From mboxrd@z Thu Jan 1 00:00:00 1970 From: bugzilla-daemon@freedesktop.org Subject: [Bug 110781] Radeon: heavy r300 performance drop regression between 11.x and 19.x Date: Fri, 07 Jun 2019 15:20:31 +0000 Message-ID: References: Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============0783805278==" Return-path: Received: from culpepper.freedesktop.org (culpepper.freedesktop.org [IPv6:2610:10:20:722:a800:ff:fe98:4b55]) by gabe.freedesktop.org (Postfix) with ESMTP id 25F2989D7D for ; Fri, 7 Jun 2019 15:20:31 +0000 (UTC) In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" To: dri-devel@lists.freedesktop.org List-Id: dri-devel@lists.freedesktop.org --===============0783805278== Content-Type: multipart/alternative; boundary="15599208310.5EEedbeB8.22781" Content-Transfer-Encoding: 7bit --15599208310.5EEedbeB8.22781 Date: Fri, 7 Jun 2019 15:20:31 +0000 MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: http://bugs.freedesktop.org/ Auto-Submitted: auto-generated https://bugs.freedesktop.org/show_bug.cgi?id=3D110781 --- Comment #54 from Richard Thier --- > The GPU has a scaler that can scale a smaller image to the native timing.= Many > LCD monitors also have a scaler built in to provide the same funct= ionality. I see, but still I see a big performance rise when I change the resolution smaller and let this scaler do its work in the last phase. I have no idea if there is a real speed difference between this two: - having 640x480 on a 4:3 aspect screen without a scaler - Having hardware pixels of 1024x768, but using the scaler in the end as 640x480 software. - Z tests run for smaller amount of pixels - Stencil and Z buffers are smaller just like the backbuffer - etc. Memory usage on the GPU should be still smaller, pixel shaders still run ra= rer times, post processing things still run on smaller amount of pixels etc... = The only reason it can be slower if for some reason the scaler circuit would be= so slow while doing its job that it cannot keep its FPS, but I think this is a highly specialized hardware not doing anything else so I hoped it is just "fast" and even if there is some loss I guess it is barely measurable - but tell me if I am wrong as that would be new to me if this has a really measurable overhead. In any ways I think it is better to use a same aspect ratio, but smaller resolution than using the biggest possible native size. Actually on my mach= ine I get faster frame rate even if I run a game in 640x480 window on the top l= eft corner the same way as if I turn the screen into 640x480 itself and even th= en I barely see a difference despite this case I feel there should be some overh= ead against using directly a video mode for that.. --=20 You are receiving this mail because: You are the assignee for the bug.= --15599208310.5EEedbeB8.22781 Date: Fri, 7 Jun 2019 15:20:31 +0000 MIME-Version: 1.0 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: http://bugs.freedesktop.org/ Auto-Submitted: auto-generated

Comme= nt # 54 on bug 11078= 1 from = Richard Thier
> The GPU has a scaler that can scale=
 a smaller image to the native timing.  Many > LCD monitors also have a =
scaler built in to provide the same functionality.

I see, but still I see a big performance rise when I change the resolution
smaller and let this scaler do its work in the last phase.

I have no idea if there is a real speed difference between this two:
- having 640x480 on a 4:3 aspect screen without a scaler
- Having hardware pixels of 1024x768, but using the scaler in the end as
640x480 software.
- Z tests run for smaller amount of pixels
- Stencil and Z buffers are smaller just like the backbuffer
- etc.

Memory usage on the GPU should be still smaller, pixel shaders still run ra=
rer
times, post processing things still run on smaller amount of pixels etc... =
The
only reason it can be slower if for some reason the scaler circuit would be=
 so
slow while doing its job that it cannot keep its FPS, but I think this is a
highly specialized hardware not doing anything else so I hoped it is just
"fast" and even if there is some loss I guess it is barely measur=
able - but
tell me if I am wrong as that would be new to me if this has a really
measurable overhead.

In any ways I think it is better to use a same aspect ratio, but smaller
resolution than using the biggest possible native size. Actually on my mach=
ine
I get faster frame rate even if I run a game in 640x480 window on the top l=
eft
corner the same way as if I turn the screen into 640x480 itself and even th=
en I
barely see a difference despite this case I feel there should be some overh=
ead
against using directly a video mode for that..


You are receiving this mail because:
  • You are the assignee for the bug.
= --15599208310.5EEedbeB8.22781-- --===============0783805278== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KZHJpLWRldmVs IG1haWxpbmcgbGlzdApkcmktZGV2ZWxAbGlzdHMuZnJlZWRlc2t0b3Aub3JnCmh0dHBzOi8vbGlz dHMuZnJlZWRlc2t0b3Aub3JnL21haWxtYW4vbGlzdGluZm8vZHJpLWRldmVs --===============0783805278==--