Just checking before trying to upgrade...
Thanks,
Jason
Yes, app store, and $4.99 or something like that to boot. Though it's
free if you already pay for their developer or iOS memberships.
>
> I can see this update potentially causing problems for the OS X App,
> if I understand the announcement correctly as being a major change in
> how things are built. I would strongly suggest testing that.
>
> Also, I think you can have more than one XCode at a time, right?
>
I don't know.
Jason
Aly Deines showed me what happened when she installed Xcode 4 then
built Sage. She build Sage-4.7.rc0 itself, and it compiled fine.
However, when she ran it, it crashed on startup. The traceback
implicates Pynac:
(gdb)
#0 0x0000000107a20c64 in PyInt_FromLong (ival=-5) at Objects/intobject.c:91
#1 0x00000001079971fa in GiNaC::Number_T::Number_T ()
#2 0x0000000107997280 in GiNaC::numeric::numeric ()
#3 0x00000001079e7bea in GiNaC::library_init::library_init ()
#4 0x00000001078708f8 in __static_initialization_and_destruction_0 ()
#5 0x00007fff5fc0d500 in
__dyld__ZN16ImageLoaderMachO18doModInitFunctionsERKN11ImageLoader11LinkContextE
()
#6 0x00007fff5fc0bcec in
__dyld__ZN11ImageLoader23recursiveInitializationERKNS_11LinkContextEj
()
#7 0x00007fff5fc0bc9d in
__dyld__ZN11ImageLoader23recursiveInitializationERKNS_11LinkContextEj
()
#8 0x00007fff5fc0bda6 in
__dyld__ZN11ImageLoader15runInitializersERKNS_11LinkContextE ()
#9 0x00007fff5fc08fbb in __dyld_dlopen ()
#10 0x00007fff86ad2360 in dlopen ()
#11 0x00000001000e88af in _PyImport_GetDynLoadFunc ()
#12 0x00000001000d2e04 in _PyImport_LoadDynamicModule ()
#13 0x00000001000d10df in import_submodule ()
Note that Ipython alone and python alone both start fine.
I should buy and install XCode4 on bsd.math.washington.edu, so it is
easier for people to debug this problem.
-- William
> --
> To post to this group, send an email to sage-...@googlegroups.com
> To unsubscribe from this group, send an email to sage-devel+...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/sage-devel
> URL: http://www.sagemath.org
>
--
William Stein
Professor of Mathematics
University of Washington
http://wstein.org
Could this be related to
http://trac.sagemath.org/sage_trac/ticket/11116
?
Solving that bug would be very nice, as it basically stops any further
progress on a 64-bit Solaris port whilst Pynac is crashing.
Dave
I thought about it too but the backtrace appears to be quite different.
Off course we cannot dismiss that solving #11116 won't have other
side effects. We alraedy know #11116 can happen on linux too.
Francois
This email may be confidential and subject to legal privilege, it may not reflect the views of the University of Canterbury, and it is not guaranteed to be virus free. If you are not an intended recipient, please notify the sender immediately and erase all copies of the message and any attachments. Please refer to http://www.canterbury.ac.nz/emaildisclaimer for more information.
Wow, that's incredible!
We should:
(1) add a remark right in the prereq spkg that checks that xcode 4
isn't being used (unless SAGE_PORT or something is set)
(2) if xcode 4 is being used (as detected by 1), the error message
should suggest xcode 3, and explain how to revert to it. (Is it easy
still?)
-- William
> For more options, visit this group at http://groups.google.com/group/sage-devel
>>> Even worse is behaviour with its clang compilers (they can't even
>>> compile several parts of Sage)
>>
>> Wow, that's incredible!
>>
>> We should:
>>
>> (1) add a remark right in the prereq spkg that checks that xcode 4 (4.0.1, to be precise)
>> isn't being used (unless SAGE_PORT or something is set)
>
> In fact, I don't know how to check this easily: it ships the same
> build of gcc as XCode 3 (3.2.6, to be precise)
> (gcc version 4.2.1 (Apple Inc. build 5666) (dot 3))
> and there is no ready way to check the version of the whole setup.
> The difference sits in libraries.
> One notable difference is that cc is not gcc, but a gcc-llvm build.
> (namely, gcc version 4.2.1 (Based on Apple Inc. build 5658) (LLVM
> build 2335.6)))
>
> So one can test that "cc -v" != "gcc -v", but this does nor seem to be
> robust enough.
>
> Probably one can use Applescript, but I have no clue about it.
Can you post the output of these two commands.
$ gcc -dM -E -xc /dev/null
$ cc -dM -E -xc /dev/null
It might help if you sorted them
to find the differences. i.e.
$ gcc -dM -E -xc /dev/null | sort > a
$ cc -dM -E -xc /dev/null | sort > b
$ diff a b
You might find there's something in there which can differentiate them. If there
is, we can add a specific test to the scripts $SAGE_LOCAL/bin/testcc.sh and
$SAGE_LOCAL/bin/testcxx.sh. Currently those scripts report one of
* GCC
* Sun_Studio
* HP_on_Tru64
* HP_on_HP-UX
* IBM_on_AIX
* HP_on_Alpha_Linux
* Unknown
Currently the scripts check if __GNUC__ is defined and if so its considered gcc
or a gcc-like compiler (e.g. Intel's icc). But if there's a need to
differentiate Clang, I'd be very surprised if there was not something we can use.
Dave
$ xcodebuild -version
Xcode 3.1.4
Component versions: DevToolsCore-1204.0; DevToolsSupport-1186.0
BuildVersion: 9M2809