2022-04-06 18:34:24

by José Expósito

[permalink] [raw]
Subject: [PATCH 1/1] drm/vc4: hdmi: Replace drm_detect_hdmi_monitor() with is_hdmi

Once EDID is parsed, the monitor HDMI support information is cached in
drm_display_info.is_hdmi by drm_parse_hdmi_vsdb_video().

This driver calls drm_detect_hdmi_monitor() to receive the same
information and stores its own cached value, which is less efficient.

Avoid calling drm_detect_hdmi_monitor() and use drm_display_info.is_hdmi
instead.

drm_detect_hdmi_monitor() is called in vc4_hdmi_connector_detect() and
vc4_hdmi_connector_get_modes(). In both cases it is safe to rely on
drm_display_info.is_hdmi as shown by ftrace:

vc4_hdmi_connector_detect:

vc4_hdmi_connector_detect() {
drm_get_edid() {
drm_connector_update_edid_property() {
drm_add_display_info() {
drm_reset_display_info();
drm_for_each_detailed_block.part.0();
drm_parse_cea_ext() {
drm_find_cea_extension();
cea_db_offsets.part.0();
cea_db_is_hdmi_vsdb.part.0();
drm_parse_hdmi_vsdb_video();
/* drm_display_info.is_hdmi is cached here */
}
}
}
}
/* drm_display_info.is_hdmi is used here */
}

vc4_hdmi_connector_get_modes:

vc4_hdmi_connector_get_modes() {
drm_get_edid() {
drm_connector_update_edid_property() {
drm_add_display_info() {
drm_reset_display_info();
drm_for_each_detailed_block.part.0();
drm_parse_cea_ext() {
drm_find_cea_extension();
cea_db_offsets.part.0();
cea_db_is_hdmi_vsdb.part.0();
drm_parse_hdmi_vsdb_video();
/* drm_display_info.is_hdmi is cached here */
}
}
}
}
/* drm_display_info.is_hdmi is used here */
drm_connector_update_edid_property();
}

Signed-off-by: José Expósito <[email protected]>
---
drivers/gpu/drm/vc4/vc4_hdmi.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c
index 6c58b0fd13fb..ecdb1ffc2023 100644
--- a/drivers/gpu/drm/vc4/vc4_hdmi.c
+++ b/drivers/gpu/drm/vc4/vc4_hdmi.c
@@ -216,7 +216,7 @@ vc4_hdmi_connector_detect(struct drm_connector *connector, bool force)

if (edid) {
cec_s_phys_addr_from_edid(vc4_hdmi->cec_adap, edid);
- vc4_hdmi->encoder.hdmi_monitor = drm_detect_hdmi_monitor(edid);
+ vc4_hdmi->encoder.hdmi_monitor = connector->display_info.is_hdmi;
kfree(edid);
}
}
@@ -255,7 +255,7 @@ static int vc4_hdmi_connector_get_modes(struct drm_connector *connector)
goto out;
}

- vc4_encoder->hdmi_monitor = drm_detect_hdmi_monitor(edid);
+ vc4_encoder->hdmi_monitor = connector->display_info.is_hdmi;

drm_connector_update_edid_property(connector, edid);
ret = drm_add_edid_modes(connector, edid);
--
2.25.1


2022-04-08 08:08:05

by Maxime Ripard

[permalink] [raw]
Subject: Re: [PATCH 1/1] drm/vc4: hdmi: Replace drm_detect_hdmi_monitor() with is_hdmi

Hi Jose,

On Wed, Apr 06, 2022 at 06:55:14PM +0200, Jos? Exp?sito wrote:
> Once EDID is parsed, the monitor HDMI support information is cached in
> drm_display_info.is_hdmi by drm_parse_hdmi_vsdb_video().
>
> This driver calls drm_detect_hdmi_monitor() to receive the same
> information and stores its own cached value, which is less efficient.
>
> Avoid calling drm_detect_hdmi_monitor() and use drm_display_info.is_hdmi
> instead.
>
> drm_detect_hdmi_monitor() is called in vc4_hdmi_connector_detect() and
> vc4_hdmi_connector_get_modes(). In both cases it is safe to rely on
> drm_display_info.is_hdmi as shown by ftrace:

How do you use ftrace to generate that?

> vc4_hdmi_connector_detect:
>
> vc4_hdmi_connector_detect() {
> drm_get_edid() {
> drm_connector_update_edid_property() {
> drm_add_display_info() {
> drm_reset_display_info();
> drm_for_each_detailed_block.part.0();
> drm_parse_cea_ext() {
> drm_find_cea_extension();
> cea_db_offsets.part.0();
> cea_db_is_hdmi_vsdb.part.0();
> drm_parse_hdmi_vsdb_video();
> /* drm_display_info.is_hdmi is cached here */
> }
> }
> }
> }
> /* drm_display_info.is_hdmi is used here */
> }
>
> vc4_hdmi_connector_get_modes:
>
> vc4_hdmi_connector_get_modes() {
> drm_get_edid() {
> drm_connector_update_edid_property() {
> drm_add_display_info() {
> drm_reset_display_info();
> drm_for_each_detailed_block.part.0();
> drm_parse_cea_ext() {
> drm_find_cea_extension();
> cea_db_offsets.part.0();
> cea_db_is_hdmi_vsdb.part.0();
> drm_parse_hdmi_vsdb_video();
> /* drm_display_info.is_hdmi is cached here */
> }
> }
> }
> }
> /* drm_display_info.is_hdmi is used here */
> drm_connector_update_edid_property();
> }
>
> Signed-off-by: Jos? Exp?sito <[email protected]>

I think what you're hinting at in the cover letter would be best though:
we should just remove that caching entirely and use is_hdmi everywhere

Maxime


Attachments:
(No filename) (2.26 kB)
signature.asc (235.00 B)
Download all attachments

2022-04-11 23:42:46

by José Expósito

[permalink] [raw]
Subject: Re: [PATCH 1/1] drm/vc4: hdmi: Replace drm_detect_hdmi_monitor() with is_hdmi

Hi Maxime,

Thanks for your comments.

On Fri, Apr 08, 2022 at 09:41:10AM +0200, Maxime Ripard wrote:
> Hi Jose,
>
> On Wed, Apr 06, 2022 at 06:55:14PM +0200, Jos? Exp?sito wrote:
> > Once EDID is parsed, the monitor HDMI support information is cached in
> > drm_display_info.is_hdmi by drm_parse_hdmi_vsdb_video().
> >
> > This driver calls drm_detect_hdmi_monitor() to receive the same
> > information and stores its own cached value, which is less efficient.
> >
> > Avoid calling drm_detect_hdmi_monitor() and use drm_display_info.is_hdmi
> > instead.
> >
> > drm_detect_hdmi_monitor() is called in vc4_hdmi_connector_detect() and
> > vc4_hdmi_connector_get_modes(). In both cases it is safe to rely on
> > drm_display_info.is_hdmi as shown by ftrace:
>
> How do you use ftrace to generate that?

I had to add noinline to a couple of relevant functions and run the
usual:

$ sudo trace-cmd record -p function_graph -l "vc4_hdmi_*" [...]

I'll add the command to v2.


> > vc4_hdmi_connector_detect:
> >
> > vc4_hdmi_connector_detect() {
> > drm_get_edid() {
> > drm_connector_update_edid_property() {
> > drm_add_display_info() {
> > drm_reset_display_info();
> > drm_for_each_detailed_block.part.0();
> > drm_parse_cea_ext() {
> > drm_find_cea_extension();
> > cea_db_offsets.part.0();
> > cea_db_is_hdmi_vsdb.part.0();
> > drm_parse_hdmi_vsdb_video();
> > /* drm_display_info.is_hdmi is cached here */
> > }
> > }
> > }
> > }
> > /* drm_display_info.is_hdmi is used here */
> > }
> >
> > vc4_hdmi_connector_get_modes:
> >
> > vc4_hdmi_connector_get_modes() {
> > drm_get_edid() {
> > drm_connector_update_edid_property() {
> > drm_add_display_info() {
> > drm_reset_display_info();
> > drm_for_each_detailed_block.part.0();
> > drm_parse_cea_ext() {
> > drm_find_cea_extension();
> > cea_db_offsets.part.0();
> > cea_db_is_hdmi_vsdb.part.0();
> > drm_parse_hdmi_vsdb_video();
> > /* drm_display_info.is_hdmi is cached here */
> > }
> > }
> > }
> > }
> > /* drm_display_info.is_hdmi is used here */
> > drm_connector_update_edid_property();
> > }
> >
> > Signed-off-by: Jos? Exp?sito <[email protected]>
>
> I think what you're hinting at in the cover letter would be best though:
> we should just remove that caching entirely and use is_hdmi everywhere

Cool, I'll work on a follow up patch to remove vc4_encoder.hdmi_monitor
and add it to v2.

Thanks,
Jose