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 c++/16590] [3.4 regression] Incorrect execution when compiling with -O2


------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-08-17 23:02 -------
This is a GCSE bug.

In the 07.addressof dump, we have:

(insn 56 104 57 1 (parallel [
            (set (reg/f:SI 77)
                (plus:SI (reg/f:SI 20 frame)
                    (const_int -11 [0xfffffff5])))
            (clobber (reg:CC 17 flags))
        ]) 138 {*addsi_1} (nil)
    (nil))

(insn 57 56 58 1 (set (mem/s:SI (plus:SI (reg/f:SI 20 frame)
                (const_int -24 [0xffffffe8])) [4 p.p+0 S4 A32])
        (reg/f:SI 77)) 36 {*movsi_1} (nil)
    (nil))

(code_label 58 57 105 2 3 "" [1 uses])

(note 105 58 62 2 [bb 2] NOTE_INSN_BASIC_BLOCK)

(insn 62 105 63 2 (set (reg:SI 78 [ p.p ])
        (mem/s:SI (plus:SI (reg/f:SI 20 frame)
                (const_int -24 [0xffffffe8])) [4 p.p+0 S4 A32])) 36 {*movsi_1} (nil)
    (nil))

(insn 63 62 64 2 (parallel [
            (set (reg:SI 79)
                (plus:SI (reg:SI 78 [ p.p ])
                    (const_int -1 [0xffffffff])))
            (clobber (reg:CC 17 flags))
        ]) 138 {*addsi_1} (nil)
    (nil))

After GCSE, we have:

(insn 56 104 114 1 (parallel [
        (set (reg/f:SI 77)
                (plus:SI (reg/f:SI 20 frame)
                    (const_int -11 [0xfffffff5])))
            (clobber (reg:CC 17 flags))
        ]) 138 {*addsi_1} (nil)
    (nil))

(insn 114 56 58 1 (set (reg/f:SI 86 [ start ])
        (reg/f:SI 77)) -1 (nil)
    (nil))                                                                     

(code_label 58 114 105 2 3 "" [1 uses])

(note 105 58 112 2 [bb 2] NOTE_INSN_BASIC_BLOCK)

(insn 112 105 63 2 (set (reg/f:SI 78 [ p.p ])
        (reg/f:SI 86 [ start ])) -1 (nil)
    (expr_list:REG_EQUAL (mem/s:SI (plus:SI (reg/f:SI 20 frame)
                (const_int -24 [0xffffffe8])) [4 p.p+0 S4 A32])
        (nil)))

Note that we now have a REG_EQUAL note on insn 112 that is a lie, since insn 114
no longer sets the memory.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16590


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