Bug#827220: [DRE-maint] Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-17 Thread Christian Hofstaedtler
* Eric Wong  [160613 23:27]:
> Looks like only the bootstrap test ("make test") is run.
> 
> That doesn't test a big chunk of functionality or standard library
> at all.
> 
> The "test-all" target should also be run and is likely to find
> more problems.
> ("make check" runs both test-all and test targets)

We've recently enabled `make check` in the ruby2.3 builds (prompted
by this thread).

I'd love any help with diagnosing/finding fixes for the resulting
build failures:

   https://buildd.debian.org/status/package.php?p=ruby2.3=sid

Please follow up to debian-ruby@lists.d.o for these things (leaving
this bug alone).

Thanks,
-- 
 ,''`.  Christian Hofstaedtler 
: :' :  Debian Developer
`. `'   7D1A CFFA D9E0 806C 9C4C  D392 5C13 D6DB 9305 2E03
  `-



signature.asc
Description: PGP signature


Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-14 Thread Eric Wong
Helge Deller  wrote:
> I'm somehow lost here now...
> Ideas?

Thanks, I'm lost here, too.  Anyways I've forwarded this
to ruby-core upstream:

https://bugs.ruby-lang.org/issues/12489

Unfortunately, I think posting to that bug tracker or replying to
Message-ID: 
on ruby-c...@ruby-lang.org requires a subscription or login;
so I'm happy to forward any messages along as needed.

I'm wondering if anybody previous experience porting similar
software to hppa could help.  Perhaps anything using a
mark-sweep GC like Boehm could help (if this is a GC bug in
Ruby).


Thanks.



Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-14 Thread Helge Deller
Hi Eric,

On 13.06.2016 23:23, Eric Wong wrote:
> Helge Deller  wrote:
>> On 13.06.2016 22:59, Eric Wong wrote:
>>> Seems to be a ruby2.3 problem.
>>> Does the ruby2.3 test suite work at all on hppa?
>>
>> No idea (I'm no ruby expert).
>> The last build log is here:
>> https://buildd.debian.org/status/fetch.php?pkg=ruby2.3=hppa=2.3.1-2=1464615629
>>
>> The log looks similar to x86_64, but I'm not sure if the test suite is run 
>> then.
> 
> Looks like only the bootstrap test ("make test") is run.
> 
> That doesn't test a big chunk of functionality or standard library
> at all.
> 
> The "test-all" target should also be run and is likely to find
> more problems.
> 
> Can you give that a shot?

Sure, I ran a "make test-all" and put the full log here: 
http://dellerweb.de/debug-logs/ruby2.3-make-test-all.log

There seems to be some faults:

# Running tests:
FFFFFF.FFF.../build/ruby2.3/ruby2.3-2.3.1/lib/forwardable.rb:189:
 [BUG] Illegal instruction at 0xfa343a5b
ruby 2.3.1p112 (2016-04-26) [hppa-linux-gnu]

-- Control frame information ---
c:0014 p: s:0041 e:40 IFUNC
c:0013 p: s:0039 e:38 CFUNC  :each
c:0012 p:0021 s:0036 e:35 METHOD 
/build/ruby2.3/ruby2.3-2.3.1/lib/forwardable.rb:189
c:0011 p:0012 s:0031 e:30 BLOCK  
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:400 [FINISH]
c:0010 p: s:0027 e:26 CFUNC  :each
c:0009 p:0011 s:0024 e:23 METHOD 
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:399 [FINISH]
c:0008 p: s:0020 e:19 CFUNC  :each
c:0007 p: s:0018 E:f9d8 CFUNC  :find
c:0006 p:0011 s:0015 E:e9e8 METHOD 
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:325
c:0005 p:0010 s:0012 e:11 METHOD 
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:633
c:0004 p:0008 s:0009 e:08 BLOCK  
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:622
c:0003 p:0015 s:0007 e:06 METHOD 
/build/ruby2.3/ruby2.3-2.3.1/lib/monitor.rb:214
c:0002 p:0025 s:0004 e:03 BLOCK  
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:621 [FINISH]
c:0001 p: s:0002 e:01 (none) [FINISH]

-- Ruby level backtrace information 
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:621:in `block in 
start_keeper'
/build/ruby2.3/ruby2.3-2.3.1/lib/monitor.rb:214:in `mon_synchronize'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:622:in `block (2 levels) 
in start_keeper'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:633:in `need_keeper?'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:325:in `has_expires?'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:325:in `find'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:325:in `each'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:399:in `each_entry'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:399:in `each'
/build/ruby2.3/ruby2.3-2.3.1/lib/rinda/tuplespace.rb:400:in `block in 
each_entry'
/build/ruby2.3/ruby2.3-2.3.1/lib/forwardable.rb:189:in `each'
/build/ruby2.3/ruby2.3-2.3.1/lib/forwardable.rb:189:in `each'

-- Other runtime information ---

* Loaded script: ./test/runner.rb: Fiddle::TestFunc#test_qsort1


I'm somehow lost here now...
Ideas?

Helge



Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-13 Thread Eric Wong
Helge Deller  wrote:
> On 13.06.2016 22:59, Eric Wong wrote:
> > Seems to be a ruby2.3 problem.
> > Does the ruby2.3 test suite work at all on hppa?
> 
> No idea (I'm no ruby expert).
> The last build log is here:
> https://buildd.debian.org/status/fetch.php?pkg=ruby2.3=hppa=2.3.1-2=1464615629
> 
> The log looks similar to x86_64, but I'm not sure if the test suite is run 
> then.

Looks like only the bootstrap test ("make test") is run.

That doesn't test a big chunk of functionality or standard library
at all.

The "test-all" target should also be run and is likely to find
more problems.

Can you give that a shot?

("make check" runs both test-all and test targets)

ruby-core actually has a good chunk of developers working on
Debian, but mainly x86/x86-64.

> > Any other ruby libraries with similar problems?
> 
> I think most ruby packages build sucessfully.
> In the package overview I find only 9 ruby packages which don't build: 
> https://buildd.debian.org/status/architecture.php?a=hppa=sid

Thanks, I'll check those at a later time.



Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-13 Thread Eric Wong
Mattia Rizzolo  wrote:
> On Mon, Jun 13, 2016 at 10:40:51PM +0200, Helge Deller wrote:
> > Program received signal SIGSEGV, Segmentation fault.
> > [Switching to Thread 0xfa5dd480 (LWP 29258)]
> > mix_id_table_lookup (tbl=0x0, id=id@entry=3265, valp=valp@entry=0xfa55f248) 
> > at id_table.c:1516

tbl=0x0 looks like a bug in Ruby; possibly GC-related.

> > 1516id_table.c: No such file or directory.
> > (gdb) bt
> > #0  mix_id_table_lookup (tbl=0x0, id=id@entry=3265, 
> > valp=valp@entry=0xfa55f248) at id_table.c:1516
> > #1  0xf9c91b04 in lookup_method_table (id=3265, klass=) at 
> > vm_method.c:190
> > #2  search_method (defined_class_ptr=, id=, 
> > klass=) at vm_method.c:701
> > #3  method_entry_get_without_cache (defined_class_ptr=, 
> > id=, klass=) at vm_method.c:726
> > #4  method_entry_get (klass=1134408, id=3265, id@entry=1134408, 
> > defined_class_ptr=defined_class_ptr@entry=0xfa55f1c8) at vm_method.c:786
> > #5  0xf9c94e90 in rb_callable_method_entry (klass=, 
> > id=1134408, id@entry=4199936456) at vm_method.c:831
> > #6  0xf9ca6048 in rb_search_method_entry (mid=4199936456, recv=1134408) at 
> > vm_eval.c:573
> > #7  rb_call0 (recv=1134408, mid=4199936456, argc=3265, argv=0xfa55f248, 
> > scope=, self=) at vm_eval.c:343
> > #8  0xf9ca6984 in rb_call (scope=CALL_FCALL, argv=, 
> > argc=, mid=, recv=) at 
> > vm_eval.c:637
> > #9  rb_funcall (recv=, mid=, n= > out>) at vm_eval.c:835
> > #10 0x in ?? ()
> > Backtrace stopped: previous frame identical to this frame (corrupt stack?)
> > 
> > I assume it's a problem in ruby2.3 itself, but since I'm no expert in 
> > ruby2.3, is there anyone who might be able to help?
> > I can provide a login to a hppa box if necessary...

Seems to be a ruby2.3 problem.
Does the ruby2.3 test suite work at all on hppa?
Any other ruby libraries with similar problems?
Thanks.

(I'm kgio upstream, and occasional ruby-core committer).



Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-13 Thread Helge Deller
On 13.06.2016 22:59, Eric Wong wrote:
> Mattia Rizzolo  wrote:
>> On Mon, Jun 13, 2016 at 10:40:51PM +0200, Helge Deller wrote:
>>> Program received signal SIGSEGV, Segmentation fault.
>>> [Switching to Thread 0xfa5dd480 (LWP 29258)]
>>> mix_id_table_lookup (tbl=0x0, id=id@entry=3265, valp=valp@entry=0xfa55f248) 
>>> at id_table.c:1516
> 
> tbl=0x0 looks like a bug in Ruby; possibly GC-related.
> 
>>> 1516id_table.c: No such file or directory.
>>> (gdb) bt
>>> #0  mix_id_table_lookup (tbl=0x0, id=id@entry=3265, 
>>> valp=valp@entry=0xfa55f248) at id_table.c:1516
>>> #1  0xf9c91b04 in lookup_method_table (id=3265, klass=) at 
>>> vm_method.c:190
>>> #2  search_method (defined_class_ptr=, id=, 
>>> klass=) at vm_method.c:701
>>> #3  method_entry_get_without_cache (defined_class_ptr=, 
>>> id=, klass=) at vm_method.c:726
>>> #4  method_entry_get (klass=1134408, id=3265, id@entry=1134408, 
>>> defined_class_ptr=defined_class_ptr@entry=0xfa55f1c8) at vm_method.c:786
>>> #5  0xf9c94e90 in rb_callable_method_entry (klass=, 
>>> id=1134408, id@entry=4199936456) at vm_method.c:831
>>> #6  0xf9ca6048 in rb_search_method_entry (mid=4199936456, recv=1134408) at 
>>> vm_eval.c:573
>>> #7  rb_call0 (recv=1134408, mid=4199936456, argc=3265, argv=0xfa55f248, 
>>> scope=, self=) at vm_eval.c:343
>>> #8  0xf9ca6984 in rb_call (scope=CALL_FCALL, argv=, 
>>> argc=, mid=, recv=) at 
>>> vm_eval.c:637
>>> #9  rb_funcall (recv=, mid=, n=>> out>) at vm_eval.c:835
>>> #10 0x in ?? ()
>>> Backtrace stopped: previous frame identical to this frame (corrupt stack?)
>>>
>>> I assume it's a problem in ruby2.3 itself, but since I'm no expert in 
>>> ruby2.3, is there anyone who might be able to help?
>>> I can provide a login to a hppa box if necessary...
> 
> Seems to be a ruby2.3 problem.
> Does the ruby2.3 test suite work at all on hppa?

No idea (I'm no ruby expert).
The last build log is here:
https://buildd.debian.org/status/fetch.php?pkg=ruby2.3=hppa=2.3.1-2=1464615629

The log looks similar to x86_64, but I'm not sure if the test suite is run then.

> Any other ruby libraries with similar problems?

I think most ruby packages build sucessfully.
In the package overview I find only 9 ruby packages which don't build: 
https://buildd.debian.org/status/architecture.php?a=hppa=sid

Helge



Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-13 Thread Mattia Rizzolo
control: reassign -1 src:ruby-kgio 2.10.0-1

On Mon, Jun 13, 2016 at 10:40:51PM +0200, Helge Deller wrote:
> Package: riby-kgio

mistypoed package name :)

> Version: 2.10.0-1+b2
> 
> This package fails only on the hppa architecture, as can be seen here:
> https://buildd.debian.org/status/package.php?p=ruby-kgio=sid
> 
> According to the last build log:
> https://buildd.debian.org/status/fetch.php?pkg=ruby-kgio=hppa=2.10.0-1%2Bb2=1461596321
> 
> it fails while running the testsuite:
> 
> Started
> TesTcpClientReadServerWrite: 
>   test_monster_trywrite:  .: (0.092000)
>   test_monster_trywritev: .: (0.292001)
>   test_monster_write: .: (0.108000)
>   test_monster_write_wait_writable:   .: (0.156000)
>   test_monster_writev:.: (0.152000)
>   test_monster_writev_wait_writable:  
> /<>/test/lib_read_write.rb:390: [BUG] Segmentation fault at 
> 0x00032d
> ruby 2.3.0p0 (2015-12-25) [hppa-linux-gnu]
> 
> I could run it in gdb with debug-symbols loaded:
> 
> (unstabletest-hppa-sbuild)root@phantom:/build/ruby-kgio/ruby-kgio-2.10.0# gdb 
> /usr/bin/ruby2.3
> GNU gdb (Debian 7.10-1.1) 7.10
> Copyright (C) 2015 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later 
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "hppa-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> .
> Find the GDB manual and other documentation resources online at:
> .
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols from /usr/bin/ruby2.3...(no debugging symbols found)...done.
>  "test/test_unix_server.rb" "test/test_unix_server_read_client_write.rb" -v
> Starting program: /usr/bin/ruby2.3 -I"test"  
> "/usr/lib/ruby/vendor_ruby/rake/rake_test_loader.rb" 
> "test/test_accept_class.rb" "test/test_accept_flags.rb" 
> "test/test_autopush.rb" "test/test_connect_fd_leak.rb" 
> "test/test_cross_thread_close.rb" "test/test_default_wait.rb" 
> "test/test_kgio_addr.rb" "test/test_no_dns_on_tcp_connect.rb" 
> "test/test_peek.rb" "test/test_pipe_popen.rb" "test/test_pipe_read_write.rb" 
> "test/test_poll.rb" "test/test_singleton_read_write.rb" "test/test_socket.rb" 
> "test/test_socketpair_read_write.rb" "test/test_syssend.rb" 
> "test/test_tcp6_client_read_server_write.rb" 
> "test/test_tcp_client_read_server_write.rb" "test/test_tcp_connect.rb" 
> "test/test_tcp_server.rb" "test/test_tcp_server_read_client_write.rb" 
> "test/test_tfo.rb" "test/test_tryopen.rb" 
> "test/test_unix_client_read_server_write.rb" "test/test_unix_connect.rb" 
> "test/test_unix_server.rb" "test/test_unix_server_read_client_write.rb" -v
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/hppa-linux-gnu/libthread_db.so.1".
> [New Thread 0xfa674480 (LWP 28760)]
> /usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in `require': 
> cannot load such file -- kgio (LoadError)
> from /usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in 
> `require'
> from /build/ruby-kgio/ruby-kgio-2.10.0/test/test_accept_class.rb:4:in 
> `'
> from /usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in 
> `require'
> from /usr/lib/ruby/2.3.0/rubygems/core_ext/kernel_require.rb:55:in 
> `require'
> from /usr/lib/ruby/vendor_ruby/rake/rake_test_loader.rb:15:in `block 
> in '
> from /usr/lib/ruby/vendor_ruby/rake/rake_test_loader.rb:4:in `select'
> from /usr/lib/ruby/vendor_ruby/rake/rake_test_loader.rb:4:in `'
> [Thread 0xfa674480 (LWP 28760) exited]
> [Inferior 1 (process 28757) exited with code 01]
> (gdb) q
>   (unstabletest-hppa-sbuild)root@phantom:/build/ruby-kgio/ruby-kgio-2.10.0# gdb 
> /usr/bin/ruby2.3
> GNU gdb (Debian 7.10-1.1) 7.10
> Copyright (C) 2015 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later 
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
> and "show warranty" for details.
> This GDB was configured as "hppa-linux-gnu".
> Type "show configuration" for configuration details.
> For bug reporting instructions, please see:
> .
> Find the GDB manual and other documentation resources online at:
> .
> For help, type "help".
> Type "apropos word" to search for commands related to "word"...
> Reading symbols 

Bug#827220: ruby-kgio FTBFS on hppa architecture

2016-06-13 Thread Helge Deller
Package: riby-kgio
Version: 2.10.0-1+b2

This package fails only on the hppa architecture, as can be seen here:
https://buildd.debian.org/status/package.php?p=ruby-kgio=sid

According to the last build log:
https://buildd.debian.org/status/fetch.php?pkg=ruby-kgio=hppa=2.10.0-1%2Bb2=1461596321

it fails while running the testsuite:

Started
TesTcpClientReadServerWrite: 
  test_monster_trywrite:.: (0.092000)
  test_monster_trywritev:   .: (0.292001)
  test_monster_write:   .: (0.108000)
  test_monster_write_wait_writable: .: (0.156000)
  test_monster_writev:  .: (0.152000)
  test_monster_writev_wait_writable:
/<>/test/lib_read_write.rb:390: [BUG] Segmentation fault at 
0x00032d
ruby 2.3.0p0 (2015-12-25) [hppa-linux-gnu]

I could run it in gdb with debug-symbols loaded:

(unstabletest-hppa-sbuild)root@phantom:/build/ruby-kgio/ruby-kgio-2.10.0# gdb 
/usr/bin/ruby2.3
GNU gdb (Debian 7.10-1.1) 7.10
Copyright (C) 2015 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later 
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "hppa-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/ruby2.3...(no debugging symbols found)...done.

This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "hppa-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
.
Find the GDB manual and other documentation resources online at:
.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/ruby2.3...(no debugging symbols found)...done.