search for: blk_ring_lock

In blkback driver, after I/O requests are submitted to Dom-0 block I/O subsystem, blkback goes to ''sleep'' effectively without letting blkfront know about it (req_event isn''t set appropriately). Hence blkfront doesn''t notify blkback when it submits a new I/O thus delaying the ''dispatch'' of the new I/O to Dom-0 block I/O subsystem. The new I/O is

...+ case BLKIF_PROTOCOL_X86_64: + memcpy(RING_GET_RESPONSE(&blk_rings->x86_64, blk_rings->x86_64.rsp_prod_pvt), + resp, sizeof(*resp)); + break; + default: + BUG(); + } - resp.id = id; - resp.operation = op; - resp.status = st; +} - spin_lock_irqsave(&blkif->blk_ring_lock, flags); - /* Place on the response ring for the relevant domain. */ +void copy_response_v2(struct xen_blkif *blkif, struct blkif_response *resp) +{ + union blkif_back_rings_v2 *blk_rings = &blkif->blk_rings_v2; + switch (blkif->blk_protocol) { case BLKIF_PROTOCOL_NATIVE: memcpy(...

...blkif *blkif, u64 id, default: BUG(); } - blk_rings->common.rsp_prod_pvt++; - RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&blk_rings->common, notify); + + blkif->ops->push_back_ring_rsp(blk_rings, nr_page, &notify); + spin_unlock_irqrestore(&blkif->blk_ring_lock, flags); if (notify) notify_remote_via_irq(blkif->irq); @@ -873,6 +939,14 @@ static int __init xen_blkif_init(void) return rc; } +struct blkback_ring_operation blkback_ring_ops = { + .get_back_ring = get_back_ring, + .copy_blkif_req = copy_blkif_req, + .copy_blkif_seg_...

The following series flushes my documentation queue and replaces previous postings of those patches. The main difference is that the xl cfg file is now formatted using POD instead of markdown and presented as a manpage. I have setup a cron job to build docs/html and publish it at http://xenbits.xen.org/docs/unstable/ (it''s a bit bare right now). The motivation for some of these patches