On 24.02.2016 19:23, Kevin Wolf wrote: > Am 24.02.2016 um 18:54 hat Max Reitz geschrieben: >> On 23.02.2016 18:16, Kevin Wolf wrote: >>> Now that we can use drive_add to create new nodes without a BB, we also >>> want to be able to delete such nodes again. >>> >>> Signed-off-by: Kevin Wolf <[email protected]> >>> --- >>> blockdev.c | 9 +++++++++ >>> 1 file changed, 9 insertions(+) >>> >>> diff --git a/blockdev.c b/blockdev.c >>> index 3f46bc1..b76b6cd 100644 >>> --- a/blockdev.c >>> +++ b/blockdev.c >>> @@ -2816,6 +2816,15 @@ void hmp_drive_del(Monitor *mon, const QDict *qdict) >>> AioContext *aio_context; >>> Error *local_err = NULL; >>> >>> + bs = bdrv_find_node(id); >>> + if (bs) { >>> + qmp_x_blockdev_del(false, NULL, true, id, &local_err); >>> + if (local_err) { >>> + error_report_err(local_err); >>> + } >>> + return; >>> + } >>> + >>> blk = blk_by_name(id); >>> if (!blk) { >>> error_report("Device '%s' not found", id); >>> >> >> It's a bit strange to require the user to specify the node name using >> "node-name" for drive_add, but the to use "id" in drive_del; especially >> because x-blockdev-del uses "node-name", too. > > Not sure I understand. For the user of drive_del that's simply a > positional parameter, so they use neither "id" nor "node-name". Am I > missing something?
No, it's just me being confused again by the way HMP works. I keep forgetting that the user doesn't specify the parameter names. So the R-b stands. :-) Max
signature.asc
Description: OpenPGP digital signature
