In D49681#1146197, @lwhsu wrote:

There is also D50108 implementing RFC 8981, I think these 2 patches can be reviewed together, at least the part of getting an interface identifier, as suggested in RFC 8981 3.3.2.

@diizzy I've tested with the ports indicated in my previous message, both with and then without the patch.

In D49662#1138986, @diizzy wrote:

Given the "recent" breakage of various things I'd like at least to see a run with and without ccmake enabled withs ports that interracts with it already in various ways. `grep -ir CCACHE --include "Makefile" /usr/ports/` gives you an idea of what ports I'm referring to.

I'd like to commit this patch, since it is fixing a an issue (avoid -DCMAKE_C_COMPILER_LAUNCHER=ccache being added when defining NO_CCACHE), but am missing portmgr approval. Who could I ask for intervention?

Fix minor issues:

Do mark the use_stableaddr sysctl as tunable set.

@glebius Thanks for the review and the endorsement!

Use u_char for the prefixlen argument.

Fixed some whitespace issues I introduced by mistake.

Fixed issues when ipv6_privacy is also enabled.

There is an issue when stablesaddr and net.inet6.ip6.use_tempaddr are simultaneously enabled.

Updating patch to fix an error with an if statement content.

Created an updated patch:

In D49681#1135624, @zarychtam_plan-b.pwste.edu.pl wrote:

I’ve just updated to the latest version of the patch, and now I no longer see a stableaddr on the interface - only an address with an EUI-64 suffix is present. The initial version of the patch was working fine for an entire week.

It’s possible there’s a conflict with another patch I’m testing[1]. However, both patches were working together as expected last week, with no apparent interference between them.

1. https://bz-attachments.freebsd.org/attachment.cgi?id=252676

I'm evaluating updating the patch to make the number of retries configurable via a sysctl (like net.inet6.ip6.stableaddr_maxretries?), defaulting to 3, obviously.

Updating the patch, addressing reported feedback.

In D49681#1133961, @emaste wrote:

We should also enable by default - probably as a separate commit (which might not be MFCd if there are POLA concerns).

I addressed all reported feedback I hope.

In D49681#1133055, @madpilot wrote:

I just realized that when checking if the generated ID is valid I simply return failure, the stack will just keep generating the same invalid address over and over, without stopping and never getting a chance to generate a valid one.

So, my solution, would be to wrap the generating function in a do while cycle, exiting on a valid ID adding an offset to the the counter, so that at each iteration (if the previous was invalid), a valid one would be generated.

I just realized that when checking if the generated ID is valid I simply return failure, the stack will just keep generating the same invalid address over and over, without stopping and never getting a chance to generate a valid one.

Thanks for the feedback!

Forgot to mention: Many thanks to cognet who helped me with suggestions, especially about how to implement the dad_failures counter.

Sorry, I pushed a completely unrelated diff I'm still working on, I should have fixed it now.

remove invalid cmake option

Remove invalid cmake option.