This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
[Cec-weeklies] ST40 tools #1517,#1518,#1519
- From: Christian Bruel <christian dot bruel at st dot com>
- To: "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 7 May 2015 10:43:51 +0200
- Subject: [Cec-weeklies] ST40 tools #1517,#1518,#1519
- Authentication-results: sourceware.org; auth=none
- References: <553108BC dot 9060705 at st dot com>
Highlights
-----------
- Beta Breakpad delived
- Upstreaming (almost accepted)
Issues
-------
- Current FSF trunk boostrap broken (delay validations)
Completed
----------
- Fixed FSF attribute optimize inlining
- Fixed FSF bad handling of ipa-cp (inlining of functions with constant
propagation)
- Respinned ARM attribute target (means a week or merging/validation)
Details
--------
- Breakpad-SH4
* Beta version delivered to customer for pre-integration testing
. No news good news. Seems to be integrated into the customer's box.
. Google upstreaming. All issues addresses. No objection for landing
. Cross and on boards testing validation.
. Rewrote syscall layer for more preprocessor factorization
(denser but more obfuscated)
- Misc ST40
obsp,seinit integration
* Received a B2147 (Alicante) Board
Board setup done.
* Meeting (Pierre Gobin, Bruno Tisseran, Seraphin Bonnaffe, Loic
Pallardy, Denis Hory, Ludovic Barre) to discuss SEINIT future
. Originally maintained in the OBSP by Bristol
. Integrated with Docsis modem using a "bpl elf patching tool" by
OS team (remoteproc)
. Knowledge seems to be lost
-> seinit.o integration ?
-> Can't rebuild seinit.o. Who is in charge of the sources
(currently embedded in the obsp) ?
proposal:
- CEC replaces the elf patching tool by a board spec and a set
of linker commands.
- CEC Volunteer to get the hands on the seinit sources. (own?)
- First Experiment to take them out of the OBSP tree (lot of
dependencies).
- Issue with MMU/TLB/OS setups. Unknown specifications
- ST40 STlinux Toolchain
Working on GCC 5.0.0 distribution port
Progressing on the packages. glibc link error fixes
- GCC generic and x-targets
- Working on ipa-cp inconsistencies (small functions not inlined)
Reduced to a generic case
x86 fixes accepted.
- Working on function alignments with attribute ((optimize))
Generic part fixed.
Need hooks into the target backends (spread to all backends)
x86, aarch64, alpha fixed
other targets still under approvals or still undetected
- Misc
Re-spined attribute-target patch
Fixed documentation