[PATCH][riscv] Fix build due to INT16_MAX issue

[Kyrill Tkachov]

Hi all,

I tried building a cc1 for riscv-x-elf and I got a build error about INT16_MAX 
not being defined on my system.
I think it's not a standard macro.  This patch fixes that with what I think is 
the equivalent definition using
GCC modes logic.

With this my build proceeds to build a cc1 successfully. No testing beyond that.

Is this ok for trunk?

Thanks,
Kyrill

P.S. Palmer, Andrew, I didn't see your names and emails in the MAINTAINERS 
file, you should update that when
you get the chance.


2016-02-07  Kyrylo Tkachov  <kyrylo.tkac...@arm.com>

    * config/riscv/riscv.c (riscv_build_integer_1): Avoid use of INT16_MAX.

diff --git a/gcc/config/riscv/riscv.c b/gcc/config/riscv/riscv.c
index 834651f..89567f7 100644
--- a/gcc/config/riscv/riscv.c
+++ b/gcc/config/riscv/riscv.c
@@ -356,7 +356,9 @@ riscv_build_integer_1 (struct riscv_integer_op codes[RISCV_MAX_INTEGER_OPS],
   /* End with ADDI.  When constructing HImode constants, do not generate any
      intermediate value that is not itself a valid HImode constant.  The
      XORI case below will handle those remaining HImode constants.  */
-  if (low_part != 0 && (mode != HImode || value - low_part <= INT16_MAX))
+  if (low_part != 0
+      && (mode != HImode
+	  || value - low_part <= ((1 << (GET_MODE_BITSIZE (HImode) - 1)) - 1)))
     {
       alt_cost = 1 + riscv_build_integer_1 (alt_codes, value - low_part, mode);
       if (alt_cost < cost)