From: Eric Anholt Subject: Re: [PATCH] Revert "drm/i915: Kill GTT mappings when moving from GTT domain" Date: Fri, 17 Jun 2011 12:06:54 -0700 Message-ID: <87mxhg5lbl.fsf@eliezer.anholt.net> References: <1308094989-3153-1-git-send-email-eric@anholt.net> <87tybrcet0.fsf@eliezer.anholt.net> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============1329982801==" Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org Errors-To: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org To: Chris Wilson , intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org --===============1329982801== Content-Type: multipart/signed; boundary="=-=-="; micalg=pgp-sha1; protocol="application/pgp-signature" --=-=-= Content-Transfer-Encoding: quoted-printable On Wed, 15 Jun 2011 17:03:58 +0100, Chris Wilson = wrote: > On Wed, 15 Jun 2011 08:08:59 -0700, Eric Anholt wrote: > > On Wed, 15 Jun 2011 10:39:06 +0100, Chris Wilson wrote: > > > For the record, on my SugarBay, this patch on top of drm-intel-fixes = and > > > mesa.git + your instruction streaming, regresses cairo-xlib 4.5s to 4= .7s > > > (4%, with a claim of .1% stddev) and improves cairo-gl from 21.5s to = 20.1s > > > (6.5%, with a claim of .5% stddev) for the firefox-talos-gfx. [This i= s a > > > system that can do under 2s for cairo-xlib using vmap and the perf > > > patches I posted before.] > > > -Chris > >=20 > > I was testing with semaphores enabled on the LLC series. I'm curious > > how this revert could possibly regress cairo-xlib, though -- that seems > > very strange. >=20 > Moving back to LLC+semaphores (2.6.39-rc2+), firefox-talos-gfx: > xlib: 4.473 > gl: 20.753 >=20 > applying the patch: > xlib: 4.472 > gl: 20.824 >=20 > I'm just not reproducing the same issue you are seeing. Are you using a > standard distro Kconfig, or if not, can you send me yours? > -Chris http://people.freedesktop.org/~anholt/dotconfig Pushed my kernel tree to "gtt-revert" branch. Fresh set of numbers on two fresh boots, just about as identical testing environments as I could produce: (without revert) anholt@eliezer:src/cairo/perf% cat /home/anholt/cairogl-before 21791099815 21677090234 22060471069 22049244399 21721362892 21966143347 (with revert) anholt@eliezer:src/cairo/perf% cat /home/anholt/cairogl-after=20 19162606198 19127697366 19462253708 19172535715 19339999910 19221539215 anholt@eliezer:src/cairo/perf% ministat ~/cairogl-before ~/cairogl-after x /home/anholt/cairogl-before + /home/anholt/cairogl-after +--------------------------------------------------------------------------= ----+ | + = x| |++ + + + xx x = x x| ||__A___| |___A_= _M_|| +--------------------------------------------------------------------------= ----+ N Min Max Median Avg Stddev x 6 2.167709e+10 2.2060471e+10 2.1966143e+10 2.1877569e+10 1.6902073e+08 + 6 1.9127697e+10 1.9462254e+10 1.9221539e+10 1.9247772e+10 1.2847426e+08 Difference at 95.0% confidence -2.6298e+09 +/- 1.93108e+08 -12.0205% +/- 0.882677% (Student's t, pooled s =3D 1.50123e+08) --=-=-= Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) iEYEARECAAYFAk37pc4ACgkQHUdvYGzw6veyYACfdGC6LurKUECioNYFfmRnmL/Y xl0AnihovYis8a3mfikcKNi5El8+J7fV =pn2g -----END PGP SIGNATURE----- --=-=-=-- --===============1329982801== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/intel-gfx --===============1329982801==--