untested changes (was re: patch applied to reload1.c)

Joern Rennecke amylaar@cygnus.co.uk
Fri Sep 25 09:45:00 GMT 1998


> There is no way these changes were tested before being installed.

Actually, they were tested for quite some time - for SH and x86 - but
some last-minute changes were made to add comments.  Unfortunately, a
line of code got lost in the process.

> I've just checked in the following fix, but I fear there may be more
> hidden treasure in these reload changes since even this obvious crash
> did not get caught.

Thanks.  However, the coding standards say to avoid assignments inside
conditions, so reverted the code to use a separate assignment.

Fri Sep 25 17:35:23 1998  J"orn Rennecke <amylaar@cygnus.co.uk>

        * reload1.c (reload_combine): Re-add line that got accidentally lost.


Index: reload1.c
===================================================================
RCS file: /egcs/carton/cvsfiles/egcs/gcc/reload1.c,v
retrieving revision 1.66
diff -p -r1.66 reload1.c
*** reload1.c	1998/09/24 23:18:31	1.66
--- reload1.c	1998/09/25 16:35:49
*************** reload_combine ()
*** 9274,9280 ****
  
  	 First, check that we have (set (REGX) (PLUS (REGX) (REGY)))
  	 and that we know all uses of REGX before it dies.  */
!       if ((set = single_set (insn)) != NULL_RTX
  	  && GET_CODE (SET_DEST (set)) == REG
  	  && (HARD_REGNO_NREGS (REGNO (SET_DEST (set)),
  				GET_MODE (SET_DEST (set)))
--- 9274,9281 ----
  
  	 First, check that we have (set (REGX) (PLUS (REGX) (REGY)))
  	 and that we know all uses of REGX before it dies.  */
!       set = single_set (insn);
!       if (set != NULL_RTX
  	  && GET_CODE (SET_DEST (set)) == REG
  	  && (HARD_REGNO_NREGS (REGNO (SET_DEST (set)),
  				GET_MODE (SET_DEST (set)))



More information about the Gcc-patches mailing list