[LLVMdev] LLVM Weekly - #79, Jul 6th 2015

[Alex Bradbury]

LLVM Weekly - #79, Jul 6th 2015
===============================

If you prefer, you can read a HTML version of this email at
<http://llvmweekly.org/issue/79>.

Welcome to the seventy-ninth 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](http://asbradbury.org). 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.

Last week I was in Berkeley for the second RISC-V conference. If you weren't
able to make it, worry not because I liveblogged both [day
one](http://www.lowrisc.org/blog/2015/06/second-risc-v-workshop-day-one/) and
[day
two](http://www.lowrisc.org/blog/2015/06/second-risc-v-workshop-day-two/).


## News and articles from around the web

Stephen Cross has [released
llvm-abi](http://article.gmane.org/gmane.comp.compilers.clang.devel/43502), a
library for generating LLVM IR that complies with platform ABIs.

This is a [rather cute implementation of Tetris in C++ header
files](http://blog.mattbierner.com/stupid-template-tricks-super-template-tetris/),
compatible with Clang.


## On the mailing lists

* Kevin Atkinson [asks whether to use MCJIT or
ORCJIT](http://article.gmane.org/gmane.comp.compilers.llvm.devel/87565). It
sounds like [ORC is working out well for the LLILC
team](http://article.gmane.org/gmane.comp.compilers.llvm.devel/87600).

* Chandler Carruth has kicked off a [discussion about the AliasAnalysis update
interface](http://article.gmane.org/gmane.comp.compilers.llvm.devel/87469) and
what should be done about it.

* In response to a question, Evgeny Astigeevich has given a useful guide to
[finding the control dependence
graph](http://article.gmane.org/gmane.comp.compilers.llvm.devel/87416).

* Manuel Klimek [is updating the LLVM Phabricator
install](http://article.gmane.org/gmane.comp.compilers.clang.devel/43558).


## LLVM commits

* The initial skeleton of the WebAssembly backend has been committed. It is
not yet functional. [r241022](http://reviews.llvm.org/rL241022).

* DIModule metadata nodes have been introduced. A DIModule is meant to be used
to record modules importaed by the current compile unit.
[r241017](http://reviews.llvm.org/rL241017).

* New exception handling intrinsics have been added for recovering and
restoring parent frames. [r241125](http://reviews.llvm.org/rL241125).


## Clang commits

* Clang gained support for the x86 builtin `__builtin_cpu_supports`.
[r240994](http://reviews.llvm.org/rL240994).

* The Clang man pages have been converted to Sphinx (from .pod).
[r241037](http://reviews.llvm.org/rL241037).


## Other project commits

* libcxx gained `shared_mutux`. [r241067](http://reviews.llvm.org/rL241067).

* LLD has gained some generally applicable optimisations. e.g. devirtualizing
SymbolBody and compacting its in-memory representation.
[r241001](http://reviews.llvm.org/rL241001).

* LLD's COFF linker can now link a working 64-bit debug build of Chrome.
chrome.dll takes 24 seconds (vs 48 seconds for linking it with MSVC).
[r241318](http://reviews.llvm.org/rL241318).

* LLDB grew an example of scripted steps in Python.
[r241216](http://reviews.llvm.org/rL241216).