xen_disk: Remove ioreq.postsync

[qemu.git] / hw / block / xen_disk.c

diff --git a/hw/block/xen_disk.c b/hw/block/xen_disk.c
index 668cc069ff2075291282d99610bf1c4d9b3d4d5e..814665034d8cd144230e5632f57afe5cde148ae3 100644 (file)
--- a/hw/block/xen_disk.c
+++ b/hw/block/xen_disk.c
@@ -24,7 +24,6 @@
 #include <stdarg.h>
 #include <string.h>
 #include <unistd.h>
 #include <stdarg.h>
 #include <string.h>
 #include <unistd.h>

-#include <signal.h>

 #include <inttypes.h>
 #include <time.h>
 #include <fcntl.h>
 #include <inttypes.h>
 #include <time.h>
 #include <fcntl.h>


@@ -39,6 +38,9 @@
 #include "hw/xen/xen_backend.h"
 #include "xen_blkif.h"
 #include "sysemu/blockdev.h"
 #include "hw/xen/xen_backend.h"
 #include "xen_blkif.h"
 #include "sysemu/blockdev.h"

+#include "sysemu/block-backend.h"
+#include "qapi/qmp/qdict.h"
+#include "qapi/qmp/qstring.h"

 
 /* ------------------------------------------------------------- */
 
 
 /* ------------------------------------------------------------- */
 


@@ -58,6 +60,13 @@ struct PersistentGrant {
 
 typedef struct PersistentGrant PersistentGrant;
 
 
 typedef struct PersistentGrant PersistentGrant;
 

+struct PersistentRegion {
+    void *addr;
+    int num;
+};
+
+typedef struct PersistentRegion PersistentRegion;
+

 struct ioreq {
     blkif_request_t     req;
     int16_t             status;
 struct ioreq {
     blkif_request_t     req;
     int16_t             status;


@@ -66,7 +75,6 @@ struct ioreq {
     off_t               start;
     QEMUIOVector        v;
     int                 presync;
     off_t               start;
     QEMUIOVector        v;
     int                 presync;

-    int                 postsync;

     uint8_t             mapped;
 
     /* grant mapping */
     uint8_t             mapped;
 
     /* grant mapping */


@@ -114,14 +122,16 @@ struct XenBlkDev {
     int                 requests_finished;
 
     /* Persistent grants extension */
     int                 requests_finished;
 
     /* Persistent grants extension */

+    gboolean            feature_discard;

     gboolean            feature_persistent;
     GTree               *persistent_gnts;
     gboolean            feature_persistent;
     GTree               *persistent_gnts;

+    GSList              *persistent_regions;

     unsigned int        persistent_gnt_count;
     unsigned int        max_grants;
 
     /* qemu block driver */
     DriveInfo           *dinfo;
     unsigned int        persistent_gnt_count;
     unsigned int        max_grants;
 
     /* qemu block driver */
     DriveInfo           *dinfo;

-    BlockDriverState    *bs;
+    BlockBackend        *blk;

     QEMUBH              *bh;
 };
 
     QEMUBH              *bh;
 };
 


@@ -133,7 +143,6 @@ static void ioreq_reset(struct ioreq *ioreq)
     ioreq->status = 0;
     ioreq->start = 0;
     ioreq->presync = 0;
     ioreq->status = 0;
     ioreq->start = 0;
     ioreq->presync = 0;

-    ioreq->postsync = 0;

     ioreq->mapped = 0;
 
     memset(ioreq->domids, 0, sizeof(ioreq->domids));
     ioreq->mapped = 0;
 
     memset(ioreq->domids, 0, sizeof(ioreq->domids));


@@ -175,6 +184,23 @@ static void destroy_grant(gpointer pgnt)
     g_free(grant);
 }
 
     g_free(grant);
 }
 

+static void remove_persistent_region(gpointer data, gpointer dev)
+{
+    PersistentRegion *region = data;
+    struct XenBlkDev *blkdev = dev;
+    XenGnttab gnt = blkdev->xendev.gnttabdev;
+
+    if (xc_gnttab_munmap(gnt, region->addr, region->num) != 0) {
+        xen_be_printf(&blkdev->xendev, 0,
+                      "xc_gnttab_munmap region %p failed: %s\n",
+                      region->addr, strerror(errno));
+    }
+    xen_be_printf(&blkdev->xendev, 3,
+                  "unmapped grant region %p with %d pages\n",
+                  region->addr, region->num);
+    g_free(region);
+}
+

 static struct ioreq *ioreq_start(struct XenBlkDev *blkdev)
 {
     struct ioreq *ioreq = NULL;
 static struct ioreq *ioreq_start(struct XenBlkDev *blkdev)
 {
     struct ioreq *ioreq = NULL;


@@ -253,6 +279,8 @@ static int ioreq_parse(struct ioreq *ioreq)
     case BLKIF_OP_WRITE:
         ioreq->prot = PROT_READ; /* from memory */
         break;
     case BLKIF_OP_WRITE:
         ioreq->prot = PROT_READ; /* from memory */
         break;

+    case BLKIF_OP_DISCARD:
+        return 0;

     default:
         xen_be_printf(&blkdev->xendev, 0, "error: unknown operation (%d)\n",
                       ioreq->req.operation);
     default:
         xen_be_printf(&blkdev->xendev, 0, "error: unknown operation (%d)\n",
                       ioreq->req.operation);


@@ -339,6 +367,7 @@ static int ioreq_map(struct ioreq *ioreq)
     void *page[BLKIF_MAX_SEGMENTS_PER_REQUEST];
     int i, j, new_maps = 0;
     PersistentGrant *grant;
     void *page[BLKIF_MAX_SEGMENTS_PER_REQUEST];
     int i, j, new_maps = 0;
     PersistentGrant *grant;

+    PersistentRegion *region;

     /* domids and refs variables will contain the information necessary
      * to map the grants that are needed to fulfill this request.
      *
     /* domids and refs variables will contain the information necessary
      * to map the grants that are needed to fulfill this request.
      *


@@ -405,6 +434,7 @@ static int ioreq_map(struct ioreq *ioreq)
                 xen_be_printf(&ioreq->blkdev->xendev, 0,
                               "can't map grant ref %d (%s, %d maps)\n",
                               refs[i], strerror(errno), ioreq->blkdev->cnt_map);
                 xen_be_printf(&ioreq->blkdev->xendev, 0,
                               "can't map grant ref %d (%s, %d maps)\n",
                               refs[i], strerror(errno), ioreq->blkdev->cnt_map);

+                ioreq->mapped = 1;

                 ioreq_unmap(ioreq);
                 return -1;
             }
                 ioreq_unmap(ioreq);
                 return -1;
             }


@@ -416,7 +446,22 @@ static int ioreq_map(struct ioreq *ioreq)
             }
         }
     }
             }
         }
     }

-    if (ioreq->blkdev->feature_persistent) {
+    if (ioreq->blkdev->feature_persistent && new_maps != 0 &&
+        (!batch_maps || (ioreq->blkdev->persistent_gnt_count + new_maps <=
+        ioreq->blkdev->max_grants))) {
+        /*
+         * If we are using persistent grants and batch mappings only
+         * add the new maps to the list of persistent grants if the whole
+         * area can be persistently mapped.
+         */
+        if (batch_maps) {
+            region = g_malloc0(sizeof(*region));
+            region->addr = ioreq->pages;
+            region->num = new_maps;
+            ioreq->blkdev->persistent_regions = g_slist_append(
+                                            ioreq->blkdev->persistent_regions,
+                                            region);
+        }

         while ((ioreq->blkdev->persistent_gnt_count < ioreq->blkdev->max_grants)
               && new_maps) {
             /* Go through the list of newly mapped grants and add as many
         while ((ioreq->blkdev->persistent_gnt_count < ioreq->blkdev->max_grants)
               && new_maps) {
             /* Go through the list of newly mapped grants and add as many


@@ -442,6 +487,7 @@ static int ioreq_map(struct ioreq *ioreq)
                           grant);
             ioreq->blkdev->persistent_gnt_count++;
         }
                           grant);
             ioreq->blkdev->persistent_gnt_count++;
         }

+        assert(!batch_maps || new_maps == 0);

     }
     for (i = 0; i < ioreq->v.niov; i++) {
         ioreq->v.iov[i].iov_base += (uintptr_t)page[i];
     }
     for (i = 0; i < ioreq->v.niov; i++) {
         ioreq->v.iov[i].iov_base += (uintptr_t)page[i];


@@ -472,17 +518,27 @@ static void qemu_aio_complete(void *opaque, int ret)
     if (ioreq->aio_inflight > 0) {
         return;
     }
     if (ioreq->aio_inflight > 0) {
         return;
     }

-    if (ioreq->postsync) {
-        ioreq->postsync = 0;
-        ioreq->aio_inflight++;
-        bdrv_aio_flush(ioreq->blkdev->bs, qemu_aio_complete, ioreq);
-        return;
-    }

 
     ioreq->status = ioreq->aio_errors ? BLKIF_RSP_ERROR : BLKIF_RSP_OKAY;
     ioreq_unmap(ioreq);
     ioreq_finish(ioreq);
 
     ioreq->status = ioreq->aio_errors ? BLKIF_RSP_ERROR : BLKIF_RSP_OKAY;
     ioreq_unmap(ioreq);
     ioreq_finish(ioreq);

-    bdrv_acct_done(ioreq->blkdev->bs, &ioreq->acct);
+    switch (ioreq->req.operation) {
+    case BLKIF_OP_WRITE:
+    case BLKIF_OP_FLUSH_DISKCACHE:
+        if (!ioreq->req.nr_segments) {
+            break;
+        }
+    case BLKIF_OP_READ:
+        if (ioreq->status == BLKIF_RSP_OKAY) {
+            block_acct_done(blk_get_stats(ioreq->blkdev->blk), &ioreq->acct);
+        } else {
+            block_acct_failed(blk_get_stats(ioreq->blkdev->blk), &ioreq->acct);
+        }
+        break;
+    case BLKIF_OP_DISCARD:
+    default:
+        break;
+    }

     qemu_bh_schedule(ioreq->blkdev->bh);
 }
 
     qemu_bh_schedule(ioreq->blkdev->bh);
 }
 


@@ -496,17 +552,18 @@ static int ioreq_runio_qemu_aio(struct ioreq *ioreq)
 
     ioreq->aio_inflight++;
     if (ioreq->presync) {
 
     ioreq->aio_inflight++;
     if (ioreq->presync) {

-        bdrv_aio_flush(ioreq->blkdev->bs, qemu_aio_complete, ioreq);
+        blk_aio_flush(ioreq->blkdev->blk, qemu_aio_complete, ioreq);

         return 0;
     }
 
     switch (ioreq->req.operation) {
     case BLKIF_OP_READ:
         return 0;
     }
 
     switch (ioreq->req.operation) {
     case BLKIF_OP_READ:

-        bdrv_acct_start(blkdev->bs, &ioreq->acct, ioreq->v.size, BDRV_ACCT_READ);
+        block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct,
+                         ioreq->v.size, BLOCK_ACCT_READ);

         ioreq->aio_inflight++;
         ioreq->aio_inflight++;

-        bdrv_aio_readv(blkdev->bs, ioreq->start / BLOCK_SIZE,
-                       &ioreq->v, ioreq->v.size / BLOCK_SIZE,
-                       qemu_aio_complete, ioreq);
+        blk_aio_readv(blkdev->blk, ioreq->start / BLOCK_SIZE,
+                      &ioreq->v, ioreq->v.size / BLOCK_SIZE,
+                      qemu_aio_complete, ioreq);

         break;
     case BLKIF_OP_WRITE:
     case BLKIF_OP_FLUSH_DISKCACHE:
         break;
     case BLKIF_OP_WRITE:
     case BLKIF_OP_FLUSH_DISKCACHE:


@@ -514,12 +571,24 @@ static int ioreq_runio_qemu_aio(struct ioreq *ioreq)
             break;
         }
 
             break;
         }
 

-        bdrv_acct_start(blkdev->bs, &ioreq->acct, ioreq->v.size, BDRV_ACCT_WRITE);
+        block_acct_start(blk_get_stats(blkdev->blk), &ioreq->acct,
+                         ioreq->v.size,
+                         ioreq->req.operation == BLKIF_OP_WRITE ?
+                         BLOCK_ACCT_WRITE : BLOCK_ACCT_FLUSH);
+        ioreq->aio_inflight++;
+        blk_aio_writev(blkdev->blk, ioreq->start / BLOCK_SIZE,
+                       &ioreq->v, ioreq->v.size / BLOCK_SIZE,
+                       qemu_aio_complete, ioreq);
+        break;
+    case BLKIF_OP_DISCARD:
+    {
+        struct blkif_request_discard *discard_req = (void *)&ioreq->req;

         ioreq->aio_inflight++;
         ioreq->aio_inflight++;

-        bdrv_aio_writev(blkdev->bs, ioreq->start / BLOCK_SIZE,
-                        &ioreq->v, ioreq->v.size / BLOCK_SIZE,
+        blk_aio_discard(blkdev->blk,
+                        discard_req->sector_number, discard_req->nr_sectors,

                         qemu_aio_complete, ioreq);
         break;
                         qemu_aio_complete, ioreq);
         break;

+    }

     default:
         /* unknown operation (shouldn't happen -- parse catches this) */
         goto err;
     default:
         /* unknown operation (shouldn't happen -- parse catches this) */
         goto err;


@@ -564,6 +633,7 @@ static int blk_send_response_one(struct ioreq *ioreq)
         break;
     default:
         dst = NULL;
         break;
     default:
         dst = NULL;

+        return 0;

     }
     memcpy(dst, &resp, sizeof(resp));
     blkdev->rings.common.rsp_prod_pvt++;
     }
     memcpy(dst, &resp, sizeof(resp));
     blkdev->rings.common.rsp_prod_pvt++;


@@ -648,6 +718,23 @@ static void blk_handle_requests(struct XenBlkDev *blkdev)
 
         /* parse them */
         if (ioreq_parse(ioreq) != 0) {
 
         /* parse them */
         if (ioreq_parse(ioreq) != 0) {

+
+            switch (ioreq->req.operation) {
+            case BLKIF_OP_READ:
+                block_acct_invalid(blk_get_stats(blkdev->blk),
+                                   BLOCK_ACCT_READ);
+                break;
+            case BLKIF_OP_WRITE:
+                block_acct_invalid(blk_get_stats(blkdev->blk),
+                                   BLOCK_ACCT_WRITE);
+                break;
+            case BLKIF_OP_FLUSH_DISKCACHE:
+                block_acct_invalid(blk_get_stats(blkdev->blk),
+                                   BLOCK_ACCT_FLUSH);
+            default:
+                break;
+            };
+

             if (blk_send_response_one(ioreq)) {
                 xen_be_send_notify(&blkdev->xendev);
             }
             if (blk_send_response_one(ioreq)) {
                 xen_be_send_notify(&blkdev->xendev);
             }


@@ -698,6 +785,21 @@ static void blk_alloc(struct XenDevice *xendev)
     }
 }
 
     }
 }
 

+static void blk_parse_discard(struct XenBlkDev *blkdev)
+{
+    int enable;
+
+    blkdev->feature_discard = true;
+
+    if (xenstore_read_be_int(&blkdev->xendev, "discard-enable", &enable) == 0) {
+        blkdev->feature_discard = !!enable;
+    }
+
+    if (blkdev->feature_discard) {
+        xenstore_write_be_int(&blkdev->xendev, "feature-discard", 1);
+    }
+}
+

 static int blk_init(struct XenDevice *xendev)
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
 static int blk_init(struct XenDevice *xendev)
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);


@@ -765,6 +867,8 @@ static int blk_init(struct XenDevice *xendev)
     xenstore_write_be_int(&blkdev->xendev, "feature-persistent", 1);
     xenstore_write_be_int(&blkdev->xendev, "info", info);
 
     xenstore_write_be_int(&blkdev->xendev, "feature-persistent", 1);
     xenstore_write_be_int(&blkdev->xendev, "info", info);
 

+    blk_parse_discard(blkdev);
+

     g_free(directiosafe);
     return 0;
 
     g_free(directiosafe);
     return 0;
 


@@ -800,40 +904,53 @@ static int blk_connect(struct XenDevice *xendev)
         qflags |= BDRV_O_RDWR;
         readonly = false;
     }
         qflags |= BDRV_O_RDWR;
         readonly = false;
     }

+    if (blkdev->feature_discard) {
+        qflags |= BDRV_O_UNMAP;
+    }

 
     /* init qemu block driver */
     index = (blkdev->xendev.dev - 202 * 256) / 16;
     blkdev->dinfo = drive_get(IF_XEN, 0, index);
     if (!blkdev->dinfo) {
 
     /* init qemu block driver */
     index = (blkdev->xendev.dev - 202 * 256) / 16;
     blkdev->dinfo = drive_get(IF_XEN, 0, index);
     if (!blkdev->dinfo) {

+        Error *local_err = NULL;
+        QDict *options = NULL;
+
+        if (strcmp(blkdev->fileproto, "<unset>")) {
+            options = qdict_new();
+            qdict_put(options, "driver", qstring_from_str(blkdev->fileproto));
+        }
+

         /* setup via xenbus -> create new block driver instance */
         xen_be_printf(&blkdev->xendev, 2, "create new bdrv (xenbus setup)\n");
         /* setup via xenbus -> create new block driver instance */
         xen_be_printf(&blkdev->xendev, 2, "create new bdrv (xenbus setup)\n");

-        blkdev->bs = bdrv_new(blkdev->dev);
-        if (blkdev->bs) {
-            BlockDriver *drv = bdrv_find_whitelisted_format(blkdev->fileproto,
-                                                           readonly);
-            if (bdrv_open(blkdev->bs,
-                          blkdev->filename, NULL, qflags, drv) != 0) {
-                bdrv_unref(blkdev->bs);
-                blkdev->bs = NULL;
-            }
-        }
-        if (!blkdev->bs) {
+        blkdev->blk = blk_new_open(blkdev->dev, blkdev->filename, NULL, options,
+                                   qflags, &local_err);
+        if (!blkdev->blk) {
+            xen_be_printf(&blkdev->xendev, 0, "error: %s\n",
+                          error_get_pretty(local_err));
+            error_free(local_err);

             return -1;
         }
     } else {
         /* setup via qemu cmdline -> already setup for us */
         xen_be_printf(&blkdev->xendev, 2, "get configured bdrv (cmdline setup)\n");
             return -1;
         }
     } else {
         /* setup via qemu cmdline -> already setup for us */
         xen_be_printf(&blkdev->xendev, 2, "get configured bdrv (cmdline setup)\n");

-        blkdev->bs = blkdev->dinfo->bdrv;
-        /* blkdev->bs is not create by us, we get a reference
-         * so we can bdrv_unref() unconditionally */
-        bdrv_ref(blkdev->bs);
+        blkdev->blk = blk_by_legacy_dinfo(blkdev->dinfo);
+        if (blk_is_read_only(blkdev->blk) && !readonly) {
+            xen_be_printf(&blkdev->xendev, 0, "Unexpected read-only drive");
+            blkdev->blk = NULL;
+            return -1;
+        }
+        /* blkdev->blk is not create by us, we get a reference
+         * so we can blk_unref() unconditionally */
+        blk_ref(blkdev->blk);

     }
     }

-    bdrv_attach_dev_nofail(blkdev->bs, blkdev);
-    blkdev->file_size = bdrv_getlength(blkdev->bs);
+    blk_attach_dev_nofail(blkdev->blk, blkdev);
+    blkdev->file_size = blk_getlength(blkdev->blk);

     if (blkdev->file_size < 0) {
     if (blkdev->file_size < 0) {

-        xen_be_printf(&blkdev->xendev, 1, "bdrv_getlength: %d (%s) | drv %s\n",
+        BlockDriverState *bs = blk_bs(blkdev->blk);
+        const char *drv_name = bs ? bdrv_get_format_name(bs) : NULL;
+        xen_be_printf(&blkdev->xendev, 1, "blk_getlength: %d (%s) | drv %s\n",

                       (int)blkdev->file_size, strerror(-blkdev->file_size),
                       (int)blkdev->file_size, strerror(-blkdev->file_size),

-                      bdrv_get_format_name(blkdev->bs) ?: "-");
+                      drv_name ?: "-");

         blkdev->file_size = 0;
     }
 
         blkdev->file_size = 0;
     }
 


@@ -907,7 +1024,10 @@ static int blk_connect(struct XenDevice *xendev)
         blkdev->max_grants = max_requests * BLKIF_MAX_SEGMENTS_PER_REQUEST;
         blkdev->persistent_gnts = g_tree_new_full((GCompareDataFunc)int_cmp,
                                              NULL, NULL,
         blkdev->max_grants = max_requests * BLKIF_MAX_SEGMENTS_PER_REQUEST;
         blkdev->persistent_gnts = g_tree_new_full((GCompareDataFunc)int_cmp,
                                              NULL, NULL,

+                                             batch_maps ?
+                                             (GDestroyNotify)g_free :

                                              (GDestroyNotify)destroy_grant);
                                              (GDestroyNotify)destroy_grant);

+        blkdev->persistent_regions = NULL;

         blkdev->persistent_gnt_count = 0;
     }
 
         blkdev->persistent_gnt_count = 0;
     }
 


@@ -924,10 +1044,10 @@ static void blk_disconnect(struct XenDevice *xendev)
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
 
 {
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
 

-    if (blkdev->bs) {
-        bdrv_detach_dev(blkdev->bs, blkdev);
-        bdrv_unref(blkdev->bs);
-        blkdev->bs = NULL;
+    if (blkdev->blk) {
+        blk_detach_dev(blkdev->blk, blkdev);
+        blk_unref(blkdev->blk);
+        blkdev->blk = NULL;

     }
     xen_be_unbind_evtchn(&blkdev->xendev);
 
     }
     xen_be_unbind_evtchn(&blkdev->xendev);
 


@@ -936,6 +1056,26 @@ static void blk_disconnect(struct XenDevice *xendev)
         blkdev->cnt_map--;
         blkdev->sring = NULL;
     }
         blkdev->cnt_map--;
         blkdev->sring = NULL;
     }

+
+    /*
+     * Unmap persistent grants before switching to the closed state
+     * so the frontend can free them.
+     *
+     * In the !batch_maps case g_tree_destroy will take care of unmapping
+     * the grant, but in the batch_maps case we need to iterate over every
+     * region in persistent_regions and unmap it.
+     */
+    if (blkdev->feature_persistent) {
+        g_tree_destroy(blkdev->persistent_gnts);
+        assert(batch_maps || blkdev->persistent_gnt_count == 0);
+        if (batch_maps) {
+            blkdev->persistent_gnt_count = 0;
+            g_slist_foreach(blkdev->persistent_regions,
+                            (GFunc)remove_persistent_region, blkdev);
+            g_slist_free(blkdev->persistent_regions);
+        }
+        blkdev->feature_persistent = false;
+    }

 }
 
 static int blk_free(struct XenDevice *xendev)
 }
 
 static int blk_free(struct XenDevice *xendev)


@@ -943,15 +1083,10 @@ static int blk_free(struct XenDevice *xendev)
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
     struct ioreq *ioreq;
 
     struct XenBlkDev *blkdev = container_of(xendev, struct XenBlkDev, xendev);
     struct ioreq *ioreq;
 

-    if (blkdev->bs || blkdev->sring) {
+    if (blkdev->blk || blkdev->sring) {

         blk_disconnect(xendev);
     }
 
         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);
     while (!QLIST_EMPTY(&blkdev->freelist)) {
         ioreq = QLIST_FIRST(&blkdev->freelist);
         QLIST_REMOVE(ioreq, list);