Re: Branch for netbsd 9 upcoming, please help and test -current

[christos%astron.com@localhost (Christos Zoulas)]

In article <20190424100614.GA488%homeworld.netbsd.org@localhost>,
 <maya%netbsd.org@localhost> wrote:
>On Wed, Apr 24, 2019 at 12:48:07PM +0300, Andreas Gustafsson wrote:
>> For starters, here's a list of PRs for regressions that have been
>> introduced since early February and are still unresolved:
>> 
>>   port-amd64/54001: call_once2_32, call_once2_static test cases
>failing on amd64 since gcc7 import

This is really @#@#!$ up. Something is wrong with TLS/locking in the
static case.

>>   lib/54002: Recent regressions in tests involving pic_profile

This is the one  the commit mentioned below fixes.

>>   bin/54060: usr.bin/gdb/t_regress/threads test case now fails on amd64

This I will look at.

>
>Try again after
>http://mail-index.netbsd.org/source-changes/2019/04/23/msg105364.html
>
>
>>   lib/54074: Test cases ending in 32 fail on sparc64 since jemalloc update
>>   port-sparc/54093: Recent test regressions on sparc
>
>I think we should revert to old jemalloc to get the branch done.
>We already have the tunable, let's use it to our advantage.

No need to do this yet. 54001 needs to be fixed first..

>>   port-sparc/53277: Many ubsan tests fail on sparc
>
>Re-test after
>https://v4.freshbsd.org/commit/netbsd/src/eTZNaSjDaT7KUvjB


christos