Thread (10 messages) 10 messages, 4 authors, 2025-10-31

Re: [PATCH v11 1/2] dt-bindings: memory: mediatek: Add SMI reset and clamp for MT8188

From: Krzysztof Kozlowski <krzk@kernel.org>
Date: 2025-10-31 12:07:56
Also in: linux-devicetree, linux-mediatek, lkml

On 31/10/2025 07:10, Friday Yang (杨阳) wrote:
On Sat, 2025-10-18 at 18:42 +0200, Krzysztof Kozlowski wrote:
quoted
External email : Please do not click links or open attachments until
you have verified the sender or the content.


On 17/09/2025 14:07, Friday Yang wrote:
quoted
Add 'resets' and 'reset-names' properties for SMI LARBs to support
SMI reset operations.
Not informative...
quoted
diff --git a/Documentation/devicetree/bindings/memory-
controllers/mediatek,smi-larb.yaml
b/Documentation/devicetree/bindings/memory-
controllers/mediatek,smi-larb.yaml
index 2e7fac4b5094..fc5feb2eac1f 100644
--- a/Documentation/devicetree/bindings/memory-
controllers/mediatek,smi-larb.yaml
+++ b/Documentation/devicetree/bindings/memory-
controllers/mediatek,smi-larb.yaml
@@ -70,6 +70,12 @@ properties:
     description: the hardware id of this larb. It's only required
when this
       hardware id is not consecutive from its M4U point of view.

+  resets:
+    maxItems: 1
+
+  reset-names:
+    const: larb
Is the reset valid for all existing devices as well? Commit msg does
not
explain that... it is pretty useless - you say what you did. We see
that
from the diff. Explain something not obvious.
Thanks, I could add more descriptions here, like below:

On the MediaTek MT8188 SoC, bus glitches may occur during MTCMOS
on/off transitions. To prevent these glitches from causing errors,
SMI requires clamp and reset operations. This issue specifically
affects the image, camera, and IPE subsystems. 
This explains why you need it. But I asked more questions than only "whY".

Read the question:
"Is the reset valid for all existing devices as well?"
Where is the answer for that? I still do not know...

Do not reply to only pieces of review.
This change adds the
'resets' and 'reset-names' properties to the SMI LARBs within
these subsystems to support the necessary reset operations.
Read again my message:
" it is pretty useless - you say what you did"
That's the same, drop.

So again you just cared to implement only piece of review.

Best regards,
Krzysztof
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help