<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEEDINFO " title="NEEDINFO - Delay in skl_disable_plane() causes a system freeze" href="https://bugs.freedesktop.org/show_bug.cgi?id=104975#c72">Comment # 72</a> on <a class="bz_bug_link bz_status_NEEDINFO " title="NEEDINFO - Delay in skl_disable_plane() causes a system freeze" href="https://bugs.freedesktop.org/show_bug.cgi?id=104975">bug 104975</a> from <a class="email" href="mailto:azhar.shaikh@intel.com" title="Azhar <azhar.shaikh@intel.com>"> Azhar</a> <pre>(In reply to Ville Syrjala from <a href="show_bug.cgi?id=104975#c71">comment #71</a>) > Came up with one more random idea: > git://github.com/vsyrjala/linux.git double_buffer_ctl_simple_2 This branch still had the spinlock issues (introduced with this patch "grab double buffer ctl around the entire update") mentioned in <a href="show_bug.cgi?id=104975#c69">comment #69</a>, have commented the schedule_timeout() in intel_pipe_update_start(), to move forward. 1. adjust the DDB and plane update seqeunces 2. irqsave/restore for skl wm 3. double buffer disable for scalers too 4. grab double buffer ctl around the entire update 5. leave the "allow" bit on, but arm the ddb/wm with a SURF write 6. clear the double buffer allow bits immediately after ddb/wm change 7. drm/i915: Use DOUBLE_BUFFER_CTL and disable planes before frobbing wm/ddb With above patches on top of Linux-stable 4.15.18 without the support of IPC on BXT+, system still crashes.</pre> </div> <hr> You are receiving this mail because: <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>