BOOTSTRAP FAILURE: segementation fault in genattrtab under hpux

John David Anglin dave@hiauly1.hia.nrc.ca
Sun Jan 7 08:33:00 GMT 2001


> 
> > 	* pa.h (hppa_rp_saved, RETURN_POINTER_REGNUM, HARD_REGNO_RENAME_OK):
> 
> Earnshaw notwithstanding, this is just plain wrong, 
> for reasons detailed elsewhere.

FWIW, the results of building under hpux 10.20 with the above and
< http://gcc.gnu.org/ml/gcc-patches/2000-12/msg01406.html > are here
< http://gcc.gnu.org/ml/gcc-testresults/2001-01/msg00081.html >.  This
was done with "-g -O3", so register renaming was in effect.

The advantage of this is that the "return" insn doesn't have to change
and all optimisations that worked before with trivial epilogues, leaf
functions, etc should still work.  I was also concerned with the old
code that regs_ever_live could change after the return pointer was
saved, affecting code generation.

Dave
-- 
J. David Anglin                                  dave.anglin@nrc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6605)


More information about the Gcc-bugs mailing list