[RESEND PATCH 0/2][SRU][DISCO][EOAN] seccomp: fix selftests compilation

Christian Brauner christian.brauner at ubuntu.com
Mon Oct 28 17:25:15 UTC 2019


On Mon, Oct 28, 2019 at 05:40:04PM +0100, Kleber Souza wrote:
> On 22.10.19 11:36, Christian Brauner wrote:
> > Hey everyone,
> > 
> > BugLink: https://bugs.launchpad.net/bugs/1849281
> > 
> > This contains two small fixes for the seccomp selftests so they can be
> > compiled on older kernels.
> > The first patch is upstream, the second one is in Kees tree targeted for
> > inclusion in v5.5. The latter one needs to be backported because we
> > backported the SECCOMP_USER_NOTIF_FLAG_CONTINUE feature
> > (cf. https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1847744)
> > 
> > Thanks!
> > Christian
> > 
> > Christian Brauner (1):
> >   UBUNTU: SAUCE: seccomp: fix SECCOMP_USER_NOTIF_FLAG_CONTINUE test
> > 
> > Tycho Andersen (1):
> >   selftests/seccomp: fix build on older kernels
> > 
> >  tools/testing/selftests/seccomp/seccomp_bpf.c | 13 +++++++++----
> >  1 file changed, 9 insertions(+), 4 deletions(-)
> > 
> 
> Hi Christian,
> 
> When compiling your seccomp testcase I get the following error
> when doing it in a disco VM (with or without these follow up
> fixes):
> 
> make[1]: Entering directory '/home/ubuntu/disco/linux/tools/testing/selftests/seccomp'
> gcc -Wl,-no-as-needed -Wall  seccomp_bpf.c -lpthread -o seccomp_bpf
> seccomp_bpf.c: In function ‘user_notification_continue’:
> seccomp_bpf.c:3074:26: warning: overflow in conversion from ‘long int’ to ‘__s32’ {aka ‘int’} changes value from ‘116983961184613’ to ‘1936943461’ [-Woverflow]
>  #define USER_NOTIF_MAGIC 116983961184613L

Ah yeah, I sent a fix for that upstream too. Yes, that's a problem
that's been around since 5.0. The fix for it is in Kees tree as
223e660bc763 ("seccomp: avoid overflow in implicit constant conversion")
https://git.kernel.org/pub/scm/linux/kernel/git/kees/linux.git/commit/?h=for-next/seccomp&id=223e660bc7638d126a0e4fbace4f33f2895788c4

Sorry, that would need to be backported as well. :(

Thanks!
Christian



More information about the kernel-team mailing list