On 24.02.2022 15:09, Roger Pau Monné wrote: > On Thu, Jan 13, 2022 at 02:17:18PM +0100, Jan Beulich wrote: >> Except in the "clocksource=tsc" case we can replace the indirect calls >> involved in accessing the platform timers by direct ones, as they get >> established once and never changed. To also cover the "tsc" case, invoke >> what read_tsc() resolves to directly. In turn read_tsc() then becomes >> unreachable and hence can move to .init.*. >> >> Signed-off-by: Jan Beulich <jbeul...@suse.com> > > Reviewed-by: Roger Pau Monné <roger....@citrix.com>
Thanks. >> --- >> TBD: Instead of adding __init to read_tsc() we could also ditch the >> function altogether, using a dedicated (non-canonical) pointer >> constant instead for the .read_counter initializer and the two >> comparisons done on plt_src.read_counter. > > I was going to suggest adding an ASSERT_UNREACHABLE, but not sure it > makes much sense if the function is init only. > > I would be fine with using a non-canonical pointer. I guess I'll put in the patch as is and do the conversion in a follow-up change adding __initconst_cf_clobber (once it's clear what to do about structures which are written to). Jan
