On 7/18/2012 1:01 PM, Yaakov (Cygwin/X) wrote:
On 2012-07-18 03:51, marco atzeri wrote:
.mex are dll not exe files.
I thought Octave modules were .oct?
standard yes.
mex are special Matlab compatible ones, in theory they should
work with both Matlab and Octave...
on previous nan-2.5.2, included in octave-forge-20120224-1
they were correctly identified as .mex
on octave-forge-20120714-1, nan are wrong and tsa are right
I will look on what caused the change, as temporary workaround
remove/copy the *.mex.exe to *.mex files.
That would be my fault. :-( This should be fixed in cygport git master
now; sorry for the inconvenience.
I noticed that outside cygport, the problem do not arise.
I am also investigating why tsa mex were not changed, and something is
clearly wrong in the octave-forge build system
$ file *.mex
covm_mex.mex: ELF 64-bit LSB shared object, x86-64, version 1
(SYSV), dynamically linked,
BuildID[sha1]=0x7b9f63a496a41bb490ebb7d6239da5c7dbb2e130, not stripped
sumskipnan_mex.mex: ELF 64-bit LSB shared object, x86-64, version 1
(SYSV), dynamically linked,
BuildID[sha1]=0x38057facd70b6f315b722c6749d6b872d425b619, not stripped
while before
$ file *.mex
covm_mex.mex: PE32 executable (DLL) (console) Intel 80386, for MS
Windows
sumskipnan_mex.mex: PE32 executable (DLL) (console) Intel 80386, for MS
Windows
it seems a upstream package bug....
Yaakov
Marco
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
