Pull request

Userspace NVMe driver patches.
 -----BEGIN PGP SIGNATURE-----
 
 iQEzBAABCAAdFiEEhpWov9P5fNqsNXdanKSrs4Grc8gFAmE3H0UACgkQnKSrs4Gr
 c8johgf/R2gB82bGdgb4A3W1kQolwCGuOJ7IMq3vIuPUf9nps4G2QXM/0cAKmwW1
 B/72XA3VZGsRByaHEdhcGxOzSWblmdDnl5PTd31oAiIKYKiu7GI+UYkBQ/t0CMHq
 fVBb4utzFbtZu0fveF9t1ViWbQ3bB/W9S9UkeG9dVjGvTi9QqqyNf8e03FaW1alv
 rPSsr7Ks/cyTTTCttpuxkwv3dNAdYHHdglm9CR/9BkyznK+LOxEX2gn7YjORNZnb
 Od8xkPptdRkNGO8gJQoIwOdA4qotSW23Uy+57a7IIY0Zwi4G2YbBDRu8+yRP/6HU
 bf+xa1oM4i0ApOB1tbbq6JJtabp1Ww==
 =Jx3A
 -----END PGP SIGNATURE-----

Merge remote-tracking branch 'remotes/stefanha-gitlab/tags/block-pull-request' into staging

Pull request

Userspace NVMe driver patches.

# gpg: Signature made Tue 07 Sep 2021 09:13:57 BST
# gpg:                using RSA key 8695A8BFD3F97CDAAC35775A9CA4ABB381AB73C8
# gpg: Good signature from "Stefan Hajnoczi <stefanha@redhat.com>" [full]
# gpg:                 aka "Stefan Hajnoczi <stefanha@gmail.com>" [full]
# Primary key fingerprint: 8695 A8BF D3F9 7CDA AC35  775A 9CA4 ABB3 81AB 73C8

* remotes/stefanha-gitlab/tags/block-pull-request:
  block/nvme: Only report VFIO error on failed retry
  util/vfio-helpers: Let qemu_vfio_do_mapping() propagate Error
  util/vfio-helpers: Simplify qemu_vfio_dma_map() returning directly
  util/vfio-helpers: Use error_setg in qemu_vfio_find_[fixed/temp]_iova
  util/vfio-helpers: Extract qemu_vfio_water_mark_reached()
  util/vfio-helpers: Pass Error handle to qemu_vfio_dma_map()
  block/nvme: Have nvme_create_queue_pair() report errors consistently
  util/vfio-helpers: Remove unreachable code in qemu_vfio_dma_map()
  util/vfio-helpers: Replace qemu_mutex_lock() calls with QEMU_LOCK_GUARD
  util/vfio-helpers: Let qemu_vfio_verify_mappings() use error_report()
  block/nvme: Use safer trace format string

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>

block/nvme.c

@@ -176,12 +176,11 @@ static bool nvme_init_queue(BDRVNVMeState *s, NVMeQueue *q,
         return false;
     }
     memset(q->queue, 0, bytes);
-    r = qemu_vfio_dma_map(s->vfio, q->queue, bytes, false, &q->iova);
+    r = qemu_vfio_dma_map(s->vfio, q->queue, bytes, false, &q->iova, errp);
     if (r) {
-        error_setg(errp, "Cannot map queue");
+        error_prepend(errp, "Cannot map queue: ");
-        return false;
     }
-    return true;
+    return r == 0;
 }
 
 static void nvme_free_queue_pair(NVMeQueuePair *q)
@@ -220,6 +219,7 @@ static NVMeQueuePair *nvme_create_queue_pair(BDRVNVMeState *s,
 
     q = g_try_new0(NVMeQueuePair, 1);
     if (!q) {
+        error_setg(errp, "Cannot allocate queue pair");
         return NULL;
     }
     trace_nvme_create_queue_pair(idx, q, size, aio_context,
@@ -228,6 +228,7 @@ static NVMeQueuePair *nvme_create_queue_pair(BDRVNVMeState *s,
                           qemu_real_host_page_size);
     q->prp_list_pages = qemu_try_memalign(qemu_real_host_page_size, bytes);
     if (!q->prp_list_pages) {
+        error_setg(errp, "Cannot allocate PRP page list");
         goto fail;
     }
     memset(q->prp_list_pages, 0, bytes);
@@ -237,8 +238,9 @@ static NVMeQueuePair *nvme_create_queue_pair(BDRVNVMeState *s,
     qemu_co_queue_init(&q->free_req_queue);
     q->completion_bh = aio_bh_new(aio_context, nvme_process_completion_bh, q);
     r = qemu_vfio_dma_map(s->vfio, q->prp_list_pages, bytes,
-                          false, &prp_list_iova);
+                          false, &prp_list_iova, errp);
     if (r) {
+        error_prepend(errp, "Cannot map buffer for DMA: ");
         goto fail;
     }
     q->free_req_head = -1;
@@ -531,9 +533,9 @@ static bool nvme_identify(BlockDriverState *bs, int namespace, Error **errp)
         error_setg(errp, "Cannot allocate buffer for identify response");
         goto out;
     }
-    r = qemu_vfio_dma_map(s->vfio, id, id_size, true, &iova);
+    r = qemu_vfio_dma_map(s->vfio, id, id_size, true, &iova, errp);
     if (r) {
-        error_setg(errp, "Cannot map buffer for DMA");
+        error_prepend(errp, "Cannot map buffer for DMA: ");
         goto out;
     }
 
@@ -1017,6 +1019,7 @@ static coroutine_fn int nvme_cmd_map_qiov(BlockDriverState *bs, NvmeCmd *cmd,
     uint64_t *pagelist = req->prp_list_page;
     int i, j, r;
     int entries = 0;
+    Error *local_err = NULL, **errp = NULL;
 
     assert(qiov->size);
     assert(QEMU_IS_ALIGNED(qiov->size, s->page_size));
@@ -1029,7 +1032,7 @@ static coroutine_fn int nvme_cmd_map_qiov(BlockDriverState *bs, NvmeCmd *cmd,
 try_map:
         r = qemu_vfio_dma_map(s->vfio,
                               qiov->iov[i].iov_base,
-                              len, true, &iova);
+                              len, true, &iova, errp);
         if (r == -ENOSPC) {
             /*
              * In addition to the -ENOMEM error, the VFIO_IOMMU_MAP_DMA
@@ -1064,6 +1067,8 @@ try_map:
                     goto fail;
                 }
             }
+            errp = &local_err;
+
             goto try_map;
         }
         if (r) {
@@ -1107,6 +1112,9 @@ fail:
      * because they are already mapped before calling this function; for
      * temporary mappings, a later nvme_cmd_(un)map_qiov will reclaim by
      * calling qemu_vfio_dma_reset_temporary when necessary. */
+    if (local_err) {
+        error_reportf_err(local_err, "Cannot map buffer for DMA: ");
+    }
     return r;
 }
 
@@ -1521,14 +1529,15 @@ static void nvme_aio_unplug(BlockDriverState *bs)
 static void nvme_register_buf(BlockDriverState *bs, void *host, size_t size)
 {
     int ret;
+    Error *local_err = NULL;
     BDRVNVMeState *s = bs->opaque;
 
-    ret = qemu_vfio_dma_map(s->vfio, host, size, false, NULL);
+    ret = qemu_vfio_dma_map(s->vfio, host, size, false, NULL, &local_err);
     if (ret) {
         /* FIXME: we may run out of IOVA addresses after repeated
          * bdrv_register_buf/bdrv_unregister_buf, because nvme_vfio_dma_unmap
          * doesn't reclaim addresses for fixed mappings. */
-        error_report("nvme_register_buf failed: %s", strerror(-ret));
+        error_reportf_err(local_err, "nvme_register_buf failed: ");
     }
 }
 

block/trace-events

@@ -156,7 +156,7 @@ nvme_dsm(void *s, uint64_t offset, uint64_t bytes) "s %p offset 0x%"PRIx64" byte
 nvme_dsm_done(void *s, uint64_t offset, uint64_t bytes, int ret) "s %p offset 0x%"PRIx64" bytes %"PRId64" ret %d"
 nvme_dma_map_flush(void *s) "s %p"
 nvme_free_req_queue_wait(void *s, unsigned q_index) "s %p q #%u"
-nvme_create_queue_pair(unsigned q_index, void *q, unsigned size, void *aio_context, int fd) "index %u q %p size %u aioctx %p fd %d"
+nvme_create_queue_pair(unsigned q_index, void *q, size_t size, void *aio_context, int fd) "index %u q %p size %zu aioctx %p fd %d"
 nvme_free_queue_pair(unsigned q_index, void *q) "index %u q %p"
 nvme_cmd_map_qiov(void *s, void *cmd, void *req, void *qiov, int entries) "s %p cmd %p req %p qiov %p entries %d"
 nvme_cmd_map_qiov_pages(void *s, int i, uint64_t page) "s %p page[%d] 0x%"PRIx64

include/qemu/vfio-helpers.h

@@ -18,7 +18,7 @@ typedef struct QEMUVFIOState QEMUVFIOState;
 QEMUVFIOState *qemu_vfio_open_pci(const char *device, Error **errp);
 void qemu_vfio_close(QEMUVFIOState *s);
 int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
-                      bool temporary, uint64_t *iova_list);
+                      bool temporary, uint64_t *iova_list, Error **errp);
 int qemu_vfio_dma_reset_temporary(QEMUVFIOState *s);
 void qemu_vfio_dma_unmap(QEMUVFIOState *s, void *host);
 void *qemu_vfio_pci_map_bar(QEMUVFIOState *s, int index,

util/vfio-helpers.c

@@ -463,13 +463,15 @@ static void qemu_vfio_ram_block_added(RAMBlockNotifier *n, void *host,
                                       size_t size, size_t max_size)
 {
     QEMUVFIOState *s = container_of(n, QEMUVFIOState, ram_notifier);
+    Error *local_err = NULL;
     int ret;
 
     trace_qemu_vfio_ram_block_added(s, host, max_size);
-    ret = qemu_vfio_dma_map(s, host, max_size, false, NULL);
+    ret = qemu_vfio_dma_map(s, host, max_size, false, NULL, &local_err);
     if (ret) {
-        error_report("qemu_vfio_dma_map(%p, %zu) failed: %s", host, max_size,
+        error_reportf_err(local_err,
-                     strerror(-ret));
+                          "qemu_vfio_dma_map(%p, %zu) failed: ",
+                          host, max_size);
     }
 }
 
@@ -608,7 +610,7 @@ static IOVAMapping *qemu_vfio_add_mapping(QEMUVFIOState *s,
 
 /* Do the DMA mapping with VFIO. */
 static int qemu_vfio_do_mapping(QEMUVFIOState *s, void *host, size_t size,
-                                uint64_t iova)
+                                uint64_t iova, Error **errp)
 {
     struct vfio_iommu_type1_dma_map dma_map = {
         .argsz = sizeof(dma_map),
@@ -620,7 +622,7 @@ static int qemu_vfio_do_mapping(QEMUVFIOState *s, void *host, size_t size,
     trace_qemu_vfio_do_mapping(s, host, iova, size);
 
     if (ioctl(s->container, VFIO_IOMMU_MAP_DMA, &dma_map)) {
-        error_report("VFIO_MAP_DMA failed: %s", strerror(errno));
+        error_setg_errno(errp, errno, "VFIO_MAP_DMA failed");
         return -errno;
     }
     return 0;
@@ -660,13 +662,13 @@ static bool qemu_vfio_verify_mappings(QEMUVFIOState *s)
     if (QEMU_VFIO_DEBUG) {
         for (i = 0; i < s->nr_mappings - 1; ++i) {
             if (!(s->mappings[i].host < s->mappings[i + 1].host)) {
-                fprintf(stderr, "item %d not sorted!\n", i);
+                error_report("item %d not sorted!", i);
                 qemu_vfio_dump_mappings(s);
                 return false;
             }
             if (!(s->mappings[i].host + s->mappings[i].size <=
                   s->mappings[i + 1].host)) {
-                fprintf(stderr, "item %d overlap with next!\n", i);
+                error_report("item %d overlap with next!", i);
                 qemu_vfio_dump_mappings(s);
                 return false;
             }
@@ -675,8 +677,8 @@ static bool qemu_vfio_verify_mappings(QEMUVFIOState *s)
     return true;
 }
 
-static int
+static bool qemu_vfio_find_fixed_iova(QEMUVFIOState *s, size_t size,
-qemu_vfio_find_fixed_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
+                                      uint64_t *iova, Error **errp)
 {
     int i;
 
@@ -691,14 +693,16 @@ qemu_vfio_find_fixed_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
             s->usable_iova_ranges[i].end - s->low_water_mark + 1 == 0) {
             *iova = s->low_water_mark;
             s->low_water_mark += size;
-            return 0;
+            return true;
         }
     }
-    return -ENOMEM;
+    error_setg(errp, "fixed iova range not found");
+
+    return false;
 }
 
-static int
+static bool qemu_vfio_find_temp_iova(QEMUVFIOState *s, size_t size,
-qemu_vfio_find_temp_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
+                                     uint64_t *iova, Error **errp)
 {
     int i;
 
@@ -713,10 +717,27 @@ qemu_vfio_find_temp_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
             s->high_water_mark - s->usable_iova_ranges[i].start + 1 == 0) {
             *iova = s->high_water_mark - size;
             s->high_water_mark = *iova;
-            return 0;
+            return true;
         }
     }
-    return -ENOMEM;
+    error_setg(errp, "temporary iova range not found");
+
+    return false;
+}
+
+/**
+ * qemu_vfio_water_mark_reached:
+ *
+ * Returns %true if high watermark has been reached, %false otherwise.
+ */
+static bool qemu_vfio_water_mark_reached(QEMUVFIOState *s, size_t size,
+                                         Error **errp)
+{
+    if (s->high_water_mark - s->low_water_mark + 1 < size) {
+        error_setg(errp, "iova exhausted (water mark reached)");
+        return true;
+    }
+    return false;
 }
 
 /* Map [host, host + size) area into a contiguous IOVA address space, and store
@@ -725,9 +746,8 @@ qemu_vfio_find_temp_iova(QEMUVFIOState *s, size_t size, uint64_t *iova)
  * mapping status within this area is not allowed).
  */
 int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
-                      bool temporary, uint64_t *iova)
+                      bool temporary, uint64_t *iova, Error **errp)
 {
-    int ret = 0;
     int index;
     IOVAMapping *mapping;
     uint64_t iova0;
@@ -735,41 +755,36 @@ int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
     assert(QEMU_PTR_IS_ALIGNED(host, qemu_real_host_page_size));
     assert(QEMU_IS_ALIGNED(size, qemu_real_host_page_size));
     trace_qemu_vfio_dma_map(s, host, size, temporary, iova);
-    qemu_mutex_lock(&s->lock);
+    QEMU_LOCK_GUARD(&s->lock);
     mapping = qemu_vfio_find_mapping(s, host, &index);
     if (mapping) {
         iova0 = mapping->iova + ((uint8_t *)host - (uint8_t *)mapping->host);
     } else {
-        if (s->high_water_mark - s->low_water_mark + 1 < size) {
+        int ret;
-            ret = -ENOMEM;
+
-            goto out;
+        if (qemu_vfio_water_mark_reached(s, size, errp)) {
+            return -ENOMEM;
         }
         if (!temporary) {
-            if (qemu_vfio_find_fixed_iova(s, size, &iova0)) {
+            if (!qemu_vfio_find_fixed_iova(s, size, &iova0, errp)) {
-                ret = -ENOMEM;
+                return -ENOMEM;
-                goto out;
             }
 
             mapping = qemu_vfio_add_mapping(s, host, size, index + 1, iova0);
-            if (!mapping) {
-                ret = -ENOMEM;
-                goto out;
-            }
             assert(qemu_vfio_verify_mappings(s));
-            ret = qemu_vfio_do_mapping(s, host, size, iova0);
+            ret = qemu_vfio_do_mapping(s, host, size, iova0, errp);
-            if (ret) {
+            if (ret < 0) {
                 qemu_vfio_undo_mapping(s, mapping, NULL);
-                goto out;
+                return ret;
             }
             qemu_vfio_dump_mappings(s);
         } else {
-            if (qemu_vfio_find_temp_iova(s, size, &iova0)) {
+            if (!qemu_vfio_find_temp_iova(s, size, &iova0, errp)) {
-                ret = -ENOMEM;
+                return -ENOMEM;
-                goto out;
             }
-            ret = qemu_vfio_do_mapping(s, host, size, iova0);
+            ret = qemu_vfio_do_mapping(s, host, size, iova0, errp);
-            if (ret) {
+            if (ret < 0) {
-                goto out;
+                return ret;
             }
         }
     }
@@ -777,9 +792,7 @@ int qemu_vfio_dma_map(QEMUVFIOState *s, void *host, size_t size,
     if (iova) {
         *iova = iova0;
     }
-out:
+    return 0;
-    qemu_mutex_unlock(&s->lock);
-    return ret;
 }
 
 /* Reset the high watermark and free all "temporary" mappings. */
@@ -813,14 +826,12 @@ void qemu_vfio_dma_unmap(QEMUVFIOState *s, void *host)
     }
 
     trace_qemu_vfio_dma_unmap(s, host);
-    qemu_mutex_lock(&s->lock);
+    QEMU_LOCK_GUARD(&s->lock);
     m = qemu_vfio_find_mapping(s, host, &index);
     if (!m) {
-        goto out;
+        return;
     }
     qemu_vfio_undo_mapping(s, m, NULL);
-out:
-    qemu_mutex_unlock(&s->lock);
 }
 
 static void qemu_vfio_reset(QEMUVFIOState *s)