[arch-commits] Commit in lib32-llvm/trunk (2 files)
Evangelos Foutras
foutrelis at archlinux.org
Sat Dec 31 14:46:04 UTC 2016
Date: Saturday, December 31, 2016 @ 14:46:04
Author: foutrelis
Revision: 203383
upgpkg: lib32-llvm 3.9.1-2
Revert commit that causes rendering issues on some GPUs (FS#52288).
Added:
lib32-llvm/trunk/AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch
Modified:
lib32-llvm/trunk/PKGBUILD
-------------------------------------------------------------------+
AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch | 140 ++++++++++
PKGBUILD | 9
2 files changed, 147 insertions(+), 2 deletions(-)
Added: AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch
===================================================================
--- AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch (rev 0)
+++ AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch 2016-12-31 14:46:04 UTC (rev 203383)
@@ -0,0 +1,140 @@
+From 25e2616626caafb896517e18cd8aa724fba2b200 Mon Sep 17 00:00:00 2001
+From: Tom Stellard <thomas.stellard at amd.com>
+Date: Tue, 29 Nov 2016 03:41:28 +0000
+Subject: [PATCH] Merging r280589:
+
+------------------------------------------------------------------------
+r280589 | nhaehnle | 2016-09-03 05:26:32 -0700 (Sat, 03 Sep 2016) | 19 lines
+
+AMDGPU: Fix an interaction between WQM and polygon stippling
+
+Summary:
+This fixes a rare bug in polygon stippling with non-monolithic pixel shaders.
+
+The underlying problem is as follows: the prolog part contains the polygon
+stippling sequence, i.e. a kill. The main part then enables WQM based on the
+_reduced_ exec mask, effectively undoing most of the polygon stippling.
+
+Since we cannot know whether polygon stippling will be used, the main part
+of a non-monolithic shader must always return to exact mode to fix this
+problem.
+
+Reviewers: arsenm, tstellarAMD, mareko
+
+Subscribers: arsenm, llvm-commits, kzhuravl
+
+Differential Revision: https://reviews.llvm.org/D23131
+
+------------------------------------------------------------------------
+
+git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_39@288105 91177308-0d34-0410-b5e6-96231b3b80d8
+---
+ lib/Target/AMDGPU/SIInstructions.td | 1 +
+ lib/Target/AMDGPU/SIWholeQuadMode.cpp | 7 -----
+ test/CodeGen/AMDGPU/wqm.ll | 49 ++++++++++++++++++++++++++++++++---
+ 3 files changed, 46 insertions(+), 11 deletions(-)
+
+diff --git a/lib/Target/AMDGPU/SIInstructions.td b/lib/Target/AMDGPU/SIInstructions.td
+index 18b7d5d..dde5f2f 100644
+--- a/lib/Target/AMDGPU/SIInstructions.td
++++ b/lib/Target/AMDGPU/SIInstructions.td
+@@ -2029,6 +2029,7 @@ def SI_RETURN : PseudoInstSI <
+ let hasSideEffects = 1;
+ let SALU = 1;
+ let hasNoSchedulingInfo = 1;
++ let DisableWQM = 1;
+ }
+
+ let Uses = [EXEC], Defs = [EXEC, VCC, M0],
+diff --git a/lib/Target/AMDGPU/SIWholeQuadMode.cpp b/lib/Target/AMDGPU/SIWholeQuadMode.cpp
+index b200c15..1534d58 100644
+--- a/lib/Target/AMDGPU/SIWholeQuadMode.cpp
++++ b/lib/Target/AMDGPU/SIWholeQuadMode.cpp
+@@ -219,13 +219,6 @@ char SIWholeQuadMode::scanInstructions(MachineFunction &MF,
+ markInstruction(MI, Flags, Worklist);
+ GlobalFlags |= Flags;
+ }
+-
+- if (WQMOutputs && MBB.succ_empty()) {
+- // This is a prolog shader. Make sure we go back to exact mode at the end.
+- Blocks[&MBB].OutNeeds = StateExact;
+- Worklist.push_back(&MBB);
+- GlobalFlags |= StateExact;
+- }
+ }
+
+ return GlobalFlags;
+diff --git a/test/CodeGen/AMDGPU/wqm.ll b/test/CodeGen/AMDGPU/wqm.ll
+index 809a7ba..41e4264 100644
+--- a/test/CodeGen/AMDGPU/wqm.ll
++++ b/test/CodeGen/AMDGPU/wqm.ll
+@@ -17,17 +17,18 @@ main_body:
+ ;CHECK-LABEL: {{^}}test2:
+ ;CHECK-NEXT: ; %main_body
+ ;CHECK-NEXT: s_wqm_b64 exec, exec
+-;CHECK: image_sample
+ ;CHECK-NOT: exec
+-;CHECK: _load_dword v0,
+-define amdgpu_ps float @test2(<8 x i32> inreg %rsrc, <4 x i32> inreg %sampler, float addrspace(1)* inreg %ptr, <4 x i32> %c) {
++define amdgpu_ps void @test2(<8 x i32> inreg %rsrc, <4 x i32> inreg %sampler, float addrspace(1)* inreg %ptr, <4 x i32> %c) {
+ main_body:
+ %c.1 = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %c, <8 x i32> %rsrc, <4 x i32> %sampler, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
+ %c.2 = bitcast <4 x float> %c.1 to <4 x i32>
+ %c.3 = extractelement <4 x i32> %c.2, i32 0
+ %gep = getelementptr float, float addrspace(1)* %ptr, i32 %c.3
+ %data = load float, float addrspace(1)* %gep
+- ret float %data
++
++ call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %data, float undef, float undef, float undef)
++
++ ret void
+ }
+
+ ; ... but disabled for stores (and, in this simple case, not re-enabled).
+@@ -414,6 +415,46 @@ entry:
+ ret void
+ }
+
++; Must return to exact at the end of a non-void returning shader,
++; otherwise the EXEC mask exported by the epilog will be wrong. This is true
++; even if the shader has no kills, because a kill could have happened in a
++; previous shader fragment.
++;
++; CHECK-LABEL: {{^}}test_nonvoid_return:
++; CHECK: s_mov_b64 [[LIVE:s\[[0-9]+:[0-9]+\]]], exec
++; CHECK: s_wqm_b64 exec, exec
++;
++; CHECK: s_and_b64 exec, exec, [[LIVE]]
++; CHECK-NOT: exec
++define amdgpu_ps <4 x float> @test_nonvoid_return() nounwind {
++ %tex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> undef, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
++ %tex.i = bitcast <4 x float> %tex to <4 x i32>
++ %dtex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %tex.i, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
++ ret <4 x float> %dtex
++}
++
++; CHECK-LABEL: {{^}}test_nonvoid_return_unreachable:
++; CHECK: s_mov_b64 [[LIVE:s\[[0-9]+:[0-9]+\]]], exec
++; CHECK: s_wqm_b64 exec, exec
++;
++; CHECK: s_and_b64 exec, exec, [[LIVE]]
++; CHECK-NOT: exec
++define amdgpu_ps <4 x float> @test_nonvoid_return_unreachable(i32 inreg %c) nounwind {
++entry:
++ %tex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> undef, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
++ %tex.i = bitcast <4 x float> %tex to <4 x i32>
++ %dtex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %tex.i, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
++
++ %cc = icmp sgt i32 %c, 0
++ br i1 %cc, label %if, label %else
++
++if:
++ store volatile <4 x float> %dtex, <4 x float>* undef
++ unreachable
++
++else:
++ ret <4 x float> %dtex
++}
+
+ declare void @llvm.amdgcn.image.store.v4i32(<4 x float>, <4 x i32>, <8 x i32>, i32, i1, i1, i1, i1) #1
+ declare void @llvm.amdgcn.buffer.store.f32(float, <4 x i32>, i32, i32, i1, i1) #1
Modified: PKGBUILD
===================================================================
--- PKGBUILD 2016-12-31 14:17:58 UTC (rev 203382)
+++ PKGBUILD 2016-12-31 14:46:04 UTC (rev 203383)
@@ -10,8 +10,8 @@
# Contributor: Gerardo Exequiel Pozzi <vmlinuz386 at yahoo.com.ar>
pkgname=('lib32-llvm' 'lib32-llvm-libs' 'lib32-clang')
-pkgver=3.9.1
-pkgrel=1
+pkgver=3.9.1
+pkgrel=2
arch=('x86_64')
url="http://llvm.org/"
license=('custom:University of Illinois/NCSA Open Source License')
@@ -21,6 +21,7 @@
source=(http://llvm.org/releases/$pkgver/llvm-$pkgver.src.tar.xz{,.sig}
http://llvm.org/releases/$pkgver/cfe-$pkgver.src.tar.xz{,.sig}
http://llvm.org/releases/$pkgver/compiler-rt-$pkgver.src.tar.xz{,.sig}
+ AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch
msan-prevent-initialization-failure-with-newer-glibc.patch)
sha256sums=('1fd90354b9cf19232e8f168faf2220e79be555df3aa743242700879e8fd329ee'
'SKIP'
@@ -28,6 +29,7 @@
'SKIP'
'd30967b1a5fa51a2503474aacc913e69fd05ae862d37bf310088955bdb13ec99'
'SKIP'
+ 'adadc41090ab0eb33fe4449a0ea3f02b15d03f186262dd89576848a2c66d0ce6'
'8e4f194c2283b91644a7fff43bc4e58c36b5507f2a4d90b72f275c0bd7511c20')
validpgpkeys=('B6C8F98282B944E3B0D5C2530FC3042E345AD05D'
'11E521D646982372EB577A1F8F0871F202119294')
@@ -41,6 +43,9 @@
mv "$srcdir/compiler-rt-$pkgver.src" projects/compiler-rt
+ # https://bugs.freedesktop.org/show_bug.cgi?id=99078
+ patch -Rp1 -i ../AMDGPU-Fix-an-interaction-between-WQM-and-polygon-stippling.patch
+
# https://reviews.llvm.org/D24736
patch -Np0 -d projects/compiler-rt <../msan-prevent-initialization-failure-with-newer-glibc.patch
More information about the arch-commits
mailing list