Re: Breakage on builder Solaris11-amd64-m64, revision 080363310650c93ad8e93018bcb6760ba5d32d1c

[Pedro Alves <palves@redhat.com>]

-- 
Thanks,
Pedro Alves
On 11/23/2018 02:16 PM, Rainer Orth wrote:
> Hi Sergio,
> 
>> On Friday, November 23 2018, I wrote:
>>
>>> On Friday, November 23 2018, Pedro Alves wrote:
>>>
>>>> On 11/22/2018 04:49 PM, gdb-buildbot@sergiodj.net wrote:
>>>>> Undefined			first referenced
>>>>>  symbol  			    in file
>>>>> is_exited(ptid_t)                   procfs.o
>>>>> ld: fatal: symbol referencing errors
>>>>> collect2: error: ld returned 1 exit status
>>>>> gmake[2]: *** [Makefile:1912: gdb] Error 1
>>>>> gmake[2]: Leaving directory
>>>>> '/opt/gdb-buildbot/home/solaris11-amd64/solaris11-amd64-m64/build/gdb'
>>>>> gmake[1]: *** [Makefile:8792: all-gdb] Error 2
>>>>> gmake[1]: Leaving directory
>>>>> '/opt/gdb-buildbot/home/solaris11-amd64/solaris11-amd64-m64/build'
>>>>> gmake: *** [Makefile:849: all] Error 2
>>>>> ============================
>>>>
>>>> Whoops, sorry about that.  I assume this works, but I haven't tried it.
>>>
>>> I took the liberty of running try builds with your patch on the Solaris
>>> builders:
>>>
>>>   https://gdb-build.sergiodj.net/builders/Solaris11-amd64-m64/builds/410
>>>   https://gdb-build.sergiodj.net/builders/Solaris11-sparcv9-m64/builds/408
>>>
>>> Unfortunately the build still fails with:
>>>
>>>   ../../binutils-gdb/gdb/sol-thread.c: In member function ‘virtual ptid_t sol_thread_target::wait(ptid_t, target_waitstatus*, int)’:
>>>   ../../binutils-gdb/gdb/sol-thread.c:465:2: error: expected ‘)’ before ‘{’ token
>>>     {
>>>     ^
>>>   ../../binutils-gdb/gdb/sol-thread.c:470:5: error: expected primary-expression before ‘}’ token
>>>        }
>>>        ^
>>>
>>> BTW, it's possible to run try builds on specific builders by using the
>>> "-b" option with "buildbot try".  E.g.:
>>>
>>>   buildbot try -b Solaris11-amd64-m64 -b Solaris11-sparcv9-m64
>>

Thanks.

>> And the actual commit failed with another reason:
>>
>> https://gdb-build.sergiodj.net/builders/Solaris11-amd64-m64/builds/411/steps/compile%20gdb/logs/stdio
>>
>> ../../binutils-gdb/gdb/procfs.c: In function ‘int procfs_notice_thread(procinfo*, procinfo*, void*)’:
>> ../../binutils-gdb/gdb/procfs.c:3136:41: error: ‘is_exited’ was not declared in this scope
>>    if (!in_thread_list (gdb_threadid) || is_exited (gdb_threadid))
>>                                          ^~~~~~~~~
>> ../../binutils-gdb/gdb/procfs.c:3136:41: note: suggested alternative: ‘inf_exited’
>>    if (!in_thread_list (gdb_threadid) || is_exited (gdb_threadid))
>>                                          ^~~~~~~~~
>>                                          inf_exited
> 
> this is weird: I still see a declaration in gdbthread.h (and also
> is_running and is_stopped), which should go as well.

That's because I meanwhile pushed this:
  https://sourceware.org/ml/gdb-patches/2018-11/msg00380.html

So before that we got a link error, and now we get a compile
error, which is expected.

> 
> With just the typo above fixed, a local build on amd64-pc-solaris2.11
> has just finished; make check running...

Great, thanks.
Pedro Alves