#include <sys/mman.h>
#include <sys/uio.h>
-#include "hw.h"
-#include "xen_backend.h"
-#include "xen_blkif.h"
+#include "hw/hw.h"
+#include "hw/xen/xen_backend.h"
+#include "hw/xen_blkif.h"
#include "sysemu/blockdev.h"
/* ------------------------------------------------------------- */
#define BLOCK_SIZE 512
#define IOCB_COUNT (BLKIF_MAX_SEGMENTS_PER_REQUEST + 2)
+struct PersistentGrant {
+ void *page;
+ struct XenBlkDev *blkdev;
+};
+
+typedef struct PersistentGrant PersistentGrant;
+
struct ioreq {
blkif_request_t req;
int16_t status;
int prot;
void *page[BLKIF_MAX_SEGMENTS_PER_REQUEST];
void *pages;
+ int num_unmap;
/* aio status */
int aio_inflight;
int requests_inflight;
int requests_finished;
+ /* Persistent grants extension */
+ gboolean feature_persistent;
+ GTree *persistent_gnts;
+ unsigned int persistent_gnt_count;
+ unsigned int max_grants;
+
/* qemu block driver */
DriveInfo *dinfo;
BlockDriverState *bs;
/* ------------------------------------------------------------- */
+static void ioreq_reset(struct ioreq *ioreq)
+{
+ memset(&ioreq->req, 0, sizeof(ioreq->req));
+ ioreq->status = 0;
+ ioreq->start = 0;
+ ioreq->presync = 0;
+ ioreq->postsync = 0;
+ ioreq->mapped = 0;
+
+ memset(ioreq->domids, 0, sizeof(ioreq->domids));
+ memset(ioreq->refs, 0, sizeof(ioreq->refs));
+ ioreq->prot = 0;
+ memset(ioreq->page, 0, sizeof(ioreq->page));
+ ioreq->pages = NULL;
+
+ ioreq->aio_inflight = 0;
+ ioreq->aio_errors = 0;
+
+ ioreq->blkdev = NULL;
+ memset(&ioreq->list, 0, sizeof(ioreq->list));
+ memset(&ioreq->acct, 0, sizeof(ioreq->acct));
+
+ qemu_iovec_reset(&ioreq->v);
+}
+
+static gint int_cmp(gconstpointer a, gconstpointer b, gpointer user_data)
+{
+ uint ua = GPOINTER_TO_UINT(a);
+ uint ub = GPOINTER_TO_UINT(b);
+ return (ua > ub) - (ua < ub);
+}
+
+static void destroy_grant(gpointer pgnt)
+{
+ PersistentGrant *grant = pgnt;
+ XenGnttab gnt = grant->blkdev->xendev.gnttabdev;
+
+ if (xc_gnttab_munmap(gnt, grant->page, 1) != 0) {
+ xen_be_printf(&grant->blkdev->xendev, 0,
+ "xc_gnttab_munmap failed: %s\n",
+ strerror(errno));
+ }
+ grant->blkdev->persistent_gnt_count--;
+ xen_be_printf(&grant->blkdev->xendev, 3,
+ "unmapped grant %p\n", grant->page);
+ g_free(grant);
+}
+
static struct ioreq *ioreq_start(struct XenBlkDev *blkdev)
{
struct ioreq *ioreq = NULL;
/* get one from freelist */
ioreq = QLIST_FIRST(&blkdev->freelist);
QLIST_REMOVE(ioreq, list);
- qemu_iovec_reset(&ioreq->v);
}
QLIST_INSERT_HEAD(&blkdev->inflight, ioreq, list);
blkdev->requests_inflight++;
struct XenBlkDev *blkdev = ioreq->blkdev;
QLIST_REMOVE(ioreq, list);
- memset(ioreq, 0, sizeof(*ioreq));
+ ioreq_reset(ioreq);
ioreq->blkdev = blkdev;
QLIST_INSERT_HEAD(&blkdev->freelist, ioreq, list);
if (finish) {
case BLKIF_OP_READ:
ioreq->prot = PROT_WRITE; /* to memory */
break;
- case BLKIF_OP_WRITE_BARRIER:
+ case BLKIF_OP_FLUSH_DISKCACHE:
+ ioreq->presync = 1;
if (!ioreq->req.nr_segments) {
- ioreq->presync = 1;
return 0;
}
- ioreq->presync = ioreq->postsync = 1;
/* fall through */
case BLKIF_OP_WRITE:
ioreq->prot = PROT_READ; /* from memory */
XenGnttab gnt = ioreq->blkdev->xendev.gnttabdev;
int i;
- if (ioreq->v.niov == 0 || ioreq->mapped == 0) {
+ if (ioreq->num_unmap == 0 || ioreq->mapped == 0) {
return;
}
if (batch_maps) {
if (!ioreq->pages) {
return;
}
- if (xc_gnttab_munmap(gnt, ioreq->pages, ioreq->v.niov) != 0) {
+ if (xc_gnttab_munmap(gnt, ioreq->pages, ioreq->num_unmap) != 0) {
xen_be_printf(&ioreq->blkdev->xendev, 0, "xc_gnttab_munmap failed: %s\n",
strerror(errno));
}
- ioreq->blkdev->cnt_map -= ioreq->v.niov;
+ ioreq->blkdev->cnt_map -= ioreq->num_unmap;
ioreq->pages = NULL;
} else {
- for (i = 0; i < ioreq->v.niov; i++) {
+ for (i = 0; i < ioreq->num_unmap; i++) {
if (!ioreq->page[i]) {
continue;
}
static int ioreq_map(struct ioreq *ioreq)
{
XenGnttab gnt = ioreq->blkdev->xendev.gnttabdev;
- int i;
+ uint32_t domids[BLKIF_MAX_SEGMENTS_PER_REQUEST];
+ uint32_t refs[BLKIF_MAX_SEGMENTS_PER_REQUEST];
+ void *page[BLKIF_MAX_SEGMENTS_PER_REQUEST];
+ int i, j, new_maps = 0;
+ PersistentGrant *grant;
+ /* domids and refs variables will contain the information necessary
+ * to map the grants that are needed to fulfill this request.
+ *
+ * After mapping the needed grants, the page array will contain the
+ * memory address of each granted page in the order specified in ioreq
+ * (disregarding if it's a persistent grant or not).
+ */
if (ioreq->v.niov == 0 || ioreq->mapped == 1) {
return 0;
}
- if (batch_maps) {
+ if (ioreq->blkdev->feature_persistent) {
+ for (i = 0; i < ioreq->v.niov; i++) {
+ grant = g_tree_lookup(ioreq->blkdev->persistent_gnts,
+ GUINT_TO_POINTER(ioreq->refs[i]));
+
+ if (grant != NULL) {
+ page[i] = grant->page;
+ xen_be_printf(&ioreq->blkdev->xendev, 3,
+ "using persistent-grant %" PRIu32 "\n",
+ ioreq->refs[i]);
+ } else {
+ /* Add the grant to the list of grants that
+ * should be mapped
+ */
+ domids[new_maps] = ioreq->domids[i];
+ refs[new_maps] = ioreq->refs[i];
+ page[i] = NULL;
+ new_maps++;
+ }
+ }
+ /* Set the protection to RW, since grants may be reused later
+ * with a different protection than the one needed for this request
+ */
+ ioreq->prot = PROT_WRITE | PROT_READ;
+ } else {
+ /* All grants in the request should be mapped */
+ memcpy(refs, ioreq->refs, sizeof(refs));
+ memcpy(domids, ioreq->domids, sizeof(domids));
+ memset(page, 0, sizeof(page));
+ new_maps = ioreq->v.niov;
+ }
+
+ if (batch_maps && new_maps) {
ioreq->pages = xc_gnttab_map_grant_refs
- (gnt, ioreq->v.niov, ioreq->domids, ioreq->refs, ioreq->prot);
+ (gnt, new_maps, domids, refs, ioreq->prot);
if (ioreq->pages == NULL) {
xen_be_printf(&ioreq->blkdev->xendev, 0,
"can't map %d grant refs (%s, %d maps)\n",
- ioreq->v.niov, strerror(errno), ioreq->blkdev->cnt_map);
+ new_maps, strerror(errno), ioreq->blkdev->cnt_map);
return -1;
}
- for (i = 0; i < ioreq->v.niov; i++) {
- ioreq->v.iov[i].iov_base = ioreq->pages + i * XC_PAGE_SIZE +
- (uintptr_t)ioreq->v.iov[i].iov_base;
+ for (i = 0, j = 0; i < ioreq->v.niov; i++) {
+ if (page[i] == NULL) {
+ page[i] = ioreq->pages + (j++) * XC_PAGE_SIZE;
+ }
}
- ioreq->blkdev->cnt_map += ioreq->v.niov;
- } else {
- for (i = 0; i < ioreq->v.niov; i++) {
+ ioreq->blkdev->cnt_map += new_maps;
+ } else if (new_maps) {
+ for (i = 0; i < new_maps; i++) {
ioreq->page[i] = xc_gnttab_map_grant_ref
- (gnt, ioreq->domids[i], ioreq->refs[i], ioreq->prot);
+ (gnt, domids[i], refs[i], ioreq->prot);
if (ioreq->page[i] == NULL) {
xen_be_printf(&ioreq->blkdev->xendev, 0,
"can't map grant ref %d (%s, %d maps)\n",
- ioreq->refs[i], strerror(errno), ioreq->blkdev->cnt_map);
+ refs[i], strerror(errno), ioreq->blkdev->cnt_map);
ioreq_unmap(ioreq);
return -1;
}
- ioreq->v.iov[i].iov_base = ioreq->page[i] + (uintptr_t)ioreq->v.iov[i].iov_base;
ioreq->blkdev->cnt_map++;
}
+ for (i = 0, j = 0; i < ioreq->v.niov; i++) {
+ if (page[i] == NULL) {
+ page[i] = ioreq->page[j++];
+ }
+ }
+ }
+ if (ioreq->blkdev->feature_persistent) {
+ while ((ioreq->blkdev->persistent_gnt_count < ioreq->blkdev->max_grants)
+ && new_maps) {
+ /* Go through the list of newly mapped grants and add as many
+ * as possible to the list of persistently mapped grants.
+ *
+ * Since we start at the end of ioreq->page(s), we only need
+ * to decrease new_maps to prevent this granted pages from
+ * being unmapped in ioreq_unmap.
+ */
+ grant = g_malloc0(sizeof(*grant));
+ new_maps--;
+ if (batch_maps) {
+ grant->page = ioreq->pages + (new_maps) * XC_PAGE_SIZE;
+ } else {
+ grant->page = ioreq->page[new_maps];
+ }
+ grant->blkdev = ioreq->blkdev;
+ xen_be_printf(&ioreq->blkdev->xendev, 3,
+ "adding grant %" PRIu32 " page: %p\n",
+ refs[new_maps], grant->page);
+ g_tree_insert(ioreq->blkdev->persistent_gnts,
+ GUINT_TO_POINTER(refs[new_maps]),
+ grant);
+ ioreq->blkdev->persistent_gnt_count++;
+ }
+ }
+ for (i = 0; i < ioreq->v.niov; i++) {
+ ioreq->v.iov[i].iov_base += (uintptr_t)page[i];
}
ioreq->mapped = 1;
+ ioreq->num_unmap = new_maps;
return 0;
}
qemu_aio_complete, ioreq);
break;
case BLKIF_OP_WRITE:
- case BLKIF_OP_WRITE_BARRIER:
+ case BLKIF_OP_FLUSH_DISKCACHE:
if (!ioreq->req.nr_segments) {
break;
}
static int blk_init(struct XenDevice *xendev)
{
struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
- int index, qflags, info = 0;
+ int info = 0;
/* read xenstore entries */
if (blkdev->params == NULL) {
}
/* read-only ? */
- qflags = BDRV_O_NOCACHE | BDRV_O_CACHE_WB | BDRV_O_NATIVE_AIO;
- if (strcmp(blkdev->mode, "w") == 0) {
- qflags |= BDRV_O_RDWR;
- } else {
+ if (strcmp(blkdev->mode, "w")) {
info |= VDISK_READONLY;
}
info |= VDISK_CDROM;
}
+ blkdev->file_blk = BLOCK_SIZE;
+
+ /* fill info
+ * blk_connect supplies sector-size and sectors
+ */
+ xenstore_write_be_int(&blkdev->xendev, "feature-flush-cache", 1);
+ xenstore_write_be_int(&blkdev->xendev, "feature-persistent", 1);
+ xenstore_write_be_int(&blkdev->xendev, "info", info);
+ return 0;
+
+out_error:
+ g_free(blkdev->params);
+ blkdev->params = NULL;
+ g_free(blkdev->mode);
+ blkdev->mode = NULL;
+ g_free(blkdev->type);
+ blkdev->type = NULL;
+ g_free(blkdev->dev);
+ blkdev->dev = NULL;
+ g_free(blkdev->devtype);
+ blkdev->devtype = NULL;
+ return -1;
+}
+
+static int blk_connect(struct XenDevice *xendev)
+{
+ struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
+ int pers, index, qflags;
+
+ /* read-only ? */
+ qflags = BDRV_O_CACHE_WB | BDRV_O_NATIVE_AIO;
+ if (strcmp(blkdev->mode, "w") == 0) {
+ qflags |= BDRV_O_RDWR;
+ }
+
/* init qemu block driver */
index = (blkdev->xendev.dev - 202 * 256) / 16;
blkdev->dinfo = drive_get(IF_XEN, 0, index);
xen_be_printf(&blkdev->xendev, 2, "create new bdrv (xenbus setup)\n");
blkdev->bs = bdrv_new(blkdev->dev);
if (blkdev->bs) {
- if (bdrv_open(blkdev->bs, blkdev->filename, qflags,
+ if (bdrv_open(blkdev->bs, blkdev->filename, NULL, qflags,
bdrv_find_whitelisted_format(blkdev->fileproto)) != 0) {
bdrv_delete(blkdev->bs);
blkdev->bs = NULL;
}
}
if (!blkdev->bs) {
- goto out_error;
+ return -1;
}
} else {
/* setup via qemu cmdline -> already setup for us */
blkdev->bs = blkdev->dinfo->bdrv;
}
bdrv_attach_dev_nofail(blkdev->bs, blkdev);
- blkdev->file_blk = BLOCK_SIZE;
blkdev->file_size = bdrv_getlength(blkdev->bs);
if (blkdev->file_size < 0) {
xen_be_printf(&blkdev->xendev, 1, "bdrv_getlength: %d (%s) | drv %s\n",
blkdev->type, blkdev->fileproto, blkdev->filename,
blkdev->file_size, blkdev->file_size >> 20);
- /* fill info */
- xenstore_write_be_int(&blkdev->xendev, "feature-barrier", 1);
- xenstore_write_be_int(&blkdev->xendev, "info", info);
- xenstore_write_be_int(&blkdev->xendev, "sector-size", blkdev->file_blk);
- xenstore_write_be_int(&blkdev->xendev, "sectors",
- blkdev->file_size / blkdev->file_blk);
- return 0;
-
-out_error:
- g_free(blkdev->params);
- blkdev->params = NULL;
- g_free(blkdev->mode);
- blkdev->mode = NULL;
- g_free(blkdev->type);
- blkdev->type = NULL;
- g_free(blkdev->dev);
- blkdev->dev = NULL;
- g_free(blkdev->devtype);
- blkdev->devtype = NULL;
- return -1;
-}
-
-static int blk_connect(struct XenDevice *xendev)
-{
- struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
+ /* Fill in number of sector size and number of sectors */
+ xenstore_write_be_int(&blkdev->xendev, "sector-size", blkdev->file_blk);
+ xenstore_write_be_int64(&blkdev->xendev, "sectors",
+ blkdev->file_size / blkdev->file_blk);
if (xenstore_read_fe_int(&blkdev->xendev, "ring-ref", &blkdev->ring_ref) == -1) {
return -1;
&blkdev->xendev.remote_port) == -1) {
return -1;
}
+ if (xenstore_read_fe_int(&blkdev->xendev, "feature-persistent", &pers)) {
+ blkdev->feature_persistent = FALSE;
+ } else {
+ blkdev->feature_persistent = !!pers;
+ }
blkdev->protocol = BLKIF_PROTOCOL_NATIVE;
if (blkdev->xendev.protocol) {
}
}
+ if (blkdev->feature_persistent) {
+ /* Init persistent grants */
+ blkdev->max_grants = max_requests * BLKIF_MAX_SEGMENTS_PER_REQUEST;
+ blkdev->persistent_gnts = g_tree_new_full((GCompareDataFunc)int_cmp,
+ NULL, NULL,
+ (GDestroyNotify)destroy_grant);
+ blkdev->persistent_gnt_count = 0;
+ }
+
xen_be_bind_evtchn(&blkdev->xendev);
xen_be_printf(&blkdev->xendev, 1, "ok: proto %s, ring-ref %d, "
blk_disconnect(xendev);
}
+ /* Free persistent grants */
+ if (blkdev->feature_persistent) {
+ g_tree_destroy(blkdev->persistent_gnts);
+ }
+
while (!QLIST_EMPTY(&blkdev->freelist)) {
ioreq = QLIST_FIRST(&blkdev->freelist);
QLIST_REMOVE(ioreq, list);
projects / qemu.git / blobdiff