similar to: [PATCH RFC 0/2] Improve virtio-blk performance

This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential read/write, random read/write IOPS boost : 28%, 24%, 21%, 16% Latency improvement: 32%, 17%, 21%, 16% 2) Fusion IO device With bio-based IO path, sequential

This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential read/write, random read/write IOPS boost : 28%, 24%, 21%, 16% Latency improvement: 32%, 17%, 21%, 16% 2) Fusion IO device With bio-based IO path, sequential

Hi folks, This version added REQ_FLUSH and REQ_FUA support as suggested by Christoph and rebased against latest linus's tree. Jens, could you please consider picking up the dependencies 1/4 and 2/4 in your tree. Thanks! This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk

Hi folks, This version added REQ_FLUSH and REQ_FUA support as suggested by Christoph and rebased against latest linus's tree. Jens, could you please consider picking up the dependencies 1/4 and 2/4 in your tree. Thanks! This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk

This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows it gives, 28%, 24%, 21%, 16% IOPS boost and 32%, 17%, 21%, 16% latency improvement for sequential read/write, random read/write respectively. Asias He (3): block: Introduce __blk_segment_map_sg() helper block: Add blk_bio_map_sg() helper virtio-blk: Add bio-based IO path for virtio-blk

This patchset implements bio-based IO path for virito-blk to improve performance. Fio test shows it gives, 28%, 24%, 21%, 16% IOPS boost and 32%, 17%, 21%, 16% latency improvement for sequential read/write, random read/write respectively. Asias He (3): block: Introduce __blk_segment_map_sg() helper block: Add blk_bio_map_sg() helper virtio-blk: Add bio-based IO path for virtio-blk

Hi, all Changes in v7: - Using vbr->flags to trace request type - Dropped unnecessary struct virtio_blk *vblk parameter - Reuse struct virtblk_req in bio done function - Added performance data on normal SATA device and the reason why make it optional Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential

Hi, all Changes in v7: - Using vbr->flags to trace request type - Dropped unnecessary struct virtio_blk *vblk parameter - Reuse struct virtblk_req in bio done function - Added performance data on normal SATA device and the reason why make it optional Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential

Hi, Jens & Rusty This version is rebased against linux-next which resolves the conflict with Paolo Bonzini's 'virtio-blk: allow toggling host cache between writeback and writethrough' patch. Patch 1/3 and 2/3 applies on linus's master as well. Since Rusty will pick up patch 3/3 so the changes to block core (adding blk_bio_map_sg()) will have a user. Jens, could you please

Hi, Jens & Rusty This version is rebased against linux-next which resolves the conflict with Paolo Bonzini's 'virtio-blk: allow toggling host cache between writeback and writethrough' patch. Patch 1/3 and 2/3 applies on linus's master as well. Since Rusty will pick up patch 3/3 so the changes to block core (adding blk_bio_map_sg()) will have a user. Jens, could you please

Hi, all This version reworked on REQ_FLUSH and REQ_FUA support as suggested by Christoph and dropped the block core bits since Jens has picked them up. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential read/write, random read/write IOPS boost : 28%, 24%, 21%, 16% Latency improvement: 32%,

Hi, all This version reworked on REQ_FLUSH and REQ_FUA support as suggested by Christoph and dropped the block core bits since Jens has picked them up. Fio test shows bio-based IO path gives the following performance improvement: 1) Ramdisk device With bio-based IO path, sequential read/write, random read/write IOPS boost : 28%, 24%, 21%, 16% Latency improvement: 32%,

Holding the vblk->lock across kick causes poor scalability in SMP guests. If one CPU is doing virtqueue kick and another CPU touches the vblk->lock it will have to spin until virtqueue kick completes. This patch reduces system% CPU utilization in SMP guests that are running multithreaded I/O-bound workloads. The improvements are small but show as iops and SMP are increased. Khoa Huynh

Holding the vblk->lock across kick causes poor scalability in SMP guests. If one CPU is doing virtqueue kick and another CPU touches the vblk->lock it will have to spin until virtqueue kick completes. This patch reduces system% CPU utilization in SMP guests that are running multithreaded I/O-bound workloads. The improvements are small but show as iops and SMP are increased. Khoa Huynh

Most device drivers do not need to perform any postprocessing on the scatterlists they receive from higher-level drivers (e.g. the block or SCSI layer), because they translate the request metadata directly from the various C structs into the data that is required by the device. virtio devices however do this translation in two steps: a device-specific step in the device driver, and generic

Most device drivers do not need to perform any postprocessing on the scatterlists they receive from higher-level drivers (e.g. the block or SCSI layer), because they translate the request metadata directly from the various C structs into the data that is required by the device. virtio devices however do this translation in two steps: a device-specific step in the device driver, and generic

The virtqueue_add_buf function has two limitations: 1) it requires the caller to provide all the buffers in a single call; 2) it does not support chained scatterlists: the buffers must be provided as an array of struct scatterlist. Because of these limitations, virtio-scsi has to copy each request into a scatterlist internal to the driver. It cannot just use the one that was prepared by the

The virtqueue_add_buf function has two limitations: 1) it requires the caller to provide all the buffers in a single call; 2) it does not support chained scatterlists: the buffers must be provided as an array of struct scatterlist. Because of these limitations, virtio-scsi has to copy each request into a scatterlist internal to the driver. It cannot just use the one that was prepared by the

Benchmark shows small performance improvement on fusion io device. Before: seq-read : io=1,024MB, bw=19,982KB/s, iops=39,964, runt= 52475msec seq-write: io=1,024MB, bw=20,321KB/s, iops=40,641, runt= 51601msec rnd-read : io=1,024MB, bw=15,404KB/s, iops=30,808, runt= 68070msec rnd-write: io=1,024MB, bw=14,776KB/s, iops=29,552, runt= 70963msec After: seq-read : io=1,024MB, bw=20,343KB/s,

Benchmark shows small performance improvement on fusion io device. Before: seq-read : io=1,024MB, bw=19,982KB/s, iops=39,964, runt= 52475msec seq-write: io=1,024MB, bw=20,321KB/s, iops=40,641, runt= 51601msec rnd-read : io=1,024MB, bw=15,404KB/s, iops=30,808, runt= 68070msec rnd-write: io=1,024MB, bw=14,776KB/s, iops=29,552, runt= 70963msec After: seq-read : io=1,024MB, bw=20,343KB/s,