Revision 1978f30a87732d4d9072a20abeded9fe17884f1b authored by Dongli Zhang on 27 March 2019, 10:36:35 UTC, committed by Jens Axboe on 10 April 2019, 14:18:25 UTC
When tag_set->nr_maps is 1, the block layer limits the number of hw queues by nr_cpu_ids. No matter how many hw queues are used by virtio-scsi, as it has (tag_set->nr_maps == 1), it can use at most nr_cpu_ids hw queues. In addition, specifically for pci scenario, when the 'num_queues' specified by qemu is more than maxcpus, virtio-scsi would not be able to allocate more than maxcpus vectors in order to have a vector for each queue. As a result, it falls back into MSI-X with one vector for config and one shared for queues. Considering above reasons, this patch limits the number of hw queues used by virtio-scsi by nr_cpu_ids. Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
1 parent bf348f9
File | Mode | Size |
---|---|---|
Kconfig | -rw-r--r-- | 625 bytes |
Makefile | -rw-r--r-- | 248 bytes |
balloc.c | -rw-r--r-- | 18.1 KB |
dir.c | -rw-r--r-- | 5.2 KB |
directory.c | -rw-r--r-- | 5.8 KB |
ecma_167.h | -rw-r--r-- | 21.3 KB |
file.c | -rw-r--r-- | 7.3 KB |
ialloc.c | -rw-r--r-- | 3.5 KB |
inode.c | -rw-r--r-- | 65.1 KB |
lowlevel.c | -rw-r--r-- | 1.6 KB |
misc.c | -rw-r--r-- | 7.7 KB |
namei.c | -rw-r--r-- | 34.3 KB |
osta_udf.h | -rw-r--r-- | 7.9 KB |
partition.c | -rw-r--r-- | 8.7 KB |
super.c | -rw-r--r-- | 65.6 KB |
symlink.c | -rw-r--r-- | 4.2 KB |
truncate.c | -rw-r--r-- | 7.9 KB |
udf_i.h | -rw-r--r-- | 1.7 KB |
udf_sb.h | -rw-r--r-- | 4.1 KB |
udfdecl.h | -rw-r--r-- | 8.8 KB |
udfend.h | -rw-r--r-- | 1.6 KB |
udftime.c | -rw-r--r-- | 3.2 KB |
unicode.c | -rw-r--r-- | 9.2 KB |
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...