[prev in list] [next in list] [prev in thread] [next in thread]
List: linux-nfs
Subject: Re: Kernel bug triggered with xfstest generic/113
From: Scott Mayhew <smayhew () redhat ! com>
Date: 2017-10-31 17:39:45
Message-ID: 20171031173945.rgsvdtulqmsa5hkz () tonberry ! usersys ! redhat ! com
[Download RAW message or body]
On Thu, 26 Oct 2017, Mkrtchyan, Tigran wrote:
>
>
> Against dCache nfs server:
>
> [ 3987.717284] ------------[ cut here ]------------ \
> [ 3987.717286] kernel BUG at fs/inode.c:567! \
> [ 3987.717292] invalid opcode: 0000 [#1] SMP \
> [ 3987.717293] Modules linked in: loop nfs_layout_nfsv41_files rpcsec_gss_krb5 \
> nfsv4 nfs lockd grace fscache binfmt_misc af_packet nf_conntrack_netbios_ns \
> nf_conntrack_broadcast xt_tcpudp xt _CT ip6t_REJECT nf_reject_ipv6 \
> nf_conntrack_ipv6 nf_defrag_ipv6 ipt_REJECT nf_reject_ipv4 nf_conntrack_ipv4 \
> nf_defrag_ipv4 xt_conntrack nf_conntrack libcrc32c nfnetlink ip6table_mangle ip6ta \
> ble_raw ip6table_security iptable_mangle iptable_raw iptable_security \
> ip6table_filter ip6_tables iptable_filter btrfs xor zstd_decompress zstd_compress \
> xxhash lzo_compress zlib_deflate raid6 _pq snd_hda_codec_idt snd_hda_codec_generic \
> snd_hda_intel snd_hda_codec snd_hda_core snd_hwdep snd_seq iTCO_wdt \
> iTCO_vendor_support snd_seq_device snd_timer lpc_ich mfd_core tpm_tis snd_pcm \
> tpm_tis_core i2c_i801 snd tpm auth_rpcgss oid_registry \
> [ 3987.717340] sunrpc ip_tables x_tables serio_raw e1000e ptp pps_core autofs4 \
> [ 3987.717349] CPU: 1 PID: 31883 Comm: rm Not tainted \
> 4.14.0-rc6-01076-gbb176f67090c #120 \
> [ 3987.717350] Hardware name: Comptronic pczW1007/DX38BT, BIOS \
> BTX3810J.86A.1893.2008.1009.1712 10/09/2008 \
> [ 3987.717353] task: ffff880107f7c480 task.stack: ffff88012a830000 \
> [ 3987.717358] RIP: 0010:evict+0x161/0x180 \
> [ 3987.717360] RSP: 0018:ffff88012a833e80 EFLAGS: 00010202 \
> [ 3987.717362] RAX: ffffffff81806110 RBX: ffff880100037800 RCX: ffff8801000378a0 \
> [ 3987.717364] RDX: ffffffff81806110 RSI: ffff8801000378a0 RDI: ffffffff81806108 \
> [ 3987.717366] RBP: ffff88012a833e98 R08: 0000000000000000 R09: ffff88012a03d6e8 \
> [ 3987.717368] R10: ffff88012a03d6f8 R11: ffff88012a03d6f8 R12: ffff880100037910 \
> [ 3987.717370] R13: ffffffffa0404540 R14: 00000000ffffff9c R15: 0000000000000000 \
> [ 3987.717372] FS: 00007faddfd6e700(0000) GS:ffff88012fc80000(0000) \
> knlGS:0000000000000000 [ 3987.717375] CS: 0010 DS: 0000 ES: 0000 CR0: \
> 0000000080050033 [ 3987.717377] CR2: \
> 00005653ab778470 CR3: 000000012840c000 CR4: 00000000000406e0 [ \
> 3987.717378] Call Trace: \
> [ 3987.717382] iput+0xf9/0x1c0 \
> [ 3987.717385] do_unlinkat+0x18d/0x2f0 \
> [ 3987.717388] SyS_unlinkat+0x16/0x30 \
> [ 3987.717392] entry_SYSCALL_64_fastpath+0x13/0x94 \
> [ 3987.717394] RIP: 0033:0x7faddf8a0c17 \
> [ 3987.717396] RSP: 002b:00007ffc44a11468 EFLAGS: 00000246 ORIG_RAX: \
> 0000000000000107 [ 3987.717399] RAX: ffffffffffffffda RBX: \
> 000055ccd0723428 RCX: 00007faddf8a0c17 [ 3987.717401] RDX: 0000000000000000 RSI: \
> 000055ccd07220f0 RDI: 00000000ffffff9c [ 3987.717403] RBP: 000055ccd0723320 R08: \
> 0000000000000003 R09: 0000000000000000 [ 3987.717404] R10: 0000000000000100 R11: \
> 0000000000000246 R12: 000055ccd0722060 [ 3987.717406] R13: 00007ffc44a115a0 R14: \
> 00007faddfb6a1e4 R15: 0000000000000000 [ 3987.717408] Code: 89 df e8 23 7d fe ff eb \
> 8f 48 83 bb 20 02 00 00 00 74 85 48 89 df e8 7f ad 01 00 0f b7 03 66 25 00 f0 e9 6b \
> ff ff ff 0f 0b 0f 0b <0f> 0b 48 8d bb 68 01 00 00 e8 a1 9e fa ff 48 89 df e8 a9 f1 \
> ff [ 3987.717443] RIP: evict+0x161/0x180 RSP: ffff88012a833e80
> [ 3987.717445] ---[ end trace ca9a0f6be0e72301 ]---
I've been seeing a similar panic. Does the attached patch help?
-Scott
>
> Looks like it happens in post-test clean-up step (rm), though some processes are \
> still there:
> root 18489 18329 0 Oct25 pts/1 00:00:00 \
> /data/git/xfstests-dev/ltp/aio-stress -t 20 -s 10 -O -I 1000 \
> /mnt/test/aiostress.18329.3 /mnt/test/aiostress.18329.3.20 \
> /mnt/test/aiostress.18329.3.19 /mnt/test/aiostress.18329.3.18 \
> /mnt/test/aiostress.18329.3.17 /mnt/test/aiostress.18329.3.16 \
> /mnt/test/aiostress.18329.3.15 /mnt/test/aiostress.18329.3.14 \
> /mnt/test/aiostress.18329.3.13 /mnt/test/aiostress.18329.3.12 \
> /mnt/test/aiostress.18329.3.11 /mnt/test/aiostress.18329.3.10 \
> /mnt/test/aiostress.18329.3.9 /mnt/test/aiostress.18329.3.8 \
> /mnt/test/aiostress.18329.3.7 /mnt/test/aiostress.18329.3.6 \
> /mnt/test/aiostress.18329.3.5 /mnt/test/aiostress.18329.3.4 \
> /mnt/test/aiostress.18329.3.3 /mnt/test/aiostress.18329.3.2
>
> I can always re-produce it.
>
> Tigran.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
["0001-nfs-make-pnfs_destroy_layout-wait-on-commit-when-ino.patch" (text/plain)]
From f432e165ebfb981608b038f4e344f5eb606691ee Mon Sep 17 00:00:00 2001
From: Scott Mayhew <smayhew@redhat.com>
Date: Thu, 26 Oct 2017 14:57:18 -0400
Subject: [PATCH] nfs: make pnfs_destroy_layout() wait on commit when inode is
being evicted
Otherwise nfs_commit_inode() can set I_DIRTY_DATASYNC in the inode's
i_state, allowing the following BUG_ON in evict() to be triggered:
BUG_ON(inode->i_state != (I_FREEING | I_CLEAR));
Fixes: 1f18b82c3437 ("pNFS: Ensure we commit the layout if it has been...")
Signed-off-by: Scott Mayhew <smayhew@redhat.com>
---
fs/nfs/filelayout/filelayout.c | 2 +-
fs/nfs/flexfilelayout/flexfilelayout.c | 2 +-
fs/nfs/nfs4super.c | 2 +-
fs/nfs/pnfs.c | 6 +++---
fs/nfs/pnfs.h | 2 +-
5 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index 508126e..8845a73 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -169,7 +169,7 @@ static int filelayout_async_handle_error(struct rpc_task *task,
* i/o and all i/o waiting on the slot table to the MDS until
* layout is destroyed and a new valid layout is obtained.
*/
- pnfs_destroy_layout(NFS_I(inode));
+ pnfs_destroy_layout(NFS_I(inode), 0);
rpc_wake_up(&tbl->slot_tbl_waitq);
goto reset;
/* RPC connection errors */
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index b0fa83a..2f0f682 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -1088,7 +1088,7 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task,
* i/o and all i/o waiting on the slot table to the MDS until
* layout is destroyed and a new valid layout is obtained.
*/
- pnfs_destroy_layout(NFS_I(inode));
+ pnfs_destroy_layout(NFS_I(inode), 0);
rpc_wake_up(&tbl->slot_tbl_waitq);
goto reset;
/* RPC connection errors */
diff --git a/fs/nfs/nfs4super.c b/fs/nfs/nfs4super.c
index 6fb7cb6..3b4a063 100644
--- a/fs/nfs/nfs4super.c
+++ b/fs/nfs/nfs4super.c
@@ -95,7 +95,7 @@ static void nfs4_evict_inode(struct inode *inode)
nfs_inode_return_delegation_noreclaim(inode);
/* Note that above delegreturn would trigger pnfs return-on-close */
pnfs_return_layout(inode);
- pnfs_destroy_layout(NFS_I(inode));
+ pnfs_destroy_layout(NFS_I(inode), FLUSH_SYNC);
/* First call standard NFS clear_inode() code */
nfs_clear_inode(inode);
}
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 3bcd669..0884f37 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -678,7 +678,7 @@ pnfs_free_lseg_list(struct list_head *free_me)
}
void
-pnfs_destroy_layout(struct nfs_inode *nfsi)
+pnfs_destroy_layout(struct nfs_inode *nfsi, int how)
{
struct pnfs_layout_hdr *lo;
LIST_HEAD(tmp_list);
@@ -692,7 +692,7 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RW_FAILED);
spin_unlock(&nfsi->vfs_inode.i_lock);
pnfs_free_lseg_list(&tmp_list);
- nfs_commit_inode(&nfsi->vfs_inode, 0);
+ nfs_commit_inode(&nfsi->vfs_inode, how);
pnfs_put_layout_hdr(lo);
} else
spin_unlock(&nfsi->vfs_inode.i_lock);
@@ -1831,7 +1831,7 @@ pnfs_update_layout(struct inode *ino,
}
/* Destroy the existing layout and start over */
if (time_after(jiffies, giveup))
- pnfs_destroy_layout(NFS_I(ino));
+ pnfs_destroy_layout(NFS_I(ino), 0);
/* Fallthrough */
case -EAGAIN:
break;
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 87f144f..51baf59 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -244,7 +244,7 @@ size_t pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio,
void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg);
struct pnfs_layout_segment *pnfs_layout_process(struct nfs4_layoutget *lgp);
void pnfs_free_lseg_list(struct list_head *tmp_list);
-void pnfs_destroy_layout(struct nfs_inode *);
+void pnfs_destroy_layout(struct nfs_inode *, int how);
void pnfs_destroy_all_layouts(struct nfs_client *);
int pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
struct nfs_fsid *fsid,
--
2.9.5
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic