This is a note to let you know that I've just added the patch titled
libceph: allow addrvecs with a single NONE/blank address
to the 5.12-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
libceph-allow-addrvecs-with-a-single-none-blank-address.patch
and it can be found in the queue-5.12 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From 3f1c6f2122fc780560f09735b6d1dbf39b44eb0f Mon Sep 17 00:00:00 2001
From: Ilya Dryomov <idryomov@xxxxxxxxx>
Date: Mon, 3 May 2021 17:09:01 +0200
Subject: libceph: allow addrvecs with a single NONE/blank address
From: Ilya Dryomov <idryomov@xxxxxxxxx>
commit 3f1c6f2122fc780560f09735b6d1dbf39b44eb0f upstream.
Normally, an unused OSD id/slot is represented by an empty addrvec.
However, it also appears to be possible to generate an osdmap where
an unused OSD id/slot has an addrvec with a single blank address of
type NONE. Allow such addrvecs and make the end result be exactly
the same as for the empty addrvec case -- leave addr intact.
Cc: stable@xxxxxxxxxxxxxxx # 5.11+
Signed-off-by: Ilya Dryomov <idryomov@xxxxxxxxx>
Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
net/ceph/decode.c | 20 ++++++++++++++------
1 file changed, 14 insertions(+), 6 deletions(-)
--- a/net/ceph/decode.c
+++ b/net/ceph/decode.c
@@ -4,6 +4,7 @@
#include <linux/inet.h>
#include <linux/ceph/decode.h>
+#include <linux/ceph/messenger.h> /* for ceph_pr_addr() */
static int
ceph_decode_entity_addr_versioned(void **p, void *end,
@@ -110,6 +111,7 @@ int ceph_decode_entity_addrvec(void **p,
}
ceph_decode_32_safe(p, end, addr_cnt, e_inval);
+ dout("%s addr_cnt %d\n", __func__, addr_cnt);
found = false;
for (i = 0; i < addr_cnt; i++) {
@@ -117,6 +119,7 @@ int ceph_decode_entity_addrvec(void **p,
if (ret)
return ret;
+ dout("%s i %d addr %s\n", __func__, i, ceph_pr_addr(&tmp_addr));
if (tmp_addr.type == my_type) {
if (found) {
pr_err("another match of type %d in addrvec\n",
@@ -128,13 +131,18 @@ int ceph_decode_entity_addrvec(void **p,
found = true;
}
}
- if (!found && addr_cnt != 0) {
- pr_err("no match of type %d in addrvec\n",
- le32_to_cpu(my_type));
- return -ENOENT;
- }
- return 0;
+ if (found)
+ return 0;
+
+ if (!addr_cnt)
+ return 0; /* normal -- e.g. unused OSD id/slot */
+
+ if (addr_cnt == 1 && !memchr_inv(&tmp_addr, 0, sizeof(tmp_addr)))
+ return 0; /* weird but effectively the same as !addr_cnt */
+
+ pr_err("no match of type %d in addrvec\n", le32_to_cpu(my_type));
+ return -ENOENT;
e_inval:
return -EINVAL;
Patches currently in stable-queue which might be from idryomov@xxxxxxxxx are
queue-5.12/libceph-allow-addrvecs-with-a-single-none-blank-address.patch
queue-5.12/libceph-don-t-set-global_id-until-we-get-an-auth-ticket.patch
queue-5.12/libceph-bump-cephxauthenticate-encoding-version.patch
