[llvm-dev] LLVM Weekly - #324, Mar 16th 2020

[Alex Bradbury via llvm-dev]

LLVM Weekly - #324, Mar 16th 2020
=================================

If you prefer, you can read a HTML version of this email at
<http://llvmweekly.org/issue/324>.

Welcome to the three hundred and twenty-fourth 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](https://www.linkedin.com/in/alex-bradbury/). Subscribe to future
issues at <http://llvmweekly.org> and pass it on to anyone else you think may
be interested. Please send any tips or feedback to <asb at asbradbury.org>, or
@llvmweekly or @asbradbury on Twitter.


## News and articles from around the web

John Regehr and Nuno Lopes have written a [blog post on Alive
2](https://blog.regehr.org/archives/1722). It will take an unoptimised
function in LLVM IR and either proves that it is refined by the optimized
version of the function, or else shows a set of inputs that illustrate a
failure of refinement.

EuroLLVM'20 [has now been cancelled due to
COVID-19](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139863.html).

LLVM 10.0.0-rc4 [has now been
tagged](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139905.html).


## On the mailing lists

* Richard Barton reports that [merging f18 in LLVM as flang has been delayed
until Mon 23rd
March](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139943.html).

* ORC JIT Weekly #8 [is
out](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139938.html). Basic
OrcV2 C Bindings have been added, and a number of MachO and COFF bugs fixed.

* Scott Constable and Zola Bridges
[wrote](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139842.html)
[separately](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139843.html)
about load value injection mitigation for x86.

* Nigel Perks [sent an update on plans from XMOS for the XCore
target](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139856.html).

* Chris Tetreault wrote an RFC on [refactoring the class hierarchy of
VectorType in the
IR](http://lists.llvm.org/pipermail/llvm-dev/2020-March/139811.html).

* Farzad Sadeghi [announced
cgrep](http://lists.llvm.org/pipermail/cfe-dev/2020-March/064836.html), a
grep-like tool for C language source files.


## LLVM commits

* The AVR backend was promoted to a standard backend, built by default.
[dfecec6](https://reviews.llvm.org/rGdfecec65e67).

* Macro fusion was implemented for the PowerPC backend.
[518292d](https://reviews.llvm.org/rG518292dbdfc).

* The attributor framework was extended to introduce propagation of known
information based on path exploration.
[bdcbdb4](https://reviews.llvm.org/rGbdcbdb48482).

* CPU definitions were added for the Fujitsu A64FX.
[c8cd1a9](https://reviews.llvm.org/rGc8cd1a994d2).

* `IEEEFloat::roundToIntegral` was made more standard conformant.
[14a1b80](https://reviews.llvm.org/rG14a1b80e044).

* Support was added for placing every basic block in a unique ELF text
section, allowing the linker to order them in an arbitrary sequence.
[4dfe92e](https://reviews.llvm.org/rG4dfe92e4654).

* AssemblerPredicates can now be combined with ORs.
[a26bd4e](https://reviews.llvm.org/rGa26bd4ec165).


## Clang commits

* Clang will now default to `-fno-common` for all targets, matching the GCC 10
default. [2c36c23](https://reviews.llvm.org/rG2c36c23f347).

* Clang's parsing of GNU C extended asm statement was improved.
[246398e](https://reviews.llvm.org/rG246398ece71).

* clang-tidy gained a new module to enforce standards specific to the
llvm-libc project. [eb41cc6](https://reviews.llvm.org/rGeb41cc61986).

## Other project commits

* libc now has a Linux implementation of `thrd_create`, `thrd_join`,
`mtc_lock`, and `mtx_unlock`.
[fd8c133](https://reviews.llvm.org/rGfd8c1336134).

* libunwind performance was improved.
[8ffdabd](https://reviews.llvm.org/rG8ffdabdb61e),
[c53c205](https://reviews.llvm.org/rGc53c2058ffb).

* PGO is now disabled for compiler-rt, as it doesn't officially support it.
[20dfcf1](https://reviews.llvm.org/rG20dfcf189d2).

* In MLIR, HasNoSideEffect was replaced with MemoryEffectInterface.
[0ddba0b](https://reviews.llvm.org/rG0ddba0bd59c).