As reported in https://gcc.gnu.org/ml/gcc-patches/2016-11/msg02388.html .
Changing the mode of a hard register can lead to problems, or at least it can make worse code if the result will need reloads. Tested on avr-elf on the test in that email, and bootstrapped and regression checked on powerpc64-linux {-m32,-m64}. Committing to trunk. Segher 2016-11-23 Segher Boessenkool <seg...@kernel.crashing.org> * combine.c (change_zero_ext): Only change the mode of a hard register destination if can_change_dest_mode holds for that. --- gcc/combine.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/gcc/combine.c b/gcc/combine.c index eef917a..c8a71eb 100644 --- a/gcc/combine.c +++ b/gcc/combine.c @@ -11287,7 +11287,8 @@ change_zero_ext (rtx pat) else if (GET_CODE (x) == ZERO_EXTEND && SCALAR_INT_MODE_P (mode) && REG_P (XEXP (x, 0)) - && HARD_REGISTER_P (XEXP (x, 0))) + && HARD_REGISTER_P (XEXP (x, 0)) + && can_change_dest_mode (XEXP (x, 0), 0, mode)) { size = GET_MODE_PRECISION (GET_MODE (XEXP (x, 0))); x = gen_rtx_REG (mode, REGNO (XEXP (x, 0))); -- 1.9.3