Explain our implementation choice for implementation-defined values
This commit is contained in:
parent
c9c3919a67
commit
10f56c9348
@ -1,4 +1,4 @@
|
||||
/* $NetBSD: fpu.h,v 1.6 2008/05/10 15:31:04 martin Exp $ */
|
||||
/* $NetBSD: fpu.h,v 1.7 2017/10/17 00:26:35 maya Exp $ */
|
||||
|
||||
/*-
|
||||
* Copyright (c) 2001 Ross Harvey
|
||||
@ -91,8 +91,8 @@
|
||||
#define IEEE_TRAP_ENABLE_INV _FP_C_DEF(1)
|
||||
|
||||
#define IEEE_INHERIT _FP_C_DEF(14)
|
||||
#define IEEE_MAP_UMZ _FP_C_DEF(13)
|
||||
#define IEEE_MAP_DMZ _FP_C_DEF(12)
|
||||
#define IEEE_MAP_UMZ _FP_C_DEF(13) /* Map underflowed outputs to zero */
|
||||
#define IEEE_MAP_DMZ _FP_C_DEF(12) /* Map denormal inputs to zero */
|
||||
|
||||
#define FP_C_MIRRORED (IEEE_STATUS_INE | IEEE_STATUS_UNF | IEEE_STATUS_OVF\
|
||||
| IEEE_STATUS_DZE | IEEE_STATUS_INV)
|
||||
|
Loading…
Reference in New Issue
Block a user