I will be sending the final 15 patches to bring all the
functionality from the LTO branch. I tried to divide the patches
along maintainer lines, but there are some overlaps
- C FE
Adds processing of -flto and -fwhopr.
- C++ FE
Adds a langhook used by need_assembler_name_p to handle
templates. This implements Jason's suggestion from
http://gcc.gnu.org/ml/gcc-patches/2009-09/msg00534.html
- Fortran FE
Adds processing of -flto and -fwhopr. Note that the
Fortran testsuite still shows ~140 failures with LTO
enabled. IIUC, most of these have corresponding bugs
filed and they are not LTO-related (in that LTO exposes
some bugs in the FE). We should discuss whether we want
to include Fortran support for LTO in 4.5.
- Java FE
There is no LTO support for Java, but since LTO uses
zlib, Java does not need to include it.
- LTO FE
This is the new front end for GIMPLE.
- Configury
Detection of libelf, LTO support, etc.
- Contrib
A minor patch to add lto-plugin to the timestamps.
- Documentation
Documentation for the different LTO options.
- Driver
The somewhat ugly changes to collect2 and the main gcc
driver to support spawning lto1 when we detect a .o file
with LTO sections in it.
- Streamer
The code needed to support the on-disk representation of
GIMPLE, CFG and cgraph.
- Langhooks
New langhooks to support manipulating ELF sections and
the new langhook for asking the FE if a DECL needs to
have an assembler name set.
- libiberty
I need help with this one. When the linker plugin is
enabled (if GCC is configured to use gold), LTO can
detect LTO objects inside archives via the callbacks it
gets from the linker. Since the linker plugin is a
shared object, and it uses libiberty functions, it needs
to use a shared libiberty.
Currently, we just force --enable-shared on libiberty,
but I would only want to do that if gold and lto are
enabled. We detect gold and lto support in the top
configure script, but how do I send that down to
libiberty's configure?
- Linker plugin
This is the set of callbacks that gold uses to talk with
lto1 and resolve symbols out of object files in the link.
- Middle end
All sorts of changes, mostly in the callgraph, IPA and
pass manager.
- Testsuite
New tests for LTO and new dejagnu support for writing LTO
tests. Most of this has already been reviewed by Janis.
Major outstanding work:
- Support for debugging information. Currently, -flto -g does
not work (no debug information is generated). This is because
free_lang_data does not preserve enough information for the
debug hooks to emit it. I will fix this during stage 3.
- Outstanding bugzilla (http://tinyurl.com/yarz5q9).
- Some 'FIXME lto' markers. These are cleanup markers, several
related to WHOPR. We could probably debate whether to include
the WHOPR functionality in 4.5. It certainly is not very
different from LTO, but it does need some TLC.
Once I get all the approvals I need, I will request trunk to be
frozen for a day or so to give me time to commit the whole thing.
I already have a merged local tree and this exact same code is in
the LTO branch.
Thanks. Diego.
