pf: Standardize rtableid

[vegeta_tuxpowered.net]

Summary

Prepare for rtableid being included in struct pfsync_state where it will be int32_t. Make variables which will be set to and from it the same width.

Sponsored by: InnoGames GmbH

[melifaro]

The new kernel code uses uint32_t as the table id, I’m curious why do you want to have rtableid signef.

[vegeta_tuxpowered.net]

In D40013#911225, @melifaro wrote:

The new kernel code uses uint32_t as the table id, I’m curious why do you want to have rtableid signef.

The rtableid variable is signed in pf. You can see it being assigned value of -1 for the default rule in pf_ioctl.c and for each rule by pfctl unless a specific value is set. Then in pf_test_rule() and a few other places M_SETFIB(); is called only if the value is >= 0. That makes 0 a valid value which can be set and -1 the default meaning don't set rtableid.

[melifaro]

In D40013#911294, @vegeta_tuxpowered.net wrote:

In D40013#911225, @melifaro wrote:

The new kernel code uses uint32_t as the table id, I’m curious why do you want to have rtableid signef.

The rtableid variable is signed in pf. You can see it being assigned value of -1 for the default rule in pf_ioctl.c and for each rule by pfctl unless a specific value is set. Then in pf_test_rule() and a few other places M_SETFIB(); is called only if the value is >= 0. That makes 0 a valid value which can be set and -1 the default meaning don't set rtableid.

Ack, thank you gor clarification!