Alexei Starovoitov [Wed, 4 Mar 2020 00:24:00 +0000 (16:24 -0800)]
Merge branch 'bpf_gso_size'
Willem de Bruijn says:
====================
See first patch for details.
Patch split across three parts { kernel feature, uapi header, tools }
following the custom for such __sk_buff changes.
====================
Acked-by: Petar Penkov <ppenkov@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Willem de Bruijn [Tue, 3 Mar 2020 20:05:03 +0000 (15:05 -0500)]
selftests/bpf: Test new __sk_buff field gso_size
Analogous to the gso_segs selftests introduced in commit
d9ff286a0f59
("bpf: allow BPF programs access skb_shared_info->gso_segs field").
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303200503.226217-4-willemdebruijn.kernel@gmail.com
Willem de Bruijn [Tue, 3 Mar 2020 20:05:02 +0000 (15:05 -0500)]
bpf: Sync uapi bpf.h to tools/
sync tools/include/uapi/linux/bpf.h to match include/uapi/linux/bpf.h
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303200503.226217-3-willemdebruijn.kernel@gmail.com
Willem de Bruijn [Tue, 3 Mar 2020 20:05:01 +0000 (15:05 -0500)]
bpf: Add gso_size to __sk_buff
BPF programs may want to know whether an skb is gso. The canonical
answer is skb_is_gso(skb), which tests that gso_size != 0.
Expose this field in the same manner as gso_segs. That field itself
is not a sufficient signal, as the comment in skb_shared_info makes
clear: gso_segs may be zero, e.g., from dodgy sources.
Also prepare net/bpf/test_run for upcoming BPF_PROG_TEST_RUN tests
of the feature.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303200503.226217-2-willemdebruijn.kernel@gmail.com
Alexei Starovoitov [Tue, 3 Mar 2020 06:06:28 +0000 (22:06 -0800)]
Merge branch 'bpf_link'
Andrii Nakryiko says:
====================
This patch series adds bpf_link abstraction, analogous to libbpf's already
existing bpf_link abstraction. This formalizes and makes more uniform existing
bpf_link-like BPF program link (attachment) types (raw tracepoint and tracing
links), which are FD-based objects that are automatically detached when last
file reference is closed. These types of BPF program links are switched to
using bpf_link framework.
FD-based bpf_link approach provides great safety guarantees, by ensuring there
is not going to be an abandoned BPF program attached, if user process suddenly
exits or forgets to clean up after itself. This is especially important in
production environment and is what all the recent new BPF link types followed.
One of the previously existing inconveniences of FD-based approach, though,
was the scenario in which user process wants to install BPF link and exit, but
let attached BPF program run. Now, with bpf_link abstraction in place, it's
easy to support pinning links in BPF FS, which is done as part of the same
patch #1. This allows FD-based BPF program links to survive exit of a user
process and original file descriptor being closed, by creating an file entry
in BPF FS. This provides great safety by default, with simple way to opt out
for cases where it's needed.
Corresponding libbpf APIs are added in the same patch set, as well as
selftests for this functionality.
Other types of BPF program attachments (XDP, cgroup, perf_event, etc) are
going to be converted in subsequent patches to follow similar approach.
v1->v2:
- use bpf_link_new_fd() uniformly (Alexei).
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Andrii Nakryiko [Tue, 3 Mar 2020 04:31:59 +0000 (20:31 -0800)]
selftests/bpf: Add link pinning selftests
Add selftests validating link pinning/unpinning and associated BPF link
(attachment) lifetime.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303043159.323675-4-andriin@fb.com
Andrii Nakryiko [Tue, 3 Mar 2020 04:31:58 +0000 (20:31 -0800)]
libbpf: Add bpf_link pinning/unpinning
With bpf_link abstraction supported by kernel explicitly, add
pinning/unpinning API for links. Also allow to create (open) bpf_link from BPF
FS file.
This API allows to have an "ephemeral" FD-based BPF links (like raw tracepoint
or fexit/freplace attachments) surviving user process exit, by pinning them in
a BPF FS, which is an important use case for long-running BPF programs.
As part of this, expose underlying FD for bpf_link. While legacy bpf_link's
might not have a FD associated with them (which will be expressed as
a bpf_link with fd=-1), kernel's abstraction is based around FD-based usage,
so match it closely. This, subsequently, allows to have a generic
pinning/unpinning API for generalized bpf_link. For some types of bpf_links
kernel might not support pinning, in which case bpf_link__pin() will return
error.
With FD being part of generic bpf_link, also get rid of bpf_link_fd in favor
of using vanialla bpf_link.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303043159.323675-3-andriin@fb.com
Andrii Nakryiko [Tue, 3 Mar 2020 04:31:57 +0000 (20:31 -0800)]
bpf: Introduce pinnable bpf_link abstraction
Introduce bpf_link abstraction, representing an attachment of BPF program to
a BPF hook point (e.g., tracepoint, perf event, etc). bpf_link encapsulates
ownership of attached BPF program, reference counting of a link itself, when
reference from multiple anonymous inodes, as well as ensures that release
callback will be called from a process context, so that users can safely take
mutex locks and sleep.
Additionally, with a new abstraction it's now possible to generalize pinning
of a link object in BPF FS, allowing to explicitly prevent BPF program
detachment on process exit by pinning it in a BPF FS and let it open from
independent other process to keep working with it.
Convert two existing bpf_link-like objects (raw tracepoint and tracing BPF
program attachments) into utilizing bpf_link framework, making them pinnable
in BPF FS. More FD-based bpf_links will be added in follow up patches.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200303043159.323675-2-andriin@fb.com
Toke Høiland-Jørgensen [Mon, 2 Mar 2020 14:53:48 +0000 (15:53 +0100)]
selftests/bpf: Declare bpf_log_buf variables as static
The cgroup selftests did not declare the bpf_log_buf variable as static, leading
to a linker error with GCC 10 (which defaults to -fno-common). Fix this by
adding the missing static declarations.
Fixes: 257c88559f36 ("selftests/bpf: Convert test_cgroup_attach to prog_tests")
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrey Ignatov <rdna@fb.com>
Link: https://lore.kernel.org/bpf/20200302145348.559177-1-toke@redhat.com
Andrii Nakryiko [Sun, 1 Mar 2020 08:10:43 +0000 (00:10 -0800)]
bpf: Reliably preserve btf_trace_xxx types
btf_trace_xxx types, crucial for tp_btf BPF programs (raw tracepoint with
verifier-checked direct memory access), have to be preserved in kernel BTF to
allow verifier do its job and enforce type/memory safety. It was reported
([0]) that for kernels built with Clang current type-casting approach doesn't
preserve these types.
This patch fixes it by declaring an anonymous union for each registered
tracepoint, capturing both struct bpf_raw_event_map information, as well as
recording btf_trace_##call type reliably. Structurally, it's still the same
content as for a plain struct bpf_raw_event_map, so no other changes are
necessary.
[0] https://github.com/iovisor/bcc/issues/2770#issuecomment-
591007692
Fixes: e8c423fb31fa ("bpf: Add typecast to raw_tracepoints to help BTF generation")
Reported-by: Wenbo Zhang <ethercflow@gmail.com>
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200301081045.3491005-2-andriin@fb.com
Alexei Starovoitov [Tue, 3 Mar 2020 00:25:15 +0000 (16:25 -0800)]
Merge branch 'move_BPF_PROG_to_libbpf'
Andrii Nakryiko says:
====================
Move BPF_PROG, BPF_KPROBE, and BPF_KRETPROBE helper macros from private
selftests helpers to public libbpf ones. These helpers are extremely helpful
for writing tracing BPF applications and have been requested to be exposed for
easy use (e.g., [0]).
As part of this move, fix up BPF_KRETPROBE to not allow for capturing input
arguments (as it's unreliable and they will be often clobbered). Also, add
vmlinux.h header guard to allow multi-time inclusion, if necessary; but also
to let PT_REGS_PARM do proper detection of struct pt_regs field names on x86
arch. See relevant patches for more details.
[0] https://github.com/iovisor/bcc/pull/2778#issue-
381642907
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Andrii Nakryiko [Sat, 29 Feb 2020 23:11:12 +0000 (15:11 -0800)]
libbpf: Merge selftests' bpf_trace_helpers.h into libbpf's bpf_tracing.h
Move BPF_PROG, BPF_KPROBE, and BPF_KRETPROBE macro into libbpf's bpf_tracing.h
header to make it available for non-selftests users.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200229231112.1240137-5-andriin@fb.com
Andrii Nakryiko [Sat, 29 Feb 2020 23:11:11 +0000 (15:11 -0800)]
selftests/bpf: Fix BPF_KRETPROBE macro and use it in attach_probe test
For kretprobes, there is no point in capturing input arguments from pt_regs,
as they are going to be, most probably, clobbered by the time probed kernel
function returns. So switch BPF_KRETPROBE to accept zero or one argument
(optional return result).
Fixes: ac065870d928 ("selftests/bpf: Add BPF_PROG, BPF_KPROBE, and BPF_KRETPROBE macros")
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200229231112.1240137-4-andriin@fb.com
Andrii Nakryiko [Sat, 29 Feb 2020 23:11:10 +0000 (15:11 -0800)]
libbpf: Fix use of PT_REGS_PARM macros with vmlinux.h
Add detection of vmlinux.h to bpf_tracing.h header for PT_REGS macro.
Currently, BPF applications have to define __KERNEL__ symbol to use correct
definition of struct pt_regs on x86 arch. This is due to different field names
under internal kernel vs UAPI conditions. To make this more transparent for
users, detect vmlinux.h by checking __VMLINUX_H__ symbol.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200229231112.1240137-3-andriin@fb.com
Andrii Nakryiko [Sat, 29 Feb 2020 23:11:09 +0000 (15:11 -0800)]
bpftool: Add header guards to generated vmlinux.h
Add canonical #ifndef/#define/#endif guard for generated vmlinux.h header with
__VMLINUX_H__ symbol. __VMLINUX_H__ is also going to play double role of
identifying whether vmlinux.h is being used, versus, say, BCC or non-CO-RE
libbpf modes with dependency on kernel headers. This will make it possible to
write helper macro/functions, agnostic to exact BPF program set up.
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200229231112.1240137-2-andriin@fb.com
Jesper Dangaard Brouer [Mon, 2 Mar 2020 13:46:28 +0000 (14:46 +0100)]
mvneta: add XDP ethtool errors stats for TX to driver
Adding ethtool stats for when XDP transmitted packets overrun the TX
queue. This is recorded separately for XDP_TX and ndo_xdp_xmit. This
is an important aid for troubleshooting XDP based setups.
It is currently a known weakness and property of XDP that there isn't
any push-back or congestion feedback when transmitting frames via XDP.
It's easy to realise when redirecting from a higher speed link into a
slower speed link, or simply two ingress links into a single egress.
The situation can also happen when Ethernet flow control is active.
For testing the patch and provoking the situation to occur on my
Espressobin board, I configured the TX-queue to be smaller (434) than
RX-queue (512) and overload network with large MTU size frames (as a
larger frame takes longer to transmit).
Hopefully the upcoming XDP TX hook can be extended to provide insight
into these TX queue overflows, to allow programmable adaptation
strategies.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 2 Mar 2020 19:16:35 +0000 (11:16 -0800)]
Merge branch 'net-zl-array'
More zero-length array transformations from Gustavo A. R. Silva.
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:28:26 +0000 (06:28 -0600)]
tehuti: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:23:05 +0000 (06:23 -0600)]
r8152: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:19:53 +0000 (06:19 -0600)]
net: atlantic: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:10:51 +0000 (06:10 -0600)]
bna: bnad: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:07:42 +0000 (06:07 -0600)]
net: inet_sock: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:06:07 +0000 (06:06 -0600)]
net: ip6_fib: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:03:52 +0000 (06:03 -0600)]
net: ip_fib: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:02:09 +0000 (06:02 -0600)]
drop_monitor: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 12:00:48 +0000 (06:00 -0600)]
net: mip6: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Mon, 2 Mar 2020 11:59:33 +0000 (05:59 -0600)]
netdevice: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 2 Mar 2020 19:13:58 +0000 (11:13 -0800)]
Merge branch 'net-thunderx-Miscellaneous-changes'
Sunil Goutham says:
====================
net: thunderx: Miscellaneous changes
This patchset has changes wrt driver performance optimization,
load time optimization. And a change to PCI device regiatration
table for timestamp device.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Prakash Brahmajyosyula [Mon, 2 Mar 2020 09:59:02 +0000 (15:29 +0530)]
net: cavium: Register driver with PCI subsys IDs
Across Cavium's ThunderX and Marvell's OcteonTx2 silicons
the PTP timestamping block's PCI device ID and vendor ID
have remained same but the HW architecture has changed.
Hence added PCI subsystem IDs to the device table to avoid
this driver from being probed on OcteonTx2 silicons.
Signed-off-by: Prakash Brahmajyosyula <bprakash@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Mon, 2 Mar 2020 09:59:01 +0000 (15:29 +0530)]
net: thunderx: Reduce mbox wait response time.
Replace msleep() with usleep_range() as internally it uses hrtimers.
This will put a cap on maximum wait time.
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Mon, 2 Mar 2020 09:59:00 +0000 (15:29 +0530)]
net: thunderx: Adjust CQE_RX drop levels for better performance
With the current RX RED/DROP levels of 192/184 for CQE_RX, when
packet incoming rate is high, LLC is getting polluted resulting
in more cache misses and higher latency in packet processing. This
slows down the whole process and performance loss. Hence reduced
the levels to 224/216 (ie for a CQ size of 1024, Rx pkts will be
red dropped or dropped when unused CQE are less than 128/160 respectively)
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Mon, 2 Mar 2020 19:08:52 +0000 (11:08 -0800)]
Merge branch 'octeontx2-Flow-control-support-and-other-misc-changes'
Sunil Goutham says:
====================
octeontx2: Flow control support and other misc changes
This patch series adds flow control support (802.3 pause frames) and
has other changes wrt generic admin function (AF) driver functionality.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Mon, 2 Mar 2020 07:19:28 +0000 (12:49 +0530)]
octeontx2-af: Modify rvu_reg_poll() to check reg atleast twice
Currently on the first check if the operation is still not
finished, the poll goes to sleep for 2-5 usecs. But if for
some reason (due to other priority stuff like interrupts etc) by
the time the poll wakes up the 10ms time is expired then we don't
check if operation is finished or not and return failure.
This patch modifies poll logic to check HW operation after sleep so
that the status is checked atleast twice.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Mon, 2 Mar 2020 07:19:27 +0000 (12:49 +0530)]
octeontx2-af: Enable PCI master
Bus mastering is enabled by firmware, but when this driver
is unbinded bus mastering gets disabled by the PCI subsystem
which results interrupts not working when driver is reloaded.
Hence set bus mastering everytime in probe().
Also
- Converted pci_set_dma_mask() and pci_set_consistent_dma_mask()
to dma_set_mask_and_coherent().
- Cleared transaction pending bit which gets set during
driver unbind due to clearing of bus mastering (ME bit).
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Sunil Goutham [Mon, 2 Mar 2020 07:19:26 +0000 (12:49 +0530)]
octeontx2-af: Set discovery ID for RVUM block
Currently there is no way for AF dependent drivers in
any domain to check if the AF driver is loaded. This
patch sets an ID for RVUM block which will automatically
reflects in PF/VFs discovery register which they can
check and defer their probe until AF is up.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Linu Cherian [Mon, 2 Mar 2020 07:19:25 +0000 (12:49 +0530)]
octeontx2-af: Optimize data retrieval from firmware
For retrieving info like interface MAC addresses, packet
parser key extraction config etc currently a command
is sent to firmware and firmware which periodically polls
for commands, processes these and returns the info.
This is resulting in interface initialization taking lot
of time. To optimize this a memory region is shared between
firmware and this driver, firmware while booting puts
static info like these into that region for driver to
read directly without using commands.
With this
- Logic for retrieving packet parser extraction config
via commands is removed and repalced with using the
shared 'fwdata' structure.
- Now RVU MSIX vector address is also retrieved from this fwdata struct
instead of from CSR. Otherwise when kexec/kdump crash kernel loads
CSR will have a IOVA setup by primary kernel which impacts
RVU PF/VF's interrupts.
- Also added a mbox handler for PF/VF interfaces to retrieve their MAC
addresses from AF.
Signed-off-by: Linu Cherian <lcherian@marvell.com>
Signed-off-by: Christina Jacob <cjacob@marvell.com>
Signed-off-by: Rakesh Babu <rsaladi2@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Mon, 2 Mar 2020 07:19:24 +0000 (12:49 +0530)]
octeontx2-pf: Support to enable/disable pause frames via ethtool
Added mailbox requests to retrieve backpressure IDs from AF and Aura,
CQ contexts are configured with these BPIDs. So that when resource
levels reach configured thresholds they assert backpressure on the
interface which is also mapped to same BPID.
Also added support to enable/disable pause frames generation via ethtool.
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Mon, 2 Mar 2020 07:19:23 +0000 (12:49 +0530)]
octeontx2-af: Pause frame configuration at cgx
CGX LMAC, the physical interface can generate pause frames when
internal resources asserts backpressure due to exhaustion.
This patch configures CGX to generate 802.3 pause frames.
Also enabled processing of received pause frames on the line which
will assert backpressure on the internal transmit path.
Also added mailbox handlers for PF drivers to enable or disable
pause frames anytime.
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Geetha sowjanya [Mon, 2 Mar 2020 07:19:22 +0000 (12:49 +0530)]
octeontx2-af: Interface backpressure configuration
Each of the interface receive channels can be backpressured by
resources upon exhaustion or reaching configured threshold levels.
Resources here are receive buffer queues (Auras) and pkt notification
descriptor queues (CQs). Resources and interface channels are mapped
using backpressure IDs (BPIDs).
HW supports upto 512 BPIDs, this patch divides these BPIDs statically
across CGX/LBK/SDP interfaces as follows.
BPIDs 0 - 191 are mapped to LMAC channels, 16 per LMAC.
BPIDs 192 - 255 are mapped to LBK channels.
BPIDs 256 - 511 are mapped to SDP channels.
Also did the needed basic configuration of BPIDs.
Added mbox handlers with which a PF device can request for a BPID which
it will use to configure Auras and CQs.
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Heiner Kallweit [Sun, 1 Mar 2020 20:57:08 +0000 (21:57 +0100)]
net: phy: mscc: add constants for used interrupt mask bits
Add constants for the used interrupts bits. This avoids the magic
number for MII_VSC85XX_INT_MASK_MASK.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 01:07:01 +0000 (19:07 -0600)]
arcnet: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 01:05:02 +0000 (19:05 -0600)]
neighbour: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 01:02:54 +0000 (19:02 -0600)]
net: flow_offload: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 01:00:40 +0000 (19:00 -0600)]
net: dn_fib: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:53:11 +0000 (18:53 -0600)]
ndisc: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:51:41 +0000 (18:51 -0600)]
net: ipv6: mld: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:50:23 +0000 (18:50 -0600)]
net: lwtunnel: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:44:10 +0000 (18:44 -0600)]
net: ip6_route: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:14:11 +0000 (18:14 -0600)]
net: nexthop: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:13:05 +0000 (18:13 -0600)]
net: sctp: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Sat, 29 Feb 2020 00:11:02 +0000 (18:11 -0600)]
net: sock_reuseport: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 1 Mar 2020 05:48:55 +0000 (21:48 -0800)]
Merge branch 'net-ethtool-Introduce-link_ksettings-API-for-virtual-network-devices'
Cris Forno says:
====================
net/ethtool: Introduce link_ksettings API for virtual network devices
This series provides an API for drivers of virtual network devices that
allows users to alter initial device speed and duplex settings to reflect
the actual capabilities of underlying hardware. The changes made include
a helper function ethtool_virtdev_set_link_ksettings, which is used to
retrieve alterable link settings. In addition, there is a new ethtool
function defined to validate those settings. These changes resolve code
duplication for existing virtual network drivers that have already
implemented this behavior. In the case of the ibmveth driver, this API is
used to provide this capability for the first time.
---
v7: - removed ethtool_validate_cmd function pointer parameter from
ethtool_virtdev_set_link_ksettings since none of the virtual drivers
pass in a custom validate function as suggested by Michal Kubecek.
v6: - removed netvsc_validate_ethtool_ss_cmd(). netvsc_drv now uses
ethtool_virtdev_validate_cmd() instead as suggested by Michal Kubecek
and approved by Haiyang Zhang.
- matched handler argument name of ethtool_virtdev_set_link_ksettings
in declaration and definition as suggested by Michal Kubecek.
- shortened validate variable assignment in
ethtool_virtdev_set_link_ksettings as suggested by Michal Kubecek.
v5: - virtdev_validate_link_ksettings is taken out of the ethtool global
structure and is instead added as an argument to
ethtool_virtdev_set_link_ksettings as suggested by Jakub Kicinski.
v4: - Cleaned up return statement in ethtool_virtdev_validate_cmd based
off of Michal Kubecek's and Thomas Falcon's suggestion.
- If the netvsc driver is using the VF device in order to get
accelerated networking, the real speed and duplex is reported by using
the VF device as suggested by Stephen Hemminger.
- The speed and duplex variables are now passed by value rather than
passed by pointer as suggested by Willem de Bruijin and Michal
Kubecek.
- Removed ethtool_virtdev_get_link_ksettings since it was too simple
to warrant a helper function.
v3: - Factored out duplicated code to core/ethtool to provide API to
virtual drivers
v2: - Updated default driver speed/duplex settings to avoid breaking
existing setups
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Cris Forno [Fri, 28 Feb 2020 20:12:05 +0000 (14:12 -0600)]
net/ethtool: Introduce link_ksettings API for virtual network devices
With the ethtool_virtdev_set_link_ksettings function in core/ethtool.c,
ibmveth, netvsc, and virtio now use the core's helper function.
Funtionality changes that pertain to ibmveth driver include:
1. Changed the initial hardcoded link speed to 1GB.
2. Added support for allowing a user to change the reported link
speed via ethtool.
Functionality changes to the netvsc driver include:
1. When netvsc_get_link_ksettings is called, it will defer to the VF
device if it exists to pull accelerated networking values, otherwise
pull default or user-defined values.
2. Similarly, if netvsc_set_link_ksettings called and a VF device
exists, the real values of speed and duplex are changed.
Signed-off-by: Cris Forno <cforno12@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cris Forno [Fri, 28 Feb 2020 20:12:04 +0000 (14:12 -0600)]
ethtool: Factored out similar ethtool link settings for virtual devices to core
Three virtual devices (ibmveth, virtio_net, and netvsc) all have
similar code to set link settings and validate ethtool command. To
eliminate duplication of code, it is factored out into core/ethtool.c.
Signed-off-by: Cris Forno <cforno12@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 1 Mar 2020 05:37:03 +0000 (21:37 -0800)]
Merge branch 'hsr-several-code-cleanup-for-hsr-module'
Taehee Yoo says:
====================
hsr: several code cleanup for hsr module
This patchset is to clean up hsr module code.
1. The first patch is to use debugfs_remove_recursive().
If it uses debugfs_remove_recursive() instead of debugfs_remove(),
hsr_priv() doesn't need to have "node_tbl_file" pointer variable.
2. The second patch is to use extack error message.
If HSR uses the extack instead of netdev_info(), users can get
error messages immediately without any checking the kernel message.
3. The third patch is to use netdev_err() instead of WARN_ONCE().
When a packet is being sent, hsr_addr_subst_dest() is called and
it tries to find the node with the ethernet destination address.
If it couldn't find a node, it warns with WARN_ONCE().
But, using WARN_ONCE() is a little bit overdoing.
So, in this patch, netdev_err() is used instead.
4. The fourth patch is to remove unnecessary rcu_read_{lock/unlock}().
There are some rcu_read_{lock/unlock}() in hsr module and some of
them are unnecessary. In this patch,
these unnecessary rcu_read_{lock/unlock}() will be removed.
5. The fifth patch is to use upper/lower device infrastructure.
netdev_upper_dev_link() is useful to manage lower/upper interfaces.
And this function internally validates looping, maximum depth.
If hsr module uses upper/lower device infrastructure,
it can prevent these above problems.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Fri, 28 Feb 2020 18:02:10 +0000 (18:02 +0000)]
hsr: use upper/lower device infrastructure
netdev_upper_dev_link() is useful to manage lower/upper interfaces.
And this function internally validates looping, maximum depth.
All or most virtual interfaces that could have a real interface
(e.g. macsec, macvlan, ipvlan etc.) use lower/upper infrastructure.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Fri, 28 Feb 2020 18:01:56 +0000 (18:01 +0000)]
hsr: remove unnecessary rcu_read_lock() in hsr module
In order to access the port list, the hsr_port_get_hsr() is used.
And this is protected by RTNL and RCU.
The hsr_fill_info(), hsr_check_carrier(), hsr_dev_open() and
hsr_get_max_mtu() are protected by RTNL.
So, rcu_read_lock() in these functions are not necessary.
The hsr_handle_frame() also uses rcu_read_lock() but this function
is called by packet path.
It's already protected by RCU.
So, the rcu_read_lock() in hsr_handle_frame() can be removed.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Fri, 28 Feb 2020 18:01:46 +0000 (18:01 +0000)]
hsr: use netdev_err() instead of WARN_ONCE()
When HSR interface is sending a frame, it finds a node with
the destination ethernet address from the list.
If there is no node, it calls WARN_ONCE().
But, using WARN_ONCE() for this situation is a little bit overdoing.
So, in this patch, the netdev_err() is used instead.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Fri, 28 Feb 2020 18:01:35 +0000 (18:01 +0000)]
hsr: use extack error message instead of netdev_info
If HSR uses the extack instead of netdev_info(), users can get
error messages immediately without any checking the kernel message.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Taehee Yoo [Fri, 28 Feb 2020 18:01:20 +0000 (18:01 +0000)]
hsr: use debugfs_remove_recursive() instead of debugfs_remove()
If it uses debugfs_remove_recursive() instead of debugfs_remove(),
hsr_priv() doesn't need to have "node_tbl_file" pointer variable.
Signed-off-by: Taehee Yoo <ap420073@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Oleksij Rempel [Fri, 28 Feb 2020 14:50:49 +0000 (15:50 +0100)]
net: ag71xx: port to phylink
The port to phylink was done as close as possible to initial
functionality.
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Acked-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sun, 1 Mar 2020 05:30:43 +0000 (21:30 -0800)]
Merge branch 'net-ll_temac-RX-TX-ring-size-and-coalesce-ethtool-parameters'
Esben Haabendal says:
====================
net: ll_temac: RX/TX ring size and coalesce ethtool parameters
This series adds support for RX/TX ring size and irq coalesce ethtool
parameters to ll_temac driver.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Esben Haabendal [Fri, 28 Feb 2020 07:57:41 +0000 (08:57 +0100)]
net: ll_temac: Add ethtool support for coalesce parameters
Please note that the delays are calculated based on typical
parameters. But as TEMAC is an HDL IP, designs may vary, and future
work might be needed to make this calculation configurable.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Esben Haabendal [Fri, 28 Feb 2020 07:57:26 +0000 (08:57 +0100)]
net: ll_temac: Make RX/TX ring sizes configurable
Add support for setting the RX and TX ring sizes for this driver using
ethtool. Also increase the default RX ring size as the previous default
was far too low for good performance in some configurations.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Esben Haabendal [Fri, 28 Feb 2020 07:57:12 +0000 (08:57 +0100)]
net: ll_temac: Remove unused start_p variable
The start_p variable was included in the initial commit,
commit
92744989533c ("net: add Xilinx ll_temac device driver"),
but has never had any real use.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Esben Haabendal [Fri, 28 Feb 2020 07:56:57 +0000 (08:56 +0100)]
net: ll_temac: Remove unused tx_bd_next struct field
The tx_bd_next field was included in the initial commit,
commit
92744989533c ("net: add Xilinx ll_temac device driver"),
but has never had any real use.
Signed-off-by: Esben Haabendal <esben@geanix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Thu, 27 Feb 2020 20:58:44 +0000 (14:58 -0600)]
net: sched: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Sat, 29 Feb 2020 23:53:35 +0000 (15:53 -0800)]
Merge git://git./linux/kernel/git/bpf/bpf-next
Alexei Starovoitov says:
====================
pull-request: bpf-next 2020-02-28
The following pull-request contains BPF updates for your *net-next* tree.
We've added 41 non-merge commits during the last 7 day(s) which contain
a total of 49 files changed, 1383 insertions(+), 499 deletions(-).
The main changes are:
1) BPF and Real-Time nicely co-exist.
2) bpftool feature improvements.
3) retrieve bpf_sk_storage via INET_DIAG.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 28 Feb 2020 20:13:02 +0000 (12:13 -0800)]
Merge branch 'net-cleanup-datagram-receive-helpers'
Paolo Abeni says:
====================
net: cleanup datagram receive helpers
Several receive helpers have an optional destructor argument, which uglify
the code a bit and is taxed by retpoline overhead.
This series refactor the code so that we can drop such optional argument,
cleaning the helpers a bit and avoiding an indirect call in fast path.
The first patch refactor a bit the caller, so that the second patch
actually dropping the argument is more straight-forward
v1 -> v2:
- call scm_stat_del() only when not peeking - Kirill
- fix build issue with CONFIG_INET_ESPINTCP
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Paolo Abeni [Fri, 28 Feb 2020 13:45:22 +0000 (14:45 +0100)]
net: datagram: drop 'destructor' argument from several helpers
The only users for such argument are the UDP protocol and the UNIX
socket family. We can safely reclaim the accounted memory directly
from the UDP code and, after the previous patch, we can do scm
stats accounting outside the datagram helpers.
Overall this cleans up a bit some datagram-related helpers, and
avoids an indirect call per packet in the UDP receive path.
v1 -> v2:
- call scm_stat_del() only when not peeking - Kirill
- fix build issue with CONFIG_INET_ESPINTCP
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Paolo Abeni [Fri, 28 Feb 2020 13:45:21 +0000 (14:45 +0100)]
unix: uses an atomic type for scm files accounting
So the scm_stat_{add,del} helper can be invoked with no
additional lock held.
This clean-up the code a bit and will make the next
patch easier.
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 14:01:43 +0000 (08:01 -0600)]
af_unix: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:56:29 +0000 (07:56 -0600)]
bonding: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:43:24 +0000 (07:43 -0600)]
net: core: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:36:41 +0000 (07:36 -0600)]
ipv6: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:33:37 +0000 (07:33 -0600)]
net: dccp: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:30:45 +0000 (07:30 -0600)]
l2tp: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
Lastly, fix the following checkpatch warning:
CHECK: Prefer kernel type 'u8' over 'uint8_t'
#50: FILE: net/l2tp/l2tp_core.h:119:
+ uint8_t priv[]; /* private data */
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:26:15 +0000 (07:26 -0600)]
net: mpls: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Gustavo A. R. Silva [Fri, 28 Feb 2020 13:19:07 +0000 (07:19 -0600)]
xdp: Replace zero-length array with flexible-array member
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit
76497732932f ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Acked-by: Jonathan Lemon <jonathan.lemon@gmail.com>
Acked-by: Björn Töpel <bjorn.topel@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Fri, 28 Feb 2020 19:59:53 +0000 (11:59 -0800)]
Merge tag 'mlx5-updates-2020-02-27' of git://git./linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-updates-2020-02-27
mlx5 misc updates and minor cleanups:
1) Use per vport tables for mirroring
2) Improve log messages for SW steering (DR)
3) Add devlink fdb_large_groups parameter
4) E-Switch, Allow goto earlier chain
5) Don't allow forwarding between uplink representors
6) Add support for devlink-port in non-representors mode
7) Minor misc cleanups
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Alexei Starovoitov [Fri, 28 Feb 2020 02:50:20 +0000 (18:50 -0800)]
Merge branch 'bpf_sk_storage_via_inet_diag'
Martin KaFai Lau says:
====================
The bpf_prog can store specific info to a sk by using bpf_sk_storage.
In other words, a sk can be extended by a bpf_prog.
This series is to support providing bpf_sk_storage data during inet_diag's
dump. The primary target is the usage like iproute2's "ss".
The first two patches are refactoring works in inet_diag to make
adding bpf_sk_storage support easier. The next two patches do
the actual work.
Please see individual patch for details.
v2:
- Add commit message for u16 to u32 change in min_dump_alloc in Patch 4 (Song)
- Add comment to explain the !skb->len check in __inet_diag_dump in Patch 4.
- Do the map->map_type check earlier in Patch 3 for readability.
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Martin KaFai Lau [Tue, 25 Feb 2020 23:04:27 +0000 (15:04 -0800)]
bpf: inet_diag: Dump bpf_sk_storages in inet_diag_dump()
This patch will dump out the bpf_sk_storages of a sk
if the request has the INET_DIAG_REQ_SK_BPF_STORAGES nlattr.
An array of SK_DIAG_BPF_STORAGE_REQ_MAP_FD can be specified in
INET_DIAG_REQ_SK_BPF_STORAGES to select which bpf_sk_storage to dump.
If no map_fd is specified, all bpf_sk_storages of a sk will be dumped.
bpf_sk_storages can be added to the system at runtime. It is difficult
to find a proper static value for cb->min_dump_alloc.
This patch learns the nlattr size required to dump the bpf_sk_storages
of a sk. If it happens to be the very first nlmsg of a dump and it
cannot fit the needed bpf_sk_storages, it will try to expand the
skb by "pskb_expand_head()".
Instead of expanding it in inet_sk_diag_fill(), it is expanded at a
sleepable context in __inet_diag_dump() so __GFP_DIRECT_RECLAIM can
be used. In __inet_diag_dump(), it will retry as long as the
skb is empty and the cb->min_dump_alloc becomes larger than before.
cb->min_dump_alloc is bounded by KMALLOC_MAX_SIZE. The min_dump_alloc
is also changed from 'u16' to 'u32' to accommodate a sk that may have
a few large bpf_sk_storages.
The updated cb->min_dump_alloc will also be used to allocate the skb in
the next dump. This logic already exists in netlink_dump().
Here is the sample output of a locally modified 'ss' and it could be made
more readable by using BTF later:
[root@arch-fb-vm1 ~]# ss --bpf-map-id 14 --bpf-map-id 13 -t6an 'dst [::1]:8989'
State Recv-Q Send-Q Local Address:Port Peer Address:PortProcess
ESTAB 0 0 [::1]:51072 [::1]:8989
bpf_map_id:14 value:[ 3feb ]
bpf_map_id:13 value:[ 3f ]
ESTAB 0 0 [::1]:51070 [::1]:8989
bpf_map_id:14 value:[ 3feb ]
bpf_map_id:13 value:[ 3f ]
[root@arch-fb-vm1 ~]# ~/devshare/github/iproute2/misc/ss --bpf-maps -t6an 'dst [::1]:8989'
State Recv-Q Send-Q Local Address:Port Peer Address:Port Process
ESTAB 0 0 [::1]:51072 [::1]:8989
bpf_map_id:14 value:[ 3feb ]
bpf_map_id:13 value:[ 3f ]
bpf_map_id:12 value:[ 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000... total:65407 ]
ESTAB 0 0 [::1]:51070 [::1]:8989
bpf_map_id:14 value:[ 3feb ]
bpf_map_id:13 value:[ 3f ]
bpf_map_id:12 value:[ 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000... total:65407 ]
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200225230427.1976129-1-kafai@fb.com
Martin KaFai Lau [Tue, 25 Feb 2020 23:04:21 +0000 (15:04 -0800)]
bpf: INET_DIAG support in bpf_sk_storage
This patch adds INET_DIAG support to bpf_sk_storage.
1. Although this series adds bpf_sk_storage diag capability to inet sk,
bpf_sk_storage is in general applicable to all fullsock. Hence, the
bpf_sk_storage logic will operate on SK_DIAG_* nlattr. The caller
will pass in its specific nesting nlattr (e.g. INET_DIAG_*) as
the argument.
2. The request will be like:
INET_DIAG_REQ_SK_BPF_STORAGES (nla_nest) (defined in latter patch)
SK_DIAG_BPF_STORAGE_REQ_MAP_FD (nla_put_u32)
SK_DIAG_BPF_STORAGE_REQ_MAP_FD (nla_put_u32)
......
Considering there could have multiple bpf_sk_storages in a sk,
instead of reusing INET_DIAG_INFO ("ss -i"), the user can select
some specific bpf_sk_storage to dump by specifying an array of
SK_DIAG_BPF_STORAGE_REQ_MAP_FD.
If no SK_DIAG_BPF_STORAGE_REQ_MAP_FD is specified (i.e. an empty
INET_DIAG_REQ_SK_BPF_STORAGES), it will dump all bpf_sk_storages
of a sk.
3. The reply will be like:
INET_DIAG_BPF_SK_STORAGES (nla_nest) (defined in latter patch)
SK_DIAG_BPF_STORAGE (nla_nest)
SK_DIAG_BPF_STORAGE_MAP_ID (nla_put_u32)
SK_DIAG_BPF_STORAGE_MAP_VALUE (nla_reserve_64bit)
SK_DIAG_BPF_STORAGE (nla_nest)
SK_DIAG_BPF_STORAGE_MAP_ID (nla_put_u32)
SK_DIAG_BPF_STORAGE_MAP_VALUE (nla_reserve_64bit)
......
4. Unlike other INET_DIAG info of a sk which is pretty static, the size
required to dump the bpf_sk_storage(s) of a sk is dynamic as the
system adding more bpf_sk_storage_map. It is hard to set a static
min_dump_alloc size.
Hence, this series learns it at the runtime and adjust the
cb->min_dump_alloc as it iterates all sk(s) of a system. The
"unsigned int *res_diag_size" in bpf_sk_storage_diag_put()
is for this purpose.
The next patch will update the cb->min_dump_alloc as it
iterates the sk(s).
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200225230421.1975729-1-kafai@fb.com
Martin KaFai Lau [Tue, 25 Feb 2020 23:04:15 +0000 (15:04 -0800)]
inet_diag: Move the INET_DIAG_REQ_BYTECODE nlattr to cb->data
The INET_DIAG_REQ_BYTECODE nlattr is currently re-found every time when
the "dump()" is re-started.
In a latter patch, it will also need to parse the new
INET_DIAG_REQ_SK_BPF_STORAGES nlattr to learn the map_fds. Thus, this
patch takes this chance to store the parsed nlattr in cb->data
during the "start" time of a dump.
By doing this, the "bc" argument also becomes unnecessary
and is removed. Also, the two copies of the INET_DIAG_REQ_BYTECODE
parsing-audit logic between compat/current version can be
consolidated to one.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200225230415.1975555-1-kafai@fb.com
Martin KaFai Lau [Tue, 25 Feb 2020 23:04:09 +0000 (15:04 -0800)]
inet_diag: Refactor inet_sk_diag_fill(), dump(), and dump_one()
In a latter patch, there is a need to update "cb->min_dump_alloc"
in inet_sk_diag_fill() as it learns the diffierent bpf_sk_storages
stored in a sk while dumping all sk(s) (e.g. tcp_hashinfo).
The inet_sk_diag_fill() currently does not take the "cb" as an argument.
One of the reason is inet_sk_diag_fill() is used by both dump_one()
and dump() (which belong to the "struct inet_diag_handler". The dump_one()
interface does not pass the "cb" along.
This patch is to make dump_one() pass a "cb". The "cb" is created in
inet_diag_cmd_exact(). The "nlh" and "in_skb" are stored in "cb" as
the dump() interface does. The total number of args in
inet_sk_diag_fill() is also cut from 10 to 7 and
that helps many callers to pass fewer args.
In particular,
"struct user_namespace *user_ns", "u32 pid", and "u32 seq"
can be replaced by accessing "cb->nlh" and "cb->skb".
A similar argument reduction is also made to
inet_twsk_diag_fill() and inet_req_diag_fill().
inet_csk_diag_dump() and inet_csk_diag_fill() are also removed.
They are mostly equivalent to inet_sk_diag_fill(). Their repeated
usages are very limited. Thus, inet_sk_diag_fill() is directly used
in those occasions.
Signed-off-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200225230409.1975173-1-kafai@fb.com
David S. Miller [Fri, 28 Feb 2020 02:31:39 +0000 (18:31 -0800)]
Merge git://git./linux/kernel/git/netdev/net
The mptcp conflict was overlapping additions.
The SMC conflict was an additional and removal happening at the same
time.
Signed-off-by: David S. Miller <davem@davemloft.net>
Roi Dayan [Thu, 13 Feb 2020 12:19:50 +0000 (14:19 +0200)]
net/mlx5e: Remove redundant comment about goto slow path
The code is self explanatory and makes the comment redundant.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Eli Cohen [Thu, 13 Feb 2020 09:12:16 +0000 (11:12 +0200)]
net/mlx5e: Reduce number of arguments in slow path handling
mlx5e_tc_offload_to_slow_path() and mlx5e_tc_unoffload_from_slow_path()
take an extra argument allocated on the stack of the caller but not used
by the caller. Avoid the extra argument and use local variable in the
function itself.
Signed-off-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Eli Cohen [Thu, 13 Feb 2020 13:18:51 +0000 (15:18 +0200)]
net/mlx5e: Remove unused argument from parse_tc_pedit_action()
parse_attr is not used by parse_tc_pedit_action() so revmove it.
Signed-off-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Roi Dayan [Tue, 18 Feb 2020 13:30:58 +0000 (15:30 +0200)]
net/mlx5e: Use NL_SET_ERR_MSG_MOD() extack for errors
This to be consistent and adds the module name to the error message.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Roi Dayan [Tue, 18 Feb 2020 13:24:39 +0000 (15:24 +0200)]
net/mlx5e: Use netdev_warn() instead of pr_err() for errors
This is for added netdev prefix that helps identify
the source of the message.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Roi Dayan [Wed, 25 Dec 2019 15:03:35 +0000 (17:03 +0200)]
net/mlx5e: Use netdev_warn() for errors for added prefix
This helps identify the source of the message.
If netdev still doesn't exists use mlx5_core_warn().
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Erez Shitrit [Sun, 12 Jan 2020 08:55:54 +0000 (10:55 +0200)]
net/mlx5: DR, Improve log messages
Few print messages are in debug level where they should be in error, and
few messages are missing.
Signed-off-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Hamdan Igbaria [Mon, 17 Feb 2020 13:53:20 +0000 (15:53 +0200)]
net/mlx5: DR, Change matcher priority parameter type
Change matcher priority parameter type from u16 to u32,
this change is needed since sometimes upper levels
create a matcher with priority bigger than 2^16.
Signed-off-by: Hamdan Igbaria <hamdani@mellanox.com>
Reviewed-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Jianbo Liu [Fri, 27 Dec 2019 06:37:07 +0000 (06:37 +0000)]
net/mlx5e: Add devlink fdb_large_groups parameter
Add a devlink parameter to control the number of large groups in a
autogrouped flow table. The default value is 15, and the range is between 1
and 1024.
The size of each large group can be calculated according to the following
formula: size = 4M / (fdb_large_groups + 1).
Examples:
- Set the number of large groups to 20.
$ devlink dev param set pci/0000:82:00.0 name fdb_large_groups \
cmode driverinit value 20
Then run devlink reload command to apply the new value.
$ devlink dev reload pci/0000:82:00.0
- Read the number of large groups in flow table.
$ devlink dev param show pci/0000:82:00.0 name fdb_large_groups
pci/0000:82:00.0:
name fdb_large_groups type driver-specific
values:
cmode driverinit value 20
Signed-off-by: Jianbo Liu <jianbol@mellanox.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Reviewed-by: Roi Dayan <roid@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Jianbo Liu [Tue, 7 Jan 2020 08:48:05 +0000 (08:48 +0000)]
net/mlx5: Change the name of steering mode param id
The prefix should be "MLX5_DEVLINK_PARAM_ID_" for all in
mlx5_devlink_param_id enum.
Signed-off-by: Jianbo Liu <jianbol@mellanox.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Vladyslav Tarasiuk [Wed, 13 Nov 2019 15:19:47 +0000 (17:19 +0200)]
net/mlx5e: Add support for devlink-port in non-representors mode
Added devlink_port field to mlx5e_priv structure and a callback to
netdev ops to enable devlink to get info about the port. The port
registration happens at driver initialization.
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Vladyslav Tarasiuk [Fri, 17 Jan 2020 10:42:53 +0000 (12:42 +0200)]
net/mlx5e: Rename representor get devlink port function
Rename representor's mlx5e_get_devlink_port() to
mlx5e_rep_get_devlink_port().
The downstream patch will add a non-representor mlx5e function called
mlx5e_get_devlink_phy_port().
Signed-off-by: Vladyslav Tarasiuk <vladyslavt@mellanox.com>
Reviewed-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Roi Dayan [Thu, 6 Feb 2020 14:06:58 +0000 (16:06 +0200)]
net/mlx5: E-Switch, Allow goto earlier chain if FW supports it
Mellanox FW can support this if ignore_flow_level capability exists.
Signed-off-by: Roi Dayan <roid@mellanox.com>
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Eli Cohen [Tue, 14 Jan 2020 15:30:41 +0000 (17:30 +0200)]
net/mlx5e: Eswitch, Use per vport tables for mirroring
When using port mirroring, we forward the traffic to another table and
use that table to forward to the mirrored vport. Since the hardware
loses the values of reg c, and in particular reg c0, we fail the match
on the input vport which previously existed in reg c0. To overcome this
situation, we use a set of per vport tables, positioned at the lowest
priority, and forward traffic to those tables. Since these tables are
per vport, we can avoid matching on reg c0.
Fixes: c01cfd0f1115 ("net/mlx5: E-Switch, Add match on vport metadata for rule in fast path")
Signed-off-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Reviewed-by: Paul Blakey <paulb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>