== This Week ==
* TCWG-72 (2/10)
- Another iteration.
* LTO (4/10)
- Building spec with different combinations of -flto -flto-partition
-flto-compression-level
- Builds cleanly on AArch64, ICE's on ARM for 447.dealII with -flto
- Looked at segfault with 483.xalanpack
* TCWG-319 benchmarking (1/10)
- fp benchmark results on:
r1-a12: -0.04 %
cortex-a53: -0.09 %
cortex-a57: -0.14%
- int benchmarks completed running for reference revision on cortex-a15
- int benchmarks with-patch runnning on cortex-a15
* TCWG-310 benchmarking (1/10)
- r1-a12: 0.11%
- Benchmark jobs failed on a53, a57.
* Misc (2/10)
- Submitted patch to fix building 450.soplex
- Looked at how to write ipa pass, ipa-pure-const and ipa-cp.
- Meetings
== Next Week ==
- tcwg-319, tcwg-310 benchmarking
- tcwg-72
- LTO
== This week ==
* Bugzilla 68543 - [AArch64] Implement overflow arithmetic standard
names (7/10)
- Implemented signed and unsigned add, subtract, and multiply
overflow standard patterns
- Investigated testing of overflow patterns
* Misc (3/10)
- Conference calls
- Misc ARM Housekeeping
- Lost connectivity due to old ARM Unix VPN Client
== Next week ==
* TCWG-317 - Exploit wide add operations when appropriate for Aarch32
- Minor code update to address upstream comment
* Vacation until rest of year
== Progress ==
* GCC (4/10)
- pr68620 (fp16 transfers in big-endian mode)
- checking regressions observed on trunk
- had to revert my cleanup patch for target
attributes tests, not sure how to handle all
possible combinations of options/defaults
* Validation (1/10)
- small improvements in reporting
* Misc (conf calls, meetings, emails, ...) (3/10)
* Internal training (2/10)
== Next ==
* Validation
* GCC: bug fixes/cleanup
* One day off on Wed. [2/10]
# Progress #
* Enable gdb core file tests when testing remotely, TCWG-171.
I've almost had the conclusion that corefile remote testing can't be
done due to limitations in dejagnu and nfs mount testing
infrastructure. Need to write them down. [2/10]
* Mutli-arch follow-up work, teach AArch64 GDBserver understand ARM
breakpoint instructions. TCWG-460. Done. [3/10]
* Review ARM GDBserver software single step patch V7. Almost OK, except
some small things. [2/10]
* Misc, meeting, email, [1/10]
# Plan #
* TCWG-424, Draft a fix for the fail in random-signal.exp.
* TCWG-156, GDB test parity between AArch64 and X86_64.
* One day off on Wed. or Fri.
Planned absence:
* Dec 24-Jan 3.
--
Yao
== Progress ==
- PR66726 (2/10)
* Testing a patch
- PR63586 (2/10)
* Posted a patch
* Revised the patch based on testing
- LuaJIT (2/10)
* Setup nginx
* Still haven't figured out how to use mongodb with nginx (config
required).
- Misc (2/10)
* gcc/bug list
* LTO
- sick (2/10)
== Plan ==
* bug reports
* LTO
Hi,
when working with the Linaro patches I found that a particular commit
breaks our aarch64 kernel build.
The patch in question is that one:
commit be09330da9d0777c4a58568d137e3f8a3dbe0a0b
Author: Yvan Roux <yvan.roux(a)linaro.org>
Date: Tue Oct 27 21:18:19 2015 +0100
One of the things it attempts to change apparently is moving the .arch
specifiers in the assembler file from a global scope to individual
functions. What also happens though is that they seem to lose some
information after that transformation.
I observed that when building arch/arm64/crypto/aes-ce-cipher.c from
the Linux kernel. This code contains inline assembly like this:
static void aes_cipher_decrypt(struct crypto_tfm *tfm, u8 dst[], u8 const src[])
{
struct crypto_aes_ctx *ctx = crypto_tfm_ctx(tfm);
struct aes_block *out = (struct aes_block *)dst;
struct aes_block const *in = (struct aes_block *)src;
void *dummy0;
int dummy1;
kernel_neon_begin_partial(4);
__asm__(" ld1 {v0.16b}, %[in] ;"
" ld1 {v1.2d}, [%[key]], #16 ;"
" cmp %w[rounds], #10 ;"
" bmi 0f ;"
" bne 3f ;"
" mov v3.16b, v1.16b ;"
" b 2f ;"
"0: mov v2.16b, v1.16b ;"
" ld1 {v3.2d}, [%[key]], #16 ;"
"1: aesd v0.16b, v2.16b ;"
" aesimc v0.16b, v0.16b ;"
"2: ld1 {v1.2d}, [%[key]], #16 ;"
" aesd v0.16b, v3.16b ;"
" aesimc v0.16b, v0.16b ;"
"3: ld1 {v2.2d}, [%[key]], #16 ;"
" subs %w[rounds], %w[rounds], #3 ;"
" aesd v0.16b, v1.16b ;"
" aesimc v0.16b, v0.16b ;"
" ld1 {v3.2d}, [%[key]], #16 ;"
" bpl 1b ;"
" aesd v0.16b, v2.16b ;"
" eor v0.16b, v0.16b, v3.16b ;"
" st1 {v0.16b}, %[out] ;"
: [out] "=Q"(*out),
[key] "=r"(dummy0),
[rounds] "=r"(dummy1)
: [in] "Q"(*in),
"1"(ctx->key_dec),
"2"(num_rounds(ctx) - 2)
: "cc");
kernel_neon_end();
}
Now without this patch the compiler behaved like the following. It was
invoked with:
aarch64-linux-gnu-gcc -Wp,-MD,arch/arm64/crypto/.aes-ce-cipher.o.d
-nostdinc -isystem
/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/bin/../lib/gcc/aarch64-linux-gnu/5.2.1/include
-I/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/arch/arm64/include
-Iarch/arm64/include/generated/uapi -Iarch/arm64/include/generated
-I/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/include
-Iinclude -I/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/arch/arm64/include/uapi
-Iarch/arm64/include/generated/uapi
-I/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/include/uapi
-Iinclude/generated/uapi -include
/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/include/linux/kconfig.h
-I/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/arch/arm64/crypto
-Iarch/arm64/crypto -D__KERNEL__ -mlittle-endian -Wall -Wundef
-Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common
-Werror-implicit-function-declaration -Wno-format-security -std=gnu89
-mgeneral-regs-only -fno-delete-null-pointer-checks -O2
--param=allow-store-data-races=0 -Wframe-larger-than=2048
-fno-stack-protector -Wno-unused-but-set-variable
-fno-omit-frame-pointer -fno-optimize-sibling-calls
-fno-var-tracking-assignments -g -Wdeclaration-after-statement
-Wno-pointer-sign -fno-strict-overflow -fconserve-stack
-Werror=implicit-int -Werror=strict-prototypes -Werror=date-time
-DCC_HAVE_ASM_GOTO -Werror -march=armv8-a+crypto
-D"KBUILD_STR(s)=#s" -D"KBUILD_BASENAME=KBUILD_STR(aes_ce_cipher)"
-D"KBUILD_MODNAME=KBUILD_STR(aes_ce_cipher)" -c -o
arch/arm64/crypto/aes-ce-cipher.o
/var/fpwork/rschiele/crossbuild/builds/aarch64-linux-gnu/linux-next/srcdir/src/linux/arch/arm64/crypto/aes-ce-cipher.c
As a result it created a file for the assembler with the global
.arch armv8-a+fp+simd+crypto
at the beginning of the file.
After the patch it created individual
.arch armv8-a
at individual places.
It is not clear to me, why the extensions (fp+simd+crypto) got lost.
Is that intended, such that the code needs special adaption for inline
assembly using those extensions or is that loss of extensions a bug of
that patch?
Greetings!
Robert
== Progress ==
o Linaro GCC (6/10)
* More backports
* Compared our various flavor of validation results
* Looked at bug reports and mailing list questions on our last snapshot.
o Upstream work (2/10)
* Continue on sanitizing gfortran testsuite
o Misc (2/10)
* Various meetings
* Discussed benchmarking infra with Bernie
== Plan ==
o GCC 5.3 branch merge
o Continue on-going tasks
o Tuesday Off
Holiday [2/10]
Port to microinstance - TCWG-432 [1/10]
* Set up reporting for CPU2006
* Learned how to generate metadata, but not how to use it
Trigger benchmarks on backports - TCWG-352 [2/10]
* Figured out the rough shape
* Created, didn't test, rough implementation
TCWG-354 [3/10]
* Build/run scaffolding for CoreMark Pro
* Working for manual runs
Misc [2/10]
* Input validation for dispatcher script
* Meeting with Ade on LAVA benchmarking
* Meetings/mail/etc
=Plan=
Review, test, debug build-triggers-benchmark job
Check CoreMark Pro run configuration, enable for automatic runs
Review security with shared uinstance/main instance code
Expose more data, benchmarks to bundles
Debug/test Jenkins job in microinstance
Create bootable image for at least 1 target, or know what the problems are
Write up noise control report (if time)
More support for SPEC-on-Android?
* TCWG-72 (2/10)
- Added new target hook to generate target-specific divmod libfunc
- Builds cleanly now on x86_64, arm and arm-linux-gnueabihf
- Sent to tcwg list for review
* LTO spec2k6 build (2/10)
- Built speck26 with LTO
* Target hook (4/10)
- Completed with ASM_OUTPUT_LABEL_REF
- In progress - SIZE_ASM_OP to data hook, ASM_OUTPUT_SIZE_DIRECTIVE,
ASM_OUTPUT_MEASURED_SIZE
* Benchmarking (1/10)
- tcwg-319: Job in progress for fp benchmarks with patch
- tcwg-310 (loop peeling): Submitted job for running 252.eon
- Both the jobs failed due to lab downtime, need to be re-run.
- Received job template from Bernie for running benchmarks on cortex-a15.
* Misc (1/10)
- Meetings
== Next Week ==
- Continue benchmarking spec2k6 with LTO
- Look at bugs exposed by speck2k6 LTO build
- Benchmarking tcwg-319, tcwg-310
