[Bug optimization/10287] [3.4 regression] [ARM] Loop/conditional store bug

[Bug optimization/10287] [3.4 regression] [ARM] Loop/conditional store bug

[drow@mvista.com]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10287



------- Additional Comments From drow@mvista.com  2003-06-02 16:10 -------
Sorry, Jan, I missed your response.  Yes, the #4 does signify auto-update.
You should be able to reproduce this with a cross compiler and the simulator.



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug optimization/10287] [3.4 regression] [ARM] Loop/conditional store bug

[dhazeghi@yahoo.com]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10287


dhazeghi@yahoo.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |WAITING
  GCC build triplet|                            |i686-pc-linux-gnu
   GCC host triplet|                            |i686-pc-linux-gnu
 GCC target triplet|                            |arm-linux


------- Additional Comments From dhazeghi@yahoo.com  2003-06-04 00:27 -------
Daniel,

the assembly I get out of gcc mainline (arm-elf targeted cross compiler) doesn't seem to be what 
you initially reported. Can you confirm whether the below code is correct? Thanks,

Dara

        .file   "test.c"
        .text
        .align  2
        .global main
        .type   main, %function
main:
        @ args = 0, pretend = 0, frame = 64
        @ frame_needed = 1, uses_anonymous_args = 0
        mov     ip, sp
        stmfd   sp!, {fp, ip, lr, pc}
        sub     fp, ip, #4
        sub     sp, sp, #64
        mov     r1, #0
        sub     r3, fp, #12
        mov     r2, #15
.L5:
        subs    r2, r2, #1
        str     r1, [r3, #-64]
        add     r3, r3, #4
        bpl     .L5
        mov     r2, #2
        sub     r1, fp, #76
        mov     r0, #1
        mov     ip, r2
.L10:
        cmp     r2, #0
        str     r0, [r1], #4
        moveq   r3, r0
        movne   r3, ip
        subs    r2, r2, #2
        str     r3, [r1], #4
        bpl     .L10
        ldr     r0, [fp, #-64]
        cmp     r0, #0
        ldmeqea fp, {fp, sp, pc}
        bl      abort
        .size   main, .-main
        .ident  "GCC: (GNU) 3.4 20030603 (experimental)"



------- You are receiving this mail because: -------
You are on the CC list for the bug, or are watching someone who is.

[Bug optimization/10287] [3.4 regression] [ARM] Loop/conditional store bug

[drow@gcc.gnu.org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10287


drow@gcc.gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2003-06-12 20:23:41
               date|                            |


------- Additional Comments From drow@gcc.gnu.org  2003-06-12 20:23 -------
Your output still contains the str/moveq/movne/str sequence (instead
of str/streq).  So the bug is still present.

[Bug optimization/10287] [3.4 regression] [ARM] Loop/conditional store bug

[rearnsha@gcc.gnu.org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10287


rearnsha@gcc.gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |DUPLICATE


------- Additional Comments From rearnsha@gcc.gnu.org  2003-06-13 10:56 -------
Duplicate of 11052, which I fixed last week.

Relevant code fragment is now:

        str     r1, [r3], #4
        cmp     r2, #0
        strne   r0, [r3], #4
        subs    r2, r2, #2
        bpl     .L10


*** This bug has been marked as a duplicate of 11052 ***