target/6221: mips-irix6 gcc-3.1 testsuite failure in gcc.c-torture/execute/20020227-1.c

[Richard Henderson]

On Sun, Apr 14, 2002 at 01:28:11PM -0400, Kaveh R. Ghazi wrote:
>  >     It's a reload problem common to _all_ 64-bit targets.
> 
> So what is the prognosis doctor?  Fixable by 3.1 release?

No.  We could paper over it by having the rtl generator not generate
the problematic case.  OTOH, unaligned complex numbers are not exactly
a common occurrance, so it's nowhere near a priority.


r~