<html>
    <head>
      <base href="https://bugs.freedesktop.org/" />
    </head>
    <body>
      <p>
        <div>
            <b><a class="bz_bug_link 
          bz_status_NEEDINFO "
   title="NEEDINFO --- - [SNB, IVB] Page flips hang indefinitely, apparently waiting for mmio writes that never happen"
   href="https://bugs.freedesktop.org/show_bug.cgi?id=75502#c22">Comment # 22</a>
              on <a class="bz_bug_link 
          bz_status_NEEDINFO "
   title="NEEDINFO --- - [SNB, IVB] Page flips hang indefinitely, apparently waiting for mmio writes that never happen"
   href="https://bugs.freedesktop.org/show_bug.cgi?id=75502">bug 75502</a>
              from <span class="vcard"><a class="email" href="mailto:simon@farnz.org.uk" title="Simon Farnsworth <simon@farnz.org.uk>"> <span class="fn">Simon Farnsworth</span></a>
</span></b>
        <pre>(In reply to <a href="show_bug.cgi?id=75502#c21">comment #21</a>)
<span class="quote">> To double check the theory that the scanout went awol, can you please try
> with

> diff --git a/drivers/gpu/drm/i915/i915_debugfs.c
> b/drivers/gpu/drm/i915/i915_debugfs.c
> index 94aabfc63c3d..c6b81e59058d 100644
> --- a/drivers/gpu/drm/i915/i915_debugfs.c
> +++ b/drivers/gpu/drm/i915/i915_debugfs.c
> @@ -611,16 +611,28 @@ static int i915_gem_pageflip_info(struct seq_file *m,
> void *data)
>                                 seq_puts(m, "Stall check waiting for page
> flip ioctl, ");
>                         seq_printf(m, "%d prepares\n",
> atomic_read(&work->pending));
>  
> +                       {
> +                               u32 addr;
> +
> +                               if (INTEL_INFO(dev)->gen >= 4)
> +                                       addr = DSPSURF(crtc->plane);
> +                               else
> +                                       addr = DSPADDR(crtc->plane);
> +
> +                               seq_printf(m, "Current scanout address
> 0x%08lx\n", 
> +                                          I915_READ(addr));
> +                       }
> +
>                         if (work->old_fb_obj) {
>                                 struct drm_i915_gem_object *obj =
> work->old_fb_obj;
> -                               seq_printf(m, "Old framebuffer gtt_offset
> 0x%08lx\n",
> +                               seq_printf(m, "Old framebuffer address
> 0x%08lx\n",
>                                            i915_gem_obj_ggtt_offset(obj));
>                         }
>                         if (work->pending_flip_obj) {
>                                 struct drm_i915_gem_object *obj =
> work->pending_flip_obj;
>                                 bool complete;
>  
> -                               seq_printf(m, "New framebuffer gtt_offset
> 0x%08lx\n",
> +                               seq_printf(m, "New framebuffer address
> 0x%08lx\n",
>                                            i915_gem_obj_ggtt_offset(obj));
>  
>                                 if (INTEL_INFO(dev)->gen >= 4) {</span >


I've got a broken machine; with this patch, I see:

Flip queued on pipe A (plane A)
Flip queued on frame 15324, now 3420744
Stall check enabled, 1 prepares
Current scanout address 0x00372000
Old framebuffer address 0x00372000
New framebuffer address 0x00684000
MMIO update completed? 0
No flip due on pipe B (plane B)

I'm going to keep the machine in broken state until you tell me otherwise.</pre>
        </div>
      </p>
      <hr>
      <span>You are receiving this mail because:</span>
      
      <ul>
          <li>You are the QA Contact for the bug.</li>
          <li>You are on the CC list for the bug.</li>
      </ul>
    </body>
</html>