[Qemu-devel] [PATCH v3 7/8] xen_disk: use a single entry iovec

[Paul Durrant <paul.durrant@citrix.com>]

Since xen_disk now always copies data to and from a guest there is no need
to maintain a vector entry corresponding to every page of a request.
This means there is less per-request state to maintain so the ioreq
structure can shrink significantly.

Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
---
Cc: Stefano Stabellini <sstabellini@kernel.org>
Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Kevin Wolf <kwolf@redhat.com>
Cc: Max Reitz <mreitz@redhat.com>

v3:
 - Un-break by fixing mis-placed qemu_iovec_add()

v2:
 - Re-based
---
 hw/block/xen_disk.c | 76 +++++++++++++++--------------------------------------
 1 file changed, 21 insertions(+), 55 deletions(-)

diff --git a/hw/block/xen_disk.c b/hw/block/xen_disk.c
index 28be8b6..28651c5 100644
--- a/hw/block/xen_disk.c
+++ b/hw/block/xen_disk.c
@@ -46,13 +46,10 @@ struct ioreq {
     /* parsed request */
     off_t               start;
     QEMUIOVector        v;
+    void                *buf;
+    size_t              size;
     int                 presync;
 
-    /* grant mapping */
-    uint32_t            refs[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-    void                *page[BLKIF_MAX_SEGMENTS_PER_REQUEST];
-    void                *pages;
-
     /* aio status */
     int                 aio_inflight;
     int                 aio_errors;
@@ -110,12 +107,10 @@ static void ioreq_reset(struct ioreq *ioreq)
     memset(&ioreq->req, 0, sizeof(ioreq->req));
     ioreq->status = 0;
     ioreq->start = 0;
+    ioreq->buf = NULL;
+    ioreq->size = 0;
     ioreq->presync = 0;
 
-    memset(ioreq->refs, 0, sizeof(ioreq->refs));
-    memset(ioreq->page, 0, sizeof(ioreq->page));
-    ioreq->pages = NULL;
-
     ioreq->aio_inflight = 0;
     ioreq->aio_errors = 0;
 
@@ -138,7 +133,7 @@ static struct ioreq *ioreq_start(struct XenBlkDev *blkdev)
         ioreq = g_malloc0(sizeof(*ioreq));
         ioreq->blkdev = blkdev;
         blkdev->requests_total++;
-        qemu_iovec_init(&ioreq->v, BLKIF_MAX_SEGMENTS_PER_REQUEST);
+        qemu_iovec_init(&ioreq->v, 1);
     } else {
         /* get one from freelist */
         ioreq = QLIST_FIRST(&blkdev->freelist);
@@ -183,7 +178,6 @@ static void ioreq_release(struct ioreq *ioreq, bool finish)
 static int ioreq_parse(struct ioreq *ioreq)
 {
     struct XenBlkDev *blkdev = ioreq->blkdev;
-    uintptr_t mem;
     size_t len;
     int i;
 
@@ -230,13 +224,10 @@ static int ioreq_parse(struct ioreq *ioreq)
             goto err;
         }
 
-        ioreq->refs[i]   = ioreq->req.seg[i].gref;
-
-        mem = ioreq->req.seg[i].first_sect * blkdev->file_blk;
         len = (ioreq->req.seg[i].last_sect - ioreq->req.seg[i].first_sect + 1) * blkdev->file_blk;
-        qemu_iovec_add(&ioreq->v, (void*)mem, len);
+        ioreq->size += len;
     }
-    if (ioreq->start + ioreq->v.size > blkdev->file_size) {
+    if (ioreq->start + ioreq->size > blkdev->file_size) {
         xen_pv_printf(&blkdev->xendev, 0, "error: access beyond end of file\n");
         goto err;
     }
@@ -247,35 +238,6 @@ err:
     return -1;
 }
 
-static void ioreq_free_copy_buffers(struct ioreq *ioreq)
-{
-    int i;
-
-    for (i = 0; i < ioreq->v.niov; i++) {
-        ioreq->page[i] = NULL;
-    }
-
-    qemu_vfree(ioreq->pages);
-}
-
-static int ioreq_init_copy_buffers(struct ioreq *ioreq)
-{
-    int i;
-
-    if (ioreq->v.niov == 0) {
-        return 0;
-    }
-
-    ioreq->pages = qemu_memalign(XC_PAGE_SIZE, ioreq->v.niov * XC_PAGE_SIZE);
-
-    for (i = 0; i < ioreq->v.niov; i++) {
-        ioreq->page[i] = ioreq->pages + i * XC_PAGE_SIZE;
-        ioreq->v.iov[i].iov_base = ioreq->page[i];
-    }
-
-    return 0;
-}
-
 static int ioreq_grant_copy(struct ioreq *ioreq)
 {
     struct XenBlkDev *blkdev = ioreq->blkdev;
@@ -284,25 +246,27 @@ static int ioreq_grant_copy(struct ioreq *ioreq)
     int i, count, rc;
     int64_t file_blk = ioreq->blkdev->file_blk;
     bool to_domain = (ioreq->req.operation == BLKIF_OP_READ);
+    void *virt = ioreq->buf;
 
-    if (ioreq->v.niov == 0) {
+    if (ioreq->req.nr_segments == 0) {
         return 0;
     }
 
-    count = ioreq->v.niov;
+    count = ioreq->req.nr_segments;
 
     for (i = 0; i < count; i++) {
         if (to_domain) {
-            segs[i].dest.foreign.ref = ioreq->refs[i];
+            segs[i].dest.foreign.ref = ioreq->req.seg[i].gref;
             segs[i].dest.foreign.offset = ioreq->req.seg[i].first_sect * file_blk;
-            segs[i].source.virt = ioreq->v.iov[i].iov_base;
+            segs[i].source.virt = virt;
         } else {
-            segs[i].source.foreign.ref = ioreq->refs[i];
+            segs[i].source.foreign.ref = ioreq->req.seg[i].gref;
             segs[i].source.foreign.offset = ioreq->req.seg[i].first_sect * file_blk;
-            segs[i].dest.virt = ioreq->v.iov[i].iov_base;
+            segs[i].dest.virt = virt;
         }
         segs[i].len = (ioreq->req.seg[i].last_sect
                        - ioreq->req.seg[i].first_sect + 1) * file_blk;
+        virt += segs[i].len;
     }
 
     rc = xen_be_copy_grant_refs(xendev, to_domain, segs, count);
@@ -348,14 +312,14 @@ static void qemu_aio_complete(void *opaque, int ret)
         if (ret == 0) {
             ioreq_grant_copy(ioreq);
         }
-        ioreq_free_copy_buffers(ioreq);
+        qemu_vfree(ioreq->buf);
         break;
     case BLKIF_OP_WRITE:
     case BLKIF_OP_FLUSH_DISKCACHE:
         if (!ioreq->req.nr_segments) {
             break;
         }
-        ioreq_free_copy_buffers(ioreq);
+        qemu_vfree(ioreq->buf);
         break;
     default:
         break;
@@ -423,12 +387,12 @@ static int ioreq_runio_qemu_aio(struct ioreq *ioreq)
 {
     struct XenBlkDev *blkdev = ioreq->blkdev;
 
-    ioreq_init_copy_buffers(ioreq);
+    ioreq->buf = qemu_memalign(XC_PAGE_SIZE, ioreq->size);
     if (ioreq->req.nr_segments &&
         (ioreq->req.operation == BLKIF_OP_WRITE ||
          ioreq->req.operation == BLKIF_OP_FLUSH_DISKCACHE) &&
         ioreq_grant_copy(ioreq)) {
-        ioreq_free_copy_buffers(ioreq);
+        qemu_vfree(ioreq->buf);
         goto err;
     }
 
@@ -440,6 +404,7 @@ static int ioreq_runio_qemu_aio(struct ioreq *ioreq)
 
     switch (ioreq->req.operation) {
     case BLKIF_OP_READ:
+        qemu_iovec_add(&ioreq->v, ioreq->buf, ioreq->size);
         block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct,
                          ioreq->v.size, BLOCK_ACCT_READ);
         ioreq->aio_inflight++;
@@ -452,6 +417,7 @@ static int ioreq_runio_qemu_aio(struct ioreq *ioreq)
             break;
         }
 
+        qemu_iovec_add(&ioreq->v, ioreq->buf, ioreq->size);
         block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct,
                          ioreq->v.size,
                          ioreq->req.operation == BLKIF_OP_WRITE ?
-- 
2.1.4