NFS: Fix an Oops in the pNFS files and flexfiles connection setup to the DS
Chris Worley reports:
RIP: 0010:[<ffffffffa0245f80>] [<ffffffffa0245f80>] rpc_new_client+0x2a0/0x2e0 [sunrpc]
RSP: 0018:ffff880158f6f548 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff880234f8bc00 RCX: 000000000000ea60
RDX: 0000000000074cc0 RSI: 000000000000ea60 RDI: ffff880234f8bcf0
RBP: ffff880158f6f588 R08: 000000000001ac80 R09: ffff880237003300
R10: ffff880201171000 R11: ffffea0000d75200 R12: ffffffffa03afc60
R13: ffff880230c18800 R14: 0000000000000000 R15: ffff880158f6f680
FS: 00007f0e32673740(0000) GS:ffff88023fc40000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000008 CR3: 0000000234886000 CR4: 00000000001406e0
Stack:
ffffffffa047a680 0000000000000000 ffff880158f6f598 ffff880158f6f680
ffff880158f6f680 ffff880234d11d00 ffff88023357f800 ffff880158f6f7d0
ffff880158f6f5b8 ffffffffa024660a ffff880158f6f5b8 ffffffffa02492ec
Call Trace:
[<ffffffffa024660a>] rpc_create_xprt+0x1a/0xb0 [sunrpc]
[<ffffffffa02492ec>] ? xprt_create_transport+0x13c/0x240 [sunrpc]
[<ffffffffa0246766>] rpc_create+0xc6/0x1a0 [sunrpc]
[<ffffffffa038e695>] nfs_create_rpc_client+0xf5/0x140 [nfs]
[<ffffffffa038f31a>] nfs_init_client+0x3a/0xd0 [nfs]
[<ffffffffa038f22f>] nfs_get_client+0x25f/0x310 [nfs]
[<ffffffffa025cef8>] ? rpc_ntop+0xe8/0x100 [sunrpc]
[<ffffffffa047512c>] nfs3_set_ds_client+0xcc/0x100 [nfsv3]
[<ffffffffa041fa10>] nfs4_pnfs_ds_connect+0x120/0x400 [nfsv4]
[<ffffffffa03d41c7>] nfs4_ff_layout_prepare_ds+0xe7/0x330 [nfs_layout_flexfiles]
[<ffffffffa03d1b1b>] ff_layout_pg_init_write+0xcb/0x280 [nfs_layout_flexfiles]
[<ffffffffa03a14dc>] __nfs_pageio_add_request+0x12c/0x490 [nfs]
[<ffffffffa03a1fa2>] nfs_pageio_add_request+0xc2/0x2a0 [nfs]
[<ffffffffa03a0365>] ? nfs_pageio_init+0x75/0x120 [nfs]
[<ffffffffa03a5b50>] nfs_do_writepage+0x120/0x270 [nfs]
[<ffffffffa03a5d31>] nfs_writepage_locked+0x61/0xc0 [nfs]
[<ffffffff813d4115>] ? __percpu_counter_add+0x55/0x70
[<ffffffffa03a6a9f>] nfs_wb_single_page+0xef/0x1c0 [nfs]
[<ffffffff811ca4a3>] ? __dec_zone_page_state+0x33/0x40
[<ffffffffa0395b21>] nfs_launder_page+0x41/0x90 [nfs]
[<ffffffff811baba0>] invalidate_inode_pages2_range+0x340/0x3a0
[<ffffffff811bac17>] invalidate_inode_pages2+0x17/0x20
[<ffffffffa039960e>] nfs_release+0x9e/0xb0 [nfs]
[<ffffffffa0399570>] ? nfs_open+0x60/0x60 [nfs]
[<ffffffffa0394dad>] nfs_file_release+0x3d/0x60 [nfs]
[<ffffffff81226e6c>] __fput+0xdc/0x1e0
[<ffffffff81226fbe>] ____fput+0xe/0x10
[<ffffffff810bf2e4>] task_work_run+0xc4/0xe0
[<ffffffff810a4188>] do_exit+0x2e8/0xb30
[<ffffffff8102471c>] ? do_audit_syscall_entry+0x6c/0x70
[<ffffffff811464e6>] ? __audit_syscall_exit+0x1e6/0x280
[<ffffffff810a4a5f>] do_group_exit+0x3f/0xa0
[<ffffffff810a4ad4>] SyS_exit_group+0x14/0x20
[<ffffffff8179b76e>] system_call_fastpath+0x12/0x71
Which seems to be due to a call to utsname() when in a task exit context
in order to determine the hostname to set in rpc_new_client().
In reality, what we want here is not the hostname of the current task, but
the hostname that was used to set up the metadata server.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index 10410e8..5fc7fbb 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -349,10 +349,10 @@
* Returns pointer to an NFS client, or an ERR_PTR value.
*/
struct nfs_client *nfs4_init_client(struct nfs_client *clp,
- const struct rpc_timeout *timeparms,
- const char *ip_addr)
+ const struct nfs_client_initdata *cl_init)
{
char buf[INET6_ADDRSTRLEN + 1];
+ const char *ip_addr = cl_init->ip_addr;
struct nfs_client *old;
int error;
@@ -370,9 +370,9 @@
__set_bit(NFS_CS_DISCRTRY, &clp->cl_flags);
__set_bit(NFS_CS_NO_RETRANS_TIMEOUT, &clp->cl_flags);
- error = nfs_create_rpc_client(clp, timeparms, RPC_AUTH_GSS_KRB5I);
+ error = nfs_create_rpc_client(clp, cl_init, RPC_AUTH_GSS_KRB5I);
if (error == -EINVAL)
- error = nfs_create_rpc_client(clp, timeparms, RPC_AUTH_UNIX);
+ error = nfs_create_rpc_client(clp, cl_init, RPC_AUTH_UNIX);
if (error < 0)
goto error;
@@ -793,10 +793,12 @@
.hostname = hostname,
.addr = addr,
.addrlen = addrlen,
+ .ip_addr = ip_addr,
.nfs_mod = &nfs_v4,
.proto = proto,
.minorversion = minorversion,
.net = net,
+ .timeparms = timeparms,
};
struct nfs_client *clp;
int error;
@@ -809,7 +811,7 @@
set_bit(NFS_CS_MIGRATION, &cl_init.init_flags);
/* Allocate or find a client reference we can use */
- clp = nfs_get_client(&cl_init, timeparms, ip_addr, authflavour);
+ clp = nfs_get_client(&cl_init, authflavour);
if (IS_ERR(clp)) {
error = PTR_ERR(clp);
goto error;
@@ -847,15 +849,18 @@
int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans,
u32 minor_version, rpc_authflavor_t au_flavor)
{
+ struct rpc_timeout ds_timeout;
struct nfs_client_initdata cl_init = {
.addr = ds_addr,
.addrlen = ds_addrlen,
+ .nodename = mds_clp->cl_rpcclient->cl_nodename,
+ .ip_addr = mds_clp->cl_ipaddr,
.nfs_mod = &nfs_v4,
.proto = ds_proto,
.minorversion = minor_version,
.net = mds_clp->cl_net,
+ .timeparms = &ds_timeout,
};
- struct rpc_timeout ds_timeout;
struct nfs_client *clp;
char buf[INET6_ADDRSTRLEN + 1];
@@ -869,8 +874,7 @@
* (section 13.1 RFC 5661).
*/
nfs_init_timeout_values(&ds_timeout, ds_proto, ds_timeo, ds_retrans);
- clp = nfs_get_client(&cl_init, &ds_timeout, mds_clp->cl_ipaddr,
- au_flavor);
+ clp = nfs_get_client(&cl_init, au_flavor);
dprintk("<-- %s %p\n", __func__, clp);
return clp;