On Tue, Feb 03, 2015 at 06:29:54AM +0000, Nicholas A. Bellinger wrote: > From: Nicholas Bellinger <nab@xxxxxxxxxxxxxxx> > > Hi MST, Paolo, Al & Co, > > Here is -v3 for adding vhost/scsi ANY_LAYOUT + VERSION_1 host feature > bit support. > > It adds a new vhost_virtqueue ->handle_kick() callback to determine the > start of protection and data payloads iovecs past starting virtio-scsi > request and response headers, based upon data_direction using iov_iter > primitives. > > It assumes request/CDB and response/sense_buffer headers may span more > than a single iovec using mm/iov_iter.c logic. > > It also allows virtio-scsi headers + T10_PI + Data SGL payloads to span > the same iovec when pinning user-space memory via get_user_pages_fast() > code. (Not tested yet) > > Based upon Al & HCH's feedback, the patch series has been converted to > use copy_*_iter() for virtio-scsi header copy. Also, patch #4 has been > updated to use iov_iter_npages() for sgl_count, and patch #5 updated to > use iov_iter_advance() for calculating prot_bytes offset to the start > of data_iter. > > v3 changelog: > - Convert memcpy_toiovecend -> copy_to_iter usage > - Update vhost_scsi_mapal + friends to use iov_iter > - Move iov_iter sanity checks into vhost_scsi_calc_sgls > - Convert vhost_scsi_calc_sgls() to iov_iter_npages() > - Convert to vhost_scsi_handle_vqal to copy_from_iter usage > - Update vhost_scsi_handle_vqal comments for iov_iter usage > - Convert prot_bytes offset to use iov_iter_advance > - Drop max_niov usage in vhost_scsi_handle_vqal > - Drop vhost_skip_iovec_bytes in favour of iov_iter > > Note the one part that has been left unchanged is vhost_scsi_map_to_sgl() > into get_user_pages_fast(), for which existing iov_iter_get_pages() code > will need to allow for a callback to perform the associated scatterlists > setup from **pages for protection + data payloads. I'm not sure it has to be a callback: maybe just add struct scatterlist *sg parameter and use it if set. > It's functioning against v3.19-rc1 virtio-scsi LLD in T10_PI mode using > TYPE-1 DIF with ANY_LAYOUT -> VERSION_1 guest feature bits enabled, using > the layout following existing convention with protection/data SGL payloads > residing within seperate iovecs. > > Also included in patch #9 is an over-due change to rename code in scsi.c > to line up with modern vhost_scsi naming convention. > > Please review. > > Thank you, > > Nicholas Bellinger (9): > vhost/scsi: Convert completion path to use copy_to_iser > vhost/scsi: Fix incorrect early vhost_scsi_handle_vq failures > vhost/scsi: Change vhost_scsi_map_to_sgl to accept iov ptr + len > vhost/scsi: Add ANY_LAYOUT iov -> sgl mapping prerequisites > vhost/scsi: Add ANY_LAYOUT vhost_virtqueue callback > vhost/scsi: Set VIRTIO_F_ANY_LAYOUT + VIRTIO_F_VERSION_1 feature bits > vhost/scsi: Drop legacy pre virtio v1.0 !ANY_LAYOUT logic > vhost/scsi: Drop left-over scsi_tcq.h include > vhost/scsi: Global tcm_vhost -> vhost_scsi rename > > drivers/vhost/scsi.c | 1073 ++++++++++++++++++++++++++------------------------ > 1 file changed, 549 insertions(+), 524 deletions(-) > > -- > 1.9.1 -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
