Re: [Intel-gfx] [PATCH] tests/kms_psr_sink_crc: commit before querying mode

[Maarten Lankhorst]

Op 20-09-16 om 02:21 schreef Rodrigo Vivi:
> From: Mika Kuoppala <mika.kuopp...@intel.com>
>
> Commit to a mode before querying it.
>
> Tested-by: Rodrigo Vivi <rodrigo.v...@intel.com>
> References: https://bugs.freedesktop.org/show_bug.cgi?id=97172
> Cc: Maarten Lankhorst <maarten.lankho...@linux.intel.com>
> Signed-off-by: Mika Kuoppala <mika.kuopp...@intel.com>
> Signed-off-by: Rodrigo Vivi <rodrigo.v...@intel.com>
> ---
>  tests/kms_psr_sink_crc.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/tests/kms_psr_sink_crc.c b/tests/kms_psr_sink_crc.c
> index 8aafedb..68310d8 100644
> --- a/tests/kms_psr_sink_crc.c
> +++ b/tests/kms_psr_sink_crc.c
> @@ -112,6 +112,14 @@ static void setup_output(data_t *data)
>                       continue;
>  
>               igt_output_set_pipe(output, pipe);
> +             igt_display_commit(display);
> +
> +             if (!output->valid) {
> +                     igt_output_set_pipe(output, PIPE_ANY);
> +                     igt_display_commit(display);
> +                     continue;
> +             }

Could you be more verbose? I'm trying to get rid of this requirement, in 
particular the output->valid branch will never happen here.
To be compatible with atomic, igt_display_commit should not be called until the 
setup is completed, with fb's set etc.

Quick hack below, does this also fix the issue?

diff --git a/lib/igt_kms.c b/lib/igt_kms.c
index 7d53360359ef..86884466c5c4 100644
--- a/lib/igt_kms.c
+++ b/lib/igt_kms.c
@@ -1215,13 +1215,16 @@ static void igt_display_log_shift(igt_display_t 
*display, int shift)
        igt_assert(display->log_shift >= 0);
 }
 
-static void igt_output_refresh(igt_output_t *output)
+static void igt_output_refresh(igt_output_t *output, bool final)
 {
        igt_display_t *display = output->display;
        unsigned long crtc_idx_mask;
 
+       crtc_idx_mask = output->pending_crtc_idx_mask;
+
        /* we mask out the pipes already in use */
-       crtc_idx_mask = output->pending_crtc_idx_mask & ~display->pipes_in_use;
+       if (final)
+               crtc_idx_mask &= ~display->pipes_in_use;
 
        kmstest_free_connector_config(&output->config);
 
@@ -1504,7 +1507,7 @@ void igt_display_init(igt_display_t *display, int drm_fd)
                output->id = resources->connectors[i];
                output->display = display;
 
-               igt_output_refresh(output);
+               igt_output_refresh(output, false);
 
                output->config.pipe_changed = true;
        }
@@ -1591,7 +1594,7 @@ static void igt_display_refresh(igt_display_t *display)
        for (i = 0; i < display->n_outputs; i++) {
                igt_output_t *output = &display->outputs[i];
 
-               igt_output_refresh(output);
+               igt_output_refresh(output, true);
        }
 }
 
@@ -2462,6 +2465,8 @@ void igt_output_set_pipe(igt_output_t *output, enum pipe 
pipe)
 
        if (pipe != output->config.pipe)
                output->config.pipe_changed = true;
+
+       igt_output_refresh(output, false);
 }
 
 void igt_output_set_scaling_mode(igt_output_t *output, uint64_t scaling_mode)



_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx