[PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p

bibek.patro@oss.qualcomm.com posted 1 patch 2 weeks, 4 days ago
drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 2 ++
1 file changed, 2 insertions(+)
[PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
Posted by bibek.patro@oss.qualcomm.com 2 weeks, 4 days ago
From: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>

The ACTLR configuration for the sa8775p MDSS client was inadvertently
dropped while reworking the commit f91879fdf70b ("iommu/arm-smmu-qcom:
Add actlr settings for mdss on Qualcomm platforms"). Without this
entry, the sa8775p MDSS block does not receive the intended default
ACTLR configuration.

Restore the missing compatible entry so that the platform receives the
expected behavior.

Signed-off-by: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
---

Hello community,

Resending this patch as the earlier version was sent as an
in‑reply‑to within an ongoing discussion thread [1], instead of being sent
as a new standalone patch submission. This may have caused it to inherit
the same message-id and appear threaded incorrectly.

Please consider this as the correct patch submission.

[1]:
https://lore.kernel.org/all/20260109204551.845959-1-bibek.patro@oss.qualcomm.com/

Thanks & regards,
Bibek

 drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c b/drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c
index 456d5146831e..718d102356d9 100644
--- a/drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c
+++ b/drivers/iommu/arm/arm-smmu/arm-smmu-qcom.c
@@ -43,6 +43,8 @@ static const struct of_device_id qcom_smmu_actlr_client_of_match[] = {
 			.data = (const void *) (PREFETCH_DEEP | CPRE | CMTLB) },
 	{ .compatible = "qcom,qcm2290-mdss",
 			.data = (const void *) (PREFETCH_SHALLOW | CPRE | CMTLB) },
+	{ .compatible = "qcom,sa8775p-mdss",
+			.data = (const void *) (PREFETCH_DEFAULT | CMTLB) },
 	{ .compatible = "qcom,sc7280-mdss",
 			.data = (const void *) (PREFETCH_SHALLOW | CPRE | CMTLB) },
 	{ .compatible = "qcom,sc7280-venus",
--
2.34.1

Re: [PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
Posted by Will Deacon 2 weeks, 2 days ago
On Tue, 20 Jan 2026 20:41:05 +0530, bibek.patro@oss.qualcomm.com wrote:
> The ACTLR configuration for the sa8775p MDSS client was inadvertently
> dropped while reworking the commit f91879fdf70b ("iommu/arm-smmu-qcom:
> Add actlr settings for mdss on Qualcomm platforms"). Without this
> entry, the sa8775p MDSS block does not receive the intended default
> ACTLR configuration.
> 
> Restore the missing compatible entry so that the platform receives the
> expected behavior.
> 
> [...]

Applied to iommu (arm/smmu/updates), thanks!

[1/1] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
      https://git.kernel.org/iommu/c/14e9a138dd02

Cheers,
-- 
Will

https://fixes.arm64.dev
https://next.arm64.dev
https://will.arm64.dev
Re: [PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
Posted by Dmitry Baryshkov 2 weeks, 3 days ago
On Tue, Jan 20, 2026 at 08:41:05PM +0530, bibek.patro@oss.qualcomm.com wrote:
> From: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
> 
> The ACTLR configuration for the sa8775p MDSS client was inadvertently
> dropped while reworking the commit f91879fdf70b ("iommu/arm-smmu-qcom:
> Add actlr settings for mdss on Qualcomm platforms"). Without this
> entry, the sa8775p MDSS block does not receive the intended default
> ACTLR configuration.
> 
> Restore the missing compatible entry so that the platform receives the
> expected behavior.
> 
> Signed-off-by: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>

Missing Fixes tag. Otherwise LGTM

-- 
With best wishes
Dmitry
Re: [PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
Posted by Bibek Kumar Patro 2 weeks, 2 days ago

On 1/21/2026 3:56 AM, Dmitry Baryshkov wrote:
> On Tue, Jan 20, 2026 at 08:41:05PM +0530, bibek.patro@oss.qualcomm.com wrote:
>> From: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
>>
>> The ACTLR configuration for the sa8775p MDSS client was inadvertently
>> dropped while reworking the commit f91879fdf70b ("iommu/arm-smmu-qcom:
>> Add actlr settings for mdss on Qualcomm platforms"). Without this
>> entry, the sa8775p MDSS block does not receive the intended default
>> ACTLR configuration.
>>
>> Restore the missing compatible entry so that the platform receives the
>> expected behavior.
>>
>> Signed-off-by: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
> 
> Missing Fixes tag. Otherwise LGTM

It looks like [1] hasn’t landed on any release branch yet - it’s still
only present in Will’s tree, which is also where the referenced “fixed
‘commit f91879fdf70b’” SHA was taken from.
Given that, do we need a Fixes: tag in this case? and should it use the
same SHA from Will’s tree.

[1]:
https://lore.kernel.org/all/176764799410.1451727.17568755330994660848.b4-ty@kernel.org/

Thanks & regards,
Bibek

> 

Re: [PATCH] iommu/arm-smmu-qcom: Restore ACTLR settings for MDSS on sa8775p
Posted by Will Deacon 2 weeks, 2 days ago
On Thu, Jan 22, 2026 at 04:52:10PM +0530, Bibek Kumar Patro wrote:
> 
> 
> On 1/21/2026 3:56 AM, Dmitry Baryshkov wrote:
> > On Tue, Jan 20, 2026 at 08:41:05PM +0530, bibek.patro@oss.qualcomm.com wrote:
> > > From: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
> > > 
> > > The ACTLR configuration for the sa8775p MDSS client was inadvertently
> > > dropped while reworking the commit f91879fdf70b ("iommu/arm-smmu-qcom:
> > > Add actlr settings for mdss on Qualcomm platforms"). Without this
> > > entry, the sa8775p MDSS block does not receive the intended default
> > > ACTLR configuration.
> > > 
> > > Restore the missing compatible entry so that the platform receives the
> > > expected behavior.
> > > 
> > > Signed-off-by: Bibek Kumar Patro <bibek.patro@oss.qualcomm.com>
> > 
> > Missing Fixes tag. Otherwise LGTM
> 
> It looks like [1] hasn’t landed on any release branch yet - it’s still
> only present in Will’s tree, which is also where the referenced “fixed
> ‘commit f91879fdf70b’” SHA was taken from.
> Given that, do we need a Fixes: tag in this case? and should it use the
> same SHA from Will’s tree.

I'll add it when I apply, no worries. Still playing catch-up with patches
:(

Will