== This Week ==
* Setting up laptop, desk, mailing lists, SSH keys, Launchpad, etc
* Livermore Loops Benchmark
* It fails on some types of builds but not others
* Failure is due to LTO, problem identified, trying to get a reduced case
* Getting LLVM building nightly and running the test-suite
* Got where we are with ARM and others regarding buildbots
* LLVM compiles and checks on pandaboard, need test-suite
* Got a Chromebook to run the test-suite
* LLVM web presence (pages, svn repositories, etc) dead
== Next Week ==
* Reduce the LTO bug on Livermore kernels
* Report it once I have more information and a reduced case
* Try to fix it
* Checkout Zorg (LLVM's build system integrator)
* Force builds to use LNT (new test harness, which passes)
* Someone else is looking into setting up consistent LTO tests to the
rest already
* Get Ubuntu on Chromebook and build LLVM on it + test-suite
* Understand the CBuild infrastructure and setup a board to do it
automatically
* Revise the current LLVM build to be continuous and check-all
* Help Nadav Rotem (Apple) with the cost model for vectorization on ARM
VFP/NEON
* Meeting with ARM and "le French" to finalize the EuroLLVM 2013 and send
the CFP
== Future ==
* Validate vectorization for ARM after cost model is reasonably accurate
* Vectorization will be turned on by default as of 3.3 on -O2 or higher
* Reduced strength with -Os, but still on
* Make Livermore Loops vectorize
* Allow vectorization to detect global structures as safe
* Allow floating point vectorization if using -unsafe-math
* Plan a workflow to report / fix bugs that our internal CBuild finds on
ARM
* Liaise with ARM on what processes can be shared / commoned-up
* Try to be less verbose on activity reports...
== Blueprints ==
Initial Current Actual
fix-gcc-multiarch-testing 31 Dec 2012 31 Jan 2013
== Progress ==
* Short week, working Weds-Fri
* Admin
* Welcomed Renato to the team, and discussions with him about LLVM
* Initial project setup for llvm-linaro.
* Finished putting current set of card drafts into Jira
* Did GCC merges
* Backport of Thumb literal pool issues to 4.7.
== Next week ==
* Prepare Cortex Strings release
* Ensure GCC backports are up to date.
* Merge reviewing week.
* Catch up on outstanding cards.
== Future ==
* Run HOT/COLD partitioning benchmarks
* Analyse ARM results
* On x86_64 to see what the actual benefit we could get
* fix-gcc-multiarch-testing
* Come up with strawman proposal for updating testsuite to handle
testing with varying command-line options.
--
Matthew Gretton-Dann
Toolchain Working Group, Linaro
Hi all!
I use the arm-linux-gnueabihf-ct-ng.config in gcc-linaro-arm-linux-gnueabihf-4.7-2012.12-20121214_win32.zip to rebuild linaro-toolchain,but failured with the fellow messages:
[INFO ] Performing some trivial sanity checks
[INFO ] Build started 20130104.110952
[INFO ] Building environment variables
[WARN ] Directory '/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/tarballs' does not exist.
[WARN ] Will not save downloaded tarballs to local storage.
[EXTRA] Preparing working directories
[EXTRA] Installing user-supplied crosstool-NG configuration
[EXTRA] =================================================================
[EXTRA] Dumping internal crosstool-NG configuration
[EXTRA] Building a toolchain for:
[EXTRA] build = i686-pc-linux-gnu
[EXTRA] host = i586-mingw32msvc
[EXTRA] target = arm-linux-gnueabihf
[EXTRA] Dumping internal crosstool-NG configuration: done in 0.26s (at 00:04)
........
[EXTRA] Saving state to restart at step 'binutils'...
[INFO ] =================================================================
[INFO ] Installing binutils
[EXTRA] Configuring binutils
[EXTRA] Building binutils
[ERROR] /home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/src/binutils-2.22/gold/arm.cc:2173: internal compiler error: in make_rtl_for_nonlocal_decl, at cp/decl.c:4971
[ERROR] make[5]: *** [arm.o] Error 1
[ERROR] make[4]: *** [all-recursive] Error 1
[ERROR] make[3]: *** [all] Error 2
[ERROR] make[2]: *** [all-gold] Error 2
[ERROR] make[1]: *** [all] Error 2
[ERROR]
[ERROR] >>
[ERROR] >> Error happened in: main[scripts/crosstool-NG.sh]
[ERROR] >>
[ERROR] >> For more info on this error, look at the file: 'build.log'
[ERROR] >> There is a list of known issues, some with workarounds, in:
[ERROR] >> 'share/doc/ct-ng-linaro-1.13.1-4.7-2012.12-20121214/B - Known issues.txt'
[ERROR]
[ERROR] Build failed in step 'Extracting and patching toolchain components'
[ERROR]
[ERROR] (elapsed: 57:51.80)
make: *** [build] 错误 2
And in build.log:
..........
[ALL ] /home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/src/binutils-2.22/gold/arm.cc:11999: instantiated from here
[ERROR] /home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/src/binutils-2.22/gold/arm.cc:2173: internal compiler error: in make_rtl_for_nonlocal_decl, at cp/decl.c:4971
[ALL ] Please submit a full bug report,
[ALL ] with preprocessed source if appropriate.
[ALL ] See <URL:http://www.mingw.org/bugs.shtml> for instructions.
[ERROR] make[5]: *** [arm.o] Error 1
[ALL ] make[5]: *** Waiting for unfinished jobs....
[ALL ] mv -f .deps/powerpc.Tpo .deps/powerpc.Po
[ALL ] make[5]: Leaving directory `/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/arm-linux-gnueabihf/build/build-binutils/gold'
[ERROR] make[4]: *** [all-recursive] Error 1
[ALL ] make[4]: Leaving directory `/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/arm-linux-gnueabihf/build/build-binutils/gold'
[ERROR] make[3]: *** [all] Error 2
[ALL ] make[3]: Leaving directory `/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/arm-linux-gnueabihf/build/build-binutils/gold'
[ERROR] make[2]: *** [all-gold] Error 2
[ALL ] make[2]: Leaving directory `/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/arm-linux-gnueabihf/build/build-binutils'
[ERROR] make[1]: *** [all] Error 2
[ALL ] make[1]: Leaving directory `/home/ljc/ct-ng/crosstool-ng-linaro-1.13.1-4.7-2012.12-20121214/mybuild/.build/arm-linux-gnueabihf/build/build-binutils'
[ERROR]
[ERROR] >>
[ERROR] >> Error happened in: main[scripts/crosstool-NG.sh]
[ERROR] >>
[ERROR] >> For more info on this error, look at the file: 'build.log'
[ERROR] >> There is a list of known issues, some with workarounds, in:
[ERROR] >> 'share/doc/ct-ng-linaro-1.13.1-4.7-2012.12-20121214/B - Known issues.txt'
[ERROR]
[ERROR] Build failed in step 'Extracting and patching toolchain components'
[ERROR]
[ERROR] (elapsed: 57:51.80)
Has anybody met this problem too?
Dear All,
When doing prelink I got following error.
/a.out
/a.out: R_ARM_TLS_DTPMOD32 reloc in executable?
Gcc version 4.6
I have following question:
1. What this relocation do. ?
2. Is it problem in tool chain ?
3. Are we need to fix this in Prelink utils
Thanks
Hi,
The following code fails to build with OE Aarch64 toolchain with
current kernel headers. While ugly, the code is a reduced testcase
from fuse build failure (
https://bugs.launchpad.net/linaro-oe/+bug/1087757 ) and the same fuse
code compiles on all other architectures. Before I send a workaround
for upstream, I'd like to know how we can end up with different
definitions for int64_t when that happens on no other architectures -
something wrong with the generic kernel headers?
Testcase:
#include <sys/types.h>
#define __s64 int64_t
#include <signal.h>
int main(int argc, char **argv)
{
int64_t x=4;
return x;
}
Failure:
/data/oe/build/tmp-eglibc/sysroots/x86_64-linux/usr/bin/aarch64-oe-linux/aarch64-oe-linux-gcc
-save-temps --sysroot=/data/oe/build/tmp-eglibc/sysroots/genericarmv8
-o test test.c
In file included from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/asm-generic/types.h:7:0,
from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/asm/types.h:1,
from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/linux/types.h:4,
from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/asm/sigcontext.h:19,
from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/bits/sigcontext.h:27,
from
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/signal.h:338,
from test.c:4:
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/asm-generic/int-ll64.h:29:44:
error: conflicting types for 'int64_t'
In file included from test.c:2:0:
/data/oe/build/tmp-eglibc/sysroots/genericarmv8/usr/include/sys/types.h:197:13:
note: previous declaration of 'int64_t' was here
Summary:
* Follow-up shrink-up and branch cost related work.
* Investigate "MSR FPEXC, r2" assemble fail issues.
Details:
1. RM toolchain related work.
2. Investigate "MSR FPEXC, r2" assemble fail issues.
* According to Reference Manual, should use VMSR/VMRS to access
FPEXC register other than MSR/MRS.
* Binutils 2.23.1 or later had enhanced VMRS/VMSR to accept FPEXC register.
3. Rebase and test the shrink-wrap patches.
4. Read codes about impact of branch-cost.
Plan:
* Follow up the shrink-wrap and branch cost related work.
Planed leaves:
* Jan. 1-3, 2013
Best Regards!
-Zhenqiang
Hi all!
I'm rebuildding linaro-toolchain with ct-ng,I expect use the configure file in linaro-toolchain-binaries ,but I can't read the options in it's arm-linux-gnueabihf-ct-ng.config.
How I can do?
Thanks!
== Progress ==
* 64-bits ops in Neon: re-posted patch after a bit of cleanup
requested by upstream.
Investigated Fortran regression: confirmed unrelated to this
patched, and fixed upstream.
* disable-peeling: launched jobs under cbuild to perform initial benchmarking
* builtin_bswap16 backport to linaro-4.7: almost OK, missing some
pending validation results.
* cbuild: investigated some reporting problems which were slowing down
the branch review process.
== Next ==
* handle 64-bits bitops in Neon feedback from upstream if any
* finish builtin_bswap16 backport
* benchmark with vectorizer cost model enabled with its default configuration
== Future ==
Back on January 7th, after the end of the world.
== Blueprints ==
Initial Current Actual
initial-aarch64-backport 31 Oct 2012 31 Dec 2012 21 Dec 2012
aarch64-baremetal-testing 31 Oct 2012 31 Dec 2012 21 Dec 2012
backport-fma-intrinsic 31 Dec 2012 Brice
fused-multiply-add-support 31 Dec 2012 Brice
gcc-investigate-lra-for-arm 31 Dec 2012 Brice
fix-gcc-multiarch-testing 31 Dec 2012 31 Jan 2013
== Progress ==
* Admin
* Interviewing
* Welcomed Brice Dobry to the working group
* Started inputting cards into cards.linaro.org
* Discussions about KVM
* Discussions about Toolchain/Platform interactions
* initial-aarch64-backport and aarch64-baremetal-testing
* Finished documentation
== Next (working) week ==
* Admin
* Finish loading card drafts into Jira.
* Welcome Renato to working group.
* Do monthly GCC merges
* Prepare Cortex Strings release
* Ensure GCC backports are up to date.
== Future ==
* Run HOT/COLD partitioning benchmarks
* Analyse ARM results
* On x86_64 to see what the actual benefit we could get
* fix-gcc-multiarch-testing
* Come up with strawman proposal for updating testsuite to handle
testing with varying command-line options.
--
Matthew Gretton-Dann
Linaro Toolchain Working Group
matthew.gretton-dann(a)linaro.org
