diff mbox series

[v2,1/4] dt-bindings: firmware: qcom,scm: Document that SCM can be dma-coherent

Message ID 20230616081440.v2.1.Ie79b5f0ed45739695c9970df121e11d724909157@changeid
State New
Headers
Series [v2,1/4] dt-bindings: firmware: qcom,scm: Document that SCM can be dma-coherent |

Commit Message

Doug Anderson June 16, 2023, 3:14 p.m. UTC 
  Trogdor devices use firmware backed by TF-A instead of Qualcomm's
normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.

Let's allow devices like trogdor to be described properly by allowing
"dma-coherent" in the SCM node.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
---

Changes in v2:
- Bindings change new for v2.

 Documentation/devicetree/bindings/firmware/qcom,scm.yaml | 2 ++
 1 file changed, 2 insertions(+)

Comments

Krzysztof Kozlowski June 18, 2023, 8:07 a.m. UTC | #1 
On 16/06/2023 17:14, Douglas Anderson wrote:
> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
> 
> Let's allow devices like trogdor to be described properly by allowing
> "dma-coherent" in the SCM node.
> 
> Signed-off-by: Douglas Anderson <dianders@chromium.org

2

Best regards,
Krzysztof
Konrad Dybcio June 19, 2023, 10 a.m. UTC | #2 
On 18.06.2023 10:07, Krzysztof Kozlowski wrote:
> On 16/06/2023 17:14, Douglas Anderson wrote:
>> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
>> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
>> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
>> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
>> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
>>
>> Let's allow devices like trogdor to be described properly by allowing
>> "dma-coherent" in the SCM node.
>>
>> Signed-off-by: Douglas Anderson <dianders@chromium.org
> 
> 2
Forgot to press alt or something

Konrad
> 
> Best regards,
> Krzysztof
>
Krzysztof Kozlowski June 19, 2023, 10:01 a.m. UTC | #3 
On 16/06/2023 17:14, Douglas Anderson wrote:
> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
> 
> Let's allow devices like trogdor to be described properly by allowing
> "dma-coherent" in the SCM node.
> 
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---


Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>

Best regards,
Krzysztof
Krzysztof Kozlowski June 19, 2023, 10:01 a.m. UTC | #4 
On 19/06/2023 12:00, Konrad Dybcio wrote:
> On 18.06.2023 10:07, Krzysztof Kozlowski wrote:
>> On 16/06/2023 17:14, Douglas Anderson wrote:
>>> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
>>> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
>>> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
>>> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
>>> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
>>>
>>> Let's allow devices like trogdor to be described properly by allowing
>>> "dma-coherent" in the SCM node.
>>>
>>> Signed-off-by: Douglas Anderson <dianders@chromium.org
>>
>> 2
> Forgot to press alt or something

D'oh!

Best regards,
Krzysztof
Doug Anderson June 21, 2023, 4:56 p.m. UTC | #5 
Bjorn,

On Fri, Jun 16, 2023 at 8:18 AM Douglas Anderson <dianders@chromium.org> wrote:
>
> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
>
> Let's allow devices like trogdor to be described properly by allowing
> "dma-coherent" in the SCM node.
>
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---
>
> Changes in v2:
> - Bindings change new for v2.
>
>  Documentation/devicetree/bindings/firmware/qcom,scm.yaml | 2 ++
>  1 file changed, 2 insertions(+)

Without this series v6.4 will have a regression where WiFi / LTE won't
work at all on trogdor devices. Any chance you can send up a "Fixes"
pull request with the 4 patches in it? ...or I could try to convince
someone on the SoC tree to land them directly?

Thanks!

-Doug
Bjorn Andersson June 22, 2023, 8:38 p.m. UTC | #6 
On Fri, 16 Jun 2023 08:14:38 -0700, Douglas Anderson wrote:
> Trogdor devices use firmware backed by TF-A instead of Qualcomm's
> normal TZ. On TF-A we end up mapping memory as cacheable. Specifically,
> you can see in Trogdor's TF-A code [1] in qti_sip_mem_assign() that we
> call qti_mmap_add_dynamic_region() with MT_RO_DATA. This translates
> down to MT_MEMORY instead of MT_NON_CACHEABLE or MT_DEVICE.
> 
> Let's allow devices like trogdor to be described properly by allowing
> "dma-coherent" in the SCM node.
> 
> [...]

Applied, thanks!

[1/4] dt-bindings: firmware: qcom,scm: Document that SCM can be dma-coherent
      commit: c0877829ada0406233aee5bd54f6813db79d5f1f
[2/4] arm64: dts: qcom: sc7180: Mark SCM as dma-coherent for IDP
      commit: 9a5f0b11e49e27f0a01a73c31d05df4a95bea3fa
[3/4] arm64: dts: qcom: sc7180: Mark SCM as dma-coherent for trogdor
      commit: a54b7fa6b9ab6b4ecb7d9aba6b1a0ce1bcc961e3
[4/4] arm64: dts: qcom: sc7280: Mark SCM as dma-coherent for chrome devices
      commit: 7b59e8ae92fe089fed8ff1b23e53442ae5b204c9

Best regards,
diff mbox series

Patch

diff --git a/Documentation/devicetree/bindings/firmware/qcom,scm.yaml b/Documentation/devicetree/bindings/firmware/qcom,scm.yaml
index 367d04ad1923..83381f3a1341 100644
--- a/Documentation/devicetree/bindings/firmware/qcom,scm.yaml
+++ b/Documentation/devicetree/bindings/firmware/qcom,scm.yaml
@@ -71,6 +71,8 @@  properties:
     minItems: 1
     maxItems: 3
 
+  dma-coherent: true
+
   interconnects:
     maxItems: 1