Skip to content

virtio_add_buf replacement.

Rusty Russell lkml virtualization 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v1 Latest6 patches 8 repliesIn review

Patches in v1 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/6virtio_ring: virtqueue_add_sgs, to add multiple sgs.drivers/virtio/virtio_ring.c include/linux/virtio.h tools/virtio/linux/scatterlist.h +1 more+132 -420 replies
1/6virtio_ring: virtqueue_add_sgs, to add multiple sgs.drivers/virtio/virtio_ring.c include/linux/virtio.h tools/virtio/linux/scatterlist.h +1 more+132 -420 replies
2/6virtio_ring: don't count elements twice for add_buf path.drivers/virtio/virtio_ring.c+80 -670 replies
3/6virtio_ring: inline internal vring functions more aggressively.drivers/virtio/virtio_ring.c+56 -352 replies
4/6virtio_ring: virtqueue_add_outbuf / virtqueue_add_inbuf.drivers/virtio/virtio_ring.c include/linux/virtio.h+54 -03 replies
5/6tools/virtio: make vringh_test use inbuf/outbuf.tools/virtio/linux/virtio.h tools/virtio/vringh_test.c+14 -40 replies
6/6virtio_scsi: use virtqueue_add_inbuf() for virtscsi_kick_event.drivers/scsi/virtio_scsi.c+2 -20 replies

Reviews trailers per patch

3 of 7 patches have a Reviewed-by or Acked-by.

PatchWanlong GaoAsias He
1/6R-b R-b R-b R-b
1/6R-b R-b R-b R-b
2/6R-b
3/6
4/6
5/6
6/6

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes.

Read the v1 thread

Cover letter [PATCH 0/6] virtio_add_buf replacement.

OK, so I've spent a few days benchmarking. Turns out 80% of virtio_add_buf cases are uni-directional (including the always-performance-sensitive networking code), and that gets no performance penalty (though tests with real networking would be appreciated!).

I'm not reposting all the "convert driver to virtio_add_outbuf()"
patches: just the scsi one which I didn't have before. I won't actually remove virtio_add_buf() until the *following* merge window, just to be sure.

One annoying thing about benchmarking is that in some cases, speeding up one side can make the whole thing slower, due to more wakeups. Device-side polling techniques might be required in future to get more performance.

Cheers,
Rusty.