This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug target/29900] [4.1/4.2/4.3] stackreg to go wrongs



------- Comment #11 from hjl at lucon dot org  2006-11-20 15:05 -------
I think regstack is all screwed up:

Basic block 2
Input stack: empty
  insn 10 input stack: empty
  insn 11 input stack: [ 10 ]
  insn 57 input stack: [ 10 11 ]

(insn 11 71 19 2 (set (reg:SF 8 st)
        (mem/c/i:SF (plus:SI (reg/f:SI 6 bp)
                (const_int 28 [0x1c])) [5 YJ+0 S4 A32])) 89 {*movsf_1} (nil)
    (expr_list:REG_EQUIV (mem/c/i:SF (plus:SI (reg/f:SI 6 bp)
                (const_int 28 [0x1c])) [5 YJ+0 S4 A32])
        (nil)))

Insn 11 loads SF 8, not SF 10.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29900


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]