glibc 2.16.0
白い熊
ShiroiKuma@ShiroiKuma.org
Mon Dec 17 19:26:00 GMT 2012
On Sun, Dec 16, 2012 at 9:22 PM, Bernhard Walle <bernhard@bwalle.de> wrote:
> after "make" and before "make install". Or just edit onfig/libc/glibc.in
> manually...
Thanks for the how-to. I edited the config/libc/glibc.in and
recompiled the toolchaing with glibc-2.16.0... So far so good.
Now, however I need serious help, since I can't solve the following issue.
I've built glibc-2.16.0 with it and compiled a host of GNU apps, for
instance tar. I'm doing this for an Android phone, so I get full glibc
and GNU utilities natively on it. Using the GNU tools then is as
simple as bindmounting glibc-2.16.1/lib to /lib and exporting
LD_LIBRARY_PATH=/path/to/gcc-4.7.2/lib
Now, the big question: For instance the tar binary I've compiled with
it doesn't run, gives me:
/system/bin/sh: ./tar: No such file or directory
which on Android signifies, the glibc is not present.
However, if I bindmount the glibc-2.14.1 to /lib, the glibc-2.14.1
being the glibc that I've crosscompiled before this with a
glibc-2.14.1 toolchain, the tar runs.
Leading me to believe the tar is built against glibc-2.14.1
But how can this be?
If I check ~/x-tools/arm-exynos-linux-gnueabi/arm-exynos-linux-gnueabi/sysroot/lib
it has ld-2.16.so libc-2.16.so and so on?
So how come the binaries run against a 2.14.1 glibc but not against a
2.16.0? I'm cracking my brains here, I'm sure there's a simple
oversight somewhere on my part, but can't figure it out.
Any ideas would be much appreciated. Many thanks in advance
--
白い熊
--
For unsubscribe information see http://sourceware.org/lists.html#faq
More information about the crossgcc
mailing list