LLVM Weekly - #298, September 16th 2019
Welcome to the two hundred and ninety-eighth issue of LLVM Weekly, a weekly newsletter (published every Monday) covering developments in LLVM, Clang, and related projects. LLVM Weekly is brought to you by Alex Bradbury. Subscribe to future issues at https://llvmweekly.org and pass it on to anyone else you think may be interested. Please send any tips or feedback to asb@asbradbury.org, or @llvmweekly or @asbradbury on Twitter.
News and articles from around the web
LLVM 9.0.0 release candidate 5 has been tagged.
On the mailing lists
-
Michael Woerister has been implementing profile-guided optimisation support in the Rust compiler. He found there was no measurable performance gain and is seeking feedback as to why this might be.
-
Philip Reames outlined plans to enhance the loop vectorizer to handle multiple exit loops.
-
Varun Gandhi has a proposal to change the behaviour of lit.py's -v flag.
-
Bardia Mahjour shares loop optimisation working group meeting minutes, covering topics such as loop invariant code motion vs loop sink strategy.
-
Daan Sprenkels kicked off a discussion on protecting against side channel attacks on values.
LLVM commits
-
Support was added to the AArch64 backend for the
arm64_32watchOS ABI. r371722. -
LLVM gained infrastructure to performance profile guided validation of performance annotations, as used in the clang-misexpect tool. r371484.
-
A parser was committed for bitstream remarks. r371429.
-
The initial patch in a series intending to overhaul SelectionDAG atomic load/store handling has landed. r371441.
-
TableGen's CodeEmitter can now support instruction widths greater than 64 bits. r371928.
Clang commits
-
The constexpr interpreter was committed. r371834.
-
clang-format gained a new IndentGotoLabels style option. r371719.
-
Codegen was improved for
thread_localvariables. r371767.