Re: mcmodel=large problem

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, Feb 17, 2012 at 7:13 PM, Ian Lance Taylor <iant@xxxxxxxxxx> wrote:
> Zeng Bin <ezengbin@xxxxxxxxx> writes:
>
>> It reported the following error message during compilation:
>>
>> ../../../src/libgcc/config/libbid/bid_decimal_globals.c: In function
>> ‘__dfp_set_round’:
>> ../../../src/libgcc/config/libbid/bid_decimal_globals.c:34:1: error:
>> unrecognizable insn:
>> (call_insn/u 6 5 7 3 (parallel [
>>             (set (reg:DI 0 ax)
>>                 (call:DI (mem:QI (symbol_ref:DI ("__tls_get_addr")) [0 S1
>> A8])
>>                     (const_int 0 [0])))
>>             (unspec:DI [
>>                     (symbol_ref:DI ("__bid_IDEC_glbround") [flags 0x10]
>> <var_decl 0x2b9b4508f000 __bid_IDEC_glbround>)
>>                 ] UNSPEC_TLS_GD)
>>         ]) ../../../src/libgcc/config/libbid/bid_decimal_globals.c:33 -1
>>      (expr_list:REG_EH_REGION (const_int -2147483648 [0xffffffff80000000])
>>         (nil))
>>     (nil))
>> ../../../src/libgcc/config/libbid/bid_decimal_globals.c:34:1: internal
>> compiler error: in extract_insn, at recog.c:2109
>> Please submit a full bug report,
>
>
> This is a compiler bug.  Please submit a bug report, as described at
> http://gcc.gnu.org/bugs.  Thanks.
>
>
>> BTW, I am using Ubuntu with gcc-4.6.1. I am trying to compile gcc-4.6.2
>> with -mcmodel=large.
>
> Relatively few people use -mcmodel=large, because of the performance
> penalty.  I suspect that it is not well-tested, hence this sort of
> problem.
>
> Another approach you can take to running a program at high addresses is
> to use -fPIE and -pie, which lets you move the program around in
> memory.
>
> Ian

Thanks a lot! The problem still exists even configured with -fPIE:

../src/configure --disable-multilib --with-multilib-list=m64
CFLAGS=-fPIE CFLAGS_FOR_TARGET=-mcmodel=large
--with-gmp=/home/bzeng/tools/gmp/gmp-5.0.4/install
--with-mpfr=/home/bzeng/tools/mpfr/mpfr-3.1.0/install
--with-mpc=/home/bzeng/tools/mpc/mpc-0.9/install
--prefix=/home/bzeng/tools/gcc/gcc-4.6.2/install-large

The error is still the same:

../../../src/libgcc/config/libbid/bid_decimal_globals.c: In function
‘__dfp_set_round’:
../../../src/libgcc/config/libbid/bid_decimal_globals.c:34:1: error:
unrecognizable insn:
(call_insn/u 6 5 7 3 (parallel [
            (set (reg:DI 0 ax)
                (call:DI (mem:QI (symbol_ref:DI ("__tls_get_addr")) [0 S1 A8])
                    (const_int 0 [0])))
            (unspec:DI [
                    (symbol_ref:DI ("__bid_IDEC_glbround") [flags
0x10] <var_decl 0x2abaca167000 __bid_IDEC_glbround>)
                ] UNSPEC_TLS_GD)
        ]) ../../../src/libgcc/config/libbid/bid_decimal_globals.c:33 -1
     (expr_list:REG_EH_REGION (const_int -2147483648 [0xffffffff80000000])
        (nil))
    (nil))
../../../src/libgcc/config/libbid/bid_decimal_globals.c:34:1: internal
compiler error: in extract_insn, at recog.c:2109
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.
make[3]: *** [bid_decimal_globals.o] Error 1
make[2]: *** [all-stage1-target-libgcc] Error 2
make[1]: *** [stage1-bubble] Error 2
make: *** [all] Error 2

BTW, it is a ubuntu box running on a core2 processor. The native gcc
is gcc-4.6.1. Trying to compile gcc-4.6.2.
Thanks a lot.



[Index of Archives]     [Linux C Programming]     [Linux Kernel]     [eCos]     [Fedora Development]     [Fedora Announce]     [Autoconf]     [The DWARVES Debugging Tools]     [Yosemite Campsites]     [Yosemite News]     [Linux GCC]

  Powered by Linux