On Wed, 02 Jun 2021 16:05:18 +0530 "Naveen N. Rao" <naveen.n....@linux.vnet.ibm.com> wrote:
> It seems to be working fine otherwise. The one limitation though is down > to how ftrace works on powerpc -- the mcount call is before a function > sets up its own stackframe. Due to this, we won't ever be able to > account for the stackframe from a leaf function -- but, that's a fairly > minor limitation. And this is true for x86 as well because it no longer uses mcount, but uses fentry instead (called before stack setup), but I figured there's not much we could do about it. -- Steve