[libvirt] [PATCH] tests: Add getuid() to virnetdevbandwidthmock

[Daniel P. Berrangé]

On Tue, Jul 09, 2019 at 02:45:54PM +0200, Andrea Bolognani wrote:
> On Tue, 2019-07-09 at 10:21 +0100, Daniel P. Berrangé wrote:
> > On Mon, Jul 08, 2019 at 06:03:10PM +0200, Andrea Bolognani wrote:
> > > When only geteuid() is mocked, the test crashes on Debian 10.
> > > 
> > >   Fatal: failed to reset uid: No such file or directory
> > > 
> > >   Program received signal SIGABRT, Aborted.
> > >   __GI_raise (sig=sig at entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
> > >   50      ../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
> > >   (gdb) t a a bt
> > > 
> > >   Thread 1 (Thread 0x7ffff3b3e080 (LWP 12003)):
> > >   #0  __GI_raise (sig=sig at entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
> > >   #1  0x00007ffff7798535 in __GI_abort () at abort.c:79
> > >   #2  0x00007ffff485ca20 in _gcry_logv (level=level at entry=40, fmt=fmt at entry=0x7ffff4929126 "failed to reset uid: %s\n", arg_ptr=arg_ptr at entry=0x7fffffffe4a0) at ../../src/misc.c:142
> > >   #3  0x00007ffff485cd61 in _gcry_log_fatal (fmt=fmt at entry=0x7ffff4929126 "failed to reset uid: %s\n") at ../../src/misc.c:218
> > >   #4  0x00007ffff48639d1 in lock_pool_pages (n=<optimized out>, p=<optimized out>) at ../../src/secmem.c:340
> > >   #5  _gcry_secmem_init_internal (n=<optimized out>) at ../../src/secmem.c:563
> > >   #6  0x00007ffff4863d78 in _gcry_secmem_init (n=4096) at ../../src/secmem.c:581
> > >   #7  0x00007ffff485e4e6 in _gcry_vcontrol (cmd=<optimized out>, arg_ptr=arg_ptr at entry=0x7fffffffe5e0) at ../../src/global.c:506
> > >   #8  0x00007ffff485a789 in gcry_control (cmd=cmd at entry=GCRYCTL_INIT_SECMEM) at ../../src/visibility.c:79
> > >   #9  0x00007ffff71af10f in ssh_crypto_init () at ./src/libgcrypt.c:621
> > >   #10 0x00007ffff7193796 in _ssh_init (constructor=constructor at entry=1) at ./src/init.c:79
> > >   #11 0x00007ffff71834de in libssh_constructor () at ./src/init.c:116
> > 
> > Ewww, so its crashing in a ELF library constructor for libssh, which
> > is in turn calling into libgcrypt.
> > 
> > Obviously nothing todo with the test cases we're actually running.
> > I guess we're confusing the code into thinking it has some wierd
> > privilege by having geteuid() return 0, while getuid() returns
> > the normal UID.
> > 
> > Mocking getuid() is ok, but I fear its just a targetting one
> > specific problem we happen to be hitting today. 
> > So I think we probably ought to make geteuid() delegate the
> > real getuid() function instead. Have a global flag we can set
> > & unset just before & after executing the real test code. That
> > way libraries will see correct UID info in general.
> 
> I tried making the mocking more selective (see attached patch) but
> that doesn't work because there are two times we need the mocking to
> be in place: when calling virNetDevBandwidthSet(), which errors out
> if it detects it's not running as root, and when the libssh is
> initialized, which happens automatically when the library is loaded.
> 
> The attached patch only covers the former, and extending it so that
> it also does the latter seems like it would be a complete mess; plus
> at that point there wouldn't be a lot of the test program still
> running without mocking, so I wonder what we'd gain by going through
> all that trouble.
> 
> Unless, of course, I've completely misunderstood what you were
> suggesting in the first place O:-)

NO, you got what I was thinking, but if it doesn't work, lets not
waste more time on it.

Go for your original quick fix:

  Reviewed-by: Daniel P. Berrangé <berrange at redhat.com>



Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|