Skip to content

L
linux
Switch branch/tag
  1. 02 Feb, 2015 28 commits
  3. 01 Dec, 2014 12 commits
    • Zefan Li's avatar
      Linux 3.4.105 · 7fd7a446
      Zefan Li authored
      7fd7a446
    • Guillaume Nault's avatar
      l2tp: fix race while getting PMTU on PPP pseudo-wire · 4290973d
      Guillaume Nault authored 
      commit eed4d839 upstream.

Use dst_entry held by sk_dst_get() to retrieve tunnel's PMTU.

The dst_mtu(__sk_dst_get(tunnel->sock)) call was racy. __sk_dst_get()
could return NULL if tunnel->sock->sk_dst_cache was reset just before the
call, thus making dst_mtu() dereference a NULL pointer:

[ 1937.661598] BUG: unable to handle kernel NULL pointer dereference at 0000000000000020
[ 1937.664005] IP: [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005] PGD daf0c067 PUD d9f93067 PMD 0
[ 1937.664005] Oops: 0000 [#1] SMP
[ 1937.664005] Modules linked in: l2tp_ppp l2tp_netlink l2tp_core ip6table_filter ip6_tables iptable_filter ip_tables ebtable_nat ebtables x_tables udp_tunnel pppoe pppox ppp_generic slhc deflate ctr twofish_generic twofish_x86_64_3way xts lrw gf128mul glue_helper twofish_x86_64 twofish_common blowfish_generic blowfish_x86_64 blowfish_common des_generic cbc xcbc rmd160 sha512_generic hmac crypto_null af_key xfrm_algo 8021q garp bridge stp llc tun atmtcp clip atm ext3 mbcache jbd iTCO_wdt coretemp kvm_intel iTCO_vendor_support kvm pcspkr evdev ehci_pci lpc_ich mfd_core i5400_edac edac_core i5k_amb shpchp button processor thermal_sys xfs crc32c_generic libcrc32c dm_mod usbhid sg hid sr_mod sd_mod cdrom crc_t10dif crct10dif_common ata_generic ahci ata_piix tg3 libahci libata uhci_hcd ptp ehci_hcd pps_core usbcore scsi_mod libphy usb_common [last unloaded: l2tp_core]
[ 1937.664005] CPU: 0 PID: 10022 Comm: l2tpstress Tainted: G           O   3.17.0-rc1 #1
[ 1937.664005] Hardware name: HP ProLiant DL160 G5, BIOS O12 08/22/2008
[ 1937.664005] task: ffff8800d8fda790 ti: ffff8800c43c4000 task.ti: ffff8800c43c4000
[ 1937.664005] RIP: 0010:[<ffffffffa049db88>]  [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005] RSP: 0018:ffff8800c43c7de8  EFLAGS: 00010282
[ 1937.664005] RAX: ffff8800da8a7240 RBX: ffff8800d8c64600 RCX: 000001c325a137b5
[ 1937.664005] RDX: 8c6318c6318c6320 RSI: 000000000000010c RDI: 0000000000000000
[ 1937.664005] RBP: ffff8800c43c7ea8 R08: 0000000000000000 R09: 0000000000000000
[ 1937.664005] R10: ffffffffa048e2c0 R11: ffff8800d8c64600 R12: ffff8800ca7a5000
[ 1937.664005] R13: ffff8800c439bf40 R14: 000000000000000c R15: 0000000000000009
[ 1937.664005] FS:  00007fd7f610f700(0000) GS:ffff88011a600000(0000) knlGS:0000000000000000
[ 1937.664005] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 1937.664005] CR2: 0000000000000020 CR3: 00000000d9d75000 CR4: 00000000000027e0
[ 1937.664005] Stack:
[ 1937.664005]  ffffffffa049da80 ffff8800d8fda790 000000000000005b ffff880000000009
[ 1937.664005]  ffff8800daf3f200 0000000000000003 ffff8800c43c7e48 ffffffff81109b57
[ 1937.664005]  ffffffff81109b0e ffffffff8114c566 0000000000000000 0000000000000000
[ 1937.664005] Call Trace:
[ 1937.664005]  [<ffffffffa049da80>] ? pppol2tp_connect+0x235/0x41e [l2tp_ppp]
[ 1937.664005]  [<ffffffff81109b57>] ? might_fault+0x9e/0xa5
[ 1937.664005]  [<ffffffff81109b0e>] ? might_fault+0x55/0xa5
[ 1937.664005]  [<ffffffff8114c566>] ? rcu_read_unlock+0x1c/0x26
[ 1937.664005]  [<ffffffff81309196>] SYSC_connect+0x87/0xb1
[ 1937.664005]  [<ffffffff813e56f7>] ? sysret_check+0x1b/0x56
[ 1937.664005]  [<ffffffff8107590d>] ? trace_hardirqs_on_caller+0x145/0x1a1
[ 1937.664005]  [<ffffffff81213dee>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[ 1937.664005]  [<ffffffff8114c262>] ? spin_lock+0x9/0xb
[ 1937.664005]  [<ffffffff813092b4>] SyS_connect+0x9/0xb
[ 1937.664005]  [<ffffffff813e56d2>] system_call_fastpath+0x16/0x1b
[ 1937.664005] Code: 10 2a 84 81 e8 65 76 bd e0 65 ff 0c 25 10 bb 00 00 4d 85 ed 74 37 48 8b 85 60 ff ff ff 48 8b 80 88 01 00 00 48 8b b8 10 02 00 00 <48> 8b 47 20 ff 50 20 85 c0 74 0f 83 e8 28 89 83 10 01 00 00 89
[ 1937.664005] RIP  [<ffffffffa049db88>] pppol2tp_connect+0x33d/0x41e [l2tp_ppp]
[ 1937.664005]  RSP <ffff8800c43c7de8>
[ 1937.664005] CR2: 0000000000000020
[ 1939.559375] ---[ end trace 82d44500f28f8708 ]---

Fixes: f34c4a35 ("l2tp: take PMTU from tunnel UDP socket")
Signed-off-by: default avatarGuillaume Nault <g.nault@alphalink.fr>
Acked-by: default avatarEric Dumazet <edumazet@google.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
Cc: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      4290973d
    • Narendra K's avatar
      ixgbevf: Prevent RX/TX statistics getting reset to zero · 9024225c
      Narendra K authored 
      commit 93659763 upstream.

The commit 4197aa7b implements 64 bit
per ring statistics. But the driver resets the 'total_bytes' and
'total_packets' from RX and TX rings in the RX and TX interrupt
handlers to zero. This results in statistics being lost and user space
reporting RX and TX statistics as zero. This patch addresses the
issue by preventing the resetting of RX and TX ring statistics to
zero.
Signed-off-by: default avatarNarendra K <narendra_k@dell.com>
Tested-by: default avatarSibai Li <sibai.li@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
Cc: Weng Meiling <wengmeiling.weng@huawei.com>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      9024225c
    • Benjamin Poirier's avatar
      net: Do not enable tx-nocache-copy by default · 10d5d534
      Benjamin Poirier authored 
      commit cdb3f4a3 upstream.

There are many cases where this feature does not improve performance or even
reduces it.

For example, here are the results from tests that I've run using 3.12.6 on one
Intel Xeon W3565 and one i7 920 connected by ixgbe adapters. The results are
from the Xeon, but they're similar on the i7. All numbers report the
mean±stddev over 10 runs of 10s.

1) latency tests similar to what is described in "c6e1a0d1 net: Allow no-cache
copy from user on transmit"
There is no statistically significant difference between tx-nocache-copy
on/off.
nic irqs spread out (one queue per cpu)

200x netperf -r 1400,1
tx-nocache-copy off
        692000±1000 tps
        50/90/95/99% latency (us): 275±2/643.8±0.4/799±1/2474.4±0.3
tx-nocache-copy on
        693000±1000 tps
        50/90/95/99% latency (us): 274±1/644.1±0.7/800±2/2474.5±0.7

200x netperf -r 14000,14000
tx-nocache-copy off
        86450±80 tps
        50/90/95/99% latency (us): 334.37±0.02/838±1/2100±20/3990±40
tx-nocache-copy on
        86110±60 tps
        50/90/95/99% latency (us): 334.28±0.01/837±2/2110±20/3990±20

2) single stream throughput tests
tx-nocache-copy leads to higher service demand

                        throughput  cpu0        cpu1        demand
                        (Gb/s)      (Gcycle)    (Gcycle)    (cycle/B)

nic irqs and netperf on cpu0 (1x netperf -T0,0 -t omni -- -d send)

tx-nocache-copy off     9402±5      9.4±0.2                 0.80±0.01
tx-nocache-copy on      9403±3      9.85±0.04               0.838±0.004

nic irqs on cpu0, netperf on cpu1 (1x netperf -T1,1 -t omni -- -d send)

tx-nocache-copy off     9401±5      5.83±0.03   5.0±0.1     0.923±0.007
tx-nocache-copy on      9404±2      5.74±0.03   5.523±0.009 0.958±0.002

As a second example, here are some results from Eric Dumazet with latest
net-next.
tx-nocache-copy also leads to higher service demand

(cpu is Intel(R) Xeon(R) CPU X5660  @ 2.80GHz)

lpq83:~# ./ethtool -K eth0 tx-nocache-copy on
lpq83:~# perf stat ./netperf -H lpq84 -c
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to lpq84.prod.google.com () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % U      us/KB   us/KB

 87380  16384  16384    10.00      9407.44   2.50     -1.00    0.522   -1.000

 Performance counter stats for './netperf -H lpq84 -c':

       4282.648396 task-clock                #    0.423 CPUs utilized
             9,348 context-switches          #    0.002 M/sec
                88 CPU-migrations            #    0.021 K/sec
               355 page-faults               #    0.083 K/sec
    11,812,797,651 cycles                    #    2.758 GHz                     [82.79%]
     9,020,522,817 stalled-cycles-frontend   #   76.36% frontend cycles idle    [82.54%]
     4,579,889,681 stalled-cycles-backend    #   38.77% backend  cycles idle    [67.33%]
     6,053,172,792 instructions              #    0.51  insns per cycle
                                             #    1.49  stalled cycles per insn [83.64%]
       597,275,583 branches                  #  139.464 M/sec                   [83.70%]
         8,960,541 branch-misses             #    1.50% of all branches         [83.65%]

      10.128990264 seconds time elapsed

lpq83:~# ./ethtool -K eth0 tx-nocache-copy off
lpq83:~# perf stat ./netperf -H lpq84 -c
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to lpq84.prod.google.com () port 0 AF_INET
Recv   Send    Send                          Utilization       Service Demand
Socket Socket  Message  Elapsed              Send     Recv     Send    Recv
Size   Size    Size     Time     Throughput  local    remote   local   remote
bytes  bytes   bytes    secs.    10^6bits/s  % S      % U      us/KB   us/KB

 87380  16384  16384    10.00      9412.45   2.15     -1.00    0.449   -1.000

 Performance counter stats for './netperf -H lpq84 -c':

       2847.375441 task-clock                #    0.281 CPUs utilized
            11,632 context-switches          #    0.004 M/sec
                49 CPU-migrations            #    0.017 K/sec
               354 page-faults               #    0.124 K/sec
     7,646,889,749 cycles                    #    2.686 GHz                     [83.34%]
     6,115,050,032 stalled-cycles-frontend   #   79.97% frontend cycles idle    [83.31%]
     1,726,460,071 stalled-cycles-backend    #   22.58% backend  cycles idle    [66.55%]
     2,079,702,453 instructions              #    0.27  insns per cycle
                                             #    2.94  stalled cycles per insn [83.22%]
       363,773,213 branches                  #  127.757 M/sec                   [83.29%]
         4,242,732 branch-misses             #    1.17% of all branches         [83.51%]

      10.128449949 seconds time elapsed

CC: Tom Herbert <therbert@google.com>
Signed-off-by: default avatarBenjamin Poirier <bpoirier@suse.de>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      10d5d534
    • Hannes Frederic Sowa's avatar
      ipv6: reuse ip6_frag_id from ip6_ufo_append_data · c58038a1
      Hannes Frederic Sowa authored 
      commit 916e4cf4 upstream.

Currently we generate a new fragmentation id on UFO segmentation. It
is pretty hairy to identify the correct net namespace and dst there.
Especially tunnels use IFF_XMIT_DST_RELEASE and thus have no skb_dst
available at all.

This causes unreliable or very predictable ipv6 fragmentation id
generation while segmentation.

Luckily we already have pregenerated the ip6_frag_id in
ip6_ufo_append_data and can use it here.
Signed-off-by: default avatarHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
[bwh: Backported to 3.2: adjust filename, indentation]
Signed-off-by: default avatarBen Hutchings <ben@decadent.org.uk>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      c58038a1
    • Ben Hutchings's avatar
      rtl8192ce: Fix null dereference in watchdog · 4b7f15c2
      Ben Hutchings authored 
      Dmitry Semyonov reported that after upgrading from 3.2.54 to
3.2.57 the rtl8192ce driver will crash when its interface is brought
up.  The oops message shows:

[ 1833.611397] BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
[ 1833.611455] IP: [<ffffffffa0410c6a>] rtl92ce_update_hal_rate_tbl+0x29/0x4db [rtl8192ce]
...
[ 1833.613326] Call Trace:
[ 1833.613346]  [<ffffffffa02ad9c6>] ? rtl92c_dm_watchdog+0xd0b/0xec9 [rtl8192c_common]
[ 1833.613391]  [<ffffffff8105b5cf>] ? process_one_work+0x161/0x269
[ 1833.613425]  [<ffffffff8105c598>] ? worker_thread+0xc2/0x145
[ 1833.613458]  [<ffffffff8105c4d6>] ? manage_workers.isra.25+0x15b/0x15b
[ 1833.613496]  [<ffffffff8105f6d9>] ? kthread+0x76/0x7e
[ 1833.613527]  [<ffffffff81356b74>] ? kernel_thread_helper+0x4/0x10
[ 1833.613563]  [<ffffffff8105f663>] ? kthread_worker_fn+0x139/0x139
[ 1833.613598]  [<ffffffff81356b70>] ? gs_change+0x13/0x13

Disassembly of rtl92ce_update_hal_rate_tbl() shows that the 'sta'
parameter was null.  None of the changes to the rtlwifi family between
3.2.54 and 3.2.57 seem to directly cause this, and reverting commit
f78bccd7 ('rtlwifi: rtl8192ce: Fix too long disable of IRQs')
doesn't fix it.

rtl92c_dm_watchdog() calls rtl92ce_update_hal_rate_tbl() via
rtl92c_dm_refresh_rate_adaptive_mask(), which does not appear in the
call trace as it was inlined.  That function has been completely
removed upstream which may explain why this crash wasn't seen there.

I'm not sure that it is sensible to completely remove
rtl92c_dm_refresh_rate_adaptive_mask() without making other
compensating changes elsewhere, so try to work around this for 3.2 by
checking for a null pointer in rtl92c_dm_refresh_rate_adaptive_mask()
and then skipping the call to rtl92ce_update_hal_rate_tbl().

References: https://bugs.debian.org/745137
References: https://bugs.debian.org/745462Reported-by: default avatarDmitry Semyonov <linulin@gmail.com>
Signed-off-by: default avatarBen Hutchings <ben@decadent.org.uk>
Cc: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Chaoming Li <chaoming_li@realsil.com.cn>
Cc: Satoshi IWAMOTO <satoshi.iwamoto@nifty.ne.jp>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      4b7f15c2
    • Marcelo Ricardo Leitner's avatar
      ipv4: disable bh while doing route gc · 50f1b3d5
      Marcelo Ricardo Leitner authored 
      Further tests revealed that after moving the garbage collector to a work
queue and protecting it with a spinlock may leave the system prone to
soft lockups if bottom half gets very busy.

It was reproced with a set of firewall rules that REJECTed packets. If
the NIC bottom half handler ends up running on the same CPU that is
running the garbage collector on a very large cache, the garbage
collector will not be able to do its job due to the amount of work
needed for handling the REJECTs and also won't reschedule.

The fix is to disable bottom half during the garbage collecting, as it
already was in the first place (most calls to it came from softirqs).
Signed-off-by: default avatarMarcelo Ricardo Leitner <mleitner@redhat.com>
Acked-by: default avatarHannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: default avatarDavid S. Miller <davem@davemloft.net>
Signed-off-by: default avatarBen Hutchings <ben@decadent.org.uk>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      50f1b3d5
    • Marcelo Ricardo Leitner's avatar
      ipv4: avoid parallel route cache gc executions · b54ca608
      Marcelo Ricardo Leitner authored 
      When rt_intern_hash() has to deal with neighbour cache overflowing,
it triggers the route cache garbage collector in an attempt to free
some references on neighbour entries.

Such call cannot be done async but should also not run in parallel with
an already-running one, so that they don't collapse fighting over the
hash lock entries.

This patch thus blocks parallel executions with spinlocks:
- A call from worker and from rt_intern_hash() are not the same, and
cannot be merged, thus they will wait each other on rt_gc_lock.
- Calls to gc from rt_intern_hash() may happen in parallel but we must
wait for it to finish in order to try again. This dedup and
synchrinozation is then performed by the locking just before calling
__do_rt_garbage_collect().
Signed-off-by: default avatarMarcelo Ricardo Leitner <mleitner@redhat.com>
Acked-by: default avatarHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: default avatarBen Hutchings <ben@decadent.org.uk>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      b54ca608
    • Marcelo Ricardo Leitner's avatar
      ipv4: move route garbage collector to work queue · b6153ead
      Marcelo Ricardo Leitner authored 
      Currently the route garbage collector gets called by dst_alloc() if it
have more entries than the threshold. But it's an expensive call, that
don't really need to be done by then.

Another issue with current way is that it allows running the garbage
collector with the same start parameters on multiple CPUs at once, which
is not optimal. A system may even soft lockup if the cache is big enough
as the garbage collectors will be fighting over the hash lock entries.

This patch thus moves the garbage collector to run asynchronously on a
work queue, much similar to how rt_expire_check runs.

There is one condition left that allows multiple executions, which is
handled by the next patch.
Signed-off-by: default avatarMarcelo Ricardo Leitner <mleitner@redhat.com>
Acked-by: default avatarHannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: default avatarBen Hutchings <ben@decadent.org.uk>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      b6153ead
    • James Bottomley's avatar
      Fix spurious request sense in error handling · 23dd72fe
      James Bottomley authored 
      commit d555a2ab upstream.

We unconditionally execute scsi_eh_get_sense() to make sure all failed
commands that should have sense attached, do.  However, the routine forgets
that some commands, because of the way they fail, will not have any sense code
... we should not bother them with a REQUEST_SENSE command.  Fix this by
testing to see if we actually got a CHECK_CONDITION return and skip asking for
sense if we don't.
Tested-by: default avatarAlan Stern <stern@rowland.harvard.edu>
Signed-off-by: default avatarJames Bottomley <JBottomley@Parallels.com>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      23dd72fe
    • Mikulas Patocka's avatar
      dm crypt: fix access beyond the end of allocated space · be73621b
      Mikulas Patocka authored 
      commit d49ec52f upstream.

The DM crypt target accesses memory beyond allocated space resulting in
a crash on 32 bit x86 systems.

This bug is very old (it dates back to 2.6.25 commit 3a7f6c99 "dm
crypt: use async crypto").  However, this bug was masked by the fact
that kmalloc rounds the size up to the next power of two.  This bug
wasn't exposed until 3.17-rc1 commit 298a9fa0 ("dm crypt: use per-bio
data").  By switching to using per-bio data there was no longer any
padding beyond the end of a dm-crypt allocated memory block.

To minimize allocation overhead dm-crypt puts several structures into one
block allocated with kmalloc.  The block holds struct ablkcipher_request,
cipher-specific scratch pad (crypto_ablkcipher_reqsize(any_tfm(cc))),
struct dm_crypt_request and an initialization vector.

The variable dmreq_start is set to offset of struct dm_crypt_request
within this memory block.  dm-crypt allocates the block with this size:
cc->dmreq_start + sizeof(struct dm_crypt_request) + cc->iv_size.

When accessing the initialization vector, dm-crypt uses the function
iv_of_dmreq, which performs this calculation: ALIGN((unsigned long)(dmreq
+ 1), crypto_ablkcipher_alignmask(any_tfm(cc)) + 1).

dm-crypt allocated "cc->iv_size" bytes beyond the end of dm_crypt_request
structure.  However, when dm-crypt accesses the initialization vector, it
takes a pointer to the end of dm_crypt_request, aligns it, and then uses
it as the initialization vector.  If the end of dm_crypt_request is not
aligned on a crypto_ablkcipher_alignmask(any_tfm(cc)) boundary the
alignment causes the initialization vector to point beyond the allocated
space.

Fix this bug by calculating the variable iv_size_padding and adding it
to the allocated size.

Also correct the alignment of dm_crypt_request.  struct dm_crypt_request
is specific to dm-crypt (it isn't used by the crypto subsystem at all),
so it is aligned on __alignof__(struct dm_crypt_request).

Also align per_bio_data_size on ARCH_KMALLOC_MINALIGN, so that it is
aligned as if the block was allocated with kmalloc.
Reported-by: default avatarKrzysztof Kolasa <kkolasa@winsoft.pl>
Tested-by: default avatarMilan Broz <gmazyland@gmail.com>
Signed-off-by: default avatarMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: default avatarMike Snitzer <snitzer@redhat.com>
[lizf: Backported by Mikulas]
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      be73621b
    • Eric W. Biederman's avatar
      mnt: Only change user settable mount flags in remount · b47d65db
      Eric W. Biederman authored 
      commit a6138db8 upstream.

Kenton Varda <kenton@sandstorm.io> discovered that by remounting a
read-only bind mount read-only in a user namespace the
MNT_LOCK_READONLY bit would be cleared, allowing an unprivileged user
to the remount a read-only mount read-write.

Correct this by replacing the mask of mount flags to preserve
with a mask of mount flags that may be changed, and preserve
all others.   This ensures that any future bugs with this mask and
remount will fail in an easy to detect way where new mount flags
simply won't change.
Acked-by: default avatarSerge E. Hallyn <serge.hallyn@ubuntu.com>
Signed-off-by: default avatar"Eric W. Biederman" <ebiederm@xmission.com>
Cc: Francis Moreau <francis.moro@gmail.com>
Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
      b47d65db
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7