The symbolic package contains a few references to lists and therefore fails to compile in octave-3.4. The attached patches were originally posted here: http://sourceforge.net/mailarchive/message.php?msg_id=27256205.
patch-src-symbols.cc.diff
Description: Binary data
patch-src-symlsolve.cc.diff
Description: Binary data
These only allow symbolic to compile but do not solve the seg fault issue
already seen on mac os x. I use MacPorts to install octave on a Macbook Pro
running Mac OS X 10.6.8 (snow leopard). The Octave port is currently @ v3.2.4;
symbolic will compile without the patches under 3.2.4.
Please let me know what you would like try next in order to find the root cause
for the seg fault issue.
-Mark
On Jul 29, 2011, at 4:26 PM, Søren Hauberg wrote:
> Hi
>
> I'm sorry about the late reply, but I've been busy with a conference
> lately.
> Also, please reply to the list (press "Reply to All" instead of "Reply"
> in your e-mail). That way your mail is more likely to get picked up by
> other helpful people :-)
>
> lør, 23 07 2011 kl. 08:16 -0500, skrev mark brethen:
>> I'd like to get to the bottom of this as well, but will need some
>> guidance, since I'm new to octave.
>
> Wonderful! We could need help.
>
>> I have installed a newer version of octave, this MacPorts distribution
>> is 'octave-devel @3.4.2_0+fltk+gcc44' and the symbolic package fails
>> to compile.
>
> Could you post the error you get?
>
>> Has anyone tried the patch below which was posted in the gnu octave
>> repository?
>
> Can you try to send this patch as an attachment instead of inside the
> body of the e-mail? I had some troubles reading the patch.
>
> Thanks
> Søren
>
>>
>> diff-Naur symbolic-1.0.9/src/symbols.cc symbolic-1.0.9.new/src/symbols.cc
>> Symbolic-1.0.9/src/symbols.cc 2009-06-07 03:45:55.000000000 -0700 ---
>> + + + 2011-03-24 17:42:29.139600452 -0700 symbolic-1.0.9.new/src/symbols.cc
>> @ @ @ @ +373.7 -373.7
>>
>>
>> gripe_wrong_type_arg ("subs", args (0));
>>
>>
>> return retval;
>>
>>
>> }
>> - If (!. (Args (1) is_list () | | args (1) is_cell ()).) {
>> + If ((args (1). Is_cell ())) {
>>
>>
>> if (! get_symbol (args (1), the_sym)) {
>>
>>
>> gripe_wrong_type_arg ("subs", args (1));
>>
>>
>> return retval;
>> diff-Naur symbolic-1.0.9/src/symlsolve.cc symbolic-1.0.9.new/src/symlsolve.cc
>> Symbolic-1.0.9/src/symlsolve.cc 2009-06-07 03:45:55.000000000 -0700 ---
>> + + + 2011-03-24 17:42:37.215600451 -0700 symbolic-1.0.9.new/src/symlsolve.cc
>> @ @ @ @ +46.7 -46.7
>>
>>
>> }
>>
>>
>>
>> try {
>> - If (. Args (0) is_list () | | args (0) is_cell (.)) {
>> + If (args (0). Is_cell ()) {
>>
>>
>> octave_value_list oct_eqn_list (. args (0) list_value ());
>>
>>
>> for (i = 0; i <oct_eqn_list.length (); i + +) {
>>
>>
>> if (! get_relation (oct_eqn_list (i), relation)) {
>> @ @ @ @ +67.7 -67.7
>>
>>
>> eqns.append (relation);
>>
>>
>> }
>>
>>
>>
>> - If (. Args (1) is_list () | | args (1) is_cell (.)) {
>> + If (args (1). Is_cell ()) {
>>
>>
>> octave_value_list oct_vars (. args (1) list_value ());
>>
>>
>> for (i = 0; i <oct_vars.length (); i + +) {
>>
>>
>> if (! get_symbol (oct_vars (i), expression)) {
>>
>>
>> -Mark
>>
>>
>>
>> On Jul 23, 2011, at 3:03 AM, Søren Hauberg wrote:
>>
>>> Hi
>>>
>>> fre, 22 07 2011 kl. 09:24 -0500, skrev mark brethen:
>>>> octave:1> symbols
>>>> octave:2> x = sym("x")
>>>> octave(95278) malloc: *** error for object 0x7fff70f01500: pointer being
>>>> freed was not allocated
>>>> *** set a breakpoint in malloc_error_break to debug
>>>> panic: Abort trap -- stopping myself...
>>>> attempting to save variables to `octave-core'...
>>>> save to `octave-core' complete
>>>> Abort trap
>>>
>>>
>>> We have had reports along this line previously from Mac users. None of
>>> us Linux developers have, however, been able to reproduce the bug, so
>>> we've had quite a hard time fixing it (we have not been able to fix it).
>>> Can you possibly help us with trying to figure out where the error
>>> occurs?
>>>
>>> Thanks
>>> Søren
>>>
>>
>
>
------------------------------------------------------------------------------ Got Input? Slashdot Needs You. Take our quick survey online. Come on, we don't ask for help often. Plus, you'll get a chance to win $100 to spend on ThinkGeek. http://p.sf.net/sfu/slashdot-survey
_______________________________________________ Octave-dev mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/octave-dev
