[PATCH] RS6000: rs6000.c checks options that have not yet been set

[Samuel Tardieu]

>>>>> "Janis" == Janis Johnson <janis187@us.ibm.com> writes:

Janis> Previously the code in rs6000.c was executed after the code in
Janis> opts.c but now it's run before that when it has no effect.

The code was wrong where it was placed as it tested flags that
happened to have an appropriate value because of their default
initialization, ignoring the user's choice. It only worked by chance.

  Sam
-- 
Samuel Tardieu -- sam@rfc1149.net -- http://www.rfc1149.net/