From: Olga Kornievskaia <kolga@netapp.com>
To: <Trond.Myklebust@primarydata.com>, <anna.schumaker@netapp.com>
Cc: <linux-nfs@vger.kernel.org>
Subject: [PATCH 1/1] PNFS fix dangling DS mount
Date: Tue, 13 Jun 2017 14:36:20 -0400 [thread overview]
Message-ID: <20170613183620.57538-1-kolga@netapp.com> (raw)
In-Reply-To: <CAN-5tyHUJW85oxaX81Zr3Drp5WFv05-tWWXGcKCQh5S+1VZ7JQ@mail.gmail.com>
There is a regression by commit 8d40b0f14846 ("NFS filelayout:call
GETDEVICEINFO after pnfs_layout_process completes"). It leaves the
DS mount dangling.
Previously, filelayout_alloc_sec() would call filelayout_check_layout()
which would call nfs4_find_get_deviceid which ups the count on the
device_id. It's only called once and it's matched by the
filelayout_free_lseg() that calls nfs4_fl_put_deviceid().
After that patch, each read/write ends up calling nfs4_find_get_deviceid
and there is no balance for that. Instead, do nfs4_fl_put_deviceid()
in the filelayout's .pg_cleanup and remove it from filelayout_free_lseg.
But we still want a reference to hold over the lifetime of the segment.
So start the recount at 2 and keep the nfs4_fl_put_deviceid() in
the filelayout_free_lseg().
Fixes: 8d40b0f14846 ("NFS filelayout:call GETDEVICEINFO after pnfs_layout_process completes")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
fs/nfs/filelayout/filelayout.c | 15 +++++++++++++--
fs/nfs/pnfs_dev.c | 2 +-
2 files changed, 14 insertions(+), 3 deletions(-)
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index acd30ba..33ebd1a 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -989,18 +989,29 @@ static void _filelayout_free_lseg(struct nfs4_filelayout_segment *fl)
nfs_pageio_reset_write_mds(pgio);
}
+static void filelayout_pg_cleanup(struct nfs_pageio_descriptor *desc)
+{
+ if (desc->pg_lseg) {
+ struct nfs4_filelayout_segment *fl =
+ FILELAYOUT_LSEG(desc->pg_lseg);
+
+ nfs4_fl_put_deviceid(fl->dsaddr);
+ }
+ pnfs_generic_pg_cleanup(desc);
+}
+
static const struct nfs_pageio_ops filelayout_pg_read_ops = {
.pg_init = filelayout_pg_init_read,
.pg_test = filelayout_pg_test,
.pg_doio = pnfs_generic_pg_readpages,
- .pg_cleanup = pnfs_generic_pg_cleanup,
+ .pg_cleanup = filelayout_pg_cleanup,
};
static const struct nfs_pageio_ops filelayout_pg_write_ops = {
.pg_init = filelayout_pg_init_write,
.pg_test = filelayout_pg_test,
.pg_doio = pnfs_generic_pg_writepages,
- .pg_cleanup = pnfs_generic_pg_cleanup,
+ .pg_cleanup = filelayout_pg_cleanup,
};
static u32 select_bucket_index(struct nfs4_filelayout_segment *fl, u32 j)
diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index 2961fcd..671d65f 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -255,7 +255,7 @@ struct nfs4_deviceid_node *
d->nfs_client = server->nfs_client;
d->flags = 0;
d->deviceid = *id;
- atomic_set(&d->ref, 1);
+ atomic_set(&d->ref, 2);
}
EXPORT_SYMBOL_GPL(nfs4_init_deviceid_node);
--
1.8.3.1
next prev parent reply other threads:[~2017-06-13 18:36 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-06-08 16:55 [PATCH 1/1] PNFS fix dangling DS mount Olga Kornievskaia
2017-06-09 20:58 ` Olga Kornievskaia
2017-06-13 18:36 ` Olga Kornievskaia [this message]
2017-06-20 19:37 ` Trond Myklebust
2017-06-20 20:40 ` Olga Kornievskaia
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170613183620.57538-1-kolga@netapp.com \
--to=kolga@netapp.com \
--cc=Trond.Myklebust@primarydata.com \
--cc=anna.schumaker@netapp.com \
--cc=linux-nfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).