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: bct_p *is* used in loop.c:strength_reduction



  In message <9810151714.AA40610@marc.watson.ibm.com>you write:
  > `unused'
  >      This attribute, attached to a variable, means that the variable is
  >      meant to be possibly unused.  GNU CC will not produce a warning
  >      for this variable.
  > 
  > I do not know what "meant to be possibly" is intended to convey.  This is
  > intended for cases where some compilations will not use the variable?
"possibly unused" from an optimizer standpoint means that you can't depend on
it being unused.  It also sounds pretty clearly to me like it was meant to
describe the exact situation we're encountering -- we have a variable that
may or may not be used depending on a variety of compile-time decisions.

If adding an unused attribute is causing us problems, then the problem must
be elsewhere, not with the attribute itself and we should go find the real
problem instead of just disabling the unused attributes.


jeff


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