nir: Handle large unsigned values in opt_algebraic.

The next patch adds an algebraic rule that uses the constant 0xff00ff00.

Without this change, the build fails with

   return hex(struct.unpack('I', struct.pack('i', self.value))[0])
   struct.error: 'i' format requires -2147483648 <= number <= 2147483647

The hex() function handles integers of any size, and assigning a
negative value to an unsigned does what we want in C. The pack/unpack is
unnecessary (and as we see, buggy).

Reviewed-by: Dylan Baker <baker.dylan.c@gmail.com>

diff --git a/src/compiler/nir/nir_algebraic.py b/src/compiler/nir/nir_algebraic.py
index 77ad35eda2a28d77c507f0faf3920ef5b92d2375..2357b57117a6abfd854a439193f432fb5289f912 100644 (file)
--- a/src/compiler/nir/nir_algebraic.py
+++ b/src/compiler/nir/nir_algebraic.py
@@ -102,13 +102,10 @@ class Constant(Value):
       self.value = val
 
    def __hex__(self):
-      # Even if it's an integer, we still need to unpack as an unsigned
-      # int.  This is because, without C99, we can only assign to the first
-      # element of a union in an initializer.
       if isinstance(self.value, (bool)):
          return 'NIR_TRUE' if self.value else 'NIR_FALSE'
       if isinstance(self.value, (int, long)):
-         return hex(struct.unpack('I', struct.pack('i', self.value))[0])
+         return hex(self.value)
       elif isinstance(self.value, float):
          return hex(struct.unpack('I', struct.pack('f', self.value))[0])
       else: