Stanislav Fomichev says:
====================
xsk: TX metadata
This series implements initial TX metadata (offloads) for AF_XDP.
See patch #2 for the main implementation and mlx5/stmmac ones for the
example on how to consume the metadata on the device side.
Starting with two types of offloads:
- request TX timestamp (and write it back into the metadata area)
- request TX checksum offload
Changes since v5:
- preserve xsk_tx_metadata flags across tx and completion by moving
them out of 'request' union (Jesper)
- fix xdp_metadata checksum failure in big endian (Alexei)
- add SPDX to xdp-rx-metadata.rst (Simon)
v5: https://lore.kernel.org/bpf/20231102225837.1141915-1-sdf@google.com/
Performance (mlx5):
I've implemented a small xskgen tool to try to saturate single tx queue:
https://github.com/fomichev/xskgen/tree/master
Here are the performance numbers with some analysis.
1. Baseline. Running with commit eb62e6aef9 ("Merge branch 'bpf:
Support bpf_get_func_ip helper in uprobes'"), nothing from this series:
- with 1400 bytes of payload: 98 gbps, 8 mpps
./xskgen -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.189130 sec, 98.357623 gbps 8.409509 mpps
- with 200 bytes of payload: 49 gbps, 23 mpps
./xskgen -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000064 packets 20960134144 bits, took 0.422235 sec, 49.640921 gbps 23.683645 mpps
2. Adding single commit that supports reserving tx_metadata_len
changes nothing numbers-wise.
- baseline for 1400
./xskgen -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.189247 sec, 98.347946 gbps 8.408682 mpps
- baseline for 200
./xskgen -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 20960000000 bits, took 0.421248 sec, 49.756913 gbps 23.738985 mpps
3. Adding -M flag causes xskgen to reserve the metadata and fill it, but
doesn't set XDP_TX_METADATA descriptor option.
- new baseline for 1400 (with only filling the metadata)
./xskgen -M -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.188767 sec, 98.387657 gbps 8.412077 mpps
- new baseline for 200 (with only filling the metadata)
./xskgen -M -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 20960000000 bits, took 0.410213 sec, 51.095407 gbps 24.377579 mpps
(the numbers go sligtly up here, not really sure why, maybe some cache-related
side-effects?
4. Next, I'm running the same test but with the commit that adds actual
general infra to parse XDP_TX_METADATA (but no driver support).
Essentially applying "xsk: add TX timestamp and TX checksum offload support"
from this series. Numbers are the same.
- fill metadata for 1400
./xskgen -M -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.188430 sec, 98.415557 gbps 8.414463 mpps
- fill metadata for 200
./xskgen -M -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 20960000000 bits, took 0.411559 sec, 50.928299 gbps 24.297853 mpps
- request metadata for 1400
./xskgen -m -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.188723 sec, 98.391299 gbps 8.412389 mpps
- request metadata for 200
./xskgen -m -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000064 packets 20960134144 bits, took 0.411240 sec, 50.968131 gbps 24.316856 mpps
5. Now, for the most interesting part, I'm adding mlx5 driver support.
The mpps for 200 bytes case goes down from 23 mpps to 19 mpps, but
_only_ when I enable the metadata. This looks like a side effect
of me pushing extra metadata pointer via mlx5e_xdpi_fifo_push.
Hence, this part is wrapped into 'if (xp_tx_metadata_enabled)'
to not affect the existing non-metadata use-cases. Since this is not
regressing existing workloads, I'm not spending any time trying to
optimize it more (and leaving it up to mlx owners to purse if
they see any good way to do it).
- same baseline
./xskgen -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.189434 sec, 98.332484 gbps 8.407360 mpps
./xskgen -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000128 packets 20960268288 bits, took 0.425254 sec, 49.288821 gbps 23.515659 mpps
- fill metadata for 1400
./xskgen -M -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.189528 sec, 98.324714 gbps 8.406696 mpps
- fill metadata for 200
./xskgen -M -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000128 packets 20960268288 bits, took 0.519085 sec, 40.379260 gbps 19.264914 mpps
- request metadata for 1400
./xskgen -m -s 1400 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000000 packets 116960000000 bits, took 1.189329 sec, 98.341165 gbps 8.408102 mpps
- request metadata for 200
./xskgen -m -s 200 -b eth3 10:70:fd:48:10:77 10:70:fd:48:10:87 fe80::1270:fdff:fe48:1077 fe80::1270:fdff:fe48:1087 1 1
sent 10000128 packets 20960268288 bits, took 0.519929 sec, 40.313713 gbps 19.233642 mpps
Acked-by: Magnus Karlsson <magnus.karlsson@intel.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
====================
Link: https://lore.kernel.org/r/20231127190319.1190813-1-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
When we get a packet on port 9091, we swap src/dst and send it out.
At this point we also request the timestamp and checksum offloads.
Checksum offload is verified by looking at the tcpdump on the other side.
The tool prints pseudo-header csum and the final one it expects.
The final checksum actually matches the incoming packets checksum
because we only flip the src/dst and don't change the payload.
Some other related changes:
- switched to zerocopy mode by default; new flag can be used to force
old behavior
- request fixed tx_metadata_len headroom
- some other small fixes (umem size, fill idx+i, etc)
mvbz3:~# ./xdp_hw_metadata eth3
...
xsk_ring_cons__peek: 1
0x19546f8: rx_desc[0]->addr=80100 addr=80100 comp_addr=80100
rx_hash: 0x80B7EA8B with RSS type:0x2A
rx_timestamp: 1697580171852147395 (sec:1697580171.8521)
HW RX-time: 1697580171852147395 (sec:1697580171.8521), delta to User RX-time sec:0.2797 (279673.082 usec)
XDP RX-time: 1697580172131699047 (sec:1697580172.1317), delta to User RX-time sec:0.0001 (121.430 usec)
0x19546f8: ping-pong with csum=3b8e (want d862) csum_start=54 csum_offset=6
0x19546f8: complete tx idx=0 addr=8
tx_timestamp: 1697580172056756493 (sec:1697580172.0568)
HW TX-complete-time: 1697580172056756493 (sec:1697580172.0568), delta to User TX-complete-time sec:0.0852 (85175.537 usec)
XDP RX-time: 1697580172131699047 (sec:1697580172.1317), delta to User TX-complete-time sec:0.0102 (10232.983 usec)
HW RX-time: 1697580171852147395 (sec:1697580171.8521), delta to HW TX-complete-time sec:0.2046 (204609.098 usec)
0x19546f8: complete rx idx=128 addr=80100
mvbz4:~# nc -Nu -q1 ${MVBZ3_LINK_LOCAL_IP}%eth3 9091
mvbz4:~# tcpdump -vvx -i eth3 udp
tcpdump: listening on eth3, link-type EN10MB (Ethernet), snapshot length 262144 bytes
12:26:09.301074 IP6 (flowlabel 0x35fa5, hlim 127, next-header UDP (17) payload length: 11) fe80::1270:fdff:fe48:1087.55807 > fe80::1270:fdff:fe48:1077.9091: [bad udp cksum 0x3b8e -> 0xde7e!] UDP, length 3
0x0000: 6003 5fa5 000b 117f fe80 0000 0000 0000
0x0010: 1270 fdff fe48 1087 fe80 0000 0000 0000
0x0020: 1270 fdff fe48 1077 d9ff 2383 000b 3b8e
0x0030: 7864 70
12:26:09.301976 IP6 (flowlabel 0x35fa5, hlim 127, next-header UDP (17) payload length: 11) fe80::1270:fdff:fe48:1077.9091 > fe80::1270:fdff:fe48:1087.55807: [udp sum ok] UDP, length 3
0x0000: 6003 5fa5 000b 117f fe80 0000 0000 0000
0x0010: 1270 fdff fe48 1077 fe80 0000 0000 0000
0x0020: 1270 fdff fe48 1087 2383 d9ff 000b de7e
0x0030: 7864 70
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20231127190319.1190813-14-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
This is the recommended way to run AF_XDP, so let's use it in the test.
Also, some unrelated changes to now blow up the log too much:
- change default mode to zerocopy and add -c to use copy mode
- small fixes for the flags/sizes/prints
- add print_tstamp_delta to print timestamp + reference
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20231127190319.1190813-13-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Accept only the flags that the kernel knows about to make
sure we can extend this field in the future. Note that only
in XDP_COPY mode we propagate the error signal back to the user
(via sendmsg). For zerocopy mode we silently skip the metadata
for the descriptors that have wrong flags (since we process
the descriptors deep in the driver).
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20231127190319.1190813-8-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
TX timestamp:
- requires passing clock, not sure I'm passing the correct one (from
cq->mdev), but the timestamp value looks convincing
TX checksum:
- looks like device does packet parsing (and doesn't accept custom
start/offset), so I'm ignoring user offsets
Cc: Saeed Mahameed <saeedm@nvidia.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20231127190319.1190813-5-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
This change actually defines the (initial) metadata layout
that should be used by AF_XDP userspace (xsk_tx_metadata).
The first field is flags which requests appropriate offloads,
followed by the offload-specific fields. The supported per-device
offloads are exported via netlink (new xsk-flags).
The offloads themselves are still implemented in a bit of a
framework-y fashion that's left from my initial kfunc attempt.
I'm introducing new xsk_tx_metadata_ops which drivers are
supposed to implement. The drivers are also supposed
to call xsk_tx_metadata_request/xsk_tx_metadata_complete in
the right places. Since xsk_tx_metadata_{request,_complete}
are static inline, we don't incur any extra overhead doing
indirect calls.
The benefit of this scheme is as follows:
- keeps all metadata layout parsing away from driver code
- makes it easy to grep and see which drivers implement what
- don't need any extra flags to maintain to keep track of what
offloads are implemented; if the callback is implemented - the offload
is supported (used by netlink reporting code)
Two offloads are defined right now:
1. XDP_TXMD_FLAGS_CHECKSUM: skb-style csum_start+csum_offset
2. XDP_TXMD_FLAGS_TIMESTAMP: writes TX timestamp back into metadata
area upon completion (tx_timestamp field)
XDP_TXMD_FLAGS_TIMESTAMP is also implemented for XDP_COPY mode: it writes
SW timestamp from the skb destructor (note I'm reusing hwtstamps to pass
metadata pointer).
The struct is forward-compatible and can be extended in the future
by appending more fields.
Reviewed-by: Song Yoong Siang <yoong.siang.song@intel.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231127190319.1190813-3-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
For zerocopy mode, tx_desc->addr can point to an arbitrary offset
and carry some TX metadata in the headroom. For copy mode, there
is no way currently to populate skb metadata.
Introduce new tx_metadata_len umem config option that indicates how many
bytes to treat as metadata. Metadata bytes come prior to tx_desc address
(same as in RX case).
The size of the metadata has mostly the same constraints as XDP:
- less than 256 bytes
- 8-byte aligned (compared to 4-byte alignment on xdp, due to 8-byte
timestamp in the completion)
- non-zero
This data is not interpreted in any way right now.
Reviewed-by: Song Yoong Siang <yoong.siang.song@intel.com>
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231127190319.1190813-2-sdf@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Akihiko Odaki says:
====================
selftests/bpf: Use pkg-config to determine ld flags
When linking statically, libraries may require other dependencies to be
included to ld flags. In particular, libelf may require libzstd. Use
pkg-config to determine such dependencies.
V4 -> V5: Introduced variables LIBELF_CFLAGS and LIBELF_LIBS.
(Daniel Borkmann)
Added patch "selftests/bpf: Choose pkg-config for the target".
V3 -> V4: Added "2> /dev/null".
V2 -> V3: Added missing "echo".
V1 -> V2: Implemented fallback, referring to HOSTPKG_CONFIG.
====================
Link: https://lore.kernel.org/r/20231125084253.85025-1-akihiko.odaki@daynix.com
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Jiri Olsa says:
====================
bpf: Add link_info support for uprobe multi link
hi,
this patchset adds support to get bpf_link_info details for
uprobe_multi links and adding support for bpftool link to
display them.
v4 changes:
- move flags field up in bpf_uprobe_multi_link [Andrii]
- include zero terminating byte in path_size [Andrii]
- return d_path error directly [Yonghong]
- use SEC(".probes") for semaphores [Yonghong]
- fix ref_ctr_offsets leak in test [Yonghong]
- other smaller fixes [Yonghong]
thanks,
jirka
---
====================
Link: https://lore.kernel.org/r/20231125193130.834322-1-jolsa@kernel.org
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Adding support to get uprobe_link details through bpf_link_info
interface.
Adding new struct uprobe_multi to struct bpf_link_info to carry
the uprobe_multi link details.
The uprobe_multi.count is passed from user space to denote size
of array fields (offsets/ref_ctr_offsets/cookies). The actual
array size is stored back to uprobe_multi.count (allowing user
to find out the actual array size) and array fields are populated
up to the user passed size.
All the non-array fields (path/count/flags/pid) are always set.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/20231125193130.834322-4-jolsa@kernel.org
Commit ef01f4e25c ("bpf: restore the ebpf program ID for BPF_AUDIT_UNLOAD
and PERF_BPF_EVENT_PROG_UNLOAD") stopped removing program's id from
idr when the offloaded/bound netdev goes away. I was supposed to
take a look and check in [0], but apparently I did not.
Martin points out it might be useful to keep it that way for
observability sake, but we at least need to mark those programs as
unusable.
Mark those programs as 'orphaned' and keep printing the list when
we encounter ENODEV.
0: unspec tag 0000000000000000
xlated 0B not jited memlock 4096B orphaned
[0]: https://lore.kernel.org/all/CAKH8qBtyR20ZWAc11z1-6pGb3Hd47AQUTbE_cfoktG59TqaJ7Q@mail.gmail.com/
v3:
* use two spaces for " orphaned" (Quentin)
Cc: netdev@vger.kernel.org
Fixes: ef01f4e25c ("bpf: restore the ebpf program ID for BPF_AUDIT_UNLOAD and PERF_BPF_EVENT_PROG_UNLOAD")
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Reviewed-by: Quentin Monnet <quentin@isovalent.com>
Link: https://lore.kernel.org/r/20231127182057.1081138-1-sdf@google.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
With latest upstream llvm18, the following test cases failed:
$ ./test_progs -j
#13/2 bpf_cookie/multi_kprobe_link_api:FAIL
#13/3 bpf_cookie/multi_kprobe_attach_api:FAIL
#13 bpf_cookie:FAIL
#77 fentry_fexit:FAIL
#78/1 fentry_test/fentry:FAIL
#78 fentry_test:FAIL
#82/1 fexit_test/fexit:FAIL
#82 fexit_test:FAIL
#112/1 kprobe_multi_test/skel_api:FAIL
#112/2 kprobe_multi_test/link_api_addrs:FAIL
[...]
#112 kprobe_multi_test:FAIL
#356/17 test_global_funcs/global_func17:FAIL
#356 test_global_funcs:FAIL
Further analysis shows llvm upstream patch [1] is responsible for the above
failures. For example, for function bpf_fentry_test7() in net/bpf/test_run.c,
without [1], the asm code is:
0000000000000400 <bpf_fentry_test7>:
400: f3 0f 1e fa endbr64
404: e8 00 00 00 00 callq 0x409 <bpf_fentry_test7+0x9>
409: 48 89 f8 movq %rdi, %rax
40c: c3 retq
40d: 0f 1f 00 nopl (%rax)
... and with [1], the asm code is:
0000000000005d20 <bpf_fentry_test7.specialized.1>:
5d20: e8 00 00 00 00 callq 0x5d25 <bpf_fentry_test7.specialized.1+0x5>
5d25: c3 retq
... and <bpf_fentry_test7.specialized.1> is called instead of <bpf_fentry_test7>
and this caused test failures for #13/#77 etc. except #356.
For test case #356/17, with [1] (progs/test_global_func17.c)), the main prog
looks like:
0000000000000000 <global_func17>:
0: b4 00 00 00 2a 00 00 00 w0 = 0x2a
1: 95 00 00 00 00 00 00 00 exit
... which passed verification while the test itself expects a verification
failure.
Let us add 'barrier_var' style asm code in both places to prevent function
specialization which caused selftests failure.
[1] https://github.com/llvm/llvm-project/pull/72903
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20231127050342.1945270-1-yonghong.song@linux.dev
Add a few test that validate BPF verifier's lazy approach to validating
global subprogs.
We check that global subprogs that are called transitively through
another global subprog is validated.
We also check that invalid global subprog is not validated, if it's not
called from the main program.
And we also check that main program is always validated first, before
any of the subprogs.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20231124035937.403208-4-andrii@kernel.org
Slightly change BPF verifier logic around eagerness and order of global
subprog validation. Instead of going over every global subprog eagerly
and validating it before main (entry) BPF program is verified, turn it
around. Validate main program first, mark subprogs that were called from
main program for later verification, but otherwise assume it is valid.
Afterwards, go over marked global subprogs and validate those,
potentially marking some more global functions as being called. Continue
this process until all (transitively) callable global subprogs are
validated. It's a BFS traversal at its heart and will always converge.
This is an important change because it allows to feature-gate some
subprograms that might not be verifiable on some older kernel, depending
on supported set of features.
E.g., at some point, global functions were allowed to accept a pointer
to memory, which size is identified by user-provided type.
Unfortunately, older kernels don't support this feature. With BPF CO-RE
approach, the natural way would be to still compile BPF object file once
and guard calls to this global subprog with some CO-RE check or using
.rodata variables. That's what people do to guard usage of new helpers
or kfuncs, and any other new BPF-side feature that might be missing on
old kernels.
That's currently impossible to do with global subprogs, unfortunately,
because they are eagerly and unconditionally validated. This patch set
aims to change this, so that in the future when global funcs gain new
features, those can be guarded using BPF CO-RE techniques in the same
fashion as any other new kernel feature.
Two selftests had to be adjusted in sync with these changes.
test_global_func12 relied on eager global subprog validation failing
before main program failure is detected (unknown return value). Fix by
making sure that main program is always valid.
verifier_subprog_precision's parent_stack_slot_precise subtest relied on
verifier checkpointing heuristic to do a checkpoint at instruction #5,
but that's no longer true because we don't have enough jumps validated
before reaching insn #5 due to global subprogs being validated later.
Other than that, no changes, as one would expect.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20231124035937.403208-3-andrii@kernel.org
Pull networking fixes from Jakub Kicinski:
"Including fixes from bpf.
Current release - regressions:
- Revert "net: r8169: Disable multicast filter for RTL8168H and
RTL8107E"
- kselftest: rtnetlink: fix ip route command typo
Current release - new code bugs:
- s390/ism: make sure ism driver implies smc protocol in kconfig
- two build fixes for tools/net
Previous releases - regressions:
- rxrpc: couple of ACK/PING/RTT handling fixes
Previous releases - always broken:
- bpf: verify bpf_loop() callbacks as if they are called unknown
number of times
- improve stability of auto-bonding with Hyper-V
- account BPF-neigh-redirected traffic in interface statistics
Misc:
- net: fill in some more MODULE_DESCRIPTION()s"
* tag 'net-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (58 commits)
tools: ynl: fix duplicate op name in devlink
tools: ynl: fix header path for nfsd
net: ipa: fix one GSI register field width
tls: fix NULL deref on tls_sw_splice_eof() with empty record
net: axienet: Fix check for partial TX checksum
vsock/test: fix SEQPACKET message bounds test
i40e: Fix adding unsupported cloud filters
ice: restore timestamp configuration after device reset
ice: unify logic for programming PFINT_TSYN_MSK
ice: remove ptp_tx ring parameter flag
amd-xgbe: propagate the correct speed and duplex status
amd-xgbe: handle the corner-case during tx completion
amd-xgbe: handle corner-case during sfp hotplug
net: veth: fix ethtool stats reporting
octeontx2-pf: Fix ntuple rule creation to direct packet to VF with higher Rx queue than its PF
net: usb: qmi_wwan: claim interface 4 for ZTE MF290
Revert "net: r8169: Disable multicast filter for RTL8168H and RTL8107E"
net/smc: avoid data corruption caused by decline
nfc: virtual_ncidev: Add variable to check if ndev is running
dpll: Fix potential msg memleak when genlmsg_put_reply failed
...
We don't support CRUD-inspired message types in YNL too well.
One aspect that currently trips us up is the fact that single
message ID can be used in multiple commands (as the response).
This leads to duplicate entries in the id-to-string tables:
devlink-user.c:19:34: warning: initialized field overwritten [-Woverride-init]
19 | [DEVLINK_CMD_PORT_NEW] = "port-new",
| ^~~~~~~~~~
devlink-user.c:19:34: note: (near initialization for ‘devlink_op_strmap[7]’)
Fixes tag points at where the code was generated, the "real" problem
is that the code generator does not support CRUD.
Fixes: f2f9dd164d ("netlink: specs: devlink: add the remaining command to generate complete split_ops")
Link: https://lore.kernel.org/r/20231123030558.1611831-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
syzkaller discovered that if tls_sw_splice_eof() is executed as part of
sendfile() when the plaintext/ciphertext sk_msg are empty, the send path
gets confused because the empty ciphertext buffer does not have enough
space for the encryption overhead. This causes tls_push_record() to go on
the `split = true` path (which is only supposed to be used when interacting
with an attached BPF program), and then get further confused and hit the
tls_merge_open_record() path, which then assumes that there must be at
least one populated buffer element, leading to a NULL deref.
It is possible to have empty plaintext/ciphertext buffers if we previously
bailed from tls_sw_sendmsg_locked() via the tls_trim_both_msgs() path.
tls_sw_push_pending_record() already handles this case correctly; let's do
the same check in tls_sw_splice_eof().
Fixes: df720d288d ("tls/sw: Use splice_eof() to flush")
Cc: stable@vger.kernel.org
Reported-by: syzbot+40d43509a099ea756317@syzkaller.appspotmail.com
Signed-off-by: Jann Horn <jannh@google.com>
Link: https://lore.kernel.org/r/20231122214447.675768-1-jannh@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tune message length calculation to make this test work on machines
where 'getpagesize()' returns >32KB. Now maximum message length is not
hardcoded (on machines above it was smaller than 'getpagesize()' return
value, thus we get negative value and test fails), but calculated at
runtime and always bigger than 'getpagesize()' result. Reproduced on
aarch64 with 64KB page size.
Fixes: 5c338112e4 ("test/vsock: rework message bounds test")
Signed-off-by: Arseniy Krasnov <avkrasnov@salutedevices.com>
Reported-by: Bogdan Marcynkov <bmarcynk@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Link: https://lore.kernel.org/r/20231121211642.163474-1-avkrasnov@salutedevices.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If a VF tries to add unsupported cloud filter through virtchnl
then i40e_add_del_cloud_filter(_big_buf) returns -ENOTSUPP but
this error code is stored in 'ret' instead of 'aq_ret' that
is used as error code sent back to VF. In this scenario where
one of the mentioned functions fails the value of 'aq_ret'
is zero so the VF will incorrectly receive a 'success'.
Use 'aq_ret' to store return value and remove 'ret' local
variable. Additionally fix the issue when filter allocation
fails, in this case no notification is sent back to the VF.
Fixes: e284fc2804 ("i40e: Add and delete cloud filter")
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: Ivan Vecera <ivecera@redhat.com>
Tested-by: Rafal Romanowski <rafal.romanowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Link: https://lore.kernel.org/r/20231121211338.3348677-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Tony Nguyen says:
====================
ice: restore timestamp config after reset
Jake Keller says:
We recently discovered during internal validation that the ice driver has
not been properly restoring Tx timestamp configuration after a device reset,
which resulted in application failures after a device reset.
After some digging, it turned out this problem is two-fold. Since the
introduction of the PTP support the driver has been clobbering the storage
of the current timestamp configuration during reset. Thus after a reset, the
driver will no longer perform Tx or Rx timestamps, and will report
timestamp configuration as disabled if SIOCGHWTSTAMP ioctl is issued.
In addition, the recently merged auxiliary bus support code missed that
PFINT_TSYN_MSK must be reprogrammed on the clock owner for E822 devices.
Failure to restore this register configuration results in the driver no
longer responding to interrupts from other ports. Depending on the traffic
pattern, this can either result in increased latency responding to
timestamps on the non-owner ports, or it can result in the driver never
reporting any timestamps. The configuration of PFINT_TSYN_MSK was only done
during initialization. Due to this, the Tx timestamp issue persists even if
userspace reconfigures timestamping.
This series fixes both issues, as well as removes a redundant Tx ring field
since we can rely on the skb flag as the primary detector for a Tx timestamp
request.
Note that I don't think this series will directly apply to older stable
releases (even v6.6) as we recently refactored a lot of the PTP code to
support auxiliary bus. Patch 2/3 only matters for the post-auxiliary bus
implementation. The principle of patch 1/3 and 3/3 could apply as far back
as the initial PTP support, but I don't think it will apply cleanly as-is.
====================
Link: https://lore.kernel.org/r/20231121211259.3348630-1-anthony.l.nguyen@intel.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The driver calls ice_ptp_cfg_timestamp() during ice_ptp_prepare_for_reset()
to disable timestamping while the device is resetting. This operation
destroys the user requested configuration. While the driver does call
ice_ptp_cfg_timestamp in ice_rebuild() to restore some hardware settings
after a reset, it unconditionally passes true or false, resulting in
failure to restore previous user space configuration.
This results in a device reset forcibly disabling timestamp configuration
regardless of current user settings.
This was not detected previously due to a quirk of the LinuxPTP ptp4l
application. If ptp4l detects a missing timestamp, it enters a fault state
and performs recovery logic which includes executing SIOCSHWTSTAMP again,
restoring the now accidentally cleared configuration.
Not every application does this, and for these applications, timestamps
will mysteriously stop after a PF reset, without being restored until an
application restart.
Fix this by replacing ice_ptp_cfg_timestamp() with two new functions:
1) ice_ptp_disable_timestamp_mode() which unconditionally disables the
timestamping logic in ice_ptp_prepare_for_reset() and ice_ptp_release()
2) ice_ptp_restore_timestamp_mode() which calls
ice_ptp_restore_tx_interrupt() to restore Tx timestamping configuration,
calls ice_set_rx_tstamp() to restore Rx timestamping configuration, and
issues an immediate TSYN_TX interrupt to ensure that timestamps which
may have occurred during the device reset get processed.
Modify the ice_ptp_set_timestamp_mode to directly save the user
configuration and then call ice_ptp_restore_timestamp_mode. This way, reset
no longer destroys the saved user configuration.
This obsoletes the ice_set_tx_tstamp() function which can now be safely
removed.
With this change, all devices should now restore Tx and Rx timestamping
functionality correctly after a PF reset without application intervention.
Fixes: 77a781155a ("ice: enable receive hardware timestamping")
Fixes: ea9b847cda ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Commit d938a8cca8 ("ice: Auxbus devices & driver for E822 TS") modified
how Tx timestamps are handled for E822 devices. On these devices, only the
clock owner handles reading the Tx timestamp data from firmware. To do
this, the PFINT_TSYN_MSK register is modified from the default value to one
which enables reacting to a Tx timestamp on all PHY ports.
The driver currently programs PFINT_TSYN_MSK in different places depending
on whether the port is the clock owner or not. For the clock owner, the
PFINT_TSYN_MSK value is programmed during ice_ptp_init_owner just before
calling ice_ptp_tx_ena_intr to program the PHY ports.
For the non-clock owner ports, the PFINT_TSYN_MSK is programmed during
ice_ptp_init_port.
If a large enough device reset occurs, the PFINT_TSYN_MSK register will be
reset to the default value in which only the PHY associated directly with
the PF will cause the Tx timestamp interrupt to trigger.
The driver lacks logic to reprogram the PFINT_TSYN_MSK register after a
device reset. For the E822 device, this results in the PF no longer
responding to interrupts for other ports. This results in failure to
deliver Tx timestamps to user space applications.
Rename ice_ptp_configure_tx_tstamp to ice_ptp_cfg_tx_interrupt, and unify
the logic for programming PFINT_TSYN_MSK and PFINT_OICR_ENA into one place.
This function will program both registers according to the combination of
user configuration and device requirements.
This ensures that PFINT_TSYN_MSK is always restored when we configure the
Tx timestamp interrupt.
Fixes: d938a8cca8 ("ice: Auxbus devices & driver for E822 TS")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Before performing a Tx timestamp in ice_stamp(), the driver checks a ptp_tx
ring variable to see if timestamping is enabled on that ring. This value is
set for all rings whenever userspace configures Tx timestamping.
Ostensibly this was done to avoid wasting cycles checking other fields when
timestamping has not been enabled. However, for Tx timestamps we already
get an individual per-SKB flag indicating whether userspace wants to
request a timestamp on that packet. We do not gain much by also having
a separate flag to check for whether timestamping was enabled.
In fact, the driver currently fails to restore the field after a PF reset.
Because of this, if a PF reset occurs, timestamps will be disabled.
Since this flag doesn't add value in the hotpath, remove it and always
provide a timestamp if the SKB flag has been set.
A following change will fix the reset path to properly restore user
timestamping configuration completely.
This went unnoticed for some time because one of the most common
applications using Tx timestamps, ptp4l, will reconfigure the socket as
part of its fault recovery logic.
Fixes: ea9b847cda ("ice: enable transmit timestamps for E810 devices")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Pucha Himasekhar Reddy <himasekharx.reddy.pucha@intel.com> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
xgbe_get_link_ksettings() does not propagate correct speed and duplex
information to ethtool during cable unplug. Due to which ethtool reports
incorrect values for speed and duplex.
Address this by propagating correct information.
Fixes: 7c12aa0877 ("amd-xgbe: Move the PHY support into amd-xgbe")
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The existing implementation uses software logic to accumulate tx
completions until the specified time (1ms) is met and then poll them.
However, there exists a tiny gap which leads to a race between
resetting and checking the tx_activate flag. Due to this the tx
completions are not reported to upper layer and tx queue timeout
kicks-in restarting the device.
To address this, introduce a tx cleanup mechanism as part of the
periodic maintenance process.
Fixes: c5aa9e3b81 ("amd-xgbe: Initial AMD 10GbE platform driver")
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Force the mode change for SFI in Fixed PHY configurations. Fixed PHY
configurations needs PLL to be enabled while doing mode set. When the
SFP module isn't connected during boot, driver assumes AN is ON and
attempts auto-negotiation. However, if the connected SFP comes up in
Fixed PHY configuration the link will not come up as PLL isn't enabled
while the initial mode set command is issued. So, force the mode change
for SFI in Fixed PHY configuration to fix link issues.
Fixes: e57f7a3fea ("amd-xgbe: Prepare for working with more than one type of phy")
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
