Correct dma sync for XDP_TX frames

In reply to Sasha Levin

From: Ilias Apalodimas <[email protected]>

[ Upstream commit d9496f3ecfe4823c1e12aecbcc29220147fa012c ]

bpf_xdp_adjust_head() can change the frame boundaries. Account for the potential shift properly by calculating the new offset before syncing the buffer to the device for XDP_TX

Fixes ba2b232108d3 ("net: netsec: add XDP support")
---
 drivers/net/ethernet/socionext/netsec.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/socionext/netsec.c +2 -2
index f9e6744d8fd6..41ddd8fff2a7 100644
@@ -847,8 +847,8 @@ static u32 netsec_xdp_queue_one(struct netsec_priv *priv,
 		enum dma_data_direction dma_dir =
 			page_pool_get_dma_dir(rx_ring->page_pool);
 
-		dma_handle = page_pool_get_dma_addr(page) +
-			NETSEC_RXBUF_HEADROOM;
+		dma_handle = page_pool_get_dma_addr(page) + xdpf->headroom +
+			sizeof(*xdpf);
 		dma_sync_single_for_device(priv->dev, dma_handle, xdpf->len,
 					   dma_dir);
 		tx_desc.buf_type = TYPE_NETSEC_XDP_TX;
Signature · 1 line
2.20.1
author maintainer
#56 lore raw

PATCH 50/205 ACPI: platform: Unregister stale platform devices

In reply to Sasha Levin

From: Andy Shevchenko <[email protected]>

[ Upstream commit cb0701acfa7e3fe9e919cf2aa2aa939b7fd603c2 ]

When commit 68bdb6773289 ("ACPI: add support for ACPI reconfiguration notifiers") introduced reconfiguration notifiers, it missed the point that the ACPI table, which might be loaded and then unloaded via ConfigFS, could contain devices that were not enumerated by their parents.

In such cases, the stale platform device is dangling in the system while the rest of the devices from the same table are already gone.

Introduce acpi_platform_device_remove_notify() notifier that, in similar way to I²C or SPI buses, unregisters the platform devices on table removal event.

Fixes 68bdb6773289 ("ACPI: add support for ACPI reconfiguration notifiers")

Depends-on: 00500147cbd3 ("drivers: Introduce device lookup variants by ACPI_COMPANION device")

[ rjw: Changelog & function rename ]
---
 drivers/acpi/acpi_platform.c | 43 ++++++++++++++++++++++++++++++++++++
 drivers/acpi/scan.c          |  1 +
 2 files changed, 44 insertions(+)
drivers/acpi/acpi_platform.c +43 -0
index 00ec4f2bf015..c05050f474cd 100644
@@ -31,6 +31,44 @@ static const struct acpi_device_id forbidden_id_list[] = {
 	{"", 0},
 };
 
+static struct platform_device *acpi_platform_device_find_by_companion(struct acpi_device *adev)
+{
+	struct device *dev;
+
+	dev = bus_find_device_by_acpi_dev(&platform_bus_type, adev);
+	return dev ? to_platform_device(dev) : NULL;
+}
+
+static int acpi_platform_device_remove_notify(struct notifier_block *nb,
+					      unsigned long value, void *arg)
+{
+	struct acpi_device *adev = arg;
+	struct platform_device *pdev;
+
+	switch (value) {
+	case ACPI_RECONFIG_DEVICE_ADD:
+		/* Nothing to do here */
+		break;
+	case ACPI_RECONFIG_DEVICE_REMOVE:
+		if (!acpi_device_enumerated(adev))
+			break;
+
+		pdev = acpi_platform_device_find_by_companion(adev);
+		if (!pdev)
+			break;
+
+		platform_device_unregister(pdev);
+		put_device(&pdev->dev);
+		break;
+	}
+
+	return NOTIFY_OK;
+}
+
+static struct notifier_block acpi_platform_notifier = {
+	.notifier_call = acpi_platform_device_remove_notify,
+};
+
 static void acpi_platform_fill_resource(struct acpi_device *adev,
 	const struct resource *src, struct resource *dest)
 {
@@ -130,3 +168,8 @@ struct platform_device *acpi_create_platform_device(struct acpi_device *adev,
 	return pdev;
 }
 EXPORT_SYMBOL_GPL(acpi_create_platform_device);
+
+void __init acpi_platform_init(void)
+{
+	acpi_reconfig_notifier_register(&acpi_platform_notifier);
+}
drivers/acpi/scan.c +1 -0
index aad6be5c0af0..915650bf519f 100644
@@ -2174,6 +2174,7 @@ int __init acpi_scan_init(void)
 	acpi_pci_root_init();
 	acpi_pci_link_init();
 	acpi_processor_init();
+	acpi_platform_init();
 	acpi_lpss_init();
 	acpi_apd_init();
 	acpi_cmos_rtc_init();
Signature · 1 line
2.20.1
author maintainer
#57 lore raw

PATCH 51/205 spi: pxa2xx: Set controller->max_transfer_size in dma mode

In reply to Sasha Levin

From: Daniel Vetter <[email protected]>

[ Upstream commit b2662a164f9dc48da8822e56600686d639056282 ]

In DMA mode we have a maximum transfer size, past that the driver falls back to PIO (see the check at the top of pxa2xx_spi_transfer_one). Falling back to PIO for big transfers defeats the point of a dma engine, hence set the max transfer size to inform spi clients that they need to do something smarter.

This was uncovered by the drm_mipi_dbi spi panel code, which does large spi transfers, but stopped splitting them after:

commit e143364b4c1774f68e923a5a0bb0fca28ac25888

Author: Noralf Trønnes <[email protected]>
Date:   Fri Jul 19 17:59:10 2019 +0200

drm/tinydrm: Remove tinydrm_spi_max_transfer_size()

After this commit the code relied on the spi core to split transfers into max dma-able blocks, which also papered over the PIO fallback issue.

Fix this by setting the overall max transfer size to the DMA limit, but only when the controller runs in DMA mode.

Fixes e143364b4c17 ("drm/tinydrm: Remove tinydrm_spi_max_transfer_size()")
Cc Sam Ravnborg <[email protected]>
Cc Noralf Trønnes <[email protected]>
Cc Andy Shevchenko <[email protected]>
Cc Daniel Mack <[email protected]>
Cc Haojian Zhuang <[email protected]>
Cc Robert Jarzmik <[email protected]>
Cc Mark Brown <[email protected]>
---
 drivers/spi/spi-pxa2xx.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/spi/spi-pxa2xx.c +7 -0
index ae95ec0bc964..9f92165fe09f 100644
@@ -1612,6 +1612,11 @@ static int pxa2xx_spi_fw_translate_cs(struct spi_controller *controller,
 	return cs;
 }
 
+static size_t pxa2xx_spi_max_dma_transfer_size(struct spi_device *spi)
+{
+	return MAX_DMA_LEN;
+}
+
 static int pxa2xx_spi_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
@@ -1717,6 +1722,8 @@ static int pxa2xx_spi_probe(struct platform_device *pdev)
 		} else {
 			controller->can_dma = pxa2xx_spi_can_dma;
 			controller->max_dma_len = MAX_DMA_LEN;
+			controller->max_transfer_size =
+				pxa2xx_spi_max_dma_transfer_size;
 		}
 	}
 
Signature · 1 line
2.20.1
author maintainer
#58 lore raw

PATCH 52/205 spi: atmel: fix handling of cs_change set on non-last xfer

In reply to Sasha Levin

From: Mans Rullgard <[email protected]>

[ Upstream commit fed8d8c7a6dc2a76d7764842853d81c770b0788e ]

The driver does the wrong thing when cs_change is set on a non-last xfer in a message. When cs_change is set, the driver deactivates the CS and leaves it off until a later xfer again has cs_change set whereas it should be briefly toggling CS off and on again.

This patch brings the behaviour of the driver back in line with the documentation and common sense. The delay of 10 us is the same as is used by the default spi_transfer_one_message() function in spi.c.

[gregory: rebased on for-5.5 from spi tree]
Fixes 8090d6d1a415 ("spi: atmel: Refactor spi-atmel to use SPI framework queue")
---
 drivers/spi/spi-atmel.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)
drivers/spi/spi-atmel.c +3 -7
index ba8eff41b746..abbc1582f457 100644
@@ -302,7 +302,6 @@ struct atmel_spi {
 	bool			use_cs_gpios;
 
 	bool			keep_cs;
-	bool			cs_active;
 
 	u32			fifo_size;
 };
@@ -1374,11 +1373,9 @@ static int atmel_spi_one_transfer(struct spi_master *master,
 				 &msg->transfers)) {
 			as->keep_cs = true;
 		} else {
-			as->cs_active = !as->cs_active;
-			if (as->cs_active)
-				cs_activate(as, msg->spi);
-			else
-				cs_deactivate(as, msg->spi);
+			cs_deactivate(as, msg->spi);
+			udelay(10);
+			cs_activate(as, msg->spi);
 		}
 	}
 
@@ -1401,7 +1398,6 @@ static int atmel_spi_transfer_one_message(struct spi_master *master,
 	atmel_spi_lock(as);
 	cs_activate(as, spi);
 
-	as->cs_active = true;
 	as->keep_cs = false;
 
 	msg->status = 0;
Signature · 1 line
2.20.1
author maintainer
#59 lore raw

PATCH 53/205 pwm: sun4i: Fix incorrect calculation of duty_cycle/period

In reply to Sasha Levin

From: Ondrej Jirman <[email protected]>

[ Upstream commit 50cc7e3e4f26e3bf5ed74a8d061195c4d2161b8b ]

Since 5.4-rc1, pwm_apply_state calls ->get_state after ->apply if available, and this revealed an issue with integer precision when calculating duty_cycle and period for the currently set state in ->get_state callback.

This issue manifested in broken backlight on several Allwinner based devices.

Previously this worked, because ->apply updated the passed state directly.

Fixes deb9c462f4e53 ("pwm: sun4i: Don't update the state for the caller of pwm_apply_state")
---
 drivers/pwm/pwm-sun4i.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/pwm/pwm-sun4i.c +2 -2
index 6f5840a1a82d..05273725a9ff 100644
@@ -137,10 +137,10 @@ static void sun4i_pwm_get_state(struct pwm_chip *chip,
 
 	val = sun4i_pwm_readl(sun4i_pwm, PWM_CH_PRD(pwm->hwpwm));
 
-	tmp = prescaler * NSEC_PER_SEC * PWM_REG_DTY(val);
+	tmp = (u64)prescaler * NSEC_PER_SEC * PWM_REG_DTY(val);
 	state->duty_cycle = DIV_ROUND_CLOSEST_ULL(tmp, clk_rate);
 
-	tmp = prescaler * NSEC_PER_SEC * PWM_REG_PRD(val);
+	tmp = (u64)prescaler * NSEC_PER_SEC * PWM_REG_PRD(val);
 	state->period = DIV_ROUND_CLOSEST_ULL(tmp, clk_rate);
 }
 
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#60 lore raw

Re: [PATCH AUTOSEL 5.4 053/205] pwm: sun4i: Fix incorrect calculation of duty_cycle/period

Quoted 12 lines · Sasha Levin

From: Ondrej Jirman <[email protected]>

[ Upstream commit 50cc7e3e4f26e3bf5ed74a8d061195c4d2161b8b ]

Since 5.4-rc1, pwm_apply_state calls ->get_state after ->apply if available, and this revealed an issue with integer precision when calculating duty_cycle and period for the currently set state in ->get_state callback.

This issue manifested in broken backlight on several Allwinner based devices.

Quoted 8 lines

Previously this worked, because ->apply updated the passed state directly.

Fixes deb9c462f4e53 ("pwm: sun4i: Don't update the state for the caller of pwm_apply_state")

Note that while the patch is still correct, the problem isn't that bad any more since commit 01ccf903edd65f6421612321648fa5a7f4b7cb10 was reverted.

So .get_state is only called once during boot where the breakage doesn't hurt that much.

I let you decide if you still want to apply this patch.

Best regards
Uwe

Signature · 2 lines
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |
author maintainer
#61 lore raw

PATCH 54/205 RDMA/hns: Release qp resources when failed to destroy qp

In reply to Sasha Levin

From: Yangyang Li <[email protected]>

[ Upstream commit d302c6e3a6895608a5856bc708c47bda1770b24d ]

Even if no response from hardware, we should make sure that qp related resources are released to avoid memory leaks.

Fixes 926a01dc000d ("RDMA/hns: Add QP operations support for hip08 SoC")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +4 -8
index e82567fcdeb7..926cb97483f9 100644
@@ -4650,16 +4650,14 @@ static int hns_roce_v2_destroy_qp_common(struct hns_roce_dev *hr_dev,
 {
 	struct hns_roce_cq *send_cq, *recv_cq;
 	struct ib_device *ibdev = &hr_dev->ib_dev;
-	int ret;
+	int ret = 0;
 
 	if (hr_qp->ibqp.qp_type == IB_QPT_RC && hr_qp->state != IB_QPS_RESET) {
 		/* Modify qp to reset before destroying qp */
 		ret = hns_roce_v2_modify_qp(&hr_qp->ibqp, NULL, 0,
 					    hr_qp->state, IB_QPS_RESET);
-		if (ret) {
+		if (ret)
 			ibdev_err(ibdev, "modify QP to Reset failed.\n");
-			return ret;
-		}
 	}
 
 	send_cq = to_hr_cq(hr_qp->ibqp.send_cq);
@@ -4715,7 +4713,7 @@ static int hns_roce_v2_destroy_qp_common(struct hns_roce_dev *hr_dev,
 		kfree(hr_qp->rq_inl_buf.wqe_list);
 	}
 
-	return 0;
+	return ret;
 }
 
 static int hns_roce_v2_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
@@ -4725,11 +4723,9 @@ static int hns_roce_v2_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 	int ret;
 
 	ret = hns_roce_v2_destroy_qp_common(hr_dev, hr_qp, udata);
-	if (ret) {
+	if (ret)
 		ibdev_err(&hr_dev->ib_dev, "Destroy qp 0x%06lx failed(%d)\n",
 			  hr_qp->qpn, ret);
-		return ret;
-	}
 
 	if (hr_qp->ibqp.qp_type == IB_QPT_GSI)
 		kfree(hr_to_hr_sqp(hr_qp));
Signature · 1 line
2.20.1
author maintainer
#62 lore raw

PATCH 55/205 ipmi: Fix memory leak in __ipmi_bmc_register

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit 4aa7afb0ee20a97fbf0c5bab3df028d5fb85fdab ]

In the impelementation of __ipmi_bmc_register() the allocated memory for bmc should be released in case ida_simple_get() fails.

Fixes 68e7e50f195f ("ipmi: Don't use BMC product/dev ids in the BMC name")

Message-Id: <[email protected]>

---
 drivers/char/ipmi/ipmi_msghandler.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/char/ipmi/ipmi_msghandler.c +4 -1
index 3c8a559506e8..0b6e7f8d9729 100644
@@ -3039,8 +3039,11 @@ static int __ipmi_bmc_register(struct ipmi_smi *intf,
 		bmc->pdev.name = "ipmi_bmc";
 
 		rv = ida_simple_get(&ipmi_bmc_ida, 0, 0, GFP_KERNEL);
-		if (rv < 0)
+		if (rv < 0) {
+			kfree(bmc);
 			goto out;
+		}
+
 		bmc->pdev.dev.driver = &ipmidriver.driver;
 		bmc->pdev.id = rv;
 		bmc->pdev.dev.release = release_bmc_device;
Signature · 1 line
2.20.1
author maintainer
#63 lore raw

PATCH 56/205 regulator: bd70528: Add MODULE_ALIAS to allow module auto loading

In reply to Sasha Levin

From: Matti Vaittinen <[email protected]>

[ Upstream commit 55d5f62c3fa005a6a8010363d7d1855909ceefbc ]

The bd70528 regulator driver is probed by MFD driver. Add MODULE_ALIAS in order to allow udev to load the module when MFD sub-device cell for regulators is added.

Fixes 99ea37bd1e7d7 ("regulator: bd70528: Support ROHM BD70528 regulator block")
---
 drivers/regulator/bd70528-regulator.c | 1 +
 1 file changed, 1 insertion(+)
drivers/regulator/bd70528-regulator.c +1 -0
index 6041839ec38c..5bf8a2dc5fe7 100644
@@ -285,3 +285,4 @@ module_platform_driver(bd70528_regulator);
 MODULE_AUTHOR("Matti Vaittinen <[email protected]>");
 MODULE_DESCRIPTION("BD70528 voltage regulator driver");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:bd70528-pmic");
Signature · 1 line
2.20.1
author maintainer
#64 lore raw

PATCH 57/205 rtlwifi: Remove unnecessary NULL check in rtl_regd_init

In reply to Sasha Levin

From: Nathan Chancellor <[email protected]>

[ Upstream commit 091c6e9c083f7ebaff00b37ad13562d51464d175 ]

When building with Clang + -Wtautological-pointer-compare:

drivers/net/wireless/realtek/rtlwifi/regd.c:389:33: warning: comparison of address of 'rtlpriv->regd' equal to a null pointer is always false [-Wtautological-pointer-compare]

        if (wiphy == NULL || &rtlpriv->regd == NULL)
                              ~~~~~~~~~^~~~    ~~~~

1 warning generated.

The address of an array member is never NULL unless it is the first struct member so remove the unnecessary check. This was addressed in the staging version of the driver in commit f986978b32b3 ("Staging:
rtlwifi: remove unnecessary NULL check").

While we are here, fix the following checkpatch warning:

CHECK: Comparison to NULL could be written "!wiphy" 35: FILE: drivers/net/wireless/realtek/rtlwifi/regd.c:389:
+ if (wiphy == NULL)

---
 drivers/net/wireless/realtek/rtlwifi/regd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/realtek/rtlwifi/regd.c +1 -1
index c10432cd703e..8be31e0ad878 100644
@@ -386,7 +386,7 @@ int rtl_regd_init(struct ieee80211_hw *hw,
 	struct wiphy *wiphy = hw->wiphy;
 	struct country_code_to_enum_rd *country = NULL;
 
-	if (wiphy == NULL || &rtlpriv->regd == NULL)
+	if (!wiphy)
 		return -EINVAL;
 
 	/* init country_code from efuse channel plan */
Signature · 1 line
2.20.1
author maintainer
#65 lore raw

PATCH 58/205 xprtrdma: Connection becomes unstable after a reconnect

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit a31b2f939219dd9bffdf01a45bd91f209f8cc369 ]

This is because xprt_request_get_cong() is allowing more than one RPC Call to be transmitted before the first Receive on the new connection. The first Receive fills the Receive Queue based on the server's credit grant. Before that Receive, there is only a single Receive WR posted because the client doesn't know the server's credit grant.

Solution is to clear rq_cong on all outstanding rpc_rqsts when the the cwnd is reset. This is because an RPC/RDMA credit is good for one connection instance only.

Fixes 75891f502f5f ("SUNRPC: Support for congestion control ... ")
---
 net/sunrpc/xprtrdma/transport.c |  3 +++
 net/sunrpc/xprtrdma/verbs.c     | 22 ++++++++++++++++++++++
 2 files changed, 25 insertions(+)
net/sunrpc/xprtrdma/transport.c +3 -0
index 160558b4135e..c67d465dc062 100644
@@ -428,8 +428,11 @@ void xprt_rdma_close(struct rpc_xprt *xprt)
 	/* Prepare @xprt for the next connection by reinitializing
 	 * its credit grant to one (see RFC 8166, Section 3.3.3).
 	 */
+	spin_lock(&xprt->transport_lock);
 	r_xprt->rx_buf.rb_credits = 1;
+	xprt->cong = 0;
 	xprt->cwnd = RPC_CWNDSHIFT;
+	spin_unlock(&xprt->transport_lock);
 
 out:
 	xprt->reestablish_timeout = 0;
net/sunrpc/xprtrdma/verbs.c +22 -0
index 3a907537e2cf..f4b136504e96 100644
@@ -75,6 +75,7 @@
  * internal functions
  */
 static void rpcrdma_sendctx_put_locked(struct rpcrdma_sendctx *sc);
+static void rpcrdma_reqs_reset(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_reps_destroy(struct rpcrdma_buffer *buf);
 static void rpcrdma_mrs_create(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf);
@@ -780,6 +781,7 @@ rpcrdma_ep_disconnect(struct rpcrdma_ep *ep, struct rpcrdma_ia *ia)
 	trace_xprtrdma_disconnect(r_xprt, rc);
 
 	rpcrdma_xprt_drain(r_xprt);
+	rpcrdma_reqs_reset(r_xprt);
 }
 
 /* Fixed-size circular FIFO queue. This implementation is wait-free and
@@ -1042,6 +1044,26 @@ struct rpcrdma_req *rpcrdma_req_create(struct rpcrdma_xprt *r_xprt, size_t size,
 	return NULL;
 }
 
+/**
+ * rpcrdma_reqs_reset - Reset all reqs owned by a transport
+ * @r_xprt: controlling transport instance
+ *
+ * ASSUMPTION: the rb_allreqs list is stable for the duration,
+ * and thus can be walked without holding rb_lock. Eg. the
+ * caller is holding the transport send lock to exclude
+ * device removal or disconnection.
+ */
+static void rpcrdma_reqs_reset(struct rpcrdma_xprt *r_xprt)
+{
+	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
+	struct rpcrdma_req *req;
+
+	list_for_each_entry(req, &buf->rb_allreqs, rl_all) {
+		/* Credits are valid only for one connection */
+		req->rl_slot.rq_cong = 0;
+	}
+}
+
 static struct rpcrdma_rep *rpcrdma_rep_create(struct rpcrdma_xprt *r_xprt,
 					      bool temp)
 {
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#66 lore raw

PATCH 59/205 xprtrdma: Fix MR list handling

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit c3700780a096fc66467c81076ddf7f3f11d639b5 ]

Close some holes introduced by commit 6dc6ec9e04c4 ("xprtrdma: Cache free MRs in each rpcrdma_req") that could result in list corruption.

In addition, the result that is tabulated in @count is no longer used, so @count is removed.

Fixes 6dc6ec9e04c4 ("xprtrdma: Cache free MRs in each rpcrdma_req")
---
 net/sunrpc/xprtrdma/verbs.c | 41 +++++++++++++++++++------------------
 1 file changed, 21 insertions(+), 20 deletions(-)
net/sunrpc/xprtrdma/verbs.c +21 -20
index f4b136504e96..be74ff8f7873 100644
@@ -79,7 +79,6 @@ static void rpcrdma_reqs_reset(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_reps_destroy(struct rpcrdma_buffer *buf);
 static void rpcrdma_mrs_create(struct rpcrdma_xprt *r_xprt);
 static void rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf);
-static void rpcrdma_mr_free(struct rpcrdma_mr *mr);
 static struct rpcrdma_regbuf *
 rpcrdma_regbuf_alloc(size_t size, enum dma_data_direction direction,
 		     gfp_t flags);
@@ -967,7 +966,7 @@ rpcrdma_mrs_create(struct rpcrdma_xprt *r_xprt)
 		mr->mr_xprt = r_xprt;
 
 		spin_lock(&buf->rb_lock);
-		list_add(&mr->mr_list, &buf->rb_mrs);
+		rpcrdma_mr_push(mr, &buf->rb_mrs);
 		list_add(&mr->mr_all, &buf->rb_all_mrs);
 		spin_unlock(&buf->rb_lock);
 	}
@@ -1185,10 +1184,19 @@ int rpcrdma_buffer_create(struct rpcrdma_xprt *r_xprt)
  */
 void rpcrdma_req_destroy(struct rpcrdma_req *req)
 {
+	struct rpcrdma_mr *mr;
+
 	list_del(&req->rl_all);
 
-	while (!list_empty(&req->rl_free_mrs))
-		rpcrdma_mr_free(rpcrdma_mr_pop(&req->rl_free_mrs));
+	while ((mr = rpcrdma_mr_pop(&req->rl_free_mrs))) {
+		struct rpcrdma_buffer *buf = &mr->mr_xprt->rx_buf;
+
+		spin_lock(&buf->rb_lock);
+		list_del(&mr->mr_all);
+		spin_unlock(&buf->rb_lock);
+
+		frwr_release_mr(mr);
+	}
 
 	rpcrdma_regbuf_free(req->rl_recvbuf);
 	rpcrdma_regbuf_free(req->rl_sendbuf);
@@ -1196,24 +1204,28 @@ void rpcrdma_req_destroy(struct rpcrdma_req *req)
 	kfree(req);
 }
 
-static void
-rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf)
+/**
+ * rpcrdma_mrs_destroy - Release all of a transport's MRs
+ * @buf: controlling buffer instance
+ *
+ * Relies on caller holding the transport send lock to protect
+ * removing mr->mr_list from req->rl_free_mrs safely.
+ */
+static void rpcrdma_mrs_destroy(struct rpcrdma_buffer *buf)
 {
 	struct rpcrdma_xprt *r_xprt = container_of(buf, struct rpcrdma_xprt,
 						   rx_buf);
 	struct rpcrdma_mr *mr;
-	unsigned int count;
 
-	count = 0;
 	spin_lock(&buf->rb_lock);
 	while ((mr = list_first_entry_or_null(&buf->rb_all_mrs,
 					      struct rpcrdma_mr,
 					      mr_all)) != NULL) {
+		list_del(&mr->mr_list);
 		list_del(&mr->mr_all);
 		spin_unlock(&buf->rb_lock);
 
 		frwr_release_mr(mr);
-		count++;
 		spin_lock(&buf->rb_lock);
 	}
 	spin_unlock(&buf->rb_lock);
@@ -1286,17 +1298,6 @@ void rpcrdma_mr_put(struct rpcrdma_mr *mr)
 	rpcrdma_mr_push(mr, &mr->mr_req->rl_free_mrs);
 }
 
-static void rpcrdma_mr_free(struct rpcrdma_mr *mr)
-{
-	struct rpcrdma_xprt *r_xprt = mr->mr_xprt;
-	struct rpcrdma_buffer *buf = &r_xprt->rx_buf;
-
-	mr->mr_req = NULL;
-	spin_lock(&buf->rb_lock);
-	rpcrdma_mr_push(mr, &buf->rb_mrs);
-	spin_unlock(&buf->rb_lock);
-}
-
 /**
  * rpcrdma_buffer_get - Get a request buffer
  * @buffers: Buffer pool from which to obtain a buffer
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#67 lore raw

PATCH 60/205 drm/tegra: Fix ordering of cleanup code

In reply to Sasha Levin

From: Thierry Reding <[email protected]>

[ Upstream commit 051172e8c1ceef8749f19faacc1d3bef65d20d8d ]

Commit Fixes: b9f8b09ce256 ("drm/tegra: Setup shared IOMMU domain after initialization") changed the initialization order of the IOMMU related bits but didn't update the cleanup path accordingly. This asymmetry can cause failures during error recovery.

Fixes b9f8b09ce256 ("drm/tegra: Setup shared IOMMU domain after initialization")
---
 drivers/gpu/drm/tegra/drm.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
drivers/gpu/drm/tegra/drm.c +7 -7
index 6fb7d74ff553..bc7cc32140f8 100644
@@ -201,19 +201,19 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 	if (tegra->hub)
 		tegra_display_hub_cleanup(tegra->hub);
 device:
-	host1x_device_exit(device);
-fbdev:
-	drm_kms_helper_poll_fini(drm);
-	tegra_drm_fb_free(drm);
-config:
-	drm_mode_config_cleanup(drm);
-
 	if (tegra->domain) {
 		mutex_destroy(&tegra->mm_lock);
 		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
 		iova_cache_put();
 	}
+
+	host1x_device_exit(device);
+fbdev:
+	drm_kms_helper_poll_fini(drm);
+	tegra_drm_fb_free(drm);
+config:
+	drm_mode_config_cleanup(drm);
 domain:
 	if (tegra->domain)
 		iommu_domain_free(tegra->domain);
Signature · 1 line
2.20.1
author maintainer
#68 lore raw

PATCH 61/205 scsi: esas2r: unlock on error in esas2r_nvram_read_direct()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 906ca6353ac09696c1bf0892513c8edffff5e0a6 ]

This error path is missing an unlock.

Fixes 26780d9e12ed ("[SCSI] esas2r: ATTO Technology ExpressSAS 6G SAS/SATA RAID Adapter Driver")
---
 drivers/scsi/esas2r/esas2r_flash.c | 1 +
 1 file changed, 1 insertion(+)
drivers/scsi/esas2r/esas2r_flash.c +1 -0
index 7bd376d95ed5..b02ac389e6c6 100644
@@ -1197,6 +1197,7 @@ bool esas2r_nvram_read_direct(struct esas2r_adapter *a)
 	if (!esas2r_read_flash_block(a, a->nvram, FLS_OFFSET_NVR,
 				     sizeof(struct esas2r_sas_nvram))) {
 		esas2r_hdebug("NVRAM read failed, using defaults");
+		up(&a->nvram_semaphore);
 		return false;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#69 lore raw

PATCH 62/205 scsi: hisi_sas: Don't create debugfs dump folder twice

In reply to Sasha Levin

From: Xiang Chen <[email protected]>

[ Upstream commit 35160421b63d4753a72e9f72ebcdd9d6f88f84b9 ]

Due to a merge error, we attempt to create 2x debugfs dump folders, which fails:

[  861.101914] debugfs: Directory 'dump' with parent '0000:74:02.0'

already present!

This breaks the dump function.

To fix, remove the superfluous attempt to create the folder.

Fixes 7ec7082c57ec ("scsi: hisi_sas: Add hisi_sas_debugfs_alloc() to centralise allocation")
---
 drivers/scsi/hisi_sas/hisi_sas_main.c | 3 ---
 1 file changed, 3 deletions(-)
drivers/scsi/hisi_sas/hisi_sas_main.c +0 -3
index 633effb09c9c..849335d76cf6 100644
@@ -3719,9 +3719,6 @@ static int hisi_sas_debugfs_alloc(struct hisi_hba *hisi_hba)
 	int p, c, d;
 	size_t sz;
 
-	hisi_hba->debugfs_dump_dentry =
-			debugfs_create_dir("dump", hisi_hba->debugfs_dir);
-
 	sz = hw->debugfs_reg_array[DEBUGFS_GLOBAL]->count * 4;
 	hisi_hba->debugfs_regs[DEBUGFS_GLOBAL] =
 				devm_kmalloc(dev, sz, GFP_KERNEL);
Signature · 1 line
2.20.1
author maintainer
#70 lore raw

PATCH 63/205 scsi: hisi_sas: Set the BIST init value before enabling BIST

In reply to Sasha Levin

From: Xiang Chen <[email protected]>

[ Upstream commit 65a3b8bd56942dc988b8c05615bd3f510a10012b ]

If set the BIST init value after enabling BIST, there may be still some few error bits. According to the process, need to set the BIST init value before enabling BIST.

Fixes 97b151e75861 ("scsi: hisi_sas: Add BIST support for phy loopback")
---
 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/scsi/hisi_sas/hisi_sas_v3_hw.c +5 -5
index ef32ee12f606..c4f76d7c29db 100644
@@ -3022,11 +3022,6 @@ static int debugfs_set_bist_v3_hw(struct hisi_hba *hisi_hba, bool enable)
 		hisi_sas_phy_write32(hisi_hba, phy_id,
 				     SAS_PHY_BIST_CTRL, reg_val);
 
-		mdelay(100);
-		reg_val |= (CFG_RX_BIST_EN_MSK | CFG_TX_BIST_EN_MSK);
-		hisi_sas_phy_write32(hisi_hba, phy_id,
-				     SAS_PHY_BIST_CTRL, reg_val);
-
 		/* set the bist init value */
 		hisi_sas_phy_write32(hisi_hba, phy_id,
 				     SAS_PHY_BIST_CODE,
@@ -3035,6 +3030,11 @@ static int debugfs_set_bist_v3_hw(struct hisi_hba *hisi_hba, bool enable)
 				     SAS_PHY_BIST_CODE1,
 				     SAS_PHY_BIST_CODE1_INIT);
 
+		mdelay(100);
+		reg_val |= (CFG_RX_BIST_EN_MSK | CFG_TX_BIST_EN_MSK);
+		hisi_sas_phy_write32(hisi_hba, phy_id,
+				     SAS_PHY_BIST_CTRL, reg_val);
+
 		/* clear error bit */
 		mdelay(100);
 		hisi_sas_phy_read32(hisi_hba, phy_id, SAS_BIST_ERR_CNT);
Signature · 1 line
2.20.1
author maintainer
#71 lore raw

PATCH 64/205 iwlwifi: mvm: consider ieee80211 station max amsdu value

In reply to Sasha Levin

From: Mordechay Goodstein <[email protected]>

[ Upstream commit ee4cce9b9d6421d037ffc002536b918fd7f4aff3 ]

debugfs amsdu_len sets only the max_amsdu_len for ieee80211 station so take it into consideration while getting max amsdu

Fixes af2984e9e625 ("iwlwifi: mvm: add a debugfs entry to set a fixed size AMSDU for all TX packets")
---
 drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c | 8 +++++++-
 drivers/net/wireless/intel/iwlwifi/mvm/tx.c    | 7 ++++++-
 2 files changed, 13 insertions(+), 2 deletions(-)
drivers/net/wireless/intel/iwlwifi/mvm/rs-fw.c +7 -1
index 8f50e2b121bd..098d48153a38 100644
@@ -350,7 +350,13 @@ void iwl_mvm_tlc_update_notif(struct iwl_mvm *mvm,
 		u16 size = le32_to_cpu(notif->amsdu_size);
 		int i;
 
-		if (WARN_ON(sta->max_amsdu_len < size))
+		/*
+		 * In debug sta->max_amsdu_len < size
+		 * so also check with orig_amsdu_len which holds the original
+		 * data before debugfs changed the value
+		 */
+		if (WARN_ON(sta->max_amsdu_len < size &&
+			    mvmsta->orig_amsdu_len < size))
 			goto out;
 
 		mvmsta->amsdu_enabled = le32_to_cpu(notif->amsdu_enabled);
drivers/net/wireless/intel/iwlwifi/mvm/tx.c +6 -1
index 8a059da7a1fa..e3b2a2bf3863 100644
@@ -935,7 +935,12 @@ static int iwl_mvm_tx_tso(struct iwl_mvm *mvm, struct sk_buff *skb,
 	    !(mvmsta->amsdu_enabled & BIT(tid)))
 		return iwl_mvm_tx_tso_segment(skb, 1, netdev_flags, mpdus_skb);
 
-	max_amsdu_len = iwl_mvm_max_amsdu_size(mvm, sta, tid);
+	/*
+	 * Take the min of ieee80211 station and mvm station
+	 */
+	max_amsdu_len =
+		min_t(unsigned int, sta->max_amsdu_len,
+		      iwl_mvm_max_amsdu_size(mvm, sta, tid));
 
 	/*
 	 * Limit A-MSDU in A-MPDU to 4095 bytes when VHT is not
Signature · 1 line
2.20.1
author maintainer
#72 lore raw

PATCH 65/205 phy: ti: gmii-sel: fix mac tx internal delay for rgmii-rxid

In reply to Sasha Levin

From: Grygorii Strashko <[email protected]>

[ Upstream commit 316b429459066215abb50060873ec0832efc4044 ]

Now phy-gmii-sel will disable MAC TX internal delay for PHY interface mode "rgmii-rxid" which is incorrect.
Hence, fix it by enabling MAC TX internal delay in the case of "rgmii-rxid" mode.

Fixes 92b58b34741f ("phy: ti: introduce phy-gmii-sel driver")
---
 drivers/phy/ti/phy-gmii-sel.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/ti/phy-gmii-sel.c +1 -1
index a52c5bb35033..a28bd15297f5 100644
@@ -69,11 +69,11 @@ static int phy_gmii_sel_mode(struct phy *phy, enum phy_mode mode, int submode)
 		break;
 
 	case PHY_INTERFACE_MODE_RGMII:
+	case PHY_INTERFACE_MODE_RGMII_RXID:
 		gmii_sel_mode = AM33XX_GMII_SEL_MODE_RGMII;
 		break;
 
 	case PHY_INTERFACE_MODE_RGMII_ID:
-	case PHY_INTERFACE_MODE_RGMII_RXID:
 	case PHY_INTERFACE_MODE_RGMII_TXID:
 		gmii_sel_mode = AM33XX_GMII_SEL_MODE_RGMII;
 		rgmii_id = 1;
Signature · 1 line
2.20.1
author maintainer
#73 lore raw

PATCH 66/205 drm/amdgpu/vi: silence an uninitialized variable warning

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 4ff17a1df7d550257972a838220a8af4611c8f2c ]

Smatch complains that we need to initialized "*cap" otherwise it can lead to an uninitialized variable bug in the caller. This seems like a reasonable warning and it doesn't hurt to silence it at least.

drivers/gpu/drm/amd/amdgpu/vi.c:767 vi_asic_reset_method() error: uninitialized symbol 'baco_reset'.

Fixes 425db2553e43 ("drm/amdgpu: expose BACO interfaces to upper level from PP")
---
 drivers/gpu/drm/amd/powerplay/amd_powerplay.c | 1 +
 1 file changed, 1 insertion(+)
drivers/gpu/drm/amd/powerplay/amd_powerplay.c +1 -0
index fa8ad7db2b3a..d306cc711997 100644
@@ -1421,6 +1421,7 @@ static int pp_get_asic_baco_capability(void *handle, bool *cap)
 {
 	struct pp_hwmgr *hwmgr = handle;
 
+	*cap = false;
 	if (!hwmgr)
 		return -EINVAL;
 
Signature · 6 lines
2.20.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#74 lore raw

PATCH 67/205 clk: imx7ulp: Correct system clock source option #7

In reply to Sasha Levin

From: Anson Huang <[email protected]>

[ Upstream commit 96ac93a7c4bea4eb4186425795c00937d2dd6085 ]

In the latest reference manual Rev.0,06/2019, the SCS's option #7 is no longer from upll, it is reserved, update clock driver accordingly.

Fixes b1260067ac3d ("clk: imx: add imx7ulp clk driver")
---
 drivers/clk/imx/clk-imx7ulp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/imx/clk-imx7ulp.c +1 -1
index a0f650150367..d2c49fbd0563 100644
@@ -24,7 +24,7 @@ static const char * const spll_pfd_sels[]	= { "spll_pfd0", "spll_pfd1", "spll_pf
 static const char * const spll_sels[]		= { "spll", "spll_pfd_sel", };
 static const char * const apll_pfd_sels[]	= { "apll_pfd0", "apll_pfd1", "apll_pfd2", "apll_pfd3", };
 static const char * const apll_sels[]		= { "apll", "apll_pfd_sel", };
-static const char * const scs_sels[]		= { "dummy", "sosc", "sirc", "firc", "dummy", "apll_sel", "spll_sel", "upll", };
+static const char * const scs_sels[]		= { "dummy", "sosc", "sirc", "firc", "dummy", "apll_sel", "spll_sel", "dummy", };
 static const char * const ddr_sels[]		= { "apll_pfd_sel", "upll", };
 static const char * const nic_sels[]		= { "firc", "ddr_clk", };
 static const char * const periph_plat_sels[]	= { "dummy", "nic1_bus_clk", "nic1_clk", "ddr_clk", "apll_pfd2", "apll_pfd1", "apll_pfd0", "upll", };
Signature · 1 line
2.20.1
author maintainer
#75 lore raw

PATCH 68/205 clk: imx7ulp: Correct DDR clock mux options

In reply to Sasha Levin

From: Anson Huang <[email protected]>

[ Upstream commit 2e2b928a04bd74ea410da72bd60e1c5b06398276 ]

In the latest reference manual Rev.0,06/2019, the DDR clock mux is extended to 2 bits, and the clock options are also changed, correct them accordingly.

Fixes b1260067ac3d ("clk: imx: add imx7ulp clk driver")
---
 drivers/clk/imx/clk-imx7ulp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/clk/imx/clk-imx7ulp.c +2 -2
index d2c49fbd0563..04a3ae979281 100644
@@ -25,7 +25,7 @@ static const char * const spll_sels[]		= { "spll", "spll_pfd_sel", };
 static const char * const apll_pfd_sels[]	= { "apll_pfd0", "apll_pfd1", "apll_pfd2", "apll_pfd3", };
 static const char * const apll_sels[]		= { "apll", "apll_pfd_sel", };
 static const char * const scs_sels[]		= { "dummy", "sosc", "sirc", "firc", "dummy", "apll_sel", "spll_sel", "dummy", };
-static const char * const ddr_sels[]		= { "apll_pfd_sel", "upll", };
+static const char * const ddr_sels[]		= { "apll_pfd_sel", "dummy", "dummy", "dummy", };
 static const char * const nic_sels[]		= { "firc", "ddr_clk", };
 static const char * const periph_plat_sels[]	= { "dummy", "nic1_bus_clk", "nic1_clk", "ddr_clk", "apll_pfd2", "apll_pfd1", "apll_pfd0", "upll", };
 static const char * const periph_bus_sels[]	= { "dummy", "sosc_bus_clk", "mpll", "firc_bus_clk", "rosc", "nic1_bus_clk", "nic1_clk", "spll_bus_clk", };
@@ -119,7 +119,7 @@ static void __init imx7ulp_clk_scg1_init(struct device_node *np)
 	clks[IMX7ULP_CLK_SYS_SEL]	= imx_clk_hw_mux2("scs_sel", base + 0x14, 24, 4, scs_sels, ARRAY_SIZE(scs_sels));
 	clks[IMX7ULP_CLK_HSRUN_SYS_SEL] = imx_clk_hw_mux2("hsrun_scs_sel", base + 0x1c, 24, 4, scs_sels, ARRAY_SIZE(scs_sels));
 	clks[IMX7ULP_CLK_NIC_SEL]	= imx_clk_hw_mux2("nic_sel", base + 0x40, 28, 1, nic_sels, ARRAY_SIZE(nic_sels));
-	clks[IMX7ULP_CLK_DDR_SEL]	= imx_clk_hw_mux_flags("ddr_sel", base + 0x30, 24, 1, ddr_sels, ARRAY_SIZE(ddr_sels), CLK_SET_RATE_PARENT | CLK_OPS_PARENT_ENABLE);
+	clks[IMX7ULP_CLK_DDR_SEL]	= imx_clk_hw_mux_flags("ddr_sel", base + 0x30, 24, 2, ddr_sels, ARRAY_SIZE(ddr_sels), CLK_SET_RATE_PARENT | CLK_OPS_PARENT_ENABLE);
 
 	clks[IMX7ULP_CLK_CORE_DIV]	= imx_clk_hw_divider_flags("divcore",	"scs_sel",  base + 0x14, 16, 4, CLK_SET_RATE_PARENT);
 	clks[IMX7ULP_CLK_HSRUN_CORE_DIV] = imx_clk_hw_divider_flags("hsrun_divcore", "hsrun_scs_sel", base + 0x1c, 16, 4, CLK_SET_RATE_PARENT);
Signature · 1 line
2.20.1
author maintainer
#76 lore raw

PATCH 69/205 arm64: dts: qcom: sdm845-cheza: delete zap-shader

In reply to Sasha Levin

From: Rob Clark <[email protected]>

[ Upstream commit 43b0a4b482478aa4fe7240230be74a79dee95679 ]

This is unused on cheza. Delete the node to get ride of the reserved- memory section, and to avoid the driver from attempting to load a zap shader that doesn't exist every time it powers up the GPU.

This also avoids a massive amount of dmesg spam about missing zap fw:
msm ae00000.mdss: [drm:adreno_request_fw] *ERROR* failed to load
qcom/a630_zap.mdt: -2
adreno 5000000.gpu: [drm:adreno_zap_shader_load] *ERROR* Unable to
load a630_zap.mdt

Cc Douglas Anderson <[email protected]>
Fixes 3fdeaee951aa ("arm64: dts: sdm845: Add zap shader region for GPU")
---
 arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi | 2 ++
 arch/arm64/boot/dts/qcom/sdm845.dtsi       | 2 +-
 2 files changed, 3 insertions(+), 1 deletion(-)
arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi +2 -0
index 34881c0113cb..99a28d64ee62 100644
@@ -165,6 +165,8 @@
 /delete-node/ &venus_mem;
 /delete-node/ &cdsp_mem;
 /delete-node/ &cdsp_pas;
+/delete-node/ &zap_shader;
+/delete-node/ &gpu_mem;
 
 /* Increase the size from 120 MB to 128 MB */
 &mpss_region {
arch/arm64/boot/dts/qcom/sdm845.dtsi +1 -1
index f406a4340b05..2287354fef86 100644
@@ -2824,7 +2824,7 @@
 
 			qcom,gmu = <&gmu>;
 
-			zap-shader {
+			zap_shader: zap-shader {
 				memory-region = <&gpu_mem>;
 			};
 
Signature · 1 line
2.20.1
author maintainer
#77 lore raw

PATCH 70/205 power: supply: bd70528: Add MODULE_ALIAS to allow module auto loading

In reply to Sasha Levin

From: Matti Vaittinen <[email protected]>

[ Upstream commit 9480029fe5c24d482efad38dc631bd555fc7afe2 ]

The bd70528 charger driver is probed by MFD driver. Add MODULE_ALIAS in order to allow udev to load the module when MFD sub-device cell for charger is added.

Fixes f8c7f7ddd8ef0 ("power: supply: Initial support for ROHM BD70528 PMIC charger block")
---
 drivers/power/supply/bd70528-charger.c | 1 +
 1 file changed, 1 insertion(+)
drivers/power/supply/bd70528-charger.c +1 -0
index 1bb32b7226d7..b8e1ec106627 100644
@@ -741,3 +741,4 @@ module_platform_driver(bd70528_power);
 MODULE_AUTHOR("Matti Vaittinen <[email protected]>");
 MODULE_DESCRIPTION("BD70528 power-supply driver");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:bd70528-power");
Signature · 1 line
2.20.1
author maintainer
#78 lore raw

PATCH 71/205 firmware: imx: Remove call to devm_of_platform_populate

In reply to Sasha Levin

From: Daniel Baluta <[email protected]>

[ Upstream commit 0e4e8cc30a2940c57448af1376e40d3c0996fb29 ]

IMX DSP device is created by SOF layer. The current call to devm_of_platform_populate is not needed and it doesn't produce any effects.

Fixes ffbf23d50353915d ("firmware: imx: Add DSP IPC protocol interface)
---
 drivers/firmware/imx/imx-dsp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/imx/imx-dsp.c +1 -1
index a43d2db5cbdb..4265e9dbed84 100644
@@ -114,7 +114,7 @@ static int imx_dsp_probe(struct platform_device *pdev)
 
 	dev_info(dev, "NXP i.MX DSP IPC initialized\n");
 
-	return devm_of_platform_populate(dev);
+	return 0;
 out:
 	kfree(chan_name);
 	for (j = 0; j < i; j++) {
Signature · 1 line
2.20.1
author maintainer
#79 lore raw

PATCH 72/205 ASoC: fsl_esai: Add spin lock to protect reset, stop and start

In reply to Sasha Levin

From: Shengjiu Wang <[email protected]>

[ Upstream commit 35dac627471938eda89fa39ee4ead1f7667e0f57 ]

xrun may happen at the end of stream, the trigger->fsl_esai_trigger_stop maybe called in the middle of fsl_esai_hw_reset, this may cause esai in wrong state after stop, and there may be endless xrun interrupt.

This issue may also happen with trigger->fsl_esai_trigger_start.

So Add spin lock to lock those functions.

---
 sound/soc/fsl/fsl_esai.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
sound/soc/fsl/fsl_esai.c +12 -0
index a78e4ab478df..c7a49d03463a 100644
@@ -33,6 +33,7 @@
  * @fsysclk: system clock source to derive HCK, SCK and FS
  * @spbaclk: SPBA clock (optional, depending on SoC design)
  * @task: tasklet to handle the reset operation
+ * @lock: spin lock between hw_reset() and trigger()
  * @fifo_depth: depth of tx/rx FIFO
  * @slot_width: width of each DAI slot
  * @slots: number of slots
@@ -56,6 +57,7 @@ struct fsl_esai {
 	struct clk *fsysclk;
 	struct clk *spbaclk;
 	struct tasklet_struct task;
+	spinlock_t lock; /* Protect hw_reset and trigger */
 	u32 fifo_depth;
 	u32 slot_width;
 	u32 slots;
@@ -676,8 +678,10 @@ static void fsl_esai_hw_reset(unsigned long arg)
 {
 	struct fsl_esai *esai_priv = (struct fsl_esai *)arg;
 	bool tx = true, rx = false, enabled[2];
+	unsigned long lock_flags;
 	u32 tfcr, rfcr;
 
+	spin_lock_irqsave(&esai_priv->lock, lock_flags);
 	/* Save the registers */
 	regmap_read(esai_priv->regmap, REG_ESAI_TFCR, &tfcr);
 	regmap_read(esai_priv->regmap, REG_ESAI_RFCR, &rfcr);
@@ -715,6 +719,8 @@ static void fsl_esai_hw_reset(unsigned long arg)
 		fsl_esai_trigger_start(esai_priv, tx);
 	if (enabled[rx])
 		fsl_esai_trigger_start(esai_priv, rx);
+
+	spin_unlock_irqrestore(&esai_priv->lock, lock_flags);
 }
 
 static int fsl_esai_trigger(struct snd_pcm_substream *substream, int cmd,
@@ -722,6 +728,7 @@ static int fsl_esai_trigger(struct snd_pcm_substream *substream, int cmd,
 {
 	struct fsl_esai *esai_priv = snd_soc_dai_get_drvdata(dai);
 	bool tx = substream->stream == SNDRV_PCM_STREAM_PLAYBACK;
+	unsigned long lock_flags;
 
 	esai_priv->channels[tx] = substream->runtime->channels;
 
@@ -729,12 +736,16 @@ static int fsl_esai_trigger(struct snd_pcm_substream *substream, int cmd,
 	case SNDRV_PCM_TRIGGER_START:
 	case SNDRV_PCM_TRIGGER_RESUME:
 	case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
+		spin_lock_irqsave(&esai_priv->lock, lock_flags);
 		fsl_esai_trigger_start(esai_priv, tx);
+		spin_unlock_irqrestore(&esai_priv->lock, lock_flags);
 		break;
 	case SNDRV_PCM_TRIGGER_SUSPEND:
 	case SNDRV_PCM_TRIGGER_STOP:
 	case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
+		spin_lock_irqsave(&esai_priv->lock, lock_flags);
 		fsl_esai_trigger_stop(esai_priv, tx);
+		spin_unlock_irqrestore(&esai_priv->lock, lock_flags);
 		break;
 	default:
 		return -EINVAL;
@@ -1002,6 +1013,7 @@ static int fsl_esai_probe(struct platform_device *pdev)
 
 	dev_set_drvdata(&pdev->dev, esai_priv);
 
+	spin_lock_init(&esai_priv->lock);
 	ret = fsl_esai_hw_init(esai_priv);
 	if (ret)
 		return ret;
Signature · 1 line
2.20.1
author maintainer
#80 lore raw

PATCH 73/205 arm64: dts: imx8mm-evk: Assigned clocks for audio plls

In reply to Sasha Levin

From: "S.j. Wang" <[email protected]>

[ Upstream commit e8b395b23643ca26e62a3081130d895e198c6154 ]

Assign clocks and clock-rates for audio plls, that audio drivers can utilize them.

Add dai-tdm-slot-num and dai-tdm-slot-width for sound-wm8524, that sai driver can generate correct bit clock.

Fixes 13f3b9fdef6c ("arm64: dts: imx8mm-evk: Enable audio codec wm8524")
---
 arch/arm64/boot/dts/freescale/imx8mm-evk.dts | 2 ++
 arch/arm64/boot/dts/freescale/imx8mm.dtsi    | 8 ++++++--
 2 files changed, 8 insertions(+), 2 deletions(-)
arch/arm64/boot/dts/freescale/imx8mm-evk.dts +2 -0
index f7a15f3904c2..13137451b438 100644
@@ -62,6 +62,8 @@
 
 		cpudai: simple-audio-card,cpu {
 			sound-dai = <&sai3>;
+			dai-tdm-slot-num = <2>;
+			dai-tdm-slot-width = <32>;
 		};
 
 		simple-audio-card,codec {
arch/arm64/boot/dts/freescale/imx8mm.dtsi +6 -2
index 23c8fad7932b..0435c64c92c8 100644
@@ -479,14 +479,18 @@
 						<&clk IMX8MM_CLK_AUDIO_AHB>,
 						<&clk IMX8MM_CLK_IPG_AUDIO_ROOT>,
 						<&clk IMX8MM_SYS_PLL3>,
-						<&clk IMX8MM_VIDEO_PLL1>;
+						<&clk IMX8MM_VIDEO_PLL1>,
+						<&clk IMX8MM_AUDIO_PLL1>,
+						<&clk IMX8MM_AUDIO_PLL2>;
 				assigned-clock-parents = <&clk IMX8MM_SYS_PLL3_OUT>,
 							 <&clk IMX8MM_SYS_PLL1_800M>;
 				assigned-clock-rates = <0>,
 							<400000000>,
 							<400000000>,
 							<750000000>,
-							<594000000>;
+							<594000000>,
+							<393216000>,
+							<361267200>;
 			};
 
 			src: reset-controller@30390000 {
Signature · 1 line
2.20.1
author maintainer
#81 lore raw

PATCH 74/205 RDMA/hns: Prevent undefined behavior in hns_roce_set_user_sq_size()

In reply to Sasha Levin

From: Jason Gunthorpe <[email protected]>

[ Upstream commit 515f60004ed985d2b2f03659365752e0b6142986 ]

The "ucmd->log_sq_bb_count" variable is a user controlled variable in the 0-255 range. If we shift more than then number of bits in an int then it's undefined behavior (it shift wraps), and potentially the int could become negative.

---
 drivers/infiniband/hw/hns/hns_roce_qp.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/infiniband/hw/hns/hns_roce_qp.c +6 -4
index bd78ff90d998..8dd2d666f687 100644
@@ -332,9 +332,8 @@ static int check_sq_size_with_integrity(struct hns_roce_dev *hr_dev,
 	u8 max_sq_stride = ilog2(roundup_sq_stride);
 
 	/* Sanity check SQ size before proceeding */
-	if ((u32)(1 << ucmd->log_sq_bb_count) > hr_dev->caps.max_wqes ||
-	     ucmd->log_sq_stride > max_sq_stride ||
-	     ucmd->log_sq_stride < HNS_ROCE_IB_MIN_SQ_STRIDE) {
+	if (ucmd->log_sq_stride > max_sq_stride ||
+	    ucmd->log_sq_stride < HNS_ROCE_IB_MIN_SQ_STRIDE) {
 		ibdev_err(&hr_dev->ib_dev, "check SQ size error!\n");
 		return -EINVAL;
 	}
@@ -358,13 +357,16 @@ static int hns_roce_set_user_sq_size(struct hns_roce_dev *hr_dev,
 	u32 max_cnt;
 	int ret;
 
+	if (check_shl_overflow(1, ucmd->log_sq_bb_count, &hr_qp->sq.wqe_cnt) ||
+	    hr_qp->sq.wqe_cnt > hr_dev->caps.max_wqes)
+		return -EINVAL;
+
 	ret = check_sq_size_with_integrity(hr_dev, cap, ucmd);
 	if (ret) {
 		ibdev_err(&hr_dev->ib_dev, "Sanity check sq size failed\n");
 		return ret;
 	}
 
-	hr_qp->sq.wqe_cnt = 1 << ucmd->log_sq_bb_count;
 	hr_qp->sq.wqe_shift = ucmd->log_sq_stride;
 
 	max_cnt = max(1U, cap->max_send_sge);
Signature · 1 line
2.20.1
author maintainer
#82 lore raw

PATCH 75/205 RDMA/hns: Fix to support 64K page for srq

In reply to Sasha Levin

From: Lijun Ou <[email protected]>

[ Upstream commit 5c7e76fb7cb5071be800c938ebf2c475e140d3f0 ]

SRQ's page size configuration of BA and buffer should depend on current PAGE_SHIFT, or it can't work in scenario of 64K page.

Fixes c7bcb13442e1 ("RDMA/hns: Add SRQ support for hip08 kernel mode")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +2 -2
index 926cb97483f9..1e1609addccf 100644
@@ -6088,11 +6088,11 @@ static void hns_roce_v2_write_srqc(struct hns_roce_dev *hr_dev,
 	roce_set_field(srq_context->byte_44_idxbufpgsz_addr,
 		       SRQC_BYTE_44_SRQ_IDX_BA_PG_SZ_M,
 		       SRQC_BYTE_44_SRQ_IDX_BA_PG_SZ_S,
-		       hr_dev->caps.idx_ba_pg_sz);
+		       hr_dev->caps.idx_ba_pg_sz + PG_SHIFT_OFFSET);
 	roce_set_field(srq_context->byte_44_idxbufpgsz_addr,
 		       SRQC_BYTE_44_SRQ_IDX_BUF_PG_SZ_M,
 		       SRQC_BYTE_44_SRQ_IDX_BUF_PG_SZ_S,
-		       hr_dev->caps.idx_buf_pg_sz);
+		       hr_dev->caps.idx_buf_pg_sz + PG_SHIFT_OFFSET);
 
 	srq_context->idx_nxt_blk_addr =
 		cpu_to_le32(mtts_idx[1] >> PAGE_ADDR_SHIFT);
Signature · 1 line
2.20.1
author maintainer
#83 lore raw

PATCH 76/205 RDMA/hns: Bugfix for qpc/cqc timer configuration

In reply to Sasha Levin

From: Yangyang Li <[email protected]>

[ Upstream commit 887803db866a7a4e1817a3cb8a3eee4e9879fed2 ]

qpc/cqc timer entry size needs one page, but currently they are fixedly configured to 4096, which is not appropriate in 64K page scenarios. So they should be modified to PAGE_SIZE.

Fixes 0e40dc2f70cd ("RDMA/hns: Add timer allocation support for hip08")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.h +2 -2
index 43219d2f7de0..76a14db7028d 100644
@@ -87,8 +87,8 @@
 #define HNS_ROCE_V2_MTT_ENTRY_SZ		64
 #define HNS_ROCE_V2_CQE_ENTRY_SIZE		32
 #define HNS_ROCE_V2_SCCC_ENTRY_SZ		32
-#define HNS_ROCE_V2_QPC_TIMER_ENTRY_SZ		4096
-#define HNS_ROCE_V2_CQC_TIMER_ENTRY_SZ		4096
+#define HNS_ROCE_V2_QPC_TIMER_ENTRY_SZ		PAGE_SIZE
+#define HNS_ROCE_V2_CQC_TIMER_ENTRY_SZ		PAGE_SIZE
 #define HNS_ROCE_V2_PAGE_SIZE_SUPPORTED		0xFFFFF000
 #define HNS_ROCE_V2_MAX_INNER_MTPT_NUM		2
 #define HNS_ROCE_INVALID_LKEY			0x100
Signature · 1 line
2.20.1
author maintainer
#84 lore raw

PATCH 77/205 libbpf: Fix compatibility for kernels without need_wakeup

In reply to Sasha Levin

From: Magnus Karlsson <[email protected]>

[ Upstream commit 94ff9ebb49a546b7f009ed840bafa235c96d4c4b ]

When the need_wakeup flag was added to AF_XDP, the format of the XDP_MMAP_OFFSETS getsockopt was extended. Code was added to the kernel to take care of compatibility issues arrising from running applications using any of the two formats. However, libbpf was not extended to take care of the case when the application/libbpf uses the new format but the kernel only supports the old format. This patch adds support in libbpf for parsing the old format, before the need_wakeup flag was added, and emulating a set of static need_wakeup flags that will always work for the application.

v2 -> v3:
* Incorporated code improvements suggested by Jonathan Lemon

v1 -> v2:
* Rebased to bpf-next
* Rewrote the code as the previous version made you blind

Fixes a4500432c2587cb2a ("libbpf: add support for need_wakeup flag in AF_XDP part")
---
 tools/lib/bpf/xsk.c | 83 ++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 71 insertions(+), 12 deletions(-)
tools/lib/bpf/xsk.c +71 -12
index 70f9e10de286..b29d37fba2b0 100644
@@ -73,6 +73,21 @@ struct xsk_nl_info {
 	int fd;
 };
 
+/* Up until and including Linux 5.3 */
+struct xdp_ring_offset_v1 {
+	__u64 producer;
+	__u64 consumer;
+	__u64 desc;
+};
+
+/* Up until and including Linux 5.3 */
+struct xdp_mmap_offsets_v1 {
+	struct xdp_ring_offset_v1 rx;
+	struct xdp_ring_offset_v1 tx;
+	struct xdp_ring_offset_v1 fr;
+	struct xdp_ring_offset_v1 cr;
+};
+
 int xsk_umem__fd(const struct xsk_umem *umem)
 {
 	return umem ? umem->fd : -EINVAL;
@@ -133,6 +148,58 @@ static int xsk_set_xdp_socket_config(struct xsk_socket_config *cfg,
 	return 0;
 }
 
+static void xsk_mmap_offsets_v1(struct xdp_mmap_offsets *off)
+{
+	struct xdp_mmap_offsets_v1 off_v1;
+
+	/* getsockopt on a kernel <= 5.3 has no flags fields.
+	 * Copy over the offsets to the correct places in the >=5.4 format
+	 * and put the flags where they would have been on that kernel.
+	 */
+	memcpy(&off_v1, off, sizeof(off_v1));
+
+	off->rx.producer = off_v1.rx.producer;
+	off->rx.consumer = off_v1.rx.consumer;
+	off->rx.desc = off_v1.rx.desc;
+	off->rx.flags = off_v1.rx.consumer + sizeof(u32);
+
+	off->tx.producer = off_v1.tx.producer;
+	off->tx.consumer = off_v1.tx.consumer;
+	off->tx.desc = off_v1.tx.desc;
+	off->tx.flags = off_v1.tx.consumer + sizeof(u32);
+
+	off->fr.producer = off_v1.fr.producer;
+	off->fr.consumer = off_v1.fr.consumer;
+	off->fr.desc = off_v1.fr.desc;
+	off->fr.flags = off_v1.fr.consumer + sizeof(u32);
+
+	off->cr.producer = off_v1.cr.producer;
+	off->cr.consumer = off_v1.cr.consumer;
+	off->cr.desc = off_v1.cr.desc;
+	off->cr.flags = off_v1.cr.consumer + sizeof(u32);
+}
+
+static int xsk_get_mmap_offsets(int fd, struct xdp_mmap_offsets *off)
+{
+	socklen_t optlen;
+	int err;
+
+	optlen = sizeof(*off);
+	err = getsockopt(fd, SOL_XDP, XDP_MMAP_OFFSETS, off, &optlen);
+	if (err)
+		return err;
+
+	if (optlen == sizeof(*off))
+		return 0;
+
+	if (optlen == sizeof(struct xdp_mmap_offsets_v1)) {
+		xsk_mmap_offsets_v1(off);
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
 int xsk_umem__create_v0_0_4(struct xsk_umem **umem_ptr, void *umem_area,
 			    __u64 size, struct xsk_ring_prod *fill,
 			    struct xsk_ring_cons *comp,
@@ -141,7 +208,6 @@ int xsk_umem__create_v0_0_4(struct xsk_umem **umem_ptr, void *umem_area,
 	struct xdp_mmap_offsets off;
 	struct xdp_umem_reg mr;
 	struct xsk_umem *umem;
-	socklen_t optlen;
 	void *map;
 	int err;
 
@@ -190,8 +256,7 @@ int xsk_umem__create_v0_0_4(struct xsk_umem **umem_ptr, void *umem_area,
 		goto out_socket;
 	}
 
-	optlen = sizeof(off);
-	err = getsockopt(umem->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
+	err = xsk_get_mmap_offsets(umem->fd, &off);
 	if (err) {
 		err = -errno;
 		goto out_socket;
@@ -499,7 +564,6 @@ int xsk_socket__create(struct xsk_socket **xsk_ptr, const char *ifname,
 	struct sockaddr_xdp sxdp = {};
 	struct xdp_mmap_offsets off;
 	struct xsk_socket *xsk;
-	socklen_t optlen;
 	int err;
 
 	if (!umem || !xsk_ptr || !rx || !tx)
@@ -558,8 +622,7 @@ int xsk_socket__create(struct xsk_socket **xsk_ptr, const char *ifname,
 		}
 	}
 
-	optlen = sizeof(off);
-	err = getsockopt(xsk->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
+	err = xsk_get_mmap_offsets(xsk->fd, &off);
 	if (err) {
 		err = -errno;
 		goto out_socket;
@@ -645,7 +708,6 @@ int xsk_socket__create(struct xsk_socket **xsk_ptr, const char *ifname,
 int xsk_umem__delete(struct xsk_umem *umem)
 {
 	struct xdp_mmap_offsets off;
-	socklen_t optlen;
 	int err;
 
 	if (!umem)
@@ -654,8 +716,7 @@ int xsk_umem__delete(struct xsk_umem *umem)
 	if (umem->refcount)
 		return -EBUSY;
 
-	optlen = sizeof(off);
-	err = getsockopt(umem->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
+	err = xsk_get_mmap_offsets(umem->fd, &off);
 	if (!err) {
 		munmap(umem->fill->ring - off.fr.desc,
 		       off.fr.desc + umem->config.fill_size * sizeof(__u64));
@@ -673,7 +734,6 @@ void xsk_socket__delete(struct xsk_socket *xsk)
 {
 	size_t desc_sz = sizeof(struct xdp_desc);
 	struct xdp_mmap_offsets off;
-	socklen_t optlen;
 	int err;
 
 	if (!xsk)
@@ -684,8 +744,7 @@ void xsk_socket__delete(struct xsk_socket *xsk)
 		close(xsk->prog_fd);
 	}
 
-	optlen = sizeof(off);
-	err = getsockopt(xsk->fd, SOL_XDP, XDP_MMAP_OFFSETS, &off, &optlen);
+	err = xsk_get_mmap_offsets(xsk->fd, &off);
 	if (!err) {
 		if (xsk->rx) {
 			munmap(xsk->rx->ring - off.rx.desc,
Signature · 1 line
2.20.1
author maintainer
#85 lore raw

PATCH 78/205 clk: samsung: exynos5420: Preserve CPU clocks configuration during suspend/resume

In reply to Sasha Levin

From: Marian Mihailescu <[email protected]>

[ Upstream commit e21be0d1d7bd7f78a77613f6bcb6965e72b22fc1 ]

Save and restore top PLL related configuration registers for big (APLL) and LITTLE (KPLL) cores during suspend/resume cycle. So far, CPU clocks were reset to default values after suspend/resume cycle and performance after system resume was affected when performance governor has been selected.

Fixes 773424326b51 ("clk: samsung: exynos5420: add more registers to restore list")
---
 drivers/clk/samsung/clk-exynos5420.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/clk/samsung/clk-exynos5420.c +2 -0
index 31466cd1842f..3b7601647d7b 100644
@@ -165,6 +165,8 @@ static const unsigned long exynos5x_clk_regs[] __initconst = {
 	GATE_BUS_CPU,
 	GATE_SCLK_CPU,
 	CLKOUT_CMU_CPU,
+	APLL_CON0,
+	KPLL_CON0,
 	CPLL_CON0,
 	DPLL_CON0,
 	EPLL_CON0,
Signature · 1 line
2.20.1
author maintainer
#86 lore raw

PATCH 79/205 tools: PCI: Fix fd leakage

In reply to Sasha Levin

From: Hewenliang <[email protected]>

[ Upstream commit 3c379a59b4795d7279d38c623e74b9790345a32b ]

We should close fd before the return of run_test.

Fixes 3f2ed8134834 ("tools: PCI: Add a userspace tool to test PCI endpoint")
---
 tools/pci/pcitest.c | 1 +
 1 file changed, 1 insertion(+)
tools/pci/pcitest.c +1 -0
index cb1e51fcc84e..32b7c6f9043d 100644
@@ -129,6 +129,7 @@ static int run_test(struct pci_test *test)
 	}
 
 	fflush(stdout);
+	close(fd);
 	return (ret < 0) ? ret : 1 - ret; /* return 0 if test succeeded */
 }
 
Signature · 1 line
2.20.1
author maintainer
#87 lore raw

PATCH 80/205 libbpf: Don't use kernel-side u32 type in xsk.c

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit a566e35f1e8b4b3be1e96a804d1cca38b578167c ]

u32 is a kernel-side typedef. User-space library is supposed to use __u32. This breaks Github's projection of libbpf. Do u32 -> __u32 fix.

Fixes 94ff9ebb49a5 ("libbpf: Fix compatibility for kernels without need_wakeup")
Cc Magnus Karlsson <[email protected]>
---
 tools/lib/bpf/xsk.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
tools/lib/bpf/xsk.c +4 -4
index b29d37fba2b0..0c7386b0e42e 100644
@@ -161,22 +161,22 @@ static void xsk_mmap_offsets_v1(struct xdp_mmap_offsets *off)
 	off->rx.producer = off_v1.rx.producer;
 	off->rx.consumer = off_v1.rx.consumer;
 	off->rx.desc = off_v1.rx.desc;
-	off->rx.flags = off_v1.rx.consumer + sizeof(u32);
+	off->rx.flags = off_v1.rx.consumer + sizeof(__u32);
 
 	off->tx.producer = off_v1.tx.producer;
 	off->tx.consumer = off_v1.tx.consumer;
 	off->tx.desc = off_v1.tx.desc;
-	off->tx.flags = off_v1.tx.consumer + sizeof(u32);
+	off->tx.flags = off_v1.tx.consumer + sizeof(__u32);
 
 	off->fr.producer = off_v1.fr.producer;
 	off->fr.consumer = off_v1.fr.consumer;
 	off->fr.desc = off_v1.fr.desc;
-	off->fr.flags = off_v1.fr.consumer + sizeof(u32);
+	off->fr.flags = off_v1.fr.consumer + sizeof(__u32);
 
 	off->cr.producer = off_v1.cr.producer;
 	off->cr.consumer = off_v1.cr.consumer;
 	off->cr.desc = off_v1.cr.desc;
-	off->cr.flags = off_v1.cr.consumer + sizeof(u32);
+	off->cr.flags = off_v1.cr.consumer + sizeof(__u32);
 }
 
 static int xsk_get_mmap_offsets(int fd, struct xdp_mmap_offsets *off)
Signature · 1 line
2.20.1
author maintainer
#88 lore raw

PATCH 81/205 RDMA/hns: Fix build error again

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit d5b60e26e86a463ca83bb5ec502dda6ea685159e ]

This is not the first attempt to fix building random configurations, unfortunately the attempt in commit a07fc0bb483e ("RDMA/hns: Fix build error") caused a new problem when CONFIG_INFINIBAND_HNS_HIP06=m and CONFIG_INFINIBAND_HNS_HIP08=y:

drivers/infiniband/hw/hns/hns_roce_main.o:(.rodata+0xe60): undefined reference to `__this_module'

Revert commits a07fc0bb483e ("RDMA/hns: Fix build error") and a3e2d4c7e766 ("RDMA/hns: remove obsolete Kconfig comment") to get back to the previous state, then fix the issues described there differently, by adding more specific dependencies: INFINIBAND_HNS can now only be built-in if at least one of HNS or HNS3 are built-in, and the individual back-ends are only available if that code is reachable from the main driver.

Fixes a07fc0bb483e ("RDMA/hns: Fix build error")
Fixes a3e2d4c7e766 ("RDMA/hns: remove obsolete Kconfig comment")
Fixes dd74282df573 ("RDMA/hns: Initialize the PCI device for hip08 RoCE")
Fixes 08805fdbeb2d ("RDMA/hns: Split hw v1 driver from hns roce driver")
---
 drivers/infiniband/hw/hns/Kconfig  | 17 ++++++++++++++---
 drivers/infiniband/hw/hns/Makefile |  8 ++++++--
 2 files changed, 20 insertions(+), 5 deletions(-)
drivers/infiniband/hw/hns/Kconfig +14 -3
index d602b698b57e..4921c1e40ccd 100644
@@ -1,23 +1,34 @@
 # SPDX-License-Identifier: GPL-2.0-only
 config INFINIBAND_HNS
-	bool "HNS RoCE Driver"
+	tristate "HNS RoCE Driver"
 	depends on NET_VENDOR_HISILICON
 	depends on ARM64 || (COMPILE_TEST && 64BIT)
+	depends on (HNS_DSAF && HNS_ENET) || HNS3
 	---help---
 	  This is a RoCE/RDMA driver for the Hisilicon RoCE engine. The engine
 	  is used in Hisilicon Hip06 and more further ICT SoC based on
 	  platform device.
 
+	  To compile HIP06 or HIP08 driver as module, choose M here.
+
 config INFINIBAND_HNS_HIP06
-	tristate "Hisilicon Hip06 Family RoCE support"
+	bool "Hisilicon Hip06 Family RoCE support"
 	depends on INFINIBAND_HNS && HNS && HNS_DSAF && HNS_ENET
+	depends on INFINIBAND_HNS=m || (HNS_DSAF=y && HNS_ENET=y)
 	---help---
 	  RoCE driver support for Hisilicon RoCE engine in Hisilicon Hip06 and
 	  Hip07 SoC. These RoCE engines are platform devices.
 
+	  To compile this driver, choose Y here: if INFINIBAND_HNS is m, this
+	  module will be called hns-roce-hw-v1
+
 config INFINIBAND_HNS_HIP08
-	tristate "Hisilicon Hip08 Family RoCE support"
+	bool "Hisilicon Hip08 Family RoCE support"
 	depends on INFINIBAND_HNS && PCI && HNS3
+	depends on INFINIBAND_HNS=m || HNS3=y
 	---help---
 	  RoCE driver support for Hisilicon RoCE engine in Hisilicon Hip08 SoC.
 	  The RoCE engine is a PCI device.
+
+	  To compile this driver, choose Y here: if INFINIBAND_HNS is m, this
+	  module will be called hns-roce-hw-v2.
drivers/infiniband/hw/hns/Makefile +6 -2
index 449a2d81319d..e105945b94a1 100644
@@ -9,8 +9,12 @@ hns-roce-objs := hns_roce_main.o hns_roce_cmd.o hns_roce_pd.o \
 	hns_roce_ah.o hns_roce_hem.o hns_roce_mr.o hns_roce_qp.o \
 	hns_roce_cq.o hns_roce_alloc.o hns_roce_db.o hns_roce_srq.o hns_roce_restrack.o
 
+ifdef CONFIG_INFINIBAND_HNS_HIP06
 hns-roce-hw-v1-objs := hns_roce_hw_v1.o $(hns-roce-objs)
-obj-$(CONFIG_INFINIBAND_HNS_HIP06) += hns-roce-hw-v1.o
+obj-$(CONFIG_INFINIBAND_HNS) += hns-roce-hw-v1.o
+endif
 
+ifdef CONFIG_INFINIBAND_HNS_HIP08
 hns-roce-hw-v2-objs := hns_roce_hw_v2.o hns_roce_hw_v2_dfx.o $(hns-roce-objs)
-obj-$(CONFIG_INFINIBAND_HNS_HIP08) += hns-roce-hw-v2.o
+obj-$(CONFIG_INFINIBAND_HNS) += hns-roce-hw-v2.o
+endif
Signature · 1 line
2.20.1
author maintainer
#89 lore raw

PATCH 82/205 rcu: Fix uninitialized variable in nocb_gp_wait()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit b8889c9c89a2655a231dfed93cc9bdca0930ea67 ]

We never set this to false. This probably doesn't affect most people's runtime because GCC will automatically initialize it to false at certain common optimization levels. But that behavior is related to a bug in GCC and obviously should not be relied on.

Fixes 5d6742b37727 ("rcu/nocb: Use rcu_segcblist for no-CBs CPUs")
---
 kernel/rcu/tree_plugin.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/rcu/tree_plugin.h +1 -1
index 2defc7fe74c3..fa08d55f7040 100644
@@ -1946,7 +1946,7 @@ static void nocb_gp_wait(struct rcu_data *my_rdp)
 	int __maybe_unused cpu = my_rdp->cpu;
 	unsigned long cur_gp_seq;
 	unsigned long flags;
-	bool gotcbs;
+	bool gotcbs = false;
 	unsigned long j = jiffies;
 	bool needwait_gp = false; // This prevents actual uninitialized use.
 	bool needwake;
Signature · 1 line
2.20.1
author maintainer
#90 lore raw

PATCH 83/205 SUNRPC: Fix svcauth_gss_proxy_init()

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit 5866efa8cbfbadf3905072798e96652faf02dbe8 ]

gss_read_proxy_verf() assumes things about the XDR buffer containing the RPC Call that are not true for buffers generated by svc_rdma_recv().

RDMA's buffers look more like what the upper layer generates for
sending: head is a kmalloc'd buffer; it does not point to a page whose contents are contiguous with the first page in the buffers' page array. The result is that ACCEPT_SEC_CONTEXT via RPC/RDMA has stopped working on Linux NFS servers that use gssproxy.

This does not affect clients that use only TCP to send their ACCEPT_SEC_CONTEXT operation (that's all Linux clients). Other clients, like Solaris NFS clients, send ACCEPT_SEC_CONTEXT on the same transport as they send all other NFS operations. Such clients can send ACCEPT_SEC_CONTEXT via RPC/RDMA.

I thought I had found every direct reference in the server RPC code to the rqstp->rq_pages field.

Bug found at the 2019 Westford NFS bake-a-thon.

Fixes 3316f0631139 ("svcrdma: Persistently allocate and DMA- ... ")
---
 net/sunrpc/auth_gss/svcauth_gss.c | 84 +++++++++++++++++++++++--------
 1 file changed, 63 insertions(+), 21 deletions(-)
net/sunrpc/auth_gss/svcauth_gss.c +63 -21
index 8be2f209982b..908b60a72d95 100644
@@ -1075,24 +1075,32 @@ gss_read_verf(struct rpc_gss_wire_cred *gc,
 	return 0;
 }
 
-/* Ok this is really heavily depending on a set of semantics in
- * how rqstp is set up by svc_recv and pages laid down by the
- * server when reading a request. We are basically guaranteed that
- * the token lays all down linearly across a set of pages, starting
- * at iov_base in rq_arg.head[0] which happens to be the first of a
- * set of pages stored in rq_pages[].
- * rq_arg.head[0].iov_base will provide us the page_base to pass
- * to the upcall.
- */
-static inline int
-gss_read_proxy_verf(struct svc_rqst *rqstp,
-		    struct rpc_gss_wire_cred *gc, __be32 *authp,
-		    struct xdr_netobj *in_handle,
-		    struct gssp_in_token *in_token)
+static void gss_free_in_token_pages(struct gssp_in_token *in_token)
 {
-	struct kvec *argv = &rqstp->rq_arg.head[0];
 	u32 inlen;
-	int res;
+	int i;
+
+	i = 0;
+	inlen = in_token->page_len;
+	while (inlen) {
+		if (in_token->pages[i])
+			put_page(in_token->pages[i]);
+		inlen -= inlen > PAGE_SIZE ? PAGE_SIZE : inlen;
+	}
+
+	kfree(in_token->pages);
+	in_token->pages = NULL;
+}
+
+static int gss_read_proxy_verf(struct svc_rqst *rqstp,
+			       struct rpc_gss_wire_cred *gc, __be32 *authp,
+			       struct xdr_netobj *in_handle,
+			       struct gssp_in_token *in_token)
+{
+	struct kvec *argv = &rqstp->rq_arg.head[0];
+	unsigned int page_base, length;
+	int pages, i, res;
+	size_t inlen;
 
 	res = gss_read_common_verf(gc, argv, authp, in_handle);
 	if (res)
@@ -1102,10 +1110,36 @@ gss_read_proxy_verf(struct svc_rqst *rqstp,
 	if (inlen > (argv->iov_len + rqstp->rq_arg.page_len))
 		return SVC_DENIED;
 
-	in_token->pages = rqstp->rq_pages;
-	in_token->page_base = (ulong)argv->iov_base & ~PAGE_MASK;
+	pages = DIV_ROUND_UP(inlen, PAGE_SIZE);
+	in_token->pages = kcalloc(pages, sizeof(struct page *), GFP_KERNEL);
+	if (!in_token->pages)
+		return SVC_DENIED;
+	in_token->page_base = 0;
 	in_token->page_len = inlen;
+	for (i = 0; i < pages; i++) {
+		in_token->pages[i] = alloc_page(GFP_KERNEL);
+		if (!in_token->pages[i]) {
+			gss_free_in_token_pages(in_token);
+			return SVC_DENIED;
+		}
+	}
 
+	length = min_t(unsigned int, inlen, argv->iov_len);
+	memcpy(page_address(in_token->pages[0]), argv->iov_base, length);
+	inlen -= length;
+
+	i = 1;
+	page_base = rqstp->rq_arg.page_base;
+	while (inlen) {
+		length = min_t(unsigned int, inlen, PAGE_SIZE);
+		memcpy(page_address(in_token->pages[i]),
+		       page_address(rqstp->rq_arg.pages[i]) + page_base,
+		       length);
+
+		inlen -= length;
+		page_base = 0;
+		i++;
+	}
 	return 0;
 }
 
@@ -1280,8 +1314,11 @@ static int svcauth_gss_proxy_init(struct svc_rqst *rqstp,
 		break;
 	case GSS_S_COMPLETE:
 		status = gss_proxy_save_rsc(sn->rsc_cache, &ud, &handle);
-		if (status)
+		if (status) {
+			pr_info("%s: gss_proxy_save_rsc failed (%d)\n",
+				__func__, status);
 			goto out;
+		}
 		cli_handle.data = (u8 *)&handle;
 		cli_handle.len = sizeof(handle);
 		break;
@@ -1292,15 +1329,20 @@ static int svcauth_gss_proxy_init(struct svc_rqst *rqstp,
 
 	/* Got an answer to the upcall; use it: */
 	if (gss_write_init_verf(sn->rsc_cache, rqstp,
-				&cli_handle, &ud.major_status))
+				&cli_handle, &ud.major_status)) {
+		pr_info("%s: gss_write_init_verf failed\n", __func__);
 		goto out;
+	}
 	if (gss_write_resv(resv, PAGE_SIZE,
 			   &cli_handle, &ud.out_token,
-			   ud.major_status, ud.minor_status))
+			   ud.major_status, ud.minor_status)) {
+		pr_info("%s: gss_write_resv failed\n", __func__);
 		goto out;
+	}
 
 	ret = SVC_COMPLETE;
 out:
+	gss_free_in_token_pages(&ud.in_token);
 	gssp_free_upcall_data(&ud);
 	return ret;
 }
Signature · 1 line
2.20.1
author maintainer
#91 lore raw

PATCH 84/205 rtw88: fix potential read outside array boundary

In reply to Sasha Levin

From: Tzu-En Huang <[email protected]>

[ Upstream commit 18a0696e85fde169e0109aa61d0505b2b935b59d ]

The level of cckpd is from 0 to 4, and it is the index of array pd_lvl[] and cs_lvl[]. However, the length of both arrays are 4, which is smaller than the possible maximum input index. Enumerate cck level to make sure the max level will not be wrong if new level is added in future.

Fixes 479c4ee931a6 ("rtw88: add dynamic cck pd mechanism")
---
 drivers/net/wireless/realtek/rtw88/phy.c      | 17 ++++++++---------
 drivers/net/wireless/realtek/rtw88/phy.h      |  9 +++++++++
 drivers/net/wireless/realtek/rtw88/rtw8822c.c |  4 ++--
 3 files changed, 19 insertions(+), 11 deletions(-)
drivers/net/wireless/realtek/rtw88/phy.c +8 -9
index d3d3f40de75e..47d199d2e7dc 100644
@@ -118,7 +118,7 @@ static void rtw_phy_cck_pd_init(struct rtw_dev *rtwdev)
 
 	for (i = 0; i <= RTW_CHANNEL_WIDTH_40; i++) {
 		for (j = 0; j < RTW_RF_PATH_MAX; j++)
-			dm_info->cck_pd_lv[i][j] = 0;
+			dm_info->cck_pd_lv[i][j] = CCK_PD_LV0;
 	}
 
 	dm_info->cck_fa_avg = CCK_FA_AVG_RESET;
@@ -461,7 +461,6 @@ static void rtw_phy_dpk_track(struct rtw_dev *rtwdev)
 		chip->ops->dpk_track(rtwdev);
 }
 
-#define CCK_PD_LV_MAX		5
 #define CCK_PD_FA_LV1_MIN	1000
 #define CCK_PD_FA_LV0_MAX	500
 
@@ -471,10 +470,10 @@ static u8 rtw_phy_cck_pd_lv_unlink(struct rtw_dev *rtwdev)
 	u32 cck_fa_avg = dm_info->cck_fa_avg;
 
 	if (cck_fa_avg > CCK_PD_FA_LV1_MIN)
-		return 1;
+		return CCK_PD_LV1;
 
 	if (cck_fa_avg < CCK_PD_FA_LV0_MAX)
-		return 0;
+		return CCK_PD_LV0;
 
 	return CCK_PD_LV_MAX;
 }
@@ -494,15 +493,15 @@ static u8 rtw_phy_cck_pd_lv_link(struct rtw_dev *rtwdev)
 	u32 cck_fa_avg = dm_info->cck_fa_avg;
 
 	if (igi > CCK_PD_IGI_LV4_VAL && rssi > CCK_PD_RSSI_LV4_VAL)
-		return 4;
+		return CCK_PD_LV4;
 	if (igi > CCK_PD_IGI_LV3_VAL && rssi > CCK_PD_RSSI_LV3_VAL)
-		return 3;
+		return CCK_PD_LV3;
 	if (igi > CCK_PD_IGI_LV2_VAL || rssi > CCK_PD_RSSI_LV2_VAL)
-		return 2;
+		return CCK_PD_LV2;
 	if (cck_fa_avg > CCK_PD_FA_LV1_MIN)
-		return 1;
+		return CCK_PD_LV1;
 	if (cck_fa_avg < CCK_PD_FA_LV0_MAX)
-		return 0;
+		return CCK_PD_LV0;
 
 	return CCK_PD_LV_MAX;
 }
drivers/net/wireless/realtek/rtw88/phy.h +9 -0
index e79b084628e7..33a5eb9637c0 100644
@@ -125,6 +125,15 @@ rtw_get_tx_power_params(struct rtw_dev *rtwdev, u8 path,
 			u8 rate, u8 bw, u8 ch, u8 regd,
 			struct rtw_power_params *pwr_param);
 
+enum rtw_phy_cck_pd_lv {
+	CCK_PD_LV0,
+	CCK_PD_LV1,
+	CCK_PD_LV2,
+	CCK_PD_LV3,
+	CCK_PD_LV4,
+	CCK_PD_LV_MAX,
+};
+
 #define	MASKBYTE0		0xff
 #define	MASKBYTE1		0xff00
 #define	MASKBYTE2		0xff0000
drivers/net/wireless/realtek/rtw88/rtw8822c.c +2 -2
index c2f6cd76a658..de0505a6a365 100644
@@ -3168,8 +3168,8 @@ rtw8822c_phy_cck_pd_set_reg(struct rtw_dev *rtwdev,
 static void rtw8822c_phy_cck_pd_set(struct rtw_dev *rtwdev, u8 new_lvl)
 {
 	struct rtw_dm_info *dm_info = &rtwdev->dm_info;
-	s8 pd_lvl[4] = {2, 4, 6, 8};
-	s8 cs_lvl[4] = {2, 2, 2, 4};
+	s8 pd_lvl[CCK_PD_LV_MAX] = {0, 2, 4, 6, 8};
+	s8 cs_lvl[CCK_PD_LV_MAX] = {0, 2, 2, 2, 4};
 	u8 cur_lvl;
 	u8 nrx, bw;
 
Signature · 1 line
2.20.1
author maintainer
#92 lore raw

PATCH 85/205 phy: lantiq: vrx200-pcie: fix error return code in ltq_vrx200_pcie_phy_power_on()

In reply to Sasha Levin

From: Wei Yongjun <[email protected]>

[ Upstream commit 82b5d164415549e74cfa1f9156ffd4463d0a76e2 ]

Fix to return a negative error code from the error handling case instead of 0, as done elsewhere in this function.

Fixes e52a632195bf ("phy: lantiq: vrx200-pcie: add a driver for the Lantiq VRX200 PCIe PHY")
---
 drivers/phy/lantiq/phy-lantiq-vrx200-pcie.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/phy/lantiq/phy-lantiq-vrx200-pcie.c +2 -1
index 544d64a84cc0..6e457967653e 100644
@@ -323,7 +323,8 @@ static int ltq_vrx200_pcie_phy_power_on(struct phy *phy)
 		goto err_disable_pdi_clk;
 
 	/* Check if we are in "startup ready" status */
-	if (ltq_vrx200_pcie_phy_wait_for_pll(phy) != 0)
+	ret = ltq_vrx200_pcie_phy_wait_for_pll(phy);
+	if (ret)
 		goto err_disable_phy_clk;
 
 	ltq_vrx200_pcie_phy_apply_workarounds(phy);
Signature · 1 line
2.20.1
author maintainer
#93 lore raw

PATCH 86/205 RDMA/mlx5: Return proper error value

In reply to Sasha Levin

From: Leon Romanovsky <[email protected]>

[ Upstream commit 546d30099ed204792083f043cd7e016de86016a3 ]

Returned value from mlx5_mr_cache_alloc() is checked to be error or real pointer. Return proper error code instead of NULL which is not checked later.

---
 drivers/infiniband/hw/mlx5/mr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mlx5/mr.c +1 -1
index 7019c12005f4..99d563dba91b 100644
@@ -428,7 +428,7 @@ struct mlx5_ib_mr *mlx5_mr_cache_alloc(struct mlx5_ib_dev *dev, int entry)
 
 	if (entry < 0 || entry >= MAX_MR_CACHE_ENTRIES) {
 		mlx5_ib_err(dev, "cache entry %d is out of range\n", entry);
-		return NULL;
+		return ERR_PTR(-EINVAL);
 	}
 
 	ent = &cache->ent[entry];
Signature · 1 line
2.20.1
author maintainer
#94 lore raw

PATCH 87/205 dpaa_eth: perform DMA unmapping before read

In reply to Sasha Levin

From: Madalin Bucur <[email protected]>

[ Upstream commit c70fd3182caef014e6c628b412f81aa57a3ef9e4 ]

DMA unmapping is required before accessing the HW provided timestamping information.

Fixes 4664856e9ca2 ("dpaa_eth: add support for hardware timestamping")
---
 .../net/ethernet/freescale/dpaa/dpaa_eth.c    | 32 ++++++++++---------
 1 file changed, 17 insertions(+), 15 deletions(-)
drivers/net/ethernet/freescale/dpaa/dpaa_eth.c +17 -15
index b4b82b9c5cd6..54ffc9d3b0a9 100644
@@ -1620,18 +1620,6 @@ static struct sk_buff *dpaa_cleanup_tx_fd(const struct dpaa_priv *priv,
 	skbh = (struct sk_buff **)phys_to_virt(addr);
 	skb = *skbh;
 
-	if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) {
-		memset(&shhwtstamps, 0, sizeof(shhwtstamps));
-
-		if (!fman_port_get_tstamp(priv->mac_dev->port[TX], (void *)skbh,
-					  &ns)) {
-			shhwtstamps.hwtstamp = ns_to_ktime(ns);
-			skb_tstamp_tx(skb, &shhwtstamps);
-		} else {
-			dev_warn(dev, "fman_port_get_tstamp failed!\n");
-		}
-	}
-
 	if (unlikely(qm_fd_get_format(fd) == qm_fd_sg)) {
 		nr_frags = skb_shinfo(skb)->nr_frags;
 		dma_unmap_single(dev, addr,
@@ -1654,14 +1642,28 @@ static struct sk_buff *dpaa_cleanup_tx_fd(const struct dpaa_priv *priv,
 			dma_unmap_page(dev, qm_sg_addr(&sgt[i]),
 				       qm_sg_entry_get_len(&sgt[i]), dma_dir);
 		}
-
-		/* Free the page frag that we allocated on Tx */
-		skb_free_frag(phys_to_virt(addr));
 	} else {
 		dma_unmap_single(dev, addr,
 				 skb_tail_pointer(skb) - (u8 *)skbh, dma_dir);
 	}
 
+	/* DMA unmapping is required before accessing the HW provided info */
+	if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) {
+		memset(&shhwtstamps, 0, sizeof(shhwtstamps));
+
+		if (!fman_port_get_tstamp(priv->mac_dev->port[TX], (void *)skbh,
+					  &ns)) {
+			shhwtstamps.hwtstamp = ns_to_ktime(ns);
+			skb_tstamp_tx(skb, &shhwtstamps);
+		} else {
+			dev_warn(dev, "fman_port_get_tstamp failed!\n");
+		}
+	}
+
+	if (qm_fd_get_format(fd) == qm_fd_sg)
+		/* Free the page frag that we allocated on Tx */
+		skb_free_frag(phys_to_virt(addr));
+
 	return skb;
 }
 
Signature · 1 line
2.20.1
author maintainer
#95 lore raw

PATCH 88/205 dpaa_eth: avoid timestamp read on error paths

In reply to Sasha Levin

From: Madalin Bucur <[email protected]>

[ Upstream commit 9a4f4f3a894ff4487f5597b7aabba9432b238292 ]

The dpaa_cleanup_tx_fd() function is called by the frame transmit confirmation callback but also on several error paths. This function is reading the transmit timestamp value. Avoid reading an invalid timestamp value on the error paths.

Fixes 4664856e9ca2 ("dpaa_eth: add support for hardware timestamping")
---
 drivers/net/ethernet/freescale/dpaa/dpaa_eth.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)
drivers/net/ethernet/freescale/dpaa/dpaa_eth.c +10 -7
index 54ffc9d3b0a9..fcbe01f61aa4 100644
@@ -1600,13 +1600,15 @@ static int dpaa_eth_refill_bpools(struct dpaa_priv *priv)
  * Skb freeing is not handled here.
  *
  * This function may be called on error paths in the Tx function, so guard
- * against cases when not all fd relevant fields were filled in.
+ * against cases when not all fd relevant fields were filled in. To avoid
+ * reading the invalid transmission timestamp for the error paths set ts to
+ * false.
  *
  * Return the skb backpointer, since for S/G frames the buffer containing it
  * gets freed here.
  */
 static struct sk_buff *dpaa_cleanup_tx_fd(const struct dpaa_priv *priv,
-					  const struct qm_fd *fd)
+					  const struct qm_fd *fd, bool ts)
 {
 	const enum dma_data_direction dma_dir = DMA_TO_DEVICE;
 	struct device *dev = priv->net_dev->dev.parent;
@@ -1648,7 +1650,8 @@ static struct sk_buff *dpaa_cleanup_tx_fd(const struct dpaa_priv *priv,
 	}
 
 	/* DMA unmapping is required before accessing the HW provided info */
-	if (priv->tx_tstamp && skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) {
+	if (ts && priv->tx_tstamp &&
+	    skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) {
 		memset(&shhwtstamps, 0, sizeof(shhwtstamps));
 
 		if (!fman_port_get_tstamp(priv->mac_dev->port[TX], (void *)skbh,
@@ -2116,7 +2119,7 @@ dpaa_start_xmit(struct sk_buff *skb, struct net_device *net_dev)
 	if (likely(dpaa_xmit(priv, percpu_stats, queue_mapping, &fd) == 0))
 		return NETDEV_TX_OK;
 
-	dpaa_cleanup_tx_fd(priv, &fd);
+	dpaa_cleanup_tx_fd(priv, &fd, false);
 skb_to_fd_failed:
 enomem:
 	percpu_stats->tx_errors++;
@@ -2162,7 +2165,7 @@ static void dpaa_tx_error(struct net_device *net_dev,
 
 	percpu_priv->stats.tx_errors++;
 
-	skb = dpaa_cleanup_tx_fd(priv, fd);
+	skb = dpaa_cleanup_tx_fd(priv, fd, false);
 	dev_kfree_skb(skb);
 }
 
@@ -2202,7 +2205,7 @@ static void dpaa_tx_conf(struct net_device *net_dev,
 
 	percpu_priv->tx_confirm++;
 
-	skb = dpaa_cleanup_tx_fd(priv, fd);
+	skb = dpaa_cleanup_tx_fd(priv, fd, true);
 
 	consume_skb(skb);
 }
@@ -2432,7 +2435,7 @@ static void egress_ern(struct qman_portal *portal,
 	percpu_priv->stats.tx_fifo_errors++;
 	count_ern(percpu_priv, msg);
 
-	skb = dpaa_cleanup_tx_fd(priv, fd);
+	skb = dpaa_cleanup_tx_fd(priv, fd, false);
 	dev_kfree_skb_any(skb);
 }
 
Signature · 1 line
2.20.1
author maintainer
#96 lore raw

PATCH 89/205 scsi: ufs: delete redundant function ufshcd_def_desc_sizes()

In reply to Sasha Levin

From: Bean Huo <[email protected]>

[ Upstream commit 059efd847a4097c67817782d8ff65397e369e69b ]

There is no need to call ufshcd_def_desc_sizes() in ufshcd_init(), since descriptor lengths will be checked and initialized later in ufshcd_init_desc_sizes().

Fixes a4b0e8a4e92b1b(scsi: ufs: Factor out ufshcd_read_desc_param)
Acked-by Avri Altman <avri.altman.wdc.com>
---
 drivers/scsi/ufs/ufshcd.c | 15 +--------------
 1 file changed, 1 insertion(+), 14 deletions(-)
drivers/scsi/ufs/ufshcd.c +1 -14
index 25a6a25b17a2..1e38bb967871 100644
@@ -6779,23 +6779,13 @@ static void ufshcd_init_desc_sizes(struct ufs_hba *hba)
 		&hba->desc_size.geom_desc);
 	if (err)
 		hba->desc_size.geom_desc = QUERY_DESC_GEOMETRY_DEF_SIZE;
+
 	err = ufshcd_read_desc_length(hba, QUERY_DESC_IDN_HEALTH, 0,
 		&hba->desc_size.hlth_desc);
 	if (err)
 		hba->desc_size.hlth_desc = QUERY_DESC_HEALTH_DEF_SIZE;
 }
 
-static void ufshcd_def_desc_sizes(struct ufs_hba *hba)
-{
-	hba->desc_size.dev_desc = QUERY_DESC_DEVICE_DEF_SIZE;
-	hba->desc_size.pwr_desc = QUERY_DESC_POWER_DEF_SIZE;
-	hba->desc_size.interc_desc = QUERY_DESC_INTERCONNECT_DEF_SIZE;
-	hba->desc_size.conf_desc = QUERY_DESC_CONFIGURATION_DEF_SIZE;
-	hba->desc_size.unit_desc = QUERY_DESC_UNIT_DEF_SIZE;
-	hba->desc_size.geom_desc = QUERY_DESC_GEOMETRY_DEF_SIZE;
-	hba->desc_size.hlth_desc = QUERY_DESC_HEALTH_DEF_SIZE;
-}
-
 static struct ufs_ref_clk ufs_ref_clk_freqs[] = {
 	{19200000, REF_CLK_FREQ_19_2_MHZ},
 	{26000000, REF_CLK_FREQ_26_MHZ},
@@ -8283,9 +8273,6 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq)
 	hba->mmio_base = mmio_base;
 	hba->irq = irq;
 
-	/* Set descriptor lengths to specification defaults */
-	ufshcd_def_desc_sizes(hba);
-
 	err = ufshcd_hba_init(hba);
 	if (err)
 		goto out_error;
Signature · 1 line
2.20.1
author maintainer
#97 lore raw

PATCH 90/205 arm64: dts: qcom: msm8998: Disable coresight by default

In reply to Sasha Levin

From: Sai Prakash Ranjan <[email protected]>

[ Upstream commit a636f93fcdb4a516e7cba6a365645ee8429602b2 ]

Boot failure has been reported on MSM8998 based laptop when coresight is enabled. This is most likely due to lack of firmware support for coresight on production device when compared to debug device like MTP where this issue is not observed. So disable coresight by default for MSM8998 and enable it only for MSM8998 MTP.

Reported-and-tested-by Jeffrey Hugo <[email protected]>
Fixes 783abfa2249a ("arm64: dts: qcom: msm8998: Add Coresight support")
---
 arch/arm64/boot/dts/qcom/msm8998-mtp.dtsi | 68 +++++++++++++++++++++++
 arch/arm64/boot/dts/qcom/msm8998.dtsi     | 51 +++++++++++------
 2 files changed, 102 insertions(+), 17 deletions(-)
arch/arm64/boot/dts/qcom/msm8998-mtp.dtsi +68 -0
index 108667ce4f31..8d15572d18e6 100644
@@ -27,6 +27,66 @@
 	status = "okay";
 };
 
+&etf {
+	status = "okay";
+};
+
+&etm1 {
+	status = "okay";
+};
+
+&etm2 {
+	status = "okay";
+};
+
+&etm3 {
+	status = "okay";
+};
+
+&etm4 {
+	status = "okay";
+};
+
+&etm5 {
+	status = "okay";
+};
+
+&etm6 {
+	status = "okay";
+};
+
+&etm7 {
+	status = "okay";
+};
+
+&etm8 {
+	status = "okay";
+};
+
+&etr {
+	status = "okay";
+};
+
+&funnel1 {
+	status = "okay";
+};
+
+&funnel2 {
+	status = "okay";
+};
+
+&funnel3 {
+	status = "okay";
+};
+
+&funnel4 {
+	status = "okay";
+};
+
+&funnel5 {
+	status = "okay";
+};
+
 &pm8005_lsid1 {
 	pm8005-regulators {
 		compatible = "qcom,pm8005-regulators";
@@ -51,6 +111,10 @@
 	vdda-phy-dpdm-supply = <&vreg_l24a_3p075>;
 };
 
+&replicator1 {
+	status = "okay";
+};
+
 &rpm_requests {
 	pm8998-regulators {
 		compatible = "qcom,rpm-pm8998-regulators";
@@ -249,6 +313,10 @@
 	pinctrl-1 = <&sdc2_clk_off &sdc2_cmd_off &sdc2_data_off &sdc2_cd_off>;
 };
 
+&stm {
+	status = "okay";
+};
+
 &ufshc {
 	vcc-supply = <&vreg_l20a_2p95>;
 	vccq-supply = <&vreg_l26a_1p2>;
arch/arm64/boot/dts/qcom/msm8998.dtsi +34 -17

189 lines not on this page · raw message

Signature · 1 line
2.20.1
author maintainer
#98 lore raw

PATCH 91/205 crypto: arm64/aes-neonbs - add return value of skcipher_walk_done() in __xts_crypt()

In reply to Sasha Levin

From: Yunfeng Ye <[email protected]>

[ Upstream commit 9b537997b669c42cec67893538037e8d1c83c91c ]

A warning is found by the static code analysis tool:
"Identical condition 'err', second condition is always false"

Fix this by adding return value of skcipher_walk_done().

Fixes 67cfa5d3b721 ("crypto: arm64/aes-neonbs - implement ciphertext stealing for XTS")
---
 arch/arm64/crypto/aes-neonbs-glue.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/crypto/aes-neonbs-glue.c +1 -1
index ea873b8904c4..e3e27349a9fe 100644
@@ -384,7 +384,7 @@ static int __xts_crypt(struct skcipher_request *req, bool encrypt,
 			goto xts_tail;
 
 		kernel_neon_end();
-		skcipher_walk_done(&walk, nbytes);
+		err = skcipher_walk_done(&walk, nbytes);
 	}
 
 	if (err || likely(!tail))
Signature · 1 line
2.20.1
author maintainer
#99 lore raw

PATCH 92/205 crypto: amcc - restore CRYPTO_AES dependency

In reply to Sasha Levin

From: Christian Lamparter <[email protected]>

[ Upstream commit 298b4c604008025b134bc6fccbc4018449945d60 ]

This patch restores the CRYPTO_AES dependency. This is necessary since some of the crypto4xx driver provided modes need functioning software fallbacks for AES-CTR/CCM and GCM.

Fixes da3e7a9715ea ("crypto: amcc - switch to AES library for GCM key derivation")
Cc Ard Biesheuvel <[email protected]>
---
 drivers/crypto/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/crypto/Kconfig +1 -0
index 8eabf7b20101..7316312935c8 100644
@@ -333,6 +333,7 @@ config CRYPTO_DEV_PPC4XX
 	depends on PPC && 4xx
 	select CRYPTO_HASH
 	select CRYPTO_AEAD
+	select CRYPTO_AES
 	select CRYPTO_LIB_AES
 	select CRYPTO_CCM
 	select CRYPTO_CTR
Signature · 1 line
2.20.1
author maintainer
#100 lore raw

PATCH 93/205 pinctrl: sh-pfc: Do not use platform_get_irq() to count interrupts

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit ad7fe1a1a35994a201497443b5140bf54b074cca ]

As platform_get_irq() now prints an error when the interrupt does not exist, counting interrupts by looping until failure causes the printing of scary messages like:

sh-pfc e6060000.pin-controller: IRQ index 0 not found

Fix this by using the platform_irq_count() helper instead.

---
 drivers/pinctrl/sh-pfc/core.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)
drivers/pinctrl/sh-pfc/core.c +6 -10
index b8640ad41bef..ce983247c9e2 100644
@@ -29,12 +29,12 @@
 static int sh_pfc_map_resources(struct sh_pfc *pfc,
 				struct platform_device *pdev)
 {
-	unsigned int num_windows, num_irqs;
 	struct sh_pfc_window *windows;
 	unsigned int *irqs = NULL;
+	unsigned int num_windows;
 	struct resource *res;
 	unsigned int i;
-	int irq;
+	int num_irqs;
 
 	/* Count the MEM and IRQ resources. */
 	for (num_windows = 0;; num_windows++) {
@@ -42,17 +42,13 @@ static int sh_pfc_map_resources(struct sh_pfc *pfc,
 		if (!res)
 			break;
 	}
-	for (num_irqs = 0;; num_irqs++) {
-		irq = platform_get_irq(pdev, num_irqs);
-		if (irq == -EPROBE_DEFER)
-			return irq;
-		if (irq < 0)
-			break;
-	}
-
 	if (num_windows == 0)
 		return -EINVAL;
 
+	num_irqs = platform_irq_count(pdev);
+	if (num_irqs < 0)
+		return num_irqs;
+
 	/* Allocate memory windows and IRQs arrays. */
 	windows = devm_kcalloc(pfc->dev, num_windows, sizeof(*windows),
 			       GFP_KERNEL);
Signature · 1 line
2.20.1
author maintainer
#101 lore raw

PATCH 94/205 soc: renesas: Add missing check for non-zero product register address

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 4194b583c104922c6141d6610bfbce26847959df ]

If the DTB for a device with an RZ/A2 SoC lacks a device node for the BSID register, the ID validation code falls back to using a register at address 0x0, which leads to undefined behavior (e.g. reading back a random value).

This could be fixed by letting fam_rza2.reg point to the actual BSID register. However, the hardcoded fallbacks were meant for backwards compatibility with old DTBs only, not for new SoCs. Hence fix this by validating renesas_family.reg before using it.

Fixes 175f435f44b724e3 ("soc: renesas: identify RZ/A2")
---
 drivers/soc/renesas/renesas-soc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/soc/renesas/renesas-soc.c +1 -1
index 3299cf5365f3..6651755e9f20 100644
@@ -326,7 +326,7 @@ static int __init renesas_soc_init(void)
 	if (np) {
 		chipid = of_iomap(np, 0);
 		of_node_put(np);
-	} else if (soc->id) {
+	} else if (soc->id && family->reg) {
 		chipid = ioremap(family->reg, 4);
 	}
 	if (chipid) {
Signature · 1 line
2.20.1
author maintainer
#102 lore raw

PATCH 95/205 scsi: core: scsi_trace: Use get_unaligned_be*()

In reply to Sasha Levin

From: Bart Van Assche <[email protected]>

[ Upstream commit b1335f5b0486f61fb66b123b40f8e7a98e49605d ]

This patch fixes an unintended sign extension on left shifts. From Colin
King: "Shifting a u8 left will cause the value to be promoted to an integer. If the top bit of the u8 is set then the following conversion to an u64 will sign extend the value causing the upper 32 bits to be set in the result."

Fix this by using get_unaligned_be*() instead.

Fixes bf8162354233 ("[SCSI] add scsi trace core functions and put trace points")
Cc Christoph Hellwig <[email protected]>
Cc Hannes Reinecke <[email protected]>
Cc Douglas Gilbert <[email protected]>
---
 drivers/scsi/scsi_trace.c | 113 +++++++++++---------------------------
 1 file changed, 33 insertions(+), 80 deletions(-)
drivers/scsi/scsi_trace.c +33 -80
index 07a2425ffa2c..ac35c301c792 100644
@@ -9,7 +9,7 @@
 #include <trace/events/scsi.h>
 
 #define SERVICE_ACTION16(cdb) (cdb[1] & 0x1f)
-#define SERVICE_ACTION32(cdb) ((cdb[8] << 8) | cdb[9])
+#define SERVICE_ACTION32(cdb) (get_unaligned_be16(&cdb[8]))
 
 static const char *
 scsi_trace_misc(struct trace_seq *, unsigned char *, int);
@@ -39,17 +39,12 @@ static const char *
 scsi_trace_rw10(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p);
-	sector_t lba = 0, txlen = 0;
+	u32 lba, txlen;
 
-	lba |= (cdb[2] << 24);
-	lba |= (cdb[3] << 16);
-	lba |= (cdb[4] << 8);
-	lba |=  cdb[5];
-	txlen |= (cdb[7] << 8);
-	txlen |=  cdb[8];
+	lba = get_unaligned_be32(&cdb[2]);
+	txlen = get_unaligned_be16(&cdb[7]);
 
-	trace_seq_printf(p, "lba=%llu txlen=%llu protect=%u",
-			 (unsigned long long)lba, (unsigned long long)txlen,
+	trace_seq_printf(p, "lba=%u txlen=%u protect=%u", lba, txlen,
 			 cdb[1] >> 5);
 
 	if (cdb[0] == WRITE_SAME)
@@ -64,19 +59,12 @@ static const char *
 scsi_trace_rw12(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p);
-	sector_t lba = 0, txlen = 0;
-
-	lba |= (cdb[2] << 24);
-	lba |= (cdb[3] << 16);
-	lba |= (cdb[4] << 8);
-	lba |=  cdb[5];
-	txlen |= (cdb[6] << 24);
-	txlen |= (cdb[7] << 16);
-	txlen |= (cdb[8] << 8);
-	txlen |=  cdb[9];
-
-	trace_seq_printf(p, "lba=%llu txlen=%llu protect=%u",
-			 (unsigned long long)lba, (unsigned long long)txlen,
+	u32 lba, txlen;
+
+	lba = get_unaligned_be32(&cdb[2]);
+	txlen = get_unaligned_be32(&cdb[6]);
+
+	trace_seq_printf(p, "lba=%u txlen=%u protect=%u", lba, txlen,
 			 cdb[1] >> 5);
 	trace_seq_putc(p, 0);
 
@@ -87,23 +75,13 @@ static const char *
 scsi_trace_rw16(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p);
-	sector_t lba = 0, txlen = 0;
-
-	lba |= ((u64)cdb[2] << 56);
-	lba |= ((u64)cdb[3] << 48);
-	lba |= ((u64)cdb[4] << 40);
-	lba |= ((u64)cdb[5] << 32);
-	lba |= (cdb[6] << 24);
-	lba |= (cdb[7] << 16);
-	lba |= (cdb[8] << 8);
-	lba |=  cdb[9];
-	txlen |= (cdb[10] << 24);
-	txlen |= (cdb[11] << 16);
-	txlen |= (cdb[12] << 8);
-	txlen |=  cdb[13];
-
-	trace_seq_printf(p, "lba=%llu txlen=%llu protect=%u",
-			 (unsigned long long)lba, (unsigned long long)txlen,
+	u64 lba;
+	u32 txlen;
+
+	lba = get_unaligned_be64(&cdb[2]);
+	txlen = get_unaligned_be32(&cdb[10]);
+
+	trace_seq_printf(p, "lba=%llu txlen=%u protect=%u", lba, txlen,
 			 cdb[1] >> 5);
 
 	if (cdb[0] == WRITE_SAME_16)
@@ -118,8 +96,8 @@ static const char *
 scsi_trace_rw32(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p), *cmd;
-	sector_t lba = 0, txlen = 0;
-	u32 ei_lbrt = 0;
+	u64 lba;
+	u32 ei_lbrt, txlen;
 
 	switch (SERVICE_ACTION32(cdb)) {
 	case READ_32:
@@ -139,26 +117,12 @@ scsi_trace_rw32(struct trace_seq *p, unsigned char *cdb, int len)
 		goto out;
 	}
 
-	lba |= ((u64)cdb[12] << 56);
-	lba |= ((u64)cdb[13] << 48);
-	lba |= ((u64)cdb[14] << 40);
-	lba |= ((u64)cdb[15] << 32);
-	lba |= (cdb[16] << 24);
-	lba |= (cdb[17] << 16);
-	lba |= (cdb[18] << 8);
-	lba |=  cdb[19];
-	ei_lbrt |= (cdb[20] << 24);
-	ei_lbrt |= (cdb[21] << 16);
-	ei_lbrt |= (cdb[22] << 8);
-	ei_lbrt |=  cdb[23];
-	txlen |= (cdb[28] << 24);
-	txlen |= (cdb[29] << 16);
-	txlen |= (cdb[30] << 8);
-	txlen |=  cdb[31];
-
-	trace_seq_printf(p, "%s_32 lba=%llu txlen=%llu protect=%u ei_lbrt=%u",
-			 cmd, (unsigned long long)lba,
-			 (unsigned long long)txlen, cdb[10] >> 5, ei_lbrt);
+	lba = get_unaligned_be64(&cdb[12]);
+	ei_lbrt = get_unaligned_be32(&cdb[20]);
+	txlen = get_unaligned_be32(&cdb[28]);
+
+	trace_seq_printf(p, "%s_32 lba=%llu txlen=%u protect=%u ei_lbrt=%u",
+			 cmd, lba, txlen, cdb[10] >> 5, ei_lbrt);
 
 	if (SERVICE_ACTION32(cdb) == WRITE_SAME_32)
 		trace_seq_printf(p, " unmap=%u", cdb[10] >> 3 & 1);
@@ -173,7 +137,7 @@ static const char *
 scsi_trace_unmap(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p);
-	unsigned int regions = cdb[7] << 8 | cdb[8];
+	unsigned int regions = get_unaligned_be16(&cdb[7]);
 
 	trace_seq_printf(p, "regions=%u", (regions - 8) / 16);
 	trace_seq_putc(p, 0);
@@ -185,8 +149,8 @@ static const char *
 scsi_trace_service_action_in(struct trace_seq *p, unsigned char *cdb, int len)
 {
 	const char *ret = trace_seq_buffer_ptr(p), *cmd;
-	sector_t lba = 0;
-	u32 alloc_len = 0;
+	u64 lba;
+	u32 alloc_len;
 
 	switch (SERVICE_ACTION16(cdb)) {
 	case SAI_READ_CAPACITY_16:
@@ -200,21 +164,10 @@ scsi_trace_service_action_in(struct trace_seq *p, unsigned char *cdb, int len)
 		goto out;
 	}
 
-	lba |= ((u64)cdb[2] << 56);
-	lba |= ((u64)cdb[3] << 48);
-	lba |= ((u64)cdb[4] << 40);
-	lba |= ((u64)cdb[5] << 32);
-	lba |= (cdb[6] << 24);
-	lba |= (cdb[7] << 16);
-	lba |= (cdb[8] << 8);
-	lba |=  cdb[9];
-	alloc_len |= (cdb[10] << 24);
-	alloc_len |= (cdb[11] << 16);
-	alloc_len |= (cdb[12] << 8);
-	alloc_len |=  cdb[13];
-
-	trace_seq_printf(p, "%s lba=%llu alloc_len=%u", cmd,
-			 (unsigned long long)lba, alloc_len);
+	lba = get_unaligned_be64(&cdb[2]);
+	alloc_len = get_unaligned_be32(&cdb[10]);
+
+	trace_seq_printf(p, "%s lba=%llu alloc_len=%u", cmd, lba, alloc_len);
 
 out:
 	trace_seq_putc(p, 0);
Signature · 1 line
2.20.1
author maintainer
#103 lore raw

PATCH 96/205 net: openvswitch: don't unlock mutex when changing the user_features fails

In reply to Sasha Levin

From: Tonghao Zhang <[email protected]>

[ Upstream commit 4c76bf696a608ea5cc555fe97ec59a9033236604 ]

Unlocking of a not locked mutex is not allowed. Other kernel thread may be in critical section while we unlock it because of setting user_feature fail.

Fixes 95a7233c4 ("net: openvswitch: Set OvS recirc_id from tc chain index")
Cc Paul Blakey <[email protected]>
---
 net/openvswitch/datapath.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/openvswitch/datapath.c +1 -1
index 23f67b8fdeaa..3eed90bfa2bf 100644
@@ -1667,6 +1667,7 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
 				ovs_dp_reset_user_features(skb, info);
 		}
 
+		ovs_unlock();
 		goto err_destroy_meters;
 	}
 
@@ -1683,7 +1684,6 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
 	return 0;
 
 err_destroy_meters:
-	ovs_unlock();
 	ovs_meters_exit(dp);
 err_destroy_ports_array:
 	kfree(dp->ports);
Signature · 1 line
2.20.1
author maintainer
#104 lore raw

PATCH 97/205 ARM: dts: imx6ul-kontron-n6310-s: Disable the snvs-poweroff driver

In reply to Sasha Levin

From: Frieder Schrempf <[email protected]>

[ Upstream commit 0ccafdf3e81bb40fe415ea13e1f42b19c585f0a0 ]

The snvs-poweroff driver can power off the system by pulling the PMIC_ON_REQ signal low, to let the PMIC disable the power. The Kontron SoMs do not have this signal connected, so let's remove the node.

This fixes a real issue when the signal is asserted at poweroff, but not actually causing the power to turn off. It was observed, that in this case the system would not shut down properly.

Fixes 1ea4b76cdfde ("ARM: dts: imx6ul-kontron-n6310: Add Kontron i.MX6UL N6310 SoM and boards")
---
 arch/arm/boot/dts/imx6ul-kontron-n6310-s.dts | 4 ----
 1 file changed, 4 deletions(-)
arch/arm/boot/dts/imx6ul-kontron-n6310-s.dts +0 -4
index 0205fd56d975..4e99e6c79a68 100644
@@ -157,10 +157,6 @@
 	status = "okay";
 };
 
-&snvs_poweroff {
-	status = "okay";
-};
-
 &uart1 {
 	pinctrl-names = "default";
 	pinctrl-0 = <&pinctrl_uart1>;
Signature · 1 line
2.20.1
author maintainer
#105 lore raw

PATCH 98/205 MIPS: Loongson: Fix return value of loongson_hwmon_init

In reply to Sasha Levin

From: Tiezhu Yang <[email protected]>

[ Upstream commit dece3c2a320b0a6d891da6ff774ab763969b6860 ]

When call function hwmon_device_register failed, use the actual return value instead of always -ENOMEM.

Fixes 64f09aa967e1 ("MIPS: Loongson-3: Add CPU Hwmon platform driver")
Cc Ralf Baechle <[email protected]>
Cc James Hogan <[email protected]>
Cc Huacai Chen <[email protected]>
Cc Jiaxun Yang <[email protected]>
---
 drivers/platform/mips/cpu_hwmon.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/platform/mips/cpu_hwmon.c +1 -1
index a7f184bb47e0..3d29a11c1d6b 100644
@@ -161,7 +161,7 @@ static int __init loongson_hwmon_init(void)
 
 	cpu_hwmon_dev = hwmon_device_register(NULL);
 	if (IS_ERR(cpu_hwmon_dev)) {
-		ret = -ENOMEM;
+		ret = PTR_ERR(cpu_hwmon_dev);
 		pr_err("hwmon_device_register fail!\n");
 		goto fail_hwmon_device_register;
 	}
Signature · 1 line
2.20.1
author maintainer
#106 lore raw

PATCH 99/205 powerpc/security: Fix debugfs data leak on 32-bit

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 3b05a1e517e1a8cfda4866ec31d28b2bc4fee4c4 ]

"powerpc_security_features" is "unsigned long", i.e. 32-bit or 64-bit, depending on the platform (PPC_FSL_BOOK3E or PPC_BOOK3S_64). Hence casting its address to "u64 *", and calling debugfs_create_x64() is wrong, and leaks 32-bit of nearby data to userspace on 32-bit platforms.

While all currently defined SEC_FTR_* security feature flags fit in 32-bit, they all have "ULL" suffixes to make them 64-bit constants. Hence fix the leak by changing the type of "powerpc_security_features" (and the parameter types of its accessors) to "u64". This also allows to drop the cast.

Fixes 398af571128fe75f ("powerpc/security: Show powerpc_security_features in debugfs")
---
 arch/powerpc/include/asm/security_features.h | 8 ++++----
 arch/powerpc/kernel/security.c               | 4 ++--
 2 files changed, 6 insertions(+), 6 deletions(-)
arch/powerpc/include/asm/security_features.h +4 -4
index ccf44c135389..7c05e95a5c44 100644
@@ -9,7 +9,7 @@
 #define _ASM_POWERPC_SECURITY_FEATURES_H
 
 
-extern unsigned long powerpc_security_features;
+extern u64 powerpc_security_features;
 extern bool rfi_flush;
 
 /* These are bit flags */
@@ -24,17 +24,17 @@ void setup_stf_barrier(void);
 void do_stf_barrier_fixups(enum stf_barrier_type types);
 void setup_count_cache_flush(void);
 
-static inline void security_ftr_set(unsigned long feature)
+static inline void security_ftr_set(u64 feature)
 {
 	powerpc_security_features |= feature;
 }
 
-static inline void security_ftr_clear(unsigned long feature)
+static inline void security_ftr_clear(u64 feature)
 {
 	powerpc_security_features &= ~feature;
 }
 
-static inline bool security_ftr_enabled(unsigned long feature)
+static inline bool security_ftr_enabled(u64 feature)
 {
 	return !!(powerpc_security_features & feature);
 }
arch/powerpc/kernel/security.c +2 -2
index d341b464f23c..1740a66cea84 100644
@@ -16,7 +16,7 @@
 #include <asm/setup.h>
 
 
-unsigned long powerpc_security_features __read_mostly = SEC_FTR_DEFAULT;
+u64 powerpc_security_features __read_mostly = SEC_FTR_DEFAULT;
 
 enum count_cache_flush_type {
 	COUNT_CACHE_FLUSH_NONE	= 0x1,
@@ -109,7 +109,7 @@ device_initcall(barrier_nospec_debugfs_init);
 static __init int security_feature_debugfs_init(void)
 {
 	debugfs_create_x64("security_features", 0400, powerpc_debugfs_root,
-			   (u64 *)&powerpc_security_features);
+			   &powerpc_security_features);
 	return 0;
 }
 device_initcall(security_feature_debugfs_init);
Signature · 1 line
2.20.1
author maintainer
#107 lore raw

PATCH 100/205 media: exynos4-is: Fix recursive locking in isp_video_release()

In reply to Sasha Levin

From: Seung-Woo Kim <[email protected]>

[ Upstream commit 704c6c80fb471d1bb0ef0d61a94617d1d55743cd ]

Quoted 1 line

From isp_video_release(), &isp->video_lock is held and subsequent

vb2_fop_release() tries to lock vdev->lock which is same with the previous one. Replace vb2_fop_release() with _vb2_fop_release() to fix the recursive locking.

Fixes 1380f5754cb0 ("[media] videobuf2: Add missing lock held on vb2_fop_release")
---
 drivers/media/platform/exynos4-is/fimc-isp-video.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/exynos4-is/fimc-isp-video.c +1 -1
index 378cc302e1f8..d2cbcdca0463 100644
@@ -313,7 +313,7 @@ static int isp_video_release(struct file *file)
 		ivc->streaming = 0;
 	}
 
-	vb2_fop_release(file);
+	_vb2_fop_release(file, NULL);
 
 	if (v4l2_fh_is_singular_file(file)) {
 		fimc_pipeline_call(&ivc->ve, close);
Signature · 1 line
2.20.1
author maintainer
#108 lore raw

PATCH 101/205 media: coda: fix deadlock between decoder picture run and start command

In reply to Sasha Levin

From: Philipp Zabel <[email protected]>

[ Upstream commit a3fd80198de6ab98a205cf7fb148d88e9e1c44bb ]

The BIT decoder picture run temporarily locks the bitstream mutex while the coda device mutex is locked, to refill the bitstream ring buffer. Consequently, the decoder start command, which locks both mutexes when flushing the bitstream ring buffer, must lock the coda device mutex first as well, to avoid an ABBA deadlock.

Fixes e7fd95849b3c ("media: coda: flush bitstream ring buffer on decoder restart")
---
 drivers/media/platform/coda/coda-common.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/platform/coda/coda-common.c +2 -2
index 73222c0615c0..834f11fe9dc2 100644
@@ -1084,16 +1084,16 @@ static int coda_decoder_cmd(struct file *file, void *fh,
 
 	switch (dc->cmd) {
 	case V4L2_DEC_CMD_START:
-		mutex_lock(&ctx->bitstream_mutex);
 		mutex_lock(&dev->coda_mutex);
+		mutex_lock(&ctx->bitstream_mutex);
 		coda_bitstream_flush(ctx);
-		mutex_unlock(&dev->coda_mutex);
 		dst_vq = v4l2_m2m_get_vq(ctx->fh.m2m_ctx,
 					 V4L2_BUF_TYPE_VIDEO_CAPTURE);
 		vb2_clear_last_buffer_dequeued(dst_vq);
 		ctx->bit_stream_param &= ~CODA_BIT_STREAM_END_FLAG;
 		coda_fill_bitstream(ctx, NULL);
 		mutex_unlock(&ctx->bitstream_mutex);
+		mutex_unlock(&dev->coda_mutex);
 		break;
 	case V4L2_DEC_CMD_STOP:
 		stream_end = false;
Signature · 1 line
2.20.1
author maintainer
#109 lore raw

PATCH 102/205 hv_netvsc: flag software created hash value

In reply to Sasha Levin

From: Stephen Hemminger <[email protected]>

[ Upstream commit df9f540ca74297a84bafacfa197e9347b20beea5 ]

When the driver needs to create a hash value because it was not done at higher level, then the hash should be marked as a software not hardware hash.

Fixes f72860afa2e3 ("hv_netvsc: Exclude non-TCP port numbers from vRSS hashing")
---
 drivers/net/hyperv/netvsc_drv.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/net/hyperv/netvsc_drv.c +3 -4
index 78e3e689a733..0dee358864f3 100644
@@ -285,9 +285,9 @@ static inline u32 netvsc_get_hash(
 		else if (flow.basic.n_proto == htons(ETH_P_IPV6))
 			hash = jhash2((u32 *)&flow.addrs.v6addrs, 8, hashrnd);
 		else
-			hash = 0;
+			return 0;
 
-		skb_set_hash(skb, hash, PKT_HASH_TYPE_L3);
+		__skb_set_sw_hash(skb, hash, false);
 	}
 
 	return hash;
@@ -795,8 +795,7 @@ static struct sk_buff *netvsc_alloc_recv_skb(struct net_device *net,
 	    skb->protocol == htons(ETH_P_IP))
 		netvsc_comp_ipcsum(skb);
 
-	/* Do L4 checksum offload if enabled and present.
-	 */
+	/* Do L4 checksum offload if enabled and present. */
 	if (csum_info && (net->features & NETIF_F_RXCSUM)) {
 		if (csum_info->receive.tcp_checksum_succeeded ||
 		    csum_info->receive.udp_checksum_succeeded)
Signature · 1 line
2.20.1
author maintainer
#110 lore raw

PATCH 103/205 arm64: dts: allwinner: a64: Re-add PMU node

In reply to Sasha Levin

From: Andre Przywara <[email protected]>

[ Upstream commit 6b832a148717f1718f57805a9a4aa7f092582d15 ]

As it was found recently, the Performance Monitoring Unit (PMU) on the Allwinner A64 SoC was not generating (the right) interrupts. With the SPI numbers from the manual the kernel did not receive any overflow interrupts, so perf was not happy at all.
It turns out that the numbers were just off by 4, so the PMU interrupts are from 148 to 151, not from 152 to 155 as the manual describes.

This was found by playing around with U-Boot, which typically does not use interrupts, so the GIC is fully available for experimentation:
With *every* PPI and SPI enabled, an overflowing PMU cycle counter was found to set a bit in one of the GICD_ISPENDR registers, with careful counting this was determined to be number 148.

Tested with perf record and perf top on a Pine64-LTS. Also tested with tasksetting to every core to confirm the assignment between IRQs and cores.

This somewhat "revert-fixes" commit ed3e9406bcbc ("arm64: dts: allwinner:
a64: Drop PMU node").

Fixes 34a97fcc71c2 ("arm64: dts: allwinner: a64: Add PMU node")
Fixes ed3e9406bcbc ("arm64: dts: allwinner: a64: Drop PMU node")
---
 arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi | 9 +++++++++
 1 file changed, 9 insertions(+)
arch/arm64/boot/dts/allwinner/sun50i-a64.dtsi +9 -0
index 70f4cce6be43..ba41c1b85887 100644
@@ -142,6 +142,15 @@
 		clock-output-names = "ext-osc32k";
 	};
 
+	pmu {
+		compatible = "arm,cortex-a53-pmu";
+		interrupts = <GIC_SPI 116 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SPI 117 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SPI 118 IRQ_TYPE_LEVEL_HIGH>,
+			     <GIC_SPI 119 IRQ_TYPE_LEVEL_HIGH>;
+		interrupt-affinity = <&cpu0>, <&cpu1>, <&cpu2>, <&cpu3>;
+	};
+
 	psci {
 		compatible = "arm,psci-0.2";
 		method = "smc";
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#111 lore raw

PATCH 104/205 rt2800: remove errornous duplicate condition

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit a1f7c2cabf701a17b1a05d6526bbdadc3d05e05c ]

Quoted 13 lines · wbob

Hello Roman,

while reading around drivers/net/wireless/ralink/rt2x00/rt2800lib.c I stumbled on what I think is an edit of yours made in error in march 2017:

https://github.com/torvalds/linux/commit/41977e86#diff-dae5dc10da180f3b055809a48118e18aR5281

RT6352 in line 5281 should not have been introduced as the "else if" below line 5291 can then not take effect for a RT6352 device. Another possibility is for line 5291 to be not for RT6352, but this seems very unlikely. Are you able to clarify still after this substantial time?

Quoted 8 lines

5277: static int rt2800_init_registers(struct rt2x00_dev *rt2x00dev)
...
5279: } else if (rt2x00_rt(rt2x00dev, RT5390) ||
5280: rt2x00_rt(rt2x00dev, RT5392) ||
5281: rt2x00_rt(rt2x00dev, RT6352)) {
...
5291: } else if (rt2x00_rt(rt2x00dev, RT6352)) {
...

Hence remove errornous line 5281 to make the driver actually execute the correct initialization routine for MT7620 chips.

As it was requested by Stanislaw Gruszka remove setting values of MIMO_PS_CFG and TX_PIN_CFG. MIMO_PS_CFG is responsible for MIMO power-safe mode (which is disabled), hence we can drop setting it. TX_PIN_CFG is set correctly in other functions, and as setting this value breaks some devices, rather don't set it here during init, but only modify it later on.

---
 drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +1 -4
index f1cdcd61c54a..c99f1912e266 100644
@@ -5839,8 +5839,7 @@ static int rt2800_init_registers(struct rt2x00_dev *rt2x00dev)
 		rt2800_register_write(rt2x00dev, TX_TXBF_CFG_0, 0x8000fc21);
 		rt2800_register_write(rt2x00dev, TX_TXBF_CFG_3, 0x00009c40);
 	} else if (rt2x00_rt(rt2x00dev, RT5390) ||
-		   rt2x00_rt(rt2x00dev, RT5392) ||
-		   rt2x00_rt(rt2x00dev, RT6352)) {
+		   rt2x00_rt(rt2x00dev, RT5392)) {
 		rt2800_register_write(rt2x00dev, TX_SW_CFG0, 0x00000404);
 		rt2800_register_write(rt2x00dev, TX_SW_CFG1, 0x00080606);
 		rt2800_register_write(rt2x00dev, TX_SW_CFG2, 0x00000000);
@@ -5854,8 +5853,6 @@ static int rt2800_init_registers(struct rt2x00_dev *rt2x00dev)
 		rt2800_register_write(rt2x00dev, TX_SW_CFG0, 0x00000401);
 		rt2800_register_write(rt2x00dev, TX_SW_CFG1, 0x000C0000);
 		rt2800_register_write(rt2x00dev, TX_SW_CFG2, 0x00000000);
-		rt2800_register_write(rt2x00dev, MIMO_PS_CFG, 0x00000002);
-		rt2800_register_write(rt2x00dev, TX_PIN_CFG, 0x00150F0F);
 		rt2800_register_write(rt2x00dev, TX_ALC_VGA3, 0x00000000);
 		rt2800_register_write(rt2x00dev, TX0_BB_GAIN_ATTEN, 0x0);
 		rt2800_register_write(rt2x00dev, TX1_BB_GAIN_ATTEN, 0x0);
Signature · 1 line
2.20.1
author maintainer
#112 lore raw

PATCH 105/205 hwmon: (pmbus/ibm-cffps) Switch LEDs to blocking brightness call

In reply to Sasha Levin

From: Eddie James <[email protected]>

[ Upstream commit 9861ff954c7e83e2f738ce16fbe15f8a1e121771 ]

Since i2c_smbus functions can sleep, the brightness setting function for this driver must be the blocking version to avoid scheduling while atomic.

Fixes ef9e1cdf419a3 ("hwmon: (pmbus/cffps) Add led class device for power supply fault led")
---
 drivers/hwmon/pmbus/ibm-cffps.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/hwmon/pmbus/ibm-cffps.c +6 -4
index d44745e498e7..aa4cdbbb100a 100644
@@ -292,8 +292,8 @@ static int ibm_cffps_read_word_data(struct i2c_client *client, int page,
 	return rc;
 }
 
-static void ibm_cffps_led_brightness_set(struct led_classdev *led_cdev,
-					 enum led_brightness brightness)
+static int ibm_cffps_led_brightness_set(struct led_classdev *led_cdev,
+					enum led_brightness brightness)
 {
 	int rc;
 	struct ibm_cffps *psu = container_of(led_cdev, struct ibm_cffps, led);
@@ -311,9 +311,11 @@ static void ibm_cffps_led_brightness_set(struct led_classdev *led_cdev,
 	rc = i2c_smbus_write_byte_data(psu->client, CFFPS_SYS_CONFIG_CMD,
 				       psu->led_state);
 	if (rc < 0)
-		return;
+		return rc;
 
 	led_cdev->brightness = brightness;
+
+	return 0;
 }
 
 static int ibm_cffps_led_blink_set(struct led_classdev *led_cdev,
@@ -351,7 +353,7 @@ static void ibm_cffps_create_led_class(struct ibm_cffps *psu)
 		 client->addr);
 	psu->led.name = psu->led_name;
 	psu->led.max_brightness = LED_FULL;
-	psu->led.brightness_set = ibm_cffps_led_brightness_set;
+	psu->led.brightness_set_blocking = ibm_cffps_led_brightness_set;
 	psu->led.blink_set = ibm_cffps_led_blink_set;
 
 	rc = devm_led_classdev_register(dev, &psu->led);
Signature · 1 line
2.20.1
author maintainer
#113 lore raw

PATCH 106/205 hwmon: (pmbus/ibm-cffps) Fix LED blink behavior

In reply to Sasha Levin

From: Eddie James <[email protected]>

[ Upstream commit 92b39ad440968bab38eb6577d63c12994601ed94 ]

The LED blink_set function incorrectly did not tell the PSU LED to blink if brightness was LED_OFF. Fix this, and also correct the LED_OFF command data, which should give control of the LED back to the PSU firmware. Also prevent I2C failures from getting the driver LED state out of sync and add some dev_dbg statements.

Fixes ef9e1cdf419a3 ("hwmon: (pmbus/cffps) Add led class device for power supply fault led")
---
 drivers/hwmon/pmbus/ibm-cffps.c | 27 +++++++++++++++++++--------
 1 file changed, 19 insertions(+), 8 deletions(-)
drivers/hwmon/pmbus/ibm-cffps.c +19 -8
index aa4cdbbb100a..929c909ac27a 100644
@@ -39,9 +39,13 @@
 #define CFFPS_MFR_VAUX_FAULT			BIT(6)
 #define CFFPS_MFR_CURRENT_SHARE_WARNING		BIT(7)
 
+/*
+ * LED off state actually relinquishes LED control to PSU firmware, so it can
+ * turn on the LED for faults.
+ */
+#define CFFPS_LED_OFF				0
 #define CFFPS_LED_BLINK				BIT(0)
 #define CFFPS_LED_ON				BIT(1)
-#define CFFPS_LED_OFF				BIT(2)
 #define CFFPS_BLINK_RATE_MS			250
 
 enum {
@@ -296,23 +300,31 @@ static int ibm_cffps_led_brightness_set(struct led_classdev *led_cdev,
 					enum led_brightness brightness)
 {
 	int rc;
+	u8 next_led_state;
 	struct ibm_cffps *psu = container_of(led_cdev, struct ibm_cffps, led);
 
 	if (brightness == LED_OFF) {
-		psu->led_state = CFFPS_LED_OFF;
+		next_led_state = CFFPS_LED_OFF;
 	} else {
 		brightness = LED_FULL;
+
 		if (psu->led_state != CFFPS_LED_BLINK)
-			psu->led_state = CFFPS_LED_ON;
+			next_led_state = CFFPS_LED_ON;
+		else
+			next_led_state = CFFPS_LED_BLINK;
 	}
 
+	dev_dbg(&psu->client->dev, "LED brightness set: %d. Command: %d.\n",
+		brightness, next_led_state);
+
 	pmbus_set_page(psu->client, 0);
 
 	rc = i2c_smbus_write_byte_data(psu->client, CFFPS_SYS_CONFIG_CMD,
-				       psu->led_state);
+				       next_led_state);
 	if (rc < 0)
 		return rc;
 
+	psu->led_state = next_led_state;
 	led_cdev->brightness = brightness;
 
 	return 0;
@@ -325,10 +337,7 @@ static int ibm_cffps_led_blink_set(struct led_classdev *led_cdev,
 	int rc;
 	struct ibm_cffps *psu = container_of(led_cdev, struct ibm_cffps, led);
 
-	psu->led_state = CFFPS_LED_BLINK;
-
-	if (led_cdev->brightness == LED_OFF)
-		return 0;
+	dev_dbg(&psu->client->dev, "LED blink set.\n");
 
 	pmbus_set_page(psu->client, 0);
 
@@ -337,6 +346,8 @@ static int ibm_cffps_led_blink_set(struct led_classdev *led_cdev,
 	if (rc < 0)
 		return rc;
 
+	psu->led_state = CFFPS_LED_BLINK;
+	led_cdev->brightness = LED_FULL;
 	*delay_on = CFFPS_BLINK_RATE_MS;
 	*delay_off = CFFPS_BLINK_RATE_MS;
 
Signature · 1 line
2.20.1
author maintainer
#114 lore raw

PATCH 107/205 net: neigh: use long type to store jiffies delta

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit 9d027e3a83f39b819e908e4e09084277a2e45e95 ]

A difference of two unsigned long needs long storage.

Fixes c7fb64db001f ("[NETLINK]: Neighbour table configuration and statistics via rtnetlink")
---
 net/core/neighbour.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/core/neighbour.c +2 -2
index f2452496ad9f..920784a9b7ff 100644
@@ -2049,8 +2049,8 @@ static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl,
 		goto nla_put_failure;
 	{
 		unsigned long now = jiffies;
-		unsigned int flush_delta = now - tbl->last_flush;
-		unsigned int rand_delta = now - tbl->last_rand;
+		long flush_delta = now - tbl->last_flush;
+		long rand_delta = now - tbl->last_rand;
 		struct neigh_hash_table *nht;
 		struct ndt_config ndc = {
 			.ndtc_key_len		= tbl->key_len,
Signature · 1 line
2.20.1
author maintainer
#115 lore raw

PATCH 108/205 tipc: reduce sensitive to retransmit failures

In reply to Sasha Levin

From: Hoang Le <[email protected]>

[ Upstream commit 426071f1f3995d7e9603246bffdcbf344cd31719 ]

With huge cluster (e.g >200nodes), the amount of that flow:
gap -> retransmit packet -> acked will take time in case of STATE_MSG dropped/delayed because a lot of traffic. This lead to 1.5 sec tolerance value criteria made link easy failure around 2nd, 3rd of failed retransmission attempts.

Instead of re-introduced criteria of 99 faled retransmissions to fix the issue, we increase failure detection timer to ten times tolerance value.

Fixes 77cf8edbc0e7 ("tipc: simplify stale link failure criteria")
Acked-by Jon
---
 net/tipc/link.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/tipc/link.c +1 -1
index 999eab592de8..a9d8a81e80cf 100644
@@ -1084,7 +1084,7 @@ static bool link_retransmit_failure(struct tipc_link *l, struct tipc_link *r,
 		return false;
 
 	if (!time_after(jiffies, TIPC_SKB_CB(skb)->retr_stamp +
-			msecs_to_jiffies(r->tolerance)))
+			msecs_to_jiffies(r->tolerance * 10)))
 		return false;
 
 	hdr = buf_msg(skb);
Signature · 1 line
2.20.1
author maintainer
#116 lore raw

PATCH 109/205 net: axienet: Fix error return code in axienet_probe()

In reply to Sasha Levin

From: Wei Yongjun <[email protected]>

[ Upstream commit eb34e98baf4ce269423948dacefea6747e963b48 ]

In the DMA memory resource get failed case, the error is not set and 0 will be returned. Fix it by removing redundant check since devm_ioremap_resource() will handle it.

Fixes 28ef9ebdb64c ("net: axienet: make use of axistream-connected attribute optional")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 4 ----
 1 file changed, 4 deletions(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +0 -4
index 676006f32f91..479325eeaf8a 100644
@@ -1790,10 +1790,6 @@ static int axienet_probe(struct platform_device *pdev)
 		/* Check for these resources directly on the Ethernet node. */
 		struct resource *res = platform_get_resource(pdev,
 							     IORESOURCE_MEM, 1);
-		if (!res) {
-			dev_err(&pdev->dev, "unable to get DMA memory resource\n");
-			goto free_netdev;
-		}
 		lp->dma_regs = devm_ioremap_resource(&pdev->dev, res);
 		lp->rx_irq = platform_get_irq(pdev, 1);
 		lp->tx_irq = platform_get_irq(pdev, 0);
Signature · 1 line
2.20.1
author maintainer
#117 lore raw

PATCH 110/205 regulator: ab8500: Remove SYSCLKREQ from enum ab8505_regulator_id

In reply to Sasha Levin

From: Stephan Gerhold <[email protected]>

[ Upstream commit 458ea3ad033fc86e291712ce50cbe60c3428cf30 ]

Those regulators are not actually supported by the AB8500 regulator driver. There is no ab8500_regulator_info for them and no entry in ab8505_regulator_match.

As such, they cannot be registered successfully, and looking them up in ab8505_regulator_match causes an out-of-bounds array read.

Fixes 547f384f33db ("regulator: ab8500: add support for ab8505")
Cc Linus Walleij <[email protected]>
---
 include/linux/regulator/ab8500.h | 2 --
 1 file changed, 2 deletions(-)
include/linux/regulator/ab8500.h +0 -2
index 505e94a6e3e8..3ab1ddf151a2 100644
@@ -42,8 +42,6 @@ enum ab8505_regulator_id {
 	AB8505_LDO_ANAMIC2,
 	AB8505_LDO_AUX8,
 	AB8505_LDO_ANA,
-	AB8505_SYSCLKREQ_2,
-	AB8505_SYSCLKREQ_4,
 	AB8505_NUM_REGULATORS,
 };
 
Signature · 1 line
2.20.1
author maintainer
#118 lore raw

PATCH 111/205 libbpf: Fix memory leak/double free issue

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit 3dc5e059821376974177cc801d377e3fcdac6712 ]

Coverity scan against Github libbpf code found the issue of not freeing memory and leaving already freed memory still referenced from bpf_program. Fix it by re-assigning successfully reallocated memory sooner.

Fixes 2993e0515bb4 ("tools/bpf: add support to read .BTF.ext sections")
---
 tools/lib/bpf/libbpf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/lib/bpf/libbpf.c +1 -1
index a267cd0c0ce2..d98838c5820c 100644
@@ -3220,6 +3220,7 @@ bpf_program__reloc_text(struct bpf_program *prog, struct bpf_object *obj,
 			pr_warning("oom in prog realloc\n");
 			return -ENOMEM;
 		}
+		prog->insns = new_insn;
 
 		if (obj->btf_ext) {
 			err = bpf_program_reloc_btf_ext(prog, obj,
@@ -3231,7 +3232,6 @@ bpf_program__reloc_text(struct bpf_program *prog, struct bpf_object *obj,
 
 		memcpy(new_insn + prog->insns_cnt, text->insns,
 		       text->insns_cnt * sizeof(*insn));
-		prog->insns = new_insn;
 		prog->main_prog_cnt = prog->insns_cnt;
 		prog->insns_cnt = new_cnt;
 		pr_debug("added %zd insn from %s to prog %s\n",
Signature · 1 line
2.20.1
author maintainer
#119 lore raw

PATCH 112/205 libbpf: Fix potential overflow issue

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit 4ee1135615713387b869dfd099ffdf8656be6784 ]

Fix a potential overflow issue found by LGTM analysis, based on Github libbpf source code.

Fixes 3d65014146c6 ("bpf: libbpf: Add btf_line_info support to libbpf")
---
 tools/lib/bpf/bpf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/lib/bpf/bpf.c +1 -1
index cbb933532981..9d0485959308 100644
@@ -189,7 +189,7 @@ static void *
 alloc_zero_tailing_info(const void *orecord, __u32 cnt,
 			__u32 actual_rec_size, __u32 expected_rec_size)
 {
-	__u64 info_len = actual_rec_size * cnt;
+	__u64 info_len = (__u64)actual_rec_size * cnt;
 	void *info, *nrecord;
 	int i;
 
Signature · 1 line
2.20.1
author maintainer
#120 lore raw

PATCH 113/205 libbpf: Fix another potential overflow issue in bpf_prog_linfo

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit dd3ab126379ec040b3edab8559f9c72de6ef9d29 ]

Fix few issues found by Coverity and LGTM.

Fixes b053b439b72a ("bpf: libbpf: bpftool: Print bpf_line_info during prog dump")
---
 tools/lib/bpf/bpf_prog_linfo.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
tools/lib/bpf/bpf_prog_linfo.c +7 -7
index 8c67561c93b0..3ed1a27b5f7c 100644
@@ -101,6 +101,7 @@ struct bpf_prog_linfo *bpf_prog_linfo__new(const struct bpf_prog_info *info)
 {
 	struct bpf_prog_linfo *prog_linfo;
 	__u32 nr_linfo, nr_jited_func;
+	__u64 data_sz;
 
 	nr_linfo = info->nr_line_info;
 
@@ -122,11 +123,11 @@ struct bpf_prog_linfo *bpf_prog_linfo__new(const struct bpf_prog_info *info)
 	/* Copy xlated line_info */
 	prog_linfo->nr_linfo = nr_linfo;
 	prog_linfo->rec_size = info->line_info_rec_size;
-	prog_linfo->raw_linfo = malloc(nr_linfo * prog_linfo->rec_size);
+	data_sz = (__u64)nr_linfo * prog_linfo->rec_size;
+	prog_linfo->raw_linfo = malloc(data_sz);
 	if (!prog_linfo->raw_linfo)
 		goto err_free;
-	memcpy(prog_linfo->raw_linfo, (void *)(long)info->line_info,
-	       nr_linfo * prog_linfo->rec_size);
+	memcpy(prog_linfo->raw_linfo, (void *)(long)info->line_info, data_sz);
 
 	nr_jited_func = info->nr_jited_ksyms;
 	if (!nr_jited_func ||
@@ -142,13 +143,12 @@ struct bpf_prog_linfo *bpf_prog_linfo__new(const struct bpf_prog_info *info)
 	/* Copy jited_line_info */
 	prog_linfo->nr_jited_func = nr_jited_func;
 	prog_linfo->jited_rec_size = info->jited_line_info_rec_size;
-	prog_linfo->raw_jited_linfo = malloc(nr_linfo *
-					     prog_linfo->jited_rec_size);
+	data_sz = (__u64)nr_linfo * prog_linfo->jited_rec_size;
+	prog_linfo->raw_jited_linfo = malloc(data_sz);
 	if (!prog_linfo->raw_jited_linfo)
 		goto err_free;
 	memcpy(prog_linfo->raw_jited_linfo,
-	       (void *)(long)info->jited_line_info,
-	       nr_linfo * prog_linfo->jited_rec_size);
+	       (void *)(long)info->jited_line_info, data_sz);
 
 	/* Number of jited_line_info per jited func */
 	prog_linfo->nr_jited_linfo_per_func = malloc(nr_jited_func *
Signature · 1 line
2.20.1
author maintainer
#121 lore raw

PATCH 114/205 libbpf: Make btf__resolve_size logic always check size error condition

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit 994021a7e08477f7e51285920aac99fc967fae8a ]

Perform size check always in btf__resolve_size. Makes the logic a bit more robust against corrupted BTF and silences LGTM/Coverity complaining about always true (size < 0) check.

Fixes 69eaab04c675 ("btf: extract BTF type size calculation")
---
 tools/lib/bpf/btf.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
tools/lib/bpf/btf.c +1 -2
index 1aa189a9112a..d606a358480d 100644
@@ -269,10 +269,9 @@ __s64 btf__resolve_size(const struct btf *btf, __u32 type_id)
 		t = btf__type_by_id(btf, type_id);
 	}
 
+done:
 	if (size < 0)
 		return -EINVAL;
-
-done:
 	if (nelems && size > UINT32_MAX / nelems)
 		return -E2BIG;
 
Signature · 1 line
2.20.1
author maintainer
#122 lore raw

PATCH 115/205 f2fs: fix potential overflow

In reply to Sasha Levin

From: Chao Yu <[email protected]>

[ Upstream commit 1f0d5c911b64165c9754139a26c8c2fad352c132 ]

We expect 64-bit calculation result from below statement, however in 32-bit machine, looped left shift operation on pgoff_t type variable may cause overflow issue, fix it by forcing type cast.

page->index << PAGE_SHIFT;

Fixes 26de9b117130 ("f2fs: avoid unnecessary updating inode during fsync")
Fixes 0a2aa8fbb969 ("f2fs: refactor __exchange_data_block for speed up")
---
 fs/f2fs/data.c | 2 +-
 fs/f2fs/file.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
fs/f2fs/data.c +1 -1
index 5755e897a5f0..2e9c73165800 100644
@@ -2098,7 +2098,7 @@ static int __write_data_page(struct page *page, bool *submitted,
 	loff_t i_size = i_size_read(inode);
 	const pgoff_t end_index = ((unsigned long long) i_size)
 							>> PAGE_SHIFT;
-	loff_t psize = (page->index + 1) << PAGE_SHIFT;
+	loff_t psize = (loff_t)(page->index + 1) << PAGE_SHIFT;
 	unsigned offset = 0;
 	bool need_balance_fs = false;
 	int err = 0;
fs/f2fs/file.c +1 -1
index 8ed8e4328bd1..fae665691481 100644
@@ -1139,7 +1139,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode,
 				}
 				dn.ofs_in_node++;
 				i++;
-				new_size = (dst + i) << PAGE_SHIFT;
+				new_size = (loff_t)(dst + i) << PAGE_SHIFT;
 				if (dst_inode->i_size < new_size)
 					f2fs_i_size_write(dst_inode, new_size);
 			} while (--ilen && (do_replace[i] || blkaddr[i] == NULL_ADDR));
Signature · 8 lines
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
author maintainer
#123 lore raw

PATCH 116/205 selftests: gen_kselftest_tar.sh: Do not clobber kselftest/

In reply to Sasha Levin

From: Kees Cook <[email protected]>

[ Upstream commit ea1bf0bb18c0bd627d7b551196453ff2fff44225 ]

The default installation location for gen_kselftest_tar.sh was still "kselftest/" which collides with the existing directory. Instead, this moves the installation target into "kselftest_install/kselftest/" and adjusts the tar creation accordingly. This also adjusts indentation and logic to be consistent.

Fixes 42d46e57ec97 ("selftests: Extract single-test shell logic from lib.mk")
---
 tools/testing/selftests/gen_kselftest_tar.sh | 21 ++++++++++-------
 tools/testing/selftests/kselftest_install.sh | 24 ++++++++++----------
 2 files changed, 25 insertions(+), 20 deletions(-)
tools/testing/selftests/gen_kselftest_tar.sh +13 -8
index a27e2eec3586..8b2b6088540d 100755
@@ -38,16 +38,21 @@ main()
 	esac
 	fi
 
-	install_dir=./kselftest
+	# Create working directory.
+	dest=`pwd`
+	install_work="$dest"/kselftest_install
+	install_name=kselftest
+	install_dir="$install_work"/"$install_name"
+	mkdir -p "$install_dir"
 
-# Run install using INSTALL_KSFT_PATH override to generate install
-# directory
-./kselftest_install.sh
-tar $copts kselftest${ext} $install_dir
-echo "Kselftest archive kselftest${ext} created!"
+	# Run install using INSTALL_KSFT_PATH override to generate install
+	# directory
+	./kselftest_install.sh "$install_dir"
+	(cd "$install_work"; tar $copts "$dest"/kselftest${ext} $install_name)
+	echo "Kselftest archive kselftest${ext} created!"
 
-# clean up install directory
-rm -rf kselftest
+	# clean up top-level install work directory
+	rm -rf "$install_work"
 }
 
 main "$@"
tools/testing/selftests/kselftest_install.sh +12 -12
index e2e1911d62d5..407af7da7037 100755
@@ -6,30 +6,30 @@
 # Author: Shuah Khan <[email protected]>
 # Copyright (C) 2015 Samsung Electronics Co., Ltd.
 
-install_loc=`pwd`
-
 main()
 {
-	if [ $(basename $install_loc) !=  "selftests" ]; then
+	base_dir=`pwd`
+	install_dir="$base_dir"/kselftest_install
+
+	# Make sure we're in the selftests top-level directory.
+	if [ $(basename "$base_dir") !=  "selftests" ]; then
 		echo "$0: Please run it in selftests directory ..."
 		exit 1;
 	fi
+
+	# Only allow installation into an existing location.
 	if [ "$#" -eq 0 ]; then
-		echo "$0: Installing in default location - $install_loc ..."
+		echo "$0: Installing in default location - $install_dir ..."
 	elif [ ! -d "$1" ]; then
 		echo "$0: $1 doesn't exist!!"
 		exit 1;
 	else
-		install_loc=$1
-		echo "$0: Installing in specified location - $install_loc ..."
+		install_dir="$1"
+		echo "$0: Installing in specified location - $install_dir ..."
 	fi
 
-	install_dir=$install_loc/kselftest_install
-
-# Create install directory
-	mkdir -p $install_dir
-# Build tests
-	KSFT_INSTALL_PATH=$install_dir make install
+	# Build tests
+	KSFT_INSTALL_PATH="$install_dir" make install
 }
 
 main "$@"
Signature · 1 line
2.20.1
author maintainer
#124 lore raw

PATCH 117/205 ASoC: core: Fix compile warning with CONFIG_DEBUG_FS=n

In reply to Sasha Levin

From: Takashi Iwai <[email protected]>

[ Upstream commit bd0b609e0c3362cb167c51d4bd4330d79fc00987 ]

Paper over a compile warning:
sound/soc/soc-pcm.c:1185:8: warning: unused variable ‘name’

Fixes 0632fa042541 ("ASoC: core: Fix pcm code debugfs error")
---
 sound/soc/soc-pcm.c | 2 ++
 1 file changed, 2 insertions(+)
sound/soc/soc-pcm.c +2 -0
index a6e96cf1d8ff..d07026a846b9 100644
@@ -1148,7 +1148,9 @@ static int dpcm_be_connect(struct snd_soc_pcm_runtime *fe,
 {
 	struct snd_soc_dpcm *dpcm;
 	unsigned long flags;
+#ifdef CONFIG_DEBUG_FS
 	char *name;
+#endif
 
 	/* only add new dpcms */
 	for_each_dpcm_be(fe, stream, dpcm) {
Signature · 6 lines
2.20.1

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author maintainer
#125 lore raw

PATCH 118/205 rtc: brcmstb-waketimer: add missed clk_disable_unprepare

In reply to Sasha Levin

From: Chuhong Yuan <[email protected]>

[ Upstream commit 94303f8930ed78aea0f189b703c9d79fff9555d7 ]

This driver forgets to disable and unprepare clock when remove. Add a call to clk_disable_unprepare to fix it.

Fixes c4f07ecee22e ("rtc: brcmstb-waketimer: Add Broadcom STB wake-timer")
---
 drivers/rtc/rtc-brcmstb-waketimer.c | 1 +
 1 file changed, 1 insertion(+)
drivers/rtc/rtc-brcmstb-waketimer.c +1 -0
index 3e9800f9878a..82d2ab0b3e9c 100644
@@ -277,6 +277,7 @@ static int brcmstb_waketmr_remove(struct platform_device *pdev)
 	struct brcmstb_waketmr *timer = dev_get_drvdata(&pdev->dev);
 
 	unregister_reboot_notifier(&timer->reboot_notifier);
+	clk_disable_unprepare(timer->clk);
 
 	return 0;
 }
Signature · 1 line
2.20.1
author maintainer
#126 lore raw

PATCH 119/205 rtc: bd70528: Add MODULE ALIAS to autoload module

In reply to Sasha Levin

From: Matti Vaittinen <[email protected]>

[ Upstream commit c3e12e66b14a043daac6b3d0559df80b9ed7679c ]

The bd70528 RTC driver is probed by MFD driver. Add MODULE_ALIAS in order to allow udev to load the module when MFD sub-device cell for RTC is added.

I'm not sure if this is a bugfix or feature addition but I guess fixes tag won't harm in this case.

Fixes 32a4a4ebf768 ("rtc: bd70528: Initial support for ROHM bd70528 RTC")
---
 drivers/rtc/rtc-bd70528.c | 1 +
 1 file changed, 1 insertion(+)
drivers/rtc/rtc-bd70528.c +1 -0
index 7744333b0f40..ddfef4d43bab 100644
@@ -491,3 +491,4 @@ module_platform_driver(bd70528_rtc);
 MODULE_AUTHOR("Matti Vaittinen <[email protected]>");
 MODULE_DESCRIPTION("BD70528 RTC driver");
 MODULE_LICENSE("GPL");
+MODULE_ALIAS("platofrm:bd70528-rtc");
Signature · 1 line
2.20.1
author maintainer
#127 lore raw

PATCH 120/205 rtc: bd70528: fix module alias to autoload module

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit afe19a7ae8b6b6032d04d3895ebd5bbac7fe9f30 ]

The module alias platform tag contains a spelling mistake. Fix it.

Fixes f33506abbcdd ("rtc: bd70528: Add MODULE ALIAS to autoload module")
---
 drivers/rtc/rtc-bd70528.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/rtc-bd70528.c +1 -1
index ddfef4d43bab..627037aa66a8 100644
@@ -491,4 +491,4 @@ module_platform_driver(bd70528_rtc);
 MODULE_AUTHOR("Matti Vaittinen <[email protected]>");
 MODULE_DESCRIPTION("BD70528 RTC driver");
 MODULE_LICENSE("GPL");
-MODULE_ALIAS("platofrm:bd70528-rtc");
+MODULE_ALIAS("platform:bd70528-rtc");
Signature · 1 line
2.20.1
author maintainer
#128 lore raw

PATCH 121/205 packet: fix data-race in fanout_flow_is_huge()

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit b756ad928d98e5ef0b74af7546a6a31a8dadde00 ]

KCSAN reported the following data-race [1]

Adding a couple of READ_ONCE()/WRITE_ONCE() should silence it.

Since the report hinted about multiple cpus using the history concurrently, I added a test avoiding writing on it if the victim slot already contains the desired value.

[1]

BUG: KCSAN: data-race in fanout_demux_rollover / fanout_demux_rollover

read to 0xffff8880b01786cc of 4 bytes by task 18921 on cpu 1:

 fanout_flow_is_huge net/packet/af_packet.c:1303 [inline]
 fanout_demux_rollover+0x33e/0x3f0 net/packet/af_packet.c:1353
 packet_rcv_fanout+0x34e/0x490 net/packet/af_packet.c:1453
 deliver_skb net/core/dev.c:1888 [inline]
 dev_queue_xmit_nit+0x15b/0x540 net/core/dev.c:1958
 xmit_one net/core/dev.c:3195 [inline]
 dev_hard_start_xmit+0x3f5/0x430 net/core/dev.c:3215
 __dev_queue_xmit+0x14ab/0x1b40 net/core/dev.c:3792
 dev_queue_xmit+0x21/0x30 net/core/dev.c:3825
 neigh_direct_output+0x1f/0x30 net/core/neighbour.c:1530
 neigh_output include/net/neighbour.h:511 [inline]
 ip6_finish_output2+0x7a2/0xec0 net/ipv6/ip6_output.c:116
 __ip6_finish_output net/ipv6/ip6_output.c:142 [inline]
 __ip6_finish_output+0x2d7/0x330 net/ipv6/ip6_output.c:127
 ip6_finish_output+0x41/0x160 net/ipv6/ip6_output.c:152
 NF_HOOK_COND include/linux/netfilter.h:294 [inline]
 ip6_output+0xf2/0x280 net/ipv6/ip6_output.c:175
 dst_output include/net/dst.h:436 [inline]
 ip6_local_out+0x74/0x90 net/ipv6/output_core.c:179
 ip6_send_skb+0x53/0x110 net/ipv6/ip6_output.c:1795
 udp_v6_send_skb.isra.0+0x3ec/0xa70 net/ipv6/udp.c:1173
 udpv6_sendmsg+0x1906/0x1c20 net/ipv6/udp.c:1471
 inet6_sendmsg+0x6d/0x90 net/ipv6/af_inet6.c:576
 sock_sendmsg_nosec net/socket.c:637 [inline]
 sock_sendmsg+0x9f/0xc0 net/socket.c:657
 ___sys_sendmsg+0x2b7/0x5d0 net/socket.c:2311
 __sys_sendmmsg+0x123/0x350 net/socket.c:2413
 __do_sys_sendmmsg net/socket.c:2442 [inline]
 __se_sys_sendmmsg net/socket.c:2439 [inline]
 __x64_sys_sendmmsg+0x64/0x80 net/socket.c:2439
 do_syscall_64+0xcc/0x370 arch/x86/entry/common.c:290
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

write to 0xffff8880b01786cc of 4 bytes by task 18922 on cpu 0:

 fanout_flow_is_huge net/packet/af_packet.c:1306 [inline]
 fanout_demux_rollover+0x3a4/0x3f0 net/packet/af_packet.c:1353
 packet_rcv_fanout+0x34e/0x490 net/packet/af_packet.c:1453
 deliver_skb net/core/dev.c:1888 [inline]
 dev_queue_xmit_nit+0x15b/0x540 net/core/dev.c:1958
 xmit_one net/core/dev.c:3195 [inline]
 dev_hard_start_xmit+0x3f5/0x430 net/core/dev.c:3215
 __dev_queue_xmit+0x14ab/0x1b40 net/core/dev.c:3792
 dev_queue_xmit+0x21/0x30 net/core/dev.c:3825
 neigh_direct_output+0x1f/0x30 net/core/neighbour.c:1530
 neigh_output include/net/neighbour.h:511 [inline]
 ip6_finish_output2+0x7a2/0xec0 net/ipv6/ip6_output.c:116
 __ip6_finish_output net/ipv6/ip6_output.c:142 [inline]
 __ip6_finish_output+0x2d7/0x330 net/ipv6/ip6_output.c:127
 ip6_finish_output+0x41/0x160 net/ipv6/ip6_output.c:152
 NF_HOOK_COND include/linux/netfilter.h:294 [inline]
 ip6_output+0xf2/0x280 net/ipv6/ip6_output.c:175
 dst_output include/net/dst.h:436 [inline]
 ip6_local_out+0x74/0x90 net/ipv6/output_core.c:179
 ip6_send_skb+0x53/0x110 net/ipv6/ip6_output.c:1795
 udp_v6_send_skb.isra.0+0x3ec/0xa70 net/ipv6/udp.c:1173
 udpv6_sendmsg+0x1906/0x1c20 net/ipv6/udp.c:1471
 inet6_sendmsg+0x6d/0x90 net/ipv6/af_inet6.c:576
 sock_sendmsg_nosec net/socket.c:637 [inline]
 sock_sendmsg+0x9f/0xc0 net/socket.c:657
 ___sys_sendmsg+0x2b7/0x5d0 net/socket.c:2311
 __sys_sendmmsg+0x123/0x350 net/socket.c:2413
 __do_sys_sendmmsg net/socket.c:2442 [inline]
 __se_sys_sendmmsg net/socket.c:2439 [inline]
 __x64_sys_sendmmsg+0x64/0x80 net/socket.c:2439
 do_syscall_64+0xcc/0x370 arch/x86/entry/common.c:290
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Reported by Kernel Concurrency Sanitizer on:

CPU: 0 PID: 18922 Comm: syz-executor.3 Not tainted 5.4.0-rc6+ #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Fixes 3b3a5b0aab5b ("packet: rollover huge flows before small flows")
Cc Willem de Bruijn <[email protected]>
---
 net/packet/af_packet.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
net/packet/af_packet.c +9 -3
index 529d4ce945db..118cd66b7516 100644
@@ -1296,15 +1296,21 @@ static void packet_sock_destruct(struct sock *sk)
 
 static bool fanout_flow_is_huge(struct packet_sock *po, struct sk_buff *skb)
 {
-	u32 rxhash;
+	u32 *history = po->rollover->history;
+	u32 victim, rxhash;
 	int i, count = 0;
 
 	rxhash = skb_get_hash(skb);
 	for (i = 0; i < ROLLOVER_HLEN; i++)
-		if (po->rollover->history[i] == rxhash)
+		if (READ_ONCE(history[i]) == rxhash)
 			count++;
 
-	po->rollover->history[prandom_u32() % ROLLOVER_HLEN] = rxhash;
+	victim = prandom_u32() % ROLLOVER_HLEN;
+
+	/* Avoid dirtying the cache line if possible */
+	if (READ_ONCE(history[victim]) != rxhash)
+		WRITE_ONCE(history[victim], rxhash);
+
 	return count > (ROLLOVER_HLEN >> 1);
 }
 
Signature · 1 line
2.20.1
author maintainer
#129 lore raw

PATCH 122/205 scsi: lpfc: Fix a kernel warning triggered by lpfc_get_sgl_per_hdwq()

In reply to Sasha Levin

From: Bart Van Assche <[email protected]>

[ Upstream commit 765ab6cdac3b681952da0e22184bf6cf1ae41cf8 ]

Fix the following kernel bug report:

BUG: using smp_processor_id() in preemptible [00000000] code: systemd-udevd/954
Fixes d79c9e9d4b3d ("scsi: lpfc: Support dynamic unbounded SGL lists on G7 hardware.")
---
 drivers/scsi/lpfc/lpfc_sli.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/scsi/lpfc/lpfc_sli.c +2 -2
index 9771638b64ba..2d75be07cd6e 100644
@@ -20430,7 +20430,7 @@ lpfc_get_sgl_per_hdwq(struct lpfc_hba *phba, struct lpfc_io_buf *lpfc_buf)
 		/* allocate more */
 		spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 		tmp = kmalloc_node(sizeof(*tmp), GFP_ATOMIC,
-				   cpu_to_node(smp_processor_id()));
+				   cpu_to_node(raw_smp_processor_id()));
 		if (!tmp) {
 			lpfc_printf_log(phba, KERN_INFO, LOG_SLI,
 					"8353 error kmalloc memory for HDWQ "
@@ -20573,7 +20573,7 @@ lpfc_get_cmd_rsp_buf_per_hdwq(struct lpfc_hba *phba,
 		/* allocate more */
 		spin_unlock_irqrestore(&hdwq->hdwq_lock, iflags);
 		tmp = kmalloc_node(sizeof(*tmp), GFP_ATOMIC,
-				   cpu_to_node(smp_processor_id()));
+				   cpu_to_node(raw_smp_processor_id()));
 		if (!tmp) {
 			lpfc_printf_log(phba, KERN_INFO, LOG_SLI,
 					"8355 error kmalloc memory for HDWQ "
Signature · 1 line
2.20.1
author maintainer
#130 lore raw

PATCH 123/205 media: cedrus: Use correct H264 8x8 scaling list

In reply to Sasha Levin

From: Jonas Karlman <[email protected]>

[ Upstream commit a6b8feae7c88343212686120740cf7551dd16e08 ]

Documentation now defines the expected order of scaling lists, change to use correct indices.

Fixes 6eb9b758e307 ("media: cedrus: Add H264 decoding support")
---
 drivers/staging/media/sunxi/cedrus/cedrus_h264.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/media/sunxi/cedrus/cedrus_h264.c +2 -2
index 08c6c9c410cc..c07526c12629 100644
@@ -244,8 +244,8 @@ static void cedrus_write_scaling_lists(struct cedrus_ctx *ctx,
 			       sizeof(scaling->scaling_list_8x8[0]));
 
 	cedrus_h264_write_sram(dev, CEDRUS_SRAM_H264_SCALING_LIST_8x8_1,
-			       scaling->scaling_list_8x8[3],
-			       sizeof(scaling->scaling_list_8x8[3]));
+			       scaling->scaling_list_8x8[1],
+			       sizeof(scaling->scaling_list_8x8[1]));
 
 	cedrus_h264_write_sram(dev, CEDRUS_SRAM_H264_SCALING_LIST_4x4,
 			       scaling->scaling_list_4x4,
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#131 lore raw

PATCH 124/205 media: aspeed-video: Fix memory leaks in aspeed_video_probe

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit c3df30a01da4955e04fa068c503cd784b31dad92 ]

In the implementation of aspeed_video_probe() the allocated memory for video should be released if either devm_ioremap_resource() or aspeed_video_init() or aspeed_video_setup_video() fails. Replace kzalloc() with devm_kzalloc to avoid explicit release for video.

Fixes d2b4387f3bdf ("media: platform: Add Aspeed Video Engine driver")
---
 drivers/media/platform/aspeed-video.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/media/platform/aspeed-video.c +2 -1
index 096a7c9a8963..4eaaf39b9223 100644
@@ -1658,7 +1658,8 @@ static int aspeed_video_probe(struct platform_device *pdev)
 {
 	int rc;
 	struct resource *res;
-	struct aspeed_video *video = kzalloc(sizeof(*video), GFP_KERNEL);
+	struct aspeed_video *video =
+		devm_kzalloc(&pdev->dev, sizeof(*video), GFP_KERNEL);
 
 	if (!video)
 		return -ENOMEM;
Signature · 1 line
2.20.1
author maintainer
#132 lore raw

PATCH 125/205 media: hantro: Set H264 FIELDPIC_FLAG_E flag correctly

In reply to Sasha Levin

From: Jonas Karlman <[email protected]>

[ Upstream commit a2cbf80a842add9663522bf898cf13cb2ac4e423 ]

The FIELDPIC_FLAG_E bit should be set when field_pic_flag exists in stream, it is currently set based on field_pic_flag of current frame. The PIC_FIELDMODE_E bit is correctly set based on the field_pic_flag.

Fix this by setting the FIELDPIC_FLAG_E bit when frame_mbs_only is not set.

Fixes dea0a82f3d22 ("media: hantro: Add support for H264 decoding on G1")
---
 drivers/staging/media/hantro/hantro_g1_h264_dec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/staging/media/hantro/hantro_g1_h264_dec.c +1 -1
index 636bf972adcf..5f29b7a836db 100644
@@ -63,7 +63,7 @@ static void set_params(struct hantro_ctx *ctx)
 	/* always use the matrix sent from userspace */
 	reg |= G1_REG_DEC_CTRL2_TYPE1_QUANT_E;
 
-	if (slices[0].flags &  V4L2_H264_SLICE_FLAG_FIELD_PIC)
+	if (!(sps->flags & V4L2_H264_SPS_FLAG_FRAME_MBS_ONLY))
 		reg |= G1_REG_DEC_CTRL2_FIELDPIC_FLAG_E;
 	vdpu_write_relaxed(vpu, reg, G1_REG_DEC_CTRL2);
 
Signature · 1 line
2.20.1
author maintainer
#133 lore raw

PATCH 126/205 mfd: intel-lpss: Add default I2C device properties for Gemini Lake

In reply to Sasha Levin

From: Jarkko Nikula <[email protected]>

[ Upstream commit 3f31bc67e4dc6a555341dffefe328ddd58e8b431 ]

It turned out Intel Gemini Lake doesn't use the same I2C timing parameters as Broxton.

I got confirmation from the Windows team that Gemini Lake systems should use updated timing parameters that differ from those used in Broxton based systems.

Fixes f80e78aa11ad ("mfd: intel-lpss: Add Intel Gemini Lake PCI IDs")
---
 drivers/mfd/intel-lpss-pci.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)
drivers/mfd/intel-lpss-pci.c +20 -8
index 9355db29d2f9..1767f30a1676 100644
@@ -122,6 +122,18 @@ static const struct intel_lpss_platform_info apl_i2c_info = {
 	.properties = apl_i2c_properties,
 };
 
+static struct property_entry glk_i2c_properties[] = {
+	PROPERTY_ENTRY_U32("i2c-sda-hold-time-ns", 313),
+	PROPERTY_ENTRY_U32("i2c-sda-falling-time-ns", 171),
+	PROPERTY_ENTRY_U32("i2c-scl-falling-time-ns", 290),
+	{ },
+};
+
+static const struct intel_lpss_platform_info glk_i2c_info = {
+	.clk_rate = 133000000,
+	.properties = glk_i2c_properties,
+};
+
 static const struct intel_lpss_platform_info cnl_i2c_info = {
 	.clk_rate = 216000000,
 	.properties = spt_i2c_properties,
@@ -174,14 +186,14 @@ static const struct pci_device_id intel_lpss_pci_ids[] = {
 	{ PCI_VDEVICE(INTEL, 0x1ac6), (kernel_ulong_t)&bxt_info },
 	{ PCI_VDEVICE(INTEL, 0x1aee), (kernel_ulong_t)&bxt_uart_info },
 	/* GLK */
-	{ PCI_VDEVICE(INTEL, 0x31ac), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31ae), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31b0), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31b2), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31b4), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31b6), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31b8), (kernel_ulong_t)&bxt_i2c_info },
-	{ PCI_VDEVICE(INTEL, 0x31ba), (kernel_ulong_t)&bxt_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31ac), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31ae), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31b0), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31b2), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31b4), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31b6), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31b8), (kernel_ulong_t)&glk_i2c_info },
+	{ PCI_VDEVICE(INTEL, 0x31ba), (kernel_ulong_t)&glk_i2c_info },
 	{ PCI_VDEVICE(INTEL, 0x31bc), (kernel_ulong_t)&bxt_uart_info },
 	{ PCI_VDEVICE(INTEL, 0x31be), (kernel_ulong_t)&bxt_uart_info },
 	{ PCI_VDEVICE(INTEL, 0x31c0), (kernel_ulong_t)&bxt_uart_info },
Signature · 1 line
2.20.1
author maintainer
#134 lore raw

PATCH 127/205 iommu/mediatek: Correct the flush_iotlb_all callback

In reply to Sasha Levin

From: Yong Wu <[email protected]>

[ Upstream commit 2009122f1d83dd8375572661961eab1e7e86bffe ]

Use the correct tlb_flush_all instead of the original one.

Fixes 4d689b619445 ("iommu/io-pgtable-arm-v7s: Convert to IOMMU API TLB sync")
---
 drivers/iommu/mtk_iommu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iommu/mtk_iommu.c +1 -1
index 67a483c1a935..76b9388cf689 100644
@@ -447,7 +447,7 @@ static size_t mtk_iommu_unmap(struct iommu_domain *domain,
 
 static void mtk_iommu_flush_iotlb_all(struct iommu_domain *domain)
 {
-	mtk_iommu_tlb_sync(mtk_iommu_get_m4u_data());
+	mtk_iommu_tlb_flush_all(mtk_iommu_get_m4u_data());
 }
 
 static void mtk_iommu_iotlb_sync(struct iommu_domain *domain,
Signature · 1 line
2.20.1
author maintainer
#135 lore raw

PATCH 128/205 iommu/mediatek: Add a new tlb_lock for tlb_flush

In reply to Sasha Levin

From: Yong Wu <[email protected]>

[ Upstream commit da3cc91b8db403728cde03c8a95cba268d8cbf1b ]

The commit 4d689b619445 ("iommu/io-pgtable-arm-v7s: Convert to IOMMU API TLB sync") help move the tlb_sync of unmap from v7s into the iommu framework. It helps add a new function "mtk_iommu_iotlb_sync", But it lacked the lock, then it will cause the variable "tlb_flush_active" may be changed unexpectedly, we could see this warning log randomly:

mtk-iommu 10205000.iommu: Partial TLB flush timed out, falling back to full flush

The HW requires tlb_flush/tlb_sync in pairs strictly, this patch adds a new tlb_lock for tlb operations to fix this issue.

Fixes 4d689b619445 ("iommu/io-pgtable-arm-v7s: Convert to IOMMU API TLB sync")
---
 drivers/iommu/mtk_iommu.c | 23 ++++++++++++++++++++++-
 drivers/iommu/mtk_iommu.h |  1 +
 2 files changed, 23 insertions(+), 1 deletion(-)
drivers/iommu/mtk_iommu.c +22 -1
index 76b9388cf689..c2f6c78fee44 100644
@@ -219,22 +219,37 @@ static void mtk_iommu_tlb_sync(void *cookie)
 static void mtk_iommu_tlb_flush_walk(unsigned long iova, size_t size,
 				     size_t granule, void *cookie)
 {
+	struct mtk_iommu_data *data = cookie;
+	unsigned long flags;
+
+	spin_lock_irqsave(&data->tlb_lock, flags);
 	mtk_iommu_tlb_add_flush_nosync(iova, size, granule, false, cookie);
 	mtk_iommu_tlb_sync(cookie);
+	spin_unlock_irqrestore(&data->tlb_lock, flags);
 }
 
 static void mtk_iommu_tlb_flush_leaf(unsigned long iova, size_t size,
 				     size_t granule, void *cookie)
 {
+	struct mtk_iommu_data *data = cookie;
+	unsigned long flags;
+
+	spin_lock_irqsave(&data->tlb_lock, flags);
 	mtk_iommu_tlb_add_flush_nosync(iova, size, granule, true, cookie);
 	mtk_iommu_tlb_sync(cookie);
+	spin_unlock_irqrestore(&data->tlb_lock, flags);
 }
 
 static void mtk_iommu_tlb_flush_page_nosync(struct iommu_iotlb_gather *gather,
 					    unsigned long iova, size_t granule,
 					    void *cookie)
 {
+	struct mtk_iommu_data *data = cookie;
+	unsigned long flags;
+
+	spin_lock_irqsave(&data->tlb_lock, flags);
 	mtk_iommu_tlb_add_flush_nosync(iova, granule, granule, true, cookie);
+	spin_unlock_irqrestore(&data->tlb_lock, flags);
 }
 
 static const struct iommu_flush_ops mtk_iommu_flush_ops = {
@@ -453,7 +468,12 @@ static void mtk_iommu_flush_iotlb_all(struct iommu_domain *domain)
 static void mtk_iommu_iotlb_sync(struct iommu_domain *domain,
 				 struct iommu_iotlb_gather *gather)
 {
-	mtk_iommu_tlb_sync(mtk_iommu_get_m4u_data());
+	struct mtk_iommu_data *data = mtk_iommu_get_m4u_data();
+	unsigned long flags;
+
+	spin_lock_irqsave(&data->tlb_lock, flags);
+	mtk_iommu_tlb_sync(data);
+	spin_unlock_irqrestore(&data->tlb_lock, flags);
 }
 
 static phys_addr_t mtk_iommu_iova_to_phys(struct iommu_domain *domain,
@@ -733,6 +753,7 @@ static int mtk_iommu_probe(struct platform_device *pdev)
 	if (ret)
 		return ret;
 
+	spin_lock_init(&data->tlb_lock);
 	list_add_tail(&data->list, &m4ulist);
 
 	if (!iommu_present(&platform_bus_type))
drivers/iommu/mtk_iommu.h +1 -0
index fc0f16eabacd..8cae22de7663 100644
@@ -58,6 +58,7 @@ struct mtk_iommu_data {
 	struct iommu_group		*m4u_group;
 	bool                            enable_4GB;
 	bool				tlb_flush_active;
+	spinlock_t			tlb_lock; /* lock for tlb range flush */
 
 	struct iommu_device		iommu;
 	const struct mtk_iommu_plat_data *plat_data;
Signature · 6 lines
2.20.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author maintainer
#136 lore raw

PATCH 129/205 i2c: stm32f7: report dma error during probe

In reply to Sasha Levin

From: Alain Volmat <[email protected]>

[ Upstream commit d77eceb2de99f5d7e0c645bad15511fe1af59e09 ]

Distinguish between the case where dma information is not provided within the DT and the case of an error during the dma init. Exit the probe with error in case of an error during dma init.

Fixes bb8822cbbc53 ("i2c: i2c-stm32: Add generic DMA API")
---
 drivers/i2c/busses/i2c-stm32.c   | 16 ++++++++--------
 drivers/i2c/busses/i2c-stm32f7.c |  9 +++++++++
 2 files changed, 17 insertions(+), 8 deletions(-)
drivers/i2c/busses/i2c-stm32.c +8 -8
index 07d5dfce68d4..1da347e6a358 100644
@@ -20,13 +20,13 @@ struct stm32_i2c_dma *stm32_i2c_dma_request(struct device *dev,
 
 	dma = devm_kzalloc(dev, sizeof(*dma), GFP_KERNEL);
 	if (!dma)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	/* Request and configure I2C TX dma channel */
-	dma->chan_tx = dma_request_slave_channel(dev, "tx");
-	if (!dma->chan_tx) {
+	dma->chan_tx = dma_request_chan(dev, "tx");
+	if (IS_ERR(dma->chan_tx)) {
 		dev_dbg(dev, "can't request DMA tx channel\n");
-		ret = -EINVAL;
+		ret = PTR_ERR(dma->chan_tx);
 		goto fail_al;
 	}
 
@@ -42,10 +42,10 @@ struct stm32_i2c_dma *stm32_i2c_dma_request(struct device *dev,
 	}
 
 	/* Request and configure I2C RX dma channel */
-	dma->chan_rx = dma_request_slave_channel(dev, "rx");
-	if (!dma->chan_rx) {
+	dma->chan_rx = dma_request_chan(dev, "rx");
+	if (IS_ERR(dma->chan_rx)) {
 		dev_err(dev, "can't request DMA rx channel\n");
-		ret = -EINVAL;
+		ret = PTR_ERR(dma->chan_rx);
 		goto fail_tx;
 	}
 
@@ -75,7 +75,7 @@ struct stm32_i2c_dma *stm32_i2c_dma_request(struct device *dev,
 	devm_kfree(dev, dma);
 	dev_info(dev, "can't use DMA\n");
 
-	return NULL;
+	return ERR_PTR(ret);
 }
 
 void stm32_i2c_dma_free(struct stm32_i2c_dma *dma)
drivers/i2c/busses/i2c-stm32f7.c +9 -0
index 84cfed17ff4f..37a76516e203 100644
@@ -1955,6 +1955,15 @@ static int stm32f7_i2c_probe(struct platform_device *pdev)
 	i2c_dev->dma = stm32_i2c_dma_request(i2c_dev->dev, phy_addr,
 					     STM32F7_I2C_TXDR,
 					     STM32F7_I2C_RXDR);
+	if (PTR_ERR(i2c_dev->dma) == -ENODEV)
+		i2c_dev->dma = NULL;
+	else if (IS_ERR(i2c_dev->dma)) {
+		ret = PTR_ERR(i2c_dev->dma);
+		if (ret != -EPROBE_DEFER)
+			dev_err(&pdev->dev,
+				"Failed to request dma error %i\n", ret);
+		goto clk_free;
+	}
 
 	platform_set_drvdata(pdev, i2c_dev);
 
Signature · 1 line
2.20.1
author maintainer
#137 lore raw

PATCH 130/205 ARM: OMAP2+: Add missing put_device() call in omapdss_init_of()

In reply to Sasha Levin

From: Markus Elfring <[email protected]>

[ Upstream commit 0b491904f053e41685162af5c5411b85b18c97a7 ]

A coccicheck run provided information like the following.

arch/arm/mach-omap2/display.c:268:2-8: ERROR: missing put_device; call of_find_device_by_node on line 258, but without a corresponding object release within this function.

Generated by: scripts/coccinelle/free/put_device.cocci

Thus add the missed function call to fix the exception handling for this function implementation.

Fixes e0c827aca0730b51f38081aa4e8ecf0912aab55f ("drm/omap: Populate DSS children in omapdss driver")
---
 arch/arm/mach-omap2/display.c | 1 +
 1 file changed, 1 insertion(+)
arch/arm/mach-omap2/display.c +1 -0
index 439e143cad7b..46012ca812f4 100644
@@ -265,6 +265,7 @@ static int __init omapdss_init_of(void)
 	r = of_platform_populate(node, NULL, NULL, &pdev->dev);
 	if (r) {
 		pr_err("Unable to populate DSS submodule devices\n");
+		put_device(&pdev->dev);
 		return r;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#138 lore raw

PATCH 131/205 s390/pkey: fix memory leak within _copy_apqns_from_user()

In reply to Sasha Levin

From: Heiko Carstens <[email protected]>

[ Upstream commit f9cac4fd8878929c6ebff0bd272317905d77c38a ]

Fixes f2bbc96e7cfad ("s390/pkey: add CCA AES cipher key support")
---
 drivers/s390/crypto/pkey_api.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/s390/crypto/pkey_api.c +3 -1
index 9de3d46b3253..e17fac20127e 100644
@@ -740,8 +740,10 @@ static void *_copy_apqns_from_user(void __user *uapqns, size_t nr_apqns)
 		kapqns = kmalloc(nbytes, GFP_KERNEL);
 		if (!kapqns)
 			return ERR_PTR(-ENOMEM);
-		if (copy_from_user(kapqns, uapqns, nbytes))
+		if (copy_from_user(kapqns, uapqns, nbytes)) {
+			kfree(kapqns);
 			return ERR_PTR(-EFAULT);
+		}
 	}
 
 	return kapqns;
Signature · 1 line
2.20.1
author maintainer
#139 lore raw

PATCH 132/205 kselftests: cgroup: Avoid the reuse of fd after it is deallocated

In reply to Sasha Levin

From: Hewenliang <[email protected]>

[ Upstream commit d671fa6393d6788fc65555d4643b71cb3a361f36 ]

It is necessary to set fd to -1 when inotify_add_watch() fails in cg_prepare_for_wait. Otherwise the fd which has been closed in cg_prepare_for_wait may be misused in other functions such as cg_enter_and_wait_for_frozen and cg_freeze_wait.

Fixes 5313bfe425c8 ("selftests: cgroup: add freezer controller self-tests")
---
 tools/testing/selftests/cgroup/test_freezer.c | 1 +
 1 file changed, 1 insertion(+)
tools/testing/selftests/cgroup/test_freezer.c +1 -0
index 0fc1b6d4b0f9..62a27ab3c2f3 100644
@@ -72,6 +72,7 @@ static int cg_prepare_for_wait(const char *cgroup)
 	if (ret == -1) {
 		debug("Error: inotify_add_watch() failed\n");
 		close(fd);
+		fd = -1;
 	}
 
 	return fd;
Signature · 1 line
2.20.1
author maintainer
#140 lore raw

PATCH 133/205 nfsd: Fix cld_net->cn_tfm initialization

In reply to Sasha Levin

From: Scott Mayhew <[email protected]>

[ Upstream commit 18b9a895e652979b70f9c20565394a69354dfebc ]

Don't assign an error pointer to cld_net->cn_tfm, otherwise an oops will occur in nfsd4_remove_cld_pipe().

Also, move the initialization of cld_net->cn_tfm so that it occurs after the check to see if nfsdcld is running. This is necessary because nfsd4_client_tracking_init() looks for -ETIMEDOUT to determine whether to use the "old" nfsdcld tracking ops.

Fixes 6ee95d1c8991 ("nfsd: add support for upcall version 2")
---
 fs/nfsd/nfs4recover.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)
fs/nfsd/nfs4recover.c +7 -5
index cdc75ad4438b..c35c0ebaf722 100644
@@ -1578,6 +1578,7 @@ nfsd4_cld_tracking_init(struct net *net)
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
 	bool running;
 	int retries = 10;
+	struct crypto_shash *tfm;
 
 	status = nfs4_cld_state_init(net);
 	if (status)
@@ -1586,11 +1587,6 @@ nfsd4_cld_tracking_init(struct net *net)
 	status = __nfsd4_init_cld_pipe(net);
 	if (status)
 		goto err_shutdown;
-	nn->cld_net->cn_tfm = crypto_alloc_shash("sha256", 0, 0);
-	if (IS_ERR(nn->cld_net->cn_tfm)) {
-		status = PTR_ERR(nn->cld_net->cn_tfm);
-		goto err_remove;
-	}
 
 	/*
 	 * rpc pipe upcalls take 30 seconds to time out, so we don't want to
@@ -1607,6 +1603,12 @@ nfsd4_cld_tracking_init(struct net *net)
 		status = -ETIMEDOUT;
 		goto err_remove;
 	}
+	tfm = crypto_alloc_shash("sha256", 0, 0);
+	if (IS_ERR(tfm)) {
+		status = PTR_ERR(tfm);
+		goto err_remove;
+	}
+	nn->cld_net->cn_tfm = tfm;
 
 	status = nfsd4_cld_get_version(nn);
 	if (status == -EOPNOTSUPP)
Signature · 1 line
2.20.1
author maintainer
#141 lore raw

PATCH 134/205 nfsd: v4 support requires CRYPTO_SHA256

In reply to Sasha Levin

From: Scott Mayhew <[email protected]>

[ Upstream commit a2e2f2dc77a18d2b0f450fb7fcb4871c9f697822 ]

The new nfsdcld client tracking operations use sha256 to compute hashes of the kerberos principals, so make sure CRYPTO_SHA256 is enabled.

Fixes 6ee95d1c8991 ("nfsd: add support for upcall version 2")
---
 fs/nfsd/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/nfsd/Kconfig +1 -1
index 10cefb0c07c7..c4b1a89b8845 100644
@@ -73,7 +73,7 @@ config NFSD_V4
 	select NFSD_V3
 	select FS_POSIX_ACL
 	select SUNRPC_GSS
-	select CRYPTO
+	select CRYPTO_SHA256
 	select GRACE_PERIOD
 	help
 	  This option enables support in your system's NFS server for
Signature · 1 line
2.20.1
author maintainer
#142 lore raw

PATCH 135/205 PCI: pciehp: Do not disable interrupt twice on suspend

In reply to Sasha Levin

From: Mika Westerberg <[email protected]>

[ Upstream commit 75fcc0ce72e5cea2e357cdde858216c5bad40442 ]

We try to keep PCIe hotplug ports runtime suspended when entering system suspend. Because the PCIe portdrv sets the DPM_FLAG_NEVER_SKIP flag, the PM core always calls system suspend/resume hooks even if the device is left runtime suspended. Since PCIe hotplug driver re-used the same function for both runtime suspend and system suspend, it ended up disabling hotplug interrupt twice and the second time following was printed:

pciehp 0000:03:01.0:pcie204: pcie_do_write_cmd: no response from device

Prevent this from happening by checking whether the device is already runtime suspended when the system suspend hook is called.

---
 drivers/pci/hotplug/pciehp_core.c | 25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)
drivers/pci/hotplug/pciehp_core.c +23 -2
index b3122c151b80..56daad828c9e 100644
@@ -253,7 +253,7 @@ static bool pme_is_native(struct pcie_device *dev)
 	return pcie_ports_native || host->native_pme;
 }
 
-static int pciehp_suspend(struct pcie_device *dev)
+static void pciehp_disable_interrupt(struct pcie_device *dev)
 {
 	/*
 	 * Disable hotplug interrupt so that it does not trigger
@@ -261,7 +261,19 @@ static int pciehp_suspend(struct pcie_device *dev)
 	 */
 	if (pme_is_native(dev))
 		pcie_disable_interrupt(get_service_data(dev));
+}
 
+#ifdef CONFIG_PM_SLEEP
+static int pciehp_suspend(struct pcie_device *dev)
+{
+	/*
+	 * If the port is already runtime suspended we can keep it that
+	 * way.
+	 */
+	if (dev_pm_smart_suspend_and_suspended(&dev->port->dev))
+		return 0;
+
+	pciehp_disable_interrupt(dev);
 	return 0;
 }
 
@@ -279,6 +291,7 @@ static int pciehp_resume_noirq(struct pcie_device *dev)
 
 	return 0;
 }
+#endif
 
 static int pciehp_resume(struct pcie_device *dev)
 {
@@ -292,6 +305,12 @@ static int pciehp_resume(struct pcie_device *dev)
 	return 0;
 }
 
+static int pciehp_runtime_suspend(struct pcie_device *dev)
+{
+	pciehp_disable_interrupt(dev);
+	return 0;
+}
+
 static int pciehp_runtime_resume(struct pcie_device *dev)
 {
 	struct controller *ctrl = get_service_data(dev);
@@ -318,10 +337,12 @@ static struct pcie_port_service_driver hpdriver_portdrv = {
 	.remove		= pciehp_remove,
 
 #ifdef	CONFIG_PM
+#ifdef	CONFIG_PM_SLEEP
 	.suspend	= pciehp_suspend,
 	.resume_noirq	= pciehp_resume_noirq,
 	.resume		= pciehp_resume,
-	.runtime_suspend = pciehp_suspend,
+#endif
+	.runtime_suspend = pciehp_runtime_suspend,
 	.runtime_resume	= pciehp_runtime_resume,
 #endif	/* PM */
 };
Signature · 1 line
2.20.1
author maintainer
#143 lore raw

PATCH 136/205 scsi: hisi_sas: Return directly if init hardware failed

In reply to Sasha Levin

From: Xiang Chen <[email protected]>

[ Upstream commit 547fde8b5a1923050f388caae4f76613b5a620e0 ]

Need to return directly if init hardware failed.

Fixes 73a4925d154c ("scsi: hisi_sas: Update all the registers after suspend and resume")
---
 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 1 +
 1 file changed, 1 insertion(+)
drivers/scsi/hisi_sas/hisi_sas_v3_hw.c +1 -0
index c4f76d7c29db..723f51c822af 100644
@@ -3423,6 +3423,7 @@ static int hisi_sas_v3_resume(struct pci_dev *pdev)
 	if (rc) {
 		scsi_remove_host(shost);
 		pci_disable_device(pdev);
+		return rc;
 	}
 	hisi_hba->hw->phys_init(hisi_hba);
 	sas_resume_ha(sha);
Signature · 1 line
2.20.1
author maintainer
#144 lore raw

PATCH 137/205 tipc: update mon's self addr when node addr generated

In reply to Sasha Levin

From: Hoang Le <[email protected]>

[ Upstream commit 46cb01eeeb86fca6afe24dda1167b0cb95424e29 ]

In commit 25b0b9c4e835 ("tipc: handle collisions of 32-bit node address hash values"), the 32-bit node address only generated after one second trial period expired. However the self's addr in struct tipc_monitor do not update according to node address generated. This lead to it is always zero as initial value. As result, sorting algorithm using this value does not work as expected, neither neighbor monitoring framework.

In this commit, we add a fix to update self's addr when 32-bit node address generated.

Fixes 25b0b9c4e835 ("tipc: handle collisions of 32-bit node address hash values")
---
 net/tipc/monitor.c | 15 +++++++++++++++
 net/tipc/monitor.h |  1 +
 net/tipc/net.c     |  2 ++
 3 files changed, 18 insertions(+)
net/tipc/monitor.c +15 -0
index 6a6eae88442f..58708b4c7719 100644
@@ -665,6 +665,21 @@ void tipc_mon_delete(struct net *net, int bearer_id)
 	kfree(mon);
 }
 
+void tipc_mon_reinit_self(struct net *net)
+{
+	struct tipc_monitor *mon;
+	int bearer_id;
+
+	for (bearer_id = 0; bearer_id < MAX_BEARERS; bearer_id++) {
+		mon = tipc_monitor(net, bearer_id);
+		if (!mon)
+			continue;
+		write_lock_bh(&mon->lock);
+		mon->self->addr = tipc_own_addr(net);
+		write_unlock_bh(&mon->lock);
+	}
+}
+
 int tipc_nl_monitor_set_threshold(struct net *net, u32 cluster_size)
 {
 	struct tipc_net *tn = tipc_net(net);
net/tipc/monitor.h +1 -0
index 2a21b93e0d04..ed63d2e650b0 100644
@@ -77,6 +77,7 @@ int __tipc_nl_add_monitor(struct net *net, struct tipc_nl_msg *msg,
 			  u32 bearer_id);
 int tipc_nl_add_monitor_peer(struct net *net, struct tipc_nl_msg *msg,
 			     u32 bearer_id, u32 *prev_node);
+void tipc_mon_reinit_self(struct net *net);
 
 extern const int tipc_max_domain_size;
 #endif
net/tipc/net.c +2 -0
index 85707c185360..2de3cec9929d 100644
@@ -42,6 +42,7 @@
 #include "node.h"
 #include "bcast.h"
 #include "netlink.h"
+#include "monitor.h"
 
 /*
  * The TIPC locking policy is designed to ensure a very fine locking
@@ -136,6 +137,7 @@ static void tipc_net_finalize(struct net *net, u32 addr)
 	tipc_set_node_addr(net, addr);
 	tipc_named_reinit(net);
 	tipc_sk_reinit(net);
+	tipc_mon_reinit_self(net);
 	tipc_nametbl_publish(net, TIPC_CFG_SRV, addr, addr,
 			     TIPC_CLUSTER_SCOPE, 0, addr);
 }
Signature · 1 line
2.20.1
author maintainer
#145 lore raw

PATCH 138/205 tty: serial: imx: use the sg count from dma_map_sg

In reply to Sasha Levin

From: Peng Fan <[email protected]>

[ Upstream commit 596fd8dffb745afcebc0ec6968e17fe29f02044c ]

The dmaengine_prep_slave_sg needs to use sg count returned by dma_map_sg, not use sport->dma_tx_nents, because the return value of dma_map_sg is not always same with "nents".

Fixes b4cdc8f61beb ("serial: imx: add DMA support for imx6q")
---
 drivers/tty/serial/imx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/tty/serial/imx.c +1 -1
index 5e08f2657b90..34f602c3a882 100644
@@ -619,7 +619,7 @@ static void imx_uart_dma_tx(struct imx_port *sport)
 		dev_err(dev, "DMA mapping error for TX.\n");
 		return;
 	}
-	desc = dmaengine_prep_slave_sg(chan, sgl, sport->dma_tx_nents,
+	desc = dmaengine_prep_slave_sg(chan, sgl, ret,
 					DMA_MEM_TO_DEV, DMA_PREP_INTERRUPT);
 	if (!desc) {
 		dma_unmap_sg(dev, sgl, sport->dma_tx_nents,
Signature · 1 line
2.20.1
author maintainer
#146 lore raw

PATCH 139/205 tty: serial: pch_uart: correct usage of dma_unmap_sg

In reply to Sasha Levin

From: Peng Fan <[email protected]>

[ Upstream commit 74887542fdcc92ad06a48c0cca17cdf09fc8aa00 ]

Per Documentation/DMA-API-HOWTO.txt,
To unmap a scatterlist, just call:

	dma_unmap_sg(dev, sglist, nents, direction);

.. note::

The 'nents' argument to the dma_unmap_sg call must be the _same_ one you passed into the dma_map_sg call, it should _NOT_ be the 'count' value _returned_ from the dma_map_sg call.

However in the driver, priv->nent is directly assigned with value returned from dma_map_sg, and dma_unmap_sg use priv->nent for unmap, this breaks the API usage.

So introduce a new entry orig_nent to remember 'nents'.

Fixes da3564ee027e ("pch_uart: add multi-scatter processing")
---
 drivers/tty/serial/pch_uart.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/tty/serial/pch_uart.c +4 -1
index 6157213a8359..c16234bca78f 100644
@@ -233,6 +233,7 @@ struct eg20t_port {
 	struct dma_chan			*chan_rx;
 	struct scatterlist		*sg_tx_p;
 	int				nent;
+	int				orig_nent;
 	struct scatterlist		sg_rx;
 	int				tx_dma_use;
 	void				*rx_buf_virt;
@@ -787,9 +788,10 @@ static void pch_dma_tx_complete(void *arg)
 	}
 	xmit->tail &= UART_XMIT_SIZE - 1;
 	async_tx_ack(priv->desc_tx);
-	dma_unmap_sg(port->dev, sg, priv->nent, DMA_TO_DEVICE);
+	dma_unmap_sg(port->dev, sg, priv->orig_nent, DMA_TO_DEVICE);
 	priv->tx_dma_use = 0;
 	priv->nent = 0;
+	priv->orig_nent = 0;
 	kfree(priv->sg_tx_p);
 	pch_uart_hal_enable_interrupt(priv, PCH_UART_HAL_TX_INT);
 }
@@ -1010,6 +1012,7 @@ static unsigned int dma_handle_tx(struct eg20t_port *priv)
 		dev_err(priv->port.dev, "%s:dma_map_sg Failed\n", __func__);
 		return 0;
 	}
+	priv->orig_nent = num;
 	priv->nent = nent;
 
 	for (i = 0; i < nent; i++, sg++) {
Signature · 1 line
2.20.1
author maintainer
#147 lore raw

PATCH 140/205 NFSv4.x: Handle bad/dead sessions correctly in nfs41_sequence_process()

In reply to Sasha Levin

From: Trond Myklebust <[email protected]>

[ Upstream commit 5c441544f045e679afd6c3c6d9f7aaf5fa5f37b0 ]

If the server returns a bad or dead session error, the we don't want to update the session slot number, but just immediately schedule recovery and allow it to proceed.

We can/should then remove handling in other places

Fixes 3453d5708b33 ("NFSv4.1: Avoid false retries when RPC calls are interrupted")
---
 fs/nfs/nfs4proc.c | 34 +++++++++++++++++++++++++---------
 1 file changed, 25 insertions(+), 9 deletions(-)
fs/nfs/nfs4proc.c +25 -9
index caacf5e7f5e1..a591aaf31071 100644
@@ -521,9 +521,7 @@ static int nfs4_do_handle_exception(struct nfs_server *server,
 		case -NFS4ERR_DEADSESSION:
 		case -NFS4ERR_SEQ_FALSE_RETRY:
 		case -NFS4ERR_SEQ_MISORDERED:
-			dprintk("%s ERROR: %d Reset session\n", __func__,
-				errorcode);
-			nfs4_schedule_session_recovery(clp->cl_session, errorcode);
+			/* Handled in nfs41_sequence_process() */
 			goto wait_on_recovery;
 #endif /* defined(CONFIG_NFS_V4_1) */
 		case -NFS4ERR_FILE_OPEN:
@@ -782,6 +780,7 @@ static int nfs41_sequence_process(struct rpc_task *task,
 	struct nfs4_session *session;
 	struct nfs4_slot *slot = res->sr_slot;
 	struct nfs_client *clp;
+	int status;
 	int ret = 1;
 
 	if (slot == NULL)
@@ -793,8 +792,13 @@ static int nfs41_sequence_process(struct rpc_task *task,
 	session = slot->table->session;
 
 	trace_nfs4_sequence_done(session, res);
+
+	status = res->sr_status;
+	if (task->tk_status == -NFS4ERR_DEADSESSION)
+		status = -NFS4ERR_DEADSESSION;
+
 	/* Check the SEQUENCE operation status */
-	switch (res->sr_status) {
+	switch (status) {
 	case 0:
 		/* Mark this sequence number as having been acked */
 		nfs4_slot_sequence_acked(slot, slot->seq_nr);
@@ -866,6 +870,10 @@ static int nfs41_sequence_process(struct rpc_task *task,
 		 */
 		slot->seq_nr = slot->seq_nr_highest_sent;
 		goto out_retry;
+	case -NFS4ERR_BADSESSION:
+	case -NFS4ERR_DEADSESSION:
+	case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
+		goto session_recover;
 	default:
 		/* Just update the slot sequence no. */
 		slot->seq_done = 1;
@@ -876,8 +884,10 @@ static int nfs41_sequence_process(struct rpc_task *task,
 out_noaction:
 	return ret;
 session_recover:
-	nfs4_schedule_session_recovery(session, res->sr_status);
-	goto retry_nowait;
+	nfs4_schedule_session_recovery(session, status);
+	dprintk("%s ERROR: %d Reset session\n", __func__, status);
+	nfs41_sequence_free_slot(res);
+	goto out;
 retry_new_seq:
 	++slot->seq_nr;
 retry_nowait:
@@ -2188,7 +2198,6 @@ static int nfs4_handle_delegation_recall_error(struct nfs_server *server, struct
 		case -NFS4ERR_BAD_HIGH_SLOT:
 		case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
 		case -NFS4ERR_DEADSESSION:
-			nfs4_schedule_session_recovery(server->nfs_client->cl_session, err);
 			return -EAGAIN;
 		case -NFS4ERR_STALE_CLIENTID:
 		case -NFS4ERR_STALE_STATEID:
@@ -7820,6 +7829,15 @@ nfs41_same_server_scope(struct nfs41_server_scope *a,
 static void
 nfs4_bind_one_conn_to_session_done(struct rpc_task *task, void *calldata)
 {
+	struct nfs41_bind_conn_to_session_args *args = task->tk_msg.rpc_argp;
+	struct nfs_client *clp = args->client;
+
+	switch (task->tk_status) {
+	case -NFS4ERR_BADSESSION:
+	case -NFS4ERR_DEADSESSION:
+		nfs4_schedule_session_recovery(clp->cl_session,
+				task->tk_status);
+	}
 }
 
 static const struct rpc_call_ops nfs4_bind_one_conn_to_session_ops = {
@@ -8867,8 +8885,6 @@ static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nf
 	case -NFS4ERR_BADSESSION:
 	case -NFS4ERR_DEADSESSION:
 	case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION:
-		nfs4_schedule_session_recovery(clp->cl_session,
-				task->tk_status);
 		break;
 	default:
 		nfs4_schedule_lease_recovery(clp);
Signature · 1 line
2.20.1
author maintainer
#148 lore raw

PATCH 141/205 firmware: arm_scmi: Fix doorbell ring logic for !CONFIG_64BIT

In reply to Sasha Levin

From: Sudeep Holla <[email protected]>

[ Upstream commit 7bd39bc6bfdf96f5df0f92199bbc1a3ee2f2adb8 ]

The logic to ring the scmi performance fastchannel ignores the value read from the doorbell register in case of !CONFIG_64BIT. This bug also shows up as warning with '-Wunused-but-set-variable' gcc flag:

drivers/firmware/arm_scmi/perf.c: In function scmi_perf_fc_ring_db:
drivers/firmware/arm_scmi/perf.c:323:7: warning: variable val set but
not used [-Wunused-but-set-variable]

Fix the same by aligning the logic with CONFIG_64BIT as used in the macro SCMI_PERF_FC_RING_DB().

Fixes 823839571d76 ("firmware: arm_scmi: Make use SCMI v2.0 fastchannel for performance protocol")
---
 drivers/firmware/arm_scmi/perf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/arm_scmi/perf.c +1 -1
index 4a8012e3cb8c..601af4edad5e 100644
@@ -323,7 +323,7 @@ static void scmi_perf_fc_ring_db(struct scmi_fc_db_info *db)
 
 		if (db->mask)
 			val = ioread64_hi_lo(db->addr) & db->mask;
-		iowrite64_hi_lo(db->set, db->addr);
+		iowrite64_hi_lo(db->set | val, db->addr);
 	}
 #endif
 }
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#149 lore raw

PATCH 142/205 iwlwifi: mvm: fix support for single antenna diversity

In reply to Sasha Levin

From: Luca Coelho <[email protected]>

[ Upstream commit bb99ff9baa02beb9216c86678999342197c849cc ]

When the single antenna diversity support was sent upstream, only some definitions were sent, due to a bad revert.

Fix this by adding the actual code.

Fixes 5952e0ec3f05 ("iwlwifi: mvm: add support for single antenna diversity")
---
 drivers/net/wireless/intel/iwlwifi/mvm/fw.c | 20 ++++++++++++++++----
 1 file changed, 16 insertions(+), 4 deletions(-)
drivers/net/wireless/intel/iwlwifi/mvm/fw.c +16 -4
index d9eb2b286438..c59cbb8cbdd7 100644
@@ -514,6 +514,18 @@ static int iwl_send_phy_cfg_cmd(struct iwl_mvm *mvm)
 	struct iwl_phy_cfg_cmd phy_cfg_cmd;
 	enum iwl_ucode_type ucode_type = mvm->fwrt.cur_fw_img;
 
+	if (iwl_mvm_has_unified_ucode(mvm) &&
+	    !mvm->trans->cfg->tx_with_siso_diversity) {
+		return 0;
+	} else if (mvm->trans->cfg->tx_with_siso_diversity) {
+		/*
+		 * TODO: currently we don't set the antenna but letting the NIC
+		 * to decide which antenna to use. This should come from BIOS.
+		 */
+		phy_cfg_cmd.phy_cfg =
+			cpu_to_le32(FW_PHY_CFG_CHAIN_SAD_ENABLED);
+	}
+
 	/* Set parameters */
 	phy_cfg_cmd.phy_cfg = cpu_to_le32(iwl_mvm_get_phy_config(mvm));
 
@@ -1344,12 +1356,12 @@ int iwl_mvm_up(struct iwl_mvm *mvm)
 		ret = iwl_send_phy_db_data(mvm->phy_db);
 		if (ret)
 			goto error;
-
-		ret = iwl_send_phy_cfg_cmd(mvm);
-		if (ret)
-			goto error;
 	}
 
+	ret = iwl_send_phy_cfg_cmd(mvm);
+	if (ret)
+		goto error;
+
 	ret = iwl_mvm_send_bt_init_conf(mvm);
 	if (ret)
 		goto error;
Signature · 1 line
2.20.1
author maintainer
#150 lore raw

PATCH 143/205 mmc: sdio: fix wl1251 vendor id

In reply to Sasha Levin

From: "H. Nikolaus Schaller" <[email protected]>

[ Upstream commit e5db673e7fe2f971ec82039a28dc0811c2100e87 ]

v4.11-rc1 did introduce a patch series that rearranged the sdio quirks into a header file. Unfortunately this did forget to handle SDIO_VENDOR_ID_TI differently between wl1251 and wl1271 with the result that although the wl1251 was found on the sdio bus, the firmware did not load any more and there was no interface registration.

This patch defines separate constants to be used by sdio quirks and drivers.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
Cc <[email protected]> # v4.11+
---
 include/linux/mmc/sdio_ids.h | 2 ++
 1 file changed, 2 insertions(+)
include/linux/mmc/sdio_ids.h +2 -0
index d1a5d5df02f5..08b25c02b5a1 100644
@@ -71,6 +71,8 @@
 
 #define SDIO_VENDOR_ID_TI			0x0097
 #define SDIO_DEVICE_ID_TI_WL1271		0x4076
+#define SDIO_VENDOR_ID_TI_WL1251		0x104c
+#define SDIO_DEVICE_ID_TI_WL1251		0x9066
 
 #define SDIO_VENDOR_ID_STE			0x0020
 #define SDIO_DEVICE_ID_STE_CW1200		0x2280
Signature · 1 line
2.20.1
author maintainer
#151 lore raw

PATCH 144/205 mmc: core: fix wl1251 sdio quirks

In reply to Sasha Levin

From: "H. Nikolaus Schaller" <[email protected]>

[ Upstream commit 16568b4a4f0c34bd35cfadac63303c7af7812764 ]

wl1251 and wl1271 have different vendor id and device id. So we need to handle both with sdio quirks.

Fixes 884f38607897 ("mmc: core: move some sdio IDs out of quirks file")
Cc <[email protected]> # v4.11+
---
 drivers/mmc/core/quirks.h | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/mmc/core/quirks.h +7 -0
index 2d2d9ea8be4f..3dba15bccce2 100644
@@ -119,7 +119,14 @@ static const struct mmc_fixup mmc_ext_csd_fixups[] = {
 	END_FIXUP
 };
 
+
 static const struct mmc_fixup sdio_fixup_methods[] = {
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI_WL1251, SDIO_DEVICE_ID_TI_WL1251,
+		   add_quirk, MMC_QUIRK_NONSTD_FUNC_IF),
+
+	SDIO_FIXUP(SDIO_VENDOR_ID_TI_WL1251, SDIO_DEVICE_ID_TI_WL1251,
+		   add_quirk, MMC_QUIRK_DISABLE_CD),
+
 	SDIO_FIXUP(SDIO_VENDOR_ID_TI, SDIO_DEVICE_ID_TI_WL1271,
 		   add_quirk, MMC_QUIRK_NONSTD_FUNC_IF),
 
Signature · 1 line
2.20.1
author maintainer
#152 lore raw

PATCH 145/205 tee: optee: Fix dynamic shm pool allocations

In reply to Sasha Levin

From: Sumit Garg <[email protected]>

[ Upstream commit a249dd200d03791cab23e47571f3e13d9c72af6c ]

In case of dynamic shared memory pool, kernel memory allocated using dmabuf_mgr pool needs to be registered with OP-TEE prior to its usage during optee_open_session() or optee_invoke_func().

So fix dmabuf_mgr pool allocations via an additional call to optee_shm_register().

Also, allow kernel pages to be registered as shared memory with OP-TEE.

Fixes 9733b072a12a ("optee: allow to work without static shared memory")
---
 drivers/tee/optee/call.c     |  7 +++++++
 drivers/tee/optee/shm_pool.c | 12 +++++++++++-
 2 files changed, 18 insertions(+), 1 deletion(-)
drivers/tee/optee/call.c +7 -0
index 13b0269a0abc..cf2367ba08d6 100644
@@ -554,6 +554,13 @@ static int check_mem_type(unsigned long start, size_t num_pages)
 	struct mm_struct *mm = current->mm;
 	int rc;
 
+	/*
+	 * Allow kernel address to register with OP-TEE as kernel
+	 * pages are configured as normal memory only.
+	 */
+	if (virt_addr_valid(start))
+		return 0;
+
 	down_read(&mm->mmap_sem);
 	rc = __check_mem_type(find_vma(mm, start),
 			      start + num_pages * PAGE_SIZE);
drivers/tee/optee/shm_pool.c +11 -1
index de1d9b8fad90..0332a5301d61 100644
@@ -17,6 +17,7 @@ static int pool_op_alloc(struct tee_shm_pool_mgr *poolm,
 {
 	unsigned int order = get_order(size);
 	struct page *page;
+	int rc = 0;
 
 	page = alloc_pages(GFP_KERNEL | __GFP_ZERO, order);
 	if (!page)
@@ -26,12 +27,21 @@ static int pool_op_alloc(struct tee_shm_pool_mgr *poolm,
 	shm->paddr = page_to_phys(page);
 	shm->size = PAGE_SIZE << order;
 
-	return 0;
+	if (shm->flags & TEE_SHM_DMA_BUF) {
+		shm->flags |= TEE_SHM_REGISTER;
+		rc = optee_shm_register(shm->ctx, shm, &page, 1 << order,
+					(unsigned long)shm->kaddr);
+	}
+
+	return rc;
 }
 
 static void pool_op_free(struct tee_shm_pool_mgr *poolm,
 			 struct tee_shm *shm)
 {
+	if (shm->flags & TEE_SHM_DMA_BUF)
+		optee_shm_unregister(shm->ctx, shm);
+
 	free_pages((unsigned long)shm->kaddr, get_order(shm->size));
 	shm->kaddr = NULL;
 }
Signature · 1 line
2.20.1
author maintainer
#153 lore raw

PATCH 146/205 tee: optee: fix device enumeration error handling

In reply to Sasha Levin

From: Jens Wiklander <[email protected]>

[ Upstream commit 03212e347f9443e524d6383c6806ac08295c1fb0 ]

Prior to this patch in optee_probe() when optee_enumerate_devices() was called the struct optee was fully initialized. If optee_enumerate_devices() returns an error optee_probe() is supposed to clean up and free the struct optee completely, but will at this late stage need to call optee_remove() instead. This isn't done and thus freeing the struct optee prematurely.

With this patch the call to optee_enumerate_devices() is done after optee_probe() has returned successfully and in case optee_enumerate_devices() fails everything is cleaned up with a call to optee_remove().

Fixes c3fa24af9244 ("tee: optee: add TEE bus device enumeration support")
---
 drivers/tee/optee/core.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)
drivers/tee/optee/core.c +12 -8
index 1854a3db7345..b830e0a87fba 100644
@@ -643,11 +643,6 @@ static struct optee *optee_probe(struct device_node *np)
 	if (optee->sec_caps & OPTEE_SMC_SEC_CAP_DYNAMIC_SHM)
 		pr_info("dynamic shared memory is enabled\n");
 
-	rc = optee_enumerate_devices();
-	if (rc)
-		goto err;
-
-	pr_info("initialized driver\n");
 	return optee;
 err:
 	if (optee) {
@@ -702,9 +697,10 @@ static struct optee *optee_svc;
 
 static int __init optee_driver_init(void)
 {
-	struct device_node *fw_np;
-	struct device_node *np;
-	struct optee *optee;
+	struct device_node *fw_np = NULL;
+	struct device_node *np = NULL;
+	struct optee *optee = NULL;
+	int rc = 0;
 
 	/* Node is supposed to be below /firmware */
 	fw_np = of_find_node_by_name(NULL, "firmware");
@@ -723,6 +719,14 @@ static int __init optee_driver_init(void)
 	if (IS_ERR(optee))
 		return PTR_ERR(optee);
 
+	rc = optee_enumerate_devices();
+	if (rc) {
+		optee_remove(optee);
+		return rc;
+	}
+
+	pr_info("initialized driver\n");
+
 	optee_svc = optee;
 
 	return 0;
Signature · 1 line
2.20.1
author maintainer
#154 lore raw

PATCH 147/205 workqueue: Add RCU annotation for pwq list walk

In reply to Sasha Levin

From: Sebastian Andrzej Siewior <[email protected]>

[ Upstream commit 49e9d1a9faf2f71fdfd80a30697ee9a15070626d ]

An additional check has been recently added to ensure that a RCU related lock is held while the RCU list is iterated.
The `pwqs' are sometimes iterated without a RCU lock but with the &wq->mutex acquired leading to a warning.

Teach list_for_each_entry_rcu() that the RCU usage is okay if &wq->mutex is acquired during the list traversal.

Fixes 28875945ba98d ("rcu: Add support for consolidated-RCU reader checking")
---
 kernel/workqueue.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
kernel/workqueue.c +2 -1
index 649687622654..e9c63b79e03f 100644
@@ -425,7 +425,8 @@ static void workqueue_sysfs_unregister(struct workqueue_struct *wq);
  * ignored.
  */
 #define for_each_pwq(pwq, wq)						\
-	list_for_each_entry_rcu((pwq), &(wq)->pwqs, pwqs_node)		\
+	list_for_each_entry_rcu((pwq), &(wq)->pwqs, pwqs_node,		\
+				lockdep_is_held(&wq->mutex))		\
 		if (({ assert_rcu_or_wq_mutex(wq); false; })) { }	\
 		else
 
Signature · 1 line
2.20.1
author maintainer
#155 lore raw

PATCH 148/205 netfilter: nf_tables_offload: release flow_rule on error from commit path

In reply to Sasha Levin

From: Pablo Neira Ayuso <[email protected]>

[ Upstream commit 23403cd8898dbc9808d3eb2f63bc1db8a340b751 ]

If hardware offload commit path fails, release all flow_rule objects.

Fixes c9626a2cbdb2 ("netfilter: nf_tables: add hardware offload support")
---
 net/netfilter/nf_tables_offload.c | 26 +++++++++++++++++++++-----
 1 file changed, 21 insertions(+), 5 deletions(-)
net/netfilter/nf_tables_offload.c +21 -5
index e743f811245f..96a64e7594a5 100644
@@ -358,14 +358,14 @@ int nft_flow_rule_offload_commit(struct net *net)
 				continue;
 
 			if (trans->ctx.flags & NLM_F_REPLACE ||
-			    !(trans->ctx.flags & NLM_F_APPEND))
-				return -EOPNOTSUPP;
-
+			    !(trans->ctx.flags & NLM_F_APPEND)) {
+				err = -EOPNOTSUPP;
+				break;
+			}
 			err = nft_flow_offload_rule(trans->ctx.chain,
 						    nft_trans_rule(trans),
 						    nft_trans_flow_rule(trans),
 						    FLOW_CLS_REPLACE);
-			nft_flow_rule_destroy(nft_trans_flow_rule(trans));
 			break;
 		case NFT_MSG_DELRULE:
 			if (!(trans->ctx.chain->flags & NFT_CHAIN_HW_OFFLOAD))
@@ -379,7 +379,23 @@ int nft_flow_rule_offload_commit(struct net *net)
 		}
 
 		if (err)
-			return err;
+			break;
+	}
+
+	list_for_each_entry(trans, &net->nft.commit_list, list) {
+		if (trans->ctx.family != NFPROTO_NETDEV)
+			continue;
+
+		switch (trans->msg_type) {
+		case NFT_MSG_NEWRULE:
+			if (!(trans->ctx.chain->flags & NFT_CHAIN_HW_OFFLOAD))
+				continue;
+
+			nft_flow_rule_destroy(nft_trans_flow_rule(trans));
+			break;
+		default:
+			break;
+		}
 	}
 
 	return err;
Signature · 1 line
2.20.1
author maintainer
#156 lore raw

PATCH 149/205 SUNRPC: Fix another issue with MIC buffer space

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit e8d70b321ecc9b23d09b8df63e38a2f73160c209 ]

xdr_shrink_pagelen() BUG's when @len is larger than buf->page_len. This can happen when xdr_buf_read_mic() is given an xdr_buf with a small page array (like, only a few bytes).

Instead, just cap the number of bytes that xdr_shrink_pagelen() will move.

Fixes 5f1bc39979d ("SUNRPC: Fix buffer handling of GSS MIC ... ")
---
 net/sunrpc/xdr.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)
net/sunrpc/xdr.c +5 -6
index 14ba9e72a204..f3104be8ff5d 100644
@@ -436,13 +436,12 @@ xdr_shrink_bufhead(struct xdr_buf *buf, size_t len)
 }
 
 /**
- * xdr_shrink_pagelen
+ * xdr_shrink_pagelen - shrinks buf->pages by up to @len bytes
  * @buf: xdr_buf
  * @len: bytes to remove from buf->pages
  *
- * Shrinks XDR buffer's page array buf->pages by
- * 'len' bytes. The extra data is not lost, but is instead
- * moved into the tail.
+ * The extra data is not lost, but is instead moved into buf->tail.
+ * Returns the actual number of bytes moved.
  */
 static unsigned int
 xdr_shrink_pagelen(struct xdr_buf *buf, size_t len)
@@ -455,8 +454,8 @@ xdr_shrink_pagelen(struct xdr_buf *buf, size_t len)
 
 	result = 0;
 	tail = buf->tail;
-	BUG_ON (len > pglen);
-
+	if (len > buf->page_len)
+		len = buf-> page_len;
 	tailbuf_len = buf->buflen - buf->head->iov_len - buf->page_len;
 
 	/* Shift the tail first */
Signature · 1 line
2.20.1
author maintainer
#157 lore raw

PATCH 150/205 affs: fix a memory leak in affs_remount

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit 450c3d4166837c496ebce03650c08800991f2150 ]

In affs_remount if data is provided it is duplicated into new_opts. The allocated memory for new_opts is only released if parse_options fails.

There's a bit of history behind new_options, originally there was save/replace options on the VFS layer so the 'data' passed must not change (thus strdup), this got cleaned up in later patches. But not completely.

There's no reason to do the strdup in cases where the filesystem does not need to reuse the 'data' again, because strsep would modify it directly.

Fixes c8f33d0bec99 ("affs: kstrdup() memory handling")
[ update changelog ]
---
 fs/affs/super.c | 6 ------
 1 file changed, 6 deletions(-)
fs/affs/super.c +0 -6
index cc463ae47c12..3812f7bc3a7f 100644
@@ -561,14 +561,9 @@ affs_remount(struct super_block *sb, int *flags, char *data)
 	int			 root_block;
 	unsigned long		 mount_flags;
 	int			 res = 0;
-	char			*new_opts;
 	char			 volume[32];
 	char			*prefix = NULL;
 
-	new_opts = kstrdup(data, GFP_KERNEL);
-	if (data && !new_opts)
-		return -ENOMEM;
-
 	pr_debug("%s(flags=0x%x,opts=\"%s\")\n", __func__, *flags, data);
 
 	sync_filesystem(sb);
@@ -579,7 +574,6 @@ affs_remount(struct super_block *sb, int *flags, char *data)
 			   &blocksize, &prefix, volume,
 			   &mount_flags)) {
 		kfree(prefix);
-		kfree(new_opts);
 		return -EINVAL;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#158 lore raw

PATCH 151/205 sched/cpufreq: Move the cfs_rq_util_change() call to cpufreq_update_util()

In reply to Sasha Levin

From: Vincent Guittot <[email protected]>

[ Upstream commit bef69dd87828ef5d8ecdab8d857cd3a33cf98675 ]

update_cfs_rq_load_avg() calls cfs_rq_util_change() every time PELT decays, which might be inefficient when the cpufreq driver has rate limitation.

When a task is attached on a CPU, we have this call path:

update_load_avg()
update_cfs_rq_load_avg()
cfs_rq_util_change -- > trig frequency update
attach_entity_load_avg()
cfs_rq_util_change -- > trig frequency update

The 1st frequency update will not take into account the utilization of the newly attached task and the 2nd one might be discarded because of rate limitation of the cpufreq driver.

update_cfs_rq_load_avg() is only called by update_blocked_averages() and update_load_avg() so we can move the call to cfs_rq_util_change/cpufreq_update_util() into these two functions.

It's also interesting to note that update_load_avg() already calls cfs_rq_util_change() directly for the !SMP case.

This change will also ensure that cpufreq_update_util() is called even when there is no more CFS rq in the leaf_cfs_rq_list to update, but only IRQ, RT or DL PELT signals.

[ mingo: Minor updates. ]

---
 kernel/sched/fair.c | 111 +++++++++++++++++++++++++-------------------
 1 file changed, 62 insertions(+), 49 deletions(-)
kernel/sched/fair.c +62 -49
index 69a81a5709ff..989352861275 100644
@@ -3504,9 +3504,6 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq)
 	cfs_rq->load_last_update_time_copy = sa->last_update_time;
 #endif
 
-	if (decayed)
-		cfs_rq_util_change(cfs_rq, 0);
-
 	return decayed;
 }
 
@@ -3616,8 +3613,12 @@ static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s
 		attach_entity_load_avg(cfs_rq, se, SCHED_CPUFREQ_MIGRATION);
 		update_tg_load_avg(cfs_rq, 0);
 
-	} else if (decayed && (flags & UPDATE_TG))
-		update_tg_load_avg(cfs_rq, 0);
+	} else if (decayed) {
+		cfs_rq_util_change(cfs_rq, 0);
+
+		if (flags & UPDATE_TG)
+			update_tg_load_avg(cfs_rq, 0);
+	}
 }
 
 #ifndef CONFIG_64BIT
@@ -7517,6 +7518,28 @@ static inline bool others_have_blocked(struct rq *rq) { return false; }
 static inline void update_blocked_load_status(struct rq *rq, bool has_blocked) {}
 #endif
 
+static bool __update_blocked_others(struct rq *rq, bool *done)
+{
+	const struct sched_class *curr_class;
+	u64 now = rq_clock_pelt(rq);
+	bool decayed;
+
+	/*
+	 * update_load_avg() can call cpufreq_update_util(). Make sure that RT,
+	 * DL and IRQ signals have been updated before updating CFS.
+	 */
+	curr_class = rq->curr->sched_class;
+
+	decayed = update_rt_rq_load_avg(now, rq, curr_class == &rt_sched_class) |
+		  update_dl_rq_load_avg(now, rq, curr_class == &dl_sched_class) |
+		  update_irq_load_avg(rq, 0);
+
+	if (others_have_blocked(rq))
+		*done = false;
+
+	return decayed;
+}
+
 #ifdef CONFIG_FAIR_GROUP_SCHED
 
 static inline bool cfs_rq_is_decayed(struct cfs_rq *cfs_rq)
@@ -7536,29 +7559,11 @@ static inline bool cfs_rq_is_decayed(struct cfs_rq *cfs_rq)
 	return true;
 }
 
-static void update_blocked_averages(int cpu)
+static bool __update_blocked_fair(struct rq *rq, bool *done)
 {
-	struct rq *rq = cpu_rq(cpu);
 	struct cfs_rq *cfs_rq, *pos;
-	const struct sched_class *curr_class;
-	struct rq_flags rf;
-	bool done = true;
-
-	rq_lock_irqsave(rq, &rf);
-	update_rq_clock(rq);
-
-	/*
-	 * update_cfs_rq_load_avg() can call cpufreq_update_util(). Make sure
-	 * that RT, DL and IRQ signals have been updated before updating CFS.
-	 */
-	curr_class = rq->curr->sched_class;
-	update_rt_rq_load_avg(rq_clock_pelt(rq), rq, curr_class == &rt_sched_class);
-	update_dl_rq_load_avg(rq_clock_pelt(rq), rq, curr_class == &dl_sched_class);
-	update_irq_load_avg(rq, 0);
-
-	/* Don't need periodic decay once load/util_avg are null */
-	if (others_have_blocked(rq))
-		done = false;
+	bool decayed = false;
+	int cpu = cpu_of(rq);
 
 	/*
 	 * Iterates the task_group tree in a bottom up fashion, see
@@ -7567,9 +7572,13 @@ static void update_blocked_averages(int cpu)
 	for_each_leaf_cfs_rq_safe(rq, cfs_rq, pos) {
 		struct sched_entity *se;
 
-		if (update_cfs_rq_load_avg(cfs_rq_clock_pelt(cfs_rq), cfs_rq))
+		if (update_cfs_rq_load_avg(cfs_rq_clock_pelt(cfs_rq), cfs_rq)) {
 			update_tg_load_avg(cfs_rq, 0);
 
+			if (cfs_rq == &rq->cfs)
+				decayed = true;
+		}
+
 		/* Propagate pending load changes to the parent, if any: */
 		se = cfs_rq->tg->se[cpu];
 		if (se && !skip_blocked_update(se))
@@ -7584,11 +7593,10 @@ static void update_blocked_averages(int cpu)
 
 		/* Don't need periodic decay once load/util_avg are null */
 		if (cfs_rq_has_blocked(cfs_rq))
-			done = false;
+			*done = false;
 	}
 
-	update_blocked_load_status(rq, !done);
-	rq_unlock_irqrestore(rq, &rf);
+	return decayed;
 }
 
 /*
@@ -7638,29 +7646,16 @@ static unsigned long task_h_load(struct task_struct *p)
 			cfs_rq_load_avg(cfs_rq) + 1);
 }
 #else
-static inline void update_blocked_averages(int cpu)
+static bool __update_blocked_fair(struct rq *rq, bool *done)
 {
-	struct rq *rq = cpu_rq(cpu);
 	struct cfs_rq *cfs_rq = &rq->cfs;
-	const struct sched_class *curr_class;
-	struct rq_flags rf;
-
-	rq_lock_irqsave(rq, &rf);
-	update_rq_clock(rq);
-
-	/*
-	 * update_cfs_rq_load_avg() can call cpufreq_update_util(). Make sure
-	 * that RT, DL and IRQ signals have been updated before updating CFS.
-	 */
-	curr_class = rq->curr->sched_class;
-	update_rt_rq_load_avg(rq_clock_pelt(rq), rq, curr_class == &rt_sched_class);
-	update_dl_rq_load_avg(rq_clock_pelt(rq), rq, curr_class == &dl_sched_class);
-	update_irq_load_avg(rq, 0);
+	bool decayed;
 
-	update_cfs_rq_load_avg(cfs_rq_clock_pelt(cfs_rq), cfs_rq);
+	decayed = update_cfs_rq_load_avg(cfs_rq_clock_pelt(cfs_rq), cfs_rq);
+	if (cfs_rq_has_blocked(cfs_rq))
+		*done = false;
 
-	update_blocked_load_status(rq, cfs_rq_has_blocked(cfs_rq) || others_have_blocked(rq));
-	rq_unlock_irqrestore(rq, &rf);
+	return decayed;
 }
 
 static unsigned long task_h_load(struct task_struct *p)
@@ -7669,6 +7664,24 @@ static unsigned long task_h_load(struct task_struct *p)
 }
 #endif
 
+static void update_blocked_averages(int cpu)
+{
+	bool decayed = false, done = true;
+	struct rq *rq = cpu_rq(cpu);
+	struct rq_flags rf;
+
+	rq_lock_irqsave(rq, &rf);
+	update_rq_clock(rq);
+
+	decayed |= __update_blocked_others(rq, &done);
+	decayed |= __update_blocked_fair(rq, &done);
+
+	update_blocked_load_status(rq, !done);
+	if (decayed)
+		cpufreq_update_util(rq, 0);
+	rq_unlock_irqrestore(rq, &rf);
+}
+
 /********** Helpers for find_busiest_group ************************/
 
 /*
Signature · 1 line
2.20.1
author maintainer
#159 lore raw

PATCH 152/205 rtc: msm6242: Fix reading of 10-hour digit

In reply to Sasha Levin

From: Kars de Jong <[email protected]>

[ Upstream commit e34494c8df0cd96fc432efae121db3212c46ae48 ]

The driver was reading the wrong register as the 10-hour digit due to a misplaced ')'. It was in fact reading the 1-second digit register due to this bug.

Also remove the use of a magic number for the hour mask and use the define for it which was already present.

---
 drivers/rtc/rtc-msm6242.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/rtc/rtc-msm6242.c +2 -1
index 1c2d3c4a4963..b1f2bedee77e 100644
@@ -133,7 +133,8 @@ static int msm6242_read_time(struct device *dev, struct rtc_time *tm)
 		      msm6242_read(priv, MSM6242_SECOND1);
 	tm->tm_min  = msm6242_read(priv, MSM6242_MINUTE10) * 10 +
 		      msm6242_read(priv, MSM6242_MINUTE1);
-	tm->tm_hour = (msm6242_read(priv, MSM6242_HOUR10 & 3)) * 10 +
+	tm->tm_hour = (msm6242_read(priv, MSM6242_HOUR10) &
+		       MSM6242_HOUR10_HR_MASK) * 10 +
 		      msm6242_read(priv, MSM6242_HOUR1);
 	tm->tm_mday = msm6242_read(priv, MSM6242_DAY10) * 10 +
 		      msm6242_read(priv, MSM6242_DAY1);
Signature · 1 line
2.20.1
author maintainer
#160 lore raw

PATCH 153/205 watchdog: sprd: Fix the incorrect pointer getting from driver data

In reply to Sasha Levin

From: Shuiqing Li <[email protected]>

[ Upstream commit 39e68d9e7ab276880980ee5386301fb218202192 ]

The device driver data saved the 'struct sprd_wdt' object, it is incorrect to get 'struct watchdog_device' object from the driver data, thus fix it.

---
 drivers/watchdog/sprd_wdt.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/watchdog/sprd_wdt.c +2 -4
index 0bb17b046140..65cb55f3916f 100644
@@ -327,10 +327,9 @@ static int sprd_wdt_probe(struct platform_device *pdev)
 
 static int __maybe_unused sprd_wdt_pm_suspend(struct device *dev)
 {
-	struct watchdog_device *wdd = dev_get_drvdata(dev);
 	struct sprd_wdt *wdt = dev_get_drvdata(dev);
 
-	if (watchdog_active(wdd))
+	if (watchdog_active(&wdt->wdd))
 		sprd_wdt_stop(&wdt->wdd);
 	sprd_wdt_disable(wdt);
 
@@ -339,7 +338,6 @@ static int __maybe_unused sprd_wdt_pm_suspend(struct device *dev)
 
 static int __maybe_unused sprd_wdt_pm_resume(struct device *dev)
 {
-	struct watchdog_device *wdd = dev_get_drvdata(dev);
 	struct sprd_wdt *wdt = dev_get_drvdata(dev);
 	int ret;
 
@@ -347,7 +345,7 @@ static int __maybe_unused sprd_wdt_pm_resume(struct device *dev)
 	if (ret)
 		return ret;
 
-	if (watchdog_active(wdd)) {
+	if (watchdog_active(&wdt->wdd)) {
 		ret = sprd_wdt_start(&wdt->wdd);
 		if (ret) {
 			sprd_wdt_disable(wdt);
Signature · 1 line
2.20.1
author maintainer
#161 lore raw

PATCH 154/205 clk: Fix memory leak in clk_unregister()

In reply to Sasha Levin

From: Kishon Vijay Abraham I <[email protected]>

[ Upstream commit 8247470772beb38822f226c99a2ed8c195f6b438 ]

Memory allocated in alloc_clk() for 'struct clk' and 'const char *con_id' while invoking clk_register() is never freed in clk_unregister(), resulting in kmemleak showing the following backtrace.

backtrace:

    [<00000000546f5dd0>] kmem_cache_alloc+0x18c/0x270
    [<0000000073a32862>] alloc_clk+0x30/0x70
    [<0000000082942480>] __clk_register+0xc8/0x760
    [<000000005c859fca>] devm_clk_register+0x54/0xb0
    [<00000000868834a8>] 0xffff800008c60950
    [<00000000d5a80534>] platform_drv_probe+0x50/0xa0
    [<000000001b3889fc>] really_probe+0x108/0x348
    [<00000000953fa60a>] driver_probe_device+0x58/0x100
    [<0000000008acc17c>] device_driver_attach+0x6c/0x90
    [<0000000022813df3>] __driver_attach+0x84/0xc8
    [<00000000448d5443>] bus_for_each_dev+0x74/0xc8
    [<00000000294aa93f>] driver_attach+0x20/0x28
    [<00000000e5e52626>] bus_add_driver+0x148/0x1f0
    [<000000001de21efc>] driver_register+0x60/0x110
    [<00000000af07c068>] __platform_driver_register+0x40/0x48
    [<0000000060fa80ee>] 0xffff800008c66020

Fix it here.

Cc Tomi Valkeinen <[email protected]>
Cc Tero Kristo <[email protected]>
Fixes 1df4046a93e0 ("clk: Combine __clk_get() and __clk_create_clk()")
---
 drivers/clk/clk.c | 1 +
 1 file changed, 1 insertion(+)
drivers/clk/clk.c +1 -0
index 27a95c86a80b..4fc294c2f9e8 100644
@@ -3886,6 +3886,7 @@ void clk_unregister(struct clk *clk)
 					__func__, clk->core->name);
 
 	kref_put(&clk->core->ref, __clk_release);
+	free_clk(clk);
 unlock:
 	clk_prepare_unlock();
 }
Signature · 1 line
2.20.1
author maintainer
#162 lore raw

PATCH 155/205 libbpf: Fix call relocation offset calculation bug

In reply to Sasha Levin

From: Andrii Nakryiko <[email protected]>

[ Upstream commit a0d7da26ce86a25e97ae191cb90574ada6daea98 ]

When relocating subprogram call, libbpf doesn't take into account relo->text_off, which comes from symbol's value. This generally works fine for subprograms implemented as static functions, but breaks for global functions.

Taking a simplified test_pkt_access.c as an example:

__attribute__ ((noinline))
static int test_pkt_access_subprog1(volatile struct __sk_buff *skb)

{
        return skb->len * 2;
}

__attribute__ ((noinline))
static int test_pkt_access_subprog2(int val, volatile struct __sk_buff *skb)

{
        return skb->len + val;
}

SEC("classifier/test_pkt_access")
int test_pkt_access(struct __sk_buff *skb)

{
        if (test_pkt_access_subprog1(skb) != skb->len * 2)
                return TC_ACT_SHOT;
        if (test_pkt_access_subprog2(2, skb) != skb->len + 2)
                return TC_ACT_SHOT;
        return TC_ACT_UNSPEC;
}

When compiled, we get two relocations, pointing to '.text' symbol. .text has st_value set to 0 (it points to the beginning of .text section):

0000000000000008  000000050000000a R_BPF_64_32            0000000000000000 .text
0000000000000040  000000050000000a R_BPF_64_32            0000000000000000 .text

test_pkt_access_subprog1 and test_pkt_access_subprog2 offsets (targets of two calls) are encoded within call instruction's imm32 part as -1 and 2, respectively:

0000000000000000 test_pkt_access_subprog1:

       0:       61 10 00 00 00 00 00 00 r0 = *(u32 *)(r1 + 0)
       1:       64 00 00 00 01 00 00 00 w0 <<= 1
       2:       95 00 00 00 00 00 00 00 exit

0000000000000018 test_pkt_access_subprog2:

       3:       61 10 00 00 00 00 00 00 r0 = *(u32 *)(r1 + 0)
       4:       04 00 00 00 02 00 00 00 w0 += 2
       5:       95 00 00 00 00 00 00 00 exit

0000000000000000 test_pkt_access:

       0:       bf 16 00 00 00 00 00 00 r6 = r1
===>   1:       85 10 00 00 ff ff ff ff call -1
       2:       bc 01 00 00 00 00 00 00 w1 = w0
       3:       b4 00 00 00 02 00 00 00 w0 = 2
       4:       61 62 00 00 00 00 00 00 r2 = *(u32 *)(r6 + 0)
       5:       64 02 00 00 01 00 00 00 w2 <<= 1
       6:       5e 21 08 00 00 00 00 00 if w1 != w2 goto +8 <LBB0_3>
       7:       bf 61 00 00 00 00 00 00 r1 = r6
===>   8:       85 10 00 00 02 00 00 00 call 2
       9:       bc 01 00 00 00 00 00 00 w1 = w0
      10:       61 62 00 00 00 00 00 00 r2 = *(u32 *)(r6 + 0)
      11:       04 02 00 00 02 00 00 00 w2 += 2
      12:       b4 00 00 00 ff ff ff ff w0 = -1
      13:       1e 21 01 00 00 00 00 00 if w1 == w2 goto +1 <LBB0_3>
      14:       b4 00 00 00 02 00 00 00 w0 = 2
0000000000000078 LBB0_3:
      15:       95 00 00 00 00 00 00 00 exit

Now, if we compile example with global functions, the setup changes. Relocations are now against specifically test_pkt_access_subprog1 and test_pkt_access_subprog2 symbols, with test_pkt_access_subprog2 pointing 24 bytes into its respective section (.text), i.e., 3 instructions in:

0000000000000008  000000070000000a R_BPF_64_32            0000000000000000 test_pkt_access_subprog1
0000000000000048  000000080000000a R_BPF_64_32            0000000000000018 test_pkt_access_subprog2

Calls instructions now encode offsets relative to function symbols and are both set ot -1:

0000000000000000 test_pkt_access_subprog1:

       0:       61 10 00 00 00 00 00 00 r0 = *(u32 *)(r1 + 0)
       1:       64 00 00 00 01 00 00 00 w0 <<= 1
       2:       95 00 00 00 00 00 00 00 exit

0000000000000018 test_pkt_access_subprog2:

       3:       61 20 00 00 00 00 00 00 r0 = *(u32 *)(r2 + 0)
       4:       0c 10 00 00 00 00 00 00 w0 += w1
       5:       95 00 00 00 00 00 00 00 exit

0000000000000000 test_pkt_access:

       0:       bf 16 00 00 00 00 00 00 r6 = r1
===>   1:       85 10 00 00 ff ff ff ff call -1
       2:       bc 01 00 00 00 00 00 00 w1 = w0
       3:       b4 00 00 00 02 00 00 00 w0 = 2
       4:       61 62 00 00 00 00 00 00 r2 = *(u32 *)(r6 + 0)
       5:       64 02 00 00 01 00 00 00 w2 <<= 1
       6:       5e 21 09 00 00 00 00 00 if w1 != w2 goto +9 <LBB2_3>
       7:       b4 01 00 00 02 00 00 00 w1 = 2
       8:       bf 62 00 00 00 00 00 00 r2 = r6
===>   9:       85 10 00 00 ff ff ff ff call -1
      10:       bc 01 00 00 00 00 00 00 w1 = w0
      11:       61 62 00 00 00 00 00 00 r2 = *(u32 *)(r6 + 0)
      12:       04 02 00 00 02 00 00 00 w2 += 2
      13:       b4 00 00 00 ff ff ff ff w0 = -1
      14:       1e 21 01 00 00 00 00 00 if w1 == w2 goto +1 <LBB2_3>
      15:       b4 00 00 00 02 00 00 00 w0 = 2
0000000000000080 LBB2_3:
      16:       95 00 00 00 00 00 00 00 exit

Thus the right formula to calculate target call offset after relocation should take into account relocation's target symbol value (offset within section), call instruction's imm32 offset, and (subtracting, to get relative instruction offset) instruction index of call instruction itself. All that is shifted by number of instructions in main program, given all sub-programs are copied over after main program.

Convert few selftests relying on bpf-to-bpf calls to use global functions instead of static ones.

---
 tools/lib/bpf/libbpf.c                             | 8 ++++++--
 tools/testing/selftests/bpf/progs/test_btf_haskv.c | 4 ++--
 tools/testing/selftests/bpf/progs/test_btf_newkv.c | 4 ++--
 tools/testing/selftests/bpf/progs/test_btf_nokv.c  | 4 ++--
 4 files changed, 12 insertions(+), 8 deletions(-)
tools/lib/bpf/libbpf.c +6 -2
index d98838c5820c..de2be6b2a748 100644
@@ -1791,9 +1791,13 @@ bpf_program__collect_reloc(struct bpf_program *prog, GElf_Shdr *shdr,
 				pr_warning("incorrect bpf_call opcode\n");
 				return -LIBBPF_ERRNO__RELOC;
 			}
+			if (sym.st_value % 8) {
+				pr_warn("bad call relo offset: %lu\n", sym.st_value);
+				return -LIBBPF_ERRNO__RELOC;
+			}
 			prog->reloc_desc[i].type = RELO_CALL;
 			prog->reloc_desc[i].insn_idx = insn_idx;
-			prog->reloc_desc[i].text_off = sym.st_value;
+			prog->reloc_desc[i].text_off = sym.st_value / 8;
 			obj->has_pseudo_calls = true;
 			continue;
 		}
@@ -3239,7 +3243,7 @@ bpf_program__reloc_text(struct bpf_program *prog, struct bpf_object *obj,
 			 prog->section_name);
 	}
 	insn = &prog->insns[relo->insn_idx];
-	insn->imm += prog->main_prog_cnt - relo->insn_idx;
+	insn->imm += relo->text_off + prog->main_prog_cnt - relo->insn_idx;
 	return 0;
 }
 
tools/testing/selftests/bpf/progs/test_btf_haskv.c +2 -2
index e5c79fe0ffdb..d65c61e64df2 100644
@@ -25,7 +25,7 @@ struct dummy_tracepoint_args {
 };
 
 __attribute__((noinline))
-static int test_long_fname_2(struct dummy_tracepoint_args *arg)
+int test_long_fname_2(struct dummy_tracepoint_args *arg)
 {
 	struct ipv_counts *counts;
 	int key = 0;
@@ -43,7 +43,7 @@ static int test_long_fname_2(struct dummy_tracepoint_args *arg)
 }
 
 __attribute__((noinline))
-static int test_long_fname_1(struct dummy_tracepoint_args *arg)
+int test_long_fname_1(struct dummy_tracepoint_args *arg)
 {
 	return test_long_fname_2(arg);
 }
tools/testing/selftests/bpf/progs/test_btf_newkv.c +2 -2
index 5ee3622ddebb..8e83317db841 100644
@@ -33,7 +33,7 @@ struct dummy_tracepoint_args {
 };
 
 __attribute__((noinline))
-static int test_long_fname_2(struct dummy_tracepoint_args *arg)
+int test_long_fname_2(struct dummy_tracepoint_args *arg)
 {
 	struct ipv_counts *counts;
 	int key = 0;
@@ -56,7 +56,7 @@ static int test_long_fname_2(struct dummy_tracepoint_args *arg)
 }
 
 __attribute__((noinline))
-static int test_long_fname_1(struct dummy_tracepoint_args *arg)
+int test_long_fname_1(struct dummy_tracepoint_args *arg)
 {
 	return test_long_fname_2(arg);
 }
tools/testing/selftests/bpf/progs/test_btf_nokv.c +2 -2
index 434188c37774..3f4422044759 100644
@@ -23,7 +23,7 @@ struct dummy_tracepoint_args {
 };
 
 __attribute__((noinline))
-static int test_long_fname_2(struct dummy_tracepoint_args *arg)
+int test_long_fname_2(struct dummy_tracepoint_args *arg)
 {
 	struct ipv_counts *counts;
 	int key = 0;
@@ -41,7 +41,7 @@ static int test_long_fname_2(struct dummy_tracepoint_args *arg)
 }
 
 __attribute__((noinline))
-static int test_long_fname_1(struct dummy_tracepoint_args *arg)
+int test_long_fname_1(struct dummy_tracepoint_args *arg)
 {
 	return test_long_fname_2(arg);
 }
Signature · 1 line
2.20.1
author maintainer
#163 lore raw

PATCH 156/205 scsi: qla4xxx: fix double free bug

In reply to Sasha Levin

From: Pan Bian <[email protected]>

[ Upstream commit 3fe3d2428b62822b7b030577cd612790bdd8c941 ]

The variable init_fw_cb is released twice, resulting in a double free bug. The call to the function dma_free_coherent() before goto is removed to get rid of potential double free.

---
 drivers/scsi/qla4xxx/ql4_mbx.c | 3 ---
 1 file changed, 3 deletions(-)
drivers/scsi/qla4xxx/ql4_mbx.c +0 -3
index dac9a7013208..02636b4785c5 100644
@@ -640,9 +640,6 @@ int qla4xxx_initialize_fw_cb(struct scsi_qla_host * ha)
 
 	if (qla4xxx_get_ifcb(ha, &mbox_cmd[0], &mbox_sts[0], init_fw_cb_dma) !=
 	    QLA_SUCCESS) {
-		dma_free_coherent(&ha->pdev->dev,
-				  sizeof(struct addr_ctrl_blk),
-				  init_fw_cb, init_fw_cb_dma);
 		goto exit_init_fw_cb;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#164 lore raw

PATCH 157/205 scsi: bnx2i: fix potential use after free

In reply to Sasha Levin

From: Pan Bian <[email protected]>

[ Upstream commit 29d28f2b8d3736ac61c28ef7e20fda63795b74d9 ]

The member hba->pcidev may be used after its reference is dropped. Move the put function to where it is never used to avoid potential use after free issues.

Fixes a77171806515 ("[SCSI] bnx2i: Removed the reference to the netdev->base_addr")
---
 drivers/scsi/bnx2i/bnx2i_iscsi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/bnx2i/bnx2i_iscsi.c +1 -1
index c5fa5f3b00e9..0b28d44d3573 100644
@@ -915,12 +915,12 @@ void bnx2i_free_hba(struct bnx2i_hba *hba)
 	INIT_LIST_HEAD(&hba->ep_ofld_list);
 	INIT_LIST_HEAD(&hba->ep_active_list);
 	INIT_LIST_HEAD(&hba->ep_destroy_list);
-	pci_dev_put(hba->pcidev);
 
 	if (hba->regview) {
 		pci_iounmap(hba->pcidev, hba->regview);
 		hba->regview = NULL;
 	}
+	pci_dev_put(hba->pcidev);
 	bnx2i_free_mp_bdt(hba);
 	bnx2i_release_free_cid_que(hba);
 	iscsi_host_free(shost);
Signature · 1 line
2.20.1
author maintainer
#165 lore raw

PATCH 158/205 scsi: target: core: Fix a pr_debug() argument

In reply to Sasha Levin

From: Bart Van Assche <[email protected]>

[ Upstream commit c941e0d172605731de9b4628bd4146d35cf2e7d6 ]

Print the string for which conversion failed instead of printing the function name twice.

Fixes 2650d71e244f ("target: move transport ID handling to the core")
Cc Christoph Hellwig <[email protected]>
---
 drivers/target/target_core_fabric_lib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/target/target_core_fabric_lib.c +1 -1
index 3c79411c4cd0..6b4b354c88aa 100644
@@ -118,7 +118,7 @@ static int srp_get_pr_transport_id(
 	memset(buf + 8, 0, leading_zero_bytes);
 	rc = hex2bin(buf + 8 + leading_zero_bytes, p, count);
 	if (rc < 0) {
-		pr_debug("hex2bin failed for %s: %d\n", __func__, rc);
+		pr_debug("hex2bin failed for %s: %d\n", p, rc);
 		return rc;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#166 lore raw

PATCH 159/205 mt76: mt76u: rely on usb_interface instead of usb_dev

In reply to Sasha Levin

From: Lorenzo Bianconi <[email protected]>

[ Upstream commit 80df01f4dc79abbed724bbe0851cab3fe8ad9d99 ]

usb drivers are supposed to communicate using usb_interface instead mt76x{0,2}u is now registering through usb_device. Fix it by passing usb_intf device to mt76_alloc_device routine.

Fixes 112f980ac8926 ("mt76usb: use usb_dev private data")
---
 drivers/net/wireless/mediatek/mt76/mt76.h       |  3 ++-
 drivers/net/wireless/mediatek/mt76/mt76x0/usb.c |  2 +-
 drivers/net/wireless/mediatek/mt76/mt76x2/usb.c |  2 +-
 drivers/net/wireless/mediatek/mt76/usb.c        | 12 +++++++++---
 4 files changed, 13 insertions(+), 6 deletions(-)
drivers/net/wireless/mediatek/mt76/mt76.h +2 -1
index 8aec7ccf2d79..e8605ae6f8a5 100644
@@ -799,7 +799,8 @@ static inline int
 mt76u_bulk_msg(struct mt76_dev *dev, void *data, int len, int *actual_len,
 	       int timeout)
 {
-	struct usb_device *udev = to_usb_device(dev->dev);
+	struct usb_interface *uintf = to_usb_interface(dev->dev);
+	struct usb_device *udev = interface_to_usbdev(uintf);
 	struct mt76_usb *usb = &dev->usb;
 	unsigned int pipe;
 
drivers/net/wireless/mediatek/mt76/mt76x0/usb.c +1 -1
index 00a445d27599..65d404e61404 100644
@@ -226,7 +226,7 @@ static int mt76x0u_probe(struct usb_interface *usb_intf,
 	u32 mac_rev;
 	int ret;
 
-	mdev = mt76_alloc_device(&usb_dev->dev, sizeof(*dev), &mt76x0u_ops,
+	mdev = mt76_alloc_device(&usb_intf->dev, sizeof(*dev), &mt76x0u_ops,
 				 &drv_ops);
 	if (!mdev)
 		return -ENOMEM;
drivers/net/wireless/mediatek/mt76/mt76x2/usb.c +1 -1
index da5e0f9a8bae..8b26c6108186 100644
@@ -39,7 +39,7 @@ static int mt76x2u_probe(struct usb_interface *intf,
 	struct mt76_dev *mdev;
 	int err;
 
-	mdev = mt76_alloc_device(&udev->dev, sizeof(*dev), &mt76x2u_ops,
+	mdev = mt76_alloc_device(&intf->dev, sizeof(*dev), &mt76x2u_ops,
 				 &drv_ops);
 	if (!mdev)
 		return -ENOMEM;
drivers/net/wireless/mediatek/mt76/usb.c +9 -3
index 20c6fe510e9d..05aa42bd9808 100644
@@ -20,7 +20,8 @@ static int __mt76u_vendor_request(struct mt76_dev *dev, u8 req,
 				  u8 req_type, u16 val, u16 offset,
 				  void *buf, size_t len)
 {
-	struct usb_device *udev = to_usb_device(dev->dev);
+	struct usb_interface *uintf = to_usb_interface(dev->dev);
+	struct usb_device *udev = interface_to_usbdev(uintf);
 	unsigned int pipe;
 	int i, ret;
 
@@ -235,7 +236,8 @@ mt76u_rd_rp(struct mt76_dev *dev, u32 base,
 
 static bool mt76u_check_sg(struct mt76_dev *dev)
 {
-	struct usb_device *udev = to_usb_device(dev->dev);
+	struct usb_interface *uintf = to_usb_interface(dev->dev);
+	struct usb_device *udev = interface_to_usbdev(uintf);
 
 	return (!disable_usb_sg && udev->bus->sg_tablesize > 0 &&
 		(udev->bus->no_sg_constraint ||
@@ -370,7 +372,8 @@ mt76u_fill_bulk_urb(struct mt76_dev *dev, int dir, int index,
 		    struct urb *urb, usb_complete_t complete_fn,
 		    void *context)
 {
-	struct usb_device *udev = to_usb_device(dev->dev);
+	struct usb_interface *uintf = to_usb_interface(dev->dev);
+	struct usb_device *udev = interface_to_usbdev(uintf);
 	unsigned int pipe;
 
 	if (dir == USB_DIR_IN)
@@ -952,6 +955,7 @@ int mt76u_init(struct mt76_dev *dev,
 		.rd_rp = mt76u_rd_rp,
 		.type = MT76_BUS_USB,
 	};
+	struct usb_device *udev = interface_to_usbdev(intf);
 	struct mt76_usb *usb = &dev->usb;
 
 	tasklet_init(&usb->rx_tasklet, mt76u_rx_tasklet, (unsigned long)dev);
@@ -965,6 +969,8 @@ int mt76u_init(struct mt76_dev *dev,
 	dev->bus = &mt76u_ops;
 	dev->queue_ops = &usb_queue_ops;
 
+	dev_set_drvdata(&udev->dev, dev);
+
 	usb->sg_en = mt76u_check_sg(dev);
 
 	return mt76u_set_endpoints(intf, usb);
Signature · 1 line
2.20.1
author maintainer
#167 lore raw

PATCH 160/205 ARM: dts: dra7: fix cpsw mdio fck clock

In reply to Sasha Levin

From: Grygorii Strashko <[email protected]>

[ Upstream commit 6af0a549c25e0d02366aa95507bfe3cad2f7b68b ]

The DRA7 CPSW MDIO functional clock (gmac_clkctrl DRA7_GMAC_GMAC_CLKCTRL 0) is specified incorrectly, which is caused incorrect MDIO bus clock configuration MDCLK. The correct CPSW MDIO functional clock is gmac_main_clk (125MHz), which is the same as CPSW fck. Hence fix it.

Fixes 1faa415c9c6e ("ARM: dts: Add fck for cpsw mdio for omap variants")
---
 arch/arm/boot/dts/dra7-l4.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/dra7-l4.dtsi +1 -1
index 5cac2dd58241..c3954e34835b 100644
@@ -3059,7 +3059,7 @@
 
 				davinci_mdio: mdio@1000 {
 					compatible = "ti,cpsw-mdio","ti,davinci_mdio";
-					clocks = <&gmac_clkctrl DRA7_GMAC_GMAC_CLKCTRL 0>;
+					clocks = <&gmac_main_clk>;
 					clock-names = "fck";
 					#address-cells = <1>;
 					#size-cells = <0>;
Signature · 1 line
2.20.1
author maintainer
#168 lore raw

PATCH 161/205 dma-direct: don't check swiotlb=force in dma_direct_map_resource

In reply to Sasha Levin

From: Christoph Hellwig <[email protected]>

[ Upstream commit 4268ac6ae5870af10a7417b22990d615f72f77e2 ]

When mapping resources we can't just use swiotlb ram for bounce buffering. Switch to a direct dma_capable check instead.

Fixes cfced786969c ("dma-mapping: remove the default map_resource implementation")
---
 kernel/dma/direct.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/dma/direct.c +1 -1
index 8402b29c280f..867fd72cb260 100644
@@ -375,7 +375,7 @@ dma_addr_t dma_direct_map_resource(struct device *dev, phys_addr_t paddr,
 {
 	dma_addr_t dma_addr = paddr;
 
-	if (unlikely(!dma_direct_possible(dev, dma_addr, size))) {
+	if (unlikely(!dma_capable(dev, dma_addr, size))) {
 		report_addr(dev, dma_addr, size);
 		return DMA_MAPPING_ERROR;
 	}
Signature · 6 lines
2.20.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author maintainer
#169 lore raw

PATCH 162/205 powerpc/powernv: Disable native PCIe port management

In reply to Sasha Levin

From: Oliver O'Halloran <[email protected]>

[ Upstream commit 9d72dcef891030545f39ad386a30cf91df517fb2 ]

On PowerNV the PCIe topology is (currently) managed by the powernv platform code in Linux in cooperation with the platform firmware. Linux's native PCIe port service drivers operate independently of both and this can cause problems.

The main issue is that the portbus driver will conflict with the platform specific hotplug driver (pnv_php) over ownership of the MSI used to notify the host when a hotplug event occurs. The portbus driver claims this MSI on behalf of the individual port services because the same interrupt is used for hotplug events, PMEs (on root ports), and link bandwidth change notifications. The portbus driver will always claim the interrupt even if the individual port service drivers, such as pciehp, are compiled out.

The second, bigger, problem is that the hotplug port service driver fundamentally does not work on PowerNV. The platform assumes that all PCI devices have a corresponding arch-specific handle derived from the DT node for the device (pci_dn) and without one the platform will not allow a PCI device to be enabled. This problem is largely due to historical baggage, but it can't be resolved without significant re-factoring of the platform PCI support.

We can fix these problems in the interim by setting the "pcie_ports_disabled" flag during platform initialisation. The flag indicates the platform owns the PCIe ports which stops the portbus driver from being registered.

This does have the side effect of disabling all port services drivers that is: AER, PME, BW notifications, hotplug, and DPC. However, this is not a huge disadvantage on PowerNV since these services are either unused or handled through other means.

Fixes 66725152fb9f ("PCI/hotplug: PowerPC PowerNV PCI hotplug driver")
---
 arch/powerpc/platforms/powernv/pci.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
arch/powerpc/platforms/powernv/pci.c +17 -0
index 2825d004dece..c0bea75ac27b 100644
@@ -945,6 +945,23 @@ void __init pnv_pci_init(void)
 	if (!firmware_has_feature(FW_FEATURE_OPAL))
 		return;
 
+#ifdef CONFIG_PCIEPORTBUS
+	/*
+	 * On PowerNV PCIe devices are (currently) managed in cooperation
+	 * with firmware. This isn't *strictly* required, but there's enough
+	 * assumptions baked into both firmware and the platform code that
+	 * it's unwise to allow the portbus services to be used.
+	 *
+	 * We need to fix this eventually, but for now set this flag to disable
+	 * the portbus driver. The AER service isn't required since that AER
+	 * events are handled via EEH. The pciehp hotplug driver can't work
+	 * without kernel changes (and portbus binding breaks pnv_php). The
+	 * other services also require some thinking about how we're going
+	 * to integrate them.
+	 */
+	pcie_ports_disabled = true;
+#endif
+
 	/* Look for IODA IO-Hubs. */
 	for_each_compatible_node(np, NULL, "ibm,ioda-hub") {
 		pnv_pci_init_ioda_hub(np);
Signature · 1 line
2.20.1
author maintainer
#170 lore raw

PATCH 163/205 afs: Remove set but not used variables 'before', 'after'

In reply to Sasha Levin

From: zhengbin <[email protected]>

[ Upstream commit 51590df4f3306cb1f43dca54e3ccdd121ab89594 ]

Fixes gcc '-Wunused-but-set-variable' warning:

fs/afs/dir_edit.c: In function afs_set_contig_bits:
fs/afs/dir_edit.c:75:20: warning: variable after set but not used [-Wunused-but-set-variable] fs/afs/dir_edit.c: In function afs_set_contig_bits:
fs/afs/dir_edit.c:75:12: warning: variable before set but not used [-Wunused-but-set-variable] fs/afs/dir_edit.c: In function afs_clear_contig_bits:
fs/afs/dir_edit.c:100:20: warning: variable after set but not used [-Wunused-but-set-variable] fs/afs/dir_edit.c: In function afs_clear_contig_bits:
fs/afs/dir_edit.c:100:12: warning: variable before set but not used [-Wunused-but-set-variable]

They are never used since commit 63a4681ff39c.

Fixes 63a4681ff39c ("afs: Locally edit directory data for mkdir/create/unlink/...")
---
 fs/afs/dir_edit.c | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)
fs/afs/dir_edit.c +2 -10
index d4fbe5f85f1b..b108528bf010 100644
@@ -68,13 +68,11 @@ static int afs_find_contig_bits(union afs_xdr_dir_block *block, unsigned int nr_
 static void afs_set_contig_bits(union afs_xdr_dir_block *block,
 				int bit, unsigned int nr_slots)
 {
-	u64 mask, before, after;
+	u64 mask;
 
 	mask = (1 << nr_slots) - 1;
 	mask <<= bit;
 
-	before = *(u64 *)block->hdr.bitmap;
-
 	block->hdr.bitmap[0] |= (u8)(mask >> 0 * 8);
 	block->hdr.bitmap[1] |= (u8)(mask >> 1 * 8);
 	block->hdr.bitmap[2] |= (u8)(mask >> 2 * 8);
@@ -83,8 +81,6 @@ static void afs_set_contig_bits(union afs_xdr_dir_block *block,
 	block->hdr.bitmap[5] |= (u8)(mask >> 5 * 8);
 	block->hdr.bitmap[6] |= (u8)(mask >> 6 * 8);
 	block->hdr.bitmap[7] |= (u8)(mask >> 7 * 8);
-
-	after = *(u64 *)block->hdr.bitmap;
 }
 
 /*
@@ -93,13 +89,11 @@ static void afs_set_contig_bits(union afs_xdr_dir_block *block,
 static void afs_clear_contig_bits(union afs_xdr_dir_block *block,
 				  int bit, unsigned int nr_slots)
 {
-	u64 mask, before, after;
+	u64 mask;
 
 	mask = (1 << nr_slots) - 1;
 	mask <<= bit;
 
-	before = *(u64 *)block->hdr.bitmap;
-
 	block->hdr.bitmap[0] &= ~(u8)(mask >> 0 * 8);
 	block->hdr.bitmap[1] &= ~(u8)(mask >> 1 * 8);
 	block->hdr.bitmap[2] &= ~(u8)(mask >> 2 * 8);
@@ -108,8 +102,6 @@ static void afs_clear_contig_bits(union afs_xdr_dir_block *block,
 	block->hdr.bitmap[5] &= ~(u8)(mask >> 5 * 8);
 	block->hdr.bitmap[6] &= ~(u8)(mask >> 6 * 8);
 	block->hdr.bitmap[7] &= ~(u8)(mask >> 7 * 8);
-
-	after = *(u64 *)block->hdr.bitmap;
 }
 
 /*
Signature · 1 line
2.20.1
author maintainer
#171 lore raw

PATCH 164/205 bpf: skmsg, fix potential psock NULL pointer dereference

In reply to Sasha Levin

From: John Fastabend <[email protected]>

[ Upstream commit 8163999db445021f2651a8a47b5632483e8722ea ]

Report from Dan Carpenter,

net/core/skmsg.c:792 sk_psock_write_space()
error: we previously assumed 'psock' could be null (see line 790)

net/core/skmsg.c
789 psock = sk_psock(sk);
790 if (likely(psock && sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)))
Check for NULL
791 schedule_work(&psock->work);

   792 write_space = psock->saved_write_space;
                     ^^^^^^^^^^^^^^^^^^^^^^^^

793 rcu_read_unlock();
794 write_space(sk);

Ensure psock dereference on line 792 only occurs if psock is not null.

Fixes 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
---
 net/core/skmsg.c | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)
net/core/skmsg.c +8 -5
index 0675d022584e..ded2d5227678 100644
@@ -793,15 +793,18 @@ static void sk_psock_strp_data_ready(struct sock *sk)
 static void sk_psock_write_space(struct sock *sk)
 {
 	struct sk_psock *psock;
-	void (*write_space)(struct sock *sk);
+	void (*write_space)(struct sock *sk) = NULL;
 
 	rcu_read_lock();
 	psock = sk_psock(sk);
-	if (likely(psock && sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)))
-		schedule_work(&psock->work);
-	write_space = psock->saved_write_space;
+	if (likely(psock)) {
+		if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED))
+			schedule_work(&psock->work);
+		write_space = psock->saved_write_space;
+	}
 	rcu_read_unlock();
-	write_space(sk);
+	if (write_space)
+		write_space(sk);
 }
 
 int sk_psock_init_strp(struct sock *sk, struct sk_psock *psock)
Signature · 1 line
2.20.1
author maintainer
#172 lore raw

PATCH 165/205 SUNRPC: Fix backchannel latency metrics

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit 8729aaba74626c4ebce3abf1b9e96bb62d2958ca ]

I noticed that for callback requests, the reported backlog latency is always zero, and the rtt value is crazy big. The problem was that rqst->rq_xtime is never set for backchannel requests.

Fixes 78215759e20d ("SUNRPC: Make RTT measurement more ... ")
---
 net/sunrpc/xprtrdma/svc_rdma_backchannel.c | 1 +
 net/sunrpc/xprtsock.c                      | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)
net/sunrpc/xprtrdma/svc_rdma_backchannel.c +1 -0
index d1fcc41d5eb5..908e78bb87c6 100644
@@ -195,6 +195,7 @@ rpcrdma_bc_send_request(struct svcxprt_rdma *rdma, struct rpc_rqst *rqst)
 	pr_info("%s: %*ph\n", __func__, 64, rqst->rq_buffer);
 #endif
 
+	rqst->rq_xtime = ktime_get();
 	rc = svc_rdma_bc_sendto(rdma, rqst, ctxt);
 	if (rc) {
 		svc_rdma_send_ctxt_put(rdma, ctxt);
net/sunrpc/xprtsock.c +2 -1
index 70e52f567b2a..5361b98f31ae 100644
@@ -2659,6 +2659,8 @@ static int bc_sendto(struct rpc_rqst *req)
 		.iov_len	= sizeof(marker),
 	};
 
+	req->rq_xtime = ktime_get();
+
 	len = kernel_sendmsg(transport->sock, &msg, &iov, 1, iov.iov_len);
 	if (len != iov.iov_len)
 		return -EAGAIN;
@@ -2684,7 +2686,6 @@ static int bc_send_request(struct rpc_rqst *req)
 	struct svc_xprt	*xprt;
 	int len;
 
-	dprintk("sending request with xid: %08x\n", ntohl(req->rq_xid));
 	/*
 	 * Get the server socket associated with this callback xprt
 	 */
Signature · 1 line
2.20.1
author maintainer
#173 lore raw

PATCH 166/205 scsi: scsi_transport_sas: Fix memory leak when removing devices

In reply to Sasha Levin

From: John Garry <[email protected]>

[ Upstream commit 82ea3e0e129e2ab913dd6684bab7a6e5e9896dee ]

Removing a non-host rphy causes a memory leak:

root@(none)$ echo 0 > /sys/devices/platform/HISI0162:01/host0/port-0:0/expander-0:0/port-0:0:10/phy-0:0:10/sas_phy/phy-0:0:10/enable
[   79.857888] hisi_sas_v2_hw HISI0162:01: dev[7:1] is gone
root@(none)$ echo scan > /sys/kernel/debug/kmemleak
[  131.656603] kmemleak: 3 new suspected memory leaks (see /sys/kernel/debug/kmemleak)
root@(none)$ more /sys/kernel/debug/kmemleak
unreferenced object 0xffff041da5c66000 (size 256):
  comm "kworker/u128:1", pid 549, jiffies 4294898543 (age 113.728s)
  hex dump (first 32 bytes):
    00 5e c6 a5 1d 04 ff ff 01 00 00 00 00 00 00 00  .^..............
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<(____ptrval____)>] kmem_cache_alloc+0x188/0x260
    [<(____ptrval____)>] bsg_setup_queue+0x48/0x1a8
    [<(____ptrval____)>] sas_rphy_add+0x108/0x2d0
    [<(____ptrval____)>] sas_probe_devices+0x168/0x208
    [<(____ptrval____)>] sas_discover_domain+0x660/0x9c8
    [<(____ptrval____)>] process_one_work+0x3f8/0x690
    [<(____ptrval____)>] worker_thread+0x70/0x6a0
    [<(____ptrval____)>] kthread+0x1b8/0x1c0
    [<(____ptrval____)>] ret_from_fork+0x10/0x18
unreferenced object 0xffff041d8c075400 (size 128):
  comm "kworker/u128:1", pid 549, jiffies 4294898543 (age 113.728s)
  hex dump (first 32 bytes):
    00 40 25 97 1d 00 ff ff 00 00 00 00 00 00 00 00  .@%.............
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<(____ptrval____)>] __kmalloc_node+0x1a8/0x2c8
    [<(____ptrval____)>] blk_mq_realloc_tag_set_tags.part.70+0x48/0xd8
    [<(____ptrval____)>] blk_mq_alloc_tag_set+0x1dc/0x530
    [<(____ptrval____)>] bsg_setup_queue+0xe8/0x1a8
    [<(____ptrval____)>] sas_rphy_add+0x108/0x2d0
    [<(____ptrval____)>] sas_probe_devices+0x168/0x208
    [<(____ptrval____)>] sas_discover_domain+0x660/0x9c8
    [<(____ptrval____)>] process_one_work+0x3f8/0x690
    [<(____ptrval____)>] worker_thread+0x70/0x6a0
    [<(____ptrval____)>] kthread+0x1b8/0x1c0
    [<(____ptrval____)>] ret_from_fork+0x10/0x18
unreferenced object 0xffff041da5c65e00 (size 256):
  comm "kworker/u128:1", pid 549, jiffies 4294898543 (age 113.728s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<(____ptrval____)>] __kmalloc_node+0x1a8/0x2c8
    [<(____ptrval____)>] blk_mq_alloc_tag_set+0x254/0x530
    [<(____ptrval____)>] bsg_setup_queue+0xe8/0x1a8
    [<(____ptrval____)>] sas_rphy_add+0x108/0x2d0
    [<(____ptrval____)>] sas_probe_devices+0x168/0x208
    [<(____ptrval____)>] sas_discover_domain+0x660/0x9c8
    [<(____ptrval____)>] process_one_work+0x3f8/0x690
    [<(____ptrval____)>] worker_thread+0x70/0x6a0
    [<(____ptrval____)>] kthread+0x1b8/0x1c0
    [<(____ptrval____)>] ret_from_fork+0x10/0x18
root@(none)$

It turns out that we don't clean up the request queue fully for bsg devices, as the blk mq tags for the request queue are not freed.

Fix by doing the queue removal in one place - in sas_rphy_remove() - instead of unregistering the queue in sas_rphy_remove() and finally cleaning up the queue in calling blk_cleanup_queue() from sas_end_device_release() or sas_expander_release().

Function bsg_remove_queue() can handle a NULL pointer q, so remove the precheck in sas_rphy_remove().

Fixes 651a013649943 ("scsi: scsi_transport_sas: switch to bsg-lib for SMP passthrough")
---
 drivers/scsi/scsi_transport_sas.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)
drivers/scsi/scsi_transport_sas.c +1 -8
index ef138c57e2a6..182fd25c7c43 100644
@@ -1391,9 +1391,6 @@ static void sas_expander_release(struct device *dev)
 	struct sas_rphy *rphy = dev_to_rphy(dev);
 	struct sas_expander_device *edev = rphy_to_expander_device(rphy);
 
-	if (rphy->q)
-		blk_cleanup_queue(rphy->q);
-
 	put_device(dev->parent);
 	kfree(edev);
 }
@@ -1403,9 +1400,6 @@ static void sas_end_device_release(struct device *dev)
 	struct sas_rphy *rphy = dev_to_rphy(dev);
 	struct sas_end_device *edev = rphy_to_end_device(rphy);
 
-	if (rphy->q)
-		blk_cleanup_queue(rphy->q);
-
 	put_device(dev->parent);
 	kfree(edev);
 }
@@ -1634,8 +1628,7 @@ sas_rphy_remove(struct sas_rphy *rphy)
 	}
 
 	sas_rphy_unlink(rphy);
-	if (rphy->q)
-		bsg_unregister_queue(rphy->q);
+	bsg_remove_queue(rphy->q);
 	transport_remove_device(dev);
 	device_del(dev);
 }
Signature · 1 line
2.20.1
author maintainer
#174 lore raw

PATCH 167/205 dmaengine: ti: edma: fix missed failure handling

In reply to Sasha Levin

From: Chuhong Yuan <[email protected]>

[ Upstream commit 340049d453682a9fe8d91fe794dd091730f4bb25 ]

When devm_kcalloc fails, it forgets to call edma_free_slot. Replace direct return with failure handler to fix it.

Fixes 1be5336bc7ba ("dmaengine: edma: New device tree binding")
---
 drivers/dma/ti/edma.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/dma/ti/edma.c +4 -2
index ba7c4f07fcd6..80b780e49971 100644
@@ -2403,8 +2403,10 @@ static int edma_probe(struct platform_device *pdev)
 
 		ecc->tc_list = devm_kcalloc(dev, ecc->num_tc,
 					    sizeof(*ecc->tc_list), GFP_KERNEL);
-		if (!ecc->tc_list)
-			return -ENOMEM;
+		if (!ecc->tc_list) {
+			ret = -ENOMEM;
+			goto err_reg1;
+		}
 
 		for (i = 0;; i++) {
 			ret = of_parse_phandle_with_fixed_args(node, "ti,tptcs",
Signature · 1 line
2.20.1
author maintainer
#175 lore raw

PATCH 168/205 crypto: sun4i-ss - fix big endian issues

In reply to Sasha Levin

From: Corentin Labbe <[email protected]>

[ Upstream commit d1d787bcebfe122a5bd443ae565696661e2e9656 ]

When testing BigEndian kernel, the sun4i-ss was failling all crypto tests.
This patch fix endian issues with it.

Fixes 6298e948215f ("crypto: sunxi-ss - Add Allwinner Security System crypto accelerator")
---
 drivers/crypto/sunxi-ss/sun4i-ss-hash.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)
drivers/crypto/sunxi-ss/sun4i-ss-hash.c +11 -10
index 1369c5fa3087..07df012893bb 100644
@@ -175,7 +175,7 @@ static int sun4i_hash(struct ahash_request *areq)
 	 */
 	unsigned int i = 0, end, fill, min_fill, nwait, nbw = 0, j = 0, todo;
 	unsigned int in_i = 0;
-	u32 spaces, rx_cnt = SS_RX_DEFAULT, bf[32] = {0}, wb = 0, v, ivmode = 0;
+	u32 spaces, rx_cnt = SS_RX_DEFAULT, bf[32] = {0}, v, ivmode = 0;
 	struct sun4i_req_ctx *op = ahash_request_ctx(areq);
 	struct crypto_ahash *tfm = crypto_ahash_reqtfm(areq);
 	struct sun4i_tfm_ctx *tfmctx = crypto_ahash_ctx(tfm);
@@ -184,6 +184,7 @@ static int sun4i_hash(struct ahash_request *areq)
 	struct sg_mapping_iter mi;
 	int in_r, err = 0;
 	size_t copied = 0;
+	__le32 wb = 0;
 
 	dev_dbg(ss->dev, "%s %s bc=%llu len=%u mode=%x wl=%u h0=%0x",
 		__func__, crypto_tfm_alg_name(areq->base.tfm),
@@ -395,7 +396,7 @@ static int sun4i_hash(struct ahash_request *areq)
 
 		nbw = op->len - 4 * nwait;
 		if (nbw) {
-			wb = *(u32 *)(op->buf + nwait * 4);
+			wb = cpu_to_le32(*(u32 *)(op->buf + nwait * 4));
 			wb &= GENMASK((nbw * 8) - 1, 0);
 
 			op->byte_count += nbw;
@@ -404,7 +405,7 @@ static int sun4i_hash(struct ahash_request *areq)
 
 	/* write the remaining bytes of the nbw buffer */
 	wb |= ((1 << 7) << (nbw * 8));
-	bf[j++] = wb;
+	bf[j++] = le32_to_cpu(wb);
 
 	/*
 	 * number of space to pad to obtain 64o minus 8(size) minus 4 (final 1)
@@ -423,13 +424,13 @@ static int sun4i_hash(struct ahash_request *areq)
 
 	/* write the length of data */
 	if (op->mode == SS_OP_SHA1) {
-		__be64 bits = cpu_to_be64(op->byte_count << 3);
-		bf[j++] = lower_32_bits(bits);
-		bf[j++] = upper_32_bits(bits);
+		__be64 *bits = (__be64 *)&bf[j];
+		*bits = cpu_to_be64(op->byte_count << 3);
+		j += 2;
 	} else {
-		__le64 bits = op->byte_count << 3;
-		bf[j++] = lower_32_bits(bits);
-		bf[j++] = upper_32_bits(bits);
+		__le64 *bits = (__le64 *)&bf[j];
+		*bits = cpu_to_le64(op->byte_count << 3);
+		j += 2;
 	}
 	writesl(ss->base + SS_RXFIFO, bf, j);
 
@@ -471,7 +472,7 @@ static int sun4i_hash(struct ahash_request *areq)
 		}
 	} else {
 		for (i = 0; i < 4; i++) {
-			v = readl(ss->base + SS_MD0 + i * 4);
+			v = cpu_to_le32(readl(ss->base + SS_MD0 + i * 4));
 			memcpy(areq->result + i * 4, &v, 4);
 		}
 	}
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel