On 12/12/19 4:19 AM, Michal Privoznik wrote:
On 12/10/19 9:50 PM, Cole Robinson wrote:
On 12/2/19 9:26 AM, Michal Privoznik wrote:
Because this is a HMP we're dealing with, there is nothing like class of reply message, so we have to do some string comparison to guess if the command fails. Well, with NVMe disks whole new class of errors comes to play because qemu needs to initialize IOMMU and VFIO for them. You can see all the messages it may produce in qemu_vfio_init_pci().
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> --- src/qemu/qemu_monitor_text.c | 7 +++++++ 1 file changed, 7 insertions(+)
diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 9054682d60..6948a5bf90 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -75,6 +75,13 @@ int qemuMonitorTextAddDrive(qemuMonitorPtr mon, goto cleanup; } + if (strstr(reply, "IOMMU") || + strstr(reply, "VFIO")) { + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + reply); + goto cleanup; + } + ret = 0; cleanup:
For the code:
Reviewed-by: Cole Robinson <crobinso@redhat.com>
Does the blockdev infrastructure have magic that turns the Props into a drive string? I don't see any -drive examples in the test output
Blockdev usues -device + -blockdev pair and obsoletes -drive. For instance:
libvirt.git $ grep "\-drive" $(git grep -l "\-blockdev" -- tests/qemuxml2argvdata/)
Neither of -blockdev files has -drive.
Right. I think my confusion is that this is a pre-blockdev/-drive code path, but I didn't see any explicit -drive handling in the series, just -blockdev. So either I 1) missed the -drive changes but in that case there should probably be test output demoing -drive output here, or 2) this feature is -blockdev only in which case this is dead code. Or some other case that I'm missing :) - Cole