Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
* Thomas Weber [EMAIL PROTECTED] [2008-03-01 12:28]: Sigh, and all this with the knowledge that probably no real person uses Octave and friends on ARM. No wonder. As Riku Voipio wrote in a post to debian-devel-announce [1]: As most ARM CPUs do not have FPUs, armel is by default softfloat. The old arm port has peculiar mixed-endian double presentation and it emulates a FPU using kernel exceptions, which is *painfully* slow. Perhaps some users will try Octave on armel? [1] http://lists.debian.org/debian-devel-announce/2008/03/msg4.html -- Rafael -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: octave3.0 crashes on arm
On 03/01/08 10:56 +0100, Soeren Sonnenburg wrote: Package: octave3.0 Version: 3.0.0-1 Severity: grave panic: Segmentation fault -- stopping myself... attempting to save variables to `octave-core'... save to `octave-core' complete Segmentation fault This turns out to be a toolchain issue; using gcc-4.1 and friends, it works in qemu. I suppose this is related to #467503 (the bug symptoms under gdb are pretty much the same). So, that's the good news. The bad news is that octave3.0 build failed on europa with an ICE. Sigh, and all this with the knowledge that probably no real person uses Octave and friends on ARM. Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
Am Donnerstag, den 10.01.2008, 10:44 +0100 schrieb Soeren Sonnenburg: On Thu, 2008-01-10 at 10:23 +0100, Rafael Laboissiere wrote: * Soeren Sonnenburg [EMAIL PROTECTED] [2008-01-08 08:47]: OK, I've recompiled octave w/ -O2 -g and it dies on loading dispatch.oct. This could be enough info for upstream to fix it. Thomas Weber has forwarded your message to bug-octave and John Eaton replied to it [1] and gave the following suggestion: Have you tried building with -O0? If that works but -O2 fails, then I would start to suspect an ARM-specific compiler bug since compiling with GCC and -O2 works on many other systems. Could you please try it? Tried already, as the original debian package builds with -O0 and without lapack ... and there it crashed too (that was my first backtrace). I only enabled -O2 and lapack as was wondering why it is not enabled... Speed reasons; compile time on ARM is a pain. Okay, I give up. After spending days (mostly waiting for qemu to finish compiling), I still don't have a clue. Mail to debian-arm went unanswered[1], #debian-arm couldn't help either. I think it's an alloca issue, but I lack the knowledge to decide this. I suggest we drop ARM from the to be built architectures. Given the nature of the bug, the fact that we had this bug for several months and no bug report about it, I don't think we have any real users out there on this architecture. Meanwhile, real testing users are told by upstream to upgrade to 3.0 (which I support, that's why there's a new release). [1] http://lists.debian.org/debian-arm/2008/01/msg00069.html Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: octave3.0 crashes on arm
* Soeren Sonnenburg [EMAIL PROTECTED] [2008-01-08 08:47]: OK, I've recompiled octave w/ -O2 -g and it dies on loading dispatch.oct. This could be enough info for upstream to fix it. Thomas Weber has forwarded your message to bug-octave and John Eaton replied to it [1] and gave the following suggestion: Have you tried building with -O0? If that works but -O2 fails, then I would start to suspect an ARM-specific compiler bug since compiling with GCC and -O2 works on many other systems. Could you please try it? Thanks, [1] http://www.cae.wisc.edu/pipermail/bug-octave/2008-January/004614.html -- Rafael -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: octave3.0 crashes on arm
On Thu, 2008-01-10 at 10:23 +0100, Rafael Laboissiere wrote: * Soeren Sonnenburg [EMAIL PROTECTED] [2008-01-08 08:47]: OK, I've recompiled octave w/ -O2 -g and it dies on loading dispatch.oct. This could be enough info for upstream to fix it. Thomas Weber has forwarded your message to bug-octave and John Eaton replied to it [1] and gave the following suggestion: Have you tried building with -O0? If that works but -O2 fails, then I would start to suspect an ARM-specific compiler bug since compiling with GCC and -O2 works on many other systems. Could you please try it? Tried already, as the original debian package builds with -O0 and without lapack ... and there it crashed too (that was my first backtrace). I only enabled -O2 and lapack as was wondering why it is not enabled... Soeren -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
On Fri, 2008-01-04 at 10:19 +0100, Thomas Weber wrote:
Am Donnerstag, den 03.01.2008, 10:56 +0100 schrieb Soeren Sonnenburg:
Package: octave3.0
Version: 3.0.0-1
Severity: grave
just try
echo version | octave
or
octave
on an arm machine:
$ uname -a
Linux arm 2.6.18-4-versatile #1 Sun Jun 17 14:34:28 UTC 2007
armv5tejl GNU/Linux
Is this real or emulated hardware?
OK, I've recompiled octave w/ -O2 -g and it dies on loading
dispatch.oct. This could be enough info for upstream to fix it.
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x40022490 (LWP 11491)]
0x40db4890 in octave_shlib::make_shlib ()
from /usr/lib/octave-3.0.0/liboctave.so
(gdb) bt
#0 0x40db4890 in octave_shlib::make_shlib ()
from /usr/lib/octave-3.0.0/liboctave.so
#1 0x401f7584 in octave_dynamic_loader::do_load_oct (
this=value optimized out, [EMAIL PROTECTED], [EMAIL PROTECTED],
relative=false) at ../liboctave/oct-shlib.h:52
#2 0x401f7f54 in octave_dynamic_loader::load_oct ([EMAIL PROTECTED],
[EMAIL PROTECTED], relative=false) at dynamic-ld.cc:508
#3 0x40469150 in load_fcn_from_file ([EMAIL PROTECTED], exec_script=true)
at parse.y:3440
#4 0x40469878 in load_fcn_from_file (sym_rec=0xf71658, exec_script=true)
at parse.y:3480
#5 0x40519554 in lookup (sym_rec=0xf71658, exec_script=true)
at variables.cc:
#6 0x408021e4 in tree_identifier::do_lookup (this=0xf6fe38,
[EMAIL PROTECTED], exec_script=true) at pt-id.cc:110
#7 0x408038e0 in tree_identifier::rvalue (this=0xf6fe38, nargout=1)
at pt-id.cc:161
#8 0x408024c4 in tree_identifier::rvalue (this=0xbedd5450) at pt-id.cc:204
#9 0x4080b72c in tree_index_expression::rvalue (this=0xf71958, nargout=0)
at pt-idx.cc:279
#10 0x4083db18 in tree_statement::eval (this=0xf6f080, silent=false,
nargout=0, in_function_body=false) at pt-stmt.cc:133
#11 0x4083e50c in tree_statement_list::eval (this=0xf66428, silent=false,
---Type return to continue, or q return to quit---
nargout=0) at pt-stmt.cc:190
#12 0x404672e8 in parse_and_execute (f=0xf60620) at parse.y:2864
#13 0x40467c68 in parse_fcn_file ([EMAIL PROTECTED], exec_script=true,
force_script=true) at parse.y:3336
#14 0x40468308 in source_file ([EMAIL PROTECTED], [EMAIL PROTECTED])
at parse.y:3613
#15 0x402eeac4 in execute_pkg_add_or_del ([EMAIL PROTECTED],
[EMAIL PROTECTED]) at load-path.cc:1232
#16 0x402eedd8 in execute_pkg_add ([EMAIL PROTECTED]) at load-path.cc:1240
#17 0x402f3960 in load_path::do_set (this=0xf38c80, [EMAIL PROTECTED],
warn=false)
at load-path.cc:454
#18 0x402f3c1c in load_path::do_initialize (this=0xf38c80,
set_initial_path=true) at load-path.cc:382
#19 0x404427b8 in octave_main (argc=1, argv=0xbedd6b14, embedded=0)
at load-path.h:51
#20 0x424de040 in __libc_start_main () from /lib/libc.so.6
Backtrace stopped: frame did not save the PC
(gdb) up 2
#2 0x401f7f54 in octave_dynamic_loader::load_oct ([EMAIL PROTECTED],
[EMAIL PROTECTED], relative=false) at dynamic-ld.cc:508
508 ? instance-do_load_oct (fcn_name, file_name, relative) : false;
Current language: auto; currently c++
(gdb) p fcn_name
$1 = (const string ) @0xbedd5664: {static npos = 4294967295,
_M_dataplus = {std::allocatorchar = {__gnu_cxx::new_allocatorchar =
{No data fields}, No data fields}, _M_p = 0xf662ec dispatch}}
(gdb) p file_name
$2 = (const string ) @0xbedd5660: {static npos = 4294967295,
_M_dataplus = {std::allocatorchar = {__gnu_cxx::new_allocatorchar =
{No data fields}, No data fields},
_M_p = 0xf7247c
/usr/lib/octave/3.0.0/oct/arm-unknown-linux-gnu/dispatch.oct}}
(gdb)
Soeren
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
On Fri, 2008-01-04 at 10:19 +0100, Thomas Weber wrote: Am Donnerstag, den 03.01.2008, 10:56 +0100 schrieb Soeren Sonnenburg: Package: octave3.0 Version: 3.0.0-1 Severity: grave just try echo version | octave or octave on an arm machine: $ uname -a Linux arm 2.6.18-4-versatile #1 Sun Jun 17 14:34:28 UTC 2007 armv5tejl GNU/Linux Is this real or emulated hardware? More infos: I've rebuild octave3.0 on my arm emulator (takes 1 day, I am currently also building it on mips but that is another issue) and octave from the resulting packages still crashes. So it is not a bug in the arm bot but in octave. gdb says: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x40022480 (LWP 944)] 0x40f94320 in octave_shlib::make_shlib () from /usr/lib/octave-3.0.0/liboctave.so (gdb) bt #0 0x40f94320 in octave_shlib::make_shlib () from /usr/lib/octave-3.0.0/liboctave.so #1 0x402bb40c in octave_shlib::octave_shlib () from /usr/lib/octave-3.0.0/liboctinterp.so #2 0x402b9e64 in octave_dynamic_loader::do_load_oct () from /usr/lib/octave-3.0.0/liboctinterp.so #3 0x402ba668 in octave_dynamic_loader::load_oct () from /usr/lib/octave-3.0.0/liboctinterp.so #4 0x405781b0 in load_fcn_from_file () from /usr/lib/octave-3.0.0/liboctinterp.so #5 0x4057879c in load_fcn_from_file () from /usr/lib/octave-3.0.0/liboctinterp.so #6 0x406256f0 in lookup () from /usr/lib/octave-3.0.0/liboctinterp.so #7 0x408cb3c0 in tree_identifier::do_lookup () from /usr/lib/octave-3.0.0/liboctinterp.so #8 0x408cba84 in tree_identifier::rvalue () from /usr/lib/octave-3.0.0/liboctinterp.so #9 0x408cadcc in tree_identifier::rvalue () from /usr/lib/octave-3.0.0/liboctinterp.so #10 0x408d1220 in tree_index_expression::rvalue () from /usr/lib/octave-3.0.0/liboctinterp.so #11 0x40907234 in tree_statement::eval () from /usr/lib/octave-3.0.0/liboctinterp.so ---Type return to continue, or q return to quit--- #12 0x40907760 in tree_statement_list::eval () from /usr/lib/octave-3.0.0/liboctinterp.so #13 0x40575f58 in parse_and_execute () from /usr/lib/octave-3.0.0/liboctinterp.so #14 0x40576c04 in ?? () from /usr/lib/octave-3.0.0/liboctinterp.so Soeren -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: octave3.0 crashes on arm
Am Samstag, den 05.01.2008, 10:14 +0100 schrieb Rafael Laboissiere: * Rafael Laboissiere [EMAIL PROTECTED] [2008-01-03 21:53]: In order to investigate why Octave 3.0 is failing to work on arm (Bug#458852, severity grave), I would need octave3.0_3.0.0-1 installed on a sid chroot on one of the arm project machines. Peter Palfrader installed octave3.0, as well as gdb and strace, on the sid chroot of agnesi.debian.org. I could replicate the segfault at startup bug. I am attaching below the output of strace and gdb run on octave. It seems to fail in the install_defaults function, which is defined in src/defaults.cc. The output of strace is quite cryptic to me. If someone has an idea on how to proceed furthur, please drop a message. Some data points from arm buildd logs: 1:2.9.19-2 [grieg] works 1:2.9.19-1 [cats]segfault 1:2.9.18-1 [hedges] segfault 1:2.9.17-1 [smackdown} works 1:2.9.16-1 [hedges] segfault 3.0.0-1[hedges] segfault buildds are named in brackets. Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
Am Donnerstag, den 03.01.2008, 10:56 +0100 schrieb Soeren Sonnenburg: Package: octave3.0 Version: 3.0.0-1 Severity: grave just try echo version | octave or octave on an arm machine: $ uname -a Linux arm 2.6.18-4-versatile #1 Sun Jun 17 14:34:28 UTC 2007 armv5tejl GNU/Linux Is this real or emulated hardware? Thomas -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: [Pkg-octave-devel] Bug#458852: octave3.0 crashes on arm
On Fri, 2008-01-04 at 10:19 +0100, Thomas Weber wrote: Am Donnerstag, den 03.01.2008, 10:56 +0100 schrieb Soeren Sonnenburg: Package: octave3.0 Version: 3.0.0-1 Severity: grave just try echo version | octave or octave on an arm machine: $ uname -a Linux arm 2.6.18-4-versatile #1 Sun Jun 17 14:34:28 UTC 2007 armv5tejl GNU/Linux Is this real or emulated hardware? Both. The arm buildbot is on real hw and there echo version | octave crashes. http://buildd.debian.org/~jeroen/status/package.php?suite=unstablep=shogun#fail-shogun-arm I could reproduce that problem on emulated hardware (following aureliens qemu arm debian howto) including the crash when one just types octave. Soeren -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#458852: octave3.0 crashes on arm
Package: octave3.0 Version: 3.0.0-1 Severity: grave just try echo version | octave or octave on an arm machine: $ uname -a Linux arm 2.6.18-4-versatile #1 Sun Jun 17 14:34:28 UTC 2007 armv5tejl GNU/Linux $ octave GNU Octave, version 3.0.0 Copyright (C) 2007 John W. Eaton and others. This is free software; see the source code for copying conditions. There is ABSOLUTELY NO WARRANTY; not even for MERCHANTIBILITY or FITNESS FOR A PARTICULAR PURPOSE. For details, type `warranty'. Octave was configured for arm-unknown-linux-gnu. Additional information about Octave is available at http://www.octave.org. Please contribute if you find this software useful. For more information, visit http://www.octave.org/help-wanted.html Report bugs to [EMAIL PROTECTED] (but first, please read http://www.octave.org/bugs.html to learn how to write a helpful report). For information about changes from previous versions, type `news'. panic: Segmentation fault -- stopping myself... attempting to save variables to `octave-core'... save to `octave-core' complete Segmentation fault -- System Information: Debian Release: lenny/sid APT prefers stable APT policy: (700, 'stable'), (650, 'testing'), (600, 'unstable'), (1, 'experimental') Architecture: i386 (i686) Kernel: Linux 2.6.24-rc6-sonne (SMP w/2 CPU cores; PREEMPT) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash Versions of packages octave3.0 depends on: ii atlas3-base [liblapack 3.6.0-20.6Automatically Tuned Linear Algebra ii atlas3-sse2 [liblapack 3.6.0-20.6Automatically Tuned Linear Algebra ii lapack3 [liblapack.so. 3.0.2531a-6.1 library of linear algebra routines ii libc6 2.7-5 GNU C Library: Shared libraries ii libcurl3-gnutls7.17.1-1 Multi-protocol file transfer libra ii libfftw3-3 3.1.2-3 library for computing Fast Fourier ii libg2c01:3.4.6-6 Runtime library for GNU Fortran 77 ii libgcc11:4.2.2-5 GCC support library ii libglpk0 4.25-1linear programming kit with intege ii libhdf5-serial-1.6.5-0 1.6.5-5+b1Hierarchical Data Format 5 (HDF5) ii libncurses55.6+20071215-1Shared libraries for terminal hand ii libpcre3 7.3-2 Perl 5 Compatible Regular Expressi ii libqhull5 2003.1-8 calculate convex hulls and related ii libreadline5 5.2-3 GNU readline and history libraries ii libstdc++6 4.2.2-5 The GNU Standard C++ Library v3 ii libsuitesparse 3.0.0-7 collection of libraries for comput ii refblas3 [libblas.so.3 1.2-8 Basic Linear Algebra Subroutines 3 ii texinfo4.11.dfsg.1-3 Documentation system for on-line i ii zlib1g 1:1.2.3.3.dfsg-8 compression library - runtime Versions of packages octave3.0 recommends: ii atlas3-base 3.6.0-20.6 Automatically Tuned Linear Algebra ii gnuplot 4.2.2-1A command-line driven interactive -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
