[Bug rtl-optimization/23837] [4.0/4.1/4.2 regression] Wrong code with -fschedule-insns

["steven at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

------- Comment #23 from steven at gcc dot gnu dot org  2005-12-15 00:00 -------
lreg decides to do this:
;; Register 95 in 25.
;; Register 97 in 28.
;; Register 99 in 22.
;; Register 102 in 21.
;; Register 104 in 25.
;; Register 111 in 28.
;; Register 112 in 19.
;; Register 113 in 28.

and reload/greg agree (note the missing page breaks in the .greg dump btw):
;; Register dispositions:
95 in 25
97 in 28
99 in 22
102 in 21
104 in 25
111 in 28
112 in 19
113 in 28

So reg 95 and reg 104 are indeed assigned the same hard register.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23837