<div dir="auto"><div><br><div class="gmail_extra"><br><div class="gmail_quote">On Jun 25, 2017 9:23 AM, "Nicolai Hähnle" <<a href="mailto:nhaehnle@gmail.com">nhaehnle@gmail.com</a>> wrote:<br type="attribution"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On <a href="tel:24.06.2017%2013" value="+12406201713" target="_blank">24.06.2017 13</a>:35, Marek Olšák wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
<br>
On Jun 24, 2017 12:01 PM, "Nicolai Hähnle" <<a href="mailto:nhaehnle@gmail.com" target="_blank">nhaehnle@gmail.com</a> <mailto:<a href="mailto:nhaehnle@gmail.com" target="_blank">nhaehnle@gmail.com</a>>> wrote:<br>
<br>
On <a href="tel:22.06.2017%2017" value="+12206201717" target="_blank">22.06.2017 17</a> <tel:22.06.2017%2017>:10, Marek Olšák wrote:<br>
<br>
From: Marek Olšák <<a href="mailto:marek.olsak@amd.com" target="_blank">marek.olsak@amd.com</a> <mailto:<a href="mailto:marek.olsak@amd.com" target="_blank">marek.olsak@amd.com</a>>><div class="quoted-text"><br>
<br>
+2.3% better score on Fiji. It might be better without HBM.<br>
<br>
<br>
Can this be a drirc option?<br>
<br>
<br>
The answer is the same as for: Can sisched be a flag defined in p_defines.h, drirc, and plumbed by st/dri? Nobody has ever put driver-specific stuff into p_defines.h.<br>
</div></blockquote>
<br>
Obviously the answer to these is no. However, some interface that provides direct generic access to drirc could be plumbed through.</blockquote></div></div></div><div dir="auto"><br></div><div dir="auto">How?</div><div dir="auto"><br></div><div dir="auto">Marek</div><div dir="auto"><br></div><div dir="auto"><div class="gmail_extra"><div class="gmail_quote"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><font color="#888888"><br>
<br>
Nicolai</font><div class="elided-text"><br>
<br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Marek<br>
<br>
<br>
<br>
<br>
<br>
---<br>
src/gallium/drivers/radeonsi/<wbr>si_pipe.c | 7 +++++++<br>
1 file changed, 7 insertions(+)<br>
<br>
diff --git a/src/gallium/drivers/radeonsi<wbr>/si_pipe.c<br>
b/src/gallium/drivers/radeonsi<wbr>/si_pipe.c<br>
index ff787ad..4088849 100644<br>
--- a/src/gallium/drivers/radeonsi<wbr>/si_pipe.c<br>
+++ b/src/gallium/drivers/radeonsi<wbr>/si_pipe.c<br>
@@ -935,20 +935,27 @@ struct pipe_screen<br>
*radeonsi_screen_create(struct radeon_winsys *ws,<br>
si_init_screen_state_function<wbr>s(sscreen);<br>
if (!r600_common_screen_init(&ssc<wbr>reen->b, ws, flags) ||<br>
!si_init_gs_info(sscreen) ||<br>
!si_init_shader_cache(<wbr>sscreen)) {<br>
FREE(sscreen);<br>
return NULL;<br>
}<br>
+ /* Enable sisched where it helps. */<br>
+ char process[128];<br>
+ if (os_get_process_name(process, sizeof(process)) &&<br>
+ /* Unigine Superposition */<br>
+ !strcmp(process, "superposition"))<br>
+ sscreen->b.debug_flags |= DBG_SI_SCHED;<br>
+<br>
/* Only enable as many threads as we have target<br>
machines, but at most<br>
* the number of CPUs - 1 if there is more than one.<br>
*/<br>
num_threads = sysconf(_SC_NPROCESSORS_ONLN);<br>
num_threads = MAX2(1, num_threads - 1);<br>
num_compiler_threads = MIN2(num_threads,<br>
ARRAY_SIZE(sscreen->tm));<br>
num_compiler_threads_lowprio =<br>
MIN2(num_threads,<br>
ARRAY_SIZE(sscreen->tm_low_pri<wbr>ority));<br>
if (!util_queue_init(&sscreen->sh<wbr>ader_compiler_queue,<br>
"si_shader",<br>
<br>
<br>
<br>
-- Lerne, wie die Welt wirklich ist,<br>
Aber vergiss niemals, wie sie sein sollte.<br>
<br>
<br>
</blockquote>
<br>
<br>
-- <br>
Lerne, wie die Welt wirklich ist,<br>
Aber vergiss niemals, wie sie sein sollte.<br>
</div></blockquote></div><br></div></div></div>