openwrt/staging/blogic.git
6 years agobpf: Improve the info.func_info and info.func_info_rec_size behavior
Martin KaFai Lau [Thu, 6 Dec 2018 01:35:43 +0000 (17:35 -0800)]
bpf: Improve the info.func_info and info.func_info_rec_size behavior

1) When bpf_dump_raw_ok() == false and the kernel can provide >=1
   func_info to the userspace, the current behavior is setting
   the info.func_info_cnt to 0 instead of setting info.func_info
   to 0.

   It is different from the behavior in jited_func_lens/nr_jited_func_lens,
   jited_ksyms/nr_jited_ksyms...etc.

   This patch fixes it. (i.e. set func_info to 0 instead of
   func_info_cnt to 0 when bpf_dump_raw_ok() == false).

2) When the userspace passed in info.func_info_cnt == 0, the kernel
   will set the expected func_info size back to the
   info.func_info_rec_size.  It is a way for the userspace to learn
   the kernel expected func_info_rec_size introduced in
   commit 838e96904ff3 ("bpf: Introduce bpf_func_info").

   An exception is the kernel expected size is not set when
   func_info is not available for a bpf_prog.  This makes the
   returned info.func_info_rec_size has different values
   depending on the returned value of info.func_info_cnt.

   This patch sets the kernel expected size to info.func_info_rec_size
   independent of the info.func_info_cnt.

3) The current logic only rejects invalid func_info_rec_size if
   func_info_cnt is non zero.  This patch also rejects invalid
   nonzero info.func_info_rec_size and not equal to the kernel
   expected size.

4) Set info.btf_id as long as prog->aux->btf != NULL.  That will
   setup the later copy_to_user() codes look the same as others
   which then easier to understand and maintain.

   prog->aux->btf is not NULL only if prog->aux->func_info_cnt > 0.

   Breaking up info.btf_id from prog->aux->func_info_cnt is needed
   for the later line info patch anyway.

   A similar change is made to bpf_get_prog_name().

Fixes: 838e96904ff3 ("bpf: Introduce bpf_func_info")
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: add a command to dump the trace pipe
Quentin Monnet [Wed, 5 Dec 2018 10:28:24 +0000 (10:28 +0000)]
tools: bpftool: add a command to dump the trace pipe

BPF programs can use the bpf_trace_printk() helper to print debug
information into the trace pipe. Add a subcommand
"bpftool prog tracelog" to simply dump this pipe to the console.

This is for a good part copied from iproute2, where the feature is
available with "tc exec bpf dbg". Changes include dumping pipe content
to stdout instead of stderr and adding JSON support (content is dumped
as an array of strings, one per line read from the pipe). This version
is dual-licensed, with Daniel's permission.

Cc: Daniel Borkmann <daniel@iogearbox.net>
Suggested-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoMerge branch 'bpf-jit-overridable-alloc'
Daniel Borkmann [Wed, 5 Dec 2018 15:36:29 +0000 (16:36 +0100)]
Merge branch 'bpf-jit-overridable-alloc'

Ard Biesheuvel says:

====================
On arm64, modules are allocated from a 128 MB window which is close to
the core kernel, so that relative direct branches are guaranteed to be
in range (except in some KASLR configurations). Also, module_alloc()
is in charge of allocating KASAN shadow memory when running with KASAN
enabled.

This means that the way BPF reuses module_alloc()/module_memfree() is
undesirable on arm64 (and potentially other architectures as well),
and so this series refactors BPF's use of those functions to permit
architectures to change this behavior.

Patch #1 breaks out the module_alloc() and module_memfree() calls into
__weak functions so they can be overridden.

Patch #2 implements the new alloc/free overrides for arm64

Changes since v3:
- drop 'const' modifier for free() hook void* argument
- move the dedicated BPF region to before the module region, putting it
  within 4GB of the module and kernel regions on non-KASLR kernels

Changes since v2:
- properly build time and runtime tested this time (log after the diffstat)
- create a dedicated 128 MB region at the top of the vmalloc space for BPF
  programs, ensuring that the programs will be in branching range of each
  other (which we currently rely upon) but at an arbitrary distance from
  the kernel and modules (which we don't care about)

Changes since v1:
- Drop misguided attempt to 'fix' and refactor the free path. Instead,
  just add another __weak wrapper for the invocation of module_memfree()
====================

Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Rick Edgecombe <rick.p.edgecombe@intel.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Jann Horn <jannh@google.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Jessica Yu <jeyu@kernel.org>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoarm64/bpf: don't allocate BPF JIT programs in module memory
Ard Biesheuvel [Fri, 23 Nov 2018 22:18:04 +0000 (23:18 +0100)]
arm64/bpf: don't allocate BPF JIT programs in module memory

The arm64 module region is a 128 MB region that is kept close to
the core kernel, in order to ensure that relative branches are
always in range. So using the same region for programs that do
not have this restriction is wasteful, and preferably avoided.

Now that the core BPF JIT code permits the alloc/free routines to
be overridden, implement them by vmalloc()/vfree() calls from a
dedicated 128 MB region set aside for BPF programs. This ensures
that BPF programs are still in branching range of each other, which
is something the JIT currently depends upon (and is not guaranteed
when using module_alloc() on KASLR kernels like we do currently).
It also ensures that placement of BPF programs does not correlate
with the placement of the core kernel or modules, making it less
likely that leaking the former will reveal the latter.

This also solves an issue under KASAN, where shadow memory is
needlessly allocated for all BPF programs (which don't require KASAN
shadow pages since they are not KASAN instrumented)

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: add __weak hook for allocating executable memory
Ard Biesheuvel [Fri, 23 Nov 2018 22:18:03 +0000 (23:18 +0100)]
bpf: add __weak hook for allocating executable memory

By default, BPF uses module_alloc() to allocate executable memory,
but this is not necessary on all arches and potentially undesirable
on some of them.

So break out the module_alloc() and module_memfree() calls into __weak
functions to allow them to be overridden in arch code.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoMerge branch 'prog_test_run-improvement'
Alexei Starovoitov [Tue, 4 Dec 2018 16:18:14 +0000 (08:18 -0800)]
Merge branch 'prog_test_run-improvement'

Lorenz Bauer says:

====================
Right now, there is no safe way to use BPF_PROG_TEST_RUN with data_out.
This is because bpf_test_finish copies the output buffer to user space
without checking its size. This can lead to the kernel overwriting
data in user space after the buffer if xdp_adjust_head and friends are
in play.

Thanks to everyone for their advice and patience with this patch set!

Changes in v5:
* Fix up libbpf.map

Changes in v4:
* Document bpf_prog_test_run and bpf_prog_test_run_xattr
* Use struct bpf_prog_test_run_attr for return values

Changes in v3:
* Introduce bpf_prog_test_run_xattr instead of modifying the existing
  function

Changes in v2:
* Make the syscall return ENOSPC if data_size_out is too small
* Make bpf_prog_test_run return EINVAL if size_out is missing
* Document the new behaviour of data_size_out
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoselftests: add a test for bpf_prog_test_run_xattr
Lorenz Bauer [Mon, 3 Dec 2018 11:31:26 +0000 (11:31 +0000)]
selftests: add a test for bpf_prog_test_run_xattr

Make sure that bpf_prog_test_run_xattr returns the correct length
and that the kernel respects the output size hint. Also check
that errno indicates ENOSPC if there is a short output buffer given.

Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agolibbpf: add bpf_prog_test_run_xattr
Lorenz Bauer [Mon, 3 Dec 2018 11:31:25 +0000 (11:31 +0000)]
libbpf: add bpf_prog_test_run_xattr

Add a new function, which encourages safe usage of the test interface.
bpf_prog_test_run continues to work as before, but should be considered
unsafe.

Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: sync uapi/linux/bpf.h
Lorenz Bauer [Mon, 3 Dec 2018 11:31:24 +0000 (11:31 +0000)]
tools: sync uapi/linux/bpf.h

Pull changes from "bpf: respect size hint to BPF_PROG_TEST_RUN if present".

Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: respect size hint to BPF_PROG_TEST_RUN if present
Lorenz Bauer [Mon, 3 Dec 2018 11:31:23 +0000 (11:31 +0000)]
bpf: respect size hint to BPF_PROG_TEST_RUN if present

Use data_size_out as a size hint when copying test output to user space.
ENOSPC is returned if the output buffer is too small.
Callers which so far did not set data_size_out are not affected.

Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agosamples: bpf: fix: seg fault with NULL pointer arg
Daniel T. Lee [Mon, 3 Dec 2018 10:39:30 +0000 (19:39 +0900)]
samples: bpf: fix: seg fault with NULL pointer arg

When NULL pointer accidentally passed to write_kprobe_events,
due to strlen(NULL), segmentation fault happens.
Changed code returns -1 to deal with this situation.

Bug issued with Smatch, static analysis.

Signed-off-by: Daniel T. Lee <danieltimlee@gmail.com>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: fix documentation for eBPF helpers
Quentin Monnet [Mon, 3 Dec 2018 12:13:35 +0000 (12:13 +0000)]
bpf: fix documentation for eBPF helpers

The missing indentation on the "Return" sections for bpf_map_pop_elem()
and bpf_map_peek_elem() helpers break RST and man pages generation. This
patch fixes them, and moves the description of those two helpers towards
the end of the list (even though they are somehow related to the three
first helpers for maps, the man page explicitly states that the helpers
are sorted in chronological order).

While at it, bring other minor formatting edits for eBPF helpers
documentation: mostly blank lines removal, RST formatting, or other
small nits for consistency.

Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: allow BPF read access to qdisc pkt_len
Petar Penkov [Mon, 3 Dec 2018 01:18:19 +0000 (20:18 -0500)]
bpf: allow BPF read access to qdisc pkt_len

The pkt_len field in qdisc_skb_cb stores the skb length as it will
appear on the wire after segmentation. For byte accounting, this value
is more accurate than skb->len. It is computed on entry to the TC
layer, so only valid there.

Allow read access to this field from BPF tc classifier and action
programs. The implementation is analogous to tc_classid, aside from
restricting to read access.

To distinguish it from skb->len and self-describe export as wire_len.

Changes v1->v2
  - Rename pkt_len to wire_len

Signed-off-by: Petar Penkov <ppenkov@google.com>
Signed-off-by: Vlad Dumitrescu <vladum@google.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agolibbpf: Fix license in README.rst
Andrey Ignatov [Sun, 2 Dec 2018 21:02:15 +0000 (13:02 -0800)]
libbpf: Fix license in README.rst

The whole libbpf is licensed as (LGPL-2.1 OR BSD-2-Clause). I missed it
while adding README.rst. Fix it and use same license as all other files
in libbpf do. Since I'm the only author of README.rst so far, no others'
permissions should be needed.

Fixes: 76d1b894c515 ("libbpf: Document API and ABI conventions")
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: Fix memleak in aux->func_info and aux->btf
Martin KaFai Lau [Sun, 2 Dec 2018 01:08:44 +0000 (17:08 -0800)]
bpf: Fix memleak in aux->func_info and aux->btf

The aux->func_info and aux->btf are leaked in the error out cases
during bpf_prog_load().  This patch fixes it.

Fixes: ba64e7d85252 ("bpf: btf: support proper non-jit func info")
Cc: Yonghong Song <yhs@fb.com>
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoMerge branch 'xdp1-improvements'
Alexei Starovoitov [Sat, 1 Dec 2018 06:06:41 +0000 (22:06 -0800)]
Merge branch 'xdp1-improvements'

Matteo Croce says:

====================
Small improvements to improve the readability and easiness
to use of the xdp1 sample.
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agosamples: bpf: get ifindex from ifname
Matteo Croce [Sat, 1 Dec 2018 00:23:06 +0000 (01:23 +0100)]
samples: bpf: get ifindex from ifname

Find the ifindex with if_nametoindex() instead of requiring the
numeric ifindex.

Signed-off-by: Matteo Croce <mcroce@redhat.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agosamples: bpf: improve xdp1 example
Matteo Croce [Sat, 1 Dec 2018 00:23:05 +0000 (01:23 +0100)]
samples: bpf: improve xdp1 example

Store only the total packet count for every protocol, instead of the
whole per-cpu array.
Use bpf_map_get_next_key() to iterate the map, instead of looking up
all the protocols.

Signed-off-by: Matteo Croce <mcroce@redhat.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoMerge branch 'improve-test-coverage-sparc'
Alexei Starovoitov [Sat, 1 Dec 2018 05:38:49 +0000 (21:38 -0800)]
Merge branch 'improve-test-coverage-sparc'

David Miller says:

====================
On sparc64 a ton of test cases in test_verifier.c fail because
the memory accesses in the test case are unaligned (or cannot
be proven to be aligned by the verifier).

Perhaps we can eventually try to (carefully) modify each test case
which has this problem to not use unaligned accesses but:

1) That is delicate work.

2) The changes might not fully respect the original
   intention of the testcase.

3) In some cases, such a transformation might not even
   be feasible at all.

So add an "any alignment" flag to tell the verifier to forcefully
disable it's alignment checks completely.

test_verifier.c is then annotated to use this flag when necessary.

The presence of the flag in each test case is good documentation to
anyone who wants to actually tackle the job of eliminating the
unaligned memory accesses in the test cases.

I've also seen several weird things in test cases, like trying to
access __skb->mark in a packet buffer.

This gets rid of 104 test_verifier.c failures on sparc64.

Changes since v1:

1) Explain the new BPF_PROG_LOAD flag in easier to understand terms.
   Suggested by Alexei.

2) Make bpf_verify_program() just take a __u32 prog_flags instead of
   just accumulating boolean arguments over and over.  Also suggested
   by Alexei.

Changes since RFC:

1) Only the admin can allow the relaxation of alignment restrictions
   on inefficient unaligned access architectures.

2) Use F_NEEDS_EFFICIENT_UNALIGNED_ACCESS instead of making a new
   flag.

3) Annotate in the output, when we have a test case that the verifier
   accepted but we did not try to execute because we are on an
   inefficient unaligned access platform.  Maybe with some arch
   machinery we can avoid this in the future.

Signed-off-by: David S. Miller <davem@davemloft.net>
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: Apply F_NEEDS_EFFICIENT_UNALIGNED_ACCESS to more ACCEPT test cases.
David Miller [Sat, 1 Dec 2018 05:08:50 +0000 (21:08 -0800)]
bpf: Apply F_NEEDS_EFFICIENT_UNALIGNED_ACCESS to more ACCEPT test cases.

If a testcase has alignment problems but is expected to be ACCEPT,
verify it using F_NEEDS_EFFICIENT_UNALIGNED_ACCESS too.

Maybe in the future if we add some architecture specific code to elide
the unaligned memory access warnings during the test, we can execute
these as well.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: Make more use of 'any' alignment in test_verifier.c
David Miller [Sat, 1 Dec 2018 05:08:39 +0000 (21:08 -0800)]
bpf: Make more use of 'any' alignment in test_verifier.c

Use F_NEEDS_EFFICIENT_UNALIGNED_ACCESS in more tests where the
expected result is REJECT.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: Adjust F_NEEDS_EFFICIENT_UNALIGNED_ACCESS handling in test_verifier.c
David Miller [Sat, 1 Dec 2018 05:08:26 +0000 (21:08 -0800)]
bpf: Adjust F_NEEDS_EFFICIENT_UNALIGNED_ACCESS handling in test_verifier.c

Make it set the flag argument to bpf_verify_program() which will relax
the alignment restrictions.

Now all such test cases will go properly through the verifier even on
inefficient unaligned access architectures.

On inefficient unaligned access architectures do not try to run such
programs, instead mark the test case as passing but annotate the
result similarly to how it is done now in the presence of this flag.

So, we get complete full coverage for all REJECT test cases, and at
least verifier level coverage for ACCEPT test cases.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: Add BPF_F_ANY_ALIGNMENT.
David Miller [Sat, 1 Dec 2018 05:08:14 +0000 (21:08 -0800)]
bpf: Add BPF_F_ANY_ALIGNMENT.

Often we want to write tests cases that check things like bad context
offset accesses.  And one way to do this is to use an odd offset on,
for example, a 32-bit load.

This unfortunately triggers the alignment checks first on platforms
that do not set CONFIG_EFFICIENT_UNALIGNED_ACCESS.  So the test
case see the alignment failure rather than what it was testing for.

It is often not completely possible to respect the original intention
of the test, or even test the same exact thing, while solving the
alignment issue.

Another option could have been to check the alignment after the
context and other validations are performed by the verifier, but
that is a non-trivial change to the verifier.

Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agobpf: Fix verifier log string check for bad alignment.
David Miller [Thu, 29 Nov 2018 06:33:53 +0000 (22:33 -0800)]
bpf: Fix verifier log string check for bad alignment.

The message got changed a lot time ago.

This was responsible for 36 test case failures on sparc64.

Fixes: f1174f77b50c ("bpf/verifier: rework value tracking")
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoMerge branch 'bpftool-fixes'
Alexei Starovoitov [Fri, 30 Nov 2018 22:06:07 +0000 (14:06 -0800)]
Merge branch 'bpftool-fixes'

Quentin Monnet says:

====================
Hi,
Several items for bpftool are included in this set: the first three patches
are fixes for bpftool itself and bash completion, while the last two
slightly improve the information obtained when dumping programs or maps, on
Daniel's suggestion. Please refer to individual commit logs for more
details.
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: add owner_prog_type and owner_jited to bpftool output
Quentin Monnet [Fri, 30 Nov 2018 16:25:48 +0000 (16:25 +0000)]
tools: bpftool: add owner_prog_type and owner_jited to bpftool output

For prog array maps, the type of the owner program, and the JIT-ed state
of that program, are available from the file descriptor information
under /proc. Add them to "bpftool map show" output. Example output:

    # bpftool map show
    158225: prog_array  name jmp_table  flags 0x0
        key 4B  value 4B  max_entries 8  memlock 4096B
        owner_prog_type flow_dissector  owner jited
    # bpftool --json --pretty map show
    [{
            "id": 1337,
            "type": "prog_array",
            "name": "jmp_table",
            "flags": 0,
            "bytes_key": 4,
            "bytes_value": 4,
            "max_entries": 8,
            "bytes_memlock": 4096,
            "owner_prog_type": "flow_dissector",
            "owner_jited": true
        }
    ]

As we move the table used for associating names to program types,
complete it with the missing types (lwt_seg6local and sk_reuseport).
Also add missing types to the help message for "bpftool prog"
(sk_reuseport and flow_dissector).

Suggested-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: mark offloaded programs more explicitly in plain output
Quentin Monnet [Fri, 30 Nov 2018 16:25:47 +0000 (16:25 +0000)]
tools: bpftool: mark offloaded programs more explicitly in plain output

In bpftool (plain) output for "bpftool prog show" or "bpftool map show",
an offloaded BPF object is simply denoted with "dev ifname", which is
not really explicit. Change it with something that clearly shows the
program is offloaded.

While at it also add an additional space, as done between other
information fields.

Example output, before:

    # bpftool prog show
    1337: xdp  tag a04f5eef06a7f555 dev foo
            loaded_at 2018-10-19T16:40:36+0100  uid 0
            xlated 16B  not jited  memlock 4096B

After:

    # bpftool prog show
    1337: xdp  tag a04f5eef06a7f555  offloaded_to foo
            loaded_at 2018-10-19T16:40:36+0100  uid 0
            xlated 16B  not jited  memlock 4096B

Suggested-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: fix bash completion for new map types (queue and stack)
Quentin Monnet [Fri, 30 Nov 2018 16:25:46 +0000 (16:25 +0000)]
tools: bpftool: fix bash completion for new map types (queue and stack)

Commit 197c2dac74e4 ("bpf: Add BPF_MAP_TYPE_QUEUE and BPF_MAP_TYPE_STACK
to bpftool-map") added support for queue and stack eBPF map types in
bpftool map handling. Let's update the bash completion accordingly.

Fixes: 197c2dac74e4 ("bpf: Add BPF_MAP_TYPE_QUEUE and BPF_MAP_TYPE_STACK to bpftool-map")
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: fix bash completion for bpftool prog (attach|detach)
Quentin Monnet [Fri, 30 Nov 2018 16:25:45 +0000 (16:25 +0000)]
tools: bpftool: fix bash completion for bpftool prog (attach|detach)

Fix bash completion for "bpftool prog (attach|detach) PROG TYPE MAP" so
that the list of indices proposed for MAP are map indices, and not PROG
indices. Also use variables for map and prog reference types ("id",
"pinned", and "tag" for programs).

Fixes: b7d3826c2ed6 ("bpf: bpftool, add support for attaching programs to maps")
Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agotools: bpftool: use "/proc/self/" i.o. crafting links with getpid()
Quentin Monnet [Fri, 30 Nov 2018 16:25:44 +0000 (16:25 +0000)]
tools: bpftool: use "/proc/self/" i.o. crafting links with getpid()

The getpid() function is called in a couple of places in bpftool to
craft links of the shape "/proc/<pid>/...". Instead, it is possible to
use the "/proc/self/" shortcut, which makes things a bit easier, in
particular in jit_disasm.c.

Do the replacement, and remove the includes of <sys/types.h> from the
relevant files, now we do not use getpid() anymore.

Signed-off-by: Quentin Monnet <quentin.monnet@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoarm64/bpf: use movn/movk/movk sequence to generate kernel addresses
Ard Biesheuvel [Fri, 23 Nov 2018 17:29:02 +0000 (18:29 +0100)]
arm64/bpf: use movn/movk/movk sequence to generate kernel addresses

On arm64, all executable code is guaranteed to reside in the vmalloc
space (or the module space), and so jump targets will only use 48
bits at most, and the remaining bits are guaranteed to be 0x1.

This means we can generate an immediate jump address using a sequence
of one MOVN (move wide negated) and two MOVK instructions, where the
first one sets the lower 16 bits but also sets all top bits to 0x1.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
David S. Miller [Fri, 30 Nov 2018 02:15:07 +0000 (18:15 -0800)]
Merge git://git./linux/kernel/git/bpf/bpf-next

Daniel Borkmann says:

====================
bpf-next 2018-11-30

The following pull-request contains BPF updates for your *net-next* tree.

(Getting out bit earlier this time to pull in a dependency from bpf.)

The main changes are:

1) Add libbpf ABI versioning and document API naming conventions
   as well as ABI versioning process, from Andrey.

2) Add a new sk_msg_pop_data() helper for sk_msg based BPF
   programs that is used in conjunction with sk_msg_push_data()
   for adding / removing meta data to the msg data, from John.

3) Optimize convert_bpf_ld_abs() for 0 offset and fix various
   lib and testsuite build failures on 32 bit, from David.

4) Make BPF prog dump for !JIT identical to how we dump subprogs
   when JIT is in use, from Yonghong.

5) Rename btf_get_from_id() to make it more conform with libbpf
   API naming conventions, from Martin.

6) Add a missing BPF kselftest config item, from Naresh.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agotools/bpf: make libbpf _GNU_SOURCE friendly
Yonghong Song [Thu, 29 Nov 2018 23:31:45 +0000 (15:31 -0800)]
tools/bpf: make libbpf _GNU_SOURCE friendly

During porting libbpf to bcc, I got some warnings like below:
  ...
  [  2%] Building C object src/cc/CMakeFiles/bpf-shared.dir/libbpf/src/libbpf.c.o
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf.c:12:0:
  warning: "_GNU_SOURCE" redefined [enabled by default]
   #define _GNU_SOURCE
  ...
  [  3%] Building C object src/cc/CMakeFiles/bpf-shared.dir/libbpf/src/libbpf_errno.c.o
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf_errno.c: In function ‘libbpf_strerror’:
  /home/yhs/work/bcc2/src/cc/libbpf/src/libbpf_errno.c:45:7:
  warning: assignment makes integer from pointer without a cast [enabled by default]
     ret = strerror_r(err, buf, size);
  ...

bcc is built with _GNU_SOURCE defined and this caused the above warning.
This patch intends to make libpf _GNU_SOURCE friendly by
  . define _GNU_SOURCE in libbpf.c unless it is not defined
  . undefine _GNU_SOURCE as non-gnu version of strerror_r is expected.

Signed-off-by: Yonghong Song <yhs@fb.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agonet: Don't default Aquantia USB driver to 'y'
David S. Miller [Fri, 30 Nov 2018 00:01:34 +0000 (16:01 -0800)]
net: Don't default Aquantia USB driver to 'y'

Reported-by: Pavel Machek <pavel@ucw.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: explain __skb_checksum_complete() with comments
Cong Wang [Mon, 26 Nov 2018 17:31:26 +0000 (09:31 -0800)]
net: explain __skb_checksum_complete() with comments

Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agotcp: remove loop to compute wscale
Eric Dumazet [Thu, 29 Nov 2018 15:56:20 +0000 (07:56 -0800)]
tcp: remove loop to compute wscale

We can remove the loop and conditional branches
and compute wscale efficiently thanks to ilog2()

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqede - Add a statistic for a case where driver drops tx packet due to memory allocati...
Michael Shteinbok [Thu, 29 Nov 2018 08:50:19 +0000 (10:50 +0200)]
qede - Add a statistic for a case where driver drops tx packet due to memory allocation failure.

skb_linearization can fail due to memory allocation failure.
In such a case, the driver will drop the packet. In such a case
The driver used to print an error message.
This patch replaces this error message by a dedicated statistic.

Signed-off-by: Michael Shteinbok <michael.shteinbok@cavium.com>
Signed-off-by: Ariel Elior <ariel.elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Add "fall through" comments
Ioana Ciocoi Radulescu [Thu, 29 Nov 2018 08:43:40 +0000 (08:43 +0000)]
dpaa2-eth: Add "fall through" comments

Add comments in the switch statement for XDP action to indicate
fallthrough is intended.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'ave-suspend-resume'
David S. Miller [Thu, 29 Nov 2018 18:34:46 +0000 (10:34 -0800)]
Merge branch 'ave-suspend-resume'

Kunihiko Hayashi says:

====================
Add suspend/resume support for AVE ethernet driver

This series adds support for suspend/resume to AVE ethernet driver.

And to avoid the error that wol state of phy hardware is enabled by default,
this sets initial wol state to disabled and add preservation the state in
suspend/resume sequence.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: ethernet: ave: Preserve wol state in suspend/resume sequence
Kunihiko Hayashi [Thu, 29 Nov 2018 08:06:33 +0000 (17:06 +0900)]
net: ethernet: ave: Preserve wol state in suspend/resume sequence

Since the wol state forces to be initialized after reset, the state should
be preserved in suspend/resume sequence.

Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: ethernet: ave: Set initial wol state to disabled
Kunihiko Hayashi [Thu, 29 Nov 2018 08:06:32 +0000 (17:06 +0900)]
net: ethernet: ave: Set initial wol state to disabled

If wol state of phy hardware is enabled after reset, phy_ethtool_get_wol()
returns that wol.wolopts is true.

However, since net_device.wol_enabled is zero and this doesn't apply wol
state until calling ethtool_set_wol(), so mdio_bus_phy_may_suspend()
returns true, that is, it's in a state where phy can suspend even though
wol state is enabled.

In this inconsistency, phy_suspend() returns -EBUSY, and at last,
suspend sequence fails with the following message:

    dpm_run_callback(): mdio_bus_phy_suspend+0x0/0x58 returns -16
    PM: Device 65000000.ethernet-ffffffff:01 failed to suspend: error -16
    PM: Some devices failed to suspend, or early wake event detected

In order to fix the above issue, this patch forces to set initial wol state
to disabled as default.

Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: ethernet: ave: Add suspend/resume support
Kunihiko Hayashi [Thu, 29 Nov 2018 08:06:31 +0000 (17:06 +0900)]
net: ethernet: ave: Add suspend/resume support

This patch introduces suspend and resume functions to ave driver.

Signed-off-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge tag 'linux-can-next-for-4.21-20181128' of git://git.kernel.org/pub/scm/linux...
David S. Miller [Thu, 29 Nov 2018 07:18:16 +0000 (23:18 -0800)]
Merge tag 'linux-can-next-for-4.21-20181128' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
This is a pull request for net-next/master consisting of 18 patches.

The first patch is by Colin Ian King and fixes the spelling in the ucan
driver.

The next three patches target the xilinx driver. YueHaibing's patch
fixes the return type of ndo_start_xmit function. Two patches by
Shubhrajyoti Datta add support for the CAN FD 2.0 controllers.

Flavio Suligoi's patch for the sja1000 driver add support for the ASEM
CAN raw hardware.

Wolfram Sang's and Kuninori Morimoto's patches switch the rcar driver to
use SPDX license identifiers.

The remaining 111 patches improve the flexcan driver. Pankaj Bansal's
patch enables the driver in Kconfig on all architectures with IOMEM
support. The next four patches by me fix indention, add missing
parentheses and comments. Aisheng Dong's patches add self wake support
and document it in the DT bindings. The remaining patches by Pankaj
Bansal first fix the loopback support and prepare the driver for the
CAN-FD support needed for the LX2160A SoC. The actual CAN-FD support
will be added in a later patch series.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
David S. Miller [Thu, 29 Nov 2018 06:10:54 +0000 (22:10 -0800)]
Merge git://git./linux/kernel/git/davem/net

Trivial conflict in net/core/filter.c, a locally computed
'sdif' is now an argument to the function.

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agobpf: Fix various lib and testsuite build failures on 32-bit.
David Miller [Wed, 28 Nov 2018 20:56:10 +0000 (12:56 -0800)]
bpf: Fix various lib and testsuite build failures on 32-bit.

Cannot cast a u64 to a pointer on 32-bit without an intervening (long)
cast otherwise GCC warns.

Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
6 years agoselftests/bpf: add config fragment CONFIG_FTRACE_SYSCALLS
Naresh Kamboju [Tue, 27 Nov 2018 15:24:44 +0000 (20:54 +0530)]
selftests/bpf: add config fragment CONFIG_FTRACE_SYSCALLS

CONFIG_FTRACE_SYSCALLS=y is required for get_cgroup_id_user test case
this test reads a file from debug trace path
/sys/kernel/debug/tracing/events/syscalls/sys_enter_nanosleep/id

Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Acked-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoMerge branch 'bpf-sk-msg-pop-data'
Daniel Borkmann [Wed, 28 Nov 2018 21:07:57 +0000 (22:07 +0100)]
Merge branch 'bpf-sk-msg-pop-data'

John Fastabend says:

====================
After being able to add metadata to messages with sk_msg_push_data we
have also found it useful to be able to "pop" this metadata off before
sending it to applications in some cases. This series adds a new helper
sk_msg_pop_data() and the associated patches to add tests and tools/lib
support.

Thanks!

v2: Daniel caught that we missed adding sk_msg_pop_data to the changes
    data helper so that the verifier ensures BPF programs revalidate
    data after using this helper. Also improve documentation adding a
    return description and using RST syntax per Quentin's comment. And
    delta calculations for DROP with pop'd data (albeit a strange set
    of operations for a program to be doing) had potential to be
    incorrect possibly confusing user space applications, so fix it.
====================

Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: test_sockmap, add options for msg_pop_data() helper
John Fastabend [Mon, 26 Nov 2018 22:16:19 +0000 (14:16 -0800)]
bpf: test_sockmap, add options for msg_pop_data() helper

Similar to msg_pull_data and msg_push_data add a set of options to
have msg_pop_data() exercised.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: add msg_pop_data helper to tools
John Fastabend [Mon, 26 Nov 2018 22:16:18 +0000 (14:16 -0800)]
bpf: add msg_pop_data helper to tools

Add the necessary header definitions to tools for new
msg_pop_data_helper.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agobpf: helper to pop data from messages
John Fastabend [Mon, 26 Nov 2018 22:16:17 +0000 (14:16 -0800)]
bpf: helper to pop data from messages

This adds a BPF SK_MSG program helper so that we can pop data from a
msg. We use this to pop metadata from a previous push data call.

Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
Linus Torvalds [Wed, 28 Nov 2018 20:53:48 +0000 (12:53 -0800)]
Merge git://git./linux/kernel/git/davem/net

Pull networking fixes from David Miller:

 1) ARM64 JIT fixes for subprog handling from Daniel Borkmann.

 2) Various sparc64 JIT bug fixes (fused branch convergance, frame
    pointer usage detection logic, PSEODU call argument handling).

 3) Fix to use BH locking in nf_conncount, from Taehee Yoo.

 4) Fix race of TX skb freeing in ipheth driver, from Bernd Eckstein.

 5) Handle return value of TX NAPI completion properly in lan743x
    driver, from Bryan Whitehead.

 6) MAC filter deletion in i40e driver clears wrong state bit, from
    Lihong Yang.

 7) Fix use after free in rionet driver, from Pan Bian.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (53 commits)
  s390/qeth: fix length check in SNMP processing
  net: hisilicon: remove unexpected free_netdev
  rapidio/rionet: do not free skb before reading its length
  i40e: fix kerneldoc for xsk methods
  ixgbe: recognize 1000BaseLX SFP modules as 1Gbps
  i40e: Fix deletion of MAC filters
  igb: fix uninitialized variables
  netfilter: nf_tables: deactivate expressions in rule replecement routine
  lan743x: Enable driver to work with LAN7431
  tipc: fix lockdep warning during node delete
  lan743x: fix return value for lan743x_tx_napi_poll
  net: via: via-velocity: fix spelling mistake "alignement" -> "alignment"
  qed: fix spelling mistake "attnetion" -> "attention"
  net: thunderx: fix NULL pointer dereference in nic_remove
  sctp: increase sk_wmem_alloc when head->truesize is increased
  firestream: fix spelling mistake: "Inititing" -> "Initializing"
  net: phy: add workaround for issue where PHY driver doesn't bind to the device
  usbnet: ipheth: fix potential recvmsg bug and recvmsg bug 2
  sparc: Adjust bpf JIT prologue for PSEUDO calls.
  bpf, doc: add entries of who looks over which jits
  ...

6 years agoMerge tag 'xtensa-20181128' of git://github.com/jcmvbkbc/linux-xtensa
Linus Torvalds [Wed, 28 Nov 2018 20:51:10 +0000 (12:51 -0800)]
Merge tag 'xtensa-20181128' of git://github.com/jcmvbkbc/linux-xtensa

Pull Xtensa fixes from Max Filippov:

 - fix kernel exception on userspace access to a currently disabled
   coprocessor

 - fix coprocessor data saving/restoring in configurations with multiple
   coprocessors

 - fix ptrace access to coprocessor data on configurations with multiple
   coprocessors with high alignment requirements

* tag 'xtensa-20181128' of git://github.com/jcmvbkbc/linux-xtensa:
  xtensa: fix coprocessor part of ptrace_{get,set}xregs
  xtensa: fix coprocessor context offset definitions
  xtensa: enable coprocessors that are being flushed

6 years agoMerge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net...
David S. Miller [Wed, 28 Nov 2018 19:33:35 +0000 (11:33 -0800)]
Merge branch '1GbE' of git://git./linux/kernel/git/jkirsher/net-queue

Jeff Kirsher says:

====================
Intel Wired LAN Driver Fixes 2018-11-28

This series contains fixes to igb, ixgbe and i40e.

Yunjian Wang from Huawei resolves a variable that could potentially be
NULL before it is used.

Lihong fixes an i40e issue which goes back to 4.17 kernels, where
deleting any of the MAC filters was causing the incorrect syncing for
the PF.

Josh Elsasser caught that there were missing enum values in the link
capabilities for x550 devices, which was preventing link for 1000BaseLX
SFP modules.

Jan fixes the function header comments for XSK methods.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agos390/qeth: fix length check in SNMP processing
Julian Wiedmann [Wed, 28 Nov 2018 15:20:50 +0000 (16:20 +0100)]
s390/qeth: fix length check in SNMP processing

The response for a SNMP request can consist of multiple parts, which
the cmd callback stages into a kernel buffer until all parts have been
received. If the callback detects that the staging buffer provides
insufficient space, it bails out with error.
This processing is buggy for the first part of the response - while it
initially checks for a length of 'data_len', it later copies an
additional amount of 'offsetof(struct qeth_snmp_cmd, data)' bytes.

Fix the calculation of 'data_len' for the first part of the response.
This also nicely cleans up the memcpy code.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Julian Wiedmann <jwi@linux.ibm.com>
Reviewed-by: Ursula Braun <ubraun@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: qualcomm: rmnet: remove set but not used variables 'ip_family, fc_seq, qos_id'
YueHaibing [Wed, 28 Nov 2018 12:31:42 +0000 (20:31 +0800)]
net: qualcomm: rmnet: remove set but not used variables 'ip_family, fc_seq, qos_id'

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:26:6:
 warning: variable 'ip_family' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:27:6:
 warning: variable 'fc_seq' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qualcomm/rmnet/rmnet_map_command.c:28:6:
 warning: variable 'qos_id' set but not used [-Wunused-but-set-variable]

It never used since introduction in commit
ceed73a2cf4a ("drivers: net: ethernet: qualcomm: rmnet: Initial implementation")

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoqlcnic: remove set but not used variables 'cur_rings, max_hw_rings, tx_desc_info'
YueHaibing [Wed, 28 Nov 2018 12:29:39 +0000 (20:29 +0800)]
qlcnic: remove set but not used variables 'cur_rings, max_hw_rings, tx_desc_info'

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:4011:5:
 warning: variable 'max_hw_rings' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:4013:6:
 warning: variable 'cur_rings' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c:2996:25:
 warning: variable 'tx_desc_info' set but not used [-Wunused-but-set-variable]

'cur_rings, max_hw_rings' never used since introduction in commit
34e8c406fda5 ("qlcnic: refactor Tx/SDS ring calculation and validation in driver.")
'tx_desc_info' never used since commit
95b3890ae39f ("qlcnic: Enhance Tx timeout debugging.")
Also 'queue_type' only can be QLCNIC_RX_QUEUE/QLCNIC_TX_QUEUE,
so make a trival cleanup on if statement.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Shahed Shaikh <shahed.shaikh@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: neterion: vxge: remove set but not used variables 'max_frags' and 'txdl_priv'
YueHaibing [Wed, 28 Nov 2018 12:24:41 +0000 (20:24 +0800)]
net: neterion: vxge: remove set but not used variables 'max_frags' and 'txdl_priv'

Fixes gcc '-Wunused-but-set-variable' warning:

drivers/net/ethernet/neterion/vxge/vxge-traffic.c:1698:35:
 warning: variable 'txdl_priv' set but not used [-Wunused-but-set-variable]
drivers/net/ethernet/neterion/vxge/vxge-traffic.c:1699:6:
 warning: variable 'max_frags' set but not used [-Wunused-but-set-variable]

It never used since introduction in
commit 113241321dcd ("Neterion: New driver: Traffic & alarm handler")

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf
David S. Miller [Wed, 28 Nov 2018 19:02:45 +0000 (11:02 -0800)]
Merge git://git./pub/scm/linux/kernel/git/pablo/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

The following patchset contains Netfilter fixes for net:

1) Disable BH while holding list spinlock in nf_conncount, from
   Taehee Yoo.

2) List corruption in nf_conncount, also from Taehee.

3) Fix race that results in leaving around an empty list node in
   nf_conncount, from Taehee Yoo.

4) Proper chain handling for inactive chains from the commit path,
   from Florian Westphal. This includes a selftest for this.

5) Do duplicate rule handles when replacing rules, also from Florian.

6) Remove net_exit path in xt_RATEEST that results in splat, from Taehee.

7) Possible use-after-free in nft_compat when releasing extensions.
   From Florian.

8) Memory leak in xt_hashlimit, from Taehee.

9) Call ip_vs_dst_notifier after ipv6_dev_notf, from Xin Long.

10) Fix cttimeout with udplite and gre, from Florian.

11) Preserve oif for IPv6 link-local generated traffic from mangle
    table, from Alin Nastac.

12) Missing error handling in masquerade notifiers, from Taehee Yoo.

13) Use mutex to protect registration/unregistration of masquerade
    extensions in order to prevent a race, from Taehee.

14) Incorrect condition check in tree_nodes_free(), also from Taehee.

15) Fix chain counter leak in rule replacement path, from Taehee.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'dpaa2-eth-Introduce-XDP-support'
David S. Miller [Wed, 28 Nov 2018 18:57:46 +0000 (10:57 -0800)]
Merge branch 'dpaa2-eth-Introduce-XDP-support'

Ioana Ciocoi Radulescu says:

====================
dpaa2-eth: Introduce XDP support

Add support for XDP programs. Only XDP_PASS, XDP_DROP and XDP_TX
actions are supported for now. Frame header changes are also
allowed.

v2: - count the XDP packets in the rx/tx inteface stats
    - add message with the maximum supported MTU value for XDP
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Add xdp counters
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:34 +0000 (16:27 +0000)]
dpaa2-eth: Add xdp counters

Add counters for xdp processed frames to the channel statistics.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Cleanup channel stats
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:33 +0000 (16:27 +0000)]
dpaa2-eth: Cleanup channel stats

Remove unused counter. Reorder fields in channel stats structure
to match the ethtool strings order and make it easier to print them
with ethtool -S.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Add support for XDP_TX
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:33 +0000 (16:27 +0000)]
dpaa2-eth: Add support for XDP_TX

Send frames back on the same port for XDP_TX action.
Since the frame buffers have been allocated by us, we can recycle
them directly into the Rx buffer pool instead of requesting a
confirmation frame upon transmission complete.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Map Rx buffers as bidirectional
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:32 +0000 (16:27 +0000)]
dpaa2-eth: Map Rx buffers as bidirectional

In order to support enqueueing Rx FDs back to hardware, we need to
DMA map Rx buffers as bidirectional.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Release buffers back to pool on XDP_DROP
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:31 +0000 (16:27 +0000)]
dpaa2-eth: Release buffers back to pool on XDP_DROP

Instead of freeing the RX buffers, release them back into the pool.
We wait for the maximum number of buffers supported by a single
release command to accumulate before issuing the command.

Also, don't unmap the Rx buffers at the beginning of the Rx routine
anymore, since that would require remapping them before release.
Instead, just do a DMA sync at first and only unmap if the frame is
meant for the stack.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Move function
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:31 +0000 (16:27 +0000)]
dpaa2-eth: Move function

We'll use function free_bufs() on the XDP path as well, so move
it higher in order to avoid a forward declaration.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Allow XDP header adjustments
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:30 +0000 (16:27 +0000)]
dpaa2-eth: Allow XDP header adjustments

Reserve XDP_PACKET_HEADROOM bytes in Rx buffers to allow XDP
programs to increase frame header size.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agodpaa2-eth: Add basic XDP support
Ioana Ciocoi Radulescu [Mon, 26 Nov 2018 16:27:29 +0000 (16:27 +0000)]
dpaa2-eth: Add basic XDP support

We keep one XDP program reference per channel. The only actions
supported for now are XDP_DROP and XDP_PASS.

Until now we didn't enforce a maximum size for Rx frames based
on MTU value. Change that, since for XDP mode we must ensure no
scatter-gather frames can be received.

Signed-off-by: Ioana Radulescu <ruxandra.radulescu@nxp.com>
Acked-by: Camelia Groza <camelia.groza@nxp.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agonet: hisilicon: remove unexpected free_netdev
Pan Bian [Wed, 28 Nov 2018 07:30:24 +0000 (15:30 +0800)]
net: hisilicon: remove unexpected free_netdev

The net device ndev is freed via free_netdev when failing to register
the device. The control flow then jumps to the error handling code
block. ndev is used and freed again. Resulting in a use-after-free bug.

Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agorapidio/rionet: do not free skb before reading its length
Pan Bian [Wed, 28 Nov 2018 06:53:19 +0000 (14:53 +0800)]
rapidio/rionet: do not free skb before reading its length

skb is freed via dev_kfree_skb_any, however, skb->len is read then. This
may result in a use-after-free bug.

Fixes: e6161d64263 ("rapidio/rionet: rework driver initialization and removal")
Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoi40e: fix kerneldoc for xsk methods
Jan Sokolowski [Tue, 27 Nov 2018 17:35:35 +0000 (09:35 -0800)]
i40e: fix kerneldoc for xsk methods

One method, xsk_umem_setup, had an incorrect kernel doc
description, which has been corrected.

Also fixes small typos found in the comments.

Signed-off-by: Jan Sokolowski <jan.sokolowski@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
6 years agoMerge tag 'for-4.20-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave...
Linus Torvalds [Wed, 28 Nov 2018 16:38:20 +0000 (08:38 -0800)]
Merge tag 'for-4.20-rc4-tag' of git://git./linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "Some of these bugs are being hit during testing so we'd like to get
  them merged, otherwise there are usual stability fixes for stable
  trees"

* tag 'for-4.20-rc4-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: relocation: set trans to be NULL after ending transaction
  Btrfs: fix race between enabling quotas and subvolume creation
  Btrfs: send, fix infinite loop due to directory rename dependencies
  Btrfs: ensure path name is null terminated at btrfs_control_ioctl
  Btrfs: fix rare chances for data loss when doing a fast fsync
  btrfs: Always try all copies when reading extent buffers

6 years agoMerge tag 'spi-fix-v4.20-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/brooni...
Linus Torvalds [Wed, 28 Nov 2018 16:33:55 +0000 (08:33 -0800)]
Merge tag 'spi-fix-v4.20-rc4' of git://git./linux/kernel/git/broonie/spi

Pull spi fixes from Mark Brown:
 "A few driver specific fixes here, nothing big or that stands out for
  anyone other than the driver users.

  The omap2-mcspi fix is for issues that started showing up with a
  change in defconfig in this release to make cpuidle get turned on by
  default"

* tag 'spi-fix-v4.20-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
  spi: omap2-mcspi: Add missing suspend and resume calls
  spi: mediatek: use correct mata->xfer_len when in fifo transfer
  spi: uniphier: fix incorrect property items

6 years agoixgbe: recognize 1000BaseLX SFP modules as 1Gbps
Josh Elsasser [Sat, 24 Nov 2018 20:57:33 +0000 (12:57 -0800)]
ixgbe: recognize 1000BaseLX SFP modules as 1Gbps

Add the two 1000BaseLX enum values to the X550's check for 1Gbps modules,
allowing the core driver code to establish a link over this SFP type.

This is done by the out-of-tree driver but the fix wasn't in mainline.

Fixes: e23f33367882 ("ixgbe: Fix 1G and 10G link stability for X550EM_x SFP+”)
Fixes: 6a14ee0cfb19 ("ixgbe: Add X550 support function pointers")
Signed-off-by: Josh Elsasser <jelsasser@appneta.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
6 years agoMerge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Linus Torvalds [Wed, 28 Nov 2018 16:29:18 +0000 (08:29 -0800)]
Merge tag 'for-linus' of git://git./virt/kvm/kvm

Pull kvm fixes from Paolo Bonzini:
 "Bugfixes, many of them reported by syzkaller and mostly predating the
  merge window"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  kvm: svm: Ensure an IBPB on all affected CPUs when freeing a vmcb
  kvm: mmu: Fix race in emulated page table writes
  KVM: nVMX: vmcs12 revision_id is always VMCS12_REVISION even when copied from eVMCS
  KVM: nVMX: Verify eVMCS revision id match supported eVMCS version on eVMCS VMPTRLD
  KVM: nVMX/nSVM: Fix bug which sets vcpu->arch.tsc_offset to L1 tsc_offset
  x86/kvm/vmx: fix old-style function declaration
  KVM: x86: fix empty-body warnings
  KVM: VMX: Update shared MSRs to be saved/restored on MSR_EFER.LMA changes
  KVM: x86: Fix kernel info-leak in KVM_HC_CLOCK_PAIRING hypercall
  KVM: nVMX: Fix kernel info-leak when enabling KVM_CAP_HYPERV_ENLIGHTENED_VMCS more than once
  svm: Add mutex_lock to protect apic_access_page_done on AMD systems
  KVM: X86: Fix scan ioapic use-before-initialization
  KVM: LAPIC: Fix pv ipis use-before-initialization
  KVM: VMX: re-add ple_gap module parameter
  KVM: PPC: Book3S HV: Fix handling for interrupted H_ENTER_NESTED

6 years agoi40e: Fix deletion of MAC filters
Lihong Yang [Wed, 21 Nov 2018 17:15:37 +0000 (09:15 -0800)]
i40e: Fix deletion of MAC filters

In __i40e_del_filter function, the flag __I40E_MACVLAN_SYNC_PENDING for
the PF state is wrongly set for the VSI. Deleting any of the MAC filters
has caused the incorrect syncing for the PF. Fix it by setting this state
flag to the intended PF.

CC: stable <stable@vger.kernel.org>
Signed-off-by: Lihong Yang <lihong.yang@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
6 years agoigb: fix uninitialized variables
Yunjian Wang [Tue, 6 Nov 2018 08:27:12 +0000 (16:27 +0800)]
igb: fix uninitialized variables

This patch fixes the variable 'phy_word' may be used uninitialized.

Signed-off-by: Yunjian Wang <wangyunjian@huawei.com>
Tested-by: Aaron Brown <aaron.f.brown@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
6 years agocan: flexcan: split the Message Buffer RAM area
Pankaj Bansal [Tue, 28 Aug 2018 17:49:12 +0000 (23:19 +0530)]
can: flexcan: split the Message Buffer RAM area

The message buffer RAM area is not a contiguous 1KB area but 2 partitions
of 512 bytes each. Till now, we used Message buffers with payload size 8
bytes, which translates to 32 MBs per partition and no spare space is left
in any partition.
However, in upcoming SOC LX2160A the message buffers can have payload size
64 bytes. This results in less than 32 MBs per partition and some empty
area is left at the end of each partition.This empty area should not be
accessed.
Therefore, split the Message Buffer RAM area into two partitions.

Signed-off-by: Pankaj Bansal <pankaj.bansal@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: Add provision for variable payload size
Pankaj Bansal [Fri, 23 Nov 2018 21:18:44 +0000 (22:18 +0100)]
can: flexcan: Add provision for variable payload size

Till now the flexcan module supported 8 byte payload size as per CAN 2.0
specifications. But now upcoming flexcan module in NXP LX2160A SOC
supports CAN FD protocol too. The Message buffers need to be configured
to have payload size 64 bytes.

Therefore, added provision in the driver for payload size to be 64 bytes.

Signed-off-by: Pankaj Bansal <pankaj.bansal@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: move rx_offload_add() from flexcan_probe() to flexcan_open()
Pankaj Bansal [Mon, 13 Aug 2018 18:20:49 +0000 (23:50 +0530)]
can: flexcan: move rx_offload_add() from flexcan_probe() to flexcan_open()

rx offload depends on number of message buffers, which in turn depends
on messgae buffer size. with the upcoming LX2160A SOC the message buffer
size can be configured to 72 bytes if it were to be used in CAN FD mode.

The current mode in which the flexcan is being operated is known at the
time of flexcan_open() but not at the time of flexcan_probe().

Therefore, move the rx_offload_add() from flexcan_probe() to
flexcan_open(). correspondingly, move rx_offload_delete() from
flexcan_remove() to flexcan_close().

Signed-off-by: Pankaj Bansal <pankaj.bansal@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: flexcan_chip_start(): enable loopback mode in flexcan
Pankaj Bansal [Mon, 13 Aug 2018 18:20:48 +0000 (23:50 +0530)]
can: flexcan: flexcan_chip_start(): enable loopback mode in flexcan

Self reception disable bit needs to be cleared for loopback mode to work
in flexcan.

Signed-off-by: Pankaj Bansal <pankaj.bansal@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: add self wakeup support
Aisheng Dong [Fri, 23 Nov 2018 08:35:33 +0000 (08:35 +0000)]
can: flexcan: add self wakeup support

If wakeup is enabled, enter stop mode, else enter disabled mode. Self wake
can only work on stop mode.

Starting from IMX6, the flexcan stop mode control bits is SoC specific,
move it out of IP driver and parse it from devicetree.

Signed-off-by: Aisheng Dong <aisheng.dong@nxp.com>
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Reviewed-by: Dong Aisheng <aisheng.dong@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agodt-bindings: can: flexcan: add stop mode property to device tree
Aisheng Dong [Fri, 23 Nov 2018 08:35:29 +0000 (08:35 +0000)]
dt-bindings: can: flexcan: add stop mode property to device tree

The FlexCAN controller can parse the stop mode property to enable CAN
self wakeup feature.

Signed-off-by: Aisheng Dong <aisheng.dong@nxp.com>
Signed-off-by: Joakim Zhang <qiangqing.zhang@nxp.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: flexcan_chip_start(): adjust comment to match the code
Marc Kleine-Budde [Fri, 17 Aug 2018 12:52:58 +0000 (14:52 +0200)]
can: flexcan: flexcan_chip_start(): adjust comment to match the code

With the conversion of the flexcan driver to support both timestamp and
FIFO mode the setup of the MCR register ("enable fifo") has been moved.

This patch moves the comment too, in order to match the code again.

Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: FLEXCAN_IFLAG_MB: add () around macro argument
Marc Kleine-Budde [Wed, 28 Nov 2018 14:31:37 +0000 (15:31 +0100)]
can: flexcan: FLEXCAN_IFLAG_MB: add () around macro argument

This patch fixes the following checkpatch warning:

| Macro argument 'x' may be better as '(x)' to avoid precedence issues

Fixes: cbffaf7aa09e ("can: flexcan: Always use last mailbox for TX")
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: flexcan_irq(): fix indention
Marc Kleine-Budde [Wed, 28 Nov 2018 14:45:27 +0000 (15:45 +0100)]
can: flexcan: flexcan_irq(): fix indention

The patch fixes the indention by replacing space by tabs, as noted by
checkpatch:

| ERROR: code indent should use tabs where possible
| #980: FILE: drivers/net/can/flexcan.c:980:

Fixes: da49a8075c00 ("can: flexcan: implement error passive state quirk")
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: flexcan_start_xmit(): fix indention
Marc Kleine-Budde [Mon, 12 Nov 2018 14:33:57 +0000 (15:33 +0100)]
can: flexcan: flexcan_start_xmit(): fix indention

This patch fixes the indentio nin flexcan_start_xmit() by alligning the
code to the opening parenthesis.

Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: flexcan: enable flexcan for all architectures
Pankaj Bansal [Wed, 1 Aug 2018 14:06:45 +0000 (19:36 +0530)]
can: flexcan: enable flexcan for all architectures

flexcan is an IP module independent of the SOC architecture.
Therefore, enable it for all architectures.

Signed-off-by: Pankaj Bansal <pankaj.bansal@nxp.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: rcar: add SPDX identifiers to Kconfig and Makefile
Kuninori Morimoto [Wed, 26 Sep 2018 01:41:53 +0000 (01:41 +0000)]
can: rcar: add SPDX identifiers to Kconfig and Makefile

This patch adds SPDX-License-Identifier to Kconfig and Makefile.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Acked-by: Ramesh Shanmugasundaram <ramesh.shanmugasundaram@bp.renesas.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: rcar: use SPDX identifier for Renesas drivers
Wolfram Sang [Thu, 23 Aug 2018 13:34:55 +0000 (15:34 +0200)]
can: rcar: use SPDX identifier for Renesas drivers

Adopt the SPDX license identifier headers to ease license compliance
management.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Reviewed-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: sja1000: plx_pci: add support for ASEM CAN raw device
Flavio Suligoi [Tue, 7 Aug 2018 07:17:19 +0000 (09:17 +0200)]
can: sja1000: plx_pci: add support for ASEM CAN raw device

This patch adds support for ASEM opto-isolated dual channels
CAN raw device (http://www.asem.it)

Signed-off-by: Flavio Suligoi <f.suligoi@asem.it>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: xilinx: add can 2.0 support
Shubhrajyoti Datta [Fri, 12 Oct 2018 04:25:08 +0000 (09:55 +0530)]
can: xilinx: add can 2.0 support

Add support for can 2.0.

Signed-off-by: Shubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agodt-bindings: can: xilinx_can: add Xilinx CAN FD 2.0 bindings
Shubhrajyoti Datta [Fri, 12 Oct 2018 04:25:09 +0000 (09:55 +0530)]
dt-bindings: can: xilinx_can: add Xilinx CAN FD 2.0 bindings

Add compatible string and new attributes to support the Xilinx CAN
FD 2.0.

Signed-off-by: Shubhrajyoti Datta <shubhrajyoti.datta@xilinx.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: xilinx: fix return type of ndo_start_xmit function
YueHaibing [Wed, 26 Sep 2018 10:32:29 +0000 (18:32 +0800)]
can: xilinx: fix return type of ndo_start_xmit function

The method ndo_start_xmit() is defined as returning an 'netdev_tx_t',
which is a typedef for an enum type, so make sure the implementation in
this driver has returns 'netdev_tx_t' value, and change the function
return type to netdev_tx_t.

Found by coccinelle.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agocan: ucan: fix spelling mistake: "resumbmitting" -> "resubmitting"
Colin Ian King [Mon, 30 Jul 2018 13:40:37 +0000 (14:40 +0100)]
can: ucan: fix spelling mistake: "resumbmitting" -> "resubmitting"

Trivial fix to spelling mistake in netdev_dbg error message

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
6 years agonetfilter: nf_tables: deactivate expressions in rule replecement routine
Taehee Yoo [Wed, 28 Nov 2018 02:27:28 +0000 (11:27 +0900)]
netfilter: nf_tables: deactivate expressions in rule replecement routine

There is no expression deactivation call from the rule replacement path,
hence, chain counter is not decremented. A few steps to reproduce the
problem:

   %nft add table ip filter
   %nft add chain ip filter c1
   %nft add chain ip filter c1
   %nft add rule ip filter c1 jump c2
   %nft replace rule ip filter c1 handle 3 accept
   %nft flush ruleset

<jump c2> expression means immediate NFT_JUMP to chain c2.
Reference count of chain c2 is increased when the rule is added.

When rule is deleted or replaced, the reference counter of c2 should be
decreased via nft_rule_expr_deactivate() which calls
nft_immediate_deactivate().

Splat looks like:
[  214.396453] WARNING: CPU: 1 PID: 21 at net/netfilter/nf_tables_api.c:1432 nf_tables_chain_destroy.isra.38+0x2f9/0x3a0 [nf_tables]
[  214.398983] Modules linked in: nf_tables nfnetlink
[  214.398983] CPU: 1 PID: 21 Comm: kworker/1:1 Not tainted 4.20.0-rc2+ #44
[  214.398983] Workqueue: events nf_tables_trans_destroy_work [nf_tables]
[  214.398983] RIP: 0010:nf_tables_chain_destroy.isra.38+0x2f9/0x3a0 [nf_tables]
[  214.398983] Code: 00 00 00 fc ff df 48 89 fa 48 c1 ea 03 80 3c 02 00 0f 85 8e 00 00 00 48 8b 7b 58 e8 e1 2c 4e c6 48 89 df e8 d9 2c 4e c6 eb 9a <0f> 0b eb 96 0f 0b e9 7e fe ff ff e8 a7 7e 4e c6 e9 a4 fe ff ff e8
[  214.398983] RSP: 0018:ffff8881152874e8 EFLAGS: 00010202
[  214.398983] RAX: 0000000000000001 RBX: ffff88810ef9fc28 RCX: ffff8881152876f0
[  214.398983] RDX: dffffc0000000000 RSI: 1ffff11022a50ede RDI: ffff88810ef9fc78
[  214.398983] RBP: 1ffff11022a50e9d R08: 0000000080000000 R09: 0000000000000000
[  214.398983] R10: 0000000000000000 R11: 0000000000000000 R12: 1ffff11022a50eba
[  214.398983] R13: ffff888114446e08 R14: ffff8881152876f0 R15: ffffed1022a50ed6
[  214.398983] FS:  0000000000000000(0000) GS:ffff888116400000(0000) knlGS:0000000000000000
[  214.398983] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  214.398983] CR2: 00007fab9bb5f868 CR3: 000000012aa16000 CR4: 00000000001006e0
[  214.398983] Call Trace:
[  214.398983]  ? nf_tables_table_destroy.isra.37+0x100/0x100 [nf_tables]
[  214.398983]  ? __kasan_slab_free+0x145/0x180
[  214.398983]  ? nf_tables_trans_destroy_work+0x439/0x830 [nf_tables]
[  214.398983]  ? kfree+0xdb/0x280
[  214.398983]  nf_tables_trans_destroy_work+0x5f5/0x830 [nf_tables]
[ ... ]

Fixes: bb7b40aecbf7 ("netfilter: nf_tables: bogus EBUSY in chain deletions")
Reported by: Christoph Anton Mitterer <calestyo@scientia.net>
Link: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=914505
Link: https://bugzilla.kernel.org/show_bug.cgi?id=201791
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
6 years agonet/ipv4: Fix missing raw_init when CONFIG_PROC_FS is disabled
David Ahern [Wed, 28 Nov 2018 04:52:24 +0000 (20:52 -0800)]
net/ipv4: Fix missing raw_init when CONFIG_PROC_FS is disabled

Randy reported when CONFIG_PROC_FS is not enabled:
    ld: net/ipv4/af_inet.o: in function `inet_init':
    af_inet.c:(.init.text+0x42d): undefined reference to `raw_init'

Fix by moving the endif up to the end of the proc entries

Fixes: 6897445fb194c ("net: provide a sysctl raw_l3mdev_accept for raw socket lookup with VRFs")
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Mike Manning <mmanning@vyatta.att-mail.com>
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agoMerge branch 'bnx2x-Popoulate-firmware-versions-in-driver-info-query'
David S. Miller [Wed, 28 Nov 2018 00:41:19 +0000 (16:41 -0800)]
Merge branch 'bnx2x-Popoulate-firmware-versions-in-driver-info-query'

Sudarsana Reddy Kalluru says:

====================
bnx2x: Popoulate firmware versions in driver info query.

The patch series populates MBI and storm firware versions in the ethtool
driver info query.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agobnx2x: Add storm FW version to ethtool driver query output.
Sudarsana Reddy Kalluru [Tue, 27 Nov 2018 06:25:57 +0000 (22:25 -0800)]
bnx2x: Add storm FW version to ethtool driver query output.

The patch populates the Storm FW version in the ethtool driver query data.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agobnx2x: Add MBI version to ethtool driver query output.
Sudarsana Reddy Kalluru [Tue, 27 Nov 2018 06:25:56 +0000 (22:25 -0800)]
bnx2x: Add MBI version to ethtool driver query output.

The patch populates the MBI version in the ethtool driver query data.
Adding 'extended_dev_info_shared_cfg' structure describing the nvram
structure, this is required to access the mbi version string.

Signed-off-by: Sudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: Ariel Elior <Ariel.Elior@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
6 years agotcp: remove hdrlen argument from tcp_queue_rcv()
Eric Dumazet [Mon, 26 Nov 2018 22:49:12 +0000 (14:49 -0800)]
tcp: remove hdrlen argument from tcp_queue_rcv()

Only one caller needs to pull TCP headers, so lets
move __skb_pull() to the caller side.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>