]> xenbits.xensource.com Git - qemu-xen.git/commitdiff
hw/block/nvme: fix zone boundary check for append
authorKlaus Jensen <k.jensen@samsung.com>
Tue, 19 Jan 2021 11:42:58 +0000 (12:42 +0100)
committerKlaus Jensen <k.jensen@samsung.com>
Mon, 8 Feb 2021 20:15:54 +0000 (21:15 +0100)
When a zone append is processed the controller checks that validity of
the write before assigning the LBA to the append command. This causes
the boundary check to be wrong.

Fix this by checking the write *after* assigning the LBA. Remove the
append special case from the nvme_check_zone_write and open code it in
nvme_do_write, assigning the slba when basic sanity checks have been
performed. Then check the validity of the resulting write like any other
write command.

In the process, also fix a missing endianness conversion for the zone
append ALBA.

Reported-by: Niklas Cassel <Niklas.Cassel@wdc.com>
Cc: Dmitry Fomichev <dmitry.fomichev@wdc.com>
Tested-by: Niklas Cassel <niklas.cassel@wdc.com>
Tested-by: Dmitry Fomichev <dmitry.fomichev@wdc.com>
Reviewed-by: Dmitry Fomichev <dmitry.fomichev@wdc.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Klaus Jensen <k.jensen@samsung.com>
hw/block/nvme.c

index e562d7467b3b3f417016d1dd08a33d3c37ddfcc8..cedb4ad9ffd3d5f07921fe5be44a8714ad8a4e7e 100644 (file)
@@ -1188,7 +1188,7 @@ static uint16_t nvme_check_zone_state_for_write(NvmeZone *zone)
 
 static uint16_t nvme_check_zone_write(NvmeCtrl *n, NvmeNamespace *ns,
                                       NvmeZone *zone, uint64_t slba,
-                                      uint32_t nlb, bool append)
+                                      uint32_t nlb)
 {
     uint16_t status;
 
@@ -1202,16 +1202,8 @@ static uint16_t nvme_check_zone_write(NvmeCtrl *n, NvmeNamespace *ns,
         trace_pci_nvme_err_zone_write_not_ok(slba, nlb, status);
     } else {
         assert(nvme_wp_is_valid(zone));
-        if (append) {
-            if (unlikely(slba != zone->d.zslba)) {
-                trace_pci_nvme_err_append_not_at_start(slba, zone->d.zslba);
-                status = NVME_INVALID_FIELD;
-            }
-            if (nvme_l2b(ns, nlb) > (n->page_size << n->zasl)) {
-                trace_pci_nvme_err_append_too_large(slba, nlb, n->zasl);
-                status = NVME_INVALID_FIELD;
-            }
-        } else if (unlikely(slba != zone->w_ptr)) {
+
+        if (unlikely(slba != zone->w_ptr)) {
             trace_pci_nvme_err_write_not_at_wp(slba, zone->d.zslba,
                                                zone->w_ptr);
             status = NVME_ZONE_INVALID_WRITE;
@@ -1349,10 +1341,9 @@ static void nvme_finalize_zoned_write(NvmeNamespace *ns, NvmeRequest *req,
     }
 }
 
-static uint64_t nvme_advance_zone_wp(NvmeNamespace *ns, NvmeZone *zone,
-                                     uint32_t nlb)
+static void nvme_advance_zone_wp(NvmeNamespace *ns, NvmeZone *zone,
+                                 uint32_t nlb)
 {
-    uint64_t result = zone->w_ptr;
     uint8_t zs;
 
     zone->w_ptr += nlb;
@@ -1368,8 +1359,6 @@ static uint64_t nvme_advance_zone_wp(NvmeNamespace *ns, NvmeZone *zone,
             nvme_assign_zone_state(ns, zone, NVME_ZONE_STATE_IMPLICITLY_OPEN);
         }
     }
-
-    return result;
 }
 
 static inline bool nvme_is_write(NvmeRequest *req)
@@ -1747,7 +1736,24 @@ static uint16_t nvme_do_write(NvmeCtrl *n, NvmeRequest *req, bool append,
     if (ns->params.zoned) {
         zone = nvme_get_zone_by_slba(ns, slba);
 
-        status = nvme_check_zone_write(n, ns, zone, slba, nlb, append);
+        if (append) {
+            if (unlikely(slba != zone->d.zslba)) {
+                trace_pci_nvme_err_append_not_at_start(slba, zone->d.zslba);
+                status = NVME_INVALID_FIELD;
+                goto invalid;
+            }
+
+            if (nvme_l2b(ns, nlb) > (n->page_size << n->zasl)) {
+                trace_pci_nvme_err_append_too_large(slba, nlb, n->zasl);
+                status = NVME_INVALID_FIELD;
+                goto invalid;
+            }
+
+            slba = zone->w_ptr;
+            res->slba = cpu_to_le64(slba);
+        }
+
+        status = nvme_check_zone_write(n, ns, zone, slba, nlb);
         if (status) {
             goto invalid;
         }
@@ -1757,11 +1763,7 @@ static uint16_t nvme_do_write(NvmeCtrl *n, NvmeRequest *req, bool append,
             goto invalid;
         }
 
-        if (append) {
-            slba = zone->w_ptr;
-        }
-
-        res->slba = nvme_advance_zone_wp(ns, zone, nlb);
+        nvme_advance_zone_wp(ns, zone, nlb);
     }
 
     data_offset = nvme_l2b(ns, slba);