https://github.com/JanekvO updated
https://github.com/llvm/llvm-project/pull/72129
>From 04b0856ae59ff367fd1b6736d49430648976cb25 Mon Sep 17 00:00:00 2001
From: Janek van Oirschot
Date: Mon, 13 Nov 2023 07:38:15 -0800
Subject: [PATCH 1/2] [AMDGPU] Avoid hitting AMDGPUAsmPrinter related asserts
https://github.com/JanekvO updated
https://github.com/llvm/llvm-project/pull/72129
>From 04b0856ae59ff367fd1b6736d49430648976cb25 Mon Sep 17 00:00:00 2001
From: Janek van Oirschot
Date: Mon, 13 Nov 2023 07:38:15 -0800
Subject: [PATCH 1/3] [AMDGPU] Avoid hitting AMDGPUAsmPrinter related asserts
JanekvO wrote:
Ping
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
JanekvO wrote:
Upstream's formatting does not seem to agree with my local `git clang-format`.
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-c
JanekvO wrote:
> Is it best to leave comments on the commit from the PR comment, i.e.
> currently
> [51f72f1](https://github.com/llvm/llvm-project/commit/51f72f115b340a092c2c9f8569911b944a4efb6d)
> ?
>
> I can't seem to start a review, so it will be individual comments (I think).
> Is there
https://github.com/JanekvO edited
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -40,12 +41,20 @@ class AMDGPUAsmPrinter final : public AsmPrinter {
AMDGPUResourceUsageAnalysis *ResourceUsage;
+ std::unique_ptr RI;
JanekvO wrote:
It requires a class inherited member for its constructor (in this case,
`OutContext`).
https://github
@@ -3025,8 +3025,8 @@ define amdgpu_kernel void @dyn_extract_v5f64_s_s(ptr
addrspace(1) %out, i32 %sel
; GPRIDX-NEXT: amd_machine_version_stepping = 0
; GPRIDX-NEXT: kernel_code_entry_byte_offset = 256
; GPRIDX-NEXT: kernel_code_prefetch_byte_size = 0
-; GPRIDX-NEX
@@ -75,10 +75,10 @@ bb.2:
store volatile i32 0, ptr addrspace(1) undef
ret void
}
-; DEFAULTSIZE: .amdhsa_private_segment_fixed_size 16
+; DEFAULTSIZE: .amdhsa_private_segment_fixed_size
kernel_non_entry_block_static_alloca_uniformly_reached_align4.private_seg_size
; DEFA
@@ -0,0 +1,225 @@
+//===- AMDGPUMCResourceInfo.cpp --- MC Resource Info
--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Ap
@@ -40,12 +41,20 @@ class AMDGPUAsmPrinter final : public AsmPrinter {
AMDGPUResourceUsageAnalysis *ResourceUsage;
+ std::unique_ptr RI;
JanekvO wrote:
Only was I was able to do so was by explicitly adding the `MCContext` as
argument for every public met
@@ -3025,8 +3025,8 @@ define amdgpu_kernel void @dyn_extract_v5f64_s_s(ptr
addrspace(1) %out, i32 %sel
; GPRIDX-NEXT: amd_machine_version_stepping = 0
; GPRIDX-NEXT: kernel_code_entry_byte_offset = 256
; GPRIDX-NEXT: kernel_code_prefetch_byte_size = 0
-; GPRIDX-NEX
@@ -0,0 +1,220 @@
+//===- AMDGPUMCResourceInfo.cpp --- MC Resource Info
--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Ap
https://github.com/JanekvO edited
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -68,82 +71,84 @@ void MCResourceInfo::assignMaxRegs() {
assignMaxRegSym(MaxSGPRSym, MaxSGPR);
}
-void MCResourceInfo::finalize() {
- assert(!finalized && "Cannot finalize ResourceInfo again.");
- finalized = true;
- assignMaxRegs();
+void MCResourceInfo::finalize(MCCon
@@ -68,82 +71,84 @@ void MCResourceInfo::assignMaxRegs() {
assignMaxRegSym(MaxSGPRSym, MaxSGPR);
}
-void MCResourceInfo::finalize() {
- assert(!finalized && "Cannot finalize ResourceInfo again.");
- finalized = true;
- assignMaxRegs();
+void MCResourceInfo::finalize(MCCon
@@ -3,6 +3,18 @@
declare i32 @llvm.amdgcn.workitem.id.x()
+define <2 x i64> @f1() #0 {
JanekvO wrote:
There were some function order changes in some tests. This function is the same
one as the function that is removed a couple of lines below.
https://githu
@@ -0,0 +1,533 @@
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -enable-ipra=0
-verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
+
+; SGPR use may not seem equal to the sgpr use provided in comments as the
latter includes extra sgprs (e.g., for vcc use).
@@ -0,0 +1,533 @@
+; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 -enable-ipra=0
-verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
+
+; SGPR use may not seem equal to the sgpr use provided in comments as the
latter includes extra sgprs (e.g., for vcc use).
+
+; Fun
https://github.com/JanekvO closed
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
JanekvO wrote:
I've added a small paragraph to AMDGPUUsage wrt the symbols being generated and
emitted which looks as follows in the browser:
https://github.com/llvm/llvm-project/pull/1029
@@ -1755,6 +1755,56 @@ As part of the AMDGPU MC layer, AMDGPU provides the
following target specific
=== =
+Function Resource Usage
+---
+
+The function resourc
JanekvO wrote:
Rebase, hopefully pull in fixes for unrelated test failures
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
JanekvO wrote:
Rebase
https://github.com/llvm/llvm-project/pull/102913
___
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -40,12 +42,19 @@ class AMDGPUAsmPrinter final : public AsmPrinter {
AMDGPUResourceUsageAnalysis *ResourceUsage;
+ MCResourceInfo RI;
+
SIProgramInfo CurrentProgramInfo;
std::unique_ptr HSAMetadataStream;
MCCodeEmitter *DumpCodeInstEmitter = nullptr;
+ //
@@ -0,0 +1,225 @@
+//===- AMDGPUMCResourceInfo.cpp --- MC Resource Info
--===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM
Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Ap
26 matches
Mail list logo
