<feed xmlns='http://www.w3.org/2005/Atom'>
<title>llvm-project.git, branch users/cdevadas/include-lr-split-insn-to-bb-prolog</title>
<subtitle>Unnamed repository; edit this file 'description' to name the repository.
</subtitle>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/'/>
<entry>
<title>[AMDGPU] Add liverange split instructions into BB Prolog</title>
<updated>2025-11-07T06:42:36+00:00</updated>
<author>
<name>Christudasan Devadasan</name>
<email>Christudasan.Devadasan@amd.com</email>
</author>
<published>2024-11-19T06:44:05+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=80232b5877a5b48e22881a45c84097299406b078'/>
<id>80232b5877a5b48e22881a45c84097299406b078</id>
<content type='text'>
The COPY inserted for liverange split during sgpr-regalloc
pipeline currently breaks the BB prolog during the subsequent
vgpr-regalloc phase while spilling and/or splitting the vector
liveranges. This patch fixes it by correctly including the
the LR split instructions during sgpr-regalloc and wwm-regalloc
pipelines into the BB prolog.
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The COPY inserted for liverange split during sgpr-regalloc
pipeline currently breaks the BB prolog during the subsequent
vgpr-regalloc phase while spilling and/or splitting the vector
liveranges. This patch fixes it by correctly including the
the LR split instructions during sgpr-regalloc and wwm-regalloc
pipelines into the BB prolog.
</pre>
</div>
</content>
</entry>
<entry>
<title>[CodeGen] Introduce MI flag for Live Range split instructions</title>
<updated>2025-11-07T05:39:32+00:00</updated>
<author>
<name>Christudasan Devadasan</name>
<email>Christudasan.Devadasan@amd.com</email>
</author>
<published>2024-11-14T08:41:44+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=a9587c1342ff0748ca19c6f35913365d09b9a635'/>
<id>a9587c1342ff0748ca19c6f35913365d09b9a635</id>
<content type='text'>
For some targets, it is required to identify the COPY instruction
corresponds to the RA inserted live range split. Adding the new
flag `MachineInstr::LRSplit` to serve the purpose.
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
For some targets, it is required to identify the COPY instruction
corresponds to the RA inserted live range split. Adding the new
flag `MachineInstr::LRSplit` to serve the purpose.
</pre>
</div>
</content>
</entry>
<entry>
<title>[libc] Add localtime_r to baremetal entrypoints (#166677)</title>
<updated>2025-11-07T05:18:26+00:00</updated>
<author>
<name>Prabhu Rajasekaran</name>
<email>prabhukr@google.com</email>
</author>
<published>2025-11-07T05:18:26+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=bf1b86698b41fcfa6875b668e8df8ea6d81081e5'/>
<id>bf1b86698b41fcfa6875b668e8df8ea6d81081e5</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>[Option] Fix simple subcommand with positional arguments (#166859)</title>
<updated>2025-11-07T04:57:56+00:00</updated>
<author>
<name>Steven Wu</name>
<email>stevenwu@apple.com</email>
</author>
<published>2025-11-07T04:57:56+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=41825fbf0e869ba357c713971f20eb1fcdafd09c'/>
<id>41825fbf0e869ba357c713971f20eb1fcdafd09c</id>
<content type='text'>
Fix subcommand detection when subcommand used with positional arguments.
When there is only one valid subcommand passed,
`ArgList::getSubCommand()` should return the correct subcommand even
there are other positionals passed.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Fix subcommand detection when subcommand used with positional arguments.
When there is only one valid subcommand passed,
`ArgList::getSubCommand()` should return the correct subcommand even
there are other positionals passed.</pre>
</div>
</content>
</entry>
<entry>
<title>[LifetimeSafety] Optimize fact storage with IDs and vector-based lookup (#165963)</title>
<updated>2025-11-07T04:27:33+00:00</updated>
<author>
<name>Utkarsh Saxena</name>
<email>usx@google.com</email>
</author>
<published>2025-11-07T04:27:33+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=c9b41699978132e83abf2ce4dd63eb21e69f846a'/>
<id>c9b41699978132e83abf2ce4dd63eb21e69f846a</id>
<content type='text'>
Optimize the FactManager and DataflowAnalysis classes by using vector-based storage with ID-based lookups instead of maps.

- Added a `FactID` type using the `utils::ID` template to uniquely identify facts
- Modified `Fact` class to store and manage IDs
- Changed `FactManager` to use vector-based storage indexed by block ID instead of a map
- Updated `DataflowAnalysis` to use vector-based storage for states instead of maps
- Modified lookups to use ID-based indexing for better performance

Improves compile time hit on long-tail targets like `tools/clang/lib/CodeGen/CMakeFiles/obj.clangCodeGen.dir/TargetBuiltins/RISCV.cpp.o`​ from [21%](http://llvm-compile-time-tracker.com/compare_clang.php?from=6e25a04027ca786b7919657c7df330a33985ceea&amp;to=20b42efa277c8b1915db757863e1fc26531cfd53&amp;stat=instructions%3Au&amp;sortBy=absolute-difference) to [3.2%](http://llvm-compile-time-tracker.com/compare_clang.php?from=6e25a04027ca786b7919657c7df330a33985ceea&amp;to=d2d1cd1109c3a85344457bfff6f092ae7b96b211&amp;stat=instructions%3Au&amp;sortBy=absolute-difference)</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Optimize the FactManager and DataflowAnalysis classes by using vector-based storage with ID-based lookups instead of maps.

- Added a `FactID` type using the `utils::ID` template to uniquely identify facts
- Modified `Fact` class to store and manage IDs
- Changed `FactManager` to use vector-based storage indexed by block ID instead of a map
- Updated `DataflowAnalysis` to use vector-based storage for states instead of maps
- Modified lookups to use ID-based indexing for better performance

Improves compile time hit on long-tail targets like `tools/clang/lib/CodeGen/CMakeFiles/obj.clangCodeGen.dir/TargetBuiltins/RISCV.cpp.o`​ from [21%](http://llvm-compile-time-tracker.com/compare_clang.php?from=6e25a04027ca786b7919657c7df330a33985ceea&amp;to=20b42efa277c8b1915db757863e1fc26531cfd53&amp;stat=instructions%3Au&amp;sortBy=absolute-difference) to [3.2%](http://llvm-compile-time-tracker.com/compare_clang.php?from=6e25a04027ca786b7919657c7df330a33985ceea&amp;to=d2d1cd1109c3a85344457bfff6f092ae7b96b211&amp;stat=instructions%3Au&amp;sortBy=absolute-difference)</pre>
</div>
</content>
</entry>
<entry>
<title>Use `llvm::unique_function` in the async APIs (#166727)</title>
<updated>2025-11-07T04:02:47+00:00</updated>
<author>
<name>Chandler Carruth</name>
<email>chandlerc@gmail.com</email>
</author>
<published>2025-11-07T04:02:47+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=5314d99d30528fde9052a39cfc7948c2919cb6b6'/>
<id>5314d99d30528fde9052a39cfc7948c2919cb6b6</id>
<content type='text'>
This is needed to allow using these APIs with callable objects that
transitively capture move-only constructs. These come up very widely
when writing concurrent code such a `std::future`, `std::promise`,
`std::unique_lock`, etc.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This is needed to allow using these APIs with callable objects that
transitively capture move-only constructs. These come up very widely
when writing concurrent code such a `std::future`, `std::promise`,
`std::unique_lock`, etc.</pre>
</div>
</content>
</entry>
<entry>
<title>[CodeGen] Fix a warning</title>
<updated>2025-11-07T03:58:59+00:00</updated>
<author>
<name>Kazu Hirata</name>
<email>kazu@google.com</email>
</author>
<published>2025-11-07T03:58:59+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=f8e9b89ae07aebf09822b61f18966d2316e7b40e'/>
<id>f8e9b89ae07aebf09822b61f18966d2316e7b40e</id>
<content type='text'>
This patch fixes:

  clang/lib/CodeGen/CGBuiltin.cpp:1216:13: error: unused variable
  'CAT' [-Werror,-Wunused-variable]
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This patch fixes:

  clang/lib/CodeGen/CGBuiltin.cpp:1216:13: error: unused variable
  'CAT' [-Werror,-Wunused-variable]
</pre>
</div>
</content>
</entry>
<entry>
<title>[RISCV] Move remat.ll test from riscv32 to riscv64. NFC</title>
<updated>2025-11-07T03:44:16+00:00</updated>
<author>
<name>Luke Lau</name>
<email>luke@igalia.com</email>
</author>
<published>2025-11-07T03:44:16+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=630f43a305b9e48d6c14ecb53a50e511402e2efb'/>
<id>630f43a305b9e48d6c14ecb53a50e511402e2efb</id>
<content type='text'>
This is to allow us to test rematting i64 lds in #166774. Checked and
we're still rematting the same lis on rv64 in this test.
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This is to allow us to test rematting i64 lds in #166774. Checked and
we're still rematting the same lis on rv64 in this test.
</pre>
</div>
</content>
</entry>
<entry>
<title>[LLDB][Editline] empty current line before `el_wgets` (#165830)</title>
<updated>2025-11-07T03:40:38+00:00</updated>
<author>
<name>Kiva</name>
<email>imkiva@islovely.icu</email>
</author>
<published>2025-11-07T03:40:38+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=ca0866bf6dcee8856a8d9a6d88ec236c9c68baa1'/>
<id>ca0866bf6dcee8856a8d9a6d88ec236c9c68baa1</id>
<content type='text'>
This PR fixes #157637 by printing ANSI sequence "\r\x1b[K" to empty the
line before calling to `el_wgets`. Later when `el_wgets` prints the real
prompt, all wrongly printed characters are removed from the terminal.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This PR fixes #157637 by printing ANSI sequence "\r\x1b[K" to empty the
line before calling to `el_wgets`. Later when `el_wgets` prints the real
prompt, all wrongly printed characters are removed from the terminal.</pre>
</div>
</content>
</entry>
<entry>
<title>[Github] Use zstd to compress CI container images (#166879)</title>
<updated>2025-11-07T03:39:23+00:00</updated>
<author>
<name>Aiden Grossman</name>
<email>aidengrossman@google.com</email>
</author>
<published>2025-11-07T03:39:23+00:00</published>
<link rel='alternate' type='text/html' href='https://git.belthelziquor.com/llvm-project.git/commit/?id=0e46b4189031c580f636b377987c950c5e38849e'/>
<id>0e46b4189031c580f636b377987c950c5e38849e</id>
<content type='text'>
zstd usually offers better compression ratios, but also offers much
higher decompression speeds, which can tangibly impact image unpack
times. Enable this by default given it is a two line change that is
transparent at image pull time given most container runtimes these days
supported zstd compressed images.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
zstd usually offers better compression ratios, but also offers much
higher decompression speeds, which can tangibly impact image unpack
times. Enable this by default given it is a two line change that is
transparent at image pull time given most container runtimes these days
supported zstd compressed images.</pre>
</div>
</content>
</entry>
</feed>
