The following changes since commit dac03af5d5482ec7ee9c23db467bb7230b33c0d9:

Merge remote-tracking branch 'remotes/rth/tags/pull-axp-20190825' into staging (2019-08-27 10:00:51 +0100)

are available in the Git repository at:

https://github.com/stefanha/qemu.git tags/block-pull-request

for you to fetch changes up to 5396234b96a2ac743f48644529771498e036e698:

block/qcow2: implement .bdrv_co_pwritev(_compressed)_part (2019-08-27 14:58:42 +0100)

----------------------------------------------------------------

Pull request

Vladimir Sementsov-Ogievskiy (12):

util/iov: introduce qemu_iovec_init_extended

util/iov: improve qemu_iovec_is_zero

block/io: refactor padding

block: define .*_part io handlers in BlockDriver

block/io: bdrv_co_do_copy_on_readv: use and support qiov_offset

block/io: bdrv_co_do_copy_on_readv: lazy allocation

block/io: bdrv_aligned_preadv: use and support qiov_offset

block/io: bdrv_aligned_pwritev: use and support qiov_offset

block/io: introduce bdrv_co_p{read, write}v_part

block/qcow2: refactor qcow2_co_preadv to use buffer-based io

block/qcow2: implement .bdrv_co_preadv_part

block/qcow2: implement .bdrv_co_pwritev(_compressed)_part

block/qcow2.h | 1 +

include/block/block_int.h | 21 ++

include/qemu/iov.h | 10 +-

block/backup.c | 2 +-

block/io.c | 541 +++++++++++++++++++++++---------------

block/qcow2-cluster.c | 14 +-

block/qcow2.c | 131 +++++----

qemu-img.c | 4 +-

util/iov.c | 153 +++++++++--

9 files changed, 568 insertions(+), 309 deletions(-)

2.21.0

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Introduce new initialization API, to create requests with padding. Will

be used in the following patch. New API uses qemu_iovec_init_buf if

resulting io vector has only one element, to avoid extra allocations.

So, we need to update qemu_iovec_destroy to support destroying such

QIOVs.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Acked-by: Stefan Hajnoczi <stefanha@redhat.com>

Message-id: 20190604161514.262241-2-vsementsov@virtuozzo.com

Message-Id: <20190604161514.262241-2-vsementsov@virtuozzo.com>

include/qemu/iov.h | 7 +++

util/iov.c | 112 +++++++++++++++++++++++++++++++++++++++++++--

2 files changed, 114 insertions(+), 5 deletions(-)

diff --git a/include/qemu/iov.h b/include/qemu/iov.h

--- a/include/qemu/iov.h

+++ b/include/qemu/iov.h

@@ -XXX,XX +XXX,XX @@ static inline void *qemu_iovec_buf(QEMUIOVector *qiov)

void qemu_iovec_init(QEMUIOVector *qiov, int alloc_hint);

void qemu_iovec_init_external(QEMUIOVector *qiov, struct iovec *iov, int niov);

+void qemu_iovec_init_extended(

+ QEMUIOVector *qiov,

+ void *head_buf, size_t head_len,

+ QEMUIOVector *mid_qiov, size_t mid_offset, size_t mid_len,

+ void *tail_buf, size_t tail_len);

+void qemu_iovec_init_slice(QEMUIOVector *qiov, QEMUIOVector *source,

+ size_t offset, size_t len);

void qemu_iovec_add(QEMUIOVector *qiov, void *base, size_t len);

void qemu_iovec_concat(QEMUIOVector *dst,

QEMUIOVector *src, size_t soffset, size_t sbytes);

diff --git a/util/iov.c b/util/iov.c

index XXXXXXX..XXXXXXX 100644

--- a/util/iov.c

+++ b/util/iov.c

@@ -XXX,XX +XXX,XX @@ void qemu_iovec_concat(QEMUIOVector *dst,

qemu_iovec_concat_iov(dst, src->iov, src->niov, soffset, sbytes);

+/*

+ * qiov_find_iov

+ *

+ * Return pointer to iovec structure, where byte at @offset in original vector

+ * @iov exactly is.

+ * Set @remaining_offset to be offset inside that iovec to the same byte.

+ */

+static struct iovec *iov_skip_offset(struct iovec *iov, size_t offset,

+ size_t *remaining_offset)

+ while (offset > 0 && offset >= iov->iov_len) {

+ offset -= iov->iov_len;

+ iov++;

+ }

+ *remaining_offset = offset;

+ return iov;

+}

+/*

+ * qiov_slice

+ *

+ * Find subarray of iovec's, containing requested range. @head would

+ * be offset in first iov (returned by the function), @tail would be

+ * count of extra bytes in last iovec (returned iov + @niov - 1).

+ */

+static struct iovec *qiov_slice(QEMUIOVector *qiov,

+ size_t offset, size_t len,

+ size_t *head, size_t *tail, int *niov)

+{

+ struct iovec *iov, *end_iov;

+ assert(offset + len <= qiov->size);

+ iov = iov_skip_offset(qiov->iov, offset, head);

+ end_iov = iov_skip_offset(iov, *head + len, tail);

+ if (*tail > 0) {

+ assert(*tail < end_iov->iov_len);

+ *tail = end_iov->iov_len - *tail;

+ end_iov++;

+ }

+ *niov = end_iov - iov;

+ return iov;

+}

+/*

+ * Compile new iovec, combining @head_buf buffer, sub-qiov of @mid_qiov,

+ * and @tail_buf buffer into new qiov.

+ */

+void qemu_iovec_init_extended(

+ QEMUIOVector *qiov,

+ void *head_buf, size_t head_len,

+ QEMUIOVector *mid_qiov, size_t mid_offset, size_t mid_len,

+ void *tail_buf, size_t tail_len)

+{

+ size_t mid_head, mid_tail;

+ int total_niov, mid_niov = 0;

+ struct iovec *p, *mid_iov;

+ if (mid_len) {

+ mid_iov = qiov_slice(mid_qiov, mid_offset, mid_len,

+ &mid_head, &mid_tail, &mid_niov);

+ }

+ total_niov = !!head_len + mid_niov + !!tail_len;

+ if (total_niov == 1) {

+ qemu_iovec_init_buf(qiov, NULL, 0);

+ p = &qiov->local_iov;

+ } else {

+ qiov->niov = qiov->nalloc = total_niov;

+ qiov->size = head_len + mid_len + tail_len;

+ p = qiov->iov = g_new(struct iovec, qiov->niov);

+ }

+ if (head_len) {

+ p->iov_base = head_buf;

+ p->iov_len = head_len;

+ p++;

+ }

+ if (mid_len) {

+ memcpy(p, mid_iov, mid_niov * sizeof(*p));

+ p[0].iov_base = (uint8_t *)p[0].iov_base + mid_head;

+ p[0].iov_len -= mid_head;

+ p[mid_niov - 1].iov_len -= mid_tail;

+ p += mid_niov;

+ }

+ if (tail_len) {

+ p->iov_base = tail_buf;

+ p->iov_len = tail_len;

/*

* Check if the contents of the iovecs are all zero

*/

@@ -XXX,XX +XXX,XX @@ bool qemu_iovec_is_zero(QEMUIOVector *qiov)

return true;

}

+void qemu_iovec_init_slice(QEMUIOVector *qiov, QEMUIOVector *source,

+ size_t offset, size_t len)

+{

+ qemu_iovec_init_extended(qiov, NULL, 0, source, offset, len, NULL, 0);

+}

+

void qemu_iovec_destroy(QEMUIOVector *qiov)

- assert(qiov->nalloc != -1);

+ if (qiov->nalloc != -1) {

+ g_free(qiov->iov);

+ }

- qemu_iovec_reset(qiov);

- g_free(qiov->iov);

- qiov->nalloc = 0;

- qiov->iov = NULL;

+ memset(qiov, 0, sizeof(*qiov));

}

void qemu_iovec_reset(QEMUIOVector *qiov)

2.21.0

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

We'll need to check a part of qiov soon, so implement it now.

Optimization with align down to 4 * sizeof(long) is dropped due to:

1. It is strange: it aligns length of the buffer, but where is a

guarantee that buffer pointer is aligned itself?

2. buffer_is_zero() is a better place for optimizations and it has

them.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Acked-by: Stefan Hajnoczi <stefanha@redhat.com>

Message-id: 20190604161514.262241-3-vsementsov@virtuozzo.com

Message-Id: <20190604161514.262241-3-vsementsov@virtuozzo.com>

include/qemu/iov.h | 2 +-

block/io.c | 2 +-

util/iov.c | 31 +++++++++++++++++++------------

3 files changed, 21 insertions(+), 14 deletions(-)

diff --git a/include/qemu/iov.h b/include/qemu/iov.h

--- a/include/qemu/iov.h

+++ b/include/qemu/iov.h

@@ -XXX,XX +XXX,XX @@ void qemu_iovec_concat(QEMUIOVector *dst,

size_t qemu_iovec_concat_iov(QEMUIOVector *dst,

struct iovec *src_iov, unsigned int src_cnt,

size_t soffset, size_t sbytes);

-bool qemu_iovec_is_zero(QEMUIOVector *qiov);

+bool qemu_iovec_is_zero(QEMUIOVector *qiov, size_t qiov_offeset, size_t bytes);

void qemu_iovec_destroy(QEMUIOVector *qiov);

void qemu_iovec_reset(QEMUIOVector *qiov);

size_t qemu_iovec_to_buf(QEMUIOVector *qiov, size_t offset,

diff --git a/block/io.c b/block/io.c

index XXXXXXX..XXXXXXX 100644

--- a/block/io.c

+++ b/block/io.c

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_aligned_pwritev(BdrvChild *child,

if (!ret && bs->detect_zeroes != BLOCKDEV_DETECT_ZEROES_OPTIONS_OFF &&

!(flags & BDRV_REQ_ZERO_WRITE) && drv->bdrv_co_pwrite_zeroes &&

- qemu_iovec_is_zero(qiov)) {

+ qemu_iovec_is_zero(qiov, 0, qiov->size)) {

flags |= BDRV_REQ_ZERO_WRITE;

if (bs->detect_zeroes == BLOCKDEV_DETECT_ZEROES_OPTIONS_UNMAP) {

flags |= BDRV_REQ_MAY_UNMAP;

diff --git a/util/iov.c b/util/iov.c

index XXXXXXX..XXXXXXX 100644

--- a/util/iov.c

+++ b/util/iov.c

@@ -XXX,XX +XXX,XX @@ void qemu_iovec_init_extended(

}

/*

- * Check if the contents of the iovecs are all zero

+ * Check if the contents of subrange of qiov data is all zeroes.

*/

-bool qemu_iovec_is_zero(QEMUIOVector *qiov)

+bool qemu_iovec_is_zero(QEMUIOVector *qiov, size_t offset, size_t bytes)

{

- int i;

- for (i = 0; i < qiov->niov; i++) {

- size_t offs = QEMU_ALIGN_DOWN(qiov->iov[i].iov_len, 4 * sizeof(long));

- uint8_t *ptr = qiov->iov[i].iov_base;

- if (offs && !buffer_is_zero(qiov->iov[i].iov_base, offs)) {

+ struct iovec *iov;

+ size_t current_offset;

+

+ assert(offset + bytes <= qiov->size);

+

+ iov = iov_skip_offset(qiov->iov, offset, &current_offset);

+

+ while (bytes) {

+ uint8_t *base = (uint8_t *)iov->iov_base + current_offset;

+ size_t len = MIN(iov->iov_len - current_offset, bytes);

+

+ if (!buffer_is_zero(base, len)) {

return false;

}

- for (; offs < qiov->iov[i].iov_len; offs++) {

- if (ptr[offs]) {

- return false;

- }

- }

+

+ current_offset = 0;

+ bytes -= len;

+ iov++;

}

+

return true;

}

2.21.0

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

We have similar padding code in bdrv_co_pwritev,

bdrv_co_do_pwrite_zeroes and bdrv_co_preadv. Let's combine and unify

it.

[Squashed in Vladimir's qemu-iotests 077 fix

--Stefan]

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Acked-by: Stefan Hajnoczi <stefanha@redhat.com>

Message-id: 20190604161514.262241-4-vsementsov@virtuozzo.com

Message-Id: <20190604161514.262241-4-vsementsov@virtuozzo.com>

block/io.c | 365 +++++++++++++++++++++++++++++------------------------

1 file changed, 200 insertions(+), 165 deletions(-)

diff --git a/block/io.c b/block/io.c

--- a/block/io.c

+++ b/block/io.c

@@ -XXX,XX +XXX,XX @@ out:

}

/*

- * Handle a read request in coroutine context

+ * Request padding

+ *

+ * |<---- align ----->| |<----- align ---->|

+ * |<- head ->|<------------- bytes ------------->|<-- tail -->|

+ * | | | | | |

+ * -*----------$-------*-------- ... --------*-----$------------*---

+ * | | | | | |

+ * | offset | | end |

+ * ALIGN_DOWN(offset) ALIGN_UP(offset) ALIGN_DOWN(end) ALIGN_UP(end)

+ * [buf ... ) [tail_buf )

+ *

+ * @buf is an aligned allocation needed to store @head and @tail paddings. @head

+ * is placed at the beginning of @buf and @tail at the @end.

+ *

+ * @tail_buf is a pointer to sub-buffer, corresponding to align-sized chunk

+ * around tail, if tail exists.

+ *

+ * @merge_reads is true for small requests,

+ * if @buf_len == @head + bytes + @tail. In this case it is possible that both

+ * head and tail exist but @buf_len == align and @tail_buf == @buf.

*/

+typedef struct BdrvRequestPadding {

+ uint8_t *buf;

+ size_t buf_len;

+ uint8_t *tail_buf;

+ size_t head;

+ size_t tail;

+ bool merge_reads;

+ QEMUIOVector local_qiov;

+} BdrvRequestPadding;

+

+static bool bdrv_init_padding(BlockDriverState *bs,

+ int64_t offset, int64_t bytes,

+ BdrvRequestPadding *pad)

+{

+ uint64_t align = bs->bl.request_alignment;

+ size_t sum;

+

+ memset(pad, 0, sizeof(*pad));

+

+ pad->head = offset & (align - 1);

+ pad->tail = ((offset + bytes) & (align - 1));

+ if (pad->tail) {

+ pad->tail = align - pad->tail;

+ }

+

+ if ((!pad->head && !pad->tail) || !bytes) {

+ return false;

+ }

+

+ sum = pad->head + bytes + pad->tail;

+ pad->buf_len = (sum > align && pad->head && pad->tail) ? 2 * align : align;

+ pad->buf = qemu_blockalign(bs, pad->buf_len);

+ pad->merge_reads = sum == pad->buf_len;

+ if (pad->tail) {

+ pad->tail_buf = pad->buf + pad->buf_len - align;

+ }

+

+ return true;

+}

+

+static int bdrv_padding_rmw_read(BdrvChild *child,

+ BdrvTrackedRequest *req,

+ BdrvRequestPadding *pad,

+ bool zero_middle)

+{

+ QEMUIOVector local_qiov;

+ BlockDriverState *bs = child->bs;

+ uint64_t align = bs->bl.request_alignment;

+ int ret;

+

+ assert(req->serialising && pad->buf);

+

+ if (pad->head || pad->merge_reads) {

+ uint64_t bytes = pad->merge_reads ? pad->buf_len : align;

+

+ qemu_iovec_init_buf(&local_qiov, pad->buf, bytes);

+

+ if (pad->head) {

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_HEAD);

+ }

+ if (pad->merge_reads && pad->tail) {

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);

+ }

+ ret = bdrv_aligned_preadv(child, req, req->overlap_offset, bytes,

+ align, &local_qiov, 0);

+ if (ret < 0) {

+ return ret;

+ }

+ if (pad->head) {

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_HEAD);

+ }

+ if (pad->merge_reads && pad->tail) {

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_TAIL);

+ }

+

+ if (pad->merge_reads) {

+ goto zero_mem;

+ }

+ }

+

+ if (pad->tail) {

+ qemu_iovec_init_buf(&local_qiov, pad->tail_buf, align);

+

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);

+ ret = bdrv_aligned_preadv(

+ child, req,

+ req->overlap_offset + req->overlap_bytes - align,

+ align, align, &local_qiov, 0);

+ if (ret < 0) {

+ return ret;

+ }

+ bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_TAIL);

+ }

+

+zero_mem:

+ if (zero_middle) {

+ memset(pad->buf + pad->head, 0, pad->buf_len - pad->head - pad->tail);

+ }

+

+ return 0;

+}

+

+static void bdrv_padding_destroy(BdrvRequestPadding *pad)

+{

+ if (pad->buf) {

+ qemu_vfree(pad->buf);

+ qemu_iovec_destroy(&pad->local_qiov);

+ }

+}

+

+/*

+ * bdrv_pad_request

+ *

+ * Exchange request parameters with padded request if needed. Don't include RMW

+ * read of padding, bdrv_padding_rmw_read() should be called separately if

+ * needed.

+ *

+ * All parameters except @bs are in-out: they represent original request at

+ * function call and padded (if padding needed) at function finish.

+ *

+ * Function always succeeds.

+ */

+static bool bdrv_pad_request(BlockDriverState *bs, QEMUIOVector **qiov,

+ int64_t *offset, unsigned int *bytes,

+ BdrvRequestPadding *pad)

+{

+ if (!bdrv_init_padding(bs, *offset, *bytes, pad)) {

+ return false;

+ }

+

+ qemu_iovec_init_extended(&pad->local_qiov, pad->buf, pad->head,

+ *qiov, 0, *bytes,

+ pad->buf + pad->buf_len - pad->tail, pad->tail);

+ *bytes += pad->head + pad->tail;

+ *offset -= pad->head;

+ *qiov = &pad->local_qiov;

+

+ return true;

+}

+

int coroutine_fn bdrv_co_preadv(BdrvChild *child,

int64_t offset, unsigned int bytes, QEMUIOVector *qiov,

BdrvRequestFlags flags)

BlockDriverState *bs = child->bs;

- BlockDriver *drv = bs->drv;

BdrvTrackedRequest req;

-

- uint64_t align = bs->bl.request_alignment;

- uint8_t *head_buf = NULL;

- uint8_t *tail_buf = NULL;

- QEMUIOVector local_qiov;

- bool use_local_qiov = false;

+ BdrvRequestPadding pad;

int ret;

- trace_bdrv_co_preadv(child->bs, offset, bytes, flags);

-

- if (!drv) {

- return -ENOMEDIUM;

- }

+ trace_bdrv_co_preadv(bs, offset, bytes, flags);

ret = bdrv_check_byte_request(bs, offset, bytes);

if (ret < 0) {

@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_preadv(BdrvChild *child,

flags |= BDRV_REQ_COPY_ON_READ;

- /* Align read if necessary by padding qiov */

- if (offset & (align - 1)) {

- head_buf = qemu_blockalign(bs, align);

- qemu_iovec_init(&local_qiov, qiov->niov + 2);

- qemu_iovec_add(&local_qiov, head_buf, offset & (align - 1));

- qemu_iovec_concat(&local_qiov, qiov, 0, qiov->size);

- use_local_qiov = true;

-

- bytes += offset & (align - 1);

- offset = offset & ~(align - 1);

- }

-

- if ((offset + bytes) & (align - 1)) {

- if (!use_local_qiov) {

- qemu_iovec_init(&local_qiov, qiov->niov + 1);

- qemu_iovec_concat(&local_qiov, qiov, 0, qiov->size);

- use_local_qiov = true;

- }

- tail_buf = qemu_blockalign(bs, align);

- qemu_iovec_add(&local_qiov, tail_buf,

- align - ((offset + bytes) & (align - 1)));

-

- bytes = ROUND_UP(bytes, align);

- }

+ bdrv_pad_request(bs, &qiov, &offset, &bytes, &pad);

tracked_request_begin(&req, bs, offset, bytes, BDRV_TRACKED_READ);

- ret = bdrv_aligned_preadv(child, &req, offset, bytes, align,

- use_local_qiov ? &local_qiov : qiov,

- flags);

+ ret = bdrv_aligned_preadv(child, &req, offset, bytes,

+ bs->bl.request_alignment,

+ qiov, flags);

tracked_request_end(&req);

bdrv_dec_in_flight(bs);

- if (use_local_qiov) {

- qemu_iovec_destroy(&local_qiov);

- qemu_vfree(head_buf);

- qemu_vfree(tail_buf);

- }

+ bdrv_padding_destroy(&pad);

return ret;

}

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,

BdrvTrackedRequest *req)

{

BlockDriverState *bs = child->bs;

- uint8_t *buf = NULL;

QEMUIOVector local_qiov;

uint64_t align = bs->bl.request_alignment;

- unsigned int head_padding_bytes, tail_padding_bytes;

int ret = 0;

+ bool padding;

+ BdrvRequestPadding pad;

- head_padding_bytes = offset & (align - 1);

- tail_padding_bytes = (align - (offset + bytes)) & (align - 1);

-

-

- assert(flags & BDRV_REQ_ZERO_WRITE);

- if (head_padding_bytes || tail_padding_bytes) {

- buf = qemu_blockalign(bs, align);

- qemu_iovec_init_buf(&local_qiov, buf, align);

- }

- if (head_padding_bytes) {

- uint64_t zero_bytes = MIN(bytes, align - head_padding_bytes);

-

- /* RMW the unaligned part before head. */

+ padding = bdrv_init_padding(bs, offset, bytes, &pad);

+ if (padding) {

mark_request_serialising(req, align);

wait_serialising_requests(req);

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_HEAD);

- ret = bdrv_aligned_preadv(child, req, offset & ~(align - 1), align,

- align, &local_qiov, 0);

- if (ret < 0) {

- goto fail;

- }

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_HEAD);

- memset(buf + head_padding_bytes, 0, zero_bytes);

- ret = bdrv_aligned_pwritev(child, req, offset & ~(align - 1), align,

- align, &local_qiov,

- flags & ~BDRV_REQ_ZERO_WRITE);

- if (ret < 0) {

- goto fail;

+ bdrv_padding_rmw_read(child, req, &pad, true);

+

+ if (pad.head || pad.merge_reads) {

+ int64_t aligned_offset = offset & ~(align - 1);

+ int64_t write_bytes = pad.merge_reads ? pad.buf_len : align;

+

+ qemu_iovec_init_buf(&local_qiov, pad.buf, write_bytes);

+ ret = bdrv_aligned_pwritev(child, req, aligned_offset, write_bytes,

+ align, &local_qiov,

+ flags & ~BDRV_REQ_ZERO_WRITE);

+ if (ret < 0 || pad.merge_reads) {

+ /* Error or all work is done */

+ goto out;

+ }

+ offset += write_bytes - pad.head;

+ bytes -= write_bytes - pad.head;

}

- offset += zero_bytes;

- bytes -= zero_bytes;

assert(!bytes || (offset & (align - 1)) == 0);

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,

ret = bdrv_aligned_pwritev(child, req, offset, aligned_bytes, align,

NULL, flags);

if (ret < 0) {

- goto fail;

+ goto out;

bytes -= aligned_bytes;

offset += aligned_bytes;

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_zero_pwritev(BdrvChild *child,

assert(!bytes || (offset & (align - 1)) == 0);

if (bytes) {

- assert(align == tail_padding_bytes + bytes);

- /* RMW the unaligned part after tail. */

- mark_request_serialising(req, align);

- wait_serialising_requests(req);

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);

- ret = bdrv_aligned_preadv(child, req, offset, align,

- align, &local_qiov, 0);

- if (ret < 0) {

- goto fail;

- }

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_TAIL);

+ assert(align == pad.tail + bytes);

- memset(buf, 0, bytes);

+ qemu_iovec_init_buf(&local_qiov, pad.tail_buf, align);

ret = bdrv_aligned_pwritev(child, req, offset, align, align,

&local_qiov, flags & ~BDRV_REQ_ZERO_WRITE);

}

-fail:

- qemu_vfree(buf);

+

+out:

+ bdrv_padding_destroy(&pad);

+

return ret;

-

/*

@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,

BlockDriverState *bs = child->bs;

BdrvTrackedRequest req;

uint64_t align = bs->bl.request_alignment;

- uint8_t *head_buf = NULL;

- uint8_t *tail_buf = NULL;

- QEMUIOVector local_qiov;

- bool use_local_qiov = false;

+ BdrvRequestPadding pad;

int ret;

trace_bdrv_co_pwritev(child->bs, offset, bytes, flags);

@@ -XXX,XX +XXX,XX @@ int coroutine_fn bdrv_co_pwritev(BdrvChild *child,

goto out;

- if (offset & (align - 1)) {

- QEMUIOVector head_qiov;

-

+ if (bdrv_pad_request(bs, &qiov, &offset, &bytes, &pad)) {

mark_request_serialising(&req, align);

wait_serialising_requests(&req);

-

- head_buf = qemu_blockalign(bs, align);

- qemu_iovec_init_buf(&head_qiov, head_buf, align);

-

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_HEAD);

- ret = bdrv_aligned_preadv(child, &req, offset & ~(align - 1), align,

- align, &head_qiov, 0);

- if (ret < 0) {

- goto fail;

- }

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_HEAD);

-

- qemu_iovec_init(&local_qiov, qiov->niov + 2);

- qemu_iovec_add(&local_qiov, head_buf, offset & (align - 1));

- qemu_iovec_concat(&local_qiov, qiov, 0, qiov->size);

- use_local_qiov = true;

-

- bytes += offset & (align - 1);

- offset = offset & ~(align - 1);

-

- /* We have read the tail already if the request is smaller

- * than one aligned block.

- */

- if (bytes < align) {

- qemu_iovec_add(&local_qiov, head_buf + bytes, align - bytes);

- bytes = align;

- }

- }

-

- if ((offset + bytes) & (align - 1)) {

- QEMUIOVector tail_qiov;

- size_t tail_bytes;

- bool waited;

-

- mark_request_serialising(&req, align);

- waited = wait_serialising_requests(&req);

- assert(!waited || !use_local_qiov);

-

- tail_buf = qemu_blockalign(bs, align);

- qemu_iovec_init_buf(&tail_qiov, tail_buf, align);

-

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_TAIL);

- ret = bdrv_aligned_preadv(child, &req, (offset + bytes) & ~(align - 1),

- align, align, &tail_qiov, 0);

- if (ret < 0) {

- goto fail;

- }

- bdrv_debug_event(bs, BLKDBG_PWRITEV_RMW_AFTER_TAIL);

-

- if (!use_local_qiov) {

- qemu_iovec_init(&local_qiov, qiov->niov + 1);

- qemu_iovec_concat(&local_qiov, qiov, 0, qiov->size);

- use_local_qiov = true;

- }

-

- tail_bytes = (offset + bytes) & (align - 1);

- qemu_iovec_add(&local_qiov, tail_buf + tail_bytes, align - tail_bytes);

-

- bytes = ROUND_UP(bytes, align);

+ bdrv_padding_rmw_read(child, &req, &pad, false);

}

ret = bdrv_aligned_pwritev(child, &req, offset, bytes, align,

- use_local_qiov ? &local_qiov : qiov,

- flags);

+ qiov, flags);

-fail:

+ bdrv_padding_destroy(&pad);

- if (use_local_qiov) {

- qemu_iovec_destroy(&local_qiov);

- }

- qemu_vfree(head_buf);

- qemu_vfree(tail_buf);

out:

tracked_request_end(&req);

bdrv_dec_in_flight(bs);

+

return ret;

2.21.0

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Add handlers supporting qiov_offset parameter:

bdrv_co_preadv_part

bdrv_co_pwritev_part

bdrv_co_pwritev_compressed_part

This is used to reduce need of defining local_qiovs and hd_qiovs in all

corners of block layer code. The following patches will increase usage

of this new API part by part.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Acked-by: Stefan Hajnoczi <stefanha@redhat.com>

Message-id: 20190604161514.262241-5-vsementsov@virtuozzo.com

Message-Id: <20190604161514.262241-5-vsementsov@virtuozzo.com>

include/block/block_int.h | 15 ++++++

block/backup.c | 2 +-

block/io.c | 96 +++++++++++++++++++++++++++++++--------

qemu-img.c | 4 +-

4 files changed, 95 insertions(+), 22 deletions(-)

diff --git a/include/block/block_int.h b/include/block/block_int.h

--- a/include/block/block_int.h

+++ b/include/block/block_int.h

@@ -XXX,XX +XXX,XX @@ struct BlockDriver {

*/

int coroutine_fn (*bdrv_co_preadv)(BlockDriverState *bs,

uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags);

+ int coroutine_fn (*bdrv_co_preadv_part)(BlockDriverState *bs,

+ uint64_t offset, uint64_t bytes,

+ QEMUIOVector *qiov, size_t qiov_offset, int flags);

int coroutine_fn (*bdrv_co_writev)(BlockDriverState *bs,

int64_t sector_num, int nb_sectors, QEMUIOVector *qiov, int flags);

/**

@@ -XXX,XX +XXX,XX @@ struct BlockDriver {

*/

int coroutine_fn (*bdrv_co_pwritev)(BlockDriverState *bs,

uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags);

+ int coroutine_fn (*bdrv_co_pwritev_part)(BlockDriverState *bs,

+ uint64_t offset, uint64_t bytes,

+ QEMUIOVector *qiov, size_t qiov_offset, int flags);

/*

* Efficiently zero a region of the disk image. Typically an image format

@@ -XXX,XX +XXX,XX @@ struct BlockDriver {

int coroutine_fn (*bdrv_co_pwritev_compressed)(BlockDriverState *bs,

uint64_t offset, uint64_t bytes, QEMUIOVector *qiov);

+ int coroutine_fn (*bdrv_co_pwritev_compressed_part)(BlockDriverState *bs,

+ uint64_t offset, uint64_t bytes, QEMUIOVector *qiov,

+ size_t qiov_offset);

int (*bdrv_snapshot_create)(BlockDriverState *bs,

QEMUSnapshotInfo *sn_info);

@@ -XXX,XX +XXX,XX @@ struct BlockDriver {

const char *const *strong_runtime_opts;

};

+static inline bool block_driver_can_compress(BlockDriver *drv)

+{

+ return drv->bdrv_co_pwritev_compressed ||

+ drv->bdrv_co_pwritev_compressed_part;

+}

+

typedef struct BlockLimits {

/* Alignment requirement, in bytes, for offset/length of I/O

* requests. Must be a power of 2 less than INT_MAX; defaults to

diff --git a/block/backup.c b/block/backup.c

index XXXXXXX..XXXXXXX 100644

--- a/block/backup.c

+++ b/block/backup.c

@@ -XXX,XX +XXX,XX @@ BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,

return NULL;

}

- if (compress && target->drv->bdrv_co_pwritev_compressed == NULL) {

+ if (compress && !block_driver_can_compress(target->drv)) {

error_setg(errp, "Compression is not supported for this drive %s",

bdrv_get_device_name(target));

return NULL;

diff --git a/block/io.c b/block/io.c

index XXXXXXX..XXXXXXX 100644

--- a/block/io.c

+++ b/block/io.c

@@ -XXX,XX +XXX,XX @@ void bdrv_refresh_limits(BlockDriverState *bs, Error **errp)

/* Default alignment based on whether driver has byte interface */

bs->bl.request_alignment = (drv->bdrv_co_preadv ||

- drv->bdrv_aio_preadv) ? 1 : 512;

+ drv->bdrv_aio_preadv ||

+ drv->bdrv_co_preadv_part) ? 1 : 512;

/* Take some limits from the children as a default */

if (bs->file) {

@@ -XXX,XX +XXX,XX @@ static void bdrv_co_io_em_complete(void *opaque, int ret)

static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,

uint64_t offset, uint64_t bytes,

- QEMUIOVector *qiov, int flags)

+ QEMUIOVector *qiov,

+ size_t qiov_offset, int flags)

{

BlockDriver *drv = bs->drv;

int64_t sector_num;

unsigned int nb_sectors;

+ QEMUIOVector local_qiov;

+ int ret;

assert(!(flags & ~BDRV_REQ_MASK));

assert(!(flags & BDRV_REQ_NO_FALLBACK));

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,

return -ENOMEDIUM;

}

+ if (drv->bdrv_co_preadv_part) {

+ return drv->bdrv_co_preadv_part(bs, offset, bytes, qiov, qiov_offset,

+ flags);

+ }

+

+ if (qiov_offset > 0 || bytes != qiov->size) {

+ qemu_iovec_init_slice(&local_qiov, qiov, qiov_offset, bytes);

+ qiov = &local_qiov;

+ }

+

if (drv->bdrv_co_preadv) {

- return drv->bdrv_co_preadv(bs, offset, bytes, qiov, flags);

+ ret = drv->bdrv_co_preadv(bs, offset, bytes, qiov, flags);

+ goto out;

}

if (drv->bdrv_aio_preadv) {

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,

acb = drv->bdrv_aio_preadv(bs, offset, bytes, qiov, flags,

bdrv_co_io_em_complete, &co);

if (acb == NULL) {

- return -EIO;

+ ret = -EIO;

+ goto out;

} else {

qemu_coroutine_yield();

- return co.ret;

+ ret = co.ret;

+ goto out;

}

}

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_driver_preadv(BlockDriverState *bs,

assert(bytes <= BDRV_REQUEST_MAX_BYTES);

assert(drv->bdrv_co_readv);

- return drv->bdrv_co_readv(bs, sector_num, nb_sectors, qiov);

+ ret = drv->bdrv_co_readv(bs, sector_num, nb_sectors, qiov);

+

+out:

+ if (qiov == &local_qiov) {

+ qemu_iovec_destroy(&local_qiov);

+ }

+

+ return ret;

}

static int coroutine_fn bdrv_driver_pwritev(BlockDriverState *bs,

uint64_t offset, uint64_t bytes,

- QEMUIOVector *qiov, int flags)

+ QEMUIOVector *qiov,

+ size_t qiov_offset, int flags)

{

BlockDriver *drv = bs->drv;

int64_t sector_num;

unsigned int nb_sectors;

+ QEMUIOVector local_qiov;

int ret;

assert(!(flags & ~BDRV_REQ_MASK));

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_driver_pwritev(BlockDriverState *bs,

return -ENOMEDIUM;

}

+ if (drv->bdrv_co_pwritev_part) {

+ ret = drv->bdrv_co_pwritev_part(bs, offset, bytes, qiov, qiov_offset,

+ flags & bs->supported_write_flags);

+ flags &= ~bs->supported_write_flags;

+ goto emulate_flags;

+ }

+

+ if (qiov_offset > 0 || bytes != qiov->size) {

+ qemu_iovec_init_slice(&local_qiov, qiov, qiov_offset, bytes);

+ qiov = &local_qiov;

+ }

+

if (drv->bdrv_co_pwritev) {

ret = drv->bdrv_co_pwritev(bs, offset, bytes, qiov,

flags & bs->supported_write_flags);

@@ -XXX,XX +XXX,XX @@ emulate_flags:

ret = bdrv_co_flush(bs);

}

+ if (qiov == &local_qiov) {

+ qemu_iovec_destroy(&local_qiov);

+ }

+

return ret;

}

static int coroutine_fn

bdrv_driver_pwritev_compressed(BlockDriverState *bs, uint64_t offset,

- uint64_t bytes, QEMUIOVector *qiov)

+ uint64_t bytes, QEMUIOVector *qiov,

+ size_t qiov_offset)

{

BlockDriver *drv = bs->drv;

+ QEMUIOVector local_qiov;

+ int ret;

if (!drv) {

return -ENOMEDIUM;

}

- if (!drv->bdrv_co_pwritev_compressed) {

+ if (!block_driver_can_compress(drv)) {

return -ENOTSUP;

}

- return drv->bdrv_co_pwritev_compressed(bs, offset, bytes, qiov);

+ if (drv->bdrv_co_pwritev_compressed_part) {

+ return drv->bdrv_co_pwritev_compressed_part(bs, offset, bytes,

+ qiov, qiov_offset);

+ }

+

+ if (qiov_offset == 0) {

+ return drv->bdrv_co_pwritev_compressed(bs, offset, bytes, qiov);

+ }

+

+ qemu_iovec_init_slice(&local_qiov, qiov, qiov_offset, bytes);

+ ret = drv->bdrv_co_pwritev_compressed(bs, offset, bytes, &local_qiov);

+ qemu_iovec_destroy(&local_qiov);

+

+ return ret;

}

static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,

@@ -XXX,XX +XXX,XX @@ static int coroutine_fn bdrv_co_do_copy_on_readv(BdrvChild *child,

qemu_iovec_init_buf(&local_qiov, bounce_buffer, pnum);

ret = bdrv_driver_preadv(bs, cluster_offset, pnum,

- &local_qiov, 0);

+ &local_qiov, 0, 0);