]> git.karo-electronics.de Git - linux-beck.git/commit
drm/i915: Actually invalidate the TLB for the SandyBridge HW contexts w/a
authorChris Wilson <chris@chris-wilson.co.uk>
Mon, 1 Oct 2012 13:27:04 +0000 (14:27 +0100)
committerDaniel Vetter <daniel.vetter@ffwll.ch>
Tue, 2 Oct 2012 08:28:18 +0000 (10:28 +0200)
commitac82ea2e97a32f9c49d0746874b4cd1d8904d10f
tree33e3ab959431fce936b581424f1b26bc13cbe820
parent3bc2913e2cb74aac8af90b46fa251dfb8d854665
drm/i915: Actually invalidate the TLB for the SandyBridge HW contexts w/a

A side-effect of commit 7d54a904285b6e780291b91a518267bec5591913
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Fri Aug 10 10:18:10 2012 +0100

    drm/i915: Apply post-sync write for pipe control invalidates

was that only a request to emit invalidate flush would result in the
TLB being invalidated (since it requires synchronisation and so incurs a
performance penalty). However, the stated w/a for hardware contexts is
that the TLBs must be invalidated prior to a MI_SET_CONTEXT, yet the w/a
itself did not request the TLBs to be invalidated...

Note this w/a does not prevent the hard system hang I experience when
using hw contexts (with rc6 enabled) on SNB GT1.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drivers/gpu/drm/i915/i915_gem_context.c