| Age | Commit message (Collapse) | Author | Files |
|---|
|
without this patch, using angstrom 2010.x on armv6 I actually get :
| checking alignment of short... configure: error:
cannot compute alignment of short, 77
| ERROR: Function do_configure failed
NOTE: package postgresql-8.1.8-r4: task do_configure: Failed
Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
commit 21cc37782c0624d1efc9b56e94d1012557dce871
Author: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
Date: Thu Mar 11 09:46:25 2010 +0100
libsndfile1: updated to 1.0.21
also fixed download location and moved two config vars to arm-common
(not sure about ac_cv_sizeof_off_t=8, this seems odd or recipe
specific, so left that one)
Problem is that ac_cv_sys_file_offset_bits should not be cached blindly
as the same var is used by glib-2.0 to detect large file system support
if cached then glib-2.0 configure tests are not perfored and it assumes
that large file support is there. uclibc is configurable to not have
large file support compiled in hence this breaks glib-2.0 in a obscure
way hence moving back to where they were.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
* gl_cv_func_unsetenv_works is not computed correctly in cross
environment so cache it.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Acked-by: Martin Jansa <Martin.Jansa@gmail.com>
|
|
need to check)
|
|
Signed-off-by: Chris Larson <chris_larson@mentor.com>
|
|
also fixed download location and moved two config vars to arm-common
(not sure about ac_cv_sizeof_off_t=8, this seems odd or recipe specific, so left that one)
Signed-off-by: Frans Meulenbroeks <fransmeulenbroeks@gmail.com>
|
|
|
|
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
* Fixes PR 5320
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
mbstring that cannot be checked
This enables UTF-8 support in OE php
|
|
The rpm build is failing on the internal db3 build because it thinks
db_cv_mutex=no is telling it that "no" is the mutext type. Setting
this to proper values for the platforms that didn't have them set
already should fix the rpm builds.
Signed-off-by: Brian Pomerantz <bapper@mvista.com>
|
|
Signed-off-by: Chris Larson <clarson@mvista.com>
|
|
Signed-off-by: Jeremy Puhlman <jpuhlman@mvista.com>
|
|
|
|
|
|
These numbers were generated using qemu and a mips32 big
endian userspace on a simulated malta. If we ever support
mips64 we will have to split up mips-common.
|
|
* Disable parallel make, mono does not like it.
* Add a patch to get mono cross compile for x86 target.
* add mono_cv_sizeof_sunpath to site files, becasue
previously it was called cv_mono_sizeof_sunpath
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
* This patch reverts the cosmetic change which used linux-gnu instead of linux
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Tested-by: Holger Freyther <zecke@selfish.org>
|
|
* After renaming the target triplet the file names
needs to be renamed because they are used in get_siteinfo_list
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
Enca's configure will fail on everything not arm or ix86.
Signed-off-by: Graham Gower <graham.gower@gmail.com>
|
|
If the other site files are accurate, almost all of the bash vars
are universal across Linux.
Did this to fix bash on x86_64. Also bump bash PRs to help track
breakage later.
Signed-off-by: Michael Smith <msmith@cbnco.com>
Acked-by: Tom Rini <trini@embeddedalley.com>
Acked-by: Khem Raj <raj.khem@gmail.com>
|
|
ix86-common had a few updates in the past couple of years; bring
them to x86_64.
Some of these could probably be moved to common-glibc.
Signed-off-by: Michael Smith <msmith@cbnco.com>
|
|
The results were found using google and applying
common sense.
|
|
ac_cv_sizeof_unsigned_char definition accidentally redefines
ac_cv_sizeof_unsigned_int to 1. Luckily, not many packages are using that
but for example it completely breakes gnutls.
Fix that.
Acked-by: Holger Freyther <zecke@selfish.org>
|
|
guile's configure is not smart to guess if csqrt works
in a cross compile environment so it ends up assuming
that it works and fails to link. Anyways for uclibc
complex functions are unimplemented so we help configure
by unsetting here in site file for uclibc.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
On powerpc it's not posible to assign va_arg by value, so
cache ac_cv_va_val_copy=no
Signed-off-by: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com>
|
|
|
|
|
|
common-uclibc: Ditto
mipsel: Add mipsel config for uclibc
|
|
- df can only be build if one of the many statvfs,fsusage
was is present. These tests work by executing a test application
and that is obviously not working out when cross compiling
- all linux (glibc, eglibc, uclibc) should have the bsize in the
struct statvfs... and this will enable the build of df.
|
|
common-linux site file.
As reported by me in bug 4854, the IFACE autoconf variables are found
in some, bug not all arch site files, but these are really common linux
variables and as such, should be in common-linux. Bump PR as there is a
change here in some arches.
|
|
- This does configure
- But it fails to link, there is something wrong with the
autoreconf...
|
|
|
|
|
|
|
|
|
|
pointing this out.
|
|
|
|
|
|
0.9.29.
|
|
actually, it's not a fix, but rather an attemt to mimick its behaviour on other
systems, which is wrong, anyway, since it attempts checks against build system.
But since it works, perhaps it should do for now. On the other hand, maybe opkg
crashes somewhere around here, as gpgme depends on pth.
BTW, long time no commits ;-)
|
|
|
|
entry
|
|
|
|
|
|
|
|
|
