Re: [RFA] Minor improvement to coremark, avoid unconditional jump to return

[Franz Sirl <Franz.Sirl-kernel@lauterbach.com>]

Am 2022-10-07 um 16:13 schrieb Jeff Law:
> 
> On 10/7/22 04:51, Franz Sirl wrote:
>> Am 2022-09-25 um 18:28 schrieb Jeff Law:
>>> This is a minor improvement for the core_list_find routine in coremark.
>>>
>>>
>>> Basically for riscv, and likely other targets, we can end up with an 
>>> unconditional jump to a return statement.    This is a result of 
>>> compensation code created by bb-reorder, and no jump optimization 
>>> pass runs after bb-reorder to clean this stuff up.
>>>
>>> This patch utilizes preexisting code to identify suitable branch 
>>> targets as well as preexisting code to emit a suitable return, so 
>>> it's pretty simple.  Note that when we arrange to do this 
>>> optimization, the original return block may become unreachable. So we 
>>> conditionally call delete_unreachable_blocks to fix that up.
>>>
>>> This triggers ~160 times during an x86_64 bootstrap.  Naturally it 
>>> bootstraps and regression tests on x86_64.
>>>
>>> I've also bootstrapped this on riscv64, regression testing with qemu 
>>> shows some regressions, but AFAICT they're actually qemu bugs with 
>>> signal handling/delivery -- qemu user mode emulation is not 
>>> consistently calling user defined signal handlers.  Given the same 
>>> binary, sometimes they'll get called and the test passes, other times 
>>> the handler isn't called and the test (of course) fails. I'll 
>>> probably spend some time to try and chase this down for the sake of 
>>> making testing easier.
>>>
>>>
>>> OK for the trunk?
>>
>> Hello Jeff,
>>
>> I've bisected this change to break a "profiledbootstrap" on x86_64 
>> like that:
>>
>> make[3]: Entering directory 
>> '/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/gcc'
>> /home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/./prev-gcc/xg++ -B/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/./prev-gcc/ -B/usr/x86_64-suse-linux/
>> bin/ -nostdinc++ 
>> -B/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/prev-x86_64-suse-linux/libstdc++-v3/src/.libs -B/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64
>> -suse-linux/prev-x86_64-suse-linux/libstdc++-v3/libsupc++/.libs 
>> -I/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/prev-x86_64-suse-linux/libstdc++-v3/include/x86_64-suse-linu
>> x 
>> -I/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/prev-x86_64-suse-linux/libstdc++-v3/include
>> -I/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/libstdc++-v3/libsupc++ -L
>> /home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/prev-x86_64-suse-linux/libstdc++-v3/src/.libs -L/home/fsirl/rpmbuild/BUILD/gcc-13.0.0+gitr13+2871/obj-x86_64-suse-linux/prev-x8
>> 6_64-suse-linux/libstdc++-v3/libsupc++/.libs  -fno-PIE -c   -O2 -g 
>> -fmessage-length=0 -D_FORTIFY_SOURCE=2 -funwind-tables 
>> -fasynchronous-unwind-tables -U_FORTIFY_SOURCE -fprofile-use 
>> -fprofile-reprod
>> ucible=parallel-runs -DIN_GCC     -fno-exceptions -fno-rtti 
>> -fasynchronous-unwind-tables -W -Wall -Wno-narrowing -Wwrite-strings 
>> -Wcast-qual -Wmissing-format-attribute -Woverloaded-virtual -pedantic
>> -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings 
>> -fno-common -DHAVE_CONFIG_H -I. -I. -I../../gcc -I../../gcc/. 
>> -I../../gcc/../include -I../../gcc/../libcpp/include 
>> -I../../gcc/../libcody
>>  -I../../gcc/../libdecnumber -I../../gcc/../libdecnumber/bid 
>> -I../libdecnumber -I../../gcc/../libbacktrace   -o cgraph.o -MT 
>> cgraph.o -MMD -MP -MF ./.deps/cgraph.TPo ../../gcc/cgraph.cc
>> ../../gcc/cgraph.cc: In member function 'cgraph_edge* 
>> cgraph_edge::first_speculative_call_target()':
>> ../../gcc/cgraph.cc:1166:1: error: EDGE_CROSSING incorrectly set 
>> across same section
>>  1166 | }
>>       | ^
>> ../../gcc/cgraph.cc:1166:1: error: No region crossing jump at section 
>> boundary in bb 19
>> during RTL pass: bbro
>> ../../gcc/cgraph.cc:1166:1: internal compiler error: verify_flow_info 
>> failed
>> 0xa7116e verify_flow_info()
>>         ../../gcc/cfghooks.cc:284
>> 0x1c64958 execute
>>         ../../gcc/bb-reorder.cc:2663
>>
>> In such a case, what do you need to reproduce it? I'm a mere user of 
>> the Suse RPM builds here, no idea if profiling needs any extra data to 
>> reproduce to bug.
> 
> Franz, it's been a long time.  Good to hear from you.
> 
> 
> I'll take a look.  I'll start with a profiledbootstrap and if that 
> doesn't reproduce I'll get in contact

Yes, long time... Things happened :-) But I'm still following GCC and 
report bugs as time permits.

I've opened 107182 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107182 
with the -freport-bug file and cgraph.gcda. I hope this is enough to 
reproduce.

Franz.