<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Wed, Mar 21, 2018 at 10:07 AM, Christian König <span dir="ltr"><<a href="mailto:christian.koenig@amd.com" target="_blank">christian.koenig@amd.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><span class="">
<div class="m_1741296731872194054moz-cite-prefix">Am 21.03.2018 um 14:57 schrieb Marek
Olšák:<br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">On Wed, Mar 21, 2018 at 4:13 AM,
Christian König <span dir="ltr"><<a href="mailto:ckoenig.leichtzumerken@gmail.com" target="_blank">ckoenig.leichtzumerken@gmail.<wbr>com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"><span>
<div class="m_1741296731872194054m_-1982752714132975532moz-cite-prefix">Am
21.03.2018 um 06:08 schrieb Marek Olšák:<br>
</div>
<blockquote type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">On Tue, Mar 20, 2018 at
4:16 PM, Christian König <span dir="ltr"><<a href="mailto:christian.koenig@amd.com" target="_blank">christian.koenig@amd.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
<div class="m_1741296731872194054m_-1982752714132975532m_4643584645760816653moz-cite-prefix">That's
what I meant with use up the otherwise
unused VRAM. I don't see any
disadvantage of always setting GTT as
second domain on APUs.<br>
<br>
My assumption was that we dropped this
in userspace for displayable surfaces,
but Marek proved that wrong.<br>
<br>
So what we should do is actually to add
GTT as fallback to all BOs on APUs in
Mesa and only make sure that the kernel
is capable of handling GTT with optimal
performance (e.g. have user huge pages
etc..).<span class="m_1741296731872194054m_-1982752714132975532HOEnZb"><font color="#888888"><br>
</font></span></div>
</div>
</blockquote>
<div><br>
</div>
<div>VRAM|GTT is practically as good as GTT.
VRAM with BO priorities and eviction
throttling is the true "VRAM|GTT".<br>
<br>
</div>
<div>I don't know how else to make use of VRAM
intelligently.</div>
</div>
</div>
</div>
</blockquote>
<br>
</span> Well why not set VRAM|GTT as default on APUs?
That should still save quite a bunch of moves even with
throttling.<br>
</div>
</blockquote>
<div><br>
</div>
<div>I explained why: <span class="m_1741296731872194054gmail-">VRAM|GTT is
practically as good as GTT.</span><br>
</div>
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF"> <br>
I mean there really shouldn't be any advantage to use
VRAM any more except that we want to use it up as long
as it is available.<br>
</div>
</blockquote>
<div><br>
</div>
<div>Why are you suggesting to use VRAM|GTT then? Let's just
only use GTT on all APUs.</div>
</div>
</div>
</div>
</blockquote>
<br></span>
Then we don't use the memory stolen for VRAM.<br>
<br>
See we want to get to a point where we have any ~16MB of stolen VRAM
on APUs and everything else in GTT.<br>
<br>
But we still have to support cases where we have 1GB stolen VRAM,
and wasting those 1GB would suck a bit.<br></div></blockquote><div><br></div>BO priorities and BO move throttling should take care of optimal VRAM usage regardless of the VRAM size. We can adjust the throttling for small VRAM, but that's about all we can do.</div><div class="gmail_quote"><br></div><div class="gmail_quote">VRAM|GTT doesn't guarantee that VRAM will be used usefully. In fact, it doesn't guarantee anything about VRAM.<br></div><div class="gmail_quote"><br></div><div class="gmail_quote">Marek<br></div></div></div>