Bug#885852: [sparc64] klibc-utils (2.0.4-10) regression, sigserv with fstype
Hello! Could we have the current version of the patch applied "as is" even though it does not meet the quality standards yet? Currently, klibc-utils is completely b0rked on sparc64 and anyone dist-upgrading their sparc64 machine will immediately break the machine upon next reboot. Adrian -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer - glaub...@debian.org `. `' Freie Universitaet Berlin - glaub...@physik.fu-berlin.de `-GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
Bug#885852: [sparc64] klibc-utils (2.0.4-10) regression, sigserv with fstype
On 15 Jul 2018, at 19:50, Ben Hutchings wrote: > > Control: tag -1 moreinfo > > On Mon, 1 Jan 2018 22:59:59 + James Clarke wrote: > [...] >> Please consider applying the patch forwarded upstream (linked in an >> earlier control message) soon; this bug means that if the current >> initramfs is updated, it will no longer boot, as run-init will segfault >> in klibc. Given sparc64 is not a release architecture I can't make this >> bug RC, otherwise I'd probably go for critical. >> >> (To be clear, the issue is in 2.0.4-10 simply because that is the first >> upload to happen since sparc64 has had PIE enabled by default in GCC) > > How exactly did you test this patch? It looks like it will cause > pipe() to crash on success. You put the first instruction of the PIC > prologue into a branch delay slot, which overwrites the register > holding the pointer used to store the system call return values. I guess nothing I ran used pipe... but you're right, I was trying to be clever with the branch delay slots (taking into account that some them already clobbered %g4 for a position-dependent errno address calculation) and hadn't noticed that pipe was using %g4 as a "caller"-saved register (it's clobbered by userspace calls...). I shall put the nops back in and somehow give it a proper test. James
Bug#885852: [sparc64] klibc-utils (2.0.4-10) regression, sigserv with fstype
Control: tag -1 moreinfo On Mon, 1 Jan 2018 22:59:59 + James Clarke wrote: [...] > Please consider applying the patch forwarded upstream (linked in an > earlier control message) soon; this bug means that if the current > initramfs is updated, it will no longer boot, as run-init will segfault > in klibc. Given sparc64 is not a release architecture I can't make this > bug RC, otherwise I'd probably go for critical. > > (To be clear, the issue is in 2.0.4-10 simply because that is the first > upload to happen since sparc64 has had PIE enabled by default in GCC) How exactly did you test this patch? It looks like it will cause pipe() to crash on success. You put the first instruction of the PIC prologue into a branch delay slot, which overwrites the register holding the pointer used to store the system call return values. Ben. -- Ben Hutchings One of the nice things about standards is that there are so many of them. signature.asc Description: This is a digitally signed message part
Bug#885852: [sparc64] klibc-utils (2.0.4-10) regression, sigserv with fstype
Control: severity -1 important On Sat, Dec 30, 2017 at 03:48:07PM +0300, Anatoly Pugachev wrote: > Package: klibc-utils > Version: 2.0.4-10 > Severity: normal > > Dear Maintainer, > > Upgrading klibc-utils from 2.0.4-9 to 2.0.4-10 started to produce sigserv in > fstype > >* What exactly did you do (or not do) that was effective (or > ineffective)? > > using latest version 2.0.4-10 : > > $ dpkg -l klibc-utils > ||/ Name VersionArchitecture >Description > +++-==-==-==-= > ii klibc-utils2.0.4-10 sparc64 >small utilities built with klibc for early boot > > $ /usr/lib/klibc/bin/fstype > Segmentation fault (core dumped) > > $ sudo /usr/lib/klibc/bin/fstype /dev/vdiska2 > Segmentation fault > > I tried with upstream klibc.git repo, but getting sigserv as well, and since > klibc.git does not have changed files almost a year now, not sure gdb > backtrace > could be relevant, please see > http://www.zytor.com/pipermail/klibc/2017-December/003965.html > > >* What outcome did you expect instead? > > using older package version of 2.0.4-9 : > > # dpkg -i *.deb > dpkg: warning: downgrading klibc-utils from 2.0.4-10 to 2.0.4-9 > (Reading database ... 68475 files and directories currently installed.) > Preparing to unpack klibc-utils_2.0.4-9_sparc64.deb ... > Unpacking klibc-utils (2.0.4-9) over (2.0.4-10) ... > dpkg: warning: downgrading libklibc from 2.0.4-10 to 2.0.4-9 > Preparing to unpack libklibc_2.0.4-9_sparc64.deb ... > Unpacking libklibc (2.0.4-9) over (2.0.4-10) ... > Setting up libklibc (2.0.4-9) ... > Setting up klibc-utils (2.0.4-9) ... > root@ttip:~/1# exit > > mator@ttip:~/linux-2.6$ dpkg -L klibc-utils | grep fstype > /usr/lib/klibc/bin/fstype > > mator@ttip:~/linux-2.6$ /usr/lib/klibc/bin/fstype > stdin: Illegal seek > > mator@ttip:~$ dpkg -l klibc-utils > ||/ Name VersionArchitecture >Description > +++-==-==-==-= > ii klibc-utils2.0.4-9sparc64 >small utilities built with klibc for early boot > > mator@ttip:~$ sudo /usr/lib/klibc/bin/fstype /dev/vdiska2 > FSTYPE=ext4 > FSSIZE=15002910720 Please consider applying the patch forwarded upstream (linked in an earlier control message) soon; this bug means that if the current initramfs is updated, it will no longer boot, as run-init will segfault in klibc. Given sparc64 is not a release architecture I can't make this bug RC, otherwise I'd probably go for critical. (To be clear, the issue is in 2.0.4-10 simply because that is the first upload to happen since sparc64 has had PIE enabled by default in GCC) Regards, James
Bug#885852: [sparc64] klibc-utils (2.0.4-10) regression, sigserv with fstype
Package: klibc-utils Version: 2.0.4-10 Severity: normal Dear Maintainer, Upgrading klibc-utils from 2.0.4-9 to 2.0.4-10 started to produce sigserv in fstype * What exactly did you do (or not do) that was effective (or ineffective)? using latest version 2.0.4-10 : $ dpkg -l klibc-utils ||/ Name VersionArchitecture Description +++-==-==-==-= ii klibc-utils2.0.4-10 sparc64 small utilities built with klibc for early boot $ /usr/lib/klibc/bin/fstype Segmentation fault (core dumped) $ sudo /usr/lib/klibc/bin/fstype /dev/vdiska2 Segmentation fault I tried with upstream klibc.git repo, but getting sigserv as well, and since klibc.git does not have changed files almost a year now, not sure gdb backtrace could be relevant, please see http://www.zytor.com/pipermail/klibc/2017-December/003965.html * What outcome did you expect instead? using older package version of 2.0.4-9 : # dpkg -i *.deb dpkg: warning: downgrading klibc-utils from 2.0.4-10 to 2.0.4-9 (Reading database ... 68475 files and directories currently installed.) Preparing to unpack klibc-utils_2.0.4-9_sparc64.deb ... Unpacking klibc-utils (2.0.4-9) over (2.0.4-10) ... dpkg: warning: downgrading libklibc from 2.0.4-10 to 2.0.4-9 Preparing to unpack libklibc_2.0.4-9_sparc64.deb ... Unpacking libklibc (2.0.4-9) over (2.0.4-10) ... Setting up libklibc (2.0.4-9) ... Setting up klibc-utils (2.0.4-9) ... root@ttip:~/1# exit mator@ttip:~/linux-2.6$ dpkg -L klibc-utils | grep fstype /usr/lib/klibc/bin/fstype mator@ttip:~/linux-2.6$ /usr/lib/klibc/bin/fstype stdin: Illegal seek mator@ttip:~$ dpkg -l klibc-utils ||/ Name VersionArchitecture Description +++-==-==-==-= ii klibc-utils2.0.4-9sparc64 small utilities built with klibc for early boot mator@ttip:~$ sudo /usr/lib/klibc/bin/fstype /dev/vdiska2 FSTYPE=ext4 FSSIZE=15002910720 -- System Information: Debian Release: buster/sid APT prefers unstable APT policy: (990, 'unstable'), (500, 'unreleased'), (500, 'experimental') Architecture: sparc64 Kernel: Linux 4.15.0-rc5-5-gbeacbc68ac3e (SMP w/1 CPU core) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8), LANGUAGE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages klibc-utils depends on: ii libklibc 2.0.4-10 klibc-utils recommends no packages. klibc-utils suggests no packages. -- no debconf information
