On Wed, Jul 26, 2017 at 1:34 PM, Ryota Ozaki <ozak...@netbsd.org> wrote: > (resending, please ignore if duplicated) > > Hi, > > I propose a debugging function of localcount, > localcount_debug_refcnt, which returns a total > reference count of a specified localcount. > > A result of localcount_debug_refcnt isn't guaranteed > to be accurate, however it's still useful for > debugging purposes, for example debugging on > ATF tests using rump kernels. > > Here is a patch: > http://www.netbsd.org/~ozaki-r/localcount_debug_refcnt.diff > > I added the following assertion in localcount_release: > KDASSERT(localcount_debug_refcnt(lc) >= 0); > > IIUC it's a proper constraint of localcount even if > a result of localcount_debug_refcnt isn't accurate. > > Any comments or suggestions?
Oops. I thought we could use percpu_foreach in softint, but I was wrong. So localcount_debug_refcnt can be used only in thread contexts. And the above assertion shouldn't be there. The proposal isn't so much worthwhile ;-< ozaki-r