This bug was fixed in the package systemd - 244-3ubuntu1
---
systemd (244-3ubuntu1) focal; urgency=medium
[ Balint Reczey ]
* Merge to Ubuntu from Debian unstable
* Refresh patches:
- Dropped changes:
* d/t/control: mark udev test skippable.
File:
** Merge proposal linked:
https://code.launchpad.net/~paelzer/ubuntu/+source/systemd/+git/systemd/+merge/376514
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Issue is in systemd, fix known (now).
I'll prep an MP to discuss, but given how much extra churn systemd updates
usually cause this most likely won't be uploaded "on its own"
** Changed in: libseccomp (Ubuntu)
Status: New => Invalid
** Changed in: systemd (Ubuntu)
Status: New =>
FYI: Discussions with both upstreams continue, I'm right now trying a
new spin with a different approach that does not use the _exact calls
and will get back to the upstreams once (if) that works.
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages,
@rbalint how would you feel about - for now - reverting debian/patches
/test-expect-mmap-to-fail-in-seccomp-test-on-s390-and-s390.patch and
instead also mark __i386__ as failing-to-protect?
That would allow the current combination of systemd in Focal to pass.
Once this issue is sorted out between
Now also systemd discussion and code started in
https://github.com/systemd/systemd/pull/14167
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Title:
hard to
Discussions with libseccomp upstream go well.
I might eventually need to go to upstream systemd and propose a fix, but before
doing so I need more clarification with libseccomp upstream and some more
testing.
Experimental branch for packaging experiments (d/p/patch file):
=>
** Changed in: libseccomp
Status: Unknown => New
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Title:
hard to reproduce issues in systemd autopkgtest
Now that I could properly describe the issue I have filed upstream bug
https://github.com/seccomp/libseccomp/issues/193
** Bug watch added: github.com/seccomp/libseccomp/issues #193
https://github.com/seccomp/libseccomp/issues/193
** Also affects: libseccomp via
I've simplified the test to a small case - running that I can reproduce the
error.
This should be enough to go to upstreams with it.
cat > test-seccomp-shmat.c << EOF
#include
#include
#include
#include
#include
/*
* Test issues with libseccomp 2.4.1 -> 2.4.2
* Derived from systemd
We know it fails at shmat, so we most likely can focus on that one.
i386 2.4.2
/* test_memory_deny_write_execute_shmat */
arch x86: SCMP_SYS(mmap) = 90
arch x86: SCMP_SYS(mmap2) = 192
arch x86: SCMP_SYS(shmget) = 395
arch x86: SCMP_SYS(shmat) = 397
arch x86: SCMP_SYS(shmdt) = 398
Operating on
I'll stay at 32bit intel for now as it is easier to get to for most
developers compared to s390x
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Title:
hard to
seccomp_memory_deny_write_execute [1] in src/shared/seccomp-util.c is
the one setting the seccomp rules called from test
test_memory_deny_write_execute_mmap
It fails at: test_memory_deny_write_execute_mmap
But later on we usually get debug info about used syscalls at:
** Tags added: update-excuse
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Title:
hard to reproduce issues in systemd autopkgtest against new libseccomp
2.4.2
Ok, confirmed the patch debian/patches/test-expect-mmap-to-fail-in-
seccomp-test-on-s390-and-s390.patch has to be taken out to make it work
on s390x.
And most likely the same needs to happen for x86-32bit - there we don't have a
patch.
But this most likely also needs to be switched to no more
Adding [1] might be worth in general - this is the patch for the arm
related discussion which seems applied upstream.
[1]:
https://github.com/systemd/systemd/commit/4df8fe8415eaf4abd5b93c3447452547c6ea9e5f
--
You received this bug notification because you are a member of Ubuntu
Touch seeded
I'm currently checking if I can build locally for quick turnaround times
of tests or if I need full LP builds every time ...
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
Uh there is something related as Ubuntu Delta:
From: Balint Reczey
Date: Tue, 22 Oct 2019 17:10:17 +0200
Subject: test: expect mmap to fail in seccomp test on s390 and s390x
(cherry picked from commit a81f7aad9a5ddeebbce002e2da36e1dd84f51b36)
---
src/test/test-seccomp.c | 2 +-
1 file changed,
in GDB with follow-fork-mode child I can check which call is actually
failing in the child:
It is this one:
p = mmap(NULL, page_size(), PROT_WRITE|PROT_EXEC, MAP_PRIVATE|MAP_ANONYMOUS,
-1,0);
assert_se(p == MAP_FAILED);
It expects a fail (due to seccomp block) but does not get that.
Take it
(gdb) bt
#0 __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
#1 0x03fffdd2b232 in __GI_abort () at abort.c:79
#2 0x03fffdb03a64 in log_assert_failed_realm () from
/lib/systemd/libsystemd-shared-243.so
#3 0x02aa746e in test_memory_deny_write_execute_mmap ()
x86 32bit is rather similar:
/* test_memory_deny_write_execute_mmap */
Operating on architecture: x86
Failed to add shmat() rule for architecture x86, skipping: Invalid argument
Assertion 'p == MAP_FAILED' failed at src/test/test-seccomp.c:493, function
test_memory_deny_write_execute_mmap().
Seems to be recreatable with
$ sudo /usr/lib/systemd/tests/test-seccomp
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
Title:
hard to reproduce issues in systemd
Isolated to the breaking test:
old: https://paste.ubuntu.com/p/n7BDf3Npwp/
bad: https://paste.ubuntu.com/p/5y5G4GrJYf/
--
You received this bug notification because you are a member of Ubuntu
Touch seeded packages, which is subscribed to systemd in Ubuntu.
https://bugs.launchpad.net/bugs/1853852
23 matches
Mail list logo