Exporting NVMe-oF devices read-only enforced on export from the NVMe-oF target.

* Exporting NVMe-oF devices read-only enforced on export from the NVMe-oF target.
@ 2020-09-29 11:34 Mark Ruijter
  2020-09-30  0:28 ` Chaitanya Kulkarni
  0 siblings, 1 reply; 6+ messages in thread
From: Mark Ruijter @ 2020-09-29 11:34 UTC (permalink / raw)
  To: Chaitanya Kulkarni; +Cc: linux-nvme

[-- Attachment #1: Type: text/plain, Size: 645 bytes --]


Hi all,

Unless I am mistaken the NVMe-oF target driver does currently not support to export a device read-only from the target system.
For exporting readonly snapshot devices I am currently using the attached patch.

Does this change, or a similar change, make sense for upstream?

    --
    With the patch and after setting the namespace to readonly:
    namespaces/1 # echo 1 > readonly
    namespaces/1 # echo 1 > enable
    
On the initiator system writes are now no longer permitted:
    root@r11i3:~# echo hello >/dev/nvme5n1 
    -bash: echo: write error: Operation not permitted
    --

Thanks,

Mark Ruijter



[-- Attachment #2: readonly.patch --]
[-- Type: application/octet-stream, Size: 1305 bytes --]

--- drivers/nvme/target/configfs.c-orig	2020-08-21 05:50:14.369462419 -0600
+++ drivers/nvme/target/configfs.c	2020-08-21 05:37:13.817449795 -0600
@@ -543,6 +543,34 @@
 
 CONFIGFS_ATTR(nvmet_ns_, buffered_io);
 
+static ssize_t nvmet_ns_readonly_show(struct config_item *item, char *page)
+{
+        return sprintf(page, "%d\n", to_nvmet_ns(item)->readonly);
+}
+
+static ssize_t nvmet_ns_readonly_store(struct config_item *item,
+                const char *page, size_t count)
+{
+        struct nvmet_ns *ns = to_nvmet_ns(item);
+        bool val;
+
+        if (strtobool(page, &val))
+                return -EINVAL;
+
+        mutex_lock(&ns->subsys->lock);
+        if (ns->enabled) {
+                pr_err("disable ns before setting readonly value.\n");
+                mutex_unlock(&ns->subsys->lock);
+                return -EINVAL;
+        }
+
+        ns->readonly = val;
+        mutex_unlock(&ns->subsys->lock);
+        return count;
+}
+
+CONFIGFS_ATTR(nvmet_ns_, readonly);
+
 static struct configfs_attribute *nvmet_ns_attrs[] = {
 	&nvmet_ns_attr_device_path,
 	&nvmet_ns_attr_device_nguid,
@@ -550,6 +578,7 @@
 	&nvmet_ns_attr_ana_grpid,
 	&nvmet_ns_attr_enable,
 	&nvmet_ns_attr_buffered_io,
+	&nvmet_ns_attr_readonly,
 #ifdef CONFIG_PCI_P2PDMA
 	&nvmet_ns_attr_p2pmem,
 #endif

[-- Attachment #3: Type: text/plain, Size: 158 bytes --]

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

^ permalink raw reply	[flat|nested] 6+ messages in thread