On 03/02/2021 14:25, Dmitry Baryshkov wrote:
On Wed, 3 Feb 2021 at 14:53, Bryan O'Donoghue
<bryan.odonoghue@xxxxxxxxxx> wrote:
- Adds some missing indices to the videocc index
- Adds videocc clock tree hirearchy for msv0
- Adds a regulator to power the videocc consistent with downstream
Once done we can move on to enabling these clocks in the DTS and switch on
the venus.
Bryan O'Donoghue (4):
dt-bindings: clock: Add missing SM8250 videoc clock indices
clk: qcom: videocc: Add sm8250 VIDEO_CC_MVS0_DIV_CLK_SRC
clk: qcom: videocc: Add sm8250 VIDEO_CC_MVS0_CLK
clk: qcom: videocc: Add gdsc mmcx-reg supply hook
I remember now the old discussion about these two clocks
(https://lore.kernel.org/linux-arm-msm/160092826778.310579.12225989905897101118@xxxxxxxxxxxxxxxxxxxxxxxxxx/).
Is the message. reported by Jonathan, solved by the mmcx-reg supply?
The stuck problem is a bandwidth problem not a regulator problem,
there's some internal clock dependency we don't have visibility of.
Both of these resolve,
Jonathan had a hack:
https://git.linaro.org/people/bryan.odonoghue/kernel.git/commit/?h=tracking-qcomlt-sm8250-09-11-20%2bvenus-flto-simple-test2&id=d4bea74282d14244127a1e97766b6108ec68e22f
Dikshita proposed this instead, which, we will send out soon:
https://git.linaro.org/people/bryan.odonoghue/kernel.git/commit/?h=tracking-qcomlt-sm8250-venus-integrated-sm8250&id=782202d611761e639af7e44d8344c0c098642b9f
Also are these clocks necessary at all?
Eh well I'm not sure what happens if I try to dump the clocks but, this
is an accurate representation of what is done downstream so..
---
bod
