mikolaj-pirog wrote:
The buildbot failure is unrelated to this change
https://github.com/llvm/llvm-project/pull/126627
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
mikolaj-pirog wrote:
> Good catch!
Thanks! I have fixed the formatting
https://github.com/llvm/llvm-project/pull/126390
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/126390
From c684a0a31ff8bc870991f1efb9a1a672cc6f0042 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Sat, 8 Feb 2025 16:12:36 -0800
Subject: [PATCH 1/2] Add missing _ph to intrinsics names
---
clan
https://github.com/mikolaj-pirog created
https://github.com/llvm/llvm-project/pull/126390
In my previous PR (#123656) to update the names of AVX10.2 intrinsics and
mnemonics, I have erroneously deleted `_ph` from few intrinsics. This PR
corrects this.
From c684a0a31ff8bc870991f1efb9a1a672cc6f
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/123656
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -24,573 +24,3369 @@
__attribute__((__always_inline__, __nodebug__, __target__("avx10.2-256"),
\
__min_vector_width__(256)))
+//clang-format off
+
+/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
+///single-precision (32-bit)
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/120766
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -24,567 +24,3243 @@
__attribute__((__always_inline__, __nodebug__, __target__("avx10.2-256"),
\
__min_vector_width__(256)))
+/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
+/// single-precision (32-bit) floating-point elements t
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/120766
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -24,567 +24,3243 @@
__attribute__((__always_inline__, __nodebug__, __target__("avx10.2-256"),
\
__min_vector_width__(256)))
+/// Convert two 128-bit vectors, \a __A and \a __B, containing packed
+/// single-precision (32-bit) floating-point elements t
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/119238
From 14ac4e104dd3adeb3c1d59d7c5ce3b0006391ffb Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Mon, 9 Dec 2024 18:59:38 +0100
Subject: [PATCH 1/2] Add comments
---
clang/lib/Headers/avx10_2co
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/119238
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog created
https://github.com/llvm/llvm-project/pull/119238
As in title
From 14ac4e104dd3adeb3c1d59d7c5ce3b0006391ffb Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Mon, 9 Dec 2024 18:59:38 +0100
Subject: [PATCH] Add comments
---
clang/lib/Headers/
@@ -1206,6 +1245,19 @@ void Writer::createMiscChunks() {
IMAGE_DLL_CHARACTERISTICS_EX_CET_COMPAT));
}
+ if (writeLTO) {
+debugRecords.emplace_back(COFF::IMAGE_DEBUG_TYPE_POGO,
+ make(ltcg));
mikolaj-pirog wrote:
> As I'm not a Windows developer, I defer to other reviewers' expertise on
> MSVC's PGO/LTO feature.
>
> However, to be honest, I'm unsure about the value of porting the strings
> given the large feature differences between Clang and MSVC on PGO and LTO.
> `clang -flto -c
mikolaj-pirog wrote:
Closing as #113528 fixed the issue
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog closed
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
mikolaj-pirog wrote:
Thanks to the review above mentioning LTO, I realized that this solution has a
problem with LTO, namely using LTO + PGO/PGU, will only emit the LTCG string in
the binary. This is because right now the magic sections are only written in
the COFF object file emission. I will
@@ -112,6 +112,8 @@ class MCTargetOptions {
// Whether or not to use full register names on PowerPC.
bool PPCUseFullRegisterNames : 1;
+ bool PgoInstrumentation = false;
mikolaj-pirog wrote:
As far as I am aware, there isn't any global metadata flag I co
@@ -1165,6 +1188,23 @@ void Writer::createMiscChunks() {
llvm::TimeTraceScope timeScope("Misc chunks");
Configuration *config = &ctx.config;
+ auto searchForPgoMagicSection = [this](char sectionName[]) {
mikolaj-pirog wrote:
I measured this by linking cl
mikolaj-pirog wrote:
Gently pinging @aganea @MaskRay @HaohaiWen. If there are no objections to this
patch, I would like to merge it
https://github.com/llvm/llvm-project/pull/114260
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.
mikolaj-pirog wrote:
Gently pinging @aganea @MaskRay
https://github.com/llvm/llvm-project/pull/114260
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,26 @@
+// This test checks if COFF file compiled with
+// -fprofile-generate has magic section ".pgi" to indicate so.
+
+// RUN: %clang --target=x86_64-pc-windows -fprofile-generate %s -c -o %t_x86
mikolaj-pirog wrote:
Okay, I see your point. I will st
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 01/13] Correct test
---
clang/test/CodeGen/debug
@@ -0,0 +1,26 @@
+// This test checks if COFF file compiled with
+// -fprofile-generate has magic section ".pgi" to indicate so.
+
+// RUN: %clang --target=x86_64-pc-windows -fprofile-generate %s -c -o %t_x86
mikolaj-pirog wrote:
There aren't any specific differe
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 01/12] Correct test
---
clang/test/CodeGen/debug
mikolaj-pirog wrote:
> > I would like to benchmark `lld` after this change, since I have added a
> > loop that goes through every section of every object file. Could someone
> > point in the direction of a good benchmark for that? I was thinking I can
> > benchmark on the linking of `clang` or
mikolaj-pirog wrote:
I would like to benchmark `lld` after this change, since I have added a loop
that goes through every section of every object file. Could someone point in
the direction of a good benchmark for that? I was thinking I can benchmark on
the linking of `clang` or any other big p
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 01/10] Correct test
---
clang/test/CodeGen/debug
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 01/11] Correct test
---
clang/test/CodeGen/debug
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
mi
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 1/9] Correct test
---
clang/test/CodeGen/debug-d
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 1/8] Correct test
---
clang/test/CodeGen/debug-d
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
+// RUN: dumpbin /H
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/114260
From f903e7e2effbd9675d0977dc1fd176ce97f11778 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Wed, 30 Oct 2024 16:30:39 +0100
Subject: [PATCH 1/5] Correct test
---
clang/test/CodeGen/debug-d
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/114260
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
mi
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
mi
@@ -1165,6 +1188,23 @@ void Writer::createMiscChunks() {
llvm::TimeTraceScope timeScope("Misc chunks");
Configuration *config = &ctx.config;
+ auto searchForPgoMagicSection = [this](char sectionName[]) {
+for (auto *obj : ctx.objFileInstances) {
+ for (auto &chun
@@ -1165,6 +1188,23 @@ void Writer::createMiscChunks() {
llvm::TimeTraceScope timeScope("Misc chunks");
Configuration *config = &ctx.config;
+ auto searchForPgoMagicSection = [this](char sectionName[]) {
mikolaj-pirog wrote:
I haven't measured that yet.
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/114260
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
+// RUN: dumpbin /H
@@ -0,0 +1,13 @@
+// This test checks if Window PE file compiled with -flto option contains a
magic
+// string "LTCG" to indicate LTO compilation.
+
+// REQUIRES: system-windows
+
+// RUN: %clang --target=x86_64-pc-windows-msvc -flto -fuse-ld=lld %s -o %t.exe
+// RUN: dumpbin /H
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/112750
From 06472c56a4a916dc2fd7b29a0c137597bbda0504 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Thu, 17 Oct 2024 10:17:09 -0700
Subject: [PATCH 1/2] Provide default value for -fprofile-sample-us
mikolaj-pirog wrote:
> > Regarding the default filename extension, it doesn't have to be `.profdata`
> > for `fprofile-sample-use`, it can be the `.afdo`, or any other extension
> > deemed proper. `fprofile-sample-use` and `fprofile-use` would behave
> > similarly then: they would look for a `
mikolaj-pirog wrote:
Regarding the default filename extension, it doesn't have to be `.profdata` for
`fprofile-sample-use`, it can be the `.afdo`, or any other extension deemed
proper. `fprofile-sample-use` and `fprofile-use` would behave similarly then:
they would look for a `default` file w
mikolaj-pirog wrote:
It's true that there isn't any compatibility argument for leaving
`-fprofile-sample/auto` (it was silently ignored afterall), but I think the
consistency argument is important. From a standpoint of a clang user, it's very
reasonable to expect that `fprofile-use` and `-fpro
mikolaj-pirog wrote:
This would close #112391
I am copying my comment from my PR #112750 which addresses the issue #112391:
I think this isn't the best solution, since it is inconsistent with the
behavior of the -fprofile-use option. Why should -fprofile-use be accepted and
-fprofile-sample-
mikolaj-pirog wrote:
> Actually I think we should drop the accepted and ignored `Flag` form of
> `-fprofile-sample-use` and `-fprofile-auto`, which is misleading at the
> least. I have a pending patch, which I will upload at night.
>
> For sample PGO, the profile file extension is usually `.af
mikolaj-pirog wrote:
Pinging @MaskRay for review, I can't add reviewers in the dedicated gh section
yet
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/lis
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/112750
From 06472c56a4a916dc2fd7b29a0c137597bbda0504 Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Thu, 17 Oct 2024 10:17:09 -0700
Subject: [PATCH 1/2] Provide default value for -fprofile-sample-us
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/112750
From 642f7ac9b1e999d8adb690b67c6303139794379d Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Thu, 17 Oct 2024 10:17:09 -0700
Subject: [PATCH 1/3] Provide default value for -fprofile-sample-us
https://github.com/mikolaj-pirog ready_for_review
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog converted_to_draft
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/112750
From 642f7ac9b1e999d8adb690b67c6303139794379d Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Thu, 17 Oct 2024 10:17:09 -0700
Subject: [PATCH 1/2] Provide default value for -fprofile-sample-us
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog updated
https://github.com/llvm/llvm-project/pull/112750
From 642f7ac9b1e999d8adb690b67c6303139794379d Mon Sep 17 00:00:00 2001
From: "Pirog, Mikolaj Maciej"
Date: Thu, 17 Oct 2024 10:17:09 -0700
Subject: [PATCH 1/2] Provide default value for -fprofile-sample-us
https://github.com/mikolaj-pirog edited
https://github.com/llvm/llvm-project/pull/112750
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/mikolaj-pirog created
https://github.com/llvm/llvm-project/pull/112750
As in title. This matches the behavior of `-fprofile-use`, which when used
without a path to the file, looks for `default.profdata` file in the
current/supplied dir.
From 642f7ac9b1e999d8adb690b67c630313
59 matches
Mail list logo
