Re: NEWBIE question - unrecognizable insn in libgcc2

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

 



"David Kane" <davidfrkane@xxxxxxxxxxx> writes:

> ../../../../source/gcc-3.4-20050121/gcc/libgcc2.c: In function `__mulsi3':
> ../../../../source/gcc-3.4-20050121/gcc/libgcc2.c:537: error: unrecognizable
> insn:
> (insn 129 69 70 2 ../../../../source/gcc-3.4-20050121/gcc/libgcc2.c:531 (set
> (subreg:HI (mem:SI (reg/f:HI 14 r14) [4 __w+0 S4 A16]) 0)
>         (reg:HI 4 r4 [54])) -1 (nil)
>     (nil))
> ../../../../source/gcc-3.4-20050121/gcc/libgcc2.c:537: internal compiler
> error: in extract_insn, at recog.c:2083

The compiler is trying to store an HImode value in a register to
memory, and it can't find an insn to do that.

If HARD_REGNO_MODE_OK is going to return true for storing HImode in
r4, then the compiler requires that you provide a way to store that
register into memory.  Either the movhi insn must be able to do this,
or you must define SECONDARY_RELOAD_CLASS to tell reload to store the
value into a different register which can be stored to memory.

It is also possible that there is some problem with the memory
address, and that it is not being recognized by
GO_IF_LEGITIMATE_ADDRESS, and is not being fixed by
LEGITIMIZE_RELOAD_ADDRESS.

Ian

[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