This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: asm constraint rule change



  In message <19981018005627.B10257@dot.cygnus.com>you write:
  > On Sat, Oct 17, 1998 at 11:45:08AM -0600, Jeffrey A Law wrote:
  > > Constraints for internal patterns are documented in this manner and the f
  > act
  > > that asm constraints do not have this documentation is a mistake.
  > 
  > A mistake, yes.  Unfortunate, yes.  Worth making a stink over?  IMHO, no.
How about a warning with the transformation?  With the transformation 
explicitly disappearing in a while (1.2, 1.3?)

Really, how much of this braindamage are we going to support and for how
long?

For example, there's still a dozen or so files in the 2.1.125 kernel that
have explicit clobbers for input/output regs.

At some point this kind of stuff has to be fixed.  I strongly suspect that
until we actually make such bogus code not compile that nobody will take
fixing the bogons seriously.


The longer we allow this kind of crud to go on, the worse it gets.

jeff


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]