ppc regression
David Edelsohn
dje@watson.ibm.com
Sun Jun 20 14:02:00 GMT 1999
I realized that in my update to the movdf_hardfloat32 pattern, I
forgot to update the insn length attribute. The attribute needs to list
the worst case instruction case.
David
Index: rs6000.md
===================================================================
RCS file: /egcs/carton/cvsfiles/egcs/gcc/config/rs6000/rs6000.md,v
retrieving revision 1.57
retrieving revision 1.58
diff -c -p -r1.57 -r1.58
*** rs6000.md 1999/06/19 00:04:59 1.57
--- rs6000.md 1999/06/20 13:59:59 1.58
***************
*** 6397,6403 ****
}
}"
[(set_attr "type" "*,load,store,*,*,*,fp,fpload,fpstore")
! (set_attr "length" "8,8,8,8,12,16,*,*,*")])
(define_insn "*movdf_softfloat32"
[(set (match_operand:DF 0 "nonimmediate_operand" "=r,r,m,r,r,r")
--- 6397,6403 ----
}
}"
[(set_attr "type" "*,load,store,*,*,*,fp,fpload,fpstore")
! (set_attr "length" "8,16,16,8,12,16,*,*,*")])
(define_insn "*movdf_softfloat32"
[(set (match_operand:DF 0 "nonimmediate_operand" "=r,r,m,r,r,r")
More information about the Gcc-patches
mailing list