[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#850447: marked as done (nmu: systemd_230-7~bpo8+2)

Your message dated Sat, 6 Apr 2019 08:03:12 +0200
with message-id <372a58c6-ca6d-0cf9-b5f3-c655586ace0f@debian.org>
and subject line Re: Bug#850447: systemd backport sections only 4K aligned, won't boot with arm64 64K kernel
has caused the Debian Bug report #850447,
regarding nmu: systemd_230-7~bpo8+2
to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact owner@bugs.debian.org
immediately.)


-- 
850447: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=850447
Debian Bug Tracking System
Contact owner@bugs.debian.org with problems
--- Begin Message --- 
Package: systemd
Version: 230-7~bpo8+2
Severity: normal

Hi,

This version of systemd won't boot on arm64 when the kernel is configured for
64K pages:
> Starting init: /etc/init exists but couldn't execute it (error -13)
> Starting init: /bin/sh exists but couldn't execute it (error -14)
> Kernel panic - not syncing: No working init found.  Try passing init= option
to kernel. See Linux Documentation/admin-guide/init.rst for guidance.
> CPU: 3 PID: 1 Comm: init Not tainted 4.10.0-rc2-00036-g5a4dd5f49931 #6712
> Hardware name: ARM Juno development board (r1) (DT)
> Call trace:
> [<ffff000008088f44>] dump_backtrace+0x0/0x25c
> [<ffff0000080891c0>] show_stack+0x20/0x28
> [<ffff0000084940b4>] dump_stack+0x94/0xb4
> [<ffff0000081af874>] panic+0x134/0x2a4
> [<ffff0000089ad834>] kernel_init+0xf4/0x104
> [<ffff000008083370>] ret_from_fork+0x10/0x20


Booting with init=/bin/bash and interrogating the linker:
> root@(none):/lib/systemd# /lib/ld-linux-aarch64.so.1 --list /sbin/init
> /sbin/init: error while loading shared libraries: /sbin/init: ELF load command
alignment not page-aligned

Comparing sytemd's LOAD sections with bash's:
> root@(none):/lib/systemd# readelf -a /sbin/init | grep -A 1 LOAD
>   LOAD           0x0000000000000000 0x0000000000000000 0x0000000000000000
>                  0x00000000000bc56c 0x00000000000bc56c  R E    1000
>   LOAD           0x00000000000bc7a8 0x00000000000bd7a8 0x00000000000bd7a8
>                  0x0000000000020a88 0x0000000000020b7d  RW     1000

> root@(none):/lib/systemd# readelf -a /bin/bash | grep -A 1 LOAD
>   LOAD           0x0000000000000000 0x0000000000400000 0x0000000000400000
>                  0x00000000000d49f4 0x00000000000d49f4  R E    10000
>   LOAD           0x00000000000d4db0 0x00000000004e4db0 0x00000000004e4db0
>                  0x0000000000008ae8 0x000000000000e728  RW     10000

(The key to these tables is:)
> Program Headers:
>   Type           Offset             VirtAddr           PhysAddr
>                  FileSiz            MemSiz              Flags  Align

The Align value for each of systemd's LOAD sections is 4K aligned, not 64K, so
the runtime linker can't load it when the kernel is built with a page-size
other than 4K.

This was reported by Basil Eljuse who was using a filesystem from Linaro.
Sanity check whether this should be reported to debian by poking around in
the original deb file:

> readlink sbin/init
> /lib/systemd/systemd

> wget
http://ftp.uk.debian.org/debian/pool/main/s/systemd/systemd_230-7~bpo8+2_arm64.deb
> ar x systemd_230-7~bpo8+2_arm64.deb
> tar xf data.tar.xz
> readelf -a lib/systemd/systemd | grep -A 1 LOAD
>   LOAD           0x0000000000000000 0x0000000000000000 0x0000000000000000
>                  0x00000000000bc56c 0x00000000000bc56c  R E    1000
>   LOAD           0x00000000000bc7a8 0x00000000000bd7a8 0x00000000000bd7a8
>                  0x0000000000020a88 0x0000000000020b7d  RW     1000

This is potentially a wider issue affecting anything else built with the same
linker that built this 'bpo' package.


Thanks,

James


-- Package-specific info:

-- System Information:
Debian Release: 8.6
  APT prefers vivid
  APT policy: (500, 'vivid'), (500, 'stable-updates'), (500, 'stable')
Architecture: arm64 (aarch64)

Kernel: Linux 4.10.0-rc2-00036-g5a4dd5f49931 (SMP w/6 CPU cores)
Locale: LANG=C.UTF-8, LC_CTYPE=C.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages systemd depends on:
ii  adduser         3.113+nmu3
ii  libacl1         2.2.52-2
ii  libapparmor1    2.10.95-4~bpo8+2
ii  libaudit1       1:2.4-1+b1
ii  libblkid1       2.27.1-1.linarojessie.1
ii  libc6           2.19-18+deb8u6
ii  libcap2         1:2.24-8
ii  libcap2-bin     1:2.24-8
ii  libcryptsetup4  2:1.6.6-5
ii  libgcrypt20     1.6.3-2+deb8u2
ii  libgpg-error0   1.17-3
ii  libidn11        1.29-1+deb8u2
ii  libkmod2        18-3
ii  liblzma5        5.1.1alpha+20120614-2+b3
ii  libmount1       2.27.1-1.linarojessie.1
ii  libpam0g        1.1.8-3.1+deb8u1+b1
ii  libseccomp2     2.2.3-3~bpo8+1
ii  libselinux1     2.3-2
ii  libsystemd0     230-7~bpo8+2
ii  mount           2.27.1-1.linarojessie.1
ii  util-linux      2.27.1-1.linarojessie.1

Versions of packages systemd recommends:
ii  dbus            1.8.20-0+deb8u1
ii  libpam-systemd  230-7~bpo8+2

Versions of packages systemd suggests:
ii  policykit-1        0.105-15~deb8u2
pn  systemd-container  <none>
pn  systemd-ui         <none>

Versions of packages systemd is related to:
ii  udev  230-7~bpo8+2

-- Configuration Files:
/etc/systemd/system.conf changed:
[Manager]
LogLevel=warning
LogTarget=journal


-- no debconf information

--- End Message ---
--- Begin Message --- 
Hi,

On Wed, 8 Feb 2017 21:34:59 +0000 Jonathan Wiltshire <jmw@debian.org> wrote:
> On Thu, Jan 12, 2017 at 06:36:55PM +0100, Michael Biebl wrote:
> > nmu systemd_230-7~bpo8+2 . arm64 . jessie-backports . -m "Rebuild with
> > fixed binutils"
> 
> Backports is outside the release team's jurisdiction; whilst wanna-build
> will probably let me monkey with it, it doesn't feel right. If there's a
> precedent I'm not aware of, I'm happy to be corrected.
> 
> I suggest contacting debian-backports@lists.debian.org in the first
> instance.

This is all moot now anyways, as jessie-backports is no more.

Paul

Attachment: signature.asc
Description: OpenPGP digital signature


--- End Message ---
Reply to: