Received-SPF: pass (google.com: domain of linux-nfs-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 31 Oct 2022 20:09:22 +0100
From:   Petr Vorel <pvorel@suse.cz>
To:     Jeff Layton <jlayton@kernel.org>
Cc:     Chuck Lever III <chuck.lever@oracle.com>,
        Linux NFS Mailing List <linux-nfs@vger.kernel.org>,
        NeilBrown <neilb@suse.com>, Steve Dickson <steved@redhat.com>,
        "ltp@lists.linux.it" <ltp@lists.linux.it>
Subject: Re: LTP test regression on exportfs in kernel 6.0
Message-ID: <Y2AdYhMwyfX9vRzu@pevik>
Reply-To: Petr Vorel <pvorel@suse.cz>
References: <Y1/P8gDAcWC/+VR3@pevik>
 <5AE322AD-FF73-4495-A3DF-60A98DBE49A7@oracle.com>
 <b735dfe6911387b7ba3a2fbfc599c9c55a455288.camel@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
In-Reply-To: <b735dfe6911387b7ba3a2fbfc599c9c55a455288.camel@kernel.org>
Precedence: bulk

Hi all,

FYI bug fixed by Jeff

https://git.kernel.org/pub/scm/linux/kernel/git/cel/linux.git/commit/?h=3Df=
or-rc&id=3D30bf340ba9d7c8059af7135e5c73f8e9d3240b29
https://lore.kernel.org/linux-nfs/20221031154921.500620-1-jlayton@kernel.or=
g/

Kind regards,
Petr

> On Mon, 2022-10-31 at 13:47 +0000, Chuck Lever III wrote:
> > [ dropping Bruce, he's no longer an NFSD maintainer ]


> > > On Oct 31, 2022, at 9:38 AM, Petr Vorel <pvorel@suse.cz> wrote:

> > > Hi all,

> > > LTP tests got broken on kernel 6.0. During test cleanup it hangs on e=
xportfs
> > > (high load). I tested it both on nfs-utils 2.6.2, the same userspace,=
 the
> > > version of the tests (obviously), it looks really like kernel related=
 regression.

> > Not oopsen, but rather RCU splats.

> > [  326.587135] rcu: INFO: rcu_preempt self-detected stall on CPU
> > [  326.587168] rcu: 	0-....: (5249 ticks this GP) idle=3Dcf74/1/0x40000=
00000000000 softirq=3D7256/7256 fqs=3D2484
> > [  326.587187] 	(t=3D5250 jiffies g=3D4009 q=3D388 ncpus=3D2)

> > and

> > watchdog: BUG: soft lockup - CPU#0 stuck for 44s! [exportfs:1180]

> > RIP: __nfsd_file_cache_purge+0x54

> > seems like the common element. Jeff, can you have a look at this?


> Sure, I'll take a look since I'm in here anyway.

> I'll see if I can set up a reproducer.


> > > # Simple reproducer
> > > git clone https://github.com/linux-test-project/ltp.git
> > > # install dependencies with your distro in ci/*.sh
> > > make autotools && ./configure && make -j$(getconf _NPROCESSORS_ONLN) =
&& make install
> > > export LTPROOT=3D/opt/ltp; PATH=3D"/opt/ltp/testcases/bin:$PATH" nfs0=
2.sh -v 4 -t tcp
> > > TMPDIR=3D"/var/tmp" nfs02.sh -v 4 -t tcp

> > > # Example on kernel oops on openSUSE
> > > [  304.772002] Initializing XFRM netlink socket
> > > [  304.833577] IPv6: ADDRCONF(NETDEV_CHANGE): ltp_ns_veth1: link beco=
mes ready
> > > [  304.833604] IPv6: ADDRCONF(NETDEV_CHANGE): ltp_ns_veth2: link beco=
mes ready
> > > [  304.957269] FS-Cache: Loaded
> > > [  305.002821] Key type dns_resolver registered
> > > [  305.153926] NFS: Registering the id_resolver key type
> > > [  305.153942] Key type id_resolver registered
> > > [  305.153943] Key type id_legacy registered
> > > [  326.587135] rcu: INFO: rcu_preempt self-detected stall on CPU
> > > [  326.587168] rcu: 	0-....: (5249 ticks this GP) idle=3Dcf74/1/0x400=
0000000000000 softirq=3D7256/7256 fqs=3D2484
> > > [  326.587187] 	(t=3D5250 jiffies g=3D4009 q=3D388 ncpus=3D2)
> > > [  326.587191] NMI backtrace for cpu 0
> > > [  326.587192] CPU: 0 PID: 1008 Comm: exportfs Not tainted 5.19.0 #4
> > > [  326.587195] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),=
 BIOS rel-1.16.0-0-gd239552-rebuilt.opensuse.org 04/01/2014
> > > [  326.587199] Call Trace:
> > > [  326.587212]  <IRQ>
> > > [  326.587216]  dump_stack_lvl+0x45/0x5e
> > > [  326.587231]  nmi_cpu_backtrace.cold+0x30/0x76
> > > [  326.587234]  ? lapic_can_unplug_cpu+0x80/0x80
> > > [  326.587241]  nmi_trigger_cpumask_backtrace+0x10c/0x140
> > > [  326.587252]  rcu_dump_cpu_stacks+0xf8/0x130
> > > [  326.587260]  rcu_sched_clock_irq.cold+0x67/0x2f8
> > > [  326.587270]  ? raw_notifier_call_chain+0x41/0x60
> > > [  326.587275]  ? timekeeping_update+0xde/0x130
> > > [  326.587280]  ? __cgroup_account_cputime_field+0x5d/0x90
> > > [  326.587286]  update_process_times+0x62/0x90
> > > [  326.587291]  tick_sched_handle+0x22/0x60
> > > [  326.587300]  tick_sched_timer+0x71/0x90
> > > [  326.587303]  ? tick_sched_do_timer+0xa0/0xa0
> > > [  326.587306]  __hrtimer_run_queues+0x127/0x2c0
> > > [  326.587310]  hrtimer_interrupt+0x106/0x220
> > > [  326.587313]  __sysvec_apic_timer_interrupt+0x7c/0x170
> > > [  326.587319]  sysvec_apic_timer_interrupt+0x9d/0xd0
> > > [  326.587328]  </IRQ>
> > > [  326.587328]  <TASK>
> > > [  326.587329]  asm_sysvec_apic_timer_interrupt+0x16/0x20
> > > [  326.587339] RIP: 0010:__nfsd_file_cache_purge+0x54/0xd0 [nfsd]
> > > [  326.587374] Code: 48 89 24 24 48 89 64 24 08 e8 88 3e e8 e5 48 8d =
7c 24 10 e8 ae 53 e8 e5 48 8d 7c 24 10 e8 04 4d e8 e5 48 89 c5 48 85 c0 74 =
27 <48> 81 fd 00 f0 ff ff 77 55 48 85 db 74 06 48 39 5d 38 75 ec 48 89
> > > [  326.587376] RSP: 0018:ffffbfe84100bd88 EFLAGS: 00000202
> > > [  326.587378] RAX: ffff9dcd01f9a3a8 RBX: ffff9dcd0af86600 RCX: fffff=
fffc0ceabc0
> > > [  326.587380] RDX: 0000000000000000 RSI: 000000000000008c RDI: ffffb=
fe84100bd98
> > > [  326.587381] RBP: ffff9dcd01f9a3a8 R08: 0000000000000000 R09: fffff=
fffa6fd2b75
> > > [  326.587382] R10: 0000000000000000 R11: ffffbfe84100bd98 R12: ffffb=
fe84100bf08
> > > [  326.587383] R13: 00007fff97dc3a20 R14: ffffbfe84100bf08 R15: 00000=
00000000000
> > > [  326.587385]  ? _raw_spin_unlock+0x15/0x30
> > > [  326.587393]  ? __nfsd_file_cache_purge+0x4c/0xd0 [nfsd]
> > > [  326.587410]  ? _raw_spin_unlock+0x15/0x30
> > > [  326.587412]  expkey_flush+0x2a/0x40 [nfsd]
> > > [  326.587427]  write_flush.constprop.0+0xf4/0x140 [sunrpc]
> > > [  326.587460]  proc_reg_write+0x4e/0x90
> > > [  326.587470]  vfs_write+0xc3/0x3d0
> > > [  326.587478]  ? preempt_count_add+0x4d/0xa0
> > > [  326.587482]  ? fd_install+0x5c/0xe0
> > > [  326.587489]  ksys_write+0x5f/0xe0
> > > [  326.587492]  do_syscall_64+0x38/0xc0
> > > [  326.587496]  entry_SYSCALL_64_after_hwframe+0x63/0xcd
> > > [  326.587501] RIP: 0033:0x7f6f908063f3
> > > [  326.587505] Code: 8b 15 21 9a 0f 00 f7 d8 64 89 02 48 c7 c0 ff ff =
ff ff eb b7 0f 1f 00 64 8b 04 25 18 00 00 00 85 c0 75 14 b8 01 00 00 00 0f =
05 <48> 3d 00 f0 ff ff 77 55 c3 0f 1f 40 00 48 83 ec 28 48 89 54 24 18
> > > [  326.587506] RSP: 002b:00007fff97dc3a08 EFLAGS: 00000246 ORIG_RAX: =
0000000000000001
> > > [  326.587508] RAX: ffffffffffffffda RBX: 000056110e2f8dd0 RCX: 00007=
f6f908063f3
> > > [  326.587509] RDX: 000000000000000b RSI: 00007fff97dc3a20 RDI: 00000=
00000000006
> > > [  326.587510] RBP: 000056110e2f500e R08: 00007f6f908b83e0 R09: 00000=
00000000000
> > > [  326.587512] R10: 0000000000000000 R11: 0000000000000246 R12: 00007=
fff97dc3ad0
> > > [  326.587514] R13: 00007fff97dc3a20 R14: 0000000000000006 R15: 00005=
6110e2f4fef
> > > [  326.587517]  </TASK>

> > > # Bug on Debian config with openSUSE setup looks fairly similar:
> > > [ 7972.371206] watchdog: BUG: soft lockup - CPU#0 stuck for 44s! [exp=
ortfs:1180]
> > > [ 7972.371250] Modules linked in: rpcsec_gss_krb5 nfsv4 dns_resolver =
nfs fscache netfs xfrm_user xfrm_algo veth intel_rapl_msr intel_rapl_common=
 intel_pmc_core_pltdrv intel_pmc_core kvm_intel kvm irqbypass ghash_clmulni=
_intel snd_hda_codec_generic ledtrig_audio aesni_intel crypto_simd cryptd r=
apl qxl evdev snd_hda_intel drm_ttm_helper snd_intel_dspcfg serio_raw pcspk=
r snd_intel_sdw_acpi ttm joydev virtio_console virtio_balloon snd_hda_codec=
 sg drm_kms_helper snd_hda_core snd_hwdep snd_pcm qemu_fw_cfg snd_timer snd=
 soundcore button nfsd auth_rpcgss nfs_acl lockd grace sunrpc drm fuse conf=
igfs virtio_rng rng_core ip_tables x_tables autofs4 btrfs blake2b_generic x=
or raid6_pq zstd_compress libcrc32c crc32c_generic hid_generic usbhid hid d=
m_mod virtio_net net_failover virtio_blk failover sr_mod cdrom crct10dif_pc=
lmul crct10dif_common crc32_pclmul crc32c_intel psmouse ata_generic uhci_hc=
d ehci_pci ehci_hcd ata_piix usbcore libata usb_common virtio_pci virtio_pc=
i_legacy_dev virtio_pci_modern_dev
> > > [ 7972.371310]  virtio virtio_ring scsi_mod i2c_piix4 scsi_common flo=
ppy
> > > [ 7972.371315] CPU: 0 PID: 1180 Comm: exportfs Not tainted 6.0.1 #2
> > > [ 7972.371317] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),=
 BIOS rel-1.16.0-0-gd239552-rebuilt.opensuse.org 04/01/2014
> > > [ 7972.371318] RIP: 0010:__nfsd_file_cache_purge+0x54/0xd0 [nfsd]
> > > [ 7972.371344] Code: 48 89 24 24 48 89 64 24 08 e8 38 23 46 d4 48 8d =
7c 24 10 e8 5e 38 46 d4 48 8d 7c 24 10 e8 b4 31 46 d4 48 89 c5 48 85 c0 74 =
27 <48> 81 fd 00 f0 ff ff 77 55 48 85 db 74 06 48 39 5d 38 75 ec 48 89
> > > [ 7972.371348] RSP: 0018:ffffc05241067d88 EFLAGS: 00000202
> > > [ 7972.371350] RAX: ffff9bd3c5380618 RBX: ffff9bd3f260d500 RCX: fffff=
fffc0b0cbc0
> > > [ 7972.371351] RDX: 0000000000000000 RSI: 00000000000000b2 RDI: ffffc=
05241067d98
> > > [ 7972.371352] RBP: ffff9bd3c5380618 R08: 0000000000000000 R09: fffff=
fff953d3aa5
> > > [ 7972.371353] R10: 0000000000000000 R11: ffffc05241067d98 R12: ffffc=
05241067f08
> > > [ 7972.371354] R13: 00007ffd820fd9f0 R14: ffffc05241067f08 R15: 00000=
00000000000
> > > [ 7972.371356] FS:  00007fd3e4d1c740(0000) GS:ffff9bd3fec00000(0000) =
knlGS:0000000000000000
> > > [ 7972.371357] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > > [ 7972.371358] CR2: 00007fd3e4d8a040 CR3: 00000000089a6005 CR4: 00000=
00000370ef0
> > > [ 7972.371362] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 00000=
00000000000
> > > [ 7972.371363] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 00000=
00000000400
> > > [ 7972.371364] Call Trace:
> > > [ 7972.371367]  <TASK>
> > > [ 7972.371376]  ? _raw_spin_unlock+0x15/0x30
> > > [ 7972.371383]  expkey_flush+0x2a/0x40 [nfsd]
> > > [ 7972.371402]  write_flush.constprop.0+0xf4/0x140 [sunrpc]
> > > [ 7972.371440]  proc_reg_write+0x4e/0x90
> > > [ 7972.371443]  vfs_write+0xc3/0x3d0
> > > [ 7972.371446]  ? preempt_count_add+0x4d/0xa0
> > > [ 7972.371450]  ? fd_install+0x5c/0xe0
> > > [ 7972.371452]  ksys_write+0x5f/0xe0
> > > [ 7972.371454]  do_syscall_64+0x38/0xc0
> > > [ 7972.371458]  entry_SYSCALL_64_after_hwframe+0x63/0xcd
> > > [ 7972.371460] RIP: 0033:0x7fd3e4e193f3
> > > [ 7972.371462] Code: 8b 15 21 9a 0f 00 f7 d8 64 89 02 48 c7 c0 ff ff =
ff ff eb b7 0f 1f 00 64 8b 04 25 18 00 00 00 85 c0 75 14 b8 01 00 00 00 0f =
05 <48> 3d 00 f0 ff ff 77 55 c3 0f 1f 40 00 48 83 ec 28 48 89 54 24 18
> > > [ 7972.371464] RSP: 002b:00007ffd820fd9d8 EFLAGS: 00000246 ORIG_RAX: =
0000000000000001
> > > [ 7972.371466] RAX: ffffffffffffffda RBX: 00005559a20e4dd0 RCX: 00007=
fd3e4e193f3
> > > [ 7972.371467] RDX: 000000000000000b RSI: 00007ffd820fd9f0 RDI: 00000=
00000000006
> > > [ 7972.371468] RBP: 00005559a20e100e R08: 00007fd3e4ecb3e0 R09: 00000=
00000000000
> > > [ 7972.371469] R10: 0000000000000000 R11: 0000000000000246 R12: 00007=
ffd820fdaa0
> > > [ 7972.371470] R13: 00007ffd820fd9f0 R14: 0000000000000006 R15: 00005=
559a20e0fef
> > > [ 7972.371473]  </TASK>

> > > I guess I don't need to turn any debugging with rpcdebug at this poin=
t (enough
> > > info from this), right?

> > > It looks like to be config related, because it's broken with openSUSE=
 config,
> > > but not with Debian config. Building with Debian config, but changing=
 just these
> > > NFS specific options it also fails:

> > > # config difference Debian (-) vs openSUSE (+)
> > > -CONFIG_NFS_DISABLE_UDP_SUPPORT=3Dy
> > > +# CONFIG_NFS_DISABLE_UDP_SUPPORT is not set
> > > -# CONFIG_NFSD_SCSILAYOUT is not set
> > > -# CONFIG_NFSD_FLEXFILELAYOUT is not set
> > > -# CONFIG_NFSD_V4_2_INTER_SSC is not set
> > > +CONFIG_NFSD_SCSILAYOUT=3Dy
> > > +CONFIG_NFSD_FLEXFILELAYOUT=3Dy
> > > +CONFIG_NFSD_V4_2_INTER_SSC=3Dy

> > > I suspect it was caused by e394ff83bbca ("Merge tag 'nfsd-6.0' of
> > > git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux") [1], but I =
need to
> > > finish bisecting to be sure (it was merged earlier than aeb6e6ac18c7 =
("Merge tag
> > > 'nfs-for-5.20-1' of git://git.linux-nfs.org/projects/trondmy/linux-nf=
s") [2],
> > > both are broken).

> > > Kind regards,
> > > Petr

> > > [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.gi=
t/commit/?id=3De394ff83bbca1c72427b1feb5c6b9d4dad832f01
> > > [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.gi=
t/commit/?id=3Daeb6e6ac18c73ec287b3b1e2c913520699358c13

> > --
> > Chuck Lever