Re: loop optimizer naivety

[Eric Botcazou <ebotcazou at libertysurf dot fr>]

> I would speculate that after PRE/LCM was completed that insn 610
> referenced reg 185 rather than 173.  ie, the code looked something like
> this:
> [...]
> Then copy-prop replaced reg 185 with reg 173 in insn 610, which
> leads to the somewhat odd looking code in your message.

That's indeed what happens.

> If you could verify that sequence of events it would be helpful in at
> least determining that LCM is operating properly.  To do this you'd put
> a breakpoint just before this hunk of code in gcse.c and dump the RTL
> to a file:

The log file is sufficient I think:

PRE:  store updated with reaching reg (reg:SI 185):
	(insn 610 609 765 3 0x40a5565c (set (mem/f:SI (plus:SI (reg/f:SI 20 frame)
                                (const_int -100 [0xffffff9c])) [5 S4 A32])
                        (reg:SI 173)) 38 {*movsi_1} (nil)
                    (expr_list:REG_EQUAL (plus:SI (reg:SI 161)
                            (const_int 1 [0x1]))
                        (nil)))

LOCAL COPY-PROP: Replacing reg 185 in insn 610 with reg 173

-- 
Eric Botcazou