Re: anonymous ifields

[Ben Elliston <bje at redhat dot com>]

>>>>> "Frank" == Frank Ch Eigler <fche@redhat.com> writes:

  Frank> :   Frank> [...] we could get cgen to automagically declare a bunch of
  Frank> :   Frank> ifields for all contiguous bitfields in an instruction word. [...]
  Frank> : 
  Frank> : In some ways, I think that would be *worse*.  Things happening
  Frank> : implicitly just gets confusing.  

  Frank> What if the automagic ifields are called something obvious like
  Frank> "bitN" and "bitsN:M" and "byteN"?

If they had such names, any advantages of reduced typing would be
lost. :-)

  Frank> Not that bad, but I prefer to avoid perlitis in the cgen's input
  Frank> language, and also worry about unforseen mechanism breakage.

As it happens, the number of single bit fields and their usage didn't
prove to be as onerous as I expected.  Forget about it ..

Ben