Created attachment 52457 [details]
dmesg.txt is the bug's dmesg information

So this is regression caused by 3-pipe code.

I can't reproduce this on my current ivb-3-pipe branch using DP and VGA.  Can you confirm?

(In reply to comment #3)
> I can't reproduce this on my current ivb-3-pipe branch using DP and VGA.  Can
> you confirm?

I have try the current ivb-3-pipe branch ,and can't reproduce this issue.

Hi Jesse,

I found the following commit on your ivb-3-pipe branch could fix this issue. But itn't on the upstream and you didn't sent it to the mail list.


commit ebf9f228b2be88635c90549e9023e9a385360118
Author: Jesse Barnes <jbarnes@virtuousgeek.org>
Date:   Thu Oct 20 14:55:49 2011 -0700

    drm/fb: don't try to pan if the fb hasn't moved relative to the CRTC

    This just causes unnecessary set_config calls.

    Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>

diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index f7c6854..4dd81f3 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -739,7 +739,8 @@ int drm_fb_helper_pan_display(struct fb_var_screeninfo *var,
        modeset->x = var->xoffset;
        modeset->y = var->yoffset;

-       if (modeset->num_connectors) {
+       if (((modeset->x != crtc->x) || (modeset->y != crtc->y)) &&
+           modeset->num_connectors) {
            ret = crtc->funcs->set_config(modeset);
            if (!ret) {
                info->var.xoffset = var->xoffset;

Hm that patch shouldn't have affected functionality, just performance a little...  current drm-intel-next still has problems w/o the change?

(In reply to comment #6)
> Hm that patch shouldn't have affected functionality, just performance a
> little...  current drm-intel-next still has problems w/o the change?
I have try the latest drm-intel-next,and can't reproduce this issue.

Ok good, thanks for confirming.

Ok, let's close this bug. The issue doesn't exist on the latest -next branch.

Closing verified+fixed.