vfio/pci: Restore zero affected bus reset devices warning

Yi notes relative to commit f6944d4a0b ("vfio/pci: Collect hot-reset
devices to local buffer") that we previously tested the resulting
device count with a WARN_ON, which was removed when we switched to
the in-loop user copy in commit b56b7aabcf ("vfio/pci: Copy hot-reset
device info to userspace in the devices loop").  Finding no devices in
the bus/slot would be an unexpected condition, so let's restore the
warning and trigger a -ERANGE error here as success with no devices
would be an unexpected result to userspace as well.

Suggested-by: Yi Liu <yi.l.liu@intel.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Link: https://lore.kernel.org/r/20240516174831.2257970-1-alex.williamson@redhat.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
This commit is contained in:
Alex Williamson 2024-05-16 11:48:30 -06:00
parent dda057ad8c
commit cbb325e77f

View File

@ -1281,6 +1281,9 @@ static int vfio_pci_ioctl_get_pci_hot_reset_info(
if (ret)
return ret;
if (WARN_ON(!count)) /* Should always be at least one */
return -ERANGE;
if (count > (hdr.argsz - sizeof(hdr)) / sizeof(*devices)) {
hdr.count = count;
ret = -ENOSPC;