| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 7 Feb 2013 14:39:45 +0100 From: Sedat Dilek <sedat.dilek@...il.com> To: Tejun Heo <tj@...nel.org> Cc: Stephen Rothwell <sfr@...b.auug.org.au>, linux-next@...r.kernel.org, linux-kernel@...r.kernel.org, Al Viro <viro@...iv.linux.org.uk>, "Eric W. Biederman" <ebiederm@...ssion.com>, Andrew Morton <akpm@...ux-foundation.org> Subject: Re: linux-next: Tree for Feb 7 (fakeroot BROKEN due to SYSV IPC support?) On Thu, Feb 7, 2013 at 12:52 PM, Sedat Dilek <sedat.dilek@...il.com> wrote: > On Thu, Feb 7, 2013 at 12:14 PM, Eric W. Biederman > <ebiederm@...ssion.com> wrote: >> Sedat Dilek <sedat.dilek@...il.com> writes: >> >>> On Thu, Feb 7, 2013 at 11:20 AM, Sedat Dilek <sedat.dilek@...il.com> wrote: >>>> On Thu, Feb 7, 2013 at 10:38 AM, Sedat Dilek <sedat.dilek@...il.com> wrote: >>>>> On Thu, Feb 7, 2013 at 7:37 AM, Stephen Rothwell <sfr@...b.auug.org.au> wrote: >>>>>> Hi all, >>>>>> >>>>>> Changes since 20130206: >>>>>> >>>>>> Removed tree: kvmtool (still present via the tip tree) >>>>>> >>>>>> The block tree lost its build failure. >>>>>> >>>>>> The tip tree gained a conflict against the s390 tree. >>>>>> >>>>>> The kvm tree gained a conflict against Linus' tree. >>>>>> >>>>>> The tty tree lost its build failure. >>>>>> >>>>>> The arm-soc tree gained conflicts against the iommu tree. >>>>>> >>>>>> The signal tree gained a conflict against the s390 tree. >>>>>> >>>>>> The akpm tree gained a conflict against the kvm tree and lost its build >>>>>> failure. >>>>>> >>>>>> ---------------------------------------------------------------------------- >>>>>> >>>>> >>>>> My build-script uses fakeroot and does no more start: >>>>> >>>>> $ ./scripts/build_linux-next.sh >>>>> make options ...... CC=gcc-4.6 -j4 >>>>> KBUILD_BUILD_USER=sedat.dilek@...il.com >>>>> LOCALVERSION=-next20130207-2-iniza-small >>>>> dep-pkg options ... KDEB_PKGVERSION=3.8.0~rc6~next20130207-2~iniza+dileks1 >>>>> fakeroot, while creating message channels: Invalid argument >>>>> This may be due to a lack of SYSV IPC support. >>>>> fakeroot: error while starting the `faked' daemon. >>>>> kill: usage: kill [-s sigspec | -n signum | -sigspec] pid | jobspec >>>>> ... or kill -l [sigspec] >>>>> >>>>> >>>>> Any hints? >>>>> ( I could run strace... ) >>>>> >>>>> - Sedat - >>>> >>>> Attached strace outputs within yesterday's (GOOD) and today's (BAD) Linux-Next. >>>> >>>> - Sedat - >>> >>> [ CCing Al and Eric ] >>> >>> I compared quickly the diff between the -next versions and saw changes >>> coming from signal and userns trees. >>> ( Sorry for re-attaching the strace outputs. ) >> >> It has been about a week and a half since I have pushed anything into >> the userns tree, and I don't have anything ipc related in my tree. >> So the reason for the suspicion seems odd. >> >> The straces are useless because all they show is that fakeroot was >> forked, but there is not a trace of fakeroot itself. >> >> Given the timing my initial suspect would be the idr_preload patches >> from Tejun that were merged via Andrew's akpm tree. I didn't see >> anything in there that would kill ipc but that is likely the most recent >> touch to the ipc code. >> > > [ CCing Tejun ] > > Hmm, still stepping in the dark... > > OK, fakeroot does PRELOADing so your assumption makes sense. > > How can I trigger fakeroot calls? > Here I have... > /usr/bin/faked-sysv > /usr/bin/faked-tcp > > I think that is the main commit but I am not sure if I can revert the > idr_preload part, otherwise it gets complicated: > > commit e2802c2defba1e5c88d7d168eb5c66813c86f249 > "idr: implement idr_preload[_end]() and idr_alloc()" > > $ grep idr ../Linux-Next-v20130206-VS-v20130207.diff | grep ^'++Applying:' > ++Applying: block: fix ext_devt_idr handling > ++Applying: idr: fix a subtle bug in idr_get_next() > ++Applying: nfsd: idr_destroy() no longer needs idr_remove_all() > ++Applying: idr: cosmetic updates to struct / initializer definitions > ++Applying: idr: relocate idr_for_each_entry() and reorganize id[r|a]_get_new() > ++Applying: idr: remove _idr_rc_to_errno() hack > ++Applying: idr: refactor idr_get_new_above() > ++Applying: idr: implement idr_preload[_end]() and idr_alloc() > ++Applying: block: convert to idr_alloc() > ++Applying: block/loop: convert to idr_alloc() > ++Applying: atm/nicstar: convert to idr_alloc() > ++Applying: drbd: convert to idr_alloc() > ++Applying: dca: convert to idr_alloc() > ++Applying: dmaengine: convert to idr_alloc() > ++Applying: firewire: convert to idr_alloc() > ++Applying: gpio: convert to idr_alloc() > ++Applying: drm: convert to idr_alloc() > ++Applying: drm/exynos: convert to idr_alloc() > ++Applying: drm/i915: convert to idr_alloc() > ++Applying: drm/sis: convert to idr_alloc() > ++Applying: drm/via: convert to idr_alloc() > ++Applying: drm/vmwgfx: convert to idr_alloc() > ++Applying: i2c: convert to idr_alloc() > ++Applying: IB/core: convert to idr_alloc() > ++Applying: IB/amso1100: convert to idr_alloc() > ++Applying: IB/cxgb3: convert to idr_alloc() > ++Applying: IB/cxgb4: convert to idr_alloc() > ++Applying: IB/ehca: convert to idr_alloc() > ++Applying: IB/ipath: convert to idr_alloc() > ++Applying: IB/mlx4: convert to idr_alloc() > ++Applying: IB/ocrdma: convert to idr_alloc() > ++Applying: IB/qib: convert to idr_alloc() > ++Applying: dm: convert to idr_alloc() > ++Applying: memstick: convert to idr_alloc() > ++Applying: mfd: convert to idr_alloc() > ++Applying: misc/c2port: convert to idr_alloc() > ++Applying: misc/tifm_core: convert to idr_alloc() > ++Applying: mmc: convert to idr_alloc() > ++Applying: mtd: convert to idr_alloc() > ++Applying: macvtap: convert to idr_alloc() > ++Applying: ppp: convert to idr_alloc() > ++Applying: power: convert to idr_alloc() > ++Applying: pps: convert to idr_alloc() > ++Applying: remoteproc: convert to idr_alloc() > ++Applying: rpmsg: convert to idr_alloc() > ++Applying: scsi/bfa: convert to idr_alloc() > ++Applying: scsi: convert to idr_alloc() > ++Applying: target/iscsi: convert to idr_alloc() > ++Applying: scsi/lpfc: convert to idr_alloc() > ++Applying: thermal: convert to idr_alloc() > ++Applying: uio: convert to idr_alloc() > ++Applying: vfio: convert to idr_alloc() > ++Applying: dlm: convert to idr_alloc() > ++Applying: inotify: convert to idr_alloc() > ++Applying: ocfs2: convert to idr_alloc() > ++Applying: ipc: convert to idr_alloc() > ++Applying: cgroup: convert to idr_alloc() > ++Applying: events: convert to idr_alloc() > ++Applying: posix-timers: convert to idr_alloc() > ++Applying: net/9p: convert to idr_alloc() > ++Applying: mac80211: convert to idr_alloc() > ++Applying: sctp: convert to idr_alloc() > ++Applying: nfs4client: convert to idr_alloc() > > - Sedat - > > [1] http://git.kernel.org/?p=linux/kernel/git/next/linux-next.git;a=commitdiff;h=e2802c2defba1e5c88d7d168eb5c66813c86f249 > >> Eric Hi Tejun, Eric was right that your idr_preload/idr_alloc patches from mmotm-tree in Linux-Next (next-20130207) caused that fakeroot crazyness. I can't say if the whole series is BORKED or the IPC part. Changelog in [1] says "compile-tested-only" :-). commitdiff aebcb5a5f97e6e567bdd1b2651253de073afa572 "ipc: convert to idr_alloc()" Reverting the series let me build my kernels with fakeroot! Tejun, can you have a look at this? Thanks! Thank you Eric for the hints and sorry for my wrong suspicions. Please, have a look at the attachments. Regards, - Sedat - [1] http://git.kernel.org/?p=linux/kernel/git/next/linux-next.git;a=commitdiff;h=aebcb5a5f97e6e567bdd1b2651253de073afa572 View attachment "strace_3.8.0-rc6-next20130207-2-iniza-small.txt" of type "text/plain" (10108 bytes) Download attachment "3.8.0-rc6-next20130207-2-iniza-small.patch" of type "application/octet-stream" (100947 bytes)
Powered by blists - more mailing lists