summaryrefslogtreecommitdiff
path: root/TAO/tao/Align.h
diff options
context:
space:
mode:
Diffstat (limited to 'TAO/tao/Align.h')
-rw-r--r--TAO/tao/Align.h84
1 files changed, 84 insertions, 0 deletions
diff --git a/TAO/tao/Align.h b/TAO/tao/Align.h
new file mode 100644
index 00000000000..f1b75a7c0b5
--- /dev/null
+++ b/TAO/tao/Align.h
@@ -0,0 +1,84 @@
+// This may look like C, but it's really -*- C++ -*-
+// @(#) $Id$
+
+// ============================================================================
+//
+// = LIBRARY
+// TAO
+//
+// = FILENAME
+// Align.h
+//
+// = DESCRIPTION
+// Pointer alignment utilities
+//
+// A "ptr_arith_t" type is defined for doing numerical operations
+// on pointers, such as aligning them. Pointer sizes vary from 2
+// to 8 bytes in today's environments; a portable data type is much
+// needed.
+//
+// = AUTHOR
+// Copyright 1994-1995 by Sun Microsystems, Inc.
+//
+// ============================================================================
+
+#ifndef TAO_ALIGN_H
+#define TAO_ALIGN_H
+
+// Type for doing arithmetic on pointers ... as elsewhere, we assume
+// that "unsigned" versions of a type are the same size as the
+// "signed" version of the same type.
+
+#if ACE_SIZEOF_VOID_P == ACE_SIZEOF_INT
+typedef u_int ptr_arith_t;
+
+#elif ACE_SIZEOF_VOID_P == ACE_SIZEOF_LONG
+typedef u_long ptr_arith_t;
+
+#elif ACE_SIZEOF_VOID_P == ACE_SIZEOF_LONG_LONG
+typedef u_long long ptr_arith_t;
+
+#else
+# error "Can't find a suitable type for doing pointer arithmetic."
+#endif /* error */
+
+// Efficiently align "value" up to "alignment", knowing that all such
+// boundaries are binary powers and that we're using two's complement
+// arithmetic.
+
+// Since the alignment is a power of two its binary representation is:
+// alignment = 0...010...0
+//
+// hence
+//
+// alignment - 1 = 0...001...1 = T1
+//
+// so the complement is:
+//
+// ~(alignment - 1) = 1...110...0 = T2
+//
+// Notice that there is a multiple of <alignment> in the range
+// [<value>,<value> + T1], also notice that if
+//
+// X = ( <value> + T1 ) & T2
+//
+// then
+//
+// <value> <= X <= <value> + T1
+//
+// because the & operator only changes the last bits, and since X is a
+// multiple of <alignment> (its last bits are zero) we have found the
+// multiple we wanted.
+//
+
+#define align_binary(ptr, alignment) \
+ ((ptr + ((ptr_arith_t)((alignment)-1))) & (~((ptr_arith_t)((alignment)-1))))
+
+// Efficiently round "ptr" up to an "alignment" boundary, knowing that
+// all such boundaries are binary powers and that we're using two's
+// complement arithmetic.
+//
+#define ptr_align_binary(ptr, alignment) \
+ ((char *) align_binary (((ptr_arith_t) (ptr)), (alignment)))
+
+#endif /* TAO_ALIGN_H */