[U-Boot-Users] libgcc pickup....
Evening...
Does someone has an idea why on OSX systems compiling u-boot picks up host's libgcc.a in the end?
Means I always have to use:
PLATFORM_LIBS += --no-warn-mismatch `$(CC) $(CFLAGS) -print-libgcc-file-name`
instead of:
PLATFORM_LIBS += --no-warn-mismatch -L $(shell dirname `$(CC) $(CFLAGS) -print-libgcc-file-name`) -lgcc
best regards rick
In message <r02010400-1037-A42C1278762611D9B21600039387ACB6@[10.0.1.1]> you wrote:
Evening...
Monring... :-)
Does someone has an idea why on OSX systems compiling u-boot
I don't even have an idea what OSX systems are.
picks up host's libgcc.a in the end?
I think I just checked in a patch for this problem:
* Removed '--no-warn-mismatch' option from Makefile. This option makes 'ld' to overlook binary objects compatibility.
* Moved $(PLATFORM_LIBS) from the library group (--start-group ... --end-group) outside of the group. This will make 'ld' to do _multiple_ search in the library group when resolving symbol references and do only a _single_ seach in libgcc.a after the group search.
Best regards,
Wolfgang Denk
In message <r02010400-1037-A42C1278762611D9B21600039387ACB6@[10.0.1.1]> you wrote:
Evening...
Monring... :-)
Does someone has an idea why on OSX systems compiling u-boot
I don't even have an idea what OSX systems are.
Those tiny little shiny and overpriced boxes from Apple? (o;
picks up host's libgcc.a in the end?
I think I just checked in a patch for this problem:
Removed '--no-warn-mismatch' option from Makefile. This option makes 'ld' to overlook binary objects compatibility.
Moved $(PLATFORM_LIBS) from the library group (--start-group ... --end-group) outside of the group. This will make 'ld' to do _multiple_ search in the library group when resolving symbol references and do only a _single_ seach in libgcc.a after the group search.
Hmm..Makefile in CVS is 3 weeks old as per now....
But with your changes suggested I still get the same:
UNDEF_SYM=`bfin-elf-objdump -x lib_generic/libgeneric.a board/stamp/libstamp.a cpu/bf533/libbf533.a lib_blackfin/libblackfin.a fs/cramfs/libcramfs.a fs/fat/libfat.a fs/fdos/libfdos.a fs/jffs2/libjffs2.a fs/reiserfs/libreiserfs.a fs/ext2/libext2fs.a net/libnet.a disk/libdisk.a rtc/librtc.a dtt/libdtt.a drivers/libdrivers.a drivers/sk98lin/libsk98lin.a post/libpost.a post/cpu/libcpu.a common/libcommon.a |sed -n -e 's/.*(__u_boot_cmd_.*)/-u\1/p'|sort|uniq`;\
bfin-elf-ld -Bstatic -T /Users/klingler/develop/blackfin/u-boot/board/stamp/u-boot.lds -Ttext 0x07FC0000 $UNDEF_SYM cpu/bf533/start.o cpu/bf533/start1.o cpu/bf533/interrupt.o \ --start-group lib_generic/libgeneric.a board/stamp/libstamp.a cpu/bf533/libbf533.a lib_blackfin/libblackfin.a fs/cramfs/libcramfs.a fs/fat/libfat.a fs/fdos/libfdos.a fs/jffs2/libjffs2.a fs/reiserfs/libreiserfs.a fs/ext2/libext2fs.a net/libnet.a disk/libdisk.a rtc/librtc.a dtt/libdtt.a drivers/libdrivers.a drivers/sk98lin/libsk98lin.a post/libpost.a post/cpu/libcpu.a common/libcommon.a --end-group \ -L /usr/local/blackfin/lib/gcc/bfin-elf/3.4.1 -lgcc -Map u-boot.map -o u-boot /usr/lib/libgcc.a: could not read symbols: Archive has no index; run ranlib to add one make: *** [u-boot] Error 1
greets from .fi rick
In message <r02010400-1037-2E5884BA763A11D9B21600039387ACB6@[10.0.1.1]> you wrote:
I don't even have an idea what OSX systems are.
Those tiny little shiny and overpriced boxes from Apple? (o;
Umm... isn't this "OS X", then?
Hmm..Makefile in CVS is 3 weeks old as per now....
I guess that's the usual delay between SF's developer machine and the anonymous access mache(s).
But with your changes suggested I still get the same:
...
/usr/lib/libgcc.a: could not read symbols: Archive has no index; run ranlib to add one
I can't help it, but this looks like a pretty helpful error message to me, and it smells somewhat like a tool problem on your build host.
Best regards,
Wolfgang Denk
participants (2)
-
richard＠uclinux.net -
Wolfgang Denk