diff options
| author | Mark Kettenis <kettenis@gnu.org> | 2000-03-26 21:21:50 +0000 |
|---|---|---|
| committer | Mark Kettenis <kettenis@gnu.org> | 2000-03-26 21:21:50 +0000 |
| commit | ac27f131b2fc0c5e0872c5855c599974b7416d07 (patch) | |
| tree | 399afc0505a8f76691e4d2f234516cd003e5961e /gdb/config/i386/tm-i386mk.h | |
| parent | 9f10757c407c9946029436682ae554f9739fa62d (diff) | |
| download | gdb-ac27f131b2fc0c5e0872c5855c599974b7416d07.zip gdb-ac27f131b2fc0c5e0872c5855c599974b7416d07.tar.gz gdb-ac27f131b2fc0c5e0872c5855c599974b7416d07.tar.bz2 | |
2000-03-26 Mark Kettenis <kettenis@gnu.org>
Provide `long double' support for most i386 targets.
* config/i386/tm-i386.h (TARGET_LONG_DOUBLE_FORMAT): Define as
&floatformat_i387_ext.
(TARGET_LONG_DOUBLE_BITS): Define as 96.
(REGISTER_VIRTUAL_TYPE): Change type for FPU registers to
`builtin_type_long_double'.
(REGISTER_CONVERT_TO_VIRTUAL): Call
i386_register_convert_to_virtual.
(REGISTER_CONVERT_TO_RAW): Call i386_register_convert_to_raw.
(i387_to_double, double_to_i387): Remove prototypes.
(i386_extract_return_value): Change prototype to match definition
in i386-tdep.c.
* config/i386/tm-i386mk.h (TARGET_LONG_DOUBLE_FORMAT): #undef.
(TARGET_LONG_DOUBLE_BITS): #undef.
* config/i386/tm-linux.h (TARGET_LONG_DOUBLE_BIT): Remove.
[HAVE_LONG_DOUBLE && HOST_I386] (LD_I387): Remove.
(i387_extract_floating, i387_store_floating): Remove prototypes.
(TARGET_EXTRACT_FLOATING, TARGET_STORE_FLOATING): Remove.
(REGISTER_CONVERT_TO_VIRTUAL, REGOISTER_CONVERT_TO_RAW): Remove.
(REGISTER_VIRTUAL_TYPE): Remove.
* i386-tdep.c (i386_register_convert_to_virtual): New function.
(i386_register_convert_to_raw): New function.
* i387-tdep.c [LD_I387] (i387_extract_floating): Remove.
(i387_store_floating): Remove.
Diffstat (limited to 'gdb/config/i386/tm-i386mk.h')
| -rw-r--r-- | gdb/config/i386/tm-i386mk.h | 14 |
1 files changed, 13 insertions, 1 deletions
diff --git a/gdb/config/i386/tm-i386mk.h b/gdb/config/i386/tm-i386mk.h index cb18e3c..de66a69 100644 --- a/gdb/config/i386/tm-i386mk.h +++ b/gdb/config/i386/tm-i386mk.h @@ -1,5 +1,5 @@ /* Macro definitions for i386, Mach 3.0, OSF 1/MK - Copyright (C) 1992 Free Software Foundation, Inc. + Copyright (C) 1992, 2000 Free Software Foundation, Inc. This file is part of GDB. @@ -24,3 +24,15 @@ #define MK67 1 #include "i386/tm-i386m3.h" + +/* FIMXE: kettenis/2000-03-26: On OSF 1, `long double' is equivalent + to `double'. However, I'm not sure what is the consequence of: + + #define TARGET_LONG_DOUBLE_FORMAT TARGET_DOUBLE_FORMAT + #define TARGET_LONG_DOUBLE_BITS TARGET_DOUBLE_BITS + + So I'll go with the current status quo instead. It looks like this + target won't compile anyway. Perhaps it should be obsoleted? */ + +#undef TARGET_LONG_DOUBLE_FORMAT +#undef TARGET_LONG_DOUBLE_BITS |