<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - [CNL-Y] GPU hang" href="https://bugs.freedesktop.org/show_bug.cgi?id=106185#c3">Comment # 3</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - [CNL-Y] GPU hang" href="https://bugs.freedesktop.org/show_bug.cgi?id=106185">bug 106185</a> from <a class="email" href="mailto:chris@chris-wilson.co.uk" title="Chris Wilson <chris@chris-wilson.co.uk>"> Chris Wilson</a> <pre>(In reply to Timo Aaltonen from <a href="show_bug.cgi?id=106185#c2">comment #2</a>) > Yes, the system hang is apparently fixed by: > > Author: Mika Kuoppala <<a href="mailto:mika.kuoppala@linux.intel.com">mika.kuoppala@linux.intel.com</a>> > Date: Thu Apr 12 17:58:02 2018 +0300 > > drm/i915/cnl: Use mmio access to context status buffer > Now replaced by commit 77dfedb5be03779f9a5d83e323a1b36e32090105 Author: Chris Wilson <<a href="mailto:chris@chris-wilson.co.uk">chris@chris-wilson.co.uk</a>> Date: Fri May 11 13:11:45 2018 +0100 drm/i915/execlists: Use rmb() to order CSB reads We assume that the CSB is written using the normal ringbuffer coherency protocols, as outlined in kernel/events/ring_buffer.c: * (HW) (DRIVER) * * if (LOAD ->data_tail) { LOAD ->data_head * (A) smp_rmb() (C) * STORE $data LOAD $data * smp_wmb() (B) smp_mb() (D) * STORE ->data_head STORE ->data_tail * } So we assume that the HW fulfils its ordering requirements (B), and so we should use a complimentary rmb (C) to ensure that our read of its WRITE pointer is completed before we start accessing the data. The final mb (D) is implied by the uncached mmio we perform to inform the HW of our READ pointer. References: <a class="bz_bug_link bz_status_CLOSED bz_closed" title="CLOSED FIXED - GPU Hang with single-channel RAM configuration" href="show_bug.cgi?id=105064">https://bugs.freedesktop.org/show_bug.cgi?id=105064</a> References: <a class="bz_bug_link bz_status_CLOSED bz_closed" title="CLOSED WORKSFORME - [CNL] System hang: GEM_BUG_ON(buf[2 * head + 1] != port->context_id) during piglit/gem_ctx_switch, masked by tracing enabled" href="show_bug.cgi?id=105888">https://bugs.freedesktop.org/show_bug.cgi?id=105888</a> References: <a class="bz_bug_link bz_status_NEW " title="NEW - [CNL-Y] GPU hang" href="show_bug.cgi?id=106185">https://bugs.freedesktop.org/show_bug.cgi?id=106185</a> Fixes: 767a983ab255 ("drm/i915/execlists: Read the context-status HEAD from the HWSP") References: 61bf9719fa17 ("drm/i915/cnl: Use mmio access to context status buffer") Suggested-by: Mika Kuoppala <<a href="mailto:mika.kuoppala@linux.intel.com">mika.kuoppala@linux.intel.com</a>> Signed-off-by: Chris Wilson <<a href="mailto:chris@chris-wilson.co.uk">chris@chris-wilson.co.uk</a>> Cc: Mika Kuoppala <<a href="mailto:mika.kuoppala@linux.intel.com">mika.kuoppala@linux.intel.com</a>> Cc: Joonas Lahtinen <<a href="mailto:joonas.lahtinen@linux.intel.com">joonas.lahtinen@linux.intel.com</a>> Cc: Tvrtko Ursulin <<a href="mailto:tvrtko.ursulin@intel.com">tvrtko.ursulin@intel.com</a>> Cc: Michał Winiarski <<a href="mailto:michal.winiarski@intel.com">michal.winiarski@intel.com</a>> Cc: Rafael Antognolli <<a href="mailto:rafael.antognolli@intel.com">rafael.antognolli@intel.com</a>> Cc: Michel Thierry <<a href="mailto:michel.thierry@intel.com">michel.thierry@intel.com</a>> Cc: Timo Aaltonen <<a href="mailto:tjaalton@ubuntu.com">tjaalton@ubuntu.com</a>> Tested-by: Timo Aaltonen <<a href="mailto:tjaalton@ubuntu.com">tjaalton@ubuntu.com</a>> Acked-by: Michel Thierry <<a href="mailto:michel.thierry@intel.com">michel.thierry@intel.com</a>> Acked-by: Mika Kuoppala <<a href="mailto:mika.kuoppala@linux.intel.com">mika.kuoppala@linux.intel.com</a>> Link: <a href="https://patchwork.freedesktop.org/patch/msgid/20180511121147.31915-1-chris@chris-wilson.co.uk">https://patchwork.freedesktop.org/patch/msgid/20180511121147.31915-1-chris@chris-wilson.co.uk</a></pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> <li>You are the assignee for the bug.</li> <li>You are on the CC list for the bug.</li> </ul> </body> </html>