This happens for me again and again, so I report this here... At nearly the last steps in emerging glibc-2.3.3.20040420-r1, I has build and install all files to /var/tmp/portage/.../image. Then it shows me the following message.. * Your version of: * /lib/modules/2.6.7-0/build/include/linux/sysctl.h * is broken (from a user space perspective). Please apply * the following patch: * ******************************************************* cat: /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch: No such file or directory * ******************************************************* * To fix, just do this: * cd /lib/modules/2.6.7-0/build/include/linux/ * patch -p3 < /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch !!! ERROR: glibc-2.3.3.20040420-r1/glibc-2.3.3.20040420-r1 failed. !!! Function pkg_setup, Line 214, Exitcode 0 !!! Broken linux/sysctl.h header included in kernel sources! I've tried 'rm -rf /var/tmp/portage{,-pkg}/glibc-2.*' And redo it, still failed it.. I after server try, I manually patch the stated kernel headers, getting the patch from /usr/portage/.../files/ but still doesn't help. How can I solve it? BTW, I know /var/tmp/portage/ is used to store files, during the build process. but what is the purpose of /var/tmp/portage-pkg? Reproducible: Always Steps to Reproduce: 1.emerge -Dv system Expected Results: glibc should be successfully merged If needed, I can attach the full log of the glibc build.
I have take a look on the glibc ebuild, the stated problems should only appear on pkg_setup stages.. I'm getting confused.. So, I attached the log for reference.. http://www.nixstyle.net/random/system-2004-09-03.elog When emerge, I do nohup emerge -Dv system 2>&1 | tee /tmp/system-`date +%F`.elog &
I have the same problem. I used stage1 bootstrap script and got it. Now using stage3 and "emerge -u system", same problem. These lines would be appended to sysctl.h with applied patch. +#ifndef __user +#define __user +#endif With stage3 the patch was not applied. With stage1+nptl the patch was applied. In both cases the same error message ?!? Of course /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch DOES NOT exist. ./sbin/ ./sbin/ldconfig ./sbin/sln >>> Done. >>> extracting info * Checking kernel headers for broken sysctl.h ... yes * Your version of: * /lib/modules/2.6.8-gentoo-r3/build/include/linux/sysctl.h * is broken (from a user space perspective). Please apply * the following patch: * ******************************************************* cat: /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch: No such file or directory * ******************************************************* * To fix, just do this: * cd /lib/modules/2.6.8-gentoo-r3/build/include/linux/ * patch -p3 < /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch !!! ERROR: glibc-2.3.3.20040420-r1/glibc-2.3.3.20040420-r1 failed. !!! Function pkg_setup, Line 214, Exitcode 0 !!! Broken linux/sysctl.h header included in kernel sources! !!! Error running pkg_setup
I think, the problem is no the ebuild of glibc, but may be emerge. I can use manually install the glibc by $ ebuild /usr/portage/sys-libs/glibc/glibc-2.3.3.20040420-r1 install $ ebuild /usr/portage/sys-libs/glibc/glibc-2.3.3.20040420-r1 qmerge $ ebuild /usr/portage/sys-libs/glibc/glibc-2.3.3.20040420-r1 package $ emerge -c So, it should be a viable solution, at this moment.. In this problem, the most critical thing is, why pkg_setup is ran again? In my case, pkg_setup has ran successfully at the begining of the emerge. But emerge fail during the process of 'install', notice that, the final error throw by emerge is: !!! ERROR: glibc-2.3.3.20040420-r1/glibc-2.3.3.20040420-r1 failed. !!! Function pkg_setup, Line 214, Exitcode 0 !!! Broken linux/sysctl.h header included in kernel sources! Another strange stuff is, it say: cat: /var/tmp/portage-pkg/glibc-2.3.3.20040420-r1/inf/files/fix-sysctl_h.patch: According to the ebuild (line 207), it should be: cat: /usr/portage/sys-libs/glibc/files/fix-sysctl_h.patch
The fix from me on bug #61856 may or may not solve this problem.
closing as I believe this isn't an issue anymore. If I am in error, please reopen .
