Re: [PATCH v7 00/13] GLIBC LoongArch PATCHES

[WANG Xuerui <i.swmail@xen0n.name>]

On 7/24/22 19:51, Xi Ruoyao wrote:
> On Sun, 2022-07-24 at 17:49 +0800, WANG Xuerui wrote:
>
>> Of these, the ifunc failures are "expected" by you, the
>> elf/check-abi-libc diff is trivial (maybe you just didn't rebase as
>> frequently):
>>
>> --- ../sysdeps/unix/sysv/linux/loongarch/lp64/libc.abilist 2022-07-23
>> 14:45:57.490029442 +0800
>> +++ /home/xenon/src/glibc/build/libc.symlist    2022-07-24
>> 13:44:10.416642655 +0800
>> @@ -496 +496 @@ GLIBC_2.36 _mcount F
>> -GLIBC_2.36 _nl_default_dirname D 0x12
>> +GLIBC_2.36 _nl_default_dirname D 0x17
>> @@ -541,0 +542,3 @@ GLIBC_2.36 alphasort64 F
>> +GLIBC_2.36 arc4random F
>> +GLIBC_2.36 arc4random_buf F
>> +GLIBC_2.36 arc4random_uniform F
> arc4random is just merged two days ago.
Yeah; I meant to alert the relevant people (patch author or committer) 
to not forget adjusting this.
>
>> The others may need some love. Of course they're possibly because of my
>> particular environment (Gentoo is a little bit different than "ordinary"
>> distros like Debian/Fedora, and I already have to symlink the
>> libgcc_s.so and libstdc++.so to pass the nptl tests at all)
> I can tell tst-nss-files-hosts-long is most likely PR24816.  For the
> others I'm not sure.
>
> By "symlink libgcc_s.so and libstdc++.so" I guess you are running tests
> in a temporary system ("Gentoo stage 1" IIRC).  In LFS we also get some
> additional test failures in the temporary system [1] so it's better to
> rerun the tests in a full system.
>
> (In earlier LFS releases we also needed to symlink libgcc_s.so and
> libstdc++.so for glibc tests, but it's no longer needed now thanks to a
> major refactoring [2].)

Of course my Gentoo devbox is fully bootstrapped and working :-)

I'm referring to the Gentoo-specific behavior of slotted gcc 
installation, where gcc-provided libraries (including the two mentioned 
here) are NOT present in the usual place, but rather in 
/usr/lib/gcc/<PACKAGE VERSION>. Additional magic is present for the 
"active" gcc to be switchable via the gcc-config tool. Binutils is 
similar; there is a binutils-config too.

But anyway, the nptl tests mostly passed, with the only 2 failures 
related to the botched gdb on LoongArch (I compiled from the upstream 
HEAD and still it doesn't work). The others are worth looking into too 
but don't seem to cause any problem so far.

>
> [1]: https://www.linuxfromscratch.org/lfs/view/systemd/chapter08/glibc.html
> [2]: https://lists.linuxfromscratch.org/sympa/arc/lfs-dev/2020-05/msg00000.html
>
> My test result on LFS (fully built) with a patched ld seems clean.  The
> patch is originally published at
> https://sourceware.org/pipermail/binutils/2022-July/121852.html, but
> it's in a large series introducing new relocation types.  I managed to
> rebased and adjusted the patch so it can be applied individually for
> binutils-gdb master branch.  The patch is attached in this mail.
>
> +Zhensong: is there any possibility to merge this patch for binutils
> master and 2.39 release branch (including 2.38 release branch would be
> even better) so at least glibc ifunc tests will be happy?
>
> XPASS: conform/UNIX98/ndbm.h/linknamespace
> XPASS: conform/XOPEN2K/ndbm.h/linknamespace
> XPASS: conform/XOPEN2K8/ndbm.h/linknamespace
> XPASS: conform/XPG42/ndbm.h/linknamespace
> UNSUPPORTED: crypt/cert
> FAIL: elf/check-abi-libc
> UNSUPPORTED: elf/tst-env-setuid
> UNSUPPORTED: elf/tst-env-setuid-tunables
> XPASS: elf/tst-protected1a
> XPASS: elf/tst-protected1b
> UNSUPPORTED: elf/tst-valgrind-smoke
> UNSUPPORTED: misc/tst-adjtimex
> UNSUPPORTED: misc/tst-clock_adjtime
> UNSUPPORTED: misc/tst-ntp_adjtime
> UNSUPPORTED: misc/tst-pkey
> UNSUPPORTED: misc/tst-rseq
> UNSUPPORTED: misc/tst-rseq-disable
> UNSUPPORTED: nptl/test-cond-printers
> UNSUPPORTED: nptl/test-condattr-printers
> UNSUPPORTED: nptl/test-mutex-printers
> UNSUPPORTED: nptl/test-mutexattr-printers
> UNSUPPORTED: nptl/test-rwlock-printers
> UNSUPPORTED: nptl/test-rwlockattr-printers
> UNSUPPORTED: nptl/tst-pthread-gdb-attach
> UNSUPPORTED: nptl/tst-pthread-gdb-attach-static
> UNSUPPORTED: nptl/tst-rseq-nptl
> UNSUPPORTED: stdlib/tst-secure-getenv
> UNSUPPORTED: time/tst-clock_settime
> UNSUPPORTED: time/tst-settimeofday
> Summary of test results:
>        1 FAIL
>     4576 PASS
>       22 UNSUPPORTED
>       12 XFAIL
>        6 XPASS
Overall this is looking good. Thanks for your testing.