[bpf-next,v9,0/2] bpf: Allow bpf_skb_event_output for more prog types
mbox series

Message ID 20190710181811.127374-1-allanzhang@google.com
Headers show
Series
  • bpf: Allow bpf_skb_event_output for more prog types
Related show

Message

Allan Zhang July 10, 2019, 6:18 p.m. UTC
Software event output is only enabled by a few prog types right now (TC,
LWT out, XDP, sockops). Many other skb based prog types need
bpf_skb_event_output to produce software event.

More prog types are enabled to access bpf_skb_event_output in this
patch.

v9 changes:
add "Acked-by" field.

v8 changes:
No actual change, just cc to netdev@vger.kernel.org and
bpf@vger.kernel.org.
v7 patches are acked by Song Liu.

v7 changes:
Reformat from hints by scripts/checkpatch.pl, including Song's comment
on signed-off-by name to captical case in cover letter.
3 of hints are ignored:
1. new file mode.
2. SPDX-License-Identifier for event_output.c since all files under
   this dir have no such line.
3. "Macros ... enclosed in parentheses" for macro in event_output.c
   due to code's nature.

Change patch 02 subject "bpf:..." to "selftests/bpf:..."

v6 changes:
Fix Signed-off-by, fix fixup map creation.

v5 changes:
Fix typos, reformat comments in event_output.c, move revision history to
cover letter.

v4 changes:
Reformating log message.

v3 changes:
Reformating log message.

v2 changes:
Reformating log message.

Allan Zhang (2):
  bpf: Allow bpf_skb_event_output for a few prog types
  selftests/bpf: Add selftests for bpf_perf_event_output

 net/core/filter.c                             |  6 ++
 tools/testing/selftests/bpf/test_verifier.c   | 12 ++-
 .../selftests/bpf/verifier/event_output.c     | 94 +++++++++++++++++++
 3 files changed, 111 insertions(+), 1 deletion(-)
 create mode 100644 tools/testing/selftests/bpf/verifier/event_output.c

Comments

Andrii Nakryiko July 10, 2019, 7:05 p.m. UTC | #1
On Wed, Jul 10, 2019 at 11:18 AM Allan Zhang <allanzhang@google.com> wrote:
>
> Software event output is only enabled by a few prog types right now (TC,
> LWT out, XDP, sockops). Many other skb based prog types need
> bpf_skb_event_output to produce software event.
>
> More prog types are enabled to access bpf_skb_event_output in this
> patch.
>
> v9 changes:
> add "Acked-by" field.

Thanks! This looks good to me. Just FYI. Not sure if you followed, but
bpf-next is closed, so this can't go in until it's open. Maintainers
might ask you to resubmit at that time, if patches don't apply
cleanly.

>
> v8 changes:
> No actual change, just cc to netdev@vger.kernel.org and
> bpf@vger.kernel.org.
> v7 patches are acked by Song Liu.
>
> v7 changes:
> Reformat from hints by scripts/checkpatch.pl, including Song's comment
> on signed-off-by name to captical case in cover letter.
> 3 of hints are ignored:
> 1. new file mode.
> 2. SPDX-License-Identifier for event_output.c since all files under
>    this dir have no such line.
> 3. "Macros ... enclosed in parentheses" for macro in event_output.c
>    due to code's nature.
>
> Change patch 02 subject "bpf:..." to "selftests/bpf:..."
>
> v6 changes:
> Fix Signed-off-by, fix fixup map creation.
>
> v5 changes:
> Fix typos, reformat comments in event_output.c, move revision history to
> cover letter.
>
> v4 changes:
> Reformating log message.
>
> v3 changes:
> Reformating log message.
>
> v2 changes:
> Reformating log message.
>
> Allan Zhang (2):
>   bpf: Allow bpf_skb_event_output for a few prog types
>   selftests/bpf: Add selftests for bpf_perf_event_output
>
>  net/core/filter.c                             |  6 ++
>  tools/testing/selftests/bpf/test_verifier.c   | 12 ++-
>  .../selftests/bpf/verifier/event_output.c     | 94 +++++++++++++++++++
>  3 files changed, 111 insertions(+), 1 deletion(-)
>  create mode 100644 tools/testing/selftests/bpf/verifier/event_output.c
>
> --
> 2.22.0.410.gd8fdbe21b5-goog
>
Daniel Borkmann July 12, 2019, 12:42 p.m. UTC | #2
On 07/10/2019 09:10 PM, Allan Zhang wrote:
> Sure, thanks. will do as suggested.

Yep, bpf-next is currently closed due to merge window, please resubmit
once it reopens.

Thanks,
Daniel