Haxe 3.0.1 'segmentation fault (core dumped)' Arch Linux 64 bit

[William Dyce]

Hi all,

Running the 'haxe' command (with or without arguments) results in a segmentation fault on my (fresh) Arch Linux 64 bit installation.

I repeat: the compiler itself won't run, no matter what it's told to compile.

I've tried installing Haxe using:

• Haxe 3.0.1 installer on openfl.org
• Haxe package from on AUR
• Compiling Haxe from the latest source

In every case running 'haxe' results in 'segmentation fault (core dumped)'. Haxe 2.1.0 runs, but since I'm trying to use OpenFL it's not really a solution.

Please help, I really don't know what else I can try :(

William

[Benjamin Dubois]

Sorry am I reading well ?

Haxe 3.0.1 installer AND Haxe 2.1.0 runs

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
---
You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/groups/opt_out.

[John Plsek]

It's quite clear he is saying haxe 3.0.1 does not work regardless of 3 different install methods (installer, package, compile from source), but haxe 2.1.0 does work, however, haxe 2.1.0 is not a viable solution as he wishes to use openfl

[Nicolas Cannasse]

Le 27/12/2013 09:12, William Dyce a �crit :

> Hi all,
>
> Running the 'haxe' command (with or without arguments) results in a
> segmentation fault on my (fresh) Arch Linux 64 bit installation.
>

> I repeat: /the compiler itself/ won't run, no matter what it's told to

> compile.
>
> I've tried installing Haxe using:
>

> * Haxe 3.0.1 installer on openfl.org
> * Haxe package from on AUR
> * Compiling Haxe from the latest source

>
> In every case running 'haxe' results in 'segmentation fault (core
> dumped)'. Haxe 2.1.0 runs, but since I'm trying to use OpenFL it's not
> really a solution.
>
> Please help, I really don't know what else I can try :(

Could you run haxe into "gdb" and use "bt" to get the stack trace ? That
would help to know where the error comes from.

Best,
Nicolas

[William Dyce]

Thanks for the replies :)

I tried gdb a while back, but it just says that there are no debug symbols. That was the official build though: perhaps if I compiled it from source in debug mode? Not sure how to go about that though unless there's a makefile option...

I'll give bt a try.

W.

On 27 December 2013 20:09, Nicolas Cannasse <ncan...@gmail.com> wrote:

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en

--- You received this message because you are subscribed to a topic in the Google Groups "Haxe" group.

For more options, visit https://groups.google.com/groups/opt_out.

--
http://wilbefast.com
@wilbefast

[William Dyce]

Oups sorry, me being a noob and not knowing how to use gdb except through a GUI :S

[william@SATAN ~]$ gdb haxe
GNU gdb (GDB) 7.6.2
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
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 "x86_64-unknown-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/lib/haxe/haxe...(no debugging symbols found)...done.
(gdb) run
Starting program: /usr/bin/haxe 
Haxe Compiler 3.0.1 - (C)2005-2013 Haxe Foundation
 Usage : haxe -main <class> [-swf|-js|-neko|-php|-cpp|-as3] <output> [options]
 Options :
  -cp <path> : add a directory to find source files
  -js <file> : compile code to JavaScript file
  -swf <file> : compile code to Flash SWF file
  -as3 <directory> : generate AS3 code into target directory
  -neko <file> : compile code to Neko Binary
  -php <directory> : generate PHP code into target directory
  -cpp <directory> : generate C++ code into target directory
  -cs <directory> : generate C# code into target directory
  -java <directory> : generate Java code into target directory
  -xml <file> : generate XML types description
  -main <class> : select startup class
  -lib <library[:version]> : use a haxelib library
  -D <var> : define a conditional compilation flag
  -v : turn on verbose mode
  -debug : add debug information to the compiled code
  -help  Display this list of options
  --help  Display this list of options
During startup program exited normally.
(gdb) quit

Wait... what?

[william@SATAN ~]$ haxe
Haxe Compiler 3.0.1 - (C)2005-2013 Haxe Foundation
 Usage : haxe -main <class> [-swf|-js|-neko|-php|-cpp|-as3] <output> [options]
 Options :
  -cp <path> : add a directory to find source files
  -js <file> : compile code to JavaScript file
  -swf <file> : compile code to Flash SWF file
  -as3 <directory> : generate AS3 code into target directory
  -neko <file> : compile code to Neko Binary
  -php <directory> : generate PHP code into target directory
  -cpp <directory> : generate C++ code into target directory
  -cs <directory> : generate C# code into target directory
  -java <directory> : generate Java code into target directory
  -xml <file> : generate XML types description
  -main <class> : select startup class
  -lib <library[:version]> : use a haxelib library
  -D <var> : define a conditional compilation flag
  -v : turn on verbose mode
  -debug : add debug information to the compiled code
  -help  Display this list of options
  --help  Display this list of options

So... euh... apparently it's not segfaulting any more. I don't know why it started or whether it will start again, but it seems to have fixed itself after a reboot. Thing is I rebooted multiple times and it didn't fix itself... though come to think of it I hadn't rebooted since trying to install from AUR and source, so maybe a side-effect of those installations fixed it with the reboot????

So if anyone if wondering how to solve this problem, the following steps may (or may not) help:

1. install haxe 2.1.0
2. install haxe from AUR
3. install haxe from source
4. install haxe from openfl.org package
5. reboot

--
http://wilbefast.com
@wilbefast

[Achmad Aulia Noorhakim]

Do you have nvidia card and using the nvidia binary driver ? That might causing the segfault. Try using nouveau. I use nouveau until nvidia get their shit together and fix their driver. So far no segfault or crash.

You received this message because you are subscribed to the Google Groups "Haxe" group.

[Paweł Stępień]

Achmad Aulia Noorhakim <aau...@gmail.com> wrote:
> Do you have nvidia card and using the nvidia binary driver ? That might
> causing the segfault. Try using nouveau. I use nouveau until nvidia get
> their shit together and fix their driver. So far no segfault or crash.

How would the video driver, of all things, make haxe segfault?

[William Dyce]

I am using Nvidia's drivers yes, might make a difference for openfl but the haxe compiler itself shouldn't be affected.

[Achmad Aulia Noorhakim]

Well if you would take time to investigate my answer further, the latest nvidia binary blob haven't been playing nice with the lateest kernel. Causing zombie/defunct provess to appear, x not starting, etc.

[Achmad Aulia Noorhakim]

https://bbs.archlinux.org/viewtopic.php?id=173024

http://www.gamersonlinux.com/forum/threads/nvidia-331-20-driver.410/

https://bbs.archlinux.org/viewtopic.php?id=173214

https://devtalk.nvidia.com/default/topic/638521/gnome-terminal-problems-ctrl-c-and-exit/

All I'm saying I have this problem before and when I asked around here (around Nov), nobody can really give an answer so I have to poke around my Arch for a few days, looking at what causing it, and turns out the problem with nvidia driver is pretty well documented since couple of months ago. The problem is different for some people. Some people with KDE can't get X started, some others noticing zombie processes, for me what give it away is that I can't "exit" normally from terminal, and it happen randomly just like haxe segfaulting. I have been using nouveau since then, and no problem. The nvidia people haven't address this issue AFAIK, maybe it haven't reach upstream, idk, but last time I check, (couple of weeks ago) it still happening.

[William Dyce]

Well I am having the same 'segmentation fault (core dumped)' error in a couple of seemingly random places - for instance I was trying to run mono game this morning... it might be a kernel issue I suppose.

W.

You received this message because you are subscribed to a topic in the Google Groups "Haxe" group.

For more options, visit https://groups.google.com/groups/opt_out.

--
http://wilbefast.com
@wilbefast

[William Dyce]

Guess what: the error is magically back again o.O

This time I had a chance to run it through gdb: 

[william@SATAN omnilith]$ gdb haxe

GNU gdb (GDB) 7.6.2
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
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 "x86_64-unknown-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/lib/haxe/haxe...(no debugging symbols found)...done.
(gdb) run
Starting program: /usr/bin/haxe 

warning: no loadable sections found in added symbol-file system-supplied DSO at 0x7ffff7ffa000
warning: Could not load shared library symbols for linux-vdso.so.1.
Do you need "set solib-search-path" or "set sysroot"?
Warning: couldn't activate thread debugging using libthread_db: Cannot find new threads: generic error
warning: File "/usr/lib/libthread_db-1.0.so" auto-loading has been declined by your `auto-load safe-path' set to "$debugdir:$datadir/auto-load".
To enable execution of this file add
add-auto-load-safe-path /usr/lib/libthread_db-1.0.so
line to your configuration file "/home/william/.gdbinit".
To completely disable this security protection add
set auto-load safe-path /
line to your configuration file "/home/william/.gdbinit".
For more information about this security protection see the
"Auto-loading safe path" section in the GDB manual.  E.g., run from the shell:
info "(gdb)Auto-loading safe path"
warning: Unable to find libthread_db matching inferior's thread library, thread debugging will not be available.
Program received signal SIGSEGV, Segmentation fault.
0x00007ffff6ba7c06 in GC_find_limit_with_bound () from /usr/lib/libgc.so.1
(gdb) 

Something about garbage collection? I'll try rebooting. 

--
http://wilbefast.com
@wilbefast

[William Dyce]

Here's the backtrace by the way :

 

(gdb) bt
#0  0x00007ffff6ba7c06 in GC_find_limit_with_bound () from /usr/lib/libgc.so.1
#1  0x00007ffff6ba7d03 in ?? () from /usr/lib/libgc.so.1
#2  0x00007ffff6ba677f in GC_init () from /usr/lib/libgc.so.1
#3  0x00007ffff6dfd43f in neko_gc_init () from /usr/lib/libneko.so
#4  0x00007ffff6dfdaee in neko_global_init () from /usr/lib/libneko.so
#5  0x00000000006d31fc in sys_dlcall0 ()
#6  0x00000000006e3cf8 in caml_c_call ()
#7  0x0000000000ae9ee0 in ?? ()
#8  0x0000000000000001 in ?? ()
#9  0xffffffffe44ca603 in ?? ()
#10 0xfffffffff539501f in ?? ()
#11 0xffffffffe46814c1 in ?? ()
#12 0xfffffffff554bedd in ?? ()
#13 0xffffffffeebb9621 in ?? ()
#14 0xfffffffface3c485 in ?? ()
#15 0xffffffffe4513c63 in ?? ()
#16 0xfffffffff53de67f in ?? ()
#17 0xffffffffe45eef45 in ?? ()
#18 0xfffffffff54b9961 in ?? ()
#19 0x0000000072bf27e3 in ?? ()
#20 0x00007ffff72ebac0 in ?? ()
#21 0xffffffffab0b5379 in ?? ()
#22 0x000000000090d283 in ?? ()
---Type <return> to continue, or q <return> to quit---
#23 0x000000000090d2a7 in ?? ()
#24 0x00000000006a4646 in camlPervasives__$5e_1112 ()
#25 0x00000000009728e0 in camlGencs__137 ()
#26 0x00000000006de1ae in hash_aux ()
#27 0x0000000000000078 in ?? ()
#28 0x00007ffff72ef370 in ?? ()
#29 0x00000000006ad894 in camlHashtbl__add_1074 ()
#30 0x00007ffff72edfd8 in ?? ()
#31 0x00000000006a9d45 in camlList__iter_1074 ()
#32 0x00007ffff72ef350 in ?? ()
#33 0x0000000000000001 in ?? ()
#34 0x0000000000000038 in ?? ()
#35 0x00000000004e958d in camlGencs__entry ()
#36 0x0000000000000000 in ?? ()

--
http://wilbefast.com
@wilbefast

[David Elahee]

We had issues with caml gc recently...maybe this one is linked.

2013/12/30 William Dyce <wilb...@gmail.com>

--
David Elahee

[William Dyce]

Reinstalling Haxe and rebooting seems to solve the problem- it seems to have cropped up again when (once again) I accidentally pulled the plug. So perhaps there's some really low-level memory corruption occurring? That wouldn't explain why reinstalling helped though, but then neither did only rebooting do anything. Very strange.

[Nicolas Cannasse]

Le 30/12/2013 10:35, William Dyce a écrit :
> Here's the backtrace by the way :
>

> *(gdb) bt
> **#0 0x00007ffff6ba7c06 in GC_find_limit_with_bound () from
> /usr/lib/libgc.so.1

Looks like an issue with Boehm GC which Neko is using.
Try to upgrade your libgc on your system, if that helps.

Best,
Nicolas