SIGN_EXTRACT feature, perhaps
Jeffrey A Law
law@upchuck.cygnus.com
Fri Apr 30 23:15:00 GMT 1999
In message < 19990407025736.C9381@cygnus.com >you write:
> On Wed, Apr 07, 1999 at 03:38:54AM -0600, Jeffrey A Law wrote:
> > > No. SIGN_EXTRACT is read-only. ZERO_EXTRACT as a destination is
> > > a bitfield insertion, and as such it doesn't matter about sign,
> > > so we just worry about the one.
> > I hate to disagree with you, but.... Note carefully "they may appear
> > on the left side of an assignment". Perhaps you're confusing them with
> > zero/sign_extend which only appear in a source operand.
>
> No, I wasn't confusing sign_extend -- but I am surprised to see
> sign_extract documented as a destination. I thought we only used
> zero_extract for that.
Me too. The only reason I chimed in is I've always thought the
sign/zero extract documentation needed some work as I don't think
they're as clear as they could be :-)
Otherwise I wouldn't have known they are documented as both being valid
for a LHS.
jeff
More information about the Gcc-bugs
mailing list