* [PATCH] nbd: allocate sufficient space for NBD_CMD_STATUS
@ 2020-06-18 21:02 Thadeu Lima de Souza Cascardo
2020-06-18 21:02 ` Thadeu Lima de Souza Cascardo
0 siblings, 1 reply; 2+ messages in thread
From: Thadeu Lima de Souza Cascardo @ 2020-06-18 21:02 UTC (permalink / raw)
To: Josef Bacik; +Cc: Jens Axboe, linux-block, nbd, linux-kernel, netdev, cascardo
The nest attribute NBD_ATTR_DEVICE_LIST was not accounted for when
allocating the message, resulting in -EMSGSIZE.
As __alloc_skb aligns size requests to SMP_CACHE_BYTES and SLUB will end up
allocating more than requested, this can hardly be reproduced on most
setups.
However, I managed to test this on a 32-bit x86 with 15 entries, by loading
with nbds_max=15. It failed with -EMSGSIZE, while it worked with 14 or 16
entries.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
---
drivers/block/nbd.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 43cff01a5a67..19551d8ca355 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -2265,6 +2265,7 @@ static int nbd_genl_status(struct sk_buff *skb, struct genl_info *info)
msg_size = nla_total_size(nla_attr_size(sizeof(u32)) +
nla_attr_size(sizeof(u8)));
msg_size *= (index == -1) ? nbd_total_devices : 1;
+ msg_size += nla_total_size(0); /* for NBD_ATTR_DEVICE_LIST */
reply = genlmsg_new(msg_size, GFP_KERNEL);
if (!reply)
--
2.25.1
^ permalink raw reply related [flat|nested] 2+ messages in thread
* [PATCH] nbd: allocate sufficient space for NBD_CMD_STATUS
2020-06-18 21:02 [PATCH] nbd: allocate sufficient space for NBD_CMD_STATUS Thadeu Lima de Souza Cascardo
@ 2020-06-18 21:02 ` Thadeu Lima de Souza Cascardo
0 siblings, 0 replies; 2+ messages in thread
From: Thadeu Lima de Souza Cascardo @ 2020-06-18 21:02 UTC (permalink / raw)
To: Josef Bacik; +Cc: Jens Axboe, linux-block, nbd, linux-kernel, netdev, cascardo
The nest attribute NBD_ATTR_DEVICE_LIST was not accounted for when
allocating the message, resulting in -EMSGSIZE.
As __alloc_skb aligns size requests to SMP_CACHE_BYTES and SLUB will end up
allocating more than requested, this can hardly be reproduced on most
setups.
However, I managed to test this on a 32-bit x86 with 15 entries, by loading
with nbds_max=15. It failed with -EMSGSIZE, while it worked with 14 or 16
entries.
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
---
drivers/block/nbd.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c
index 43cff01a5a67..19551d8ca355 100644
--- a/drivers/block/nbd.c
+++ b/drivers/block/nbd.c
@@ -2265,6 +2265,7 @@ static int nbd_genl_status(struct sk_buff *skb, struct genl_info *info)
msg_size = nla_total_size(nla_attr_size(sizeof(u32)) +
nla_attr_size(sizeof(u8)));
msg_size *= (index == -1) ? nbd_total_devices : 1;
+ msg_size += nla_total_size(0); /* for NBD_ATTR_DEVICE_LIST */
reply = genlmsg_new(msg_size, GFP_KERNEL);
if (!reply)
--
2.25.1
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-06-18 21:03 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-06-18 21:02 [PATCH] nbd: allocate sufficient space for NBD_CMD_STATUS Thadeu Lima de Souza Cascardo
2020-06-18 21:02 ` Thadeu Lima de Souza Cascardo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).