Thanks, it's getting funny ;-) with your adjustments, the output of uname -m is now i48664 i'm not very good at c, could you help me one again, please? thanks so much!
toby Joe Ciccone wrote: > Tobias Vogel wrote: > >> Hi, >> thanks for that patch, >> i compiled it, but it seems not to work properly, instead of i486 >> uname now shows x46_64, so neither x86_64 nor i486. >> is there any other kernel-hack method? >> >> thanks, >> >> toby >> >> Joe Ciccone wrote: >> >> >>> Ken Moffat wrote: >>> >>> >>> >>>> On Mon, Oct 22, 2007 at 05:31:36PM +0200, Tobias Vogel wrote: >>>> >>>> >>>> >>>> >>>> >>>>> the configure-output seems quite normal to me: >>>>> >>>>> root:/sources/glibc-build# ../glibc-2.4/configure --prefix=/usr >>>>> --disable-profile --enable-add-ons --enable-kernel=2.6.0 >>>>> --libexecdir=/usr/lib/glibc >>>>> checking build system type... x86_64-unknown-linux-gnu >>>>> checking host system type... x86_64-unknown-linux-gnu >>>>> >>>>> >>>>> >>>>> >>>> Those last 2 lines are only normal if you are building for x86_64. >>>> And yes, in chroot on x86_64 the system will respond to 'uname -m' >>>> with x86_64 if the kernel is 64-bit. >>>> >>>> Basically, if you are following the 1.0.0 x86 book on x86_64, you >>>> should use the 'if you are going to boot' option and build a 32-bit >>>> kernel, copy it over, and then build chapters 9 and 10 on the target >>>> board. >>>> >>>> >>>> >>>> >>> Or, load a uname hack into the kernel right before you chroot so uname >>> -m reports i486 on the target system. >>> >>> http://www.cross-lfs.org/~jciccone/uname_ix86.c for 2.6.16 (I think) and >>> newer kernels or http://www.cross-lfs.org/~jciccone/uname_hack.tar.bz2 >>> for older 2.6 kernels. >>> _______________________________________________ >>> Clfs-support mailing list >>> Clfs-support@lists.cross-lfs.org >>> http://lists.cross-lfs.org/cgi-bin/mailman/listinfo/clfs-support >>> >>> >>> >>> >> _______________________________________________ >> Clfs-support mailing list >> Clfs-support@lists.cross-lfs.org >> http://lists.cross-lfs.org/cgi-bin/mailman/listinfo/clfs-support >> >> >> > Sorry about that, > > Add #include <linux/string.h> to the includes up at the top and change > utsname()->machine[1] = UNAME_DUMB_STEPPING; > -to- > memcpy(utsname()->machine, "i486", 4); > > The hack was originally designed for i?86. as i'm sure you can tell by > the change it made. > _______________________________________________ > Clfs-support mailing list > Clfs-support@lists.cross-lfs.org > http://lists.cross-lfs.org/cgi-bin/mailman/listinfo/clfs-support > > _______________________________________________ Clfs-support mailing list Clfs-support@lists.cross-lfs.org http://lists.cross-lfs.org/cgi-bin/mailman/listinfo/clfs-support