On 3 Jun 2026, at 22:08, Gleb Smirnoff wrote:
> On Wed, Jun 03, 2026 at 01:15:41PM +0000, Kristof Provost wrote:
> K> The branch main has been updated by kp:
> K>
> K> URL: 
> https://cgit.FreeBSD.org/src/commit/?id=017690e509134422dbfa80316eab56d8d3f87065
> K>
> K> commit 017690e509134422dbfa80316eab56d8d3f87065
> K> Author:     Ishan Agrawal <[email protected]>
> K> AuthorDate: 2026-06-01 12:24:05 +0000
> K> Commit:     Kristof Provost <[email protected]>
> K> CommitDate: 2026-06-03 08:52:06 +0000
> K>
> K>     netpfil: add PF netlink command decoding support
> K>
> K>     Convert PFNL_CMD values in pf_nl.h from an enum to #define constants,
> K>     add a pfnl_cmd table definition for mktable, and implement the
> K>     corresponding command decoding helpers in libsysdecode.
> K>
> K>     This allows mktable to generate PF netlink command lookup tables and
> K>     enables symbolic decoding of PF netlink commands.
> K>
> K>     Reviewed by:    kp
> K>     Signed-off-by:  Ishan Agrawal <[email protected]>
> K>     Sponsored by:   Google LLC (GSoC 2026)
>
> Sorry for not reviewing in time.  I got question: why does this
> require ditching the enum and regressing down to preprocessor
> defines?
>
The mktables script in libsysdecode expects to work on defines: 
https://github.com/freebsd/freebsd-src/blob/07d733963ecd1957999f5381b438b8c369686750/lib/libsysdecode/mktables#L76

It’s possible to change that, but this being an enum doesn’t really buy us much 
either, and this was the less invasive path.

Best regards,
Kristof

Reply via email to