platform/kernel/linux-starfive.git
16 months agoMerge branch 'ptp-adjphase-cleanups'
David S. Miller [Tue, 20 Jun 2023 08:02:33 +0000 (09:02 +0100)]
Merge branch 'ptp-adjphase-cleanups'

Rahul Rameshbabu says:

====================
ptp .adjphase cleanups

The goal of this patch series is to improve documentation of .adjphase, add
a new callback .getmaxphase to enable advertising the max phase offset a
device PHC can support, and support invoking .adjphase from the testptp
kselftest.

Changes:
  v2->v1:
    * Removes arbitrary rule that the PHC servo must restore the frequency
      to the value used in the last .adjfine call if any other PHC
      operation is used after a .adjphase operation.
    * Removes a macro introduced in v1 for adding PTP sysfs device
      attribute nodes using a callback for populating the data.

Link: https://lore.kernel.org/netdev/20230120160609.19160723@kernel.org/
Link: https://lore.kernel.org/netdev/20230510205306.136766-1-rrameshbabu@nvidia.com/
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoptp: ocp: Add .getmaxphase ptp_clock_info callback
Rahul Rameshbabu [Mon, 12 Jun 2023 21:15:00 +0000 (14:15 -0700)]
ptp: ocp: Add .getmaxphase ptp_clock_info callback

Add a function that advertises a maximum offset of zero supported by
ptp_clock_info .adjphase in the OCP null ptp implementation.

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Jonathan Lemon <jonathan.lemon@gmail.com>
Cc: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoptp: idt82p33: Add .getmaxphase ptp_clock_info callback
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:59 +0000 (14:14 -0700)]
ptp: idt82p33: Add .getmaxphase ptp_clock_info callback

Advertise the maximum offset the .adjphase callback is capable of
supporting in nanoseconds for IDT devices.

Refactor the negation of the offset stored in the register to be after the
boundary check of the offset value rather than before. Boundary check based
on the intended value rather than its device-specific representation.
Depend on ptp_clock_adjtime for handling out-of-range offsets.
ptp_clock_adjtime returns -ERANGE instead of clamping out-of-range offsets.

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Min Li <min.li.xe@renesas.com>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoptp: ptp_clockmatrix: Add .getmaxphase ptp_clock_info callback
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:58 +0000 (14:14 -0700)]
ptp: ptp_clockmatrix: Add .getmaxphase ptp_clock_info callback

Advertise the maximum offset the .adjphase callback is capable of
supporting in nanoseconds for IDT ClockMatrix devices. Depend on
ptp_clock_adjtime for handling out-of-range offsets. ptp_clock_adjtime
returns -ERANGE instead of clamping out-of-range offsets.

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Vincent Cheng <vincent.cheng.xh@renesas.com>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agonet/mlx5: Add .getmaxphase ptp_clock_info callback
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:57 +0000 (14:14 -0700)]
net/mlx5: Add .getmaxphase ptp_clock_info callback

Implement .getmaxphase callback of ptp_clock_info in mlx5 driver. No longer
do a range check in .adjphase callback implementation. Handled by the ptp
stack.

Cc: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoptp: Add .getmaxphase callback to ptp_clock_info
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:56 +0000 (14:14 -0700)]
ptp: Add .getmaxphase callback to ptp_clock_info

Enables advertisement of the maximum offset supported by the phase control
functionality of PHCs. The callback is used to return an error if an offset
not supported by the PHC is used in ADJ_OFFSET. The ioctls
PTP_CLOCK_GETCAPS and PTP_CLOCK_GETCAPS2 now advertise the maximum offset a
PHC's phase control functionality is capable of supporting. Introduce new
sysfs node, max_phase_adjustment.

Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Maciek Machnikowski <maciek@machnikowski.net>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agotestptp: Add support for testing ptp_clock_info .adjphase callback
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:55 +0000 (14:14 -0700)]
testptp: Add support for testing ptp_clock_info .adjphase callback

Invoke clock_adjtime syscall with tx.modes set with ADJ_OFFSET when testptp
is invoked with a phase adjustment offset value. Support seconds and
nanoseconds for the offset value.

Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Maciek Machnikowski <maciek@machnikowski.net>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agotestptp: Remove magic numbers related to nanosecond to second conversion
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:54 +0000 (14:14 -0700)]
testptp: Remove magic numbers related to nanosecond to second conversion

Use existing NSEC_PER_SEC declaration in place of hardcoded magic numbers.

Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Maciek Machnikowski <maciek@machnikowski.net>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agodocs: ptp.rst: Add information about NVIDIA Mellanox devices
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:53 +0000 (14:14 -0700)]
docs: ptp.rst: Add information about NVIDIA Mellanox devices

The mlx5_core driver has implemented ptp clock driver functionality but
lacked documentation about the PTP devices. This patch adds information
about the Mellanox device family.

Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoptp: Clarify ptp_clock_info .adjphase expects an internal servo to be used
Rahul Rameshbabu [Mon, 12 Jun 2023 21:14:52 +0000 (14:14 -0700)]
ptp: Clarify ptp_clock_info .adjphase expects an internal servo to be used

.adjphase expects a PHC to use an internal servo algorithm to correct the
provided phase offset target in the callback. Implementation of the
internal servo algorithm are defined by the individual devices.

Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoMerge branch 'ipv6-random-cleanup-for-extension-header'
Jakub Kicinski [Mon, 19 Jun 2023 18:33:08 +0000 (11:33 -0700)]
Merge branch 'ipv6-random-cleanup-for-extension-header'

Kuniyuki Iwashima says:

====================
ipv6: Random cleanup for Extension Header.

This series (1) cleans up pskb_may_pull() in some functions, where needed
data are already pulled by their caller, (2) removes redundant multicast
test, and (3) optimises reload timing of the header.
====================

Link: https://lore.kernel.org/r/20230614230107.22301-1-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoipv6: exthdrs: Remove redundant skb_headlen() check in ip6_parse_tlv().
Kuniyuki Iwashima [Wed, 14 Jun 2023 23:01:07 +0000 (16:01 -0700)]
ipv6: exthdrs: Remove redundant skb_headlen() check in ip6_parse_tlv().

ipv6_destopt_rcv() and ipv6_parse_hopopts() pulls these data

  - Hop-by-Hop/Destination Options Header : 8
  - Hdr Ext Len                           : skb_transport_header(skb)[1] << 3

and calls ip6_parse_tlv(), so it need not check if skb_headlen() is less
than skb_transport_offset(skb) + (skb_transport_header(skb)[1] << 3).

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoipv6: exthdrs: Reload hdr only when needed in ipv6_srh_rcv().
Kuniyuki Iwashima [Wed, 14 Jun 2023 23:01:06 +0000 (16:01 -0700)]
ipv6: exthdrs: Reload hdr only when needed in ipv6_srh_rcv().

We need not reload hdr in ipv6_srh_rcv() unless we call
pskb_expand_head().

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoipv6: exthdrs: Replace pskb_pull() with skb_pull() in ipv6_srh_rcv().
Kuniyuki Iwashima [Wed, 14 Jun 2023 23:01:05 +0000 (16:01 -0700)]
ipv6: exthdrs: Replace pskb_pull() with skb_pull() in ipv6_srh_rcv().

ipv6_rthdr_rcv() pulls these data

  - Segment Routing Header : 8
  - Hdr Ext Len            : skb_transport_header(skb)[1] << 3

needed by ipv6_srh_rcv(), so pskb_pull() in ipv6_srh_rcv() never
fails and can be replaced with skb_pull().

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoipv6: rpl: Remove redundant multicast tests in ipv6_rpl_srh_rcv().
Kuniyuki Iwashima [Wed, 14 Jun 2023 23:01:04 +0000 (16:01 -0700)]
ipv6: rpl: Remove redundant multicast tests in ipv6_rpl_srh_rcv().

ipv6_rpl_srh_rcv() checks if ipv6_hdr(skb)->daddr or ohdr->rpl_segaddr[i]
is the multicast address with ipv6_addr_type().

We have the same check for ipv6_hdr(skb)->daddr in ipv6_rthdr_rcv(), so we
need not recheck it in ipv6_rpl_srh_rcv().

Also, we should use ipv6_addr_is_multicast() for ohdr->rpl_segaddr[i]
instead of ipv6_addr_type().

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoipv6: rpl: Remove pskb(_may)?_pull() in ipv6_rpl_srh_rcv().
Kuniyuki Iwashima [Wed, 14 Jun 2023 23:01:03 +0000 (16:01 -0700)]
ipv6: rpl: Remove pskb(_may)?_pull() in ipv6_rpl_srh_rcv().

As Eric Dumazet pointed out [0], ipv6_rthdr_rcv() pulls these data

  - Segment Routing Header : 8
  - Hdr Ext Len            : skb_transport_header(skb)[1] << 3

needed by ipv6_rpl_srh_rcv().  We can remove pskb_may_pull() and
replace pskb_pull() with skb_pull() in ipv6_rpl_srh_rcv().

Link: https://lore.kernel.org/netdev/CANn89iLboLwLrHXeHJucAqBkEL_S0rJFog68t7wwwXO-aNf5Mg@mail.gmail.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
16 months agoMerge tag 'mlx5-updates-2023-06-16' of git://git.kernel.org/pub/scm/linux/kernel...
David S. Miller [Sun, 18 Jun 2023 17:52:58 +0000 (18:52 +0100)]
Merge tag 'mlx5-updates-2023-06-16' of git://git./linux/kernel/git/saeed/linux

mlx5-updates-2023-06-16

1) Added a new event handler to firmware sync reset, which is used to
   support firmware sync reset flow on smart NIC. Adding this new stage to
   the flow enables the firmware to ensure host PFs unload before ECPFs
   unload, to avoid race of PFs recovery.

2) Debugfs for mlx5 eswitch bridge offloads

3) Added two new counters for vport stats

4) Minor Fixups and cleanups for net-next branch

Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agogro: move the tc_ext comparison to a helper
Jakub Kicinski [Fri, 16 Jun 2023 20:49:39 +0000 (13:49 -0700)]
gro: move the tc_ext comparison to a helper

The double ifdefs (one for the variable declaration and
one around the code) are quite aesthetically displeasing.
Factor this code out into a helper for easier wrapping.

This will become even more ugly when another skb ext
comparison is added in the future.

The resulting machine code looks the same, the compiler
seems to try to use %rax more and some blocks more around
but I haven't spotted minor differences.

Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agonet: phy: at803x: Use devm_regulator_get_enable_optional()
Christophe JAILLET [Sat, 17 Jun 2023 14:24:37 +0000 (16:24 +0200)]
net: phy: at803x: Use devm_regulator_get_enable_optional()

Use devm_regulator_get_enable_optional() instead of hand writing it. It
saves some line of code.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agodt-bindings: net: phy: gpy2xx: more precise description
Michael Walle [Fri, 16 Jun 2023 10:45:57 +0000 (12:45 +0200)]
dt-bindings: net: phy: gpy2xx: more precise description

Mention that the interrupt line is just asserted for a random period of
time, not the entire time.

Suggested-by: Rob Herring <robh@kernel.org>
Signed-off-by: Michael Walle <mwalle@kernel.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agoipv6: also use netdev_hold() in ip6_route_check_nh()
Eric Dumazet [Fri, 16 Jun 2023 08:57:52 +0000 (08:57 +0000)]
ipv6: also use netdev_hold() in ip6_route_check_nh()

In blamed commit, we missed the fact that ip6_validate_gw()
could change dev under us from ip6_route_check_nh()

In this fix, I use GFP_ATOMIC in order to not pass too many additional
arguments to ip6_validate_gw() and ip6_route_check_nh() only
for a rarely used debug feature.

syzbot reported:

refcount_t: decrement hit 0; leaking memory.
WARNING: CPU: 0 PID: 5006 at lib/refcount.c:31 refcount_warn_saturate+0x1d7/0x1f0 lib/refcount.c:31
Modules linked in:
CPU: 0 PID: 5006 Comm: syz-executor403 Not tainted 6.4.0-rc5-syzkaller-01229-g97c5209b3d37 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/27/2023
RIP: 0010:refcount_warn_saturate+0x1d7/0x1f0 lib/refcount.c:31
Code: 05 fb 8e 51 0a 01 e8 98 95 38 fd 0f 0b e9 d3 fe ff ff e8 ac d9 70 fd 48 c7 c7 00 d3 a6 8a c6 05 d8 8e 51 0a 01 e8 79 95 38 fd <0f> 0b e9 b4 fe ff ff 48 89 ef e8 1a d7 c3 fd e9 5c fe ff ff 0f 1f
RSP: 0018:ffffc900039df6b8 EFLAGS: 00010282
RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
RDX: ffff888026d71dc0 RSI: ffffffff814c03b7 RDI: 0000000000000001
RBP: ffff888146a505fc R08: 0000000000000001 R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000001 R12: 1ffff9200073bedc
R13: 00000000ffffffef R14: ffff888146a505fc R15: ffff8880284eb5a8
FS: 0000555556c88300(0000) GS:ffff8880b9800000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000004585c0 CR3: 000000002b1b1000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
__refcount_dec include/linux/refcount.h:344 [inline]
refcount_dec include/linux/refcount.h:359 [inline]
ref_tracker_free+0x539/0x820 lib/ref_tracker.c:236
netdev_tracker_free include/linux/netdevice.h:4097 [inline]
netdev_put include/linux/netdevice.h:4114 [inline]
netdev_put include/linux/netdevice.h:4110 [inline]
fib6_nh_init+0xb96/0x1bd0 net/ipv6/route.c:3624
ip6_route_info_create+0x10f3/0x1980 net/ipv6/route.c:3791
ip6_route_add+0x28/0x150 net/ipv6/route.c:3835
ipv6_route_ioctl+0x3fc/0x570 net/ipv6/route.c:4459
inet6_ioctl+0x246/0x290 net/ipv6/af_inet6.c:569
sock_do_ioctl+0xcc/0x230 net/socket.c:1189
sock_ioctl+0x1f8/0x680 net/socket.c:1306
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__x64_sys_ioctl+0x197/0x210 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]

Fixes: 70f7457ad6d6 ("net: create device lookup API with reference tracking")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: David Ahern <dsahern@kernel.org>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agocrypto: Fix af_alg_sendmsg(MSG_SPLICE_PAGES) sglist limit
David Howells [Thu, 15 Jun 2023 21:08:54 +0000 (22:08 +0100)]
crypto: Fix af_alg_sendmsg(MSG_SPLICE_PAGES) sglist limit

When af_alg_sendmsg() calls extract_iter_to_sg(), it passes MAX_SGL_ENTS as
the maximum number of elements that may be written to, but some of the
elements may already have been used (as recorded in sgl->cur), so
extract_iter_to_sg() may end up overrunning the scatterlist.

Fix this to limit the number of elements to "MAX_SGL_ENTS - sgl->cur".

Note: It probably makes sense in future to alter the behaviour of
extract_iter_to_sg() to stop if "sgtable->nents >= sg_max" instead, but
this is a smaller fix for now.

The bug causes errors looking something like:

BUG: KASAN: slab-out-of-bounds in sg_assign_page include/linux/scatterlist.h:109 [inline]
BUG: KASAN: slab-out-of-bounds in sg_set_page include/linux/scatterlist.h:139 [inline]
BUG: KASAN: slab-out-of-bounds in extract_bvec_to_sg lib/scatterlist.c:1183 [inline]
BUG: KASAN: slab-out-of-bounds in extract_iter_to_sg lib/scatterlist.c:1352 [inline]
BUG: KASAN: slab-out-of-bounds in extract_iter_to_sg+0x17a6/0x1960 lib/scatterlist.c:1339

Fixes: bf63e250c4b1 ("crypto: af_alg: Support MSG_SPLICE_PAGES")
Reported-by: syzbot+6efc50cc1f8d718d6cb7@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/000000000000b2585a05fdeb8379@google.com/
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: syzbot+6efc50cc1f8d718d6cb7@syzkaller.appspotmail.com
cc: Herbert Xu <herbert@gondor.apana.org.au>
cc: "David S. Miller" <davem@davemloft.net>
cc: Eric Dumazet <edumazet@google.com>
cc: Jakub Kicinski <kuba@kernel.org>
cc: Paolo Abeni <pabeni@redhat.com>
cc: Jens Axboe <axboe@kernel.dk>
cc: Matthew Wilcox <willy@infradead.org>
cc: linux-crypto@vger.kernel.org
cc: netdev@vger.kernel.org
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
16 months agotcp: Use per-vma locking for receive zerocopy
Arjun Roy [Fri, 16 Jun 2023 19:34:27 +0000 (12:34 -0700)]
tcp: Use per-vma locking for receive zerocopy

Per-VMA locking allows us to lock a struct vm_area_struct without
taking the process-wide mmap lock in read mode.

Consider a process workload where the mmap lock is taken constantly in
write mode. In this scenario, all zerocopy receives are periodically
blocked during that period of time - though in principle, the memory
ranges being used by TCP are not touched by the operations that need
the mmap write lock. This results in performance degradation.

Now consider another workload where the mmap lock is never taken in
write mode, but there are many TCP connections using receive zerocopy
that are concurrently receiving. These connections all take the mmap
lock in read mode, but this does induce a lot of contention and atomic
ops for this process-wide lock. This results in additional CPU
overhead caused by contending on the cache line for this lock.

However, with per-vma locking, both of these problems can be avoided.

As a test, I ran an RPC-style request/response workload with 4KB
payloads and receive zerocopy enabled, with 100 simultaneous TCP
connections. I measured perf cycles within the
find_tcp_vma/mmap_read_lock/mmap_read_unlock codepath, with and
without per-vma locking enabled.

When using process-wide mmap semaphore read locking, about 1% of
measured perf cycles were within this path. With per-VMA locking, this
value dropped to about 0.45%.

Signed-off-by: Arjun Roy <arjunroy@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agotcp: enforce receive buffer memory limits by allowing the tcp window to shrink
mfreemon@cloudflare.com [Mon, 12 Jun 2023 03:05:24 +0000 (22:05 -0500)]
tcp: enforce receive buffer memory limits by allowing the tcp window to shrink

Under certain circumstances, the tcp receive buffer memory limit
set by autotuning (sk_rcvbuf) is increased due to incoming data
packets as a result of the window not closing when it should be.
This can result in the receive buffer growing all the way up to
tcp_rmem[2], even for tcp sessions with a low BDP.

To reproduce:  Connect a TCP session with the receiver doing
nothing and the sender sending small packets (an infinite loop
of socket send() with 4 bytes of payload with a sleep of 1 ms
in between each send()).  This will cause the tcp receive buffer
to grow all the way up to tcp_rmem[2].

As a result, a host can have individual tcp sessions with receive
buffers of size tcp_rmem[2], and the host itself can reach tcp_mem
limits, causing the host to go into tcp memory pressure mode.

The fundamental issue is the relationship between the granularity
of the window scaling factor and the number of byte ACKed back
to the sender.  This problem has previously been identified in
RFC 7323, appendix F [1].

The Linux kernel currently adheres to never shrinking the window.

In addition to the overallocation of memory mentioned above, the
current behavior is functionally incorrect, because once tcp_rmem[2]
is reached when no remediations remain (i.e. tcp collapse fails to
free up any more memory and there are no packets to prune from the
out-of-order queue), the receiver will drop in-window packets
resulting in retransmissions and an eventual timeout of the tcp
session.  A receive buffer full condition should instead result
in a zero window and an indefinite wait.

In practice, this problem is largely hidden for most flows.  It
is not applicable to mice flows.  Elephant flows can send data
fast enough to "overrun" the sk_rcvbuf limit (in a single ACK),
triggering a zero window.

But this problem does show up for other types of flows.  Examples
are websockets and other type of flows that send small amounts of
data spaced apart slightly in time.  In these cases, we directly
encounter the problem described in [1].

RFC 7323, section 2.4 [2], says there are instances when a retracted
window can be offered, and that TCP implementations MUST ensure
that they handle a shrinking window, as specified in RFC 1122,
section 4.2.2.16 [3].  All prior RFCs on the topic of tcp window
management have made clear that sender must accept a shrunk window
from the receiver, including RFC 793 [4] and RFC 1323 [5].

This patch implements the functionality to shrink the tcp window
when necessary to keep the right edge within the memory limit by
autotuning (sk_rcvbuf).  This new functionality is enabled with
the new sysctl: net.ipv4.tcp_shrink_window

Additional information can be found at:
https://blog.cloudflare.com/unbounded-memory-usage-by-tcp-for-receive-buffers-and-how-we-fixed-it/

[1] https://www.rfc-editor.org/rfc/rfc7323#appendix-F
[2] https://www.rfc-editor.org/rfc/rfc7323#section-2.4
[3] https://www.rfc-editor.org/rfc/rfc1122#page-91
[4] https://www.rfc-editor.org/rfc/rfc793
[5] https://www.rfc-editor.org/rfc/rfc1323

Signed-off-by: Mike Freemon <mfreemon@cloudflare.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agodevlink: report devlink_port_type_warn source device
Petr Oros [Thu, 15 Jun 2023 09:54:47 +0000 (11:54 +0200)]
devlink: report devlink_port_type_warn source device

devlink_port_type_warn is scheduled for port devlink and warning
when the port type is not set. But from this warning it is not easy
found out which device (driver) has no devlink port set.

[ 3709.975552] Type was not set for devlink port.
[ 3709.975579] WARNING: CPU: 1 PID: 13092 at net/devlink/leftover.c:6775 devlink_port_type_warn+0x11/0x20
[ 3709.993967] Modules linked in: openvswitch nf_conncount nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 nfnetlink bluetooth rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace fscache netfs vhost_net vhost vhost_iotlb tap tun bridge stp llc qrtr intel_rapl_msr intel_rapl_common i10nm_edac nfit libnvdimm x86_pkg_temp_thermal mlx5_ib intel_powerclamp coretemp dell_wmi ledtrig_audio sparse_keymap ipmi_ssif kvm_intel ib_uverbs rfkill ib_core video kvm iTCO_wdt acpi_ipmi intel_vsec irqbypass ipmi_si iTCO_vendor_support dcdbas ipmi_devintf mei_me ipmi_msghandler rapl mei intel_cstate isst_if_mmio isst_if_mbox_pci dell_smbios intel_uncore isst_if_common i2c_i801 dell_wmi_descriptor wmi_bmof i2c_smbus intel_pch_thermal pcspkr acpi_power_meter xfs libcrc32c sd_mod sg nvme_tcp mgag200 i2c_algo_bit nvme_fabrics drm_shmem_helper drm_kms_helper nvme syscopyarea ahci sysfillrect sysimgblt nvme_core fb_sys_fops crct10dif_pclmul libahci mlx5_core sfc crc32_pclmul nvme_common drm
[ 3709.994030]  crc32c_intel mtd t10_pi mlxfw libata tg3 mdio megaraid_sas psample ghash_clmulni_intel pci_hyperv_intf wmi dm_multipath sunrpc dm_mirror dm_region_hash dm_log dm_mod be2iscsi bnx2i cnic uio cxgb4i cxgb4 tls libcxgbi libcxgb qla4xxx iscsi_boot_sysfs iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi fuse
[ 3710.108431] CPU: 1 PID: 13092 Comm: kworker/1:1 Kdump: loaded Not tainted 5.14.0-319.el9.x86_64 #1
[ 3710.108435] Hardware name: Dell Inc. PowerEdge R750/0PJ80M, BIOS 1.8.2 09/14/2022
[ 3710.108437] Workqueue: events devlink_port_type_warn
[ 3710.108440] RIP: 0010:devlink_port_type_warn+0x11/0x20
[ 3710.108443] Code: 84 76 fe ff ff 48 c7 03 20 0e 1a ad 31 c0 e9 96 fd ff ff 66 0f 1f 44 00 00 0f 1f 44 00 00 48 c7 c7 18 24 4e ad e8 ef 71 62 ff <0f> 0b c3 cc cc cc cc 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 f6 87
[ 3710.108445] RSP: 0018:ff3b6d2e8b3c7e90 EFLAGS: 00010282
[ 3710.108447] RAX: 0000000000000000 RBX: ff366d6580127080 RCX: 0000000000000027
[ 3710.108448] RDX: 0000000000000027 RSI: 00000000ffff86de RDI: ff366d753f41f8c8
[ 3710.108449] RBP: ff366d658ff5a0c0 R08: ff366d753f41f8c0 R09: ff3b6d2e8b3c7e18
[ 3710.108450] R10: 0000000000000001 R11: 0000000000000023 R12: ff366d753f430600
[ 3710.108451] R13: ff366d753f436900 R14: 0000000000000000 R15: ff366d753f436905
[ 3710.108452] FS:  0000000000000000(0000) GS:ff366d753f400000(0000) knlGS:0000000000000000
[ 3710.108453] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 3710.108454] CR2: 00007f1c57bc74e0 CR3: 000000111d26a001 CR4: 0000000000773ee0
[ 3710.108456] PKRU: 55555554
[ 3710.108457] Call Trace:
[ 3710.108458]  <TASK>
[ 3710.108459]  process_one_work+0x1e2/0x3b0
[ 3710.108466]  ? rescuer_thread+0x390/0x390
[ 3710.108468]  worker_thread+0x50/0x3a0
[ 3710.108471]  ? rescuer_thread+0x390/0x390
[ 3710.108473]  kthread+0xdd/0x100
[ 3710.108477]  ? kthread_complete_and_exit+0x20/0x20
[ 3710.108479]  ret_from_fork+0x1f/0x30
[ 3710.108485]  </TASK>
[ 3710.108486] ---[ end trace 1b4b23cd0c65d6a0 ]---

After patch:
[  402.473064] ice 0000:41:00.0: Type was not set for devlink port.
[  402.473064] ice 0000:41:00.1: Type was not set for devlink port.

Signed-off-by: Petr Oros <poros@redhat.com>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20230615095447.8259-1-poros@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: mctp: remove redundant RTN_UNICAST check
Lin Ma [Thu, 15 Jun 2023 15:22:40 +0000 (23:22 +0800)]
net: mctp: remove redundant RTN_UNICAST check

Current mctp_newroute() contains two exactly same check against
rtm->rtm_type

static int mctp_newroute(...)
{
...
    if (rtm->rtm_type != RTN_UNICAST) { // (1)
        NL_SET_ERR_MSG(extack, "rtm_type must be RTN_UNICAST");
        return -EINVAL;
    }
...
    if (rtm->rtm_type != RTN_UNICAST) // (2)
        return -EINVAL;
...
}

This commits removes the (2) check as it is redundant.

Signed-off-by: Lin Ma <linma@zju.edu.cn>
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Acked-by: Jeremy Kerr <jk@codeconstruct.com.au>
Link: https://lore.kernel.org/r/20230615152240.1749428-1-linma@zju.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonetlink: specs: fixup openvswitch specs for code generation
Donald Hunter [Thu, 15 Jun 2023 15:14:05 +0000 (16:14 +0100)]
netlink: specs: fixup openvswitch specs for code generation

Refine the ovs_* specs to align exactly with the ovs netlink UAPI
definitions to enable code generation.

Signed-off-by: Donald Hunter <donald.hunter@gmail.com>
Link: https://lore.kernel.org/r/20230615151405.77649-1-donald.hunter@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: sched: Remove unused qdisc_l2t()
YueHaibing [Thu, 15 Jun 2023 12:48:10 +0000 (20:48 +0800)]
net: sched: Remove unused qdisc_l2t()

This is unused since switch to psched_l2t_ns().

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230615124810.34020-1-yuehaibing@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agokcm: Fix unnecessary psock unreservation.
David Howells [Thu, 15 Jun 2023 11:32:02 +0000 (12:32 +0100)]
kcm: Fix unnecessary psock unreservation.

kcm_write_msgs() calls unreserve_psock() to release its hold on the
underlying TCP socket if it has run out of things to transmit, but if we
have nothing in the write queue on entry (e.g. because someone did a
zero-length sendmsg), we don't actually go into the transmission loop and
as a consequence don't call reserve_psock().

Fix this by skipping the call to unreserve_psock() if we didn't reserve a
psock.

Fixes: c31a25e1db48 ("kcm: Send multiple frags in one sendmsg()")
Reported-by: syzbot+dd1339599f1840e4cc65@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/000000000000a61ffe05fe0c3d08@google.com/
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: syzbot+dd1339599f1840e4cc65@syzkaller.appspotmail.com
cc: Tom Herbert <tom@herbertland.com>
cc: Tom Herbert <tom@quantonium.net>
cc: Jens Axboe <axboe@kernel.dk>
cc: Matthew Wilcox <willy@infradead.org>
Link: https://lore.kernel.org/r/20787.1686828722@warthog.procyon.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet/mlx5: Remove unused ecpu field from struct mlx5_sf_table
Jiri Pirko [Fri, 2 Jun 2023 13:34:00 +0000 (15:34 +0200)]
net/mlx5: Remove unused ecpu field from struct mlx5_sf_table

"ecpu" field in struct mlx5_sf_table is not used anywhere. Remove it.

Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Add header file for events
Juhee Kang [Mon, 5 Jun 2023 07:51:36 +0000 (16:51 +0900)]
net/mlx5: Add header file for events

Separate the event API defined in the generic mlx5.h header into
a dedicated header. And remove the TODO comment in commit
69c1280b1f3b ("net/mlx5: Device events, Use async events chain").

Signed-off-by: Juhee Kang <claudiajkang@gmail.com>
Reviewed-by: Larysa Zaremba <larysa.zaremba@intel.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: DR, update query of HCA caps for EC VFs
Daniel Jurgens [Tue, 13 Jun 2023 19:30:49 +0000 (22:30 +0300)]
net/mlx5: DR, update query of HCA caps for EC VFs

This change is needed to use EC VFs with metadata based steering.

There was an assumption that vport was equal to function ID. That's
not the case for EC VF functions. Adjust to function ID and set the
ec_vf_function bit accordingly.

Fixes: 9ac0b128248e ("net/mlx5: Update vport caps query/set for EC VFs")
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Fix the macro for accessing EC VF vports
Daniel Jurgens [Tue, 13 Jun 2023 19:26:43 +0000 (22:26 +0300)]
net/mlx5: Fix the macro for accessing EC VF vports

The last value is not set correctly. This results in representors not
being created for all EC VFs when the base value is higher than 0.

Fixes: a7719b29a821 ("net/mlx5: Add management of EC VF vports")
Signed-off-by: Daniel Jurgens <danielj@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: Add local loopback counter to vport stats
Or Har-Toov [Thu, 23 Mar 2023 16:11:50 +0000 (18:11 +0200)]
net/mlx5e: Add local loopback counter to vport stats

Add counter for number of unicast, multicast and broadcast packets/
octets that were loopback.

Signed-off-by: Or Har-Toov <ohartoov@nvidia.com>
Reviewed-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Expose bits for local loopback counter
Or Har-Toov [Thu, 23 Mar 2023 15:52:03 +0000 (17:52 +0200)]
net/mlx5: Expose bits for local loopback counter

Add needed HW bits for querying local loopback counter and the
HCA capability for it.

Signed-off-by: Or Har-Toov <ohartoov@nvidia.com>
Reviewed-by: Avihai Horon <avihaih@nvidia.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: Remove mlx5e_dbg() and msglvl support
Gal Pressman [Sun, 23 Apr 2023 11:29:26 +0000 (14:29 +0300)]
net/mlx5e: Remove mlx5e_dbg() and msglvl support

The msglvl support was implemented using the mlx5e_dbg() macro which is
rarely used in the driver, and is not very useful when you can just use
dynamic debug instead.
Remove mlx5e_dbg() and convert its usages to netdev_dbg().

Signed-off-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: E-Switch, remove redundant else statements
Saeed Mahameed [Wed, 7 Jun 2023 21:12:10 +0000 (14:12 -0700)]
net/mlx5: E-Switch, remove redundant else statements

These else statement blocks are redundant since the if block already
jumps to the function abort label.

Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Reviewed-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
17 months agonet/mlx5: Bridge, expose FDB state via debugfs
Vlad Buslov [Thu, 25 May 2023 13:19:00 +0000 (15:19 +0200)]
net/mlx5: Bridge, expose FDB state via debugfs

For debugging purposes expose offloaded FDB state (flags, counters, etc.)
via debugfs inside 'esw' root directory. Example debugfs file output:

$ cat mlx5/0000\:08\:00.0/esw/bridge/bridge1/fdb
DEV              MAC               VLAN              PACKETS                BYTES              LASTUSE FLAGS
enp8s0f0_1       e4:0a:05:08:00:06    2                    2                  204           4295567112   0x0
enp8s0f0_0       e4:0a:05:08:00:03    2                    3                  278           4295567112   0x0

Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Bridge, pass net device when linking vport to bridge
Vlad Buslov [Fri, 26 May 2023 06:55:15 +0000 (08:55 +0200)]
net/mlx5: Bridge, pass net device when linking vport to bridge

Following patch requires access to additional data in bridge net_device.
Pass the whole structure down the stack instead of adding necessary fields
as function arguments one-by-one.

Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Create eswitch debugfs root directory
Vlad Buslov [Thu, 1 Jun 2023 07:34:35 +0000 (09:34 +0200)]
net/mlx5: Create eswitch debugfs root directory

Following patch in series uses the new directory for bridge FDB debugfs.
The new directory is intended for all future eswitch-specific debugfs
files.

Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Handle sync reset unload event
Moshe Shemesh [Wed, 17 May 2023 13:07:40 +0000 (16:07 +0300)]
net/mlx5: Handle sync reset unload event

Added a new event handler to firmware sync reset, which is used to
support firmware sync reset flow on smart NIC. Adding this new stage to
the flow enables the firmware to ensure host PFs unload before ECPFs
unload, to avoid race of PFs recovery.

If firmware sends sync_reset_unload event to driver the driver should
unload and close all HW resources of the function. Once the driver
finishes unloading part, it can't get any more events from firmware as
event queues are closed, so it polls the reset state field to know when
to continue to next stage of the sync reset flow.

Added capability bit for supporting sync_reset_unload event.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Check DTOR entry value is not zero
Moshe Shemesh [Wed, 14 Jun 2023 03:47:58 +0000 (06:47 +0300)]
net/mlx5: Check DTOR entry value is not zero

The Default Timeout Register (DTOR) provides timeout values to driver
for flows that are device dependent. Zero value for DTOR entry is not
valid and should not be used. In case of reading zero value from DTOR,
the driver should use the hard coded SW default value instead.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Expose timeout for sync reset unload stage
Moshe Shemesh [Wed, 17 May 2023 08:53:50 +0000 (11:53 +0300)]
net/mlx5: Expose timeout for sync reset unload stage

Expose new timoueout in Default Timeouts Register to be used on sync
reset flow running on smart NIC. In this flow the driver should know how
much time to wait from getting unload request till firmware will ask the
PF to continue to next stage of the flow.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Ack on sync_reset_request only if PF can do reset_now
Moshe Shemesh [Sat, 20 May 2023 08:33:14 +0000 (11:33 +0300)]
net/mlx5: Ack on sync_reset_request only if PF can do reset_now

Verify at reset_request stage that PF is capable to do reset_now. In
case PF is not capable, notify the firmware that the sync reset can not
happen and so firmware will abort the sync reset at early stage and will
not send reset_now event to any PF.

Signed-off-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agoip, ip6: Fix splice to raw and ping sockets
David Howells [Wed, 14 Jun 2023 08:04:16 +0000 (09:04 +0100)]
ip, ip6: Fix splice to raw and ping sockets

Splicing to SOCK_RAW sockets may set MSG_SPLICE_PAGES, but in such a case,
__ip_append_data() will call skb_splice_from_iter() to access the 'from'
data, assuming it to point to a msghdr struct with an iter, instead of
using the provided getfrag function to access it.

In the case of raw_sendmsg(), however, this is not the case and 'from' will
point to a raw_frag_vec struct and raw_getfrag() will be the frag-getting
function.  A similar issue may occur with rawv6_sendmsg().

Fix this by ignoring MSG_SPLICE_PAGES if getfrag != ip_generic_getfrag as
ip_generic_getfrag() expects "from" to be a msghdr*, but the other getfrags
don't.  Note that this will prevent MSG_SPLICE_PAGES from being effective
for udplite.

This likely affects ping sockets too.  udplite looks like it should be okay
as it expects "from" to be a msghdr.

Signed-off-by: David Howells <dhowells@redhat.com>
Reported-by: syzbot+d8486855ef44506fd675@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/000000000000ae4cbf05fdeb8349@google.com/
Fixes: 2dc334f1a63a ("splice, net: Use sendmsg(MSG_SPLICE_PAGES) rather than ->sendpage()")
Tested-by: syzbot+d8486855ef44506fd675@syzkaller.appspotmail.com
cc: David Ahern <dsahern@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Matthew Wilcox <willy@infradead.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/1410156.1686729856@warthog.procyon.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agosfc: do not try to call tc functions when CONFIG_SFC_SRIOV=n
Edward Cree [Thu, 15 Jun 2023 21:52:43 +0000 (22:52 +0100)]
sfc: do not try to call tc functions when CONFIG_SFC_SRIOV=n

Functions efx_tc_netdev_event and efx_tc_netevent_event do not exist
 in that case as object files tc_bindings.o and tc_encap_actions.o
 are not built, so the calls to them from ef100_netdev_event and
 ef100_netevent_event cause link errors.
Wrap the corresponding header files (tc_bindings.h, tc_encap_actions.h)
 with #if IS_ENABLED(CONFIG_SFC_SRIOV), and add an #else with static
 inline stubs for these two functions.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202306102026.ISK5JfUQ-lkp@intel.com/
Fixes: 7e5e7d800011 ("sfc: neighbour lookup for TC encap action offload")
Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
Reviewed-by: Martin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agos390/net: lcs: use IS_ENABLED() for kconfig detection
Randy Dunlap [Thu, 15 Jun 2023 22:21:52 +0000 (15:21 -0700)]
s390/net: lcs: use IS_ENABLED() for kconfig detection

When CONFIG_ETHERNET=m or CONFIG_FDDI=m, lcs.s has build errors or
warnings:

../drivers/s390/net/lcs.c:40:2: error: #error Cannot compile lcs.c without some net devices switched on.
   40 | #error Cannot compile lcs.c without some net devices switched on.
../drivers/s390/net/lcs.c: In function 'lcs_startlan_auto':
../drivers/s390/net/lcs.c:1601:13: warning: unused variable 'rc' [-Wunused-variable]
 1601 |         int rc;

Solve this by using IS_ENABLED(CONFIG_symbol) instead of ifdef
CONFIG_symbol. The latter only works for builtin (=y) values
while IS_ENABLED() works for builtin or modular values.

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Alexandra Winter <wintera@linux.ibm.com>
Cc: Wenjia Zhang <wenjia@linux.ibm.com>
Cc: linux-s390@vger.kernel.org
Cc: netdev@vger.kernel.org
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Christian Borntraeger <borntraeger@linux.ibm.com>
Cc: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: ethernet: litex: add support for 64 bit stats
Jisheng Zhang [Wed, 14 Jun 2023 16:20:35 +0000 (00:20 +0800)]
net: ethernet: litex: add support for 64 bit stats

Implement 64 bit per cpu stats to fix the overflow of netdev->stats
on 32 bit platforms. To simplify the code, we use net core
pcpu_sw_netstats infrastructure. One small drawback is some memory
overhead because litex uses just one queue, but we allocate the
counters per cpu.

Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Gabriel Somlo <gsomlo@gmail.com>
Link: https://lore.kernel.org/r/20230614162035.300-1-jszhang@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoMerge branch 'optimize-procedure-of-changing-mac-address-on-interface'
Jakub Kicinski [Fri, 16 Jun 2023 05:54:55 +0000 (22:54 -0700)]
Merge branch 'optimize-procedure-of-changing-mac-address-on-interface'

Piotr Gardocki says:

====================
optimize procedure of changing MAC address on interface

The first patch adds an if statement in core to skip early when
the MAC address is not being changes.
The remaining patches remove such checks from Intel drivers
as they're redundant at this point.
====================

Link: https://lore.kernel.org/r/20230614145302.902301-1-piotrx.gardocki@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoice: remove unnecessary check for old MAC == new MAC
Piotr Gardocki [Wed, 14 Jun 2023 14:53:02 +0000 (16:53 +0200)]
ice: remove unnecessary check for old MAC == new MAC

The check has been moved to core. The ndo_set_mac_address callback
is not being called with new MAC address equal to the old one anymore.

Signed-off-by: Piotr Gardocki <piotrx.gardocki@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoi40e: remove unnecessary check for old MAC == new MAC
Piotr Gardocki [Wed, 14 Jun 2023 14:53:01 +0000 (16:53 +0200)]
i40e: remove unnecessary check for old MAC == new MAC

The check has been moved to core. The ndo_set_mac_address callback
is not being called with new MAC address equal to the old one anymore.

Signed-off-by: Piotr Gardocki <piotrx.gardocki@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: add check for current MAC address in dev_set_mac_address
Piotr Gardocki [Wed, 14 Jun 2023 14:53:00 +0000 (16:53 +0200)]
net: add check for current MAC address in dev_set_mac_address

In some cases it is possible for kernel to come with request
to change primary MAC address to the address that is already
set on the given interface.

Add proper check to return fast from the function in these cases.

An example of such case is adding an interface to bonding
channel in balance-alb mode:
modprobe bonding mode=balance-alb miimon=100 max_bonds=1
ip link set bond0 up
ifenslave bond0 <eth>

Signed-off-by: Piotr Gardocki <piotrx.gardocki@intel.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoeth: fs_enet: fix print format for resource size
Jakub Kicinski [Thu, 15 Jun 2023 03:52:31 +0000 (20:52 -0700)]
eth: fs_enet: fix print format for resource size

Randy reported that linux-next build warns on PowerPC:

drivers/net/ethernet/freescale/fs_enet/mii-fec.c: In function 'fs_enet_mdio_probe':
drivers/net/ethernet/freescale/fs_enet/mii-fec.c:130:50: warning: format '%x' expects argument of type 'unsigned int', but argument 4 has type 'resource_size_t' {aka 'long long unsigned int'} [-Wformat=]
  130 |         snprintf(new_bus->id, MII_BUS_ID_SIZE, "%x", res.start);
      |                                                 ~^   ~~~~~~~~~
      |                                                  |      |
      |                                                  |      resource_size_t {aka long long unsigned int}
      |                                                  unsigned int
      |                                                 %llx

Use the right print format.

Link: https://lore.kernel.org/all/8f9f8d38-d9c7-9f1b-feb0-103d76902d14@infradead.org/
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/r/20230615035231.2184880-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agosplice, net: Fix splice_to_socket() to handle pipe bufs larger than a page
David Howells [Wed, 14 Jun 2023 10:09:48 +0000 (11:09 +0100)]
splice, net: Fix splice_to_socket() to handle pipe bufs larger than a page

splice_to_socket() assumes that a pipe_buffer won't hold more than a single
page of data - but this assumption can be violated by skb_splice_bits()
when it splices from a socket into a pipe.

The problem is that splice_to_socket() doesn't advance the pipe_buffer
length and offset when transcribing from the pipe buf into a bio_vec, so if
the buf is >PAGE_SIZE, it keeps repeating the same initial chunk and
doesn't advance the tail index.  It then subtracts this from "remain" and
overcounts the amount of data to be sent.

The cleanup phase then tries to overclean the pipe, hits an unused pipe buf
and a NULL-pointer dereference occurs.

Fix this by not restricting the bio_vec size to PAGE_SIZE and instead
transcribing the entirety of each pipe_buffer into a single bio_vec and
advancing the tail index if remain hasn't hit zero yet.

Large bio_vecs will then be split up by iterator functions such as
iov_iter_extract_pages().

This resulted in a KASAN report looking like:

general protection fault, probably for non-canonical address 0xdffffc0000000001: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000008-0x000000000000000f]
...
RIP: 0010:pipe_buf_release include/linux/pipe_fs_i.h:203 [inline]
RIP: 0010:splice_to_socket+0xa91/0xe30 fs/splice.c:933

Fixes: 2dc334f1a63a ("splice, net: Use sendmsg(MSG_SPLICE_PAGES) rather than ->sendpage()")
Reported-by: syzbot+f9e28a23426ac3b24f20@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/0000000000000900e905fdeb8e39@google.com/
Tested-by: syzbot+f9e28a23426ac3b24f20@syzkaller.appspotmail.com
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
cc: David Ahern <dsahern@kernel.org>
cc: Jens Axboe <axboe@kernel.dk>
cc: Matthew Wilcox <willy@infradead.org>
cc: Christian Brauner <brauner@kernel.org>
cc: Alexander Viro <viro@zeniv.linux.org.uk>
Link: https://lore.kernel.org/r/1428985.1686737388@warthog.procyon.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agosunvnet: fix sparc64 build error after gso code split
Stephen Rothwell [Tue, 13 Jun 2023 06:46:39 +0000 (16:46 +1000)]
sunvnet: fix sparc64 build error after gso code split

After merging the net-next tree, today's linux-next build (sparc64
defconfig) failed like this:

drivers/net/ethernet/sun/sunvnet_common.c: In function 'vnet_handle_offloads':
drivers/net/ethernet/sun/sunvnet_common.c:1277:16: error: implicit declaration of function 'skb_gso_segment'; did you mean 'skb_gso_reset'? [-Werror=implicit-function-declaration]
 1277 |         segs = skb_gso_segment(skb, dev->features & ~NETIF_F_TSO);
      |                ^~~~~~~~~~~~~~~
      |                skb_gso_reset
drivers/net/ethernet/sun/sunvnet_common.c:1277:14: warning: assignment to 'struct sk_buff *' from 'int' makes pointer from integer without a cast [-Wint-conversion]
 1277 |         segs = skb_gso_segment(skb, dev->features & ~NETIF_F_TSO);
      |              ^

Fixes: d457a0e329b0 ("net: move gso declarations and functions to their own files")
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230613164639.164b2991@canb.auug.org.au
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: ena: Add dynamic recycling mechanism for rx buffers
David Arinzon [Mon, 12 Jun 2023 12:14:48 +0000 (12:14 +0000)]
net: ena: Add dynamic recycling mechanism for rx buffers

The current implementation allocates page-sized rx buffers.
As traffic may consist of different types and sizes of packets,
in various cases, buffers are not fully used.

This change (Dynamic RX Buffers - DRB) uses part of the allocated rx
page needed for the incoming packet, and returns the rest of the
unused page to be used again as an rx buffer for future packets.
A threshold of 2K for unused space has been set in order to declare
whether the remainder of the page can be reused again as an rx buffer.

As a page may be reused, dma_sync_single_for_cpu() is added in order
to sync the memory to the CPU side after it was owned by the HW.
In addition, when the rx page can no longer be reused, it is being
unmapped using dma_page_unmap(), which implicitly syncs and then
unmaps the entire page. In case the kernel still handles the skbs
pointing to the previous buffers from that rx page, it may access
garbage pointers, caused by the implicit sync overwriting them.
The implicit dma sync is removed by replacing dma_page_unmap() with
dma_unmap_page_attrs() with DMA_ATTR_SKIP_CPU_SYNC flag.

The functionality is disabled for XDP traffic to avoid handling
several descriptors per packet.

Signed-off-by: Arthur Kiyanovski <akiyano@amazon.com>
Signed-off-by: Shay Agroskin <shayagr@amazon.com>
Signed-off-by: David Arinzon <darinzon@amazon.com>
Link: https://lore.kernel.org/r/20230612121448.28829-1-darinzon@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: ioctl: Use kernel memory on protocol ioctl callbacks
Breno Leitao [Fri, 9 Jun 2023 15:27:42 +0000 (08:27 -0700)]
net: ioctl: Use kernel memory on protocol ioctl callbacks

Most of the ioctls to net protocols operates directly on userspace
argument (arg). Usually doing get_user()/put_user() directly in the
ioctl callback.  This is not flexible, because it is hard to reuse these
functions without passing userspace buffers.

Change the "struct proto" ioctls to avoid touching userspace memory and
operate on kernel buffers, i.e., all protocol's ioctl callbacks is
adapted to operate on a kernel memory other than on userspace (so, no
more {put,get}_user() and friends being called in the ioctl callback).

This changes the "struct proto" ioctl format in the following way:

    int                     (*ioctl)(struct sock *sk, int cmd,
-                                        unsigned long arg);
+                                        int *karg);

(Important to say that this patch does not touch the "struct proto_ops"
protocols)

So, the "karg" argument, which is passed to the ioctl callback, is a
pointer allocated to kernel space memory (inside a function wrapper).
This buffer (karg) may contain input argument (copied from userspace in
a prep function) and it might return a value/buffer, which is copied
back to userspace if necessary. There is not one-size-fits-all format
(that is I am using 'may' above), but basically, there are three type of
ioctls:

1) Do not read from userspace, returns a result to userspace
2) Read an input parameter from userspace, and does not return anything
  to userspace
3) Read an input from userspace, and return a buffer to userspace.

The default case (1) (where no input parameter is given, and an "int" is
returned to userspace) encompasses more than 90% of the cases, but there
are two other exceptions. Here is a list of exceptions:

* Protocol RAW:
   * cmd = SIOCGETVIFCNT:
     * input and output = struct sioc_vif_req
   * cmd = SIOCGETSGCNT
     * input and output = struct sioc_sg_req
   * Explanation: for the SIOCGETVIFCNT case, userspace passes the input
     argument, which is struct sioc_vif_req. Then the callback populates
     the struct, which is copied back to userspace.

* Protocol RAW6:
   * cmd = SIOCGETMIFCNT_IN6
     * input and output = struct sioc_mif_req6
   * cmd = SIOCGETSGCNT_IN6
     * input and output = struct sioc_sg_req6

* Protocol PHONET:
  * cmd == SIOCPNADDRESOURCE | SIOCPNDELRESOURCE
     * input int (4 bytes)
  * Nothing is copied back to userspace.

For the exception cases, functions sock_sk_ioctl_inout() will
copy the userspace input, and copy it back to kernel space.

The wrapper that prepare the buffer and put the buffer back to user is
sk_ioctl(), so, instead of calling sk->sk_prot->ioctl(), the callee now
calls sk_ioctl(), which will handle all cases.

Signed-off-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Link: https://lore.kernel.org/r/20230609152800.830401-1-leitao@debian.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Jakub Kicinski [Fri, 16 Jun 2023 05:18:58 +0000 (22:18 -0700)]
Merge git://git./linux/kernel/git/netdev/net

Cross-merge networking fixes after downstream PR.

Conflicts:

include/linux/mlx5/driver.h
  617f5db1a626 ("RDMA/mlx5: Fix affinity assignment")
  dc13180824b7 ("net/mlx5: Enable devlink port for embedded cpu VF vports")
https://lore.kernel.org/all/20230613125939.595e50b8@canb.auug.org.au/

tools/testing/selftests/net/mptcp/mptcp_join.sh
  47867f0a7e83 ("selftests: mptcp: join: skip check if MIB counter not supported")
  425ba803124b ("selftests: mptcp: join: support RM_ADDR for used endpoints or not")
  45b1a1227a7a ("mptcp: introduces more address related mibs")
  0639fa230a21 ("selftests: mptcp: add explicit check for new mibs")
https://lore.kernel.org/netdev/20230609-upstream-net-20230610-mptcp-selftests-support-old-kernels-part-3-v1-0-2896fe2ee8a3@tessares.net/

No adjacent changes.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoMerge tag 'net-6.4-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Fri, 16 Jun 2023 04:11:17 +0000 (21:11 -0700)]
Merge tag 'net-6.4-rc7' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from wireless, and netfilter.

  Selftests excluded - we have 58 patches and diff of +442/-199, which
  isn't really small but perhaps with the exception of the WiFi locking
  change it's old(ish) bugs.

  We have no known problems with v6.4.

  The selftest changes are rather large as MPTCP folks try to apply
  Greg's guidance that selftest from torvalds/linux should be able to
  run against stable kernels.

  Last thing I should call out is the DCCP/UDP-lite deprecation notices.
  We are fairly sure those are dead, but if we're wrong reverting them
  back in won't be fun.

  Current release - regressions:

   - wifi:
      - cfg80211: fix double lock bug in reg_wdev_chan_valid()
      - iwlwifi: mvm: spin_lock_bh() to fix lockdep regression

  Current release - new code bugs:

   - handshake: remove fput() that causes use-after-free

  Previous releases - regressions:

   - sched: cls_u32: fix reference counter leak leading to overflow

   - sched: cls_api: fix lockup on flushing explicitly created chain

  Previous releases - always broken:

   - nf_tables: integrate pipapo into commit protocol

   - nf_tables: incorrect error path handling with NFT_MSG_NEWRULE, fix
     dangling pointer on failure

   - ping6: fix send to link-local addresses with VRF

   - sched: act_pedit: parse L3 header for L4 offset, the skb may not
     have the offset saved

   - sched: act_ct: fix promotion of offloaded unreplied tuple

   - sched: refuse to destroy an ingress and clsact Qdiscs if there are
     lockless change operations in flight

   - wifi: mac80211: fix handful of bugs in multi-link operation

   - ipvlan: fix bound dev checking for IPv6 l3s mode

   - eth: enetc: correct the indexes of highest and 2nd highest TCs

   - eth: ice: fix XDP memory leak when NIC is brought up and down

  Misc:

   - add deprecation notices for UDP-lite and DCCP

   - selftests: mptcp: skip tests not supported by old kernels

   - sctp: handle invalid error codes without calling BUG()"

* tag 'net-6.4-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (91 commits)
  dccp: Print deprecation notice.
  udplite: Print deprecation notice.
  octeon_ep: Add missing check for ioremap
  selftests/ptp: Fix timestamp printf format for PTP_SYS_OFFSET
  net: ethernet: stmicro: stmmac: fix possible memory leak in __stmmac_open
  net: tipc: resize nlattr array to correct size
  sfc: fix XDP queues mode with legacy IRQ
  net: macsec: fix double free of percpu stats
  net: lapbether: only support ethernet devices
  MAINTAINERS: add reviewers for SMC Sockets
  s390/ism: Fix trying to free already-freed IRQ by repeated ism_dev_exit()
  net: dsa: felix: fix taprio guard band overflow at 10Mbps with jumbo frames
  net/sched: cls_api: Fix lockup on flushing explicitly created chain
  ice: Fix ice module unload
  net/handshake: remove fput() that causes use-after-free
  selftests: forwarding: hw_stats_l3: Set addrgenmode in a separate step
  net/sched: qdisc_destroy() old ingress and clsact Qdiscs before grafting
  net/sched: Refactor qdisc_graft() for ingress and clsact Qdiscs
  net/sched: act_ct: Fix promotion of offloaded unreplied tuple
  wifi: iwlwifi: mvm: spin_lock_bh() to fix lockdep regression
  ...

17 months agoMerge tag 'loongarch-fixes-6.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 16 Jun 2023 03:56:39 +0000 (20:56 -0700)]
Merge tag 'loongarch-fixes-6.4-1' of git://git./linux/kernel/git/chenhuacai/linux-loongson

Pull LoongArch fixes from Huacai Chen:
 "Some trivial bug fixes for v6.4-rc7"

* tag 'loongarch-fixes-6.4-1' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
  LoongArch: Fix debugfs_create_dir() error checking
  LoongArch: Avoid uninitialized alignment_mask
  LoongArch: Fix perf event id calculation
  LoongArch: Fix the write_fcsr() macro
  LoongArch: Let pmd_present() return true when splitting pmd

17 months agoMerge tag 'for-6.4/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device...
Linus Torvalds [Fri, 16 Jun 2023 03:19:21 +0000 (20:19 -0700)]
Merge tag 'for-6.4/dm-fixes' of git://git./linux/kernel/git/device-mapper/linux-dm

Pull device mapper fixes from Mike Snitzer:

 - Fix DM thinp discard performance regression introduced during this
   merge window where DM core was splitting large discards every 128K
   (max_sectors_kb) rather than every 64M (discard_max_bytes).

 - Extend DM core LOCKFS fix, made during 6.4 merge, to also fix race
   between do_mount and dm's do_suspend (in addition to the earlier
   fix's do_mount race with dm's do_resume).

 - Fix DM thin metadata operations to first check if the thin-pool is in
   "fail_io" mode; otherwise UAF can occur.

 - Fix DM thinp's call to __blkdev_issue_discard to use GFP_NOIO rather
   than GFP_NOWAIT (__blkdev_issue_discard cannot handle NULL return
   from bio_alloc).

* tag 'for-6.4/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
  dm: use op specific max_sectors when splitting abnormal io
  dm thin: fix issue_discard to pass GFP_NOIO to __blkdev_issue_discard
  dm thin metadata: check fail_io before using data_sm
  dm: don't lock fs when the map is NULL during suspend or resume

17 months agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Linus Torvalds [Fri, 16 Jun 2023 03:13:56 +0000 (20:13 -0700)]
Merge tag 'for-linus' of git://git./linux/kernel/git/rdma/rdma

Pull rdma fixes from Jason Gunthorpe:
 "This is an unusually large bunch of bug fixes for the later rc cycle,
  rxe and mlx5 both dumped a lot of things at once. rxe continues to fix
  itself, and mlx5 is fixing a bunch of "queue counters" related bugs.

  There is one highly notable bug fix regarding the qkey. This small
  security check was missed in the original 2005 implementation and it
  allows some significant issues.

  Summary:

   - Two rtrs bug fixes for error unwind bugs

   - Several rxe bug fixes:
      * Incorrect Rx packet validation
      * Using memory without a refcount
      * Syzkaller found use before initialization
      * Regression fix for missing locking with the tasklet conversion
        from this merge window

   - Have bnxt report the correct link properties to userspace, this was
     a regression in v6.3

   - Several mlx5 bug fixes:
      * Kernel crash triggerable by userspace for the RAW ethernet
        profile
      * Defend against steering refcounting issues created by userspace
      * Incorrect change of QP port affinity parameters in some LAG
        configurations

   - Fix mlx5 Q counters:
      * Do not over allocate Q counters to allow userspace to use the
        full port capacity
      * Kernel crash triggered by eswitch due to mis-use of Q counters
      * Incorrect mlx5_device for Q counters in some LAG configurations

   - Properly implement the IBA spec restricting privileged qkeys to
     root

   - Always an error when reading from a disassociated device's event
     queue

   - isert bug fixes:
      * Avoid a deadlock with the CM handler and CM ID destruction
      * Correct list corruption due to incorrect locking
      * Fix a use after free around connection tear down"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/rxe: Fix rxe_cq_post
  IB/isert: Fix incorrect release of isert connection
  IB/isert: Fix possible list corruption in CMA handler
  IB/isert: Fix dead lock in ib_isert
  RDMA/mlx5: Fix affinity assignment
  IB/uverbs: Fix to consider event queue closing also upon non-blocking mode
  RDMA/uverbs: Restrict usage of privileged QKEYs
  RDMA/cma: Always set static rate to 0 for RoCE
  RDMA/mlx5: Fix Q-counters query in LAG mode
  RDMA/mlx5: Remove vport Q-counters dependency on normal Q-counters
  RDMA/mlx5: Fix Q-counters per vport allocation
  RDMA/mlx5: Create an indirect flow table for steering anchor
  RDMA/mlx5: Initiate dropless RQ for RAW Ethernet functions
  RDMA/rxe: Fix the use-before-initialization error of resp_pkts
  RDMA/bnxt_re: Fix reporting active_{speed,width} attributes
  RDMA/rxe: Fix ref count error in check_rkey()
  RDMA/rxe: Fix packet length checks
  RDMA/rtrs: Fix rxe_dealloc_pd warning
  RDMA/rtrs: Fix the last iu->buf leak in err path

17 months agoMerge tag 'spi-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi
Linus Torvalds [Fri, 16 Jun 2023 03:03:15 +0000 (20:03 -0700)]
Merge tag 'spi-fix-v6.4-rc6' of git://git./linux/kernel/git/broonie/spi

Pull spi fixes from Mark Brown:
 "A few more driver specific fixes.

  The DesignWare fix is for an issue introduced by conversion to the
  chip select accessor functions and is pretty important but the other
  two are less severe"

* tag 'spi-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: dw: Replace incorrect spi_get_chipselect with set
  spi: fsl-dspi: avoid SCK glitches with continuous transfers
  spi: cadence-quadspi: Add missing check for dma_set_mask

17 months agoMerge tag 'regulator-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 16 Jun 2023 02:54:58 +0000 (19:54 -0700)]
Merge tag 'regulator-fix-v6.4-rc6' of git://git./linux/kernel/git/broonie/regulator

Pull regulator fix from Mark Brown:
 "The set of regulators described for the Qualcomm PM8550 just seems to
  have been completely wrong and would likely not have worked at all if
  anything tried to actually configure anything except for enabling and
  disabling at runtime"

* tag 'regulator-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:
  regulator: qcom-rpmh: Fix regulators for PM8550

17 months agoMerge tag 'regmap-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 16 Jun 2023 02:50:57 +0000 (19:50 -0700)]
Merge tag 'regmap-fix-v6.4-rc6' of git://git./linux/kernel/git/broonie/regmap

Pull regmap fix from Mark Brown:
 "Another fix for the maple tree cache, Takashi noticed that unlike
  other caches the maple tree cache didn't check for read only registers
  before trying to sync which would result in spurious syncs for read
  only registers where we don't have a default.

  This was due to the check being open coded in the caches, we now check
  in the shared 'does this register need sync' function so that is fixed
  for this and future caches"

* tag 'regmap-fix-v6.4-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regmap:
  regmap: regcache: Don't sync read-only registers

17 months agoMerge tag 'media/v6.4-6' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Fri, 16 Jun 2023 02:13:45 +0000 (19:13 -0700)]
Merge tag 'media/v6.4-6' of git://git./linux/kernel/git/mchehab/linux-media

Pull media fixes from Mauro Carvalho Chehab:
 "A fix for dvb-core to avoid a race condition during DVB board
  registration"

* tag 'media/v6.4-6' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
  Revert "media: dvb-core: Fix use-after-free on race condition at dvb_frontend"

17 months agoMerge tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Thu, 15 Jun 2023 22:40:58 +0000 (15:40 -0700)]
Merge tag 'ext4_for_linus_stable' of git://git./linux/kernel/git/tytso/ext4

Pull ext4 fixes from Ted Ts'o:
 "Fix two regressions in ext4, one report by syzkaller[1], and reported
  by multiple users (and tracked by regzbot[2])"

[1] https://syzkaller.appspot.com/bug?extid=4acc7d910e617b360859
[2] https://linux-regtracking.leemhuis.info/regzbot/regression/ZIauBR7YiV3rVAHL@glitch/

* tag 'ext4_for_linus_stable' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
  ext4: drop the call to ext4_error() from ext4_get_group_info()
  Revert "ext4: remove unnecessary check in ext4_bg_num_gdb_nometa"

17 months agoMerge tag '6.4-rc6-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Thu, 15 Jun 2023 22:24:33 +0000 (15:24 -0700)]
Merge tag '6.4-rc6-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6

Pull smb client fixes from Steve French:
 "Eight, mostly small, smb3 client fixes:

   - important fix for deferred close oops (race with unmount) found
     with xfstest generic/098 to some servers

   - important reconnect fix

   - fix problem with max_credits mount option

   - two multichannel (interface related) fixes

   - one trivial removal of confusing comment

   - two small debugging improvements (to better spot crediting
     problems)"

* tag '6.4-rc6-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: add a warning when the in-flight count goes negative
  cifs: fix lease break oops in xfstest generic/098
  cifs: fix max_credits implementation
  cifs: fix sockaddr comparison in iface_cmp
  smb/client: print "Unknown" instead of bogus link speed value
  cifs: print all credit counters in DebugData
  cifs: fix status checks in cifs_tree_connect
  smb: remove obsolete comment

17 months agoMerge branch 'udplite-dccp-print-deprecation-notice'
Jakub Kicinski [Thu, 15 Jun 2023 22:09:00 +0000 (15:09 -0700)]
Merge branch 'udplite-dccp-print-deprecation-notice'

Kuniyuki Iwashima says:

====================
udplite/dccp: Print deprecation notice.

UDP-Lite is assumed to have no users for 7 years, and DCCP is
orphaned for 7 years too.

Let's add deprecation notice and see if anyone responds to it.
====================

Link: https://lore.kernel.org/r/20230614194705.90673-1-kuniyu@amazon.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agodccp: Print deprecation notice.
Kuniyuki Iwashima [Wed, 14 Jun 2023 19:47:05 +0000 (12:47 -0700)]
dccp: Print deprecation notice.

DCCP was marked as Orphan in the MAINTAINERS entry 2 years ago in commit
054c4610bd05 ("MAINTAINERS: dccp: move Gerrit Renker to CREDITS").  It says
we haven't heard from the maintainer for five years, so DCCP is not well
maintained for 7 years now.

Recently DCCP only receives updates for bugs, and major distros disable it
by default.

Removing DCCP would allow for better organisation of TCP fields to reduce
the number of cache lines hit in the fast path.

Let's add a deprecation notice when DCCP socket is created and schedule its
removal to 2025.

Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoudplite: Print deprecation notice.
Kuniyuki Iwashima [Wed, 14 Jun 2023 19:47:04 +0000 (12:47 -0700)]
udplite: Print deprecation notice.

Recently syzkaller reported a 7-year-old null-ptr-deref [0] that occurs
when a UDP-Lite socket tries to allocate a buffer under memory pressure.

Someone should have stumbled on the bug much earlier if UDP-Lite had been
used in a real app.  Also, we do not always need a large UDP-Lite workload
to hit the bug since UDP and UDP-Lite share the same memory accounting
limit.

Removing UDP-Lite would simplify UDP code removing a bunch of conditionals
in fast path.

Let's add a deprecation notice when UDP-Lite socket is created and schedule
its removal to 2025.

Link: https://lore.kernel.org/netdev/20230523163305.66466-1-kuniyu@amazon.com/
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoocteon_ep: Add missing check for ioremap
Jiasheng Jiang [Thu, 15 Jun 2023 03:34:00 +0000 (11:34 +0800)]
octeon_ep: Add missing check for ioremap

Add check for ioremap() and return the error if it fails in order to
guarantee the success of ioremap().

Fixes: 862cd659a6fb ("octeon_ep: Add driver framework and device initialization")
Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Link: https://lore.kernel.org/r/20230615033400.2971-1-jiasheng@iscas.ac.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoselftests/ptp: Fix timestamp printf format for PTP_SYS_OFFSET
Alex Maftei [Thu, 15 Jun 2023 08:34:04 +0000 (09:34 +0100)]
selftests/ptp: Fix timestamp printf format for PTP_SYS_OFFSET

Previously, timestamps were printed using "%lld.%u" which is incorrect
for nanosecond values lower than 100,000,000 as they're fractional
digits, therefore leading zeros are meaningful.

This patch changes the format strings to "%lld.%09u" in order to add
leading zeros to the nanosecond value.

Fixes: 568ebc5985f5 ("ptp: add the PTP_SYS_OFFSET ioctl to the testptp program")
Fixes: 4ec54f95736f ("ptp: Fix compiler warnings in the testptp utility")
Fixes: 6ab0e475f1f3 ("Documentation: fix misc. warnings")
Signed-off-by: Alex Maftei <alex.maftei@amd.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Link: https://lore.kernel.org/r/20230615083404.57112-1-alex.maftei@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: ethernet: stmicro: stmmac: fix possible memory leak in __stmmac_open
Christian Marangi [Wed, 14 Jun 2023 09:17:14 +0000 (11:17 +0200)]
net: ethernet: stmicro: stmmac: fix possible memory leak in __stmmac_open

Fix a possible memory leak in __stmmac_open when stmmac_init_phy fails.
It's also needed to free everything allocated by stmmac_setup_dma_desc
and not just the dma_conf struct.

Drop free_dma_desc_resources from __stmmac_open and correctly call
free_dma_desc_resources on each user of __stmmac_open on error.

Reported-by: Jose Abreu <Jose.Abreu@synopsys.com>
Fixes: ba39b344e924 ("net: ethernet: stmicro: stmmac: generate stmmac dma conf before open")
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Cc: stable@vger.kernel.org
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Jose Abreu <Jose.Abreu@synopsys.com>
Link: https://lore.kernel.org/r/20230614091714.15912-1-ansuelsmth@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet: tipc: resize nlattr array to correct size
Lin Ma [Wed, 14 Jun 2023 12:06:04 +0000 (20:06 +0800)]
net: tipc: resize nlattr array to correct size

According to nla_parse_nested_deprecated(), the tb[] is supposed to the
destination array with maxtype+1 elements. In current
tipc_nl_media_get() and __tipc_nl_media_set(), a larger array is used
which is unnecessary. This patch resize them to a proper size.

Fixes: 1e55417d8fc6 ("tipc: add media set to new netlink api")
Fixes: 46f15c6794fb ("tipc: add media get/dump to new netlink api")
Signed-off-by: Lin Ma <linma@zju.edu.cn>
Reviewed-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Tung Nguyen <tung.q.nguyen@dektech.com.au>
Link: https://lore.kernel.org/r/20230614120604.1196377-1-linma@zju.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agodm: use op specific max_sectors when splitting abnormal io
Mike Snitzer [Thu, 15 Jun 2023 01:47:46 +0000 (21:47 -0400)]
dm: use op specific max_sectors when splitting abnormal io

Split abnormal IO in terms of the corresponding operation specific
max_sectors (max_discard_sectors, max_secure_erase_sectors or
max_write_zeroes_sectors).

This fixes a significant dm-thinp discard performance regression that
was introduced with commit e2dd8aca2d76 ("dm bio prison v1: improve
concurrent IO performance"). Relative to discard: max_discard_sectors
is used instead of max_sectors; which fixes excessive discard splitting
(e.g. max_sectors=128K vs max_discard_sectors=64M).

Tested by discarding an 1 Petabyte dm-thin device:
lvcreate -V 1125899906842624B -T test/pool -n thin
time blkdiscard /dev/test/thin

Before this fix (splitting discards every 128K): ~116m
 After this fix (splitting discards every 64M) : 0m33.460s

Reported-by: Zorro Lang <zlang@redhat.com>
Fixes: 06961c487a33 ("dm: split discards further if target sets max_discard_granularity")
Requires: 13f6facf3fae ("dm: allow targets to require splitting WRITE_ZEROES and SECURE_ERASE")
Fixes: e2dd8aca2d76 ("dm bio prison v1: improve concurrent IO performance")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
17 months agodm thin: fix issue_discard to pass GFP_NOIO to __blkdev_issue_discard
Mike Snitzer [Wed, 14 Jun 2023 00:05:34 +0000 (20:05 -0400)]
dm thin: fix issue_discard to pass GFP_NOIO to __blkdev_issue_discard

issue_discard() passes GFP_NOWAIT to __blkdev_issue_discard() despite
its code assuming bio_alloc() always succeeds.

Commit 3dba53a958a75 ("dm thin: use __blkdev_issue_discard for async
discard support") clearly shows where things went bad:

Before commit 3dba53a958a75, dm-thin.c's open-coded
__blkdev_issue_discard_async() properly handled using GFP_NOWAIT.
Unfortunately __blkdev_issue_discard() doesn't and it was missed
during review.

Cc: stable@vger.kernel.org
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
17 months agodm thin metadata: check fail_io before using data_sm
Li Lingfeng [Tue, 6 Jun 2023 12:20:24 +0000 (20:20 +0800)]
dm thin metadata: check fail_io before using data_sm

Must check pmd->fail_io before using pmd->data_sm since
pmd->data_sm may be destroyed by other processes.

       P1(kworker)                             P2(message)
do_worker
 process_prepared
  process_prepared_discard_passdown_pt2
   dm_pool_dec_data_range
                                    pool_message
                                     commit
                                      dm_pool_commit_metadata
                                        ↓
                                       // commit failed
                                      metadata_operation_failed
                                       abort_transaction
                                        dm_pool_abort_metadata
                                         __open_or_format_metadata
                                           ↓
                                          dm_sm_disk_open
                                            ↓
                                           // open failed
                                           // pmd->data_sm is NULL
    dm_sm_dec_blocks
      ↓
     // try to access pmd->data_sm --> UAF

As shown above, if dm_pool_commit_metadata() and
dm_pool_abort_metadata() fail in pool_message process, kworker may
trigger UAF.

Fixes: be500ed721a6 ("dm space maps: improve performance with inc/dec on ranges of blocks")
Cc: stable@vger.kernel.org
Signed-off-by: Li Lingfeng <lilingfeng3@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
17 months agodm: don't lock fs when the map is NULL during suspend or resume
Li Lingfeng [Thu, 1 Jun 2023 06:14:23 +0000 (14:14 +0800)]
dm: don't lock fs when the map is NULL during suspend or resume

As described in commit 38d11da522aa ("dm: don't lock fs when the map is
NULL in process of resume"), a deadlock may be triggered between
do_resume() and do_mount().

This commit preserves the fix from commit 38d11da522aa but moves it to
where it also serves to fix a similar deadlock between do_suspend()
and do_mount().  It does so, if the active map is NULL, by clearing
DM_SUSPEND_LOCKFS_FLAG in dm_suspend() which is called by both
do_suspend() and do_resume().

Fixes: 38d11da522aa ("dm: don't lock fs when the map is NULL in process of resume")
Signed-off-by: Li Lingfeng <lilingfeng3@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
17 months agoleds: trigger: netdev: uninitialized variable in netdev_trig_activate()
Dan Carpenter [Wed, 14 Jun 2023 07:03:59 +0000 (10:03 +0300)]
leds: trigger: netdev: uninitialized variable in netdev_trig_activate()

The qca8k_cled_hw_control_get() function which implements ->hw_control_get
sets the appropriate bits but does not clear them.  This leads to an
uninitialized variable bug.  Fix this by setting mode to zero at the
start.

Fixes: e0256648c831 ("net: dsa: qca8k: implement hw_control ops")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Acked-by: Lee Jones <lee@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agotools: ynl: work around stale system headers
Jakub Kicinski [Wed, 14 Jun 2023 00:28:00 +0000 (17:28 -0700)]
tools: ynl: work around stale system headers

The inability to include the uAPI headers directly in tools/
is one of the bigger annoyances of compiling user space code.
Most projects trade the pain for smaller inconvenience of having
to copy the headers under tools/include.

In case of netlink headers I think that we can avoid both.
Netlink family headers are simple and should be self-contained.
We can try to twiddle the Makefile a little to force-include
just the family header, and use system headers for the rest.

This works fairly well. There are two warts - for some reason
if we specify -include $path/family.h as a compilation flag,
the #ifdef header guard does not seem to work. So we need
to throw the guard in on the command line as well. Seems like
GCC detects that the header is different and tries to include
both. Second problem is that make wants hash sign to be escaped
or not depending on the version. Sigh.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agosfc: fix XDP queues mode with legacy IRQ
Íñigo Huguet [Tue, 13 Jun 2023 13:38:54 +0000 (15:38 +0200)]
sfc: fix XDP queues mode with legacy IRQ

In systems without MSI-X capabilities, xdp_txq_queues_mode is calculated
in efx_allocate_msix_channels, but when enabling MSI-X fails, it was not
changed to a proper default value. This was leading to the driver
thinking that it has dedicated XDP queues, when it didn't.

Fix it by setting xdp_txq_queues_mode to the correct value if the driver
fallbacks to MSI or legacy IRQ mode. The correct value is
EFX_XDP_TX_QUEUES_BORROWED because there are no XDP dedicated queues.

The issue can be easily visible if the kernel is started with pci=nomsi,
then a call trace is shown. It is not shown only with sfc's modparam
interrupt_mode=2. Call trace example:
 WARNING: CPU: 2 PID: 663 at drivers/net/ethernet/sfc/efx_channels.c:828 efx_set_xdp_channels+0x124/0x260 [sfc]
 [...skip...]
 Call Trace:
  <TASK>
  efx_set_channels+0x5c/0xc0 [sfc]
  efx_probe_nic+0x9b/0x15a [sfc]
  efx_probe_all+0x10/0x1a2 [sfc]
  efx_pci_probe_main+0x12/0x156 [sfc]
  efx_pci_probe_post_io+0x18/0x103 [sfc]
  efx_pci_probe.cold+0x154/0x257 [sfc]
  local_pci_probe+0x42/0x80

Fixes: 6215b608a8c4 ("sfc: last resort fallback for lack of xdp tx queues")
Reported-by: Yanghang Liu <yanghliu@redhat.com>
Signed-off-by: Íñigo Huguet <ihuguet@redhat.com>
Acked-by: Martin Habets <habetsm.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: macsec: fix double free of percpu stats
Fedor Pchelkin [Tue, 13 Jun 2023 19:22:20 +0000 (22:22 +0300)]
net: macsec: fix double free of percpu stats

Inside macsec_add_dev() we free percpu macsec->secy.tx_sc.stats and
macsec->stats on some of the memory allocation failure paths. However, the
net_device is already registered to that moment: in macsec_newlink(), just
before calling macsec_add_dev(). This means that during unregister process
its priv_destructor - macsec_free_netdev() - will be called and will free
the stats again.

Remove freeing percpu stats inside macsec_add_dev() because
macsec_free_netdev() will correctly free the already allocated ones. The
pointers to unallocated stats stay NULL, and free_percpu() treats that
correctly.

Found by Linux Verification Center (linuxtesting.org) with Syzkaller.

Fixes: 0a28bfd4971f ("net/macsec: Add MACsec skb_metadata_dst Tx Data path support")
Fixes: c09440f7dcb3 ("macsec: introduce IEEE 802.1AE driver")
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: tls: make the offload check helper take skb not socket
Jakub Kicinski [Tue, 13 Jun 2023 20:50:06 +0000 (13:50 -0700)]
net: tls: make the offload check helper take skb not socket

All callers of tls_is_sk_tx_device_offloaded() currently do
an equivalent of:

 if (skb->sk && tls_is_skb_tx_device_offloaded(skb->sk))

Have the helper accept skb and do the skb->sk check locally.
Two drivers have local static inlines with similar wrappers
already.

While at it change the ifdef condition to TLS_DEVICE.
Only TLS_DEVICE selects SOCK_VALIDATE_XMIT, so the two are
equivalent. This makes removing the duplicated IS_ENABLED()
check in funeth more obviously correct.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Acked-by: Maxim Mikityanskiy <maxtram95@gmail.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Acked-by: Dimitris Michailidis <dmichail@fungible.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: lapbether: only support ethernet devices
Eric Dumazet [Wed, 14 Jun 2023 16:18:02 +0000 (16:18 +0000)]
net: lapbether: only support ethernet devices

It probbaly makes no sense to support arbitrary network devices
for lapbether.

syzbot reported:

skbuff: skb_under_panic: text:ffff80008934c100 len:44 put:40 head:ffff0000d18dd200 data:ffff0000d18dd1ea tail:0x16 end:0x140 dev:bond1
kernel BUG at net/core/skbuff.c:200 !
Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP
Modules linked in:
CPU: 0 PID: 5643 Comm: dhcpcd Not tainted 6.4.0-rc5-syzkaller-g4641cff8e810 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/25/2023
pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : skb_panic net/core/skbuff.c:196 [inline]
pc : skb_under_panic+0x13c/0x140 net/core/skbuff.c:210
lr : skb_panic net/core/skbuff.c:196 [inline]
lr : skb_under_panic+0x13c/0x140 net/core/skbuff.c:210
sp : ffff8000973b7260
x29: ffff8000973b7270 x28: ffff8000973b7360 x27: dfff800000000000
x26: ffff0000d85d8150 x25: 0000000000000016 x24: ffff0000d18dd1ea
x23: ffff0000d18dd200 x22: 000000000000002c x21: 0000000000000140
x20: 0000000000000028 x19: ffff80008934c100 x18: ffff8000973b68a0
x17: 0000000000000000 x16: ffff80008a43bfbc x15: 0000000000000202
x14: 0000000000000000 x13: 0000000000000001 x12: 0000000000000001
x11: 0000000000000201 x10: 0000000000000000 x9 : f22f7eb937cced00
x8 : f22f7eb937cced00 x7 : 0000000000000001 x6 : 0000000000000001
x5 : ffff8000973b6b78 x4 : ffff80008df9ee80 x3 : ffff8000805974f4
x2 : 0000000000000001 x1 : 0000000100000201 x0 : 0000000000000086
Call trace:
skb_panic net/core/skbuff.c:196 [inline]
skb_under_panic+0x13c/0x140 net/core/skbuff.c:210
skb_push+0xf0/0x108 net/core/skbuff.c:2409
ip6gre_header+0xbc/0x738 net/ipv6/ip6_gre.c:1383
dev_hard_header include/linux/netdevice.h:3137 [inline]
lapbeth_data_transmit+0x1c4/0x298 drivers/net/wan/lapbether.c:257
lapb_data_transmit+0x8c/0xb0 net/lapb/lapb_iface.c:447
lapb_transmit_buffer+0x178/0x204 net/lapb/lapb_out.c:149
lapb_send_control+0x220/0x320 net/lapb/lapb_subr.c:251
lapb_establish_data_link+0x94/0xec
lapb_device_event+0x348/0x4e0
notifier_call_chain+0x1a4/0x510 kernel/notifier.c:93
raw_notifier_call_chain+0x3c/0x50 kernel/notifier.c:461
__dev_notify_flags+0x2bc/0x544
dev_change_flags+0xd0/0x15c net/core/dev.c:8643
devinet_ioctl+0x858/0x17e4 net/ipv4/devinet.c:1150
inet_ioctl+0x2ac/0x4d8 net/ipv4/af_inet.c:979
sock_do_ioctl+0x134/0x2dc net/socket.c:1201
sock_ioctl+0x4ec/0x858 net/socket.c:1318
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__arm64_sys_ioctl+0x14c/0x1c8 fs/ioctl.c:856
__invoke_syscall arch/arm64/kernel/syscall.c:38 [inline]
invoke_syscall+0x98/0x2c0 arch/arm64/kernel/syscall.c:52
el0_svc_common+0x138/0x244 arch/arm64/kernel/syscall.c:142
do_el0_svc+0x64/0x198 arch/arm64/kernel/syscall.c:191
el0_svc+0x4c/0x160 arch/arm64/kernel/entry-common.c:647
el0t_64_sync_handler+0x84/0xfc arch/arm64/kernel/entry-common.c:665
el0t_64_sync+0x190/0x194 arch/arm64/kernel/entry.S:591
Code: aa1803e6 aa1903e7 a90023f5 947730f5 (d4210000)

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Martin Schiller <ms@dev.tdt.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoMAINTAINERS: add reviewers for SMC Sockets
Jan Karcher [Wed, 14 Jun 2023 06:54:56 +0000 (08:54 +0200)]
MAINTAINERS: add reviewers for SMC Sockets

adding three people from Alibaba as reviewers for SMC.
They are currently working on improving SMC on other architectures than
s390 and help with reviewing patches on top.

Thank you D. Wythe, Tony Lu and Wen Gu for your contributions and
collaboration and welcome on board as reviewers!

Reviewed-by: Wenjia Zhang <wenjia@linux.ibm.com>
Signed-off-by: Jan Karcher <jaka@linux.ibm.com>
Acked-by: Tony Lu <tonylu@linux.alibaba.com>
Acked-by: Wen Gu <guwen@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agos390/ism: Fix trying to free already-freed IRQ by repeated ism_dev_exit()
Julian Ruess [Tue, 13 Jun 2023 12:25:37 +0000 (14:25 +0200)]
s390/ism: Fix trying to free already-freed IRQ by repeated ism_dev_exit()

This patch prevents the system from crashing when unloading the ISM module.

How to reproduce: Attach an ISM device and execute 'rmmod ism'.

Error-Log:
- Trying to free already-free IRQ 0
- WARNING: CPU: 1 PID: 966 at kernel/irq/manage.c:1890 free_irq+0x140/0x540

After calling ism_dev_exit() for each ISM device in the exit routine,
pci_unregister_driver() will execute ism_remove() for each ISM device.
Because ism_remove() also calls ism_dev_exit(),
free_irq(pci_irq_vector(pdev, 0), ism) is called twice for each ISM
device. This results in a crash with the error
'Trying to free already-free IRQ'.

In the exit routine, it is enough to call pci_unregister_driver()
because it ensures that ism_dev_exit() is called once per
ISM device.

Cc: <stable@vger.kernel.org> # 6.3+
Fixes: 89e7d2ba61b7 ("net/ism: Add new API for client registration")
Reviewed-by: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Julian Ruess <julianr@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoMerge branch 'macb-partial-store-and-forward'
David S. Miller [Thu, 15 Jun 2023 07:45:29 +0000 (08:45 +0100)]
Merge branch 'macb-partial-store-and-forward'

Pranavi Somisetty says:

====================
Add support for partial store and forward

Add support for partial store and forward mode in Cadence MACB.

Link for v1:
https://lore.kernel.org/all/20221213121245.13981-1-pranavi.somisetty@amd.com/

Changes v2:
1. Removed all the changes related to validating FCS when Rx checksum
offload is disabled.
2. Instead of using a platform dependent number (0xFFF) for the reset
value of rx watermark, derive it from designcfg_debug2 register.
3. Added a check to see if partial s/f is supported, by reading the
designcfg_debug6 register.
4. Added devicetree bindings for "rx-watermark" property.
Link for v2:
https://lore.kernel.org/all/20230511071214.18611-1-pranavi.somisetty@amd.com/

Changes v3:
1. Fixed DT schema error: "scalar properties shouldn't have array keywords"
2. Modified description of rx-watermark in to include units of the watermark value
3. Modified the DT property name corresponding to rx_watermark in pbuf_rxcutthru to
"cdns,rx-watermark".
4. Followed reverse christmas tree pattern in declaring variables.
5. Return -EINVAL when an invalid watermark value is set.
6. Removed netdev_info when partial store and forward is not enabled.
7. Validating the rx-watermark value in probe itself and only write to the register
in init.
8. Writing a reset value to the pbuf_cuthru register before disabing partial store
and forward is redundant. So removing it.
9. Removed the platform caps flag.
10. Instead of reading rx-watermark from DT in macb_configure_caps,
reading it in probe.
11. Changed Signed-Off-By and author names on the macb driver patch.
Link for v3:
https://lore.kernel.org/all/20230530095138.1302-1-pranavi.somisetty@amd.com/

Changes v4:
1. Modified description for "rx-watermark" property in the DT bindings.
2. Changed the width of the rx-watermark property to uint32.
3. Removed redundant code and unused variables.
4. When the rx-watermark value is invalid, instead of returning EINVAL,
do not enable partial store and forward.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: macb: Add support for partial store and forward
Maulik Jodhani [Tue, 13 Jun 2023 05:43:40 +0000 (23:43 -0600)]
net: macb: Add support for partial store and forward

When the receive partial store and forward mode is activated, the
receiver will only begin to forward the packet to the external AHB
or AXI slave when enough packet data is stored in the packet buffer.
The amount of packet data required to activate the forwarding process
is programmable via watermark registers which are located at the same
address as the partial store and forward enable bits. Adding support to
read this rx-watermark value from device-tree, to program the watermark
registers and enable partial store and forwarding.

Signed-off-by: Maulik Jodhani <maulik.jodhani@xilinx.com>
Signed-off-by: Pranavi Somisetty <pranavi.somisetty@amd.com>
Reviewed-by: Claudiu Beznea <claudiu.beznea@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agodt-bindings: net: cdns,macb: Add rx-watermark property
Pranavi Somisetty [Tue, 13 Jun 2023 05:43:39 +0000 (23:43 -0600)]
dt-bindings: net: cdns,macb: Add rx-watermark property

watermark value is the minimum amount of packet data
required to activate the forwarding process. The watermark
implementation and maximum size is dependent on the device
where Cadence MACB/GEM is used.

Signed-off-by: Pranavi Somisetty <pranavi.somisetty@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoMerge branch 'netdev-tracking'
David S. Miller [Thu, 15 Jun 2023 07:21:12 +0000 (08:21 +0100)]
Merge branch 'netdev-tracking'

Jakub Kicinski says:

====================
net: create device lookup API with reference tracking

We still see dev_hold() / dev_put() calls without reference tracker
getting added in new code. dev_get_by_name() / dev_get_by_index()
seem to be one of the sources of those. Provide appropriate helpers.
Allocating the tracker can obviously be done with an additional call
to netdev_tracker_alloc(), but a single API feels cleaner.

v2:
 - fix a dev_put() in ethtool
v1: https://lore.kernel.org/all/20230609183207.1466075-1-kuba@kernel.org/
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonetpoll: allocate netdev tracker right away
Jakub Kicinski [Mon, 12 Jun 2023 21:49:44 +0000 (14:49 -0700)]
netpoll: allocate netdev tracker right away

Commit 5fa5ae605821 ("netpoll: add net device refcount tracker to struct netpoll")
was part of one of the initial netdev tracker introduction patches.
It added an explicit netdev_tracker_alloc() for netpoll, presumably
because the flow of the function is somewhat odd.
After most of the core networking stack was converted to use
the tracking hold() variants, netpoll's call to old dev_hold()
stands out a bit.

np is allocated by the caller and ready to use, we can use
netdev_hold() here, even tho np->ndev will only be set to
ndev inside __netpoll_setup().

Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agonet: create device lookup API with reference tracking
Jakub Kicinski [Mon, 12 Jun 2023 21:49:43 +0000 (14:49 -0700)]
net: create device lookup API with reference tracking

New users of dev_get_by_index() and dev_get_by_name() keep
getting added and it would be nice to steer them towards
the APIs with reference tracking.

Add variants of those calls which allocate the reference
tracker and use them in a couple of places.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoLoongArch: Fix debugfs_create_dir() error checking
Immad Mir [Thu, 15 Jun 2023 06:35:56 +0000 (14:35 +0800)]
LoongArch: Fix debugfs_create_dir() error checking

The debugfs_create_dir() returns ERR_PTR in case of an error and the
correct way of checking it is using the IS_ERR_OR_NULL inline function
rather than the simple null comparision. This patch fixes the issue.

Cc: stable@vger.kernel.org
Suggested-By: Ivan Orlov <ivan.orlov0322@gmail.com>
Signed-off-by: Immad Mir <mirimmad17@gmail.com>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
17 months agoLoongArch: Avoid uninitialized alignment_mask
Qing Zhang [Thu, 15 Jun 2023 06:35:52 +0000 (14:35 +0800)]
LoongArch: Avoid uninitialized alignment_mask

The hardware monitoring points for instruction fetching and load/store
operations need to align 4 bytes and 1/2/4/8 bytes respectively.

Reported-by: Colin King <colin.i.king@gmail.com>
Signed-off-by: Qing Zhang <zhangqing@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
17 months agoLoongArch: Fix perf event id calculation
Huacai Chen [Thu, 15 Jun 2023 06:35:52 +0000 (14:35 +0800)]
LoongArch: Fix perf event id calculation

LoongArch PMCFG has 10bit event id rather than 8 bit, so fix it.

Cc: stable@vger.kernel.org
Signed-off-by: Jun Yi <yijun@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
17 months agoLoongArch: Fix the write_fcsr() macro
Qi Hu [Thu, 15 Jun 2023 06:35:52 +0000 (14:35 +0800)]
LoongArch: Fix the write_fcsr() macro

The "write_fcsr()" macro uses wrong the positions for val and dest in
asm. Fix it!

Reported-by: Miao HAO <haomiao19@mails.ucas.ac.cn>
Signed-off-by: Qi Hu <huqi@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
17 months agoLoongArch: Let pmd_present() return true when splitting pmd
Hongchen Zhang [Thu, 15 Jun 2023 06:35:52 +0000 (14:35 +0800)]
LoongArch: Let pmd_present() return true when splitting pmd

When we split a pmd into ptes, pmd_present() and pmd_trans_huge() should
return true, otherwise it would be treated as a swap pmd.

This is the same as arm64 does in commit b65399f6111b ("arm64/mm: Change
THP helpers to comply with generic MM semantics"), we also add a new bit
named _PAGE_PRESENT_INVALID for LoongArch.

Signed-off-by: Hongchen Zhang <zhanghongchen@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
17 months agonet: dsa: felix: fix taprio guard band overflow at 10Mbps with jumbo frames
Vladimir Oltean [Tue, 13 Jun 2023 17:09:07 +0000 (20:09 +0300)]
net: dsa: felix: fix taprio guard band overflow at 10Mbps with jumbo frames

The DEV_MAC_MAXLEN_CFG register contains a 16-bit value - up to 65535.
Plus 2 * VLAN_HLEN (4), that is up to 65543.

The picos_per_byte variable is the largest when "speed" is lowest -
SPEED_10 = 10. In that case it is (1000000L * 8) / 10 = 800000.

Their product - 52434400000 - exceeds 32 bits, which is a problem,
because apparently, a multiplication between two 32-bit factors is
evaluated as 32-bit before being assigned to a 64-bit variable.
In fact it's a problem for any MTU value larger than 5368.

Cast one of the factors of the multiplication to u64 to force the
multiplication to take place on 64 bits.

Issue found by Coverity.

Fixes: 55a515b1f5a9 ("net: dsa: felix: drop oversized frames with tc-taprio instead of hanging the port")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230613170907.2413559-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet/sched: cls_api: Fix lockup on flushing explicitly created chain
Vlad Buslov [Mon, 12 Jun 2023 09:34:26 +0000 (11:34 +0200)]
net/sched: cls_api: Fix lockup on flushing explicitly created chain

Mingshuai Ren reports:

When a new chain is added by using tc, one soft lockup alarm will be
 generated after delete the prio 0 filter of the chain. To reproduce
 the problem, perform the following steps:
(1) tc qdisc add dev eth0 root handle 1: htb default 1
(2) tc chain add dev eth0
(3) tc filter del dev eth0 chain 0 parent 1: prio 0
(4) tc filter add dev eth0 chain 0 parent 1:

Fix the issue by accounting for additional reference to chains that are
explicitly created by RTM_NEWCHAIN message as opposed to implicitly by
RTM_NEWTFILTER message.

Fixes: 726d061286ce ("net: sched: prevent insertion of new classifiers during chain flush")
Reported-by: Mingshuai Ren <renmingshuai@huawei.com>
Closes: https://lore.kernel.org/lkml/87legswvi3.fsf@nvidia.com/T/
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Link: https://lore.kernel.org/r/20230612093426.2867183-1-vladbu@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>