Port-arm archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: RFH: let us know if PCRE2 with JIT will work for you



On Tue, 2023-11-28 at 04:11 -0800, Carlo Arenas wrote:
> Thanks,
> 
> Could you send me off-list the contents from RunTest.log and
> pcre2_jit_test.log so I could try to produce a fix?

No need. I got the test to pass. I forgot to run paxctl +m on the
pcre2_jit_test binary.

	$ paxctl +m .libs/*
	$ make check
	make  check-am
	make  
	make  all-am
	make  check-TESTS
	PASS: pcre2posix_test
	PASS: pcre2_jit_test
	PASS: RunTest
	PASS: RunGrepTest
	==============================================================
==============
	Testsuite summary for PCRE2 10.43-DEV
	==============================================================
==============
	# TOTAL: 4
	# PASS:  4
	# SKIP:  0
	# XFAIL: 0
	# FAIL:  0
	# XPASS: 0
	# ERROR: 0
	==============================================================
==============

> Carlo
> 
> PS. I think I have an idea where test64 might be going wrong and it
> is
> not armv6 specific, if you could comment out "test_src/sljitTest.c"
> line 7974 and run it again might help finding the issues that
> affected
> PCRE2,

I did that and the sljit tests seem to have all passed

	$ ./bin/sljit_test -v
	Run executable allocator test
	Run test1
	Run test2
	...
	no simd available, simd tests are skipped
	SLJIT tests: 1 (0%) tests FAILED on ARMv6 32bit (little endian
+ aligned) ABI:softfp (with fpu)

There were no error messages and the count of one failure is because
test64 was commented out, leaving the number of successful tests one
short from the expected total.

For completeness, here is some information about the test system I ran
these tests on.

	$ uname -a
	NetBSD  10.99.9 NetBSD 10.99.9 (RPI) #0: Fri Sep 29 12:27:57
NZDT 2023 
lloyd%riftsweeper.must-have-coffee.gen.nz@localhost:/home/lloyd/NetBSD/build/trun
k/objdir.evbarm/sys/arch/evbarm/compile/RPI evbarm
	$ sysctl hw
	hw.machine = evbarm
	hw.model = raspberrypi,model-b
	hw.ncpu = 1
	hw.byteorder = 1234
	hw.physmem = 469762048
	hw.usermem = 469762048
	hw.pagesize = 4096
	hw.disknames = ld0 sd0
	hw.machine_arch = earmv6
	...
	$ sysctl machdep
	machdep.debug = 0
	machdep.booted_device = ld0
	machdep.powersave = 1
	machdep.firmware_revision = 1576077262
	machdep.board_model = 0
	machdep.board_revision = 15
	machdep.cpu_arch = 6ZK
	machdep.cpu_id = 1091549031
	machdep.fpu_id = 1090592949
	machdep.fpu_present = 1
	machdep.hwdiv_present = 0
	machdep.neon_present = 0
	machdep.simd_present = 0
	...


Cheers,
Lloyd



Home | Main Index | Thread Index | Old Index