is subreg with subreg_byte outside source valid?

[Jan Hubicka]

> 	I am not as familiar with the SVR4 mode.  What jumps out at me is
> that PowerPC is a RISC architecture: loading a register from that many
> operands cannot be valid.  You've been working on Hammer code generation
> for too long (:^).  I would guess that there is something wrong with the
:) Give me another new chip to play with and I will write code for it as well :)
> legitimate address code to allow a MEM like that to exist so late.
I've already found the problem, but it is also rather nontrivial. See my
last post to gcc-patches.

Is the powerpc breakage I caused by the patch serious for you? I hope
these two issues to see sorted out till tommorow evening, but I may
try to commit some very conservative fixes as obvious until thinks settle
down if required.
> 
> David