Swig is used to generate C source and it has trouble processing opensslconf.h
sometimes. So, we help it out when multilib variants exist.
For native builds, the variant header will never exist.
Specific multi-lib header files might be named *-32.h or *-n32.h, so we check
for both names. We also might check for *-n64.h although that will never exist.
It appears that *-o??.h will never exist, so we don't check for that.
Signed-off-by: Joe Slater <jslater@windriver.com>
Signed-off-by: Chong Lu <Chong.Lu@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Changes:
- rename SUMMARY with length > 80 to DESCRIPTION
- rename DESCRIPTION with length < 80 to (non present tag) SUMMARY
- drop final point character at the end of SUMMARY string
- remove trailing whitespace of SUMMARY line
Note: don't bump PR
Signed-off-by: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
After python-m2crypto was reported as failed task in bitbake world
on 2014-02-08 by Martin Jansa, I investigated python-m2crypto with
MACHINE = "qemux86-64" setting. When compiling python-m2crypto for
qemux86-64, the setup.py aborts in the swig call with:
| swig -python -I[...]/tmp-eglibc/sysroots/qemux86-64/usr/include/python2.7 -I[...]/tmp-eglibc/sysroots/qemux86-64/usr/include -includeall -o SWIG/_m2crypto_wrap.c SWIG/_m2crypto.i
| [...]/tmp-eglibc/sysroots/qemux86-64/usr/include/openssl/opensslconf.h:41: Error: Unable to find 'openssl/opensslconf-32.h'
| error: command 'swig' failed with exit status 1
| ERROR: python setup.py build_ext execution failed.
| WARNING: exit code 1 from a shell command.
To compile for x86-64 architecture, opensslconf.h shall include
opensslconf-64.h, not opensslconf-32.h. Properly including
opensslconf-64.h can be configured through the bits/wordsize.h
header and setting the __x86_64__ directive in the swig call. To
set the directive, the SWIG_FEATURES variable for x86-64 is
set and exported in the recipe.
Compiling python-m2crypto for qemumips still fails and needs
further adjustments, but is not addressed here.
This patch follows Khem Raj's suggestion on the openembedded-devel
mailing list after submission of the first commit to this issue.
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@oss.bmw-carit.de>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
