[Bug middle-end/59420] arm: broken code generated (memset from newlib 2.0)

pinskia at gcc dot gnu.org gcc-bugzilla@gcc.gnu.org
Sun Dec 8 00:56:00 GMT 2013


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59420

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
          Component|c                           |middle-end
         Resolution|---                         |INVALID

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Richard Earnshaw from comment #2)
> You'll get more attention paid to this if you can describe why you think the
> code generated is incorrect.

The issue is there is a loop inside of memset that gets converted to a call to
memset; that is if I read the assembly code correctly.  if that is the case
then this is a bug in the newlib makefiles as they should be using
-fno-tree-loop-distribute-patterns or like what glibc does:
/* Add the compiler optimization to inhibit loop transformation to library
   calls.  This is used to avoid recursive calls in memset and memmove
   default implementations.  */
#ifdef HAVE_CC_INHIBIT_LOOP_TO_LIBCALL
# define inhibit_loop_to_libcall \
    __attribute__ ((__optimize__ ("-fno-tree-loop-distribute-patterns")))
#else
# define inhibit_loop_to_libcall
#endif

So closing as invalid, please report this to newlib.

Note the work around is not to use -O3 or use -O3
-fno-tree-loop-distribute-patterns .



More information about the Gcc-bugs mailing list