[PATCH] test/eal: remove PCI probing for vdev tests

[David Marchand]

On Mon, 23 Feb 2026 at 18:16, Marat Khalili <marat.khalili at huawei.com> wrote:
> > @@ -422,17 +422,17 @@ test_invalid_vdev_flag(void)
> >
> >       /* Test with invalid vdev option */
> >       const char *vdevinval[] = {prgname, prefix, no_huge, eal_debug_logs,
> > -                             bus_debug_logs, vdev, "eth_dummy"};
> > +                             bus_debug_logs, no_pci, vdev, "eth_dummy"};
> >
> >       /* Test with valid vdev option */
> >       const char *vdevval1[] = {prgname, prefix, no_huge, eal_debug_logs,
> > -                             bus_debug_logs, vdev, "net_ring0"};
> > +                             bus_debug_logs, no_pci, vdev, "net_ring0"};
>
> Could use this opportunity to replace tab before no_pci with space.

Oh yes, good catch.

>
> >
> >       const char *vdevval2[] = {prgname, prefix, no_huge, eal_debug_logs,
> > -                             bus_debug_logs, vdev, "net_ring0,args=test"};
> > +                             bus_debug_logs, no_pci, vdev, "net_ring0,args=test"};
> >
> >       const char *vdevval3[] = {prgname, prefix, no_huge, eal_debug_logs,
> > -                             bus_debug_logs, vdev, "net_ring0,nodeaction=r1:0:CREATE"};
> > +                             bus_debug_logs, no_pci, vdev, "net_ring0,nodeaction=r1:0:CREATE"};
> >
> >       if (launch_proc(vdevinval) == 0) {
> >               printf("Error (line %d) - process did run ok with invalid vdev parameter\n",
> > --
> > 2.53.0
>
> Acked-by: Marat Khalili <marat.khalili at huawei.com>
> Tested-by: Marat Khalili <marat.khalili at huawei.com>
>
> Definitely not a problem of this patch, but creating a separate define for each command-line argument like --no-huge does nothing to save us from repeating `prgname, prefix, eal_debug_logs, ...` bazillion times. If tokens are burning a hole in someone's pocket, perhaps instead we could move this whole command prefix into some kind of function or macro with parameters (possibly one launching subprocess as well).
>
> Another thought, should we have NOPCI_OK/NOPCI_SKIP similar to ASAN and NOHUGE?

It would be simpler to always disable pci and keep it enabled for
those tests that do require it.
I wonder what unit tests do require PCI in the fast-tests suite...
probably really few.


-- 
David Marchand