From: bugzilla-daemon@freedesktop.org
To: dri-devel@lists.freedesktop.org
Subject: [Bug 110781] Radeon: heavy r300 performance drop regression between 11.x and 19.x
Date: Tue, 04 Jun 2019 10:02:02 +0000 [thread overview]
Message-ID: <bug-110781-502-eQiZTj1vvq@http.bugs.freedesktop.org/> (raw)
In-Reply-To: <bug-110781-502@http.bugs.freedesktop.org/>
[-- Attachment #1.1: Type: text/plain, Size: 1798 bytes --]
https://bugs.freedesktop.org/show_bug.cgi?id=110781
--- Comment #28 from Richard Thier <u9vata@gmail.com> ---
Hi Rui!
We should also thank Marek for him proposing a possible fix. I am currently
creating a "less hacky" solution based on that attachment and a better idea to
not always force VRAM domain, but only if the r300 caller codes ask "VRAM or
something" and forcing GTT if it asks "GTT or something - but not VRAM". This
basically lead to the driver favoring the domain choice to be in VRAM, but
still supporting other things if the caller asks other domains directly.
Also incorporating the parts that Marek proposes is much better and cleaner
than my hack that just sets the flag right before it is read.
After I am completed with these changes and their testing I will publish it
here and only maybe then have any time to compile an x86_64 version for other
people. I highly suggest however for you and others to just grab the latest
revision from git and apply the changesets until the solution arrives in mesa
releases. I am on a single core, 32bit laptop with 1512-128MB of memory I got
in 2007 and I think most of the hardware is like from 2005 technology in it so
I might be some among the slowest cross-compiling person so please give a shot
to put together a build environment yourself.
It is not that hard to build mesa, this was my first time doing so too! Just
believe in yourself and read the nice docs on the mesa page about building with
meson. If you have any questions of the build process look around phoronix or
mesa docs and forums or if you give up just you can ask any people. Even I can
help now and you will be able too after you build yourself once ;-)
--
You are receiving this mail because:
You are the assignee for the bug.
[-- Attachment #1.2: Type: text/html, Size: 2648 bytes --]
[-- Attachment #2: Type: text/plain, Size: 159 bytes --]
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
next prev parent reply other threads:[~2019-06-04 10:02 UTC|newest]
Thread overview: 71+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-05-28 11:51 [Bug 110781] Radeon: heavy r300 performance drop regression between 11.x and 19.x bugzilla-daemon
2019-05-28 11:54 ` bugzilla-daemon
2019-05-28 16:19 ` bugzilla-daemon
2019-05-29 10:16 ` bugzilla-daemon
2019-05-29 10:32 ` bugzilla-daemon
2019-05-29 10:35 ` bugzilla-daemon
2019-05-29 15:49 ` bugzilla-daemon
2019-05-29 23:17 ` bugzilla-daemon
2019-05-30 12:23 ` bugzilla-daemon
2019-05-31 8:21 ` bugzilla-daemon
2019-06-02 20:42 ` bugzilla-daemon
2019-06-03 8:47 ` bugzilla-daemon
2019-06-03 10:11 ` bugzilla-daemon
2019-06-03 10:49 ` bugzilla-daemon
2019-06-03 12:35 ` bugzilla-daemon
2019-06-03 12:50 ` bugzilla-daemon
2019-06-03 12:50 ` bugzilla-daemon
2019-06-03 13:36 ` bugzilla-daemon
2019-06-03 18:52 ` bugzilla-daemon
2019-06-03 20:19 ` bugzilla-daemon
2019-06-03 20:51 ` bugzilla-daemon
2019-06-04 4:22 ` bugzilla-daemon
2019-06-04 5:50 ` bugzilla-daemon
2019-06-04 6:30 ` bugzilla-daemon
2019-06-04 6:48 ` bugzilla-daemon
2019-06-04 6:53 ` bugzilla-daemon
2019-06-04 7:56 ` bugzilla-daemon
2019-06-04 8:24 ` bugzilla-daemon
2019-06-04 9:43 ` bugzilla-daemon
2019-06-04 10:02 ` bugzilla-daemon [this message]
2019-06-04 10:23 ` bugzilla-daemon
2019-06-04 10:27 ` bugzilla-daemon
2019-06-04 10:33 ` bugzilla-daemon
2019-06-04 10:37 ` bugzilla-daemon
2019-06-04 12:43 ` bugzilla-daemon
2019-06-04 23:09 ` bugzilla-daemon
2019-06-05 9:52 ` bugzilla-daemon
2019-06-05 15:10 ` bugzilla-daemon
2019-06-05 16:14 ` bugzilla-daemon
2019-06-05 20:35 ` bugzilla-daemon
2019-06-05 20:50 ` bugzilla-daemon
2019-06-05 21:13 ` bugzilla-daemon
2019-06-05 22:47 ` bugzilla-daemon
2019-06-06 3:02 ` bugzilla-daemon
2019-06-06 17:46 ` bugzilla-daemon
2019-06-06 19:50 ` bugzilla-daemon
2019-06-07 1:48 ` bugzilla-daemon
2019-06-07 6:36 ` bugzilla-daemon
2019-06-07 8:05 ` bugzilla-daemon
2019-06-07 8:22 ` bugzilla-daemon
2019-06-07 8:38 ` bugzilla-daemon
2019-06-07 8:49 ` bugzilla-daemon
2019-06-07 14:43 ` bugzilla-daemon
2019-06-07 14:56 ` bugzilla-daemon
2019-06-07 15:10 ` bugzilla-daemon
2019-06-07 15:20 ` bugzilla-daemon
2019-06-07 15:25 ` bugzilla-daemon
2019-06-07 16:23 ` bugzilla-daemon
2019-06-07 19:23 ` bugzilla-daemon
2019-06-07 20:12 ` bugzilla-daemon
2019-06-08 3:28 ` bugzilla-daemon
2019-06-08 6:51 ` bugzilla-daemon
2019-06-08 16:43 ` bugzilla-daemon
2019-06-10 22:34 ` bugzilla-daemon
2019-06-11 11:23 ` bugzilla-daemon
2019-06-11 17:12 ` bugzilla-daemon
2019-06-11 17:33 ` bugzilla-daemon
2019-06-11 18:24 ` bugzilla-daemon
2019-06-11 23:14 ` bugzilla-daemon
2019-06-12 0:46 ` bugzilla-daemon
2019-06-24 8:44 ` bugzilla-daemon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=bug-110781-502-eQiZTj1vvq@http.bugs.freedesktop.org/ \
--to=bugzilla-daemon@freedesktop.org \
--cc=dri-devel@lists.freedesktop.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).