Catherine Sullivan [Thu, 29 Jul 2021 15:52:58 +0000 (08:52 -0700)]
gve: Update MAINTAINERS list
The team maintaining the gve driver has undergone some changes,
this updates the MAINTAINERS file accordingly.
Signed-off-by: Catherine Sullivan <csully@google.com>
Signed-off-by: Jon Olson <jonolson@google.com>
Signed-off-by: David Awogbemila <awogbemila@google.com>
Signed-off-by: Jeroen de Borst <jeroendb@google.com>
Link: https://lore.kernel.org/r/20210729155258.442650-1-csully@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
David S. Miller [Wed, 28 Jul 2021 23:53:32 +0000 (00:53 +0100)]
Merge git://git./pub/scm/linux/kernel/git/bpf/bpf
Daniel Borkmann says:
====================
pull-request: bpf 2021-07-29
The following pull-request contains BPF updates for your *net* tree.
We've added 9 non-merge commits during the last 14 day(s) which contain
a total of 20 files changed, 446 insertions(+), 138 deletions(-).
The main changes are:
1) Fix UBSAN out-of-bounds splat for showing XDP link fdinfo, from Lorenz Bauer.
2) Fix insufficient Spectre v4 mitigation in BPF runtime, from Daniel Borkmann,
Piotr Krysiuk and Benedict Schlueter.
3) Batch of fixes for BPF sockmap found under stress testing, from John Fastabend.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Daniel Borkmann [Tue, 13 Jul 2021 08:18:31 +0000 (08:18 +0000)]
bpf: Fix leakage due to insufficient speculative store bypass mitigation
Spectre v4 gadgets make use of memory disambiguation, which is a set of
techniques that execute memory access instructions, that is, loads and
stores, out of program order; Intel's optimization manual, section 2.4.4.5:
A load instruction micro-op may depend on a preceding store. Many
microarchitectures block loads until all preceding store addresses are
known. The memory disambiguator predicts which loads will not depend on
any previous stores. When the disambiguator predicts that a load does
not have such a dependency, the load takes its data from the L1 data
cache. Eventually, the prediction is verified. If an actual conflict is
detected, the load and all succeeding instructions are re-executed.
af86ca4e3088 ("bpf: Prevent memory disambiguation attack") tried to mitigate
this attack by sanitizing the memory locations through preemptive "fast"
(low latency) stores of zero prior to the actual "slow" (high latency) store
of a pointer value such that upon dependency misprediction the CPU then
speculatively executes the load of the pointer value and retrieves the zero
value instead of the attacker controlled scalar value previously stored at
that location, meaning, subsequent access in the speculative domain is then
redirected to the "zero page".
The sanitized preemptive store of zero prior to the actual "slow" store is
done through a simple ST instruction based on r10 (frame pointer) with
relative offset to the stack location that the verifier has been tracking
on the original used register for STX, which does not have to be r10. Thus,
there are no memory dependencies for this store, since it's only using r10
and immediate constant of zero; hence
af86ca4e3088 /assumed/ a low latency
operation.
However, a recent attack demonstrated that this mitigation is not sufficient
since the preemptive store of zero could also be turned into a "slow" store
and is thus bypassed as well:
[...]
// r2 = oob address (e.g. scalar)
// r7 = pointer to map value
31: (7b) *(u64 *)(r10 -16) = r2
// r9 will remain "fast" register, r10 will become "slow" register below
32: (bf) r9 = r10
// JIT maps BPF reg to x86 reg:
// r9 -> r15 (callee saved)
// r10 -> rbp
// train store forward prediction to break dependency link between both r9
// and r10 by evicting them from the predictor's LRU table.
33: (61) r0 = *(u32 *)(r7 +24576)
34: (63) *(u32 *)(r7 +29696) = r0
35: (61) r0 = *(u32 *)(r7 +24580)
36: (63) *(u32 *)(r7 +29700) = r0
37: (61) r0 = *(u32 *)(r7 +24584)
38: (63) *(u32 *)(r7 +29704) = r0
39: (61) r0 = *(u32 *)(r7 +24588)
40: (63) *(u32 *)(r7 +29708) = r0
[...]
543: (61) r0 = *(u32 *)(r7 +25596)
544: (63) *(u32 *)(r7 +30716) = r0
// prepare call to bpf_ringbuf_output() helper. the latter will cause rbp
// to spill to stack memory while r13/r14/r15 (all callee saved regs) remain
// in hardware registers. rbp becomes slow due to push/pop latency. below is
// disasm of bpf_ringbuf_output() helper for better visual context:
//
//
ffffffff8117ee20: 41 54 push r12
//
ffffffff8117ee22: 55 push rbp
//
ffffffff8117ee23: 53 push rbx
//
ffffffff8117ee24: 48 f7 c1 fc ff ff ff test rcx,0xfffffffffffffffc
//
ffffffff8117ee2b: 0f 85 af 00 00 00 jne
ffffffff8117eee0 <-- jump taken
// [...]
//
ffffffff8117eee0: 49 c7 c4 ea ff ff ff mov r12,0xffffffffffffffea
//
ffffffff8117eee7: 5b pop rbx
//
ffffffff8117eee8: 5d pop rbp
//
ffffffff8117eee9: 4c 89 e0 mov rax,r12
//
ffffffff8117eeec: 41 5c pop r12
//
ffffffff8117eeee: c3 ret
545: (18) r1 = map[id:4]
547: (bf) r2 = r7
548: (b7) r3 = 0
549: (b7) r4 = 4
550: (85) call bpf_ringbuf_output#194288
// instruction 551 inserted by verifier \
551: (7a) *(u64 *)(r10 -16) = 0 | /both/ are now slow stores here
// storing map value pointer r7 at fp-16 | since value of r10 is "slow".
552: (7b) *(u64 *)(r10 -16) = r7 /
// following "fast" read to the same memory location, but due to dependency
// misprediction it will speculatively execute before insn 551/552 completes.
553: (79) r2 = *(u64 *)(r9 -16)
// in speculative domain contains attacker controlled r2. in non-speculative
// domain this contains r7, and thus accesses r7 +0 below.
554: (71) r3 = *(u8 *)(r2 +0)
// leak r3
As can be seen, the current speculative store bypass mitigation which the
verifier inserts at line 551 is insufficient since /both/, the write of
the zero sanitation as well as the map value pointer are a high latency
instruction due to prior memory access via push/pop of r10 (rbp) in contrast
to the low latency read in line 553 as r9 (r15) which stays in hardware
registers. Thus, architecturally, fp-16 is r7, however, microarchitecturally,
fp-16 can still be r2.
Initial thoughts to address this issue was to track spilled pointer loads
from stack and enforce their load via LDX through r10 as well so that /both/
the preemptive store of zero /as well as/ the load use the /same/ register
such that a dependency is created between the store and load. However, this
option is not sufficient either since it can be bypassed as well under
speculation. An updated attack with pointer spill/fills now _all_ based on
r10 would look as follows:
[...]
// r2 = oob address (e.g. scalar)
// r7 = pointer to map value
[...]
// longer store forward prediction training sequence than before.
2062: (61) r0 = *(u32 *)(r7 +25588)
2063: (63) *(u32 *)(r7 +30708) = r0
2064: (61) r0 = *(u32 *)(r7 +25592)
2065: (63) *(u32 *)(r7 +30712) = r0
2066: (61) r0 = *(u32 *)(r7 +25596)
2067: (63) *(u32 *)(r7 +30716) = r0
// store the speculative load address (scalar) this time after the store
// forward prediction training.
2068: (7b) *(u64 *)(r10 -16) = r2
// preoccupy the CPU store port by running sequence of dummy stores.
2069: (63) *(u32 *)(r7 +29696) = r0
2070: (63) *(u32 *)(r7 +29700) = r0
2071: (63) *(u32 *)(r7 +29704) = r0
2072: (63) *(u32 *)(r7 +29708) = r0
2073: (63) *(u32 *)(r7 +29712) = r0
2074: (63) *(u32 *)(r7 +29716) = r0
2075: (63) *(u32 *)(r7 +29720) = r0
2076: (63) *(u32 *)(r7 +29724) = r0
2077: (63) *(u32 *)(r7 +29728) = r0
2078: (63) *(u32 *)(r7 +29732) = r0
2079: (63) *(u32 *)(r7 +29736) = r0
2080: (63) *(u32 *)(r7 +29740) = r0
2081: (63) *(u32 *)(r7 +29744) = r0
2082: (63) *(u32 *)(r7 +29748) = r0
2083: (63) *(u32 *)(r7 +29752) = r0
2084: (63) *(u32 *)(r7 +29756) = r0
2085: (63) *(u32 *)(r7 +29760) = r0
2086: (63) *(u32 *)(r7 +29764) = r0
2087: (63) *(u32 *)(r7 +29768) = r0
2088: (63) *(u32 *)(r7 +29772) = r0
2089: (63) *(u32 *)(r7 +29776) = r0
2090: (63) *(u32 *)(r7 +29780) = r0
2091: (63) *(u32 *)(r7 +29784) = r0
2092: (63) *(u32 *)(r7 +29788) = r0
2093: (63) *(u32 *)(r7 +29792) = r0
2094: (63) *(u32 *)(r7 +29796) = r0
2095: (63) *(u32 *)(r7 +29800) = r0
2096: (63) *(u32 *)(r7 +29804) = r0
2097: (63) *(u32 *)(r7 +29808) = r0
2098: (63) *(u32 *)(r7 +29812) = r0
// overwrite scalar with dummy pointer; same as before, also including the
// sanitation store with 0 from the current mitigation by the verifier.
2099: (7a) *(u64 *)(r10 -16) = 0 | /both/ are now slow stores here
2100: (7b) *(u64 *)(r10 -16) = r7 | since store unit is still busy.
// load from stack intended to bypass stores.
2101: (79) r2 = *(u64 *)(r10 -16)
2102: (71) r3 = *(u8 *)(r2 +0)
// leak r3
[...]
Looking at the CPU microarchitecture, the scheduler might issue loads (such
as seen in line 2101) before stores (line 2099,2100) because the load execution
units become available while the store execution unit is still busy with the
sequence of dummy stores (line 2069-2098). And so the load may use the prior
stored scalar from r2 at address r10 -16 for speculation. The updated attack
may work less reliable on CPU microarchitectures where loads and stores share
execution resources.
This concludes that the sanitizing with zero stores from
af86ca4e3088 ("bpf:
Prevent memory disambiguation attack") is insufficient. Moreover, the detection
of stack reuse from
af86ca4e3088 where previously data (STACK_MISC) has been
written to a given stack slot where a pointer value is now to be stored does
not have sufficient coverage as precondition for the mitigation either; for
several reasons outlined as follows:
1) Stack content from prior program runs could still be preserved and is
therefore not "random", best example is to split a speculative store
bypass attack between tail calls, program A would prepare and store the
oob address at a given stack slot and then tail call into program B which
does the "slow" store of a pointer to the stack with subsequent "fast"
read. From program B PoV such stack slot type is STACK_INVALID, and
therefore also must be subject to mitigation.
2) The STACK_SPILL must not be coupled to register_is_const(&stack->spilled_ptr)
condition, for example, the previous content of that memory location could
also be a pointer to map or map value. Without the fix, a speculative
store bypass is not mitigated in such precondition and can then lead to
a type confusion in the speculative domain leaking kernel memory near
these pointer types.
While brainstorming on various alternative mitigation possibilities, we also
stumbled upon a retrospective from Chrome developers [0]:
[...] For variant 4, we implemented a mitigation to zero the unused memory
of the heap prior to allocation, which cost about 1% when done concurrently
and 4% for scavenging. Variant 4 defeats everything we could think of. We
explored more mitigations for variant 4 but the threat proved to be more
pervasive and dangerous than we anticipated. For example, stack slots used
by the register allocator in the optimizing compiler could be subject to
type confusion, leading to pointer crafting. Mitigating type confusion for
stack slots alone would have required a complete redesign of the backend of
the optimizing compiler, perhaps man years of work, without a guarantee of
completeness. [...]
From BPF side, the problem space is reduced, however, options are rather
limited. One idea that has been explored was to xor-obfuscate pointer spills
to the BPF stack:
[...]
// preoccupy the CPU store port by running sequence of dummy stores.
[...]
2106: (63) *(u32 *)(r7 +29796) = r0
2107: (63) *(u32 *)(r7 +29800) = r0
2108: (63) *(u32 *)(r7 +29804) = r0
2109: (63) *(u32 *)(r7 +29808) = r0
2110: (63) *(u32 *)(r7 +29812) = r0
// overwrite scalar with dummy pointer; xored with random 'secret' value
// of
943576462 before store ...
2111: (b4) w11 =
943576462
2112: (af) r11 ^= r7
2113: (7b) *(u64 *)(r10 -16) = r11
2114: (79) r11 = *(u64 *)(r10 -16)
2115: (b4) w2 =
943576462
2116: (af) r2 ^= r11
// ... and restored with the same 'secret' value with the help of AX reg.
2117: (71) r3 = *(u8 *)(r2 +0)
[...]
While the above would not prevent speculation, it would make data leakage
infeasible by directing it to random locations. In order to be effective
and prevent type confusion under speculation, such random secret would have
to be regenerated for each store. The additional complexity involved for a
tracking mechanism that prevents jumps such that restoring spilled pointers
would not get corrupted is not worth the gain for unprivileged. Hence, the
fix in here eventually opted for emitting a non-public BPF_ST | BPF_NOSPEC
instruction which the x86 JIT translates into a lfence opcode. Inserting the
latter in between the store and load instruction is one of the mitigations
options [1]. The x86 instruction manual notes:
[...] An LFENCE that follows an instruction that stores to memory might
complete before the data being stored have become globally visible. [...]
The latter meaning that the preceding store instruction finished execution
and the store is at minimum guaranteed to be in the CPU's store queue, but
it's not guaranteed to be in that CPU's L1 cache at that point (globally
visible). The latter would only be guaranteed via sfence. So the load which
is guaranteed to execute after the lfence for that local CPU would have to
rely on store-to-load forwarding. [2], in section 2.3 on store buffers says:
[...] For every store operation that is added to the ROB, an entry is
allocated in the store buffer. This entry requires both the virtual and
physical address of the target. Only if there is no free entry in the store
buffer, the frontend stalls until there is an empty slot available in the
store buffer again. Otherwise, the CPU can immediately continue adding
subsequent instructions to the ROB and execute them out of order. On Intel
CPUs, the store buffer has up to 56 entries. [...]
One small upside on the fix is that it lifts constraints from
af86ca4e3088
where the sanitize_stack_off relative to r10 must be the same when coming
from different paths. The BPF_ST | BPF_NOSPEC gets emitted after a BPF_STX
or BPF_ST instruction. This happens either when we store a pointer or data
value to the BPF stack for the first time, or upon later pointer spills.
The former needs to be enforced since otherwise stale stack data could be
leaked under speculation as outlined earlier. For non-x86 JITs the BPF_ST |
BPF_NOSPEC mapping is currently optimized away, but others could emit a
speculation barrier as well if necessary. For real-world unprivileged
programs e.g. generated by LLVM, pointer spill/fill is only generated upon
register pressure and LLVM only tries to do that for pointers which are not
used often. The program main impact will be the initial BPF_ST | BPF_NOSPEC
sanitation for the STACK_INVALID case when the first write to a stack slot
occurs e.g. upon map lookup. In future we might refine ways to mitigate
the latter cost.
[0] https://arxiv.org/pdf/1902.05178.pdf
[1] https://msrc-blog.microsoft.com/2018/05/21/analysis-and-mitigation-of-speculative-store-bypass-cve-2018-3639/
[2] https://arxiv.org/pdf/1905.05725.pdf
Fixes:
af86ca4e3088 ("bpf: Prevent memory disambiguation attack")
Fixes:
f7cf25b2026d ("bpf: track spill/fill of constants")
Co-developed-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Benedict Schlueter <benedict.schlueter@rub.de>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Signed-off-by: Benedict Schlueter <benedict.schlueter@rub.de>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Daniel Borkmann [Tue, 13 Jul 2021 08:18:31 +0000 (08:18 +0000)]
bpf: Introduce BPF nospec instruction for mitigating Spectre v4
In case of JITs, each of the JIT backends compiles the BPF nospec instruction
/either/ to a machine instruction which emits a speculation barrier /or/ to
/no/ machine instruction in case the underlying architecture is not affected
by Speculative Store Bypass or has different mitigations in place already.
This covers both x86 and (implicitly) arm64: In case of x86, we use 'lfence'
instruction for mitigation. In case of arm64, we rely on the firmware mitigation
as controlled via the ssbd kernel parameter. Whenever the mitigation is enabled,
it works for all of the kernel code with no need to provide any additional
instructions here (hence only comment in arm64 JIT). Other archs can follow
as needed. The BPF nospec instruction is specifically targeting Spectre v4
since i) we don't use a serialization barrier for the Spectre v1 case, and
ii) mitigation instructions for v1 and v4 might be different on some archs.
The BPF nospec is required for a future commit, where the BPF verifier does
annotate intermediate BPF programs with speculation barriers.
Co-developed-by: Piotr Krysiuk <piotras@gmail.com>
Co-developed-by: Benedict Schlueter <benedict.schlueter@rub.de>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Piotr Krysiuk <piotras@gmail.com>
Signed-off-by: Benedict Schlueter <benedict.schlueter@rub.de>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Wang Hai [Wed, 28 Jul 2021 12:11:07 +0000 (20:11 +0800)]
sis900: Fix missing pci_disable_device() in probe and remove
Replace pci_enable_device() with pcim_enable_device(),
pci_disable_device() and pci_release_regions() will be
called in release automatically.
Fixes:
1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
zhang kai [Wed, 28 Jul 2021 10:54:18 +0000 (18:54 +0800)]
net: let flow have same hash in two directions
using same source and destination ip/port for flow hash calculation
within the two directions.
Signed-off-by: zhang kai <zhangkaiheb@126.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Krzysztof Kozlowski [Wed, 28 Jul 2021 06:49:09 +0000 (08:49 +0200)]
nfc: nfcsim: fix use after free during module unload
There is a use after free memory corruption during module exit:
- nfcsim_exit()
- nfcsim_device_free(dev0)
- nfc_digital_unregister_device()
This iterates over command queue and frees all commands,
- dev->up = false
- nfcsim_link_shutdown()
- nfcsim_link_recv_wake()
This wakes the sleeping thread nfcsim_link_recv_skb().
- nfcsim_link_recv_skb()
Wake from wait_event_interruptible_timeout(),
call directly the deb->cb callback even though (dev->up == false),
- digital_send_cmd_complete()
Dereference of "struct digital_cmd" cmd which was freed earlier by
nfc_digital_unregister_device().
This causes memory corruption shortly after (with unrelated stack
trace):
nfc nfc0: NFC: nfcsim_recv_wq: Device is down
llcp: nfc_llcp_recv: err -19
nfc nfc1: NFC: nfcsim_recv_wq: Device is down
BUG: unable to handle page fault for address:
ffffffffffffffed
Call Trace:
fsnotify+0x54b/0x5c0
__fsnotify_parent+0x1fe/0x300
? vfs_write+0x27c/0x390
vfs_write+0x27c/0x390
ksys_write+0x63/0xe0
do_syscall_64+0x3b/0x90
entry_SYSCALL_64_after_hwframe+0x44/0xae
KASAN report:
BUG: KASAN: use-after-free in digital_send_cmd_complete+0x16/0x50
Write of size 8 at addr
ffff88800a05f720 by task kworker/0:2/71
Workqueue: events nfcsim_recv_wq [nfcsim]
Call Trace:
dump_stack_lvl+0x45/0x59
print_address_description.constprop.0+0x21/0x140
? digital_send_cmd_complete+0x16/0x50
? digital_send_cmd_complete+0x16/0x50
kasan_report.cold+0x7f/0x11b
? digital_send_cmd_complete+0x16/0x50
? digital_dep_link_down+0x60/0x60
digital_send_cmd_complete+0x16/0x50
nfcsim_recv_wq+0x38f/0x3d5 [nfcsim]
? nfcsim_in_send_cmd+0x4a/0x4a [nfcsim]
? lock_is_held_type+0x98/0x110
? finish_wait+0x110/0x110
? rcu_read_lock_sched_held+0x9c/0xd0
? rcu_read_lock_bh_held+0xb0/0xb0
? lockdep_hardirqs_on_prepare+0x12e/0x1f0
This flow of calling digital_send_cmd_complete() callback on driver exit
is specific to nfcsim which implements reading and sending work queues.
Since the NFC digital device was unregistered, the callback should not
be called.
Fixes:
204bddcb508f ("NFC: nfcsim: Make use of the Digital layer")
Cc: <stable@vger.kernel.org>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Wang Hai [Wed, 28 Jul 2021 07:43:13 +0000 (15:43 +0800)]
tulip: windbond-840: Fix missing pci_disable_device() in probe and remove
Replace pci_enable_device() with pcim_enable_device(),
pci_disable_device() and pci_release_regions() will be
called in release automatically.
Fixes:
1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Marcelo Ricardo Leitner [Wed, 28 Jul 2021 02:40:54 +0000 (23:40 -0300)]
sctp: fix return value check in __sctp_rcv_asconf_lookup
As Ben Hutchings noticed, this check should have been inverted: the call
returns true in case of success.
Reported-by: Ben Hutchings <ben@decadent.org.uk>
Fixes:
0c5dc070ff3d ("sctp: validate from_addr_param return")
Signed-off-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Tang Bin [Wed, 28 Jul 2021 01:49:25 +0000 (09:49 +0800)]
nfc: s3fwrn5: fix undefined parameter values in dev_err()
In the function s3fwrn5_fw_download(), the 'ret' is not assigned,
so the correct value should be given in dev_err function.
Fixes:
a0302ff5906a ("nfc: s3fwrn5: remove unnecessary label")
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 28 Jul 2021 08:16:37 +0000 (09:16 +0100)]
Merge tag 'mlx5-fixes-2021-07-27' of git://git./linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5 fixes 2021-07-27
This series introduces some fixes to mlx5 driver.
Please pull and let me know if there is any problem.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Chris Mi [Mon, 26 Apr 2021 03:06:37 +0000 (11:06 +0800)]
net/mlx5: Fix mlx5_vport_tbl_attr chain from u16 to u32
The offending refactor commit uses u16 chain wrongly. Actually, it
should be u32.
Fixes:
c620b772152b ("net/mlx5: Refactor tc flow attributes structure")
CC: Ariel Levkovich <lariel@nvidia.com>
Signed-off-by: Chris Mi <cmi@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Dima Chumak [Mon, 26 Apr 2021 12:16:26 +0000 (15:16 +0300)]
net/mlx5e: Fix nullptr in mlx5e_hairpin_get_mdev()
The result of __dev_get_by_index() is not checked for NULL and then gets
dereferenced immediately.
Also, __dev_get_by_index() must be called while holding either RTNL lock
or @dev_base_lock, which isn't satisfied by mlx5e_hairpin_get_mdev() or
its callers. This makes the underlying hlist_for_each_entry() loop not
safe, and can have adverse effects in itself.
Fix by using dev_get_by_index() and handling nullptr return value when
ifindex device is not found. Update mlx5e_hairpin_get_mdev() callers to
check for possible PTR_ERR() result.
Fixes:
77ab67b7f0f9 ("net/mlx5e: Basic setup of hairpin object")
Addresses-Coverity: ("Dereference null return value")
Signed-off-by: Dima Chumak <dchumak@nvidia.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Aya Levin [Wed, 16 Jun 2021 16:11:03 +0000 (19:11 +0300)]
net/mlx5: Unload device upon firmware fatal error
When fw_fatal reporter reports an error, the firmware in not responding.
Unload the device to ensure that the driver closes all its resources,
even if recovery is not due (user disabled auto-recovery or reporter is
in grace period). On successful recovery the device is loaded back up.
Fixes:
b3bd076f7501 ("net/mlx5: Report devlink health on FW fatal issues")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Aya Levin [Tue, 22 Jun 2021 07:24:17 +0000 (10:24 +0300)]
net/mlx5e: Fix page allocation failure for ptp-RQ over SF
Set the correct pci-device pointer to the ptp-RQ. This allows access to
dma_mask and avoids allocation request with wrong pci-device.
Fixes:
a099da8ffcf6 ("net/mlx5e: Add RQ to PTP channel")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Aya Levin [Mon, 21 Jun 2021 15:04:07 +0000 (18:04 +0300)]
net/mlx5e: Fix page allocation failure for trap-RQ over SF
Set the correct device pointer to the trap-RQ, to allow access to
dma_mask and avoid allocation request with the wrong pci-dev.
WARNING: CPU: 1 PID: 12005 at kernel/dma/mapping.c:151 dma_map_page_attrs+0x139/0x1c0
...
all Trace:
<IRQ>
? __page_pool_alloc_pages_slow+0x5a/0x210
mlx5e_post_rx_wqes+0x258/0x400 [mlx5_core]
mlx5e_trap_napi_poll+0x44/0xc0 [mlx5_core]
__napi_poll+0x24/0x150
net_rx_action+0x22b/0x280
__do_softirq+0xc7/0x27e
do_softirq+0x61/0x80
</IRQ>
__local_bh_enable_ip+0x4b/0x50
mlx5e_handle_action_trap+0x2dd/0x4d0 [mlx5_core]
blocking_notifier_call_chain+0x5a/0x80
mlx5_devlink_trap_action_set+0x8b/0x100 [mlx5_core]
Fixes:
5543e989fe5e ("net/mlx5e: Add trap entity to ETH driver")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Aya Levin [Wed, 30 Jun 2021 08:17:12 +0000 (11:17 +0300)]
net/mlx5e: Consider PTP-RQ when setting RX VLAN stripping
Add PTP-RQ to the loop when setting rx-vlan-offload feature via ethtool.
On PTP-RQ's creation, set rx-vlan-offload into its parameters.
Fixes:
a099da8ffcf6 ("net/mlx5e: Add RQ to PTP channel")
Signed-off-by: Aya Levin <ayal@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Maxim Mikityanskiy [Wed, 30 Jun 2021 10:33:31 +0000 (13:33 +0300)]
net/mlx5e: Add NETIF_F_HW_TC to hw_features when HTB offload is available
If a feature flag is only present in features, but not in hw_features,
the user can't reset it. Although hw_features may contain NETIF_F_HW_TC
by the point where the driver checks whether HTB offload is supported,
this flag is controlled by another condition that may not hold. Set it
explicitly to make sure the user can disable it.
Fixes:
214baf22870c ("net/mlx5e: Support HTB offload")
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Tariq Toukan [Wed, 30 Jun 2021 10:45:05 +0000 (13:45 +0300)]
net/mlx5e: RX, Avoid possible data corruption when relaxed ordering and LRO combined
When HW aggregates packets for an LRO session, it writes the payload
of two consecutive packets of a flow contiguously, so that they usually
share a cacheline.
The first byte of a packet's payload is written immediately after
the last byte of the preceding packet.
In this flow, there are two consecutive write requests to the shared
cacheline:
1. Regular write for the earlier packet.
2. Read-modify-write for the following packet.
In case of relaxed-ordering on, these two writes might be re-ordered.
Using the end padding optimization (to avoid partial write for the last
cacheline of a packet) becomes problematic if the two writes occur
out-of-order, as the padding would overwrite payload that belongs to
the following packet, causing data corruption.
Avoid this by disabling the end padding optimization when both
LRO and relaxed-ordering are enabled.
Fixes:
17347d5430c4 ("net/mlx5e: Add support for PCI relaxed ordering")
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Roi Dayan [Wed, 2 Jun 2021 11:17:07 +0000 (14:17 +0300)]
net/mlx5: E-Switch, handle devcom events only for ports on the same device
This is the same check as LAG mode checks if to enable lag.
This will fix adding peer miss rules if lag is not supported
and even an incorrect rules in socket direct mode.
Also fix the incorrect comment on mlx5_get_next_phys_dev() as flow #1
doesn't exists.
Fixes:
ac004b832128 ("net/mlx5e: E-Switch, Add peer miss rules")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Maor Dickman [Tue, 22 Jun 2021 14:07:02 +0000 (17:07 +0300)]
net/mlx5: E-Switch, Set destination vport vhca id only when merged eswitch is supported
Destination vport vhca id is valid flag is set only merged eswitch isn't supported.
Change destination vport vhca id value to be set also only when merged eswitch
is supported.
Fixes:
e4ad91f23f10 ("net/mlx5e: Split offloaded eswitch TC rules for port mirroring")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Maor Dickman [Thu, 8 Jul 2021 12:24:58 +0000 (15:24 +0300)]
net/mlx5e: Disable Rx ntuple offload for uplink representor
Rx ntuple offload is not supported in switchdev mode.
Tryng to enable it cause kernel panic.
BUG: kernel NULL pointer dereference, address:
0000000000000008
#PF: supervisor read access in kernel mode
#PF: error_code(0x0000) - not-present page
PGD
80000001065a5067 P4D
80000001065a5067 PUD
106594067 PMD 0
Oops: 0000 [#1] SMP PTI
CPU: 7 PID: 1089 Comm: ethtool Not tainted 5.13.0-rc7_for_upstream_min_debug_2021_06_23_16_44 #1
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:mlx5e_arfs_enable+0x70/0xd0 [mlx5_core]
Code: 44 24 10 00 00 00 00 48 c7 44 24 18 00 00 00 00 49 63 c4 48 89 e2 44 89 e6 48 69 c0 20 08 00 00 48 89 ef 48 03 85 68 ac 00 00 <48> 8b 40 08 48 89 44 24 08 e8 d2 aa fd ff 48 83 05 82 96 18 00 01
RSP: 0018:
ffff8881047679e0 EFLAGS:
00010246
RAX:
0000000000000000 RBX:
0000004000000000 RCX:
0000004000000000
RDX:
ffff8881047679e0 RSI:
0000000000000000 RDI:
ffff888115100880
RBP:
ffff888115100880 R08:
ffffffffa00f6cb0 R09:
ffff888104767a18
R10:
ffff8881151000a0 R11:
ffff888109479540 R12:
0000000000000000
R13:
ffff888104767bb8 R14:
ffff888115100000 R15:
ffff8881151000a0
FS:
00007f41a64ab740(0000) GS:
ffff8882f5dc0000(0000) knlGS:
0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0:
0000000080050033
CR2:
0000000000000008 CR3:
0000000104cbc005 CR4:
0000000000370ea0
DR0:
0000000000000000 DR1:
0000000000000000 DR2:
0000000000000000
DR3:
0000000000000000 DR6:
00000000fffe0ff0 DR7:
0000000000000400
Call Trace:
set_feature_arfs+0x1e/0x40 [mlx5_core]
mlx5e_handle_feature+0x43/0xa0 [mlx5_core]
mlx5e_set_features+0x139/0x1b0 [mlx5_core]
__netdev_update_features+0x2b3/0xaf0
ethnl_set_features+0x176/0x3a0
? __nla_parse+0x22/0x30
genl_family_rcv_msg_doit+0xe2/0x140
genl_rcv_msg+0xde/0x1d0
? features_reply_size+0xe0/0xe0
? genl_get_cmd+0xd0/0xd0
netlink_rcv_skb+0x4e/0xf0
genl_rcv+0x24/0x40
netlink_unicast+0x1f6/0x2b0
netlink_sendmsg+0x225/0x450
sock_sendmsg+0x33/0x40
__sys_sendto+0xd4/0x120
? __sys_recvmsg+0x4e/0x90
? exc_page_fault+0x219/0x740
__x64_sys_sendto+0x25/0x30
do_syscall_64+0x3f/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f41a65b0cba
Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb b8 0f 1f 00 f3 0f 1e fa 41 89 ca 64 8b 04 25 18 00 00 00 85 c0 75 15 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 76 c3 0f 1f 44 00 00 55 48 83 ec 30 44 89 4c
RSP: 002b:
00007ffd8d688358 EFLAGS:
00000246 ORIG_RAX:
000000000000002c
RAX:
ffffffffffffffda RBX:
00000000010f42a0 RCX:
00007f41a65b0cba
RDX:
0000000000000058 RSI:
00000000010f43b0 RDI:
0000000000000003
RBP:
000000000047ae60 R08:
00007f41a667c000 R09:
000000000000000c
R10:
0000000000000000 R11:
0000000000000246 R12:
00000000010f4340
R13:
00000000010f4350 R14:
00007ffd8d688400 R15:
00000000010f42a0
Modules linked in: mlx5_vdpa vhost_iotlb vdpa xt_conntrack xt_MASQUERADE nf_conntrack_netlink nfnetlink xt_addrtype iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 br_netfilter rpcrdma rdma_ucm ib_iser libiscsi scsi_transport_iscsi ib_umad ib_ipoib rdma_cm iw_cm ib_cm mlx5_ib ib_uverbs ib_core overlay mlx5_core ptp pps_core fuse
CR2:
0000000000000008
---[ end trace
c66523f2aba94b43 ]---
Fixes:
7a9fb35e8c3a ("net/mlx5e: Do not reload ethernet ports when changing eswitch mode")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Maor Gottlieb [Mon, 26 Jul 2021 06:20:14 +0000 (09:20 +0300)]
net/mlx5: Fix flow table chaining
Fix a bug when flow table is created in priority that already
has other flow tables as shown in the below diagram.
If the new flow table (FT-B) has the lowest level in the priority,
we need to connect the flow tables from the previous priority (p0)
to this new table. In addition when this flow table is destroyed
(FT-B), we need to connect the flow tables from the previous
priority (p0) to the next level flow table (FT-C) in the same
priority of the destroyed table (if exists).
---------
|root_ns|
---------
|
--------------------------------
| | |
---------- ---------- ---------
|p(prio)-x| | p-y | | p-n |
---------- ---------- ---------
| |
---------------- ------------------
|ns(e.g bypass)| |ns(e.g. kernel) |
---------------- ------------------
| | |
------- ------ ----
| p0 | | p1 | |p2|
------- ------ ----
| | \
-------- ------- ------
| FT-A | |FT-B | |FT-C|
-------- ------- ------
Fixes:
f90edfd279f3 ("net/mlx5_core: Connect flow tables")
Signed-off-by: Maor Gottlieb <maorg@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Andrii Nakryiko [Tue, 27 Jul 2021 21:54:45 +0000 (14:54 -0700)]
Merge branch 'sockmap fixes picked up by stress tests'
John Fastabend says:
====================
Running stress tests with recent patch to remove an extra lock in sockmap
resulted in a couple new issues popping up. It seems only one of them
is actually related to the patch:
799aa7f98d53 ("skmsg: Avoid lock_sock() in sk_psock_backlog()")
The other two issues had existed long before, but I guess the timing
with the serialization we had before was too tight to get any of
our tests or deployments to hit it.
With attached series stress testing sockmap+TCP with workloads that
create lots of short-lived connections no more splats like below were
seen on upstream bpf branch.
[224913.935822] WARNING: CPU: 3 PID: 32100 at net/core/stream.c:208 sk_stream_kill_queues+0x212/0x220
[224913.935841] Modules linked in: fuse overlay bpf_preload x86_pkg_temp_thermal intel_uncore wmi_bmof squashfs sch_fq_codel efivarfs ip_tables x_tables uas xhci_pci ixgbe mdio xfrm_algo xhci_hcd wmi
[224913.935897] CPU: 3 PID: 32100 Comm: fgs-bench Tainted: G I 5.14.0-rc1alu+ #181
[224913.935908] Hardware name: Dell Inc. Precision 5820 Tower/002KVM, BIOS 1.9.2 01/24/2019
[224913.935914] RIP: 0010:sk_stream_kill_queues+0x212/0x220
[224913.935923] Code: 8b 83 20 02 00 00 85 c0 75 20 5b 5d 41 5c 41 5d 41 5e 41 5f c3 48 89 df e8 2b 11 fe ff eb c3 0f 0b e9 7c ff ff ff 0f 0b eb ce <0f> 0b 5b 5d 41 5c 41 5d 41 5e 41 5f c3 90 0f 1f 44 00 00 41 57 41
[224913.935932] RSP: 0018:
ffff88816271fd38 EFLAGS:
00010206
[224913.935941] RAX:
0000000000000ae8 RBX:
ffff88815acd5240 RCX:
dffffc0000000000
[224913.935948] RDX:
0000000000000003 RSI:
0000000000000ae8 RDI:
ffff88815acd5460
[224913.935954] RBP:
ffff88815acd5460 R08:
ffffffff955c0ae8 R09:
fffffbfff2e6f543
[224913.935961] R10:
ffffffff9737aa17 R11:
fffffbfff2e6f542 R12:
ffff88815acd5390
[224913.935967] R13:
ffff88815acd5480 R14:
ffffffff98d0c080 R15:
ffffffff96267500
[224913.935974] FS:
00007f86e6bd1700(0000) GS:
ffff888451cc0000(0000) knlGS:
0000000000000000
[224913.935981] CS: 0010 DS: 0000 ES: 0000 CR0:
0000000080050033
[224913.935988] CR2:
000000c0008eb000 CR3:
00000001020e0005 CR4:
00000000003706e0
[224913.935994] DR0:
0000000000000000 DR1:
0000000000000000 DR2:
0000000000000000
[224913.936000] DR3:
0000000000000000 DR6:
00000000fffe0ff0 DR7:
0000000000000400
[224913.936007] Call Trace:
[224913.936016] inet_csk_destroy_sock+0xba/0x1f0
[224913.936033] __tcp_close+0x620/0x790
[224913.936047] tcp_close+0x20/0x80
[224913.936056] inet_release+0x8f/0xf0
[224913.936070] __sock_release+0x72/0x120
v3: make sock_drop inline in skmsg.h
v2: init skb to null and fix a space/tab issue. Added Jakub's acks.
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
John Fastabend [Tue, 27 Jul 2021 16:05:00 +0000 (09:05 -0700)]
bpf, sockmap: Fix memleak on ingress msg enqueue
If backlog handler is running during a tear down operation we may enqueue
data on the ingress msg queue while tear down is trying to free it.
sk_psock_backlog()
sk_psock_handle_skb()
skb_psock_skb_ingress()
sk_psock_skb_ingress_enqueue()
sk_psock_queue_msg(psock,msg)
spin_lock(ingress_lock)
sk_psock_zap_ingress()
_sk_psock_purge_ingerss_msg()
_sk_psock_purge_ingress_msg()
-- free ingress_msg list --
spin_unlock(ingress_lock)
spin_lock(ingress_lock)
list_add_tail(msg,ingress_msg) <- entry on list with no one
left to free it.
spin_unlock(ingress_lock)
To fix we only enqueue from backlog if the ENABLED bit is set. The tear
down logic clears the bit with ingress_lock set so we wont enqueue the
msg in the last step.
Fixes:
799aa7f98d53 ("skmsg: Avoid lock_sock() in sk_psock_backlog()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210727160500.1713554-4-john.fastabend@gmail.com
John Fastabend [Tue, 27 Jul 2021 16:04:59 +0000 (09:04 -0700)]
bpf, sockmap: On cleanup we additionally need to remove cached skb
Its possible if a socket is closed and the receive thread is under memory
pressure it may have cached a skb. We need to ensure these skbs are
free'd along with the normal ingress_skb queue.
Before
799aa7f98d53 ("skmsg: Avoid lock_sock() in sk_psock_backlog()") tear
down and backlog processing both had sock_lock for the common case of
socket close or unhash. So it was not possible to have both running in
parrallel so all we would need is the kfree in those kernels.
But, latest kernels include the commit
799aa7f98d5e and this requires a
bit more work. Without the ingress_lock guarding reading/writing the
state->skb case its possible the tear down could run before the state
update causing it to leak memory or worse when the backlog reads the state
it could potentially run interleaved with the tear down and we might end up
free'ing the state->skb from tear down side but already have the reference
from backlog side. To resolve such races we wrap accesses in ingress_lock
on both sides serializing tear down and backlog case. In both cases this
only happens after an EAGAIN error case so having an extra lock in place
is likely fine. The normal path will skip the locks.
Note, we check state->skb before grabbing lock. This works because
we can only enqueue with the mutex we hold already. Avoiding a race
on adding state->skb after the check. And if tear down path is running
that is also fine if the tear down path then removes state->skb we
will simply set skb=NULL and the subsequent goto is skipped. This
slight complication avoids locking in normal case.
With this fix we no longer see this warning splat from tcp side on
socket close when we hit the above case with redirect to ingress self.
[224913.935822] WARNING: CPU: 3 PID: 32100 at net/core/stream.c:208 sk_stream_kill_queues+0x212/0x220
[224913.935841] Modules linked in: fuse overlay bpf_preload x86_pkg_temp_thermal intel_uncore wmi_bmof squashfs sch_fq_codel efivarfs ip_tables x_tables uas xhci_pci ixgbe mdio xfrm_algo xhci_hcd wmi
[224913.935897] CPU: 3 PID: 32100 Comm: fgs-bench Tainted: G I 5.14.0-rc1alu+ #181
[224913.935908] Hardware name: Dell Inc. Precision 5820 Tower/002KVM, BIOS 1.9.2 01/24/2019
[224913.935914] RIP: 0010:sk_stream_kill_queues+0x212/0x220
[224913.935923] Code: 8b 83 20 02 00 00 85 c0 75 20 5b 5d 41 5c 41 5d 41 5e 41 5f c3 48 89 df e8 2b 11 fe ff eb c3 0f 0b e9 7c ff ff ff 0f 0b eb ce <0f> 0b 5b 5d 41 5c 41 5d 41 5e 41 5f c3 90 0f 1f 44 00 00 41 57 41
[224913.935932] RSP: 0018:
ffff88816271fd38 EFLAGS:
00010206
[224913.935941] RAX:
0000000000000ae8 RBX:
ffff88815acd5240 RCX:
dffffc0000000000
[224913.935948] RDX:
0000000000000003 RSI:
0000000000000ae8 RDI:
ffff88815acd5460
[224913.935954] RBP:
ffff88815acd5460 R08:
ffffffff955c0ae8 R09:
fffffbfff2e6f543
[224913.935961] R10:
ffffffff9737aa17 R11:
fffffbfff2e6f542 R12:
ffff88815acd5390
[224913.935967] R13:
ffff88815acd5480 R14:
ffffffff98d0c080 R15:
ffffffff96267500
[224913.935974] FS:
00007f86e6bd1700(0000) GS:
ffff888451cc0000(0000) knlGS:
0000000000000000
[224913.935981] CS: 0010 DS: 0000 ES: 0000 CR0:
0000000080050033
[224913.935988] CR2:
000000c0008eb000 CR3:
00000001020e0005 CR4:
00000000003706e0
[224913.935994] DR0:
0000000000000000 DR1:
0000000000000000 DR2:
0000000000000000
[224913.936000] DR3:
0000000000000000 DR6:
00000000fffe0ff0 DR7:
0000000000000400
[224913.936007] Call Trace:
[224913.936016] inet_csk_destroy_sock+0xba/0x1f0
[224913.936033] __tcp_close+0x620/0x790
[224913.936047] tcp_close+0x20/0x80
[224913.936056] inet_release+0x8f/0xf0
[224913.936070] __sock_release+0x72/0x120
[224913.936083] sock_close+0x14/0x20
Fixes:
a136678c0bdbb ("bpf: sk_msg, zap ingress queue on psock down")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210727160500.1713554-3-john.fastabend@gmail.com
John Fastabend [Tue, 27 Jul 2021 16:04:58 +0000 (09:04 -0700)]
bpf, sockmap: Zap ingress queues after stopping strparser
We don't want strparser to run and pass skbs into skmsg handlers when
the psock is null. We just sk_drop them in this case. When removing
a live socket from map it means extra drops that we do not need to
incur. Move the zap below strparser close to avoid this condition.
This way we stop the stream parser first stopping it from processing
packets and then delete the psock.
Fixes:
a136678c0bdbb ("bpf: sk_msg, zap ingress queue on psock down")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210727160500.1713554-2-john.fastabend@gmail.com
Yufeng Mo [Tue, 27 Jul 2021 14:03:50 +0000 (22:03 +0800)]
net: hns3: change the method of obtaining default ptp cycle
The ptp cycle is related to the hardware, so it may cause compatibility
issues if a fixed value is used in driver. Therefore, the method of
obtaining this value is changed to read from the register rather than
use a fixed value in driver.
Fixes:
0bf5eb788512 ("net: hns3: add support for PTP")
Signed-off-by: Yufeng Mo <moyufeng@huawei.com>
Signed-off-by: Guangbin Huang <huangguangbin2@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Tang Bin [Tue, 27 Jul 2021 12:25:06 +0000 (20:25 +0800)]
nfc: s3fwrn5: fix undefined parameter values in dev_err()
In the function s3fwrn5_fw_download(), the 'ret' is not assigned,
so the correct value should be given in dev_err function.
Fixes:
a0302ff5906a ("nfc: s3fwrn5: remove unnecessary label")
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Signed-off-by: Tang Bin <tangbin@cmss.chinamobile.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pavel Skripkin [Sat, 24 Jul 2021 21:11:59 +0000 (00:11 +0300)]
net: llc: fix skb_over_panic
Syzbot reported skb_over_panic() in llc_pdu_init_as_xid_cmd(). The
problem was in wrong LCC header manipulations.
Syzbot's reproducer tries to send XID packet. llc_ui_sendmsg() is
doing following steps:
1. skb allocation with size = len + header size
len is passed from userpace and header size
is 3 since addr->sllc_xid is set.
2. skb_reserve() for header_len = 3
3. filling all other space with memcpy_from_msg()
Ok, at this moment we have fully loaded skb, only headers needs to be
filled.
Then code comes to llc_sap_action_send_xid_c(). This function pushes 3
bytes for LLC PDU header and initializes it. Then comes
llc_pdu_init_as_xid_cmd(). It initalizes next 3 bytes *AFTER* LLC PDU
header and call skb_push(skb, 3). This looks wrong for 2 reasons:
1. Bytes rigth after LLC header are user data, so this function
was overwriting payload.
2. skb_push(skb, 3) call can cause skb_over_panic() since
all free space was filled in llc_ui_sendmsg(). (This can
happen is user passed 686 len: 686 + 14 (eth header) + 3 (LLC
header) = 703. SKB_DATA_ALIGN(703) = 704)
So, in this patch I added 2 new private constansts: LLC_PDU_TYPE_U_XID
and LLC_PDU_LEN_U_XID. LLC_PDU_LEN_U_XID is used to correctly reserve
header size to handle LLC + XID case. LLC_PDU_TYPE_U_XID is used by
llc_pdu_header_init() function to push 6 bytes instead of 3. And finally
I removed skb_push() call from llc_pdu_init_as_xid_cmd().
This changes should not affect other parts of LLC, since after
all steps we just transmit buffer.
Fixes:
1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-and-tested-by: syzbot+5e5a981ad7cc54c4b2b4@syzkaller.appspotmail.com
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Sun, 25 Jul 2021 13:24:52 +0000 (18:54 +0530)]
octeontx2-af: Do NIX_RX_SW_SYNC twice
NIX_RX_SW_SYNC ensures all existing transactions are finished and
pkts are written to LLC/DRAM, queues should be teared down after
successful SW_SYNC. Due to a HW errata, in some rare scenarios
an existing transaction might end after SW_SYNC operation. To
ensure operation is fully done, do the SW_SYNC twice.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Somnath Kotur [Mon, 26 Jul 2021 18:52:48 +0000 (14:52 -0400)]
bnxt_en: Fix static checker warning in bnxt_fw_reset_task()
Now that we return when bnxt_open() fails in bnxt_fw_reset_task(),
there is no need to check for 'rc' value again before invoking
bnxt_reenable_sriov().
Fixes:
3958b1da725a ("bnxt_en: fix error path of FW reset")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Letu Ren [Sun, 25 Jul 2021 13:45:12 +0000 (21:45 +0800)]
net/qla3xxx: fix schedule while atomic in ql_wait_for_drvr_lock and ql_adapter_reset
When calling the 'ql_wait_for_drvr_lock' and 'ql_adapter_reset', the driver
has already acquired the spin lock, so the driver should not call 'ssleep'
in atomic context.
This bug can be fixed by using 'mdelay' instead of 'ssleep'.
Reported-by: Letu Ren <fantasquex@gmail.com>
Signed-off-by: Letu Ren <fantasquex@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Chen Shen [Mon, 26 Jul 2021 05:47:34 +0000 (13:47 +0800)]
sctp: delete addr based on sin6_scope_id
sctp_inet6addr_event deletes 'addr' from 'local_addr_list' when setting
netdev down, but it is possible to delete the incorrect entry (match
the first one with the same ipaddr, but the different 'ifindex'), if
there are some netdevs with the same 'local-link' ipaddr added already.
It should delete the entry depending on 'sin6_addr' and 'sin6_scope_id'
both. otherwise, the endpoint will call 'sctp_sf_ootb' if it can't find
the according association when receives 'heartbeat', and finally will
reply 'abort'.
For example:
1.when linux startup
the entries in local_addr_list:
ifindex:35 addr:fe80::40:43ff:fe80:0 (eths0.201)
ifindex:36 addr:fe80::40:43ff:fe80:0 (eths0.209)
ifindex:37 addr:fe80::40:43ff:fe80:0 (eths0.210)
the route table:
local fe80::40:43ff:fe80:0 dev eths0.201
local fe80::40:43ff:fe80:0 dev eths0.209
local fe80::40:43ff:fe80:0 dev eths0.210
2.after 'ifconfig eths0.209 down'
the entries in local_addr_list:
ifindex:36 addr:fe80::40:43ff:fe80:0 (eths0.209)
ifindex:37 addr:fe80::40:43ff:fe80:0 (eths0.210)
the route table:
local fe80::40:43ff:fe80:0 dev eths0.201
local fe80::40:43ff:fe80:0 dev eths0.210
3.asoc not found for src:[fe80::40:43ff:fe80:0]:37381 dst:[:1]:53335
::1->fe80::40:43ff:fe80:0 HEARTBEAT
fe80::40:43ff:fe80:0->::1 ABORT
Signed-off-by: Chen Shen <peterchenshen@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Mohammad Athari Bin Ismail [Mon, 26 Jul 2021 02:20:20 +0000 (10:20 +0800)]
net: stmmac: add est_irq_status callback function for GMAC 4.10 and 5.10
Assign dwmac5_est_irq_status to est_irq_status callback function for
GMAC 4.10 and 5.10. With this, EST related interrupts could be handled
properly.
Fixes:
e49aa315cb01 ("net: stmmac: EST interrupts handling and error reporting")
Cc: <stable@vger.kernel.org> # 5.13.x
Signed-off-by: Mohammad Athari Bin Ismail <mohammad.athari.ismail@intel.com>
Acked-by: Wong Vee Khee <vee.khee.wong@linux.intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 25 Jul 2021 22:06:21 +0000 (23:06 +0100)]
Merge branch 'sctp-pmtu-probe'
Xin Long says:
====================
sctp: improve the pmtu probe in Search Complete state
Timo recently suggested to use the loss of (data) packets as
indication to send pmtu probe for Search Complete state, which
should also be implied by RFC8899. This patchset is to change
the current one that is doing probe with current pmtu all the
time.
v1->v2:
- see Patch 2/2.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Sun, 25 Jul 2021 17:42:51 +0000 (13:42 -0400)]
sctp: send pmtu probe only if packet loss in Search Complete state
This patch is to introduce last_rtx_chunks into sctp_transport to detect
if there's any packet retransmission/loss happened by checking against
asoc's rtx_data_chunks in sctp_transport_pl_send().
If there is, namely, transport->last_rtx_chunks != asoc->rtx_data_chunks,
the pmtu probe will be sent out. Otherwise, increment the pl.raise_count
and return when it's in Search Complete state.
With this patch, if in Search Complete state, which is a long period, it
doesn't need to keep probing the current pmtu unless there's data packet
loss. This will save quite some traffic.
v1->v2:
- add the missing Fixes tag.
Fixes:
0dac127c0557 ("sctp: do black hole detection in search complete state")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Sun, 25 Jul 2021 17:42:50 +0000 (13:42 -0400)]
sctp: improve the code for pmtu probe send and recv update
This patch does 3 things:
- make sctp_transport_pl_send() and sctp_transport_pl_recv()
return bool type to decide if more probe is needed to send.
- pr_debug() only when probe is really needed to send.
- count pl.raise_count in sctp_transport_pl_send() instead of
sctp_transport_pl_recv(), and it's only incremented for the
1st probe for the same size.
These are preparations for the next patch to make probes happen
only when there's packet loss in Search Complete state.
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Harshvardhan Jha [Sun, 25 Jul 2021 17:58:04 +0000 (23:28 +0530)]
net: qede: Fix end of loop tests for list_for_each_entry
The list_for_each_entry() iterator, "vlan" in this code, can never be
NULL so the warning will never be printed.
Signed-off-by: Harshvardhan Jha <harshvardhan.jha@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jiapeng Chong [Fri, 23 Jul 2021 10:36:09 +0000 (18:36 +0800)]
mlx4: Fix missing error code in mlx4_load_one()
The error code is missing in this code scenario, add the error code
'-EINVAL' to the return value 'err'.
Eliminate the follow smatch warning:
drivers/net/ethernet/mellanox/mlx4/main.c:3538 mlx4_load_one() warn:
missing error code 'err'.
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Fixes:
7ae0e400cd93 ("net/mlx4_core: Flexible (asymmetric) allocation of EQs and MSI-X vectors for PF/VFs")
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Kevin Lo [Fri, 23 Jul 2021 13:59:27 +0000 (21:59 +0800)]
net: phy: broadcom: re-add check for PHY_BRCM_DIS_TXCRXC_NOENRGY on the BCM54811 PHY
Restore PHY_ID_BCM54811 accidently removed by commit
5d4358ede8eb.
Fixes:
5d4358ede8eb ("net: phy: broadcom: Allow BCM54210E to configure APD")
Signed-off-by: Kevin Lo <kevlo@kevlo.org>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Parav Pandit [Fri, 23 Jul 2021 14:56:00 +0000 (17:56 +0300)]
devlink: Fix phys_port_name of virtual port and merge error
Merge commit cited in fixes tag was incorrect. Due to it phys_port_name
of the virtual port resulted in incorrect name.
Also the phys_port_name of the physical port was written twice due to
the merge error.
Fix it by removing the old code and inserting back the misplaced code.
Related commits of interest in net and net-next branches that resulted
in merge conflict are:
in net-next branch:
commit
f285f37cb1e6 ("devlink: append split port number to the port name")
in net branch:
commit
b28d8f0c25a9 ("devlink: Correct VIRTUAL port to not have phys_port attributes")
Fixes:
126285651b7 ("Merge ra.kernel.org:/pub/scm/linux/kernel/git/netdev/net")
Signed-off-by: Parav Pandit <parav@nvidia.com>
Reported-by: Niklas Schnelle <schnelle@linux.ibm.com>
Tested-by: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Hariprasad Kelam [Sun, 25 Jul 2021 07:59:37 +0000 (13:29 +0530)]
octeontx2-pf: Dont enable backpressure on LBK links
Avoid configure backpressure for LBK links as they
don't support it and enable lmacs before configuration
pause frames.
Fixes:
75f36270990c ("octeontx2-pf: Support to enable/disable pause frames via ethtool")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Hariprasad Kelam <hkelam@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Sun, 25 Jul 2021 07:59:03 +0000 (13:29 +0530)]
octeontx2-pf: Fix interface down flag on error
In the existing code while changing the number of TX/RX
queues using ethtool the PF/VF interface resources are
freed and reallocated (otx2_stop and otx2_open is called)
if the device is in running state. If any resource allocation
fails in otx2_open, driver free already allocated resources
and return. But again, when the number of queues changes
as the device state still running oxt2_stop is called.
In which we try to free already freed resources leading
to driver crash.
This patch fixes the issue by setting the INTF_DOWN flag on
error and free the resources in otx2_stop only if the flag is
not set.
Fixes:
50fe6c02e5ad ("octeontx2-pf: Register and handle link notifications")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <Sunil.Goutham@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Sun, 25 Jul 2021 07:58:24 +0000 (13:28 +0530)]
octeontx2-af: Fix PKIND overlap between LBK and LMAC interfaces
Currently PKINDs are not assigned to LBK channels.
The default value of LBK_CHX_PKIND (channel to PKIND mapping) register
is zero, which is resulting in a overlap of pkind between LBK and CGX
LMACs. When KPU1 parser config is modified when PTP timestamping is
enabled on the CGX LMAC interface it is impacting traffic on LBK
interfaces as well.
This patch fixes the issue by reserving the PKIND#0 for LBK devices.
CGX mapped PF pkind starts from 1 and also fixes the max pkind available.
Fixes:
421572175ba5 ("octeontx2-af: Support to enable/disable HW timestamping")
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Fri, 23 Jul 2021 21:53:48 +0000 (17:53 -0400)]
bnxt_en: Add missing periodic PHC overflow check
We use the timecounter APIs for the 48-bit PHC and packet timestamps.
We must periodically update the timecounter at roughly half the
overflow interval. The overflow interval is about 78 hours, so
update it every 19 hours (1/4 interval) for some extra margins.
Fixes:
390862f45c85 ("bnxt_en: Get the full 48-bit hardware timestamp periodically")
Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Acked-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Fri, 23 Jul 2021 22:46:01 +0000 (18:46 -0400)]
tipc: do not write skb_shinfo frags when doing decrytion
One skb's skb_shinfo frags are not writable, and they can be shared with
other skbs' like by pskb_copy(). To write the frags may cause other skb's
data crash.
So before doing en/decryption, skb_cow_data() should always be called for
a cloned or nonlinear skb if req dst is using the same sg as req src.
While at it, the likely branch can be removed, as it will be covered
by skb_cow_data().
Note that esp_input() has the same issue, and I will fix it in another
patch. tipc_aead_encrypt() doesn't have this issue, as it only processes
linear data in the unlikely branch.
Fixes:
fc1b6d6de220 ("tipc: introduce TIPC encryption & authentication")
Reported-by: Shuang Li <shuali@redhat.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 24 Jul 2021 18:33:23 +0000 (19:33 +0100)]
Merge tag 'linux-can-fixes-for-5.14-
20210724' of git://git./linux/kernel/git/mkl/linux-can
linux-can-fixes-for-5.14-
20210724
Marc Kleine-Budde says:
====================
pull-request: can 2021-07-24
this is a pull request of 6 patches for net/master.
The first patch is by Joakim Zhang targets the imx8mp device tree. It
removes the imx6 fallback from the flexcan binding, as the imx6 is not
compatible with the imx8mp.
Ziyang Xuan contributes a patch to fix a use-after-free in the CAN
raw's raw_setsockopt().
The next two patches target the CAN J1939 protocol. The first one is
by Oleksij Rempel and clarifies the lifetime of session object in
j1939_session_deactivate(). Zhang Changzhong's patch fixes the timeout
value between consecutive TP.DT.
Stephane Grosjean contributes a patch for the peak_usb driver to fix
reading of the rxerr/txerr values.
The last patch is by me for the mcp251xfd driver. It stops the
timestamp worker in case of a fatal error in the IRQ handler.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Marc Kleine-Budde [Sat, 24 Apr 2021 14:20:39 +0000 (16:20 +0200)]
can: mcp251xfd: mcp251xfd_irq(): stop timestamping worker in case error in IRQ
In case an error occurred in the IRQ handler, the chip status is
dumped via devcoredump and all IRQs are disabled, but the chip stays
powered for further analysis.
The chip is in an undefined state and will not receive any CAN frames,
so shut down the timestamping worker, which reads the TBC register
regularly, too. This avoids any CRC read error messages if there is a
communication problem with the chip.
Fixes:
efd8d98dfb90 ("can: mcp251xfd: add HW timestamp infrastructure")
Link: https://lore.kernel.org/r/20210724155131.471303-1-mkl@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Stephane Grosjean [Fri, 25 Jun 2021 13:09:29 +0000 (15:09 +0200)]
can: peak_usb: pcan_usb_handle_bus_evt(): fix reading rxerr/txerr values
This patch fixes an incorrect way of reading error counters in messages
received for this purpose from the PCAN-USB interface. These messages
inform about the increase or decrease of the error counters, whose values
are placed in bytes 1 and 2 of the message data (not 0 and 1).
Fixes:
ea8b33bde76c ("can: pcan_usb: add support of rxerr/txerr counters")
Link: https://lore.kernel.org/r/20210625130931.27438-4-s.grosjean@peak-system.com
Cc: linux-stable <stable@vger.kernel.org>
Signed-off-by: Stephane Grosjean <s.grosjean@peak-system.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Zhang Changzhong [Tue, 6 Jul 2021 11:00:08 +0000 (19:00 +0800)]
can: j1939: j1939_xtp_rx_dat_one(): fix rxtimer value between consecutive TP.DT to 750ms
For receive side, the max time interval between two consecutive TP.DT
should be 750ms.
Fixes:
9d71dd0c7009 ("can: add support of SAE J1939 protocol")
Link: https://lore.kernel.org/r/1625569210-47506-1-git-send-email-zhangchangzhong@huawei.com
Cc: linux-stable <stable@vger.kernel.org>
Signed-off-by: Zhang Changzhong <zhangchangzhong@huawei.com>
Acked-by: Oleksij Rempel <o.rempel@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Oleksij Rempel [Wed, 14 Jul 2021 11:16:02 +0000 (13:16 +0200)]
can: j1939: j1939_session_deactivate(): clarify lifetime of session object
The j1939_session_deactivate() is decrementing the session ref-count and
potentially can free() the session. This would cause use-after-free
situation.
However, the code calling j1939_session_deactivate() does always hold
another reference to the session, so that it would not be free()ed in
this code path.
This patch adds a comment to make this clear and a WARN_ON, to ensure
that future changes will not violate this requirement. Further this
patch avoids dereferencing the session pointer as a precaution to avoid
use-after-free if the session is actually free()ed.
Fixes:
9d71dd0c7009 ("can: add support of SAE J1939 protocol")
Link: https://lore.kernel.org/r/20210714111602.24021-1-o.rempel@pengutronix.de
Reported-by: Xiaochen Zou <xzou017@ucr.edu>
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Ziyang Xuan [Thu, 22 Jul 2021 07:08:19 +0000 (15:08 +0800)]
can: raw: raw_setsockopt(): fix raw_rcv panic for sock UAF
We get a bug during ltp can_filter test as following.
===========================================
[60919.264984] BUG: unable to handle kernel NULL pointer dereference at
0000000000000010
[60919.265223] PGD
8000003dda726067 P4D
8000003dda726067 PUD
3dda727067 PMD 0
[60919.265443] Oops: 0000 [#1] SMP PTI
[60919.265550] CPU: 30 PID: 3638365 Comm: can_filter Kdump: loaded Tainted: G W 4.19.90+ #1
[60919.266068] RIP: 0010:selinux_socket_sock_rcv_skb+0x3e/0x200
[60919.293289] RSP: 0018:
ffff8d53bfc03cf8 EFLAGS:
00010246
[60919.307140] RAX:
0000000000000000 RBX:
000000000000001d RCX:
0000000000000007
[60919.320756] RDX:
0000000000000001 RSI:
ffff8d5104a8ed00 RDI:
ffff8d53bfc03d30
[60919.334319] RBP:
ffff8d9338056800 R08:
ffff8d53bfc29d80 R09:
0000000000000001
[60919.347969] R10:
ffff8d53bfc03ec0 R11:
ffffb8526ef47c98 R12:
ffff8d53bfc03d30
[60919.350320] perf: interrupt took too long (3063 > 2500), lowering kernel.perf_event_max_sample_rate to 65000
[60919.361148] R13:
0000000000000001 R14:
ffff8d53bcf90000 R15:
0000000000000000
[60919.361151] FS:
00007fb78b6b3600(0000) GS:
ffff8d53bfc00000(0000) knlGS:
0000000000000000
[60919.400812] CS: 0010 DS: 0000 ES: 0000 CR0:
0000000080050033
[60919.413730] CR2:
0000000000000010 CR3:
0000003e3f784006 CR4:
00000000007606e0
[60919.426479] DR0:
0000000000000000 DR1:
0000000000000000 DR2:
0000000000000000
[60919.439339] DR3:
0000000000000000 DR6:
00000000fffe0ff0 DR7:
0000000000000400
[60919.451608] PKRU:
55555554
[60919.463622] Call Trace:
[60919.475617] <IRQ>
[60919.487122] ? update_load_avg+0x89/0x5d0
[60919.498478] ? update_load_avg+0x89/0x5d0
[60919.509822] ? account_entity_enqueue+0xc5/0xf0
[60919.520709] security_sock_rcv_skb+0x2a/0x40
[60919.531413] sk_filter_trim_cap+0x47/0x1b0
[60919.542178] ? kmem_cache_alloc+0x38/0x1b0
[60919.552444] sock_queue_rcv_skb+0x17/0x30
[60919.562477] raw_rcv+0x110/0x190 [can_raw]
[60919.572539] can_rcv_filter+0xbc/0x1b0 [can]
[60919.582173] can_receive+0x6b/0xb0 [can]
[60919.591595] can_rcv+0x31/0x70 [can]
[60919.600783] __netif_receive_skb_one_core+0x5a/0x80
[60919.609864] process_backlog+0x9b/0x150
[60919.618691] net_rx_action+0x156/0x400
[60919.627310] ? sched_clock_cpu+0xc/0xa0
[60919.635714] __do_softirq+0xe8/0x2e9
[60919.644161] do_softirq_own_stack+0x2a/0x40
[60919.652154] </IRQ>
[60919.659899] do_softirq.part.17+0x4f/0x60
[60919.667475] __local_bh_enable_ip+0x60/0x70
[60919.675089] __dev_queue_xmit+0x539/0x920
[60919.682267] ? finish_wait+0x80/0x80
[60919.689218] ? finish_wait+0x80/0x80
[60919.695886] ? sock_alloc_send_pskb+0x211/0x230
[60919.702395] ? can_send+0xe5/0x1f0 [can]
[60919.708882] can_send+0xe5/0x1f0 [can]
[60919.715037] raw_sendmsg+0x16d/0x268 [can_raw]
It's because raw_setsockopt() concurrently with
unregister_netdevice_many(). Concurrent scenario as following.
cpu0 cpu1
raw_bind
raw_setsockopt unregister_netdevice_many
unlist_netdevice
dev_get_by_index raw_notifier
raw_enable_filters ......
can_rx_register
can_rcv_list_find(..., net->can.rx_alldev_list)
......
sock_close
raw_release(sock_a)
......
can_receive
can_rcv_filter(net->can.rx_alldev_list, ...)
raw_rcv(skb, sock_a)
BUG
After unlist_netdevice(), dev_get_by_index() return NULL in
raw_setsockopt(). Function raw_enable_filters() will add sock
and can_filter to net->can.rx_alldev_list. Then the sock is closed.
Followed by, we sock_sendmsg() to a new vcan device use the same
can_filter. Protocol stack match the old receiver whose sock has
been released on net->can.rx_alldev_list in can_rcv_filter().
Function raw_rcv() uses the freed sock. UAF BUG is triggered.
We can find that the key issue is that net_device has not been
protected in raw_setsockopt(). Use rtnl_lock to protect net_device
in raw_setsockopt().
Fixes:
c18ce101f2e4 ("[CAN]: Add raw protocol")
Link: https://lore.kernel.org/r/20210722070819.1048263-1-william.xuanziyang@huawei.com
Cc: linux-stable <stable@vger.kernel.org>
Signed-off-by: Ziyang Xuan <william.xuanziyang@huawei.com>
Acked-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Joakim Zhang [Mon, 19 Jul 2021 07:34:37 +0000 (15:34 +0800)]
arm64: dts: imx8mp: remove fallback compatible string for FlexCAN
FlexCAN on i.MX8MP is not derived from i.MX6Q, instead reuses from
i.MX8QM with extra ECC added and default is enabled, so that the FlexCAN
would be put into freeze mode without FLEXCAN_QUIRK_DISABLE_MECR quirk.
This patch removes "fsl,imx6q-flexcan" fallback compatible string since
it's not compatible with the i.MX6Q.
Link: https://lore.kernel.org/r/20210719073437.32078-1-qiangqing.zhang@nxp.com
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
David S. Miller [Fri, 23 Jul 2021 20:57:52 +0000 (21:57 +0100)]
Merge branch 'ionic-fixes'
Shannon Nelson says:
====================
ionic: bug fixes
Fix a thread race in rx_mode, remove unnecessary log message,
fix dynamic coalescing issues, and count all csum_none cases.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Shannon Nelson [Fri, 23 Jul 2021 18:02:49 +0000 (11:02 -0700)]
ionic: count csum_none when offload enabled
Be sure to count the csum_none cases when csum offload is
enabled.
Fixes:
0f3154e6bcb3 ("ionic: Add Tx and Rx handling")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Shannon Nelson [Fri, 23 Jul 2021 18:02:48 +0000 (11:02 -0700)]
ionic: fix up dim accounting for tx and rx
We need to count the correct Tx and/or Rx packets for dynamic
interrupt moderation, depending on which we're processing on
the queue interrupt.
Fixes:
04a834592bf5 ("ionic: dynamic interrupt moderation")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Shannon Nelson [Fri, 23 Jul 2021 18:02:47 +0000 (11:02 -0700)]
ionic: remove intr coalesce update from napi
Move the interrupt coalesce value update out of the napi
thread and into the dim_work thread and set it only when it
has actually changed.
Fixes:
04a834592bf5 ("ionic: dynamic interrupt moderation")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Shannon Nelson [Fri, 23 Jul 2021 18:02:46 +0000 (11:02 -0700)]
ionic: catch no ptp support earlier
If PTP configuration is attempted on ports that don't support
it, such as VF ports, the driver will return an error status
-95, or EOPNOSUPP and print an error message
enp98s0: hwstamp set failed: -95
Because some daemons can retry every few seconds, this can end
up filling the dmesg log and pushing out other more useful
messages.
We can catch this issue earlier in our handling and return
the error without a log message.
Fixes:
829600ce5e4e ("ionic: add ts_config replay")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
Shannon Nelson [Fri, 23 Jul 2021 18:02:45 +0000 (11:02 -0700)]
ionic: make all rx_mode work threadsafe
Move the bulk of the code from ionic_set_rx_mode(), which
can be called from atomic context, into ionic_lif_rx_mode()
which is a safe context.
A call from the stack will get pushed off into a work thread,
but it is also possible to simultaneously have a call driven
by a queue reconfig request from an ethtool command or fw
recovery event. We add a mutex around the rx_mode work to be
sure they don't collide.
Fixes:
81dbc24147f9 ("ionic: change set_rx_mode from_ndo to can_sleep")
Signed-off-by: Shannon Nelson <snelson@pensando.io>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 23 Jul 2021 20:21:42 +0000 (21:21 +0100)]
Merge branch '40GbE' of git://git./linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2021-07-23
This series contains updates to i40e driver only.
Arkadiusz corrects the order of calls for disabling queues to resolve
a false error message and adds a better message to the user when
transitioning FW LLDP back on while the firmware is still processing
the off request.
Lukasz adds additional information regarding possible incorrect cable
use when a PHY type error occurs.
Jedrzej adds ndo_select_queue support to resolve incorrect queue
selection when SW DCB is used and adds a warning when there are not
enough queues for desired TC configuration.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 23 Jul 2021 16:57:09 +0000 (17:57 +0100)]
Merge tag 'mac80211-for-net-2021-07-23' of git://git./linux/kernel/git/jberg/mac80211
Couple of fixes:
* fix aggregation on mesh
* fix late enabling of 4-addr mode
* leave monitor SKBs with some headroom
* limit band information for old applications
* fix virt-wifi WARN_ON
* fix memory leak in cfg80211 BSS list maintenance
Paul Jakma [Fri, 23 Jul 2021 15:13:04 +0000 (16:13 +0100)]
NIU: fix incorrect error return, missed in previous revert
Commit
7930742d6, reverting 26fd962, missed out on reverting an incorrect
change to a return value. The niu_pci_vpd_scan_props(..) == 1 case appears
to be a normal path - treating it as an error and return -EINVAL was
breaking VPD_SCAN and causing the driver to fail to load.
Fix, so my Neptune card works again.
Cc: Kangjie Lu <kjlu@umn.edu>
Cc: Shannon Nelson <shannon.lee.nelson@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: stable <stable@vger.kernel.org>
Fixes:
7930742d ('Revert "niu: fix missing checks of niu_pci_eeprom_read"')
Signed-off-by: Paul Jakma <paul@jakma.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pavel Skripkin [Fri, 23 Jul 2021 15:31:32 +0000 (18:31 +0300)]
net: qrtr: fix memory leaks
Syzbot reported memory leak in qrtr. The problem was in unputted
struct sock. qrtr_local_enqueue() function calls qrtr_port_lookup()
which takes sock reference if port was found. Then there is the following
check:
if (!ipc || &ipc->sk == skb->sk) {
...
return -ENODEV;
}
Since we should drop the reference before returning from this function and
ipc can be non-NULL inside this if, we should add qrtr_port_put() inside
this if.
The similar corner case is in qrtr_endpoint_post() as Manivannan
reported. In case of sock_queue_rcv_skb() failure we need to put
port reference to avoid leaking struct sock pointer.
Fixes:
e04df98adf7d ("net: qrtr: Remove receive worker")
Fixes:
bdabad3e363d ("net: Add Qualcomm IPC router")
Reported-and-tested-by: syzbot+35a511c72ea7356cdcf3@syzkaller.appspotmail.com
Signed-off-by: Pavel Skripkin <paskripkin@gmail.com>
Reviewed-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 23 Jul 2021 16:46:05 +0000 (17:46 +0100)]
Merge git://git./pub/scm/linux/kernel/git/pablo/nf
Pablo Neira Ayusosays:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net:
1) Memleak in commit audit error path, from Dongliang Mu.
2) Avoid possible false sharing for flowtable timeout updates
and nft_last use.
3) Adjust conntrack timestamp due to garbage collection delay,
from Florian Westphal.
4) Fix nft_nat without layer 3 address for the inet family.
5) Fix compilation warning in nfnl_hook when ingress support
is disabled, from Arnd Bergmann.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Subbaraya Sundeep [Fri, 23 Jul 2021 08:06:18 +0000 (13:36 +0530)]
octeontx2-af: Fix uninitialized variables in rvu_switch
Get the number of VFs of a PF correctly by calling
rvu_get_pf_numvfs in rvu_switch_disable function.
Also hwvf is not required hence remove it.
Fixes:
23109f8dd06d ("octeontx2-af: Introduce internal packet switching")
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Loic Poulain [Thu, 22 Jul 2021 18:21:05 +0000 (20:21 +0200)]
wwan: core: Fix missing RTM_NEWLINK event for default link
A wwan link created via the wwan_create_default_link procedure is
never notified to the user (RTM_NEWLINK), causing issues with user
tools relying on such event to track network links (NetworkManager).
This is because the procedure misses a call to rtnl_configure_link(),
which sets the link as initialized and notifies the new link (cf
proper usage in __rtnl_newlink()).
Cc: stable@vger.kernel.org
Fixes:
ca374290aaad ("wwan: core: support default netdev creation")
Suggested-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Acked-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Vladimir Oltean [Thu, 22 Jul 2021 13:05:51 +0000 (16:05 +0300)]
net: dsa: mv88e6xxx: silently accept the deletion of VID 0 too
The blamed commit modified the driver to accept the addition of VID 0
without doing anything, but deleting that VID still fails:
[ 32.080780] mv88e6085
d0032004.mdio-mii:10 lan8: failed to kill vid 0081/0
Modify mv88e6xxx_port_vlan_leave() to do the same thing as the addition.
Fixes:
b8b79c414eca ("net: dsa: mv88e6xxx: Fix adding vlan 0")
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Kangmin Park [Thu, 22 Jul 2021 17:44:43 +0000 (02:44 +0900)]
ipv6: decrease hop limit counter in ip6_forward()
Decrease hop limit counter when deliver skb to ndp proxy.
Signed-off-by: Kangmin Park <l4stpr0gr4m@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gilad Naaman [Thu, 22 Jul 2021 17:01:28 +0000 (20:01 +0300)]
net: Set true network header for ECN decapsulation
In cases where the header straight after the tunnel header was
another ethernet header (TEB), instead of the network header,
the ECN decapsulation code would treat the ethernet header as if
it was an IP header, resulting in mishandling and possible
wrong drops or corruption of the IP header.
In this case, ECT(1) is sent, so IP_ECN_decapsulate tries to copy it to the
inner IPv4 header, and correct its checksum.
The offset of the ECT bits in an IPv4 header corresponds to the
lower 2 bits of the second octet of the destination MAC address
in the ethernet header.
The IPv4 checksum corresponds to end of the source address.
In order to reproduce:
$ ip netns add A
$ ip netns add B
$ ip -n A link add _v0 type veth peer name _v1 netns B
$ ip -n A link set _v0 up
$ ip -n A addr add dev _v0 10.254.3.1/24
$ ip -n A route add default dev _v0 scope global
$ ip -n B link set _v1 up
$ ip -n B addr add dev _v1 10.254.1.6/24
$ ip -n B route add default dev _v1 scope global
$ ip -n B link add gre1 type gretap local 10.254.1.6 remote 10.254.3.1 key 0x49000000
$ ip -n B link set gre1 up
# Now send an IPv4/GRE/Eth/IPv4 frame where the outer header has ECT(1),
# and the inner header has no ECT bits set:
$ cat send_pkt.py
#!/usr/bin/env python3
from scapy.all import *
pkt = IP(b'E\x01\x00\xa7\x00\x00\x00\x00@/`%\n\xfe\x03\x01\n\xfe\x01\x06 \x00eXI\x00'
b'\x00\x00\x18\xbe\x92\xa0\xee&\x18\xb0\x92\xa0l&\x08\x00E\x00\x00}\x8b\x85'
b'@\x00\x01\x01\xe4\xf2\x82\x82\x82\x01\x82\x82\x82\x02\x08\x00d\x11\xa6\xeb'
b'3\x1e\x1e\\xf3\\xf7`\x00\x00\x00\x00ZN\x00\x00\x00\x00\x00\x00\x10\x11\x12'
b'\x13\x14\x15\x16\x17\x18\x19\x1a\x1b\x1c\x1d\x1e\x1f !"#$%&\'()*+,-./01234'
b'56789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ')
send(pkt)
$ sudo ip netns exec B tcpdump -neqlllvi gre1 icmp & ; sleep 1
$ sudo ip netns exec A python3 send_pkt.py
In the original packet, the source/destinatio MAC addresses are
dst=18:be:92:a0:ee:26 src=18:b0:92:a0:6c:26
In the received packet, they are
dst=18:bd:92:a0:ee:26 src=18:b0:92:a0:6c:27
Thanks to Lahav Schlesinger <lschlesinger@drivenets.com> and Isaac Garzon <isaac@speed.io>
for helping me pinpoint the origin.
Fixes:
b723748750ec ("tunnel: Propagate ECT(1) when decapsulating as recommended by RFC6040")
Cc: David S. Miller <davem@davemloft.net>
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Cc: David Ahern <dsahern@kernel.org>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Gilad Naaman <gnaaman@drivenets.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Hoang Le [Fri, 23 Jul 2021 02:25:34 +0000 (09:25 +0700)]
tipc: fix sleeping in tipc accept routine
The release_sock() is blocking function, it would change the state
after sleeping. In order to evaluate the stated condition outside
the socket lock context, switch to use wait_woken() instead.
Fixes:
6398e23cdb1d8 ("tipc: standardize accept routine")
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: Hoang Le <hoang.h.le@dektech.com.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Xin Long [Thu, 22 Jul 2021 16:05:41 +0000 (12:05 -0400)]
tipc: fix implicit-connect for SYN+
For implicit-connect, when it's either SYN- or SYN+, an ACK should
be sent back to the client immediately. It's not appropriate for
the client to enter established state only after receiving data
from the server.
On client side, after the SYN is sent out, tipc_wait_for_connect()
should be called to wait for the ACK if timeout is set.
This patch also restricts __tipc_sendstream() to call __sendmsg()
only when it's in TIPC_OPEN state, so that the client can program
in a single loop doing both connecting and data sending like:
for (...)
sendmsg(dest, buf);
This makes the implicit-connect more implicit.
Fixes:
b97bf3fd8f6a ("[TIPC] Initial merge")
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Acked-by: Jon Maloy <jmaloy@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Thu, 22 Jul 2021 12:45:51 +0000 (18:15 +0530)]
octeontx2-af: Remove unnecessary devm_kfree
Remove devm_kfree of memory where VLAN entry to RVU PF mapping
info is saved. This will be freed anyway at driver exit.
Having this could result in warning from devm_kfree() if
the memory is not allocated due to errors in rvu_nix_block_init()
before nix_setup_txvlan().
Fixes:
9a946def264d ("octeontx2-af: Modify nix_vtag_cfg mailbox to support TX VTAG entries")
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jedrzej Jagielski [Fri, 18 Jun 2021 08:49:49 +0000 (08:49 +0000)]
i40e: Fix log TC creation failure when max num of queues is exceeded
Fix missing failed message if driver does not have enough queues to
complete TC command. Without this fix no message is displayed in dmesg.
Fixes:
a9ce82f744dc ("i40e: Enable 'channel' mode in mqprio for TC configs")
Signed-off-by: Grzegorz Szczurek <grzegorzx.szczurek@intel.com>
Signed-off-by: Jedrzej Jagielski <jedrzej.jagielski@intel.com>
Tested-by: Imam Hassan Reza Biswas <imam.hassan.reza.biswas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Jedrzej Jagielski [Wed, 2 Jun 2021 00:47:03 +0000 (00:47 +0000)]
i40e: Fix queue-to-TC mapping on Tx
In SW DCB mode the packets sent receive incorrect UP tags. They are
constructed correctly and put into tx_ring, but UP is later remapped by
HW on the basis of TCTUPR register contents according to Tx queue
selected, and BW used is consistent with the new UP values. This is
caused by Tx queue selection in kernel not taking into account DCB
configuration. This patch fixes the issue by implementing the
ndo_select_queue NDO callback.
Fixes:
fd0a05ce74ef ("i40e: transmit, receive, and NAPI")
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Signed-off-by: Jedrzej Jagielski <jedrzej.jagielski@intel.com>
Tested-by: Imam Hassan Reza Biswas <imam.hassan.reza.biswas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Lukasz Cieplicki [Mon, 31 May 2021 16:55:49 +0000 (16:55 +0000)]
i40e: Add additional info to PHY type error
In case of PHY type error occurs, the message was too generic.
Add additional info to PHY type error indicating that it can be
wrong cable connected.
Fixes:
124ed15bf126 ("i40e: Add dual speed module support")
Signed-off-by: Lukasz Cieplicki <lukaszx.cieplicki@intel.com>
Signed-off-by: Michal Maloszewski <michal.maloszewski@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Arkadiusz Kubalewski [Fri, 21 May 2021 16:41:26 +0000 (18:41 +0200)]
i40e: Fix firmware LLDP agent related warning
Make warning meaningful for the user.
Previously the trace:
"Starting FW LLDP agent failed: error: I40E_ERR_ADMIN_QUEUE_ERROR, I40E_AQ_RC_EAGAIN"
was produced when user tried to start Firmware LLDP agent,
just after it was stopped with sequence:
ethtool --set-priv-flags <dev> disable-fw-lldp on
ethtool --set-priv-flags <dev> disable-fw-lldp off
(without any delay between the commands)
At that point the firmware is still processing stop command, the behavior
is expected.
Fixes:
c1041d070437 ("i40e: Missing response checks in driver when starting/stopping FW LLDP")
Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Tested-by: Imam Hassan Reza Biswas <imam.hassan.reza.biswas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Arkadiusz Kubalewski [Thu, 29 Apr 2021 17:49:47 +0000 (19:49 +0200)]
i40e: Fix logic of disabling queues
Correct the message flow between driver and firmware when disabling
queues.
Previously in case of PF reset (due to required reinit after reconfig),
the error like: "VSI seid 397 Tx ring 60 disable timeout" could show up
occasionally. The error was not a real issue of hardware or firmware,
it was caused by wrong sequence of messages invoked by the driver.
Fixes:
41c445ff0f48 ("i40e: main driver core")
Signed-off-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com>
Signed-off-by: Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com>
Tested-by: Tony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Arnd Bergmann [Wed, 21 Jul 2021 15:22:32 +0000 (17:22 +0200)]
netfilter: nfnl_hook: fix unused variable warning
The only user of this variable is in an #ifdef:
net/netfilter/nfnetlink_hook.c: In function 'nfnl_hook_entries_head':
net/netfilter/nfnetlink_hook.c:177:28: error: unused variable 'netdev' [-Werror=unused-variable]
Fixes:
e2cf17d3774c ("netfilter: add new hook nfnl subsystem")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso [Tue, 20 Jul 2021 16:22:50 +0000 (18:22 +0200)]
netfilter: nft_nat: allow to specify layer 4 protocol NAT only
nft_nat reports a bogus EAFNOSUPPORT if no layer 3 information is specified.
Fixes:
d07db9884a5f ("netfilter: nf_tables: introduce nft_validate_register_load()")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Sun, 18 Jul 2021 16:36:00 +0000 (18:36 +0200)]
netfilter: conntrack: adjust stop timestamp to real expiry value
In case the entry is evicted via garbage collection there is
delay between the timeout value and the eviction event.
This adjusts the stop value based on how much time has passed.
Fixes:
b87a2f9199ea82 ("netfilter: conntrack: add gc worker to remove timed-out entries")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso [Sat, 17 Jul 2021 08:20:08 +0000 (10:20 +0200)]
netfilter: nft_last: avoid possible false sharing
Use the idiom described in:
https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE#it-may-improve-performance
Moreover, prevent a compiler optimization.
Fixes:
836382dc2471 ("netfilter: nf_tables: add last expression")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso [Sat, 17 Jul 2021 08:10:29 +0000 (10:10 +0200)]
netfilter: flowtable: avoid possible false sharing
The flowtable follows the same timeout approach as conntrack, use the
same idiom as in
cc16921351d8 ("netfilter: conntrack: avoid same-timeout
update") but also include the fix provided by
e37542ba111f ("netfilter:
conntrack: avoid possible false sharing").
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Nguyen Dinh Phi [Mon, 28 Jun 2021 13:23:34 +0000 (21:23 +0800)]
cfg80211: Fix possible memory leak in function cfg80211_bss_update
When we exceed the limit of BSS entries, this function will free the
new entry, however, at this time, it is the last door to access the
inputed ies, so these ies will be unreferenced objects and cause memory
leak.
Therefore we should free its ies before deallocating the new entry, beside
of dropping it from hidden_list.
Signed-off-by: Nguyen Dinh Phi <phind.uet@gmail.com>
Link: https://lore.kernel.org/r/20210628132334.851095-1-phind.uet@gmail.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Johannes Berg [Mon, 12 Jul 2021 19:53:30 +0000 (21:53 +0200)]
nl80211: limit band information in non-split data
In non-split data, we shouldn't be adding S1G and 6 GHz
data (or future bands) since we're really close to the
4k message size limit. Remove those bands, any modern
userspace that can use S1G or 6 GHz should already be
using split dumps, and if not then it needs to update.
Link: https://lore.kernel.org/r/20210712215329.31444162a2c2.I5555312e4a074c84f8b4e7ad79dc4d1fbfc5126c@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Matteo Croce [Tue, 6 Jul 2021 15:44:23 +0000 (17:44 +0200)]
virt_wifi: fix error on connect
When connecting without first doing a scan, the BSS list is empty
and __cfg80211_connect_result() generates this warning:
$ iw dev wlan0 connect -w VirtWifi
[ 15.371989] ------------[ cut here ]------------
[ 15.372179] WARNING: CPU: 0 PID: 92 at net/wireless/sme.c:756 __cfg80211_connect_result+0x402/0x440
[ 15.372383] CPU: 0 PID: 92 Comm: kworker/u2:2 Not tainted 5.13.0-kvm #444
[ 15.372512] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-3.fc34 04/01/2014
[ 15.372597] Workqueue: cfg80211 cfg80211_event_work
[ 15.372756] RIP: 0010:__cfg80211_connect_result+0x402/0x440
[ 15.372818] Code: 48 2b 04 25 28 00 00 00 75 59 48 8b 3b 48 8b 76 10 48 8d 65 e0 5b 41 5c 41 5d 41 5e 5d 49 8d 65 f0 41 5d e9 d0 d4 fd ff 0f 0b <0f> 0b e9 f6 fd ff ff e8 f2 4a b4 ff e9 ec fd ff ff 0f 0b e9 19 fd
[ 15.372966] RSP: 0018:
ffffc900005cbdc0 EFLAGS:
00010246
[ 15.373022] RAX:
0000000000000000 RBX:
ffff8880028e2400 RCX:
ffff8880028e2472
[ 15.373088] RDX:
0000000000000002 RSI:
00000000fffffe01 RDI:
ffffffff815335ba
[ 15.373149] RBP:
ffffc900005cbe00 R08:
0000000000000008 R09:
ffff888002bdf8b8
[ 15.373209] R10:
ffff88803ec208f0 R11:
ffffffffffffe9ae R12:
ffff88801d687d98
[ 15.373280] R13:
ffff88801b5fe000 R14:
ffffc900005cbdc0 R15:
dead000000000100
[ 15.373330] FS:
0000000000000000(0000) GS:
ffff88803ec00000(0000) knlGS:
0000000000000000
[ 15.373382] CS: 0010 DS: 0000 ES: 0000 CR0:
0000000080050033
[ 15.373425] CR2:
000056421c468958 CR3:
000000001b458001 CR4:
0000000000170eb0
[ 15.373478] DR0:
0000000000000000 DR1:
0000000000000000 DR2:
0000000000000000
[ 15.373529] DR3:
0000000000000000 DR6:
00000000fffe0ff0 DR7:
0000000000000400
[ 15.373580] Call Trace:
[ 15.373611] ? cfg80211_process_wdev_events+0x10e/0x170
[ 15.373743] cfg80211_process_wdev_events+0x10e/0x170
[ 15.373783] cfg80211_process_rdev_events+0x21/0x40
[ 15.373846] cfg80211_event_work+0x20/0x30
[ 15.373892] process_one_work+0x1e9/0x340
[ 15.373956] worker_thread+0x4b/0x3f0
[ 15.374017] ? process_one_work+0x340/0x340
[ 15.374053] kthread+0x11f/0x140
[ 15.374089] ? set_kthread_struct+0x30/0x30
[ 15.374153] ret_from_fork+0x1f/0x30
[ 15.374187] ---[ end trace
321ef0cb7e9c0be1 ]---
wlan0 (phy #0): connected to 00:00:00:00:00:00
Add the fake bss just before the connect so that cfg80211_get_bss()
finds the virtual network.
As some code was duplicated, move it in a common function.
Signed-off-by: Matteo Croce <mcroce@microsoft.com>
Link: https://lore.kernel.org/r/20210706154423.11065-1-mcroce@linux.microsoft.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Felix Fietkau [Fri, 2 Jul 2021 05:01:11 +0000 (07:01 +0200)]
mac80211: fix enabling 4-address mode on a sta vif after assoc
Notify the driver about the 4-address mode change and also send a nulldata
packet to the AP to notify it about the change
Fixes:
1ff4e8f2dec8 ("mac80211: notify the driver when a sta uses 4-address mode")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Link: https://lore.kernel.org/r/20210702050111.47546-1-nbd@nbd.name
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Felix Fietkau [Tue, 29 Jun 2021 11:28:53 +0000 (13:28 +0200)]
mac80211: fix starting aggregation sessions on mesh interfaces
The logic for starting aggregation sessions was recently moved from minstrel_ht
to mac80211, into the subif tx handler just after the sta lookup.
Unfortunately this didn't work for mesh interfaces, since the sta lookup is
deferred until a much later point in time on those.
Fix this by also calling the aggregation check right after the deferred sta
lookup.
Fixes:
08a46c642001 ("mac80211: move A-MPDU session check from minstrel_ht to mac80211")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Link: https://lore.kernel.org/r/20210629112853.29785-1-nbd@nbd.name
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Johan Almbladh [Mon, 28 Jun 2021 12:37:13 +0000 (14:37 +0200)]
mac80211: Do not strip skb headroom on monitor frames
When a monitor interface is present together with other interfaces, a
received skb is copied and received on the monitor netdev. Before, the
copied skb was allocated with exactly the amount of space needed for
the radiotap header, resulting in an skb without any headroom at all
being received on the monitor netdev. With the introduction of eBPF
and XDP in the kernel, skbs may be processed by custom eBPF programs.
However, since the skb cannot be reallocated in the eBPF program, no
more data or headers can be pushed. The old code made sure the final
headroom was zero regardless of the value of NET_SKB_PAD, so increasing
that constant would have no effect.
Now we allocate monitor skb copies with a headroom of NET_SKB_PAD bytes
before the radiotap header. Monitor interfaces now behave in the same
way as other netdev interfaces that honor the NET_SKB_PAD constant.
Signed-off-by: Johan Almbladh <johan.almbladh@anyfinetworks.com>
Link: https://lore.kernel.org/r/20210628123713.2070753-1-johan.almbladh@anyfinetworks.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Linus Torvalds [Thu, 22 Jul 2021 17:38:19 +0000 (10:38 -0700)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux
Pull arm64 fixes from Will Deacon:
"A pair of arm64 fixes for -rc3. The straightforward one is a fix to
our firmware calling stub, which accidentally started corrupting the
link register on machines with SVE. Since these machines don't really
exist yet, it wasn't spotted in -next.
The other fix is a revert-and-a-bit of a patch originally intended to
allow PTE-level huge mappings for the VMAP area on 32-bit PPC 8xx. A
side-effect of this change was that our pXd_set_huge() implementations
could be replaced with generic dummy functions depending on the levels
of page-table being used, which in turn broke the boot if we fail to
create the linear mapping as a result of using these functions to
operate on the pgd. Huge thanks to Michael Ellerman for modifying the
revert so as not to regress PPC 8xx in terms of functionality.
Anyway, that's the background and it's also available in the commit
message along with Link tags pointing at all of the fun.
Summary:
- Fix hang when issuing SMC on SVE-capable system due to
clobbered LR
- Fix boot failure due to missing block mappings with folded
page-table"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
Revert "mm/pgtable: add stubs for {pmd/pub}_{set/clear}_huge"
arm64: smccc: Save lr before calling __arm_smccc_sve_check()
Linus Torvalds [Thu, 22 Jul 2021 17:22:52 +0000 (10:22 -0700)]
Merge tag 'hyperv-fixes-signed-
20210722' of git://git./linux/kernel/git/hyperv/linux
Pull hyperv fixes from Wei Liu:
- bug fix from Haiyang for vmbus CPU assignment
- revert of a bogus patch that went into 5.14-rc1
* tag 'hyperv-fixes-signed-
20210722' of git://git.kernel.org/pub/scm/linux/kernel/git/hyperv/linux:
Revert "x86/hyperv: fix logical processor creation"
Drivers: hv: vmbus: Fix duplicate CPU assignments within a device
Linus Torvalds [Thu, 22 Jul 2021 17:11:27 +0000 (10:11 -0700)]
Merge git://git./linux/kernel/git/netdev/net
Pull networking fixes from David Miller:
1) Fix type of bind option flag in af_xdp, from Baruch Siach.
2) Fix use after free in bpf_xdp_link_release(), from Xuan Zhao.
3) PM refcnt imbakance in r8152, from Takashi Iwai.
4) Sign extension ug in liquidio, from Colin Ian King.
5) Mising range check in s390 bpf jit, from Colin Ian King.
6) Uninit value in caif_seqpkt_sendmsg(), from Ziyong Xuan.
7) Fix skb page recycling race, from Ilias Apalodimas.
8) Fix memory leak in tcindex_partial_destroy_work, from Pave Skripkin.
9) netrom timer sk refcnt issues, from Nguyen Dinh Phi.
10) Fix data races aroun tcp's tfo_active_disable_stamp, from Eric
Dumazet.
11) act_skbmod should only operate on ethernet packets, from Peilin Ye.
12) Fix slab out-of-bpunds in fib6_nh_flush_exceptions(),, from Psolo
Abeni.
13) Fix sparx5 dependencies, from Yajun Deng.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (74 commits)
dpaa2-switch: seed the buffer pool after allocating the swp
net: sched: cls_api: Fix the the wrong parameter
net: sparx5: fix unmet dependencies warning
net: dsa: tag_ksz: dont let the hardware process the layer 4 checksum
net: dsa: ensure linearized SKBs in case of tail taggers
ravb: Remove extra TAB
ravb: Fix a typo in comment
net: dsa: sja1105: make VID 4095 a bridge VLAN too
tcp: disable TFO blackhole logic by default
sctp: do not update transport pathmtu if SPP_PMTUD_ENABLE is not set
net: ixp46x: fix ptp build failure
ibmvnic: Remove the proper scrq flush
selftests: net: add ESP-in-UDP PMTU test
udp: check encap socket in __udp_lib_err
sctp: update active_key for asoc when old key is being replaced
r8169: Avoid duplicate sysfs entry creation error
ixgbe: Fix packet corruption due to missing DMA sync
Revert "qed: fix possible unpaired spin_{un}lock_bh in _qed_mcp_cmd_and_union()"
ipv6: fix another slab-out-of-bounds in fib6_nh_flush_exceptions
fsl/fman: Add fibre support
...
Linus Torvalds [Thu, 22 Jul 2021 16:51:38 +0000 (09:51 -0700)]
Merge tag 'mmc-v5.14-rc1' of git://git./linux/kernel/git/ulfh/mmc
Pull MMC fixes from Ulf Hansson:
- Use kref to fix KASAN splats triggered during card removal
- Don't allocate IDA for OF aliases
* tag 'mmc-v5.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
mmc: core: Don't allocate IDA for OF aliases
mmc: core: Use kref in place of struct mmc_blk_data::usage
Ioana Ciornei [Thu, 22 Jul 2021 12:15:51 +0000 (15:15 +0300)]
dpaa2-switch: seed the buffer pool after allocating the swp
Any interraction with the buffer pool (seeding a buffer, acquire one) is
made through a software portal (SWP, a DPIO object).
There are circumstances where the dpaa2-switch driver probes on a DPSW
before any DPIO devices have been probed. In this case, seeding of the
buffer pool will lead to a panic since no SWPs are initialized.
To fix this, seed the buffer pool after making sure that the software
portals have been probed and are ready to be used.
Fixes:
0b1b71370458 ("staging: dpaa2-switch: handle Rx path on control interface")
Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Yajun Deng [Thu, 22 Jul 2021 03:23:43 +0000 (11:23 +0800)]
net: sched: cls_api: Fix the the wrong parameter
The 4th parameter in tc_chain_notify() should be flags rather than seq.
Let's change it back correctly.
Fixes:
32a4f5ecd738 ("net: sched: introduce chain object to uapi")
Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Signed-off-by: David S. Miller <davem@davemloft.net>
Randy Dunlap [Wed, 21 Jul 2021 22:33:36 +0000 (15:33 -0700)]
net: sparx5: fix unmet dependencies warning
WARNING: unmet direct dependencies detected for PHY_SPARX5_SERDES
Depends on [n]: (ARCH_SPARX5 || COMPILE_TEST [=n]) && OF [=y] && HAS_IOMEM [=y]
Selected by [y]:
- SPARX5_SWITCH [=y] && NETDEVICES [=y] && ETHERNET [=y] && NET_VENDOR_MICROCHIP [=y] && NET_SWITCHDEV [=y] && HAS_IOMEM [=y] && OF [=y]
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Lars Povlsen <lars.povlsen@microchip.com>
Cc: Steen Hegelund <Steen.Hegelund@microchip.com>
Cc: UNGLinuxDriver@microchip.com
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Thu, 22 Jul 2021 06:15:06 +0000 (23:15 -0700)]
Merge branch 'ksz-dsa-fixes'
Lino Sanfilippo says:
====================
Fixes for KSZ DSA switch
These patches fix issues I encountered while using a KSZ9897 as a DSA
switch with a broadcom GENET network device as the DSA master device.
PATCH 1 fixes an invalid access to an SKB in case it is scattered.
PATCH 2 fixes incorrect hardware checksum calculation caused by the DSA
tag.
Changes in v2:
- instead of linearizing the SKBs only for KSZ switches ensure linearized
SKBs for all tail taggers by clearing the feature flags NETIF_F_HW_SG and
NETIF_F_FRAGLIST (suggested by Vladimir Oltean)
The patches have been tested with a KSZ9897 and apply against net-next.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Lino Sanfilippo [Wed, 21 Jul 2021 21:56:42 +0000 (23:56 +0200)]
net: dsa: tag_ksz: dont let the hardware process the layer 4 checksum
If the checksum calculation is offloaded to the network device (e.g due to
NETIF_F_HW_CSUM inherited from the DSA master device), the calculated
layer 4 checksum is incorrect. This is since the DSA tag which is placed
after the layer 4 data is considered as being part of the daa and thus
errorneously included into the checksum calculation.
To avoid this, always calculate the layer 4 checksum in software.
Signed-off-by: Lino Sanfilippo <LinoSanfilippo@gmx.de>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Lino Sanfilippo [Wed, 21 Jul 2021 21:56:41 +0000 (23:56 +0200)]
net: dsa: ensure linearized SKBs in case of tail taggers
The function skb_put() that is used by tail taggers to make room for the
DSA tag must only be called for linearized SKBS. However in case that the
slave device inherited features like NETIF_F_HW_SG or NETIF_F_FRAGLIST the
SKB passed to the slaves transmit function may not be linearized.
Avoid those SKBs by clearing the NETIF_F_HW_SG and NETIF_F_FRAGLIST flags
for tail taggers.
Furthermore since the tagging protocol can be changed at runtime move the
code for setting up the slaves features into dsa_slave_setup_tagger().
Suggested-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Lino Sanfilippo <LinoSanfilippo@gmx.de>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Biju Das [Wed, 21 Jul 2021 18:21:26 +0000 (19:21 +0100)]
ravb: Remove extra TAB
Align the member description comments for struct ravb_desc by
removing the extra TAB.
Signed-off-by: Biju Das <biju.das.jz@bp.renesas.com>
Reviewed-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
Reviewed-by: Sergei Shtylyov <sergei.shtylyov@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>