Issues of Sage 6.9 for Mac OS X 10.10

148 views
Skip to first unread message

Yoshihiro Sato

unread,
Nov 7, 2015, 2:47:23 AM11/7/15
to sage-support
I tried to install Sage 6.9 for Mac OS X 10.10. However it does not work.The log is as follows.
I hope it will be updated.
Thanks,
Y.Sato

$ sage
┌────────────────────────────────────────────────────────────────────┐
│ SageMath Version 6.9, Release Date: 2015-10-10                     │
│ Type "notebook()" for the browser-based notebook interface.        │
│ Type "help()" for help.                                            │
└────────────────────────────────────────────────────────────────────┘
The Sage installation tree has moved
from /Users/buildslave-sage/slave/sage_git/build
  to /Applications/Sage-6.9.app/Contents/Resources/sage
Updating various hardcoded paths...
(Please wait at most a few minutes.)
DO NOT INTERRUPT THIS.
Done updating paths.
------------------------------------------------------------------------
Unhandled SIGILL: An illegal instruction occurred in Sage.
This probably occurred because a *compiled* component of Sage has a bug
in it and is not properly wrapped with sig_on(), sig_off().
Sage will now terminate.
------------------------------------------------------------------------
/usr/local/bin/sage: line 2:  3829 Illegal instruction: 4  $SAGE_ROOT/sage $@
$

Dima Pasechnik

unread,
Nov 7, 2015, 7:11:40 PM11/7/15
to sage-support
Hi,
this is a typical error caused by an attempt to use an executable for a higher end CPU.
What is the CPU of your computer?

I guess we need a build for Core2 Duo...

Cheers,
Dima

Yoshihiro Sato

unread,
Nov 8, 2015, 3:26:32 AM11/8/15
to sage-support
Hi,

The CPU of my computer is Intel Core i7.

Sage 6.7 binary works on my pc.

Thanks,
Y.Sato

2015年11月8日日曜日 9時11分40秒 UTC+9 Dima Pasechnik:

Dima Pasechnik

unread,
Nov 8, 2015, 6:03:35 AM11/8/15
to sage-support


On Sunday, 8 November 2015 08:26:32 UTC, Yoshihiro Sato wrote:
Hi,

The CPU of my computer is Intel Core i7.

and you're trying sage-6.9-x86_64-Darwin-OSX_10.10_x86_64.dmg, I presume?

could we have more details?
(which are usually to be found in the output of "About this Mac")

Thanks!

Yoshihiro Sato

unread,
Nov 8, 2015, 9:29:25 AM11/8/15
to sage-support
Hi,

Thank you.
I have two Mac pc.

1.  Sage-6.9 works well on the ffirst pc.

OS X Yosemite  version 10.10.5
MacBook Pro (17-inch, Late 2011)
CPU:  2.4 GHz Intel Core i7
memory:  16 GB 1600 MHz DDR3
Graphics:  AMD Radeon HD 6770M 1024 MB

2.  In the second pc Sage-6.9 does not work.  Sage-6.7 works well.

OS X Yosemite version 10.10.5
Mac Pro (Mid 2012)
CPU: 2 x 2.4 GHz 6-Core Intel Xenon
Memory: 28 GB 1333 MHz DDR3 ECC
Graphics: ATI Radeon HD 5770 1025 MB


$ sage
┌───────────────   ────────────────┐
│ SageMath Version 6.9, Release Date: 2015-10-10                                  │
│ Type "notebook()" for the browser-based notebook interface.                │
│ Type "help()" for help.                                                                               │
└────────────────────────────────┘
------------------------------------------------------------------------
Unhandled SIGILL: An illegal instruction occurred in Sage.
This probably occurred because a *compiled* component of Sage has a bug
in it and is not properly wrapped with sig_on(), sig_off().
Sage will now terminate.
------------------------------------------------------------------------
/usr/local/bin/sage: line 2:  1517 Illegal instruction: 4  $SAGE_ROOT/sage $@
$

Thanks,
Y.Sato

2015年11月8日日曜日 20時03分35秒 UTC+9 Dima Pasechnik:

Volker Braun

unread,
Nov 8, 2015, 1:42:31 PM11/8/15
to sage-support
Can somebody debug where the sigill is coming from? Use gdb to narrow down the library? (e.g. sage -f gdb and then run sage -gdb as root on OSX)

Dima Pasechnik

unread,
Nov 8, 2015, 3:24:40 PM11/8/15
to sage-support


On Sunday, 8 November 2015 14:29:25 UTC, Yoshihiro Sato wrote:
Hi,

Thank you.
I have two Mac pc.

1.  Sage-6.9 works well on the ffirst pc.

OS X Yosemite  version 10.10.5
MacBook Pro (17-inch, Late 2011)
CPU:  2.4 GHz Intel Core i7
memory:  16 GB 1600 MHz DDR3
Graphics:  AMD Radeon HD 6770M 1024 MB

2.  In the second pc Sage-6.9 does not work.  Sage-6.7 works well.

OS X Yosemite version 10.10.5
Mac Pro (Mid 2012)
CPU: 2 x 2.4 GHz 6-Core Intel Xenon
Memory: 28 GB 1333 MHz DDR3 ECC
Graphics: ATI Radeon HD 5770 1025 MB

OK, so it's on Xenon CPU that you have a problem (Xenon != i7)

Dima Pasechnik

unread,
Nov 8, 2015, 3:26:07 PM11/8/15
to sage-support


On Sunday, 8 November 2015 18:42:31 UTC, Volker Braun wrote:
Can somebody debug where the sigill is coming from? Use gdb to narrow down the library? (e.g. sage -f gdb and then run sage -gdb as root on OSX)

I'll try on a Core2 Duo running 10.11 - not sure it will give us the right info though...
 

Dima Pasechnik

unread,
Nov 8, 2015, 4:38:46 PM11/8/15
to sage-support


On Sunday, 8 November 2015 20:26:07 UTC, Dima Pasechnik wrote:


On Sunday, 8 November 2015 18:42:31 UTC, Volker Braun wrote:
Can somebody debug where the sigill is coming from? Use gdb to narrow down the library? (e.g. sage -f gdb and then run sage -gdb as root on OSX)

I'll try on a Core2 Duo running 10.11 - not sure it will give us the right info though...

Sorry, this does not fly - I get an internal compiler error while trying to  sage -f gdb...

Christian Nassau

unread,
Nov 9, 2015, 8:00:16 AM11/9/15
to sage-s...@googlegroups.com
On 08.11.2015 19:42, Volker Braun wrote:
> Can somebody debug where the sigill is coming from? Use gdb to narrow
> down the library? (e.g. sage -f gdb and then run sage -gdb as root on OSX)

I cannot test the 6.9 on my system (without rebooting with disabled
security). But the current sage-6.10.beta1-x86_64-Darwin (with md5sum =
f5eba161c01c37aa0422519dd4f028d6) also has problems, even though it has
presumably been built with SAGE_FAT_BINARY=yes

Program received signal SIGILL, Illegal instruction.
0x000000010ab1e0b0 in __gmpf_set_d ()
from
/Volumes/sage-6.10.beta1-x86_64-Darwin/sage/local/lib/libmpir.16.dylib

On my system, I get this from the mpir platform detection code:

el-capitan:mpir-2.7.0 cn$ /bin/sh ./configfsf.guess
x86_64-apple-darwin15.0.0
el-capitan:mpir-2.7.0 cn$ /bin/sh ./config.guess
core2-apple-darwin15.0.0

The first is what's used in SAGE_FAT_BINARY=yes, the 2nd gives a working
configuration for me. Maybe we should force configfsf.guess to return
"core2" also... ?

Volker Braun

unread,
Nov 9, 2015, 8:53:39 AM11/9/15
to sage-support
I've opened http://trac.sagemath.org/ticket/19560 for this issue
Reply all
Reply to author
Forward
0 new messages