Messages in this thread |  | | From | (Eric W. Biederman) | | Date | Fri, 18 Nov 2011 20:53:12 -0800 | | Subject | Re: user namespace: make signal.c respect user namespaces (v5) |
| |
Serge Hallyn <serge.hallyn@canonical.com> writes:
> Quoting Eric W. Biederman (ebiederm@xmission.com): >> "Serge E. Hallyn" <serge.hallyn@canonical.com> writes: >> >> > Quoting Eric W. Biederman (ebiederm@xmission.com): >> >> There is a small bug here. You want to fixup q->info, not info. >> >> Otherwise you might try dereferencing one of the special signals and get >> >> a NULL pointer dereference. >> > >> > Thanks, Eric. Oddly I've not seen this happen in quite a bit of >> > testing with the kernel, but you certainly must be right. I sent >> > out a new patch to fix that. >> >> You clearly have a different test case than I do. > > I ran a good chunk of ltp... and it passed. I can't explain it.
I guess my test case as mostly dinking around and hitting ctrl-c because something wasn't behaving as I would like.
Still I a tad surprised that ltp doesn't seem to test that one.
>> I managed to trigger the oops within about 5 minutes of just fooling >> around. >> >> You want to say &q->info not q->info in your updated patch. > > Oh, yes, thanks. Sorry, I shouldn't have sent that one as I wasn't able > to compile and test until tonight.
No problem.
Eric
|  |