Re: [net-next RESEND v2] net: core: use shared sysctl macro

[Luis Chamberlain <mcgrof@kernel.org>]

On Wed, Apr 06, 2022 at 12:16:11PM -0700, Jakub Kicinski wrote:
> On Wed, 6 Apr 2022 09:20:24 -0700 Luis Chamberlain wrote:
> > On Wed, Apr 06, 2022 at 08:42:08PM +0800, xiangxia.m.yue@gmail.com wrote:
> > > From: Tonghao Zhang <xiangxia.m.yue@gmail.com>
> > > 
> > > This patch introdues the SYSCTL_THREE, and replace the
> > > two, three and long_one to SYSCTL_XXX accordingly.
> > > 
> > >  KUnit:
> > >  [23:03:58] ================ sysctl_test (10 subtests) =================
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_null_tbl_data
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_table_maxlen_unset
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_table_len_is_zero
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_table_read_but_position_set
> > >  [23:03:58] [PASSED] sysctl_test_dointvec_read_happy_single_positive
> > >  [23:03:58] [PASSED] sysctl_test_dointvec_read_happy_single_negative
> > >  [23:03:58] [PASSED] sysctl_test_dointvec_write_happy_single_positive
> > >  [23:03:58] [PASSED] sysctl_test_dointvec_write_happy_single_negative
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_write_single_less_int_min
> > >  [23:03:58] [PASSED] sysctl_test_api_dointvec_write_single_greater_int_max
> > >  [23:03:58] =================== [PASSED] sysctl_test ===================
> > > 
> > >  ./run_kselftest.sh -c sysctl
> > >  ...
> > >  # Running test: sysctl_test_0006 - run #49
> > >  # Checking bitmap handler... ok
> > >  # Wed Mar 16 14:58:41 UTC 2022
> > >  # Running test: sysctl_test_0007 - run #0
> > >  # Boot param test only possible sysctl_test is built-in, not module:
> > >  # CONFIG_TEST_SYSCTL=m
> > >  ok 1 selftests: sysctl: sysctl.sh
> >
> > I can take this through sysctl-next [0] if folks are OK with that. There are
> > quite a bit of changes already queued there for sysctl.
> > 
> > Jakub?
> > 
> > [0] https://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux.git/log/?h=sysctl-next
> 
> sysctl-next makes a lot of sense, but I'm worried about conflicts.

I can try to deal with them as I can send the pull request to Linus towards
the end of the merge window.

> Would you be able to spin up a stable branch based on -rc1 so we
> can pull it into net-next as well?

Yes, absolutely. Just pushed, but I should note that linux-next already
takes in sysctl-next. And there are non-networking changes that are
in sysctl-next. Does net-next go to Linus or is it just to help with
developers so they get something more close to linux-next but not as
insane?

  Luis