Created attachment 322918 [details] FEATURES="--sandbox --usersandbox" emerge --info sys-fs/zfs sys-fs/zfs-0.6.0-rc10 and sys-fs/zfs-9999 fail to emerge when they attempt to make the contents of /usr/src/linux under kernels 3.4.9 and 3.5.3. It appears as though the make is aborting when it encounters a new kernel option, that should already have been configured. Also, sandbox is not happy about the kernel portion of the sys-fs/zfs ebuild. make[3]: Leaving directory `/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999/cmd' make[2]: Leaving directory `/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999/cmd' Making all in /usr/src/linux make[2]: Entering directory `/usr/src/linux-3.5.3-gentoo' scripts/kconfig/conf --silentoldconfig Kconfig * * Restart config... * * * RCU Subsystem * RCU Implementation > 1. Tree-based hierarchical RCU (TREE_RCU) choice[1]: 1 Tree-based hierarchical RCU fanout value (RCU_FANOUT) [32] (NEW) aborted! Console input/output is redirected. Run 'make oldconfig' to update configuration. make[4]: *** [silentoldconfig] Error 1 make[3]: *** [silentoldconfig] Error 2 make[3]: Nothing to be done for `all'. make[2]: *** No rule to make target `include/config/auto.conf', needed by `include/config/kernel.release'. Stop. make[2]: *** Waiting for unfinished jobs.... make[3]: Nothing to be done for `relocs'. make[2]: Leaving directory `/usr/src/linux-3.5.3-gentoo' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999' make: *** [all] Error 2 * ERROR: sys-fs/zfs-9999 failed (compile phase): * emake failed * * If you need support, post the output of `emerge --info '=sys-fs/zfs-9999'`, * the complete build log and the output of `emerge -pqv '=sys-fs/zfs-9999'`. * The complete build log is located at '/var/tmp/portage/sys-fs/zfs-9999/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/sys-fs/zfs-9999/temp/environment'. * Working directory: '/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999' * S: '/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999' >>> Failed to emerge sys-fs/zfs-9999, Log file: >>> '/var/tmp/portage/sys-fs/zfs-9999/temp/build.log' * Messages for package sys-fs/zfs-9999: * This profile is merely a convenience for people who require a more * minimal profile, yet are unable to use hardened due to restrictions in * the software being used on the server. If you seek a secure * production server profile, please check the Hardened project * (http://hardened.gentoo.org) * ERROR: sys-fs/zfs-9999 failed (compile phase): * emake failed * * If you need support, post the output of `emerge --info '=sys-fs/zfs-9999'`, * the complete build log and the output of `emerge -pqv '=sys-fs/zfs-9999'`. * The complete build log is located at '/var/tmp/portage/sys-fs/zfs-9999/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/sys-fs/zfs-9999/temp/environment'. * Working directory: '/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999' * S: '/var/tmp/portage/sys-fs/zfs-9999/work/zfs-9999'
Please attach the entire build log to this bug report.
Created attachment 322926 [details] build.log for FEATURES="-sandbox -usersandbox" emerge sys-fs/zfs
Wouldn't this happen only when the source trees in question haven't been previously configured?
One would assume yes, however, the kernel tree was configured, built and running before getting this error. An moving from 3.4.9 to 3.5.3 was also done, with a make oldconfig run and all options resolved, specifically the one that seems to be causing the error.
Issue resolved. KERNEL_DIR was set to /usr/src/linux in make.conf. sys-fs/zfs seems to use this as an override for something related to the kernel. It would be good to note somewhere that this behavior is happening, preferably right where it would start deviating from normal behavior by attempting to write to/build at the location KERNEL_DIR refers to.
(In reply to comment #5) > Issue resolved. > > KERNEL_DIR was set to /usr/src/linux in make.conf. sys-fs/zfs seems to use > this as an override for something related to the kernel. > > It would be good to note somewhere that this behavior is happening, > preferably right where it would start deviating from normal behavior by > attempting to write to/build at the location KERNEL_DIR refers to. This is an upstream bug. I will leave it open, but I am too busy to work on it right now. I have informed upstream about this, but I expect that this will be a low priority for them too.
(In reply to comment #6) > This is an upstream bug. I will leave it open, but I am too busy to work on > it right now. I have informed upstream about this, but I expect that this > will be a low priority for them too. I did not understand autotools at the time this bug was filed. That has since changed and I have sent a patch upstream to address this issue: https://github.com/zfsonlinux/zfs/pull/1151 I will backport it to sys-fs/zfs after upstream accepts it. I regret not fixing this sooner.
This bug has been fixed since sys-fs/zfs-kmod-0.6.0_rc13 entered the tree on December 23. I did not backport it because it was rolled into the next release by upstream shortly after being accepted.
