Re: gmp

[Jim Pick <jim@jimpick.com>]

Alan Post <alan.post@usa.net> writes:

> OK, the problem is in -fschedule-insns and -fschedule-insns2:
> 
> `-fschedule-insns'
>      If supported for the target machine, attempt to reorder
>      instructions to eliminate execution stalls due to required data
>      being unavailable.  This helps machines that have slow floating
>      point or memory load instructions by allowing other instructions
>      to be issued until the result of the load or floating point
>      instruction is required.
> 
> `-fschedule-insns2'
>      Similar to `-fschedule-insns', but requests an additional pass of
>      instruction scheduling after register allocation has been done.
>      This is especially useful on machines with a relatively small
>      number of registers and where memory load instructions take more
>      than one cycle.
> 
> Adding -fno-schedule-insns -fno-schedule-insns2 to the CFLAGS in debian/rules
> (in all three places) compiles cleanly.
> 
> This unfortunately is the limit of my troubleshooting ability for this
> problem (within any reasonable time period :).

Great detective work!

What's the proper way to fix this?  Compile gmp with flags?

Or is this really a bug in gcc that should be corrected?

Cheers,

 - Jim