[PATCH 6/8] drm/i915/scaler: Pimp scaler debugs
Jani Nikula
jani.nikula at linux.intel.com
Tue Dec 31 13:47:13 UTC 2024
On Tue, 31 Dec 2024, Gustavo Sousa <gustavo.sousa at intel.com> wrote:
> Quoting Jani Nikula (2024-12-30 09:58:31-03:00)
>>On Tue, 24 Dec 2024, Gustavo Sousa <gustavo.sousa at intel.com> wrote:
>>> Quoting Ville Syrjala (2024-12-19 10:08:25-03:00)
>>>>From: Ville Syrjälä <ville.syrjala at linux.intel.com>
>>>>
>>>>Include the standard "[CRTC:...]" information in the scaler debugs
>>>>to make life easier.
>>>
>>> Drive-by comment (and going a bit off-topic):
>>>
>>> $ git grep '\[[A-Z]\+:%d:%s]' -- drivers/gpu/drm | wc -l
>>> 600
>>>
>>> Has someone already considered creating DRM_*_FMT and DRM_*_ARG() for
>>> those? E.g. DRM_CRTC_FMT and CRM_CRTC_ARG(crtc->base), which, IMO, would
>>> be easier to use and arguably more concise.
>>
>>Personally, I'm not in favour. I dislike having to concatenate the
>>string constants all over the place.
>>
>>> I tried doing a quick search on lore.kernel.org/dri-devel, but I'm not
>>> sure what would be good search terms to find previous attempts to see
>>> possible arguments against it.
>>
>>We've gone through this a number of times, and some of the other options
>>are:
>
> Thanks for the recap!
>
>>
>>- Add allocated debug string to the objects, e.g. crtc->debug or
>> crtc->id and print it with %s. It also works when you want to print
>> the info of e.g. both the connector and the encoder.
>
> This one seems interesting, although there would be some increased
> memory usage. Maybe that's not too bad?
>
> Was the increase in memory usage what caused it not to be adopted?
No, we just never made a decision, nor did anyone post patches for any
of the alternatives as far as I remember.
BR,
Jani.
>
> --
> Gustavo Sousa
>
>>
>>- Add function pointers for debug logging in the drm objects, and have
>> the drm_dbg* family of functions use them based on the type passed to
>> it with generics. You'd do drm_dbg_kms(connector, "foo\n"); and that
>> would add the [CONNECTOR:...] prefix. Falls short when you want to
>> print the info of multiple objects.
>>
>>- Object specific debug logging macros. connector_dbg() etc. I'm
>> strongly against the proliferation of logging macros. All the variants
>> (like once, ratelimited, etc.) get multiplied by the number of object
>> types. (Yes, I also dislike the gt/guc macros, but I digress.) And
>> this also doesn't solve the logging of multiple objects at once.
>>
>>- Add printk format specifiers. There's apparently no way to do this in
>> a human readable way, as anything nice throws off compiler printf
>> format checks. So you end up with stuff like %pXYZ where the XYZ are
>> meaningless magic letters. And they'd need to be implemented in kernel
>> core.
>>
>>
>>BR,
>>Jani.
>>
>>
>>>
>>> --
>>> Gustavo Sousa
>>>
>>>>
>>>>Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
>>>>---
>>>> drivers/gpu/drm/i915/display/skl_scaler.c | 25 +++++++++++++++--------
>>>> 1 file changed, 16 insertions(+), 9 deletions(-)
>>>>
>>>>diff --git a/drivers/gpu/drm/i915/display/skl_scaler.c b/drivers/gpu/drm/i915/display/skl_scaler.c
>>>>index cbc71e44fcbb..f6d76ef1a854 100644
>>>>--- a/drivers/gpu/drm/i915/display/skl_scaler.c
>>>>+++ b/drivers/gpu/drm/i915/display/skl_scaler.c
>>>>@@ -166,7 +166,8 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
>>>> if (DISPLAY_VER(display) >= 9 && crtc_state->hw.enable &&
>>>> need_scaler && adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) {
>>>> drm_dbg_kms(display->drm,
>>>>- "Pipe/Plane scaling not supported with IF-ID mode\n");
>>>>+ "[CRTC:%d:%s] scaling not supported with IF-ID mode\n",
>>>>+ crtc->base.base.id, crtc->base.name);
>>>> return -EINVAL;
>>>> }
>>>>
>>>>@@ -186,8 +187,9 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
>>>> scaler_state->scalers[*scaler_id].in_use = false;
>>>>
>>>> drm_dbg_kms(display->drm,
>>>>- "scaler_user index %u.%u: "
>>>>+ "[CRTC:%d:%s] scaler_user index %u.%u: "
>>>> "Staged freeing scaler id %d scaler_users = 0x%x\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> crtc->pipe, scaler_user, *scaler_id,
>>>> scaler_state->scaler_users);
>>>> *scaler_id = -1;
>>>>@@ -207,8 +209,9 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
>>>> src_w > max_src_w || src_h > max_src_h ||
>>>> dst_w > max_dst_w || dst_h > max_dst_h) {
>>>> drm_dbg_kms(display->drm,
>>>>- "scaler_user index %u.%u: src %ux%u dst %ux%u "
>>>>+ "[CRTC:%d:%s] scaler_user index %u.%u: src %ux%u dst %ux%u "
>>>> "size is out of scaler range\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> crtc->pipe, scaler_user, src_w, src_h,
>>>> dst_w, dst_h);
>>>> return -EINVAL;
>>>>@@ -224,16 +227,18 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
>>>> */
>>>> if (pipe_src_w > max_dst_w || pipe_src_h > max_dst_h) {
>>>> drm_dbg_kms(display->drm,
>>>>- "scaler_user index %u.%u: pipe src size %ux%u "
>>>>+ "[CRTC:%d:%s] scaler_user index %u.%u: pipe src size %ux%u "
>>>> "is out of scaler range\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> crtc->pipe, scaler_user, pipe_src_w, pipe_src_h);
>>>> return -EINVAL;
>>>> }
>>>>
>>>> /* mark this plane as a scaler user in crtc_state */
>>>> scaler_state->scaler_users |= (1 << scaler_user);
>>>>- drm_dbg_kms(display->drm, "scaler_user index %u.%u: "
>>>>+ drm_dbg_kms(display->drm, "[CRTC:%d:%s] scaler_user index %u.%u: "
>>>> "staged scaling request for %ux%u->%ux%u scaler_users = 0x%x\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> crtc->pipe, scaler_user, src_w, src_h, dst_w, dst_h,
>>>> scaler_state->scaler_users);
>>>>
>>>>@@ -421,8 +426,8 @@ static int intel_atomic_setup_scaler(struct intel_crtc_scaler_state *scaler_stat
>>>>
>>>> if (hscale < 0 || vscale < 0) {
>>>> drm_dbg_kms(display->drm,
>>>>- "Scaler %d doesn't support required plane scaling\n",
>>>>- *scaler_id);
>>>>+ "[CRTC:%d:%s] scaler %d doesn't support required plane scaling\n",
>>>>+ crtc->base.base.id, crtc->base.name, *scaler_id);
>>>> drm_rect_debug_print("src: ", src, true);
>>>> drm_rect_debug_print("dst: ", dst, false);
>>>>
>>>>@@ -430,7 +435,8 @@ static int intel_atomic_setup_scaler(struct intel_crtc_scaler_state *scaler_stat
>>>> }
>>>> }
>>>>
>>>>- drm_dbg_kms(display->drm, "Attached scaler id %u.%u to %s:%d\n",
>>>>+ drm_dbg_kms(display->drm, "[CRTC:%d:%s] attached scaler id %u.%u to %s:%d\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> crtc->pipe, *scaler_id, name, idx);
>>>> scaler_state->scalers[*scaler_id].mode = mode;
>>>>
>>>>@@ -530,7 +536,8 @@ int intel_atomic_setup_scalers(struct intel_atomic_state *state,
>>>> /* fail if required scalers > available scalers */
>>>> if (num_scalers_need > crtc->num_scalers) {
>>>> drm_dbg_kms(display->drm,
>>>>- "Too many scaling requests %d > %d\n",
>>>>+ "[CRTC:%d:%s] too many scaling requests %d > %d\n",
>>>>+ crtc->base.base.id, crtc->base.name,
>>>> num_scalers_need, crtc->num_scalers);
>>>> return -EINVAL;
>>>> }
>>>>--
>>>>2.45.2
>>>>
>>
>>--
>>Jani Nikula, Intel
--
Jani Nikula, Intel
More information about the Intel-gfx
mailing list