OpenHarmony/third_party_mesa3d
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

intel/fs: Track single accumulator in scoreboard lowering pass.

Browse Source 
This change reduces the precision of the scoreboard data structure for
accumulator registers, because the rules determining the aliasing of
accumulator registers are non-trivial and poorly documented (e.g. acc0
overlaps the storage of acc1 when the former is accessed with an
integer type).  We could implement those rules but it wouldn't have
any practical benefit since we currently only use acc0-1, and for the
most part we can rely on the hardware's accumulator dependency
tracking.  Instead make our lives easier by representing it as a
single register.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11433>
This commit is contained in:
Francisco Jerez
2021-05-25 14:41:26 -07:00
committed by Marge Bot
parent 231337a13a
commit 385da1fe36
1 changed files with 9 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
src/intel/compiler/brw_fs_scoreboard.cpp
View File

@@ -43,6 +43,8 @@
* - ip instruction pointer
* - tm0 timestamp register
* - dbg0 debug register
* - acc2-9 special accumulator registers on TGL
* - mme0-7 math macro extended accumulator registers
*
* The following ARF registers don't need to be tracked here because data
* coherency is still provided transparently by the hardware:
@@ -615,9 +617,7 @@ namespace {
sb.grf_deps[i] = merge(eq, sb0.grf_deps[i], sb1.grf_deps[i]);
sb.addr_dep = merge(eq, sb0.addr_dep, sb1.addr_dep);
for (unsigned i = 0; i < ARRAY_SIZE(sb.accum_deps); i++)
sb.accum_deps[i] = merge(eq, sb0.accum_deps[i], sb1.accum_deps[i]);
sb.accum_dep = merge(eq, sb0.accum_dep, sb1.accum_dep);
return sb;
}
@@ -635,9 +635,7 @@ namespace {
sb.grf_deps[i] = shadow(sb0.grf_deps[i], sb1.grf_deps[i]);
sb.addr_dep = shadow(sb0.addr_dep, sb1.addr_dep);
for (unsigned i = 0; i < ARRAY_SIZE(sb.accum_deps); i++)
sb.accum_deps[i] = shadow(sb0.accum_deps[i], sb1.accum_deps[i]);
sb.accum_dep = shadow(sb0.accum_dep, sb1.accum_dep);
return sb;
}
@@ -655,9 +653,7 @@ namespace {
sb.grf_deps[i] = transport(sb0.grf_deps[i], delta);
sb.addr_dep = transport(sb0.addr_dep, delta);
for (unsigned i = 0; i < ARRAY_SIZE(sb.accum_deps); i++)
sb.accum_deps[i] = transport(sb0.accum_deps[i], delta);
sb.accum_dep = transport(sb0.accum_dep, delta);
return sb;
}
@@ -673,10 +669,8 @@ namespace {
if (sb0.addr_dep != sb1.addr_dep)
return false;
for (unsigned i = 0; i < ARRAY_SIZE(sb0.accum_deps); i++) {
if (sb0.accum_deps[i] != sb1.accum_deps[i])
return false;
}
if (sb0.accum_dep != sb1.accum_dep)
return false;
return true;
}
@@ -690,7 +684,7 @@ namespace {
private:
dependency grf_deps[BRW_MAX_GRF];
dependency addr_dep;
dependency accum_deps[10];
dependency accum_dep;
dependency *
dep(const fs_reg &r)
@@ -703,8 +697,7 @@ namespace {
r.file == ARF && reg >= BRW_ARF_ADDRESS &&
reg < BRW_ARF_ACCUMULATOR ? &addr_dep :
r.file == ARF && reg >= BRW_ARF_ACCUMULATOR &&
reg < BRW_ARF_FLAG ? &accum_deps[
reg - BRW_ARF_ACCUMULATOR] :
reg < BRW_ARF_FLAG ? &accum_dep :
NULL);
}
};