Re: [PATCH 2/2] drm/msm/dpu: Set input_sel bit for INTF

[Dmitry Baryshkov <dmitry.baryshkov@linaro.org>]

On 01/12/2023 23:29, Abhinav Kumar wrote:
> 
> 
> On 11/30/2023 11:36 PM, Dmitry Baryshkov wrote:
>> On Fri, 1 Dec 2023 at 03:31, Jessica Zhang <quic_jesszhan@quicinc.com> 
>> wrote:
>>>
>>> Set the input_sel bit for encoders as it was missed in the initial
>>> implementation.
>>>
>>> Reported-by: Rob Clark <robdclark@gmail.com>
>>> Fixes: 91143873a05d ("drm/msm/dpu: Add MISR register support for 
>>> interface")
>>> Closes: https://gitlab.freedesktop.org/drm/msm/-/issues/39
>>> Signed-off-by: Jessica Zhang <quic_jesszhan@quicinc.com>
>>> ---
>>>   drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c | 2 +-
>>>   drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c   | 2 +-
>>>   drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c | 7 ++++++-
>>>   drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h | 4 +++-
>>>   4 files changed, 11 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c 
>>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c
>>> index 3442cf65b86f..d0884997ecb7 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_intf.c
>>> @@ -320,7 +320,7 @@ static u32 dpu_hw_intf_get_line_count(struct 
>>> dpu_hw_intf *intf)
>>>
>>>   static void dpu_hw_intf_setup_misr(struct dpu_hw_intf *intf)
>>>   {
>>> -       dpu_hw_setup_misr(&intf->hw, INTF_MISR_CTRL);
>>> +       dpu_hw_setup_misr(&intf->hw, INTF_MISR_CTRL, true);
>>>   }
>>>
>>>   static int dpu_hw_intf_collect_misr(struct dpu_hw_intf *intf, u32 
>>> *misr_value)
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c 
>>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c
>>> index f38473e68f79..77b14107c84a 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_lm.c
>>> @@ -83,7 +83,7 @@ static void dpu_hw_lm_setup_border_color(struct 
>>> dpu_hw_mixer *ctx,
>>>
>>>   static void dpu_hw_lm_setup_misr(struct dpu_hw_mixer *ctx)
>>>   {
>>> -       dpu_hw_setup_misr(&ctx->hw, LM_MISR_CTRL);
>>> +       dpu_hw_setup_misr(&ctx->hw, LM_MISR_CTRL, false);
>>>   }
>>>
>>>   static int dpu_hw_lm_collect_misr(struct dpu_hw_mixer *ctx, u32 
>>> *misr_value)
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c 
>>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c
>>> index a8a0a4e76b94..f441df47fdde 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.c
>>> @@ -481,7 +481,8 @@ void _dpu_hw_setup_qos_lut(struct 
>>> dpu_hw_blk_reg_map *c, u32 offset,
>>>                        cfg->danger_safe_en ? 
>>> QOS_QOS_CTRL_DANGER_SAFE_EN : 0);
>>>   }
>>>
>>> -void dpu_hw_setup_misr(struct dpu_hw_blk_reg_map *c, u32 
>>> misr_ctrl_offset)
>>> +void dpu_hw_setup_misr(struct dpu_hw_blk_reg_map *c, u32 
>>> misr_ctrl_offset,
>>> +               bool set_input_sel)
>>>   {
>>>          u32 config = 0;
>>>
>>> @@ -491,6 +492,10 @@ void dpu_hw_setup_misr(struct dpu_hw_blk_reg_map 
>>> *c, u32 misr_ctrl_offset)
>>>          wmb();
>>>
>>>          config = MISR_FRAME_COUNT | MISR_CTRL_ENABLE | 
>>> MISR_CTRL_FREE_RUN_MASK;
>>> +
>>> +       if (set_input_sel)
>>> +               config |= MISR_CTRL_INPUT_SEL;
>>> +
>>>          DPU_REG_WRITE(c, misr_ctrl_offset, config);
>>>   }
>>>
>>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h 
>>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h
>>> index bb496ebe283b..793670d62414 100644
>>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h
>>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_hw_util.h
>>> @@ -17,6 +17,7 @@
>>>   #define MISR_CTRL_ENABLE                BIT(8)
>>>   #define MISR_CTRL_STATUS                BIT(9)
>>>   #define MISR_CTRL_STATUS_CLEAR          BIT(10)
>>> +#define MISR_CTRL_INPUT_SEL             BIT(24)
>>
>> The public apq8916 TRM documents this as a 4-bit field. I think this
>> was followed into the later generations. Can we please document it
>> correctly and use an uint instead of just bool for set_input_sel?
>>
> 
> Can you pls point us to this document you are referring?

I have this link in my bookmarks, which doesn't seem to work no longer:

https://developer.qualcomm.com/download/sd410/snapdragon-410e-technical-reference-manual.pdf?referrer=node/29241

96boards forum has several links and mentions of this doc.

> 
> I was not aware that bit level details are revealed in external 
> documents :)
> 
> Even though its a 4-bit field, it only takes a 0 or 1 as others are 
> undefined.
> 
> Exposing all the bits will only cause more confusion like it did for 
> others thinking that input select is actually configurable when its not.
> 
> I think what we should do is just pass "misr_type" to this API to tell 
> whether its lm misr or intf misr and set BIT(24) based on that.

This would be another simplification. Can we instead just use values 0 
and 1 instead and maybe document that by default everybody should use 0.

> 
> 
>>>   #define MISR_CTRL_FREE_RUN_MASK         BIT(31)
>>>
>>>   /*
>>> @@ -357,7 +358,8 @@ void _dpu_hw_setup_qos_lut(struct 
>>> dpu_hw_blk_reg_map *c, u32 offset,
>>>                             bool qos_8lvl,
>>>                             const struct dpu_hw_qos_cfg *cfg);
>>>
>>> -void dpu_hw_setup_misr(struct dpu_hw_blk_reg_map *c, u32 
>>> misr_ctrl_offset);
>>> +void dpu_hw_setup_misr(struct dpu_hw_blk_reg_map *c, u32 
>>> misr_ctrl_offset,
>>> +                      bool set_input_sel);
>>>
>>>   int dpu_hw_collect_misr(struct dpu_hw_blk_reg_map *c,
>>>                  u32 misr_ctrl_offset,
>>>
>>> -- 
>>> 2.43.0
>>>
>>
>>

-- 
With best wishes
Dmitry