http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46317

--- Comment #4 from end3er at gmail dot com 2011-01-28 21:04:45 UTC ---
(In reply to comment #3)
> (In reply to comment #2)
> > (In reply to comment #1)
> > > Does -fno-tree-sink fixes the issue?
> > 
> > No it doesn't. The only flags that works are:
> > -O0 or
> > -fno-inline or
> > -fno-unit-at-a-time or
> > -fno-toplevel-reorder
> > 
> > I can also make it work more or less by deactivating a certain number of
> > optimisation flag, but the output code does not really work... (I have tried
> > something like 58 different flags...)
> 
> Out of interest, did you try any newer releases of gcc?

I finally took the time to compile gcc-4.5, and the issue is still there:
Using built-in specs.
COLLECT_GCC=arm-none-eabi-c++
Target: arm-none-eabi
Configured with: ../configure --prefix=/opt/gcc-4.5/arm-none-eabi
--target=arm-none-eabi --enable-interwork --enable-multilib
--enable-languages=c,c++ --with-newlib --disable-nls --disable-shared
--disable-threads --with-gnu-as --with-gnu-ld
--with-gmp=/home/build/gcc-4.5_arm/gmp --with-mpfr=/home/build/gcc-4.5_arm/mpfr
--with-mpc=/home/build/gcc-4.5_arm/mpc
Thread model: single
gcc version 4.5.0 (GCC)


00008268 <D::D(int)>:
    8268:       e92d4038        push    {r3, r4, r5, lr}
    826c:       e1a04000        mov     r4, r0
    8270:       e1a02001        mov     r2, r1
    8274:       e1a03000        mov     r3, r0
    8278:       e59f103c        ldr     r1, [pc, #60]   ; 82bc <D::D(int)+0x54>
    827c:       e4831008        str     r1, [r3], #8
    8280:       e59f5038        ldr     r5, [pc, #56]   ; 82c0 <D::D(int)+0x58>
    8284:       e595102c        ldr     r1, [r5, #44]   ; 0x2c
    8288:       e5801008        str     r1, [r0, #8]
    828c:       e5111010        ldr     r1, [r1, #-16]
    8290:       e595c030        ldr     ip, [r5, #48]   ; 0x30
    8294:       e783c001        str     ip, [r3, r1]
    8298:       e285101c        add     r1, r5, #28
    829c:       ebffffca        bl      81cc <C::C(int)>

...

    82bc:       0001d3e8
    82c0:       0001d418
...
r1 = [0x1d418, #44]
   1d444:       0001d4b0
...
r1 = 0x1d4b0, #-16
0001d4a0 <construction vtable for B-in-D>:
   1d4a0:       fffffff8
...
000081cc <C::C(int)>:
    81cc:       e591c000        ldr     ip, [r1]


the correct r1 value should be:

0001d488 <construction vtable for C-in-D>:
        ...
   1d494:       0001d4f8        strdeq  sp, [r1], -r8
   1d498:       000081b8        undefined instruction 0x000081b8
   1d49c:       00000000        andeq   r0, r0, r0


When compiling with -O0, r1 is set to a correct address belonging to C-in-D:

00008448 <D::D(int)>:
    8448:       e92d4800        push    {fp, lr}
    844c:       e28db004        add     fp, sp, #4
    8450:       e24dd008        sub     sp, sp, #8
    8454:       e50b0008        str     r0, [fp, #-8]
    8458:       e50b100c        str     r1, [fp, #-12]
    845c:       e51b3008        ldr     r3, [fp, #-8]
    8460:       e1a00003        mov     r0, r3
    8464:       eb000100        bl      886c <A::A()>
    8468:       e51b3008        ldr     r3, [fp, #-8]
    846c:       e2832008        add     r2, r3, #8
    8470:       e59f3070        ldr     r3, [pc, #112]  ; 84e8 <D::D(int)+0xa0>
    8474:       e2833014        add     r3, r3, #20
    8478:       e1a00002        mov     r0, r2
    847c:       e1a01003        mov     r1, r3
    8480:       eb000106        bl      88a0 <B::B()>
    8484:       e51b2008        ldr     r2, [fp, #-8]
    8488:       e59f3058        ldr     r3, [pc, #88]   ; 84e8 <D::D(int)+0xa0>
    848c:       e2833004        add     r3, r3, #4
    8490:       e1a00002        mov     r0, r2
    8494:       e1a01003        mov     r1, r3
    8498:       e51b200c        ldr     r2, [fp, #-12]
    849c:       ebffff45        bl      81b8 <C::C(int)>
...
    84e8:       0001d950
...
0001d950, #4
    1d954:       0001d980
...
0001d970 <construction vtable for C-in-D>:
        ...
   1d97c:       0001d9f4        strdeq  sp, [r1], -r4
   1d980:       00008374        andeq   r8, r0, r4, ror r3

Reply via email to