SLTozer wrote:
Split into two new reviews: https://github.com/llvm/llvm-project/pull/11 &
https://github.com/llvm/llvm-project/pull/110102
https://github.com/llvm/llvm-project/pull/106724
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
htt
@@ -22,6 +23,67 @@ namespace llvm {
class LLVMContext;
class raw_ostream;
class DILocation;
+ class Function;
+
+#if ENABLE_DEBUGLOC_COVERAGE_TRACKING
+ // Used to represent different "kinds" of DebugLoc, expressing that a
DebugLoc
+ // is either ordinary, containing
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/107279
>From e45d7e68a371a09ea766c4accf8edc6c030fd7fd Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 4 Sep 2024 12:09:50 +0100
Subject: [PATCH 1/3] Add CMake option to enable expensive line number origin
trac
@@ -1279,6 +1279,9 @@ void Instruction::swapProfMetadata() {
void Instruction::copyMetadata(const Instruction &SrcInst,
ArrayRef WL) {
+ if (WL.empty() || is_contained(WL, LLVMContext::MD_dbg))
+setDebugLoc(SrcInst.getDebugLoc());
--
https://github.com/SLTozer created
https://github.com/llvm/llvm-project/pull/11
Following the commit that added the fake use intrinsic to LLVM, this patch adds
a pair of flags for the clang frontend that emit fake use intrinsics, for the
purpose of extending the lifetime of variables (eith
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/11
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
SLTozer wrote:
Ping - if anyone isn't able to review themselves but has an idea of someone who
might, adding/pinging them would also be appreciated.
https://github.com/llvm/llvm-project/pull/11
___
cfe-commits mailing list
cfe-commits@lists.llvm.o
@@ -1834,6 +1834,14 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
Opts.setInlining(CodeGenOptions::NormalInlining);
}
+ // If we have specified -Og and have not set any explicit -fextend-lifetimes
+ // value, then default to -fexten
@@ -0,0 +1,30 @@
+/// Check that we generate fake uses only when -fextend-lifetimes is set and we
+/// are not setting optnone, or when we have optimizations set to -Og and we
have
+/// not passed -fno-extend-lifetimes.
+// RUN: %clang_cc1 -emit-llvm -disable-llvm-passes -O0 -fex
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/118026
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,61 @@
+// RUN: %clang_cc1 %s -O0 -disable-O0-optnone -emit-llvm -fextend-lifetimes
-fsanitize=null -fsanitize-trap=null -o - | FileCheck
--check-prefixes=CHECK,NULL --implicit-check-not=ubsantrap %s
+// RUN: %clang_cc1 %s -O0 -disable-O0-optnone -emit-llvm -fextend-li
SLTozer wrote:
> As far as I understand it, these are driver options that will be passed
> through to cc1? In that case we can at least test the passthrough, i.e. that
> `-fextend-lifetimes` appears in the output of `-###`.
I think my preferred approach would be to merge both patches simultane
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/118026
>From 7e42a82d73e3be98aab8e523b16a2e43d31b0552 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Thu, 28 Nov 2024 13:53:20 +
Subject: [PATCH 1/2] Enable -fextend-lifetimes at -Og
---
clang/docs/CommandGui
@@ -1353,6 +1353,19 @@ void CodeGenFunction::EmitLifetimeEnd(llvm::Value *Size,
llvm::Value *Addr) {
C->setDoesNotThrow();
}
+void CodeGenFunction::EmitFakeUse(Address Addr) {
+ // We do not emit a fake use if we want to apply optnone to this function,
+ // even if we mig
@@ -1664,6 +1710,17 @@ CodeGenFunction::EmitAutoVarAlloca(const VarDecl &D) {
emission.getOriginalAllocatedAddress(),
emission.getSizeForLifetimeMarkers());
+ // Analogous to lifetime markers,
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/110102
>From fdfa695bc74847f5cc366bfcbf142bd5c2e3937f Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 16:55:39 +0100
Subject: [PATCH 1/3] [Clang] Add fake use emission to Clang with
-fextend-lifeti
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/110102
>From fdfa695bc74847f5cc366bfcbf142bd5c2e3937f Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 16:55:39 +0100
Subject: [PATCH 1/4] [Clang] Add fake use emission to Clang with
-fextend-lifeti
@@ -2217,6 +2217,11 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
Args.getAllArgValues(OPT_fsanitize_trap_EQ), Diags,
Opts.SanitizeTrap);
+ Opts.ExtendThisPtr =
+ Opts.OptimizationLevel > 0
SLTozer wrote:
This patch has been changed as a result of some other patches that have removed
the need for the attribute that this patch once added. Since this patch does
not emit fake uses (the actual emission of fake uses is added in the [next
patch](https://github.com/llvm/llvm-project/pul
@@ -0,0 +1,22 @@
+// RUN: %clang_cc1 %s -O2 -emit-llvm -fextend-lifetimes -o - | FileCheck %s
+// Make sure we don't generate fake.use for non-scalar variables.
+// Make sure we don't generate fake.use for volatile variables
+// and parameters even when they are scalar.
+
+struct
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/11
>From a33566974fbf260181d20d3f1b67081d21493506 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/2] [Clang] Add "extend lifetime" flags and release note
Follow
https://github.com/SLTozer created
https://github.com/llvm/llvm-project/pull/118026
This patch follows two other ongoing reviews,
https://github.com/llvm/llvm-project/pull/11 and
https://github.com/llvm/llvm-project/pull/110102, which add a new feature to
Clang - `-fextend-lifetimes`, whi
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/110102
>From 7d8646edb373c38ee761de2c338c45b967a423e7 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 16:55:39 +0100
Subject: [PATCH 1/2] [Clang] Add fake use emission to Clang with
-fextend-lifeti
@@ -0,0 +1,29 @@
+// RUN: %clang_cc1 %s -O2 -emit-llvm -fextend-lifetimes -o - | FileCheck %s
+// Check that fake use calls are emitted at the correct locations, i.e.
+// at the end of lexical blocks and at the end of the function.
+
+extern int use(int);
+int glob1;
+int glob2;
+
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -0,0 +1,43 @@
+// RUN: %clang_cc1 %s -triple=%itanium_abi_triple -O1 -emit-llvm
-fextend-lifetimes -o - | FileCheck %s
+// Make sure we don't crash compiling a lambda that is not nested in a
function.
+// We also check that fake uses are properly issued in lambdas.
+
+int glo
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/11
>From 6a873f5c487a936344f6cd226b7d525b406f34b2 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/4] [Clang] Add "extend lifetime" flags and release note
Follow
@@ -1353,6 +1353,19 @@ void CodeGenFunction::EmitLifetimeEnd(llvm::Value *Size,
llvm::Value *Addr) {
C->setDoesNotThrow();
}
+void CodeGenFunction::EmitFakeUse(Address Addr) {
+ // We do not emit a fake use if we want to apply optnone to this function,
+ // even if we mig
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -1834,6 +1834,14 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
Opts.setInlining(CodeGenOptions::NormalInlining);
}
+ // If we have specified -Og and have not explicitly set
-fno-extend-lifetimes,
+ // then default to -fextend-li
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -1664,6 +1710,17 @@ CodeGenFunction::EmitAutoVarAlloca(const VarDecl &D) {
emission.getOriginalAllocatedAddress(),
emission.getSizeForLifetimeMarkers());
+ // Analogous to lifetime markers,
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/11
>From 6a873f5c487a936344f6cd226b7d525b406f34b2 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/6] [Clang] Add "extend lifetime" flags and release note
Follow
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/118026
>From efd7f58e421b0afdc886688128f72170e0c4a970 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/2] [Clang] Add "extend lifetime" flags and release note
Follow
@@ -1834,6 +1834,14 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions
&Opts, ArgList &Args,
Opts.setInlining(CodeGenOptions::NormalInlining);
}
+ // If we have specified -Og and have not explicitly set
-fno-extend-lifetimes,
+ // then default to -fextend-li
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/11
>From 6a873f5c487a936344f6cd226b7d525b406f34b2 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/7] [Clang] Add "extend lifetime" flags and release note
Follow
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/110102
>From fdfa695bc74847f5cc366bfcbf142bd5c2e3937f Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 16:55:39 +0100
Subject: [PATCH 1/5] [Clang] Add fake use emission to Clang with
-fextend-lifeti
SLTozer wrote:
I've added a test for this patch that tests the driver - since we will now be
relying on the driver to decide whether to pass these flags to the frontend,
rather than trivially always doing so (so that [later
on](https://github.com/llvm/llvm-project/pull/118026) we can use the d
SLTozer wrote:
I've removed the logic that disables this flag at O0 - this has also removed
the need for `-disable-O0-optnone` in any tests. Following the prior patch
updating this flag to be an `=` flag represented by an enum, the latest commit
changes this patch to handle that type. Finally,
SLTozer wrote:
I've updated this patch, and made it dependent on a [previous
PR](https://github.com/llvm/llvm-project/pull/11) - the logic for enabling
the flag is coupled with code from the earlier PR, and also we now slightly
extend the test added in that PR instead of creating a new one
SLTozer wrote:
Post-holiday ping - I believe all the comments have been addressed now, does
the new form look acceptable?
https://github.com/llvm/llvm-project/pull/11
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/c
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/110102
>From fdfa695bc74847f5cc366bfcbf142bd5c2e3937f Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 16:55:39 +0100
Subject: [PATCH 1/6] [Clang] Add fake use emission to Clang with
-fextend-lifeti
@@ -412,6 +412,20 @@ New Compiler Flags
only for thread-local variables, and none (which corresponds to the
existing ``-fno-c++-static-destructors`` flag) skips all static
destructors registration.
+- The ``-fextend-variable-liveness`` flag has been added to allow for imp
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/11
>From 6a873f5c487a936344f6cd226b7d525b406f34b2 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 25 Sep 2024 15:08:39 +0100
Subject: [PATCH 1/9] [Clang] Add "extend lifetime" flags and release note
Follow
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/123737
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SLTozer approved this pull request.
Code changes LGTM. Regarding the direction of the patch, it seems a little
unfortunate to completely remove `getFirstNonPHI()` - sometimes we want the
instruction and sometimes we want the iterator, and forcing all of them to go
through `g
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/123737
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -173,8 +173,9 @@ Value *SSAUpdater::GetValueInMiddleOfBlock(BasicBlock *BB) {
// Set the DebugLoc of the inserted PHI, if available.
DebugLoc DL;
- if (const Instruction *I = BB->getFirstNonPHI())
- DL = I->getDebugLoc();
+ BasicBlock::iterator It = BB->getFirstN
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/123737
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
SLTozer wrote:
Ping
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
@@ -3595,15 +3595,26 @@ static llvm::StoreInst
*findDominatingStoreToReturnValue(CodeGenFunction &CGF) {
llvm::BasicBlock *IP = CGF.Builder.GetInsertBlock();
if (IP->empty()) return nullptr;
-// Look at directly preceding instruction, skipping bitcasts and lifetim
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
SLTozer wrote:
With the prior patches having landed, this is now in a state where it could
land; I've rebased it onto main, meaning all of the code diff in this PR is now
relevant.
https://github.com/llvm/llvm-project/pull/118026
___
cfe-commits mail
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/118026
>From beb5e1955f9b0aa106512bb10d4543232e4d4cfa Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Thu, 28 Nov 2024 13:53:20 +
Subject: [PATCH] Enable -fextend-lifetimes at -Og
---
clang/docs/CommandGuide/c
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/124767
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
SLTozer wrote:
Apologies, I didn't expect this PR to trigger a failure and didn't catch the
triggered failures; thank you for catching this so quickly!
https://github.com/llvm/llvm-project/pull/124767
___
cfe-commits mailing list
cfe-commits@lists.llv
SLTozer wrote:
I think the forward fix has been identified, and importantly it will take a bit
longer to revert than to submit the fix (adding triples to the tests), because
there are 4 commits that need to be reverted and I need to run a quick build to
ensure that nothing is broken by the rev
Author: Stephen Tozer
Date: 2025-01-28T19:34:41Z
New Revision: de9b0ddedc43302117b15518ca21f3341cf6b5ff
URL:
https://github.com/llvm/llvm-project/commit/de9b0ddedc43302117b15518ca21f3341cf6b5ff
DIFF:
https://github.com/llvm/llvm-project/commit/de9b0ddedc43302117b15518ca21f3341cf6b5ff.diff
LOG:
SLTozer wrote:
Merged in
[71ab44a8](https://github.com/llvm/llvm-project/commit/71ab44a8193c56e4ef9aede4d9bac8e14760c6c6),
via command line so as to provide proper commit author attribution to
@wolfy1961 (as this doesn't seem to be possible via the web interface).
https://github.com/llvm/llvm
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/11
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: Wolfgang Pieb
Date: 2025-01-28T12:30:31Z
New Revision: 4424c44c8c4ec8e071f5c5999fba216d36fb92c9
URL:
https://github.com/llvm/llvm-project/commit/4424c44c8c4ec8e071f5c5999fba216d36fb92c9
DIFF:
https://github.com/llvm/llvm-project/commit/4424c44c8c4ec8e071f5c5999fba216d36fb92c9.diff
LOG:
Author: Wolfgang Pieb
Date: 2025-01-28T10:08:02Z
New Revision: 71ab44a8193c56e4ef9aede4d9bac8e14760c6c6
URL:
https://github.com/llvm/llvm-project/commit/71ab44a8193c56e4ef9aede4d9bac8e14760c6c6
DIFF:
https://github.com/llvm/llvm-project/commit/71ab44a8193c56e4ef9aede4d9bac8e14760c6c6.diff
LOG:
SLTozer wrote:
As with the previous commit #11, this has been merged in
[4424c44c](https://github.com/llvm/llvm-project/commit/4424c44c8c4ec8e071f5c5999fba216d36fb92c9),
via command line so as to provide proper commit author attribution to
@wolfy1961 (as this doesn't seem to be possible vi
Author: Stephen Tozer
Date: 2025-01-28T22:51:30Z
New Revision: 548ecde42886149dd4d69366d7c2dc02076a7083
URL:
https://github.com/llvm/llvm-project/commit/548ecde42886149dd4d69366d7c2dc02076a7083
DIFF:
https://github.com/llvm/llvm-project/commit/548ecde42886149dd4d69366d7c2dc02076a7083.diff
LOG:
SLTozer wrote:
I made a mistake and missed one of the tests in the fix-forward commit - the
[build with the fix
commit](https://luci-milo.appspot.com/ui/p/fuchsia/builders/toolchain.ci/clang-windows-x64/b8724466517233166081/overview)
only recently finished, so I've pushed up a second fix now -
Author: Stephen Tozer
Date: 2025-01-28T13:21:41Z
New Revision: 8ad9e1ecb7e565c2f99b4ef67517d1f37d1c80d0
URL:
https://github.com/llvm/llvm-project/commit/8ad9e1ecb7e565c2f99b4ef67517d1f37d1c80d0
DIFF:
https://github.com/llvm/llvm-project/commit/8ad9e1ecb7e565c2f99b4ef67517d1f37d1c80d0.diff
LOG:
SLTozer wrote:
Pushed up a fix-forward commit,
[8ad9e1e](https://github.com/llvm/llvm-project/commit/8ad9e1ecb7e565c2f99b4ef67517d1f37d1c80d0),
that should fix the above issues; if issues still remain, I'll revert and
recommit after fixing offline.
https://github.com/llvm/llvm-project/pull/11
https://github.com/SLTozer created
https://github.com/llvm/llvm-project/pull/124767
This patch contains a number of changes relating to the above flag; primarily
it updates comment references to the old flag names, "-fextend-lifetimes" and
"-fextend-this-ptr" to refer to the new names,
"-fext
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/124767
>From 3afa9cafe2f39ae28dd0c017c9dcd00dc6656143 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Tue, 28 Jan 2025 15:22:22 +
Subject: [PATCH 1/2] [Clang][LLVM] Cleanup docs and comments relating to
-fexten
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/110102
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
SLTozer wrote:
Looks like the fix commit is passing everywhere, marking this one closed by
[4424c44c](https://github.com/llvm/llvm-project/commit/4424c44c8c4ec8e071f5c5999fba216d36fb92c9)
and
[8ad9e1e](https://github.com/llvm/llvm-project/commit/8ad9e1ecb7e565c2f99b4ef67517d1f37d1c80d0).
http
SLTozer wrote:
> Commented on the commit but I'm not sure if that sends notifications.
I saw it, thanks - I'll likely update the tests afterwards, though for now I'm
keeping an eye on the bots/merge to make sure things turn/stay green.
https://github.com/llvm/llvm-project/pull/110102
___
@@ -412,6 +412,20 @@ New Compiler Flags
only for thread-local variables, and none (which corresponds to the
existing ``-fno-c++-static-destructors`` flag) skips all static
destructors registration.
+- The ``-fextend-variable-liveness`` flag has been added to allow for imp
SLTozer wrote:
Post-holiday ping; I think all the outstanding comments have been addressed,
with some minor changes since the last round of review - primarily referencing
the new CodeGenOpts settings, and removing the O0-related logic.
https://github.com/llvm/llvm-project/pull/110102
_
@@ -113,6 +113,8 @@ static void substituteOperandWithArgument(Function *OldF,
for (Use *Op : OpsToReplace)
UniqueValues.insert(Op->get());
+ // TODO: Preserve range metadata with ranges
+
SLTozer wrote:
Unrelated change sneaking in?
https://github.com
https://github.com/SLTozer edited
https://github.com/llvm/llvm-project/pull/131743
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SLTozer approved this pull request.
Some minor comments, LGTM once addressed.
https://github.com/llvm/llvm-project/pull/131743
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cf
@@ -1837,13 +1837,13 @@ attributes #2 = { nounwind "target-cpu"="tahiti" }
!llvm.module.flags = !{!0}
!0 = !{i32 1, !"amdhsa_code_object_version", i32 500}
;.
-; HSA: attributes #[[ATTR0:[0-9]+]] = { nounwind }
+; HSA: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nosync n
SLTozer wrote:
> IIRC, there are still some open review threads in this PR, mainly related to
> reducing duplication in the symbolication bits.
Yes, this part won't land quite yet; as I'm kicking these reviews back into
action, I'll be pushing up some other reviews to come before this one in t
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/107279
>From 55115af37710a80abbc2ec2ca5d96d1eea5827b4 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 4 Sep 2024 12:23:52 +0100
Subject: [PATCH 1/4] Add conditionally-enabled DebugLocKinds
---
clang/lib/CodeG
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/118026
>From c0369e1b36594d5d6f6c29d7ebb7319d79edf55b Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Thu, 28 Nov 2024 13:53:20 +
Subject: [PATCH 1/2] Enable -fextend-lifetimes at -Og
---
clang/docs/CommandGui
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/118026
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/107279
>From 55115af37710a80abbc2ec2ca5d96d1eea5827b4 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 4 Sep 2024 12:23:52 +0100
Subject: [PATCH 1/3] Add conditionally-enabled DebugLocKinds
---
clang/lib/CodeG
SLTozer wrote:
> Any ideas why we might be hitting this issue? This happens during a built
> with `-Og`, and issue does not remove if I remove `-Og` flag.
At a glance that looks like it could be related, we generate (no-op) cleanup as
part of -fextend-variable-liveness. I'll look into this, bu
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/136867
>From 90c09c8326077a1ba6797519bbefd014f32b5beb Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 23 Apr 2025 14:24:50 +0100
Subject: [PATCH 1/3] [Clang] Emit Fake Uses before musttail calls
Fixes the issu
Author: Stephen Tozer
Date: 2025-04-25T00:36:28+01:00
New Revision: fdbf073a86573c9ac4d595fac8e06d252ce1469f
URL:
https://github.com/llvm/llvm-project/commit/fdbf073a86573c9ac4d595fac8e06d252ce1469f
DIFF:
https://github.com/llvm/llvm-project/commit/fdbf073a86573c9ac4d595fac8e06d252ce1469f.diff
Author: Stephen Tozer
Date: 2025-04-30T11:39:29+01:00
New Revision: 92195f6fc873cd27a5aa0852252dfe44ccdc6ea0
URL:
https://github.com/llvm/llvm-project/commit/92195f6fc873cd27a5aa0852252dfe44ccdc6ea0
DIFF:
https://github.com/llvm/llvm-project/commit/92195f6fc873cd27a5aa0852252dfe44ccdc6ea0.diff
SLTozer wrote:
> It would be great to also have some documentation that shows step by step how
> you've used these abilities to find the bugs you've found.
SGTM - I think that should come in a separate patch. I have a branch with all
my changes on it (not stable, I rebase semi-regularly)
[her
SLTozer wrote:
> Is this ready to go? I can't recall if we ultimately accepted the RFC.
My interpretation of the RFC is that there was a general acceptance of `-Og =
-O1 + -fextend-variable-liveness`, so with the acceptance here I believe this
is ready!
https://github.com/llvm/llvm-project/pu
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/118026
>From 351971bbff4e77b0f36cd92cd1a881584d17a9e7 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Thu, 28 Nov 2024 13:53:20 +
Subject: [PATCH 1/2] Enable -fextend-lifetimes at -Og
---
clang/docs/CommandGui
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/136867
>From 90c09c8326077a1ba6797519bbefd014f32b5beb Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 23 Apr 2025 14:24:50 +0100
Subject: [PATCH 1/2] [Clang] Emit Fake Uses before musttail calls
Fixes the issu
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/136867
>From 90c09c8326077a1ba6797519bbefd014f32b5beb Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 23 Apr 2025 14:24:50 +0100
Subject: [PATCH 1/4] [Clang] Emit Fake Uses before musttail calls
Fixes the issu
SLTozer wrote:
> I did some downstream testing with "-Og" since this patch and noticed that
> e.g.
Thanks for catching these - in terms of legalizer support it looks like
everything relevant is implemented except for `SoftenFloatOperand`, so that
should be a straightforward fix. The target-sp
@@ -169,6 +169,47 @@ See the discussion in the section about
:ref:`merging locations` for examples of when the rule for
dropping locations applies.
+.. _NewInstLocations:
+
+Setting locations for new instructions
+--
+
+Whenever a new instru
SLTozer wrote:
> Hi, I believe this merge is breaking external projects because you added
> `#include "llvm/Config/config.h"` which is not available (see comments at the
> top of that file).
Reverted for now, will look into relanding without the inclusion issue later.
https://github.com/llvm/
https://github.com/SLTozer updated
https://github.com/llvm/llvm-project/pull/107279
>From 55115af37710a80abbc2ec2ca5d96d1eea5827b4 Mon Sep 17 00:00:00 2001
From: Stephen Tozer
Date: Wed, 4 Sep 2024 12:23:52 +0100
Subject: [PATCH 1/6] Add conditionally-enabled DebugLocKinds
---
clang/lib/CodeG
Author: Stephen Tozer
Date: 2025-04-25T12:55:10+01:00
New Revision: 2dc6e98169baeb1f73036da0ea50fd828d8323d0
URL:
https://github.com/llvm/llvm-project/commit/2dc6e98169baeb1f73036da0ea50fd828d8323d0
DIFF:
https://github.com/llvm/llvm-project/commit/2dc6e98169baeb1f73036da0ea50fd828d8323d0.diff
https://github.com/SLTozer closed
https://github.com/llvm/llvm-project/pull/136867
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
101 - 200 of 223 matches
Mail list logo
