On Tue, Aug 4, 2015 at 12:29 PM, Segher Boessenkool <seg...@kernel.crashing.org> wrote: > On Tue, Aug 04, 2015 at 11:50:00AM -0700, H.J. Lu wrote: >> >> The motivation of __builtin_stack_top is that frame_address requires a >> >> frame pointer register, which isn't desirable for x86. >> >> __builtin_stack_top >> >> doesn't require a frame pointer register. >> > >> > If the target just returns frame_pointer_rtx from >> > INITIAL_FRAME_ADDRESS_RTX, >> > you don't get crtl->accesses_prior_frames set either, and as far as I can >> > see everything works fine? For __builtin_frame_address(0). >> > >> > You might have a reason why you want the entry stack address instead of the >> > frame address, but you didn't really explain I think? Or I missed it. >> > >> >> expand_builtin_return_addr sets >> >> crtl->accesses_prior_frames = 1; >> >> for __builtin_frame_address, which requires a frame pointer register. >> __builtin_stack_top doesn't set crtl->accesses_prior_frames and frame >> pointer register isn't required. > > Not if you have INITIAL_FRAME_ADDRESS_RTX. I don't see why the generic code > cannot just use frame_pointer_rtx (instead of hard_frame_pointer_rtx) for > a count of 0; but making it target-specific is certainly more conservative. > > You say i386 doesn't have that target macro defined currently. Yes I know; > so change that? Or change the generic code, but that is much more testing.
There is another issue with x86, maybe other targets. You can't get the real stack top when stack is realigned and -maccumulate-outgoing-args isn't used since ix86_expand_prologue will create and return another stack frame for __builtin_frame_address and __builtin_return_address. It will be wrong for __builtin_stack_top, which should return the real stack address. -- H.J.
