On 03/05/2023 02:19, Abhinav Kumar wrote:
On 5/2/2023 4:15 PM, Dmitry Baryshkov wrote:
On Wed, 3 May 2023 at 02:04, Abhinav Kumar <quic_abhinavk@xxxxxxxxxxx>
wrote:
On 4/30/2023 4:57 PM, Dmitry Baryshkov wrote:
Remove intf_idx and wb_idx fields from struct dpu_encoder_phys and
struct dpu_enc_phys_init_params. Set the hw_intf and hw_wb directly and
use them to get the instance index.
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx>
---
From whatever I can see, this will not affect functionality of intf or
wb and cleans it up well , so I am fine with this. Hence,
Reviewed-by: Abhinav Kumar <quic_abhinavk@xxxxxxxxxxx>
One minor comment/question.
<snipped other parts>
@@ -761,7 +761,7 @@ struct dpu_encoder_phys *dpu_encoder_phys_cmd_init(
struct dpu_encoder_phys_cmd *cmd_enc = NULL;
int ret = 0;
- DPU_DEBUG("intf %d\n", p->intf_idx - INTF_0);
Was it intentional to drop the index in this log?
We don't have p->intf_idx at this point. I think we can use
p->hw_intf->idx instead, I'll fix that for v2.
Yes, I was aware that. In all other logs, intf_idx was replaced with
hw_intf->idx except this one. So I was not sure if it was intentional or
just removed accidentally.
Most likely it was a leftover from the interim patch where I didn't have
p->hw_intf. Initially I kept the code which manually set the
phys->hw_intf (and hw_wb), but then it was just easier to pass that
through the params and let the caller set it.
+ DPU_DEBUG("intf\n");
cmd_enc = kzalloc(sizeof(*cmd_enc), GFP_KERNEL);
--
With best wishes
Dmitry
