On 2016/7/21 12:18, Al Viro wrote:
Hash insertion does smp_store_release(). Hash chain traversal -
smp_read_barrier_depends(). On ppc the former is lwsync, while the latter
is no-op, so it boils down to
store dentry->d_name.name
lwsync
store mangled address of dentry into hash chain
vs.
fetch mangled address of dentry
demangle it
fetch dentry->d_name.name
which should be enough - lwsync paired with address dependency gives the
ordering. IOW, it's not about the barriers in __d_alloc(), it's those in
hlist_bl_add_head_rcu() and hlist_bl_for_each_entry_rcu().
And it couldn't be a missing barrier anyway - crash dump shows that
sucker with NULL ->d_name.name.
Maybe this is useful: there's a warning that in dmesg before the oops:
http://paste.ubuntu.com/20279712/
[379630.827833] ------------[ cut here ]------------
[379630.827834] WARNING: at lib/list_debug.c:59
[379630.827835] Modules linked in: iptable_mangle iptable_nat nf_nat_ipv4 nf_nat iptable_raw iptable_filter ip_tables binfmt_misc nf_conntrack_ipv4 nf_defrag_ipv4 xt_conntrack xfs libcrc32c ses enclosure sg ipmi_powernv ipmi_msghandler rtc_opal i2c_opal i2c_core powernv_rng shpchp ext4 mbcache jbd2 dm_service_time sd_mod sr_mod crc_t10dif crct10dif_generic cdrom crct10dif_common ipr tg3 cxgb4 libata ptp pps_core dm_mirror dm_region_hash dm_log dm_multipath nf_conntrack_ftp dm_mod nf_conntrack [last unloaded: ip_tables]
[379630.827876] CPU: 120 PID: 1371 Comm: kswapd17 Not tainted 3.10.0-327.18.2.el7.ppc64le #1
[379630.827877] task: c000000fe4c49c80 ti: c000000fe4ca0000 task.ti: c000000fe4ca0000
[379630.827879] NIP: c0000000004c9964 LR: c0000000004c9960 CTR: c0000000004b2860
[379630.827880] REGS: c000000fe4ca36a0 TRAP: 0700 Not tainted (3.10.0-327.18.2.el7.ppc64le)
[379630.827881] MSR: 9000000100029033 <SF,HV,EE,ME,IR,DR,RI,LE> CR: 28044044 XER: 20000000
[379630.827886] CFAR: c000000000956fb8 SOFTE: 1
GPR00: c0000000004c9960 c000000fe4ca3920 c0000000011231b0 0000000000000054
GPR04: c00000010ce08018 c00000010ce18bf0 0000000000000000 0000000000000002
GPR08: c000000000cb31b0 0000000000000000 0000000000000000 3239343331313632
GPR12: 0000000042044042 c000000007b63800 0000000000000000 0000000000000d63
GPR16: c000000001055848 00000000000001d4 00000000000f0b03 c000000001055868
GPR20: c000000fa0a2fbd8 00000000000f0b02 0000000000000040 0000000000000078
GPR24: 0000000000000000 c000000fe4ca39b0 c000000fe4ca39a0 0000000000000000
GPR28: c000000001162580 c000000fa0a2f8d0 c000000fa0a2f800 c000002611349200
[379630.827904] NIP [c0000000004c9964] __list_del_entry+0xb4/0xe0
[379630.827906] LR [c0000000004c9960] __list_del_entry+0xb0/0xe0
[379630.827907] Call Trace:
[379630.827909] [c000000fe4ca3920] [c0000000004c9960] __list_del_entry+0xb0/0xe0 (unreliable)
[379630.827913] [c000000fe4ca3980] [c0000000003270dc] prune_dcache_sb+0x9c/0x250
[379630.827915] [c000000fe4ca3a10] [c0000000002ff558] prune_super+0x1b8/0x200
[379630.827918] [c000000fe4ca3a50] [c00000000025adf4] shrink_slab+0x204/0x3b0
[379630.827920] [c000000fe4ca3b50] [c00000000025f00c] balance_pgdat+0x8ac/0xb20
[379630.827922] [c000000fe4ca3c90] [c00000000025f444] kswapd+0x1c4/0x740
[379630.827924] [c000000fe4ca3d80] [c00000000011023c] kthread+0xec/0x100
[379630.827927] [c000000fe4ca3e30] [c00000000000a474] ret_from_kernel_thread+0x5c/0x68
[379630.827928] Instruction dump:
[379630.827929] 0fe00000 4bffffd8 3c62ffa5 38639668 4848d5f9 60000000 0fe00000 4bffffc0
[379630.827931] 3c62ffa5 38639628 4848d5e1 60000000 <0fe00000> 4bffffa8 3c62ffa5 7d254b78
[379630.827935] ---[ end trace 7c1c505a25279a31 ]---
Thanks
Fiona
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
