gcc-2.95pre: Internal compiler error in `gen_add2_insn'

[Franz Sirl]

At 19:34 26.05.99 , David Edelsohn wrote:
>         gen_reload() is designed to handle a failure from
>emit_insn(gen_add2_insn()), but it is designed to handle a different type
>of failure.  The call to gen_add2_insn() is not protected from predicate
>mismatches by that stage.
>
>         Maybe gen_add2_insn() and gen_sub2_insn() should return
>CODE_FOR_nothing instead of aborting.  I am not sure of the consequences
>elsewhere in the compiler and their ability to handle that type of return
>value.  On the other hand, that is what they should get for another type
>of failure mode that they seem to be able to handle.

Actually, having looked at it again, I still don't see where the problem is 
with my patch. It may need tweaking, but basically I copied the code from a 
few lines above in gen_reload(), where it tries to check whether the 
addition can be handled in a single 3-operand add.

Franz.