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]
Other format: [Raw text]

Re: SEQUENCEs (PR optimization/9745)


   From: David Edelsohn <dje at watson dot ibm dot com>
   Date: Tue, 25 Feb 2003 22:05:58 -0500
   
   	The anamoly is loop.c:loop_regs_update() changed from
   
     if (GET_CODE (seq) == SEQUENCE)
       {
       }
     else if (GET_CODE (seq) == SET ...)
       {
       }
   
   to
   
     if (INSN_P (seq))
       {
       }
     else if (GET_CODE (seq) == SET ...)
       {
       }
   
   GET_CODE (seq) == SET depends on INSN_P (seq) being true, which was tested
   earlier, so the latter case seems useless.  One could change the first
   test to 
   
     if (INSN_P (seq) && NEXT_INSN (seq) != NULL_RTX)
   
   which mainly avoids the single_set() test for some instructions.
   
The top-level ELSE block may be safely deleted.


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