summaryrefslogtreecommitdiffstats
path: root/gcc/config/i960/i960.h
diff options
context:
space:
mode:
Diffstat (limited to 'gcc/config/i960/i960.h')
-rw-r--r--gcc/config/i960/i960.h1404
1 files changed, 1404 insertions, 0 deletions
diff --git a/gcc/config/i960/i960.h b/gcc/config/i960/i960.h
new file mode 100644
index 00000000000..67c34e25031
--- /dev/null
+++ b/gcc/config/i960/i960.h
@@ -0,0 +1,1404 @@
+/* Definitions of target machine for GNU compiler, for Intel 80960
+ Copyright (C) 1992, 1993, 1995, 1996, 1998, 1999, 2000, 2001, 2002
+ Free Software Foundation, Inc.
+ Contributed by Steven McGeady, Intel Corp.
+ Additional Work by Glenn Colon-Bonet, Jonathan Shapiro, Andy Wilson
+ Converted to GCC 2.0 by Jim Wilson and Michael Tiemann, Cygnus Support.
+
+This file is part of GCC.
+
+GCC is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2, or (at your option)
+any later version.
+
+GCC is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with GCC; see the file COPYING. If not, write to
+the Free Software Foundation, 59 Temple Place - Suite 330,
+Boston, MA 02111-1307, USA. */
+
+/* Note that some other tm.h files may include this one and then override
+ many of the definitions that relate to assembler syntax. */
+
+/* Target CPU builtins. */
+#define TARGET_CPU_CPP_BUILTINS() \
+ do \
+ { \
+ builtin_define_std ("i960"); \
+ builtin_define_std ("I960"); \
+ builtin_define_std ("i80960"); \
+ builtin_define_std ("I80960"); \
+ builtin_assert ("cpu=i960"); \
+ builtin_assert ("machine=i960"); \
+ } \
+ while (0)
+
+#define MULTILIB_DEFAULTS { "mnumerics" }
+
+/* Name to predefine in the preprocessor for processor variations.
+ -mic* options make characters signed by default. */
+#define CPP_SPEC "%{mic*:-D__i960 -fsigned-char\
+ %{mka:-D__i960KA}%{mkb:-D__i960KB}\
+ %{mja:-D__i960JA}%{mjd:-D__i960JD}%{mjf:-D__i960JF}\
+ %{mrp:-D__i960RP}\
+ %{msa:-D__i960SA}%{msb:-D__i960SB}\
+ %{mmc:-D__i960MC}\
+ %{mca:-D__i960CA}%{mcc:-D__i960CC}\
+ %{mcf:-D__i960CF}}\
+ %{msoft-float:-D_SOFT_FLOAT}\
+ %{mka:-D__i960KA__ -D__i960_KA__}\
+ %{mkb:-D__i960KB__ -D__i960_KB__}\
+ %{msa:-D__i960SA__ -D__i960_SA__}\
+ %{msb:-D__i960SB__ -D__i960_SB__}\
+ %{mmc:-D__i960MC__ -D__i960_MC__}\
+ %{mca:-D__i960CA__ -D__i960_CA__}\
+ %{mcc:-D__i960CC__ -D__i960_CC__}\
+ %{mcf:-D__i960CF__ -D__i960_CF__}\
+ %{!mka:%{!mkb:%{!msa:%{!msb:%{!mmc:%{!mca:\
+ %{!mcc:%{!mcf:-D__i960_KB -D__i960KB__ %{mic*:-D__i960KB}}}}}}}}}\
+ %{mlong-double-64:-D__LONG_DOUBLE_64__}"
+
+/* Specs for the compiler, to handle processor variations.
+ If the user gives an explicit -gstabs or -gcoff option, then do not
+ try to add an implicit one, as this will fail.
+ -mic* options make characters signed by default. */
+#define CC1_SPEC \
+ "%{mic*:-fsigned-char}\
+%{!mka:%{!mkb:%{!msa:%{!msb:%{!mmc:%{!mca:%{!mcc:%{!mcf:%{!mja:%{!mjd:%{!mjf:%{!mrp:-mka}}}}}}}}}}}}\
+ %{!gs*:%{!gc*:%{mbout:%{g*:-gstabs}}\
+ %{mcoff:%{g*:-gcoff}}\
+ %{!mbout:%{!mcoff:%{g*:-gstabs}}}}}"
+
+/* Specs for the assembler, to handle processor variations.
+ For compatibility with Intel's gnu960 tool chain, pass -A options to
+ the assembler. */
+#define ASM_SPEC \
+ "%{mka:-AKA}%{mkb:-AKB}%{msa:-ASA}%{msb:-ASB}\
+ %{mmc:-AMC}%{mca:-ACA}%{mcc:-ACC}%{mcf:-ACF}\
+ %{mja:-AJX}%{mjd:-AJX}%{mjf:-AJX}%{mrp:-AJX}\
+ %{!mka:%{!mkb:%{!msa:%{!msb:%{!mmc:%{!mca:%{!mcc:%{!mcf:%{!mja:%{!mjd:%{!mjf:%{!mrp:-AKB}}}}}}}}}}}}\
+ %{mlink-relax:-linkrelax}"
+
+/* Specs for the linker, to handle processor variations.
+ For compatibility with Intel's gnu960 tool chain, pass -F and -A options
+ to the linker. */
+#define LINK_SPEC \
+ "%{mka:-AKA}%{mkb:-AKB}%{msa:-ASA}%{msb:-ASB}\
+ %{mmc:-AMC}%{mca:-ACA}%{mcc:-ACC}%{mcf:-ACF}\
+ %{mja:-AJX}%{mjd:-AJX}%{mjf:-AJX}%{mrp:-AJX}\
+ %{mbout:-Fbout}%{mcoff:-Fcoff}\
+ %{mlink-relax:-relax}"
+
+/* Specs for the libraries to link with, to handle processor variations.
+ Compatible with Intel's gnu960 tool chain. */
+#define LIB_SPEC "%{!nostdlib:-lcg %{p:-lprof}%{pg:-lgprof}\
+ %{mka:-lfpg}%{msa:-lfpg}%{mca:-lfpg}%{mcf:-lfpg} -lgnu}"
+
+/* Defining the macro shows we can debug even without a frame pointer.
+ Actually, we can debug without FP. But defining the macro results in
+ that -O means FP elimination. Addressing through sp requires
+ negative offset and more one word addressing in the most cases
+ (offsets except for 0-4095 require one more word). Therefore we've
+ not defined the macro. */
+/*#define CAN_DEBUG_WITHOUT_FP*/
+
+/* Do leaf procedure and tail call optimizations for -O2 and higher. */
+#define OPTIMIZATION_OPTIONS(LEVEL,SIZE) \
+{ \
+ if ((LEVEL) >= 2) \
+ { \
+ target_flags |= TARGET_FLAG_LEAFPROC; \
+ target_flags |= TARGET_FLAG_TAILCALL; \
+ } \
+}
+
+/* Print subsidiary information on the compiler version in use. */
+#define TARGET_VERSION fprintf (stderr," (intel 80960)");
+
+/* Generate DBX debugging information. */
+#define DBX_DEBUGGING_INFO 1
+
+/* Generate SDB style debugging information. */
+#define SDB_DEBUGGING_INFO 1
+#define EXTENDED_SDB_BASIC_TYPES
+
+/* Generate DBX_DEBUGGING_INFO by default. */
+#define PREFERRED_DEBUGGING_TYPE DBX_DEBUG
+
+/* Redefine this to print in hex. No value adjustment is necessary
+ anymore. */
+#define PUT_SDB_TYPE(A) \
+ fprintf (asm_out_file, "\t.type\t0x%x;", A)
+
+/* Handle pragmas for compatibility with Intel's compilers. */
+
+extern int i960_maxbitalignment;
+extern int i960_last_maxbitalignment;
+
+#define REGISTER_TARGET_PRAGMAS() do { \
+ c_register_pragma (0, "align", i960_pr_align); \
+ c_register_pragma (0, "noalign", i960_pr_noalign); \
+} while (0)
+
+/* Run-time compilation parameters selecting different hardware subsets. */
+
+/* 960 architecture with floating-point. */
+#define TARGET_FLAG_NUMERICS 0x01
+#define TARGET_NUMERICS (target_flags & TARGET_FLAG_NUMERICS)
+
+/* 960 architecture with memory management. */
+/* ??? Not used currently. */
+#define TARGET_FLAG_PROTECTED 0x02
+#define TARGET_PROTECTED (target_flags & TARGET_FLAG_PROTECTED)
+
+/* The following three are mainly used to provide a little sanity checking
+ against the -mARCH flags given. The Jx series, for the purposes of
+ gcc, is a Kx with a data cache. */
+
+/* Nonzero if we should generate code for the KA and similar processors.
+ No FPU, no microcode instructions. */
+#define TARGET_FLAG_K_SERIES 0x04
+#define TARGET_K_SERIES (target_flags & TARGET_FLAG_K_SERIES)
+
+/* Nonzero if we should generate code for the MC processor.
+ Not really different from KB for our purposes. */
+#define TARGET_FLAG_MC 0x08
+#define TARGET_MC (target_flags & TARGET_FLAG_MC)
+
+/* Nonzero if we should generate code for the CA processor.
+ Enables different optimization strategies. */
+#define TARGET_FLAG_C_SERIES 0x10
+#define TARGET_C_SERIES (target_flags & TARGET_FLAG_C_SERIES)
+
+/* Nonzero if we should generate leaf-procedures when we find them.
+ You may not want to do this because leaf-proc entries are
+ slower when not entered via BAL - this would be true when
+ a linker not supporting the optimization is used. */
+#define TARGET_FLAG_LEAFPROC 0x20
+#define TARGET_LEAFPROC (target_flags & TARGET_FLAG_LEAFPROC)
+
+/* Nonzero if we should perform tail-call optimizations when we find them.
+ You may not want to do this because the detection of cases where
+ this is not valid is not totally complete. */
+#define TARGET_FLAG_TAILCALL 0x40
+#define TARGET_TAILCALL (target_flags & TARGET_FLAG_TAILCALL)
+
+/* Nonzero if use of a complex addressing mode is a win on this implementation.
+ Complex addressing modes are probably not worthwhile on the K-series,
+ but they definitely are on the C-series. */
+#define TARGET_FLAG_COMPLEX_ADDR 0x80
+#define TARGET_COMPLEX_ADDR (target_flags & TARGET_FLAG_COMPLEX_ADDR)
+
+/* Align code to 8 byte boundaries for faster fetching. */
+#define TARGET_FLAG_CODE_ALIGN 0x100
+#define TARGET_CODE_ALIGN (target_flags & TARGET_FLAG_CODE_ALIGN)
+
+/* Append branch prediction suffixes to branch opcodes. */
+/* ??? Not used currently. */
+#define TARGET_FLAG_BRANCH_PREDICT 0x200
+#define TARGET_BRANCH_PREDICT (target_flags & TARGET_FLAG_BRANCH_PREDICT)
+
+/* Forces prototype and return promotions. */
+/* ??? This does not work. */
+#define TARGET_FLAG_CLEAN_LINKAGE 0x400
+#define TARGET_CLEAN_LINKAGE (target_flags & TARGET_FLAG_CLEAN_LINKAGE)
+
+/* For compatibility with iC960 v3.0. */
+#define TARGET_FLAG_IC_COMPAT3_0 0x800
+#define TARGET_IC_COMPAT3_0 (target_flags & TARGET_FLAG_IC_COMPAT3_0)
+
+/* For compatibility with iC960 v2.0. */
+#define TARGET_FLAG_IC_COMPAT2_0 0x1000
+#define TARGET_IC_COMPAT2_0 (target_flags & TARGET_FLAG_IC_COMPAT2_0)
+
+/* If no unaligned accesses are to be permitted. */
+#define TARGET_FLAG_STRICT_ALIGN 0x2000
+#define TARGET_STRICT_ALIGN (target_flags & TARGET_FLAG_STRICT_ALIGN)
+
+/* For compatibility with iC960 assembler. */
+#define TARGET_FLAG_ASM_COMPAT 0x4000
+#define TARGET_ASM_COMPAT (target_flags & TARGET_FLAG_ASM_COMPAT)
+
+/* For compatibility with the gcc960 v1.2 compiler. Use the old structure
+ alignment rules. Also, turns on STRICT_ALIGNMENT. */
+#define TARGET_FLAG_OLD_ALIGN 0x8000
+#define TARGET_OLD_ALIGN (target_flags & TARGET_FLAG_OLD_ALIGN)
+
+/* Nonzero if long doubles are to be 64 bits. Useful for soft-float targets
+ if 80 bit long double support is missing. */
+#define TARGET_FLAG_LONG_DOUBLE_64 0x10000
+#define TARGET_LONG_DOUBLE_64 (target_flags & TARGET_FLAG_LONG_DOUBLE_64)
+
+extern int target_flags;
+
+/* Macro to define tables used to set the flags.
+ This is a list in braces of pairs in braces,
+ each pair being { "NAME", VALUE }
+ where VALUE is the bits to set or minus the bits to clear.
+ An empty string NAME is used to identify the default VALUE. */
+
+/* ??? Not all ten of these architecture variations actually exist, but I
+ am not sure which are real and which aren't. */
+
+#define TARGET_SWITCHES \
+ { {"sa", (TARGET_FLAG_K_SERIES|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate SA code")}, \
+ {"sb", (TARGET_FLAG_NUMERICS|TARGET_FLAG_K_SERIES| \
+ TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate SB code")}, \
+/* {"sc", (TARGET_FLAG_NUMERICS|TARGET_FLAG_PROTECTED| \
+ TARGET_FLAG_MC|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate SC code")}, */ \
+ {"ka", (TARGET_FLAG_K_SERIES|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate KA code")}, \
+ {"kb", (TARGET_FLAG_NUMERICS|TARGET_FLAG_K_SERIES| \
+ TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate KB code")}, \
+/* {"kc", (TARGET_FLAG_NUMERICS|TARGET_FLAG_PROTECTED| \
+ TARGET_FLAG_MC|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate KC code")}, */ \
+ {"ja", (TARGET_FLAG_K_SERIES|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate JA code")}, \
+ {"jd", (TARGET_FLAG_K_SERIES|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate JD code")}, \
+ {"jf", (TARGET_FLAG_NUMERICS|TARGET_FLAG_K_SERIES| \
+ TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate JF code")}, \
+ {"rp", (TARGET_FLAG_K_SERIES|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("generate RP code")}, \
+ {"mc", (TARGET_FLAG_NUMERICS|TARGET_FLAG_PROTECTED| \
+ TARGET_FLAG_MC|TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Generate MC code")}, \
+ {"ca", (TARGET_FLAG_C_SERIES|TARGET_FLAG_BRANCH_PREDICT| \
+ TARGET_FLAG_CODE_ALIGN|TARGET_FLAG_COMPLEX_ADDR),\
+ N_("Generate CA code")}, \
+/* {"cb", (TARGET_FLAG_NUMERICS|TARGET_FLAG_C_SERIES| \
+ TARGET_FLAG_BRANCH_PREDICT|TARGET_FLAG_CODE_ALIGN),\
+ N_("Generate CB code")}, \
+ {"cc", (TARGET_FLAG_NUMERICS|TARGET_FLAG_PROTECTED| \
+ TARGET_FLAG_C_SERIES|TARGET_FLAG_BRANCH_PREDICT|\
+ TARGET_FLAG_CODE_ALIGN), \
+ N_("Generate CC code")}, */ \
+ {"cf", (TARGET_FLAG_C_SERIES|TARGET_FLAG_BRANCH_PREDICT| \
+ TARGET_FLAG_CODE_ALIGN|TARGET_FLAG_COMPLEX_ADDR),\
+ N_("Generate CF code")}, \
+ {"numerics", (TARGET_FLAG_NUMERICS), \
+ N_("Use hardware floating point instructions")}, \
+ {"soft-float", -(TARGET_FLAG_NUMERICS), \
+ N_("Use software floating point")}, \
+ {"leaf-procedures", TARGET_FLAG_LEAFPROC, \
+ N_("Use alternate leaf function entries")}, \
+ {"no-leaf-procedures", -(TARGET_FLAG_LEAFPROC), \
+ N_("Do not use alternate leaf function entries")}, \
+ {"tail-call", TARGET_FLAG_TAILCALL, \
+ N_("Perform tail call optimization")}, \
+ {"no-tail-call", -(TARGET_FLAG_TAILCALL), \
+ N_("Do not perform tail call optimization")}, \
+ {"complex-addr", TARGET_FLAG_COMPLEX_ADDR, \
+ N_("Use complex addressing modes")}, \
+ {"no-complex-addr", -(TARGET_FLAG_COMPLEX_ADDR), \
+ N_("Do not use complex addressing modes")}, \
+ {"code-align", TARGET_FLAG_CODE_ALIGN, \
+ N_("Align code to 8 byte boundary")}, \
+ {"no-code-align", -(TARGET_FLAG_CODE_ALIGN), \
+ N_("Do not align code to 8 byte boundary")}, \
+/* {"clean-linkage", (TARGET_FLAG_CLEAN_LINKAGE), \
+ N_("Force use of prototypes")}, \
+ {"no-clean-linkage", -(TARGET_FLAG_CLEAN_LINKAGE), \
+ N_("Do not force use of prototypes")}, */ \
+ {"ic-compat", TARGET_FLAG_IC_COMPAT2_0, \
+ N_("Enable compatibility with iC960 v2.0")}, \
+ {"ic2.0-compat", TARGET_FLAG_IC_COMPAT2_0, \
+ N_("Enable compatibility with iC960 v2.0")}, \
+ {"ic3.0-compat", TARGET_FLAG_IC_COMPAT3_0, \
+ N_("Enable compatibility with iC960 v3.0")}, \
+ {"asm-compat", TARGET_FLAG_ASM_COMPAT, \
+ N_("Enable compatibility with ic960 assembler")}, \
+ {"intel-asm", TARGET_FLAG_ASM_COMPAT, \
+ N_("Enable compatibility with ic960 assembler")}, \
+ {"strict-align", TARGET_FLAG_STRICT_ALIGN, \
+ N_("Do not permit unaligned accesses")}, \
+ {"no-strict-align", -(TARGET_FLAG_STRICT_ALIGN), \
+ N_("Permit unaligned accesses")}, \
+ {"old-align", (TARGET_FLAG_OLD_ALIGN|TARGET_FLAG_STRICT_ALIGN), \
+ N_("Layout types like Intel's v1.3 gcc")}, \
+ {"no-old-align", -(TARGET_FLAG_OLD_ALIGN|TARGET_FLAG_STRICT_ALIGN), \
+ N_("Do not layout types like Intel's v1.3 gcc")}, \
+ {"long-double-64", TARGET_FLAG_LONG_DOUBLE_64, \
+ N_("Use 64 bit long doubles")}, \
+ {"link-relax", 0, \
+ N_("Enable linker relaxation")}, \
+ {"no-link-relax", 0, \
+ N_("Do not enable linker relaxation")}, \
+ SUBTARGET_SWITCHES \
+ { "", TARGET_DEFAULT, \
+ NULL}}
+
+/* This are meant to be redefined in the host dependent files */
+#define SUBTARGET_SWITCHES
+
+/* Override conflicting target switch options.
+ Doesn't actually detect if more than one -mARCH option is given, but
+ does handle the case of two blatantly conflicting -mARCH options. */
+#define OVERRIDE_OPTIONS i960_initialize ()
+
+/* Don't enable anything by default. The user is expected to supply a -mARCH
+ option. If none is given, then -mka is added by CC1_SPEC. */
+#define TARGET_DEFAULT 0
+
+/* Target machine storage layout. */
+
+/* Define this if most significant bit is lowest numbered
+ in instructions that operate on numbered bit-fields. */
+#define BITS_BIG_ENDIAN 0
+
+/* Define this if most significant byte of a word is the lowest numbered.
+ The i960 case be either big endian or little endian. We only support
+ little endian, which is the most common. */
+#define BYTES_BIG_ENDIAN 0
+
+/* Define this if most significant word of a multiword number is lowest
+ numbered. */
+#define WORDS_BIG_ENDIAN 0
+
+/* Bitfields cannot cross word boundaries. */
+#define BITFIELD_NBYTES_LIMITED 1
+
+/* Width of a word, in units (bytes). */
+#define UNITS_PER_WORD 4
+
+/* Width in bits of a long double. */
+#define LONG_DOUBLE_TYPE_SIZE (TARGET_LONG_DOUBLE_64 ? 64 : 128)
+#define MAX_LONG_DOUBLE_TYPE_SIZE 128
+
+/* Define this to set long double type size to use in libgcc2.c, which can
+ not depend on target_flags. */
+#if defined(__LONG_DOUBLE_64__)
+#define LIBGCC2_LONG_DOUBLE_TYPE_SIZE 64
+#else
+#define LIBGCC2_LONG_DOUBLE_TYPE_SIZE 128
+#endif
+
+/* Allocation boundary (in *bits*) for storing pointers in memory. */
+#define POINTER_BOUNDARY 32
+
+/* Allocation boundary (in *bits*) for storing arguments in argument list. */
+#define PARM_BOUNDARY 32
+
+/* Boundary (in *bits*) on which stack pointer should be aligned. */
+#define STACK_BOUNDARY 128
+
+/* Allocation boundary (in *bits*) for the code of a function. */
+#define FUNCTION_BOUNDARY 128
+
+/* Alignment of field after `int : 0' in a structure. */
+#define EMPTY_FIELD_BOUNDARY 32
+
+/* This makes zero-length anonymous fields lay the next field
+ at a word boundary. It also makes the whole struct have
+ at least word alignment if there are any bitfields at all. */
+#define PCC_BITFIELD_TYPE_MATTERS 1
+
+/* Every structure's size must be a multiple of this. */
+#define STRUCTURE_SIZE_BOUNDARY 8
+
+/* No data type wants to be aligned rounder than this.
+ Extended precision floats gets 4-word alignment. */
+#define BIGGEST_ALIGNMENT 128
+
+/* Define this if move instructions will actually fail to work
+ when given unaligned data.
+ 80960 will work even with unaligned data, but it is slow. */
+#define STRICT_ALIGNMENT TARGET_STRICT_ALIGN
+
+/* Specify alignment for string literals (which might be higher than the
+ base type's minimal alignment requirement. This allows strings to be
+ aligned on word boundaries, and optimizes calls to the str* and mem*
+ library functions. */
+#define CONSTANT_ALIGNMENT(EXP, ALIGN) \
+ (TREE_CODE (EXP) == STRING_CST \
+ && i960_object_bytes_bitalign (int_size_in_bytes (TREE_TYPE (EXP))) > (int)(ALIGN) \
+ ? i960_object_bytes_bitalign (int_size_in_bytes (TREE_TYPE (EXP))) \
+ : (int)(ALIGN))
+
+/* Macros to determine size of aggregates (structures and unions
+ in C). Normally, these may be defined to simply return the maximum
+ alignment and simple rounded-up size, but on some machines (like
+ the i960), the total size of a structure is based on a non-trivial
+ rounding method. */
+
+#define ROUND_TYPE_ALIGN(TYPE, COMPUTED, SPECIFIED) \
+ i960_round_align (MAX ((COMPUTED), (SPECIFIED)), TYPE)
+
+/* Standard register usage. */
+
+/* Number of actual hardware registers.
+ The hardware registers are assigned numbers for the compiler
+ from 0 to just below FIRST_PSEUDO_REGISTER.
+ All registers that the compiler knows about must be given numbers,
+ even those that are not normally considered general registers.
+
+ Registers 0-15 are the global registers (g0-g15).
+ Registers 16-31 are the local registers (r0-r15).
+ Register 32-35 are the fp registers (fp0-fp3).
+ Register 36 is the condition code register.
+ Register 37 is unused. */
+
+#define FIRST_PSEUDO_REGISTER 38
+
+/* 1 for registers that have pervasive standard uses and are not available
+ for the register allocator. On 80960, this includes the frame pointer
+ (g15), the previous FP (r0), the stack pointer (r1), the return
+ instruction pointer (r2), and the argument pointer (g14). */
+#define FIXED_REGISTERS \
+ {0, 0, 0, 0, 0, 0, 0, 0, \
+ 0, 0, 0, 0, 0, 0, 1, 1, \
+ 1, 1, 1, 0, 0, 0, 0, 0, \
+ 0, 0, 0, 0, 0, 0, 0, 0, \
+ 0, 0, 0, 0, 1, 1}
+
+/* 1 for registers not available across function calls.
+ These must include the FIXED_REGISTERS and also any
+ registers that can be used without being saved.
+ The latter must include the registers where values are returned
+ and the register where structure-value addresses are passed.
+ Aside from that, you can include as many other registers as you like. */
+
+/* On the 80960, note that:
+ g0..g3 are used for return values,
+ g0..g7 may always be used for parameters,
+ g8..g11 may be used for parameters, but are preserved if they aren't,
+ g12 is the static chain if needed, otherwise is preserved
+ g13 is the struct return ptr if used, or temp, but may be trashed,
+ g14 is the leaf return ptr or the arg block ptr otherwise zero,
+ must be reset to zero before returning if it was used,
+ g15 is the frame pointer,
+ r0 is the previous FP,
+ r1 is the stack pointer,
+ r2 is the return instruction pointer,
+ r3-r15 are always available,
+ r3 is clobbered by calls in functions that use the arg pointer
+ r4-r11 may be clobbered by the mcount call when profiling
+ r4-r15 if otherwise unused may be used for preserving global registers
+ fp0..fp3 are never available. */
+#define CALL_USED_REGISTERS \
+ {1, 1, 1, 1, 1, 1, 1, 1, \
+ 0, 0, 0, 0, 0, 1, 1, 1, \
+ 1, 1, 1, 0, 0, 0, 0, 0, \
+ 0, 0, 0, 0, 0, 0, 0, 0, \
+ 1, 1, 1, 1, 1, 1}
+
+/* If no fp unit, make all of the fp registers fixed so that they can't
+ be used. */
+#define CONDITIONAL_REGISTER_USAGE \
+ if (! TARGET_NUMERICS) { \
+ fixed_regs[32] = fixed_regs[33] = fixed_regs[34] = fixed_regs[35] = 1;\
+ } \
+
+/* Return number of consecutive hard regs needed starting at reg REGNO
+ to hold something of mode MODE.
+ This is ordinarily the length in words of a value of mode MODE
+ but can be less for certain modes in special long registers.
+
+ On 80960, ordinary registers hold 32 bits worth, but can be ganged
+ together to hold double or extended precision floating point numbers,
+ and the floating point registers hold any size floating point number */
+#define HARD_REGNO_NREGS(REGNO, MODE) \
+ ((REGNO) < 32 \
+ ? (((MODE) == VOIDmode) \
+ ? 1 : ((GET_MODE_SIZE (MODE) + UNITS_PER_WORD - 1) / UNITS_PER_WORD)) \
+ : ((REGNO) < FIRST_PSEUDO_REGISTER) ? 1 : 0)
+
+/* Value is 1 if hard register REGNO can hold a value of machine-mode MODE.
+ On 80960, the cpu registers can hold any mode but the float registers
+ can only hold SFmode, DFmode, or TFmode. */
+#define HARD_REGNO_MODE_OK(REGNO, MODE) hard_regno_mode_ok ((REGNO), (MODE))
+
+/* Value is 1 if it is a good idea to tie two pseudo registers
+ when one has mode MODE1 and one has mode MODE2.
+ If HARD_REGNO_MODE_OK could produce different values for MODE1 and MODE2,
+ for any hard reg, then this must be 0 for correct output. */
+
+#define MODES_TIEABLE_P(MODE1, MODE2) \
+ ((MODE1) == (MODE2) || GET_MODE_CLASS (MODE1) == GET_MODE_CLASS (MODE2))
+
+/* Specify the registers used for certain standard purposes.
+ The values of these macros are register numbers. */
+
+/* 80960 pc isn't overloaded on a register that the compiler knows about. */
+/* #define PC_REGNUM */
+
+/* Register to use for pushing function arguments. */
+#define STACK_POINTER_REGNUM 17
+
+/* Actual top-of-stack address is same as
+ the contents of the stack pointer register. */
+#define STACK_POINTER_OFFSET (-current_function_outgoing_args_size)
+
+/* Base register for access to local variables of the function. */
+#define FRAME_POINTER_REGNUM 15
+
+/* Value should be nonzero if functions must have frame pointers.
+ Zero means the frame pointer need not be set up (and parms
+ may be accessed via the stack pointer) in functions that seem suitable.
+ This is computed in `reload', in reload1.c. */
+/* ??? It isn't clear to me why this is here. Perhaps because of a bug (since
+ fixed) in the definition of INITIAL_FRAME_POINTER_OFFSET which would have
+ caused this to fail. */
+/* ??? Must check current_function_has_nonlocal_goto, otherwise frame pointer
+ elimination messes up nonlocal goto sequences. I think this works for other
+ targets because they use indirect jumps for the return which disables fp
+ elimination. */
+#define FRAME_POINTER_REQUIRED \
+ (! leaf_function_p () || current_function_has_nonlocal_goto)
+
+/* Definitions for register eliminations.
+
+ This is an array of structures. Each structure initializes one pair
+ of eliminable registers. The "from" register number is given first,
+ followed by "to". Eliminations of the same "from" register are listed
+ in order of preference.. */
+
+#define ELIMINABLE_REGS {{FRAME_POINTER_REGNUM, STACK_POINTER_REGNUM}}
+
+/* Given FROM and TO register numbers, say whether this elimination is allowed.
+ Frame pointer elimination is automatically handled. */
+#define CAN_ELIMINATE(FROM, TO) 1
+
+/* Define the offset between two registers, one to be eliminated, and
+ the other its replacement, at the start of a routine.
+
+ Since the stack grows upward on the i960, this must be a negative number.
+ This includes the 64 byte hardware register save area and the size of
+ the frame. */
+
+#define INITIAL_ELIMINATION_OFFSET(FROM, TO, OFFSET) \
+ do { (OFFSET) = - (64 + compute_frame_size (get_frame_size ())); } while (0)
+
+/* Base register for access to arguments of the function. */
+#define ARG_POINTER_REGNUM 14
+
+/* Register in which static-chain is passed to a function.
+ On i960, we use g12. We can't use any local register, because we need
+ a register that can be set before a call or before a jump. */
+#define STATIC_CHAIN_REGNUM 12
+
+/* Functions which return large structures get the address
+ to place the wanted value at in g13. */
+
+#define STRUCT_VALUE_REGNUM 13
+
+/* The order in which to allocate registers. */
+
+#define REG_ALLOC_ORDER \
+{ 4, 5, 6, 7, 0, 1, 2, 3, 13, /* g4, g5, g6, g7, g0, g1, g2, g3, g13 */ \
+ 20, 21, 22, 23, 24, 25, 26, 27,/* r4, r5, r6, r7, r8, r9, r10, r11 */ \
+ 28, 29, 30, 31, 19, 8, 9, 10, /* r12, r13, r14, r15, r3, g8, g9, g10 */ \
+ 11, 12, /* g11, g12 */ \
+ 32, 33, 34, 35, /* fp0, fp1, fp2, fp3 */ \
+ /* We can't actually allocate these. */ \
+ 16, 17, 18, 14, 15, 36, 37} /* r0, r1, r2, g14, g15, cc */
+
+/* Define the classes of registers for register constraints in the
+ machine description. Also define ranges of constants.
+
+ One of the classes must always be named ALL_REGS and include all hard regs.
+ If there is more than one class, another class must be named NO_REGS
+ and contain no registers.
+
+ The name GENERAL_REGS must be the name of a class (or an alias for
+ another name such as ALL_REGS). This is the class of registers
+ that is allowed by "g" or "r" in a register constraint.
+ Also, registers outside this class are allocated only when
+ instructions express preferences for them.
+
+ The classes must be numbered in nondecreasing order; that is,
+ a larger-numbered class must never be contained completely
+ in a smaller-numbered class.
+
+ For any two classes, it is very desirable that there be another
+ class that represents their union. */
+
+/* The 80960 has four kinds of registers, global, local, floating point,
+ and condition code. The cc register is never allocated, so no class
+ needs to be defined for it. */
+
+enum reg_class { NO_REGS, GLOBAL_REGS, LOCAL_REGS, LOCAL_OR_GLOBAL_REGS,
+ FP_REGS, ALL_REGS, LIM_REG_CLASSES };
+
+/* 'r' includes floating point registers if TARGET_NUMERICS. 'd' never
+ does. */
+#define GENERAL_REGS ((TARGET_NUMERICS) ? ALL_REGS : LOCAL_OR_GLOBAL_REGS)
+
+#define N_REG_CLASSES (int) LIM_REG_CLASSES
+
+/* Give names of register classes as strings for dump file. */
+
+#define REG_CLASS_NAMES \
+{ "NO_REGS", "GLOBAL_REGS", "LOCAL_REGS", "LOCAL_OR_GLOBAL_REGS", \
+ "FP_REGS", "ALL_REGS" }
+
+/* Define which registers fit in which classes.
+ This is an initializer for a vector of HARD_REG_SET
+ of length N_REG_CLASSES. */
+
+#define REG_CLASS_CONTENTS \
+{ {0, 0}, {0x0ffff, 0}, {0xffff0000, 0}, {-1,0}, {0, -1}, {-1,-1}}
+
+/* The same information, inverted:
+ Return the class number of the smallest class containing
+ reg number REGNO. This could be a conditional expression
+ or could index an array. */
+
+#define REGNO_REG_CLASS(REGNO) \
+ ((REGNO) < 16 ? GLOBAL_REGS \
+ : (REGNO) < 32 ? LOCAL_REGS \
+ : (REGNO) < 36 ? FP_REGS \
+ : NO_REGS)
+
+/* The class value for index registers, and the one for base regs.
+ There is currently no difference between base and index registers on the
+ i960, but this distinction may one day be useful. */
+#define INDEX_REG_CLASS LOCAL_OR_GLOBAL_REGS
+#define BASE_REG_CLASS LOCAL_OR_GLOBAL_REGS
+
+/* Get reg_class from a letter such as appears in the machine description.
+ 'f' is a floating point register (fp0..fp3)
+ 'l' is a local register (r0-r15)
+ 'b' is a global register (g0-g15)
+ 'd' is any local or global register
+ 'r' or 'g' are pre-defined to the class GENERAL_REGS. */
+/* 'l' and 'b' are probably never used. Note that 'd' and 'r' are *not*
+ the same thing, since 'r' may include the fp registers. */
+#define REG_CLASS_FROM_LETTER(C) \
+ (((C) == 'f') && (TARGET_NUMERICS) ? FP_REGS : ((C) == 'l' ? LOCAL_REGS : \
+ (C) == 'b' ? GLOBAL_REGS : ((C) == 'd' ? LOCAL_OR_GLOBAL_REGS : NO_REGS)))
+
+/* The letters I, J, K, L and M in a register constraint string
+ can be used to stand for particular ranges of immediate operands.
+ This macro defines what the ranges are.
+ C is the letter, and VALUE is a constant value.
+ Return 1 if VALUE is in the range specified by C.
+
+ For 80960:
+ 'I' is used for literal values 0..31
+ 'J' means literal 0
+ 'K' means 0..-31. */
+
+#define CONST_OK_FOR_LETTER_P(VALUE, C) \
+ ((C) == 'I' ? (((unsigned) (VALUE)) <= 31) \
+ : (C) == 'J' ? ((VALUE) == 0) \
+ : (C) == 'K' ? ((VALUE) >= -31 && (VALUE) <= 0) \
+ : (C) == 'M' ? ((VALUE) >= -32 && (VALUE) <= 0) \
+ : 0)
+
+/* Similar, but for floating constants, and defining letters G and H.
+ Here VALUE is the CONST_DOUBLE rtx itself.
+ For the 80960, G is 0.0 and H is 1.0. */
+
+#define CONST_DOUBLE_OK_FOR_LETTER_P(VALUE, C) \
+ ((TARGET_NUMERICS) && \
+ (((C) == 'G' && (VALUE) == CONST0_RTX (GET_MODE (VALUE))) \
+ || ((C) == 'H' && ((VALUE) == CONST1_RTX (GET_MODE (VALUE))))))
+
+/* Given an rtx X being reloaded into a reg required to be
+ in class CLASS, return the class of reg to actually use.
+ In general this is just CLASS; but on some machines
+ in some cases it is preferable to use a more restrictive class. */
+
+/* On 960, can't load constant into floating-point reg except
+ 0.0 or 1.0.
+
+ Any hard reg is ok as a src operand of a reload insn. */
+
+#define PREFERRED_RELOAD_CLASS(X,CLASS) \
+ (GET_CODE (X) == REG && REGNO (X) < FIRST_PSEUDO_REGISTER \
+ ? (CLASS) \
+ : ((CLASS) == FP_REGS && CONSTANT_P (X) \
+ && (X) != CONST0_RTX (DFmode) && (X) != CONST1_RTX (DFmode)\
+ && (X) != CONST0_RTX (SFmode) && (X) != CONST1_RTX (SFmode)\
+ ? NO_REGS \
+ : (CLASS) == ALL_REGS ? LOCAL_OR_GLOBAL_REGS : (CLASS)))
+
+#define SECONDARY_RELOAD_CLASS(CLASS,MODE,IN) \
+ secondary_reload_class (CLASS, MODE, IN)
+
+/* Return the maximum number of consecutive registers
+ needed to represent mode MODE in a register of class CLASS. */
+/* On 80960, this is the size of MODE in words,
+ except in the FP regs, where a single reg is always enough. */
+#define CLASS_MAX_NREGS(CLASS, MODE) \
+ ((CLASS) == FP_REGS ? 1 : HARD_REGNO_NREGS (0, (MODE)))
+
+/* Stack layout; function entry, exit and calling. */
+
+/* Define this if pushing a word on the stack
+ makes the stack pointer a smaller address. */
+/* #define STACK_GROWS_DOWNWARD */
+
+/* Define this if the nominal address of the stack frame
+ is at the high-address end of the local variables;
+ that is, each additional local variable allocated
+ goes at a more negative offset in the frame. */
+/* #define FRAME_GROWS_DOWNWARD */
+
+/* Offset within stack frame to start allocating local variables at.
+ If FRAME_GROWS_DOWNWARD, this is the offset to the END of the
+ first local allocated. Otherwise, it is the offset to the BEGINNING
+ of the first local allocated.
+
+ The i960 has a 64 byte register save area, plus possibly some extra
+ bytes allocated for varargs functions. */
+#define STARTING_FRAME_OFFSET 64
+
+/* If we generate an insn to push BYTES bytes,
+ this says how many the stack pointer really advances by.
+ On 80960, don't define this because there are no push insns. */
+/* #define PUSH_ROUNDING(BYTES) BYTES */
+
+/* Offset of first parameter from the argument pointer register value. */
+#define FIRST_PARM_OFFSET(FNDECL) 0
+
+/* When a parameter is passed in a register, no stack space is
+ allocated for it. However, when args are passed in the
+ stack, space is allocated for every register parameter. */
+#define MAYBE_REG_PARM_STACK_SPACE 48
+#define FINAL_REG_PARM_STACK_SPACE(CONST_SIZE, VAR_SIZE) \
+ i960_final_reg_parm_stack_space (CONST_SIZE, VAR_SIZE);
+#define REG_PARM_STACK_SPACE(DECL) i960_reg_parm_stack_space (DECL)
+#define OUTGOING_REG_PARM_STACK_SPACE
+
+/* Keep the stack pointer constant throughout the function. */
+#define ACCUMULATE_OUTGOING_ARGS 1
+
+/* Value is 1 if returning from a function call automatically
+ pops the arguments described by the number-of-args field in the call.
+ FUNDECL is the declaration node of the function (as a tree),
+ FUNTYPE is the data type of the function (as a tree),
+ or for a library call it is an identifier node for the subroutine name. */
+
+#define RETURN_POPS_ARGS(FUNDECL,FUNTYPE,SIZE) 0
+
+/* Define how to find the value returned by a library function
+ assuming the value has mode MODE. */
+
+#define LIBCALL_VALUE(MODE) gen_rtx_REG ((MODE), 0)
+
+/* 1 if N is a possible register number for a function value
+ as seen by the caller.
+ On 80960, returns are in g0..g3 */
+
+#define FUNCTION_VALUE_REGNO_P(N) ((N) == 0)
+
+/* 1 if N is a possible register number for function argument passing.
+ On 80960, parameters are passed in g0..g11 */
+
+#define FUNCTION_ARG_REGNO_P(N) ((N) < 12)
+
+/* Perform any needed actions needed for a function that is receiving a
+ variable number of arguments.
+
+ CUM is as above.
+
+ MODE and TYPE are the mode and type of the current parameter.
+
+ PRETEND_SIZE is a variable that should be set to the amount of stack
+ that must be pushed by the prolog to pretend that our caller pushed
+ it.
+
+ Normally, this macro will push all remaining incoming registers on the
+ stack and set PRETEND_SIZE to the length of the registers pushed. */
+
+#define SETUP_INCOMING_VARARGS(CUM,MODE,TYPE,PRETEND_SIZE,NO_RTL) \
+ i960_setup_incoming_varargs(&CUM,MODE,TYPE,&PRETEND_SIZE,NO_RTL)
+
+/* Implement `va_start' for varargs and stdarg. */
+#define EXPAND_BUILTIN_VA_START(valist, nextarg) \
+ i960_va_start (valist, nextarg)
+
+/* Implement `va_arg'. */
+#define EXPAND_BUILTIN_VA_ARG(valist, type) \
+ i960_va_arg (valist, type)
+
+/* Define a data type for recording info about an argument list
+ during the scan of that argument list. This data type should
+ hold all necessary information about the function itself
+ and about the args processed so far, enough to enable macros
+ such as FUNCTION_ARG to determine where the next arg should go.
+
+ On 80960, this is two integers, which count the number of register
+ parameters and the number of stack parameters seen so far. */
+
+struct cum_args { int ca_nregparms; int ca_nstackparms; };
+
+#define CUMULATIVE_ARGS struct cum_args
+
+/* Define the number of registers that can hold parameters.
+ This macro is used only in macro definitions below and/or i960.c. */
+#define NPARM_REGS 12
+
+/* Define how to round to the next parameter boundary.
+ This macro is used only in macro definitions below and/or i960.c. */
+#define ROUND_PARM(X, MULTIPLE_OF) \
+ ((((X) + (MULTIPLE_OF) - 1) / (MULTIPLE_OF)) * MULTIPLE_OF)
+
+/* Initialize a variable CUM of type CUMULATIVE_ARGS
+ for a call to a function whose data type is FNTYPE.
+ For a library call, FNTYPE is 0.
+
+ On 80960, the offset always starts at 0; the first parm reg is g0. */
+
+#define INIT_CUMULATIVE_ARGS(CUM, FNTYPE, LIBNAME, INDIRECT, N_NAMED_ARGS) \
+ ((CUM).ca_nregparms = 0, (CUM).ca_nstackparms = 0)
+
+/* Update the data in CUM to advance over an argument
+ of mode MODE and data type TYPE.
+ CUM should be advanced to align with the data type accessed and
+ also the size of that data type in # of regs.
+ (TYPE is null for libcalls where that information may not be available.) */
+
+#define FUNCTION_ARG_ADVANCE(CUM, MODE, TYPE, NAMED) \
+ i960_function_arg_advance(&CUM, MODE, TYPE, NAMED)
+
+/* Indicate the alignment boundary for an argument of the specified mode and
+ type. */
+#define FUNCTION_ARG_BOUNDARY(MODE, TYPE) \
+ (((TYPE) != 0) \
+ ? ((TYPE_ALIGN (TYPE) <= PARM_BOUNDARY) \
+ ? PARM_BOUNDARY \
+ : TYPE_ALIGN (TYPE)) \
+ : ((GET_MODE_ALIGNMENT (MODE) <= PARM_BOUNDARY) \
+ ? PARM_BOUNDARY \
+ : GET_MODE_ALIGNMENT (MODE)))
+
+/* Determine where to put an argument to a function.
+ Value is zero to push the argument on the stack,
+ or a hard register in which to store the argument.
+
+ MODE is the argument's machine mode.
+ TYPE is the data type of the argument (as a tree).
+ This is null for libcalls where that information may
+ not be available.
+ CUM is a variable of type CUMULATIVE_ARGS which gives info about
+ the preceding args and about the function being called.
+ NAMED is nonzero if this argument is a named parameter
+ (otherwise it is an extra parameter matching an ellipsis). */
+
+#define FUNCTION_ARG(CUM, MODE, TYPE, NAMED) \
+ i960_function_arg(&CUM, MODE, TYPE, NAMED)
+
+/* Define how to find the value returned by a function.
+ VALTYPE is the data type of the value (as a tree).
+ If the precise function being called is known, FUNC is its FUNCTION_DECL;
+ otherwise, FUNC is 0. */
+
+#define FUNCTION_VALUE(TYPE, FUNC) \
+ gen_rtx_REG (TYPE_MODE (TYPE), 0)
+
+/* Force aggregates and objects larger than 16 bytes to be returned in memory,
+ since we only have 4 registers available for return values. */
+
+#define RETURN_IN_MEMORY(TYPE) \
+ (TYPE_MODE (TYPE) == BLKmode || int_size_in_bytes (TYPE) > 16)
+
+/* Don't default to pcc-struct-return, because we have already specified
+ exactly how to return structures in the RETURN_IN_MEMORY macro. */
+#define DEFAULT_PCC_STRUCT_RETURN 0
+
+/* For an arg passed partly in registers and partly in memory,
+ this is the number of registers used.
+ This never happens on 80960. */
+
+#define FUNCTION_ARG_PARTIAL_NREGS(CUM, MODE, TYPE, NAMED) 0
+
+/* Output the label for a function definition.
+ This handles leaf functions and a few other things for the i960. */
+
+#define ASM_DECLARE_FUNCTION_NAME(FILE, NAME, DECL) \
+ i960_function_name_declare (FILE, NAME, DECL)
+
+/* Output assembler code to FILE to increment profiler label # LABELNO
+ for profiling a function entry. */
+
+#define FUNCTION_PROFILER(FILE, LABELNO) \
+ output_function_profiler ((FILE), (LABELNO));
+
+/* EXIT_IGNORE_STACK should be nonzero if, when returning from a function,
+ the stack pointer does not matter. The value is tested only in
+ functions that have frame pointers.
+ No definition is equivalent to always zero. */
+
+#define EXIT_IGNORE_STACK 1
+
+/* Addressing modes, and classification of registers for them. */
+
+/* Macros to check register numbers against specific register classes. */
+
+/* These assume that REGNO is a hard or pseudo reg number.
+ They give nonzero only if REGNO is a hard reg of the suitable class
+ or a pseudo reg currently allocated to a suitable hard reg.
+ Since they use reg_renumber, they are safe only once reg_renumber
+ has been allocated, which happens in local-alloc.c. */
+
+#define REGNO_OK_FOR_INDEX_P(REGNO) \
+ ((REGNO) < 32 || (unsigned) reg_renumber[REGNO] < 32)
+#define REGNO_OK_FOR_BASE_P(REGNO) \
+ ((REGNO) < 32 || (unsigned) reg_renumber[REGNO] < 32)
+#define REGNO_OK_FOR_FP_P(REGNO) \
+ ((REGNO) < 36 || (unsigned) reg_renumber[REGNO] < 36)
+
+/* Now macros that check whether X is a register and also,
+ strictly, whether it is in a specified class.
+
+ These macros are specific to the 960, and may be used only
+ in code for printing assembler insns and in conditions for
+ define_optimization. */
+
+/* 1 if X is an fp register. */
+
+#define FP_REG_P(X) (REGNO (X) >= 32 && REGNO (X) < 36)
+
+/* Maximum number of registers that can appear in a valid memory address. */
+#define MAX_REGS_PER_ADDRESS 2
+
+#define CONSTANT_ADDRESS_P(X) \
+ (GET_CODE (X) == LABEL_REF || GET_CODE (X) == SYMBOL_REF \
+ || GET_CODE (X) == CONST_INT || GET_CODE (X) == CONST \
+ || GET_CODE (X) == HIGH)
+
+/* LEGITIMATE_CONSTANT_P is nonzero if the constant value X
+ is a legitimate general operand.
+ It is given that X satisfies CONSTANT_P.
+
+ Anything but a CONST_DOUBLE can be made to work, excepting 0.0 and 1.0.
+
+ ??? This probably should be defined to 1. */
+
+#define LEGITIMATE_CONSTANT_P(X) \
+ ((GET_CODE (X) != CONST_DOUBLE) || fp_literal ((X), GET_MODE (X)))
+
+/* The macros REG_OK_FOR..._P assume that the arg is a REG rtx
+ and check its validity for a certain class.
+ We have two alternate definitions for each of them.
+ The usual definition accepts all pseudo regs; the other rejects
+ them unless they have been allocated suitable hard regs.
+ The symbol REG_OK_STRICT causes the latter definition to be used.
+
+ Most source files want to accept pseudo regs in the hope that
+ they will get allocated to the class that the insn wants them to be in.
+ Source files for reload pass need to be strict.
+ After reload, it makes no difference, since pseudo regs have
+ been eliminated by then. */
+
+#ifndef REG_OK_STRICT
+
+/* Nonzero if X is a hard reg that can be used as an index
+ or if it is a pseudo reg. */
+#define REG_OK_FOR_INDEX_P(X) \
+ (REGNO (X) < 32 || REGNO (X) >= FIRST_PSEUDO_REGISTER)
+/* Nonzero if X is a hard reg that can be used as a base reg
+ or if it is a pseudo reg. */
+#define REG_OK_FOR_BASE_P(X) \
+ (REGNO (X) < 32 || REGNO (X) >= FIRST_PSEUDO_REGISTER)
+
+#define REG_OK_FOR_INDEX_P_STRICT(X) REGNO_OK_FOR_INDEX_P (REGNO (X))
+#define REG_OK_FOR_BASE_P_STRICT(X) REGNO_OK_FOR_BASE_P (REGNO (X))
+
+#else
+
+/* Nonzero if X is a hard reg that can be used as an index. */
+#define REG_OK_FOR_INDEX_P(X) REGNO_OK_FOR_INDEX_P (REGNO (X))
+/* Nonzero if X is a hard reg that can be used as a base reg. */
+#define REG_OK_FOR_BASE_P(X) REGNO_OK_FOR_BASE_P (REGNO (X))
+
+#endif
+
+/* GO_IF_LEGITIMATE_ADDRESS recognizes an RTL expression
+ that is a valid memory address for an instruction.
+ The MODE argument is the machine mode for the MEM expression
+ that wants to use this address.
+
+ On 80960, legitimate addresses are:
+ base ld (g0),r0
+ disp (12 or 32 bit) ld foo,r0
+ base + index ld (g0)[g1*1],r0
+ base + displ ld 0xf00(g0),r0
+ base + index*scale + displ ld 0xf00(g0)[g1*4],r0
+ index*scale + base ld (g0)[g1*4],r0
+ index*scale + displ ld 0xf00[g1*4],r0
+ index*scale ld [g1*4],r0
+ index + base + displ ld 0xf00(g0)[g1*1],r0
+
+ In each case, scale can be 1, 2, 4, 8, or 16. */
+
+/* Returns 1 if the scale factor of an index term is valid. */
+#define SCALE_TERM_P(X) \
+ (GET_CODE (X) == CONST_INT \
+ && (INTVAL (X) == 1 || INTVAL (X) == 2 || INTVAL (X) == 4 \
+ || INTVAL(X) == 8 || INTVAL (X) == 16))
+
+
+#ifdef REG_OK_STRICT
+#define GO_IF_LEGITIMATE_ADDRESS(MODE, X, ADDR) \
+ { if (legitimate_address_p (MODE, X, 1)) goto ADDR; }
+#else
+#define GO_IF_LEGITIMATE_ADDRESS(MODE, X, ADDR) \
+ { if (legitimate_address_p (MODE, X, 0)) goto ADDR; }
+#endif
+
+/* Try machine-dependent ways of modifying an illegitimate address
+ to be legitimate. If we find one, return the new, valid address.
+ This macro is used in only one place: `memory_address' in explow.c.
+
+ OLDX is the address as it was before break_out_memory_refs was called.
+ In some cases it is useful to look at this to decide what needs to be done.
+
+ MODE and WIN are passed so that this macro can use
+ GO_IF_LEGITIMATE_ADDRESS.
+
+ It is always safe for this macro to do nothing. It exists to recognize
+ opportunities to optimize the output. */
+
+/* On 80960, convert non-canonical addresses to canonical form. */
+
+#define LEGITIMIZE_ADDRESS(X, OLDX, MODE, WIN) \
+{ rtx orig_x = (X); \
+ (X) = legitimize_address (X, OLDX, MODE); \
+ if ((X) != orig_x && memory_address_p (MODE, X)) \
+ goto WIN; }
+
+/* Go to LABEL if ADDR (a legitimate address expression)
+ has an effect that depends on the machine mode it is used for.
+ On the 960 this is never true. */
+
+#define GO_IF_MODE_DEPENDENT_ADDRESS(ADDR,LABEL)
+
+/* Specify the machine mode that this machine uses
+ for the index in the tablejump instruction. */
+#define CASE_VECTOR_MODE SImode
+
+/* Define as C expression which evaluates to nonzero if the tablejump
+ instruction expects the table to contain offsets from the address of the
+ table.
+ Do not define this if the table should contain absolute addresses. */
+/* #define CASE_VECTOR_PC_RELATIVE 1 */
+
+/* Define this as 1 if `char' should by default be signed; else as 0. */
+#define DEFAULT_SIGNED_CHAR 0
+
+/* Max number of bytes we can move from memory to memory
+ in one reasonably fast instruction. */
+#define MOVE_MAX 16
+
+/* Define if operations between registers always perform the operation
+ on the full register even if a narrower mode is specified. */
+#define WORD_REGISTER_OPERATIONS
+
+/* Define if loading in MODE, an integral mode narrower than BITS_PER_WORD
+ will either zero-extend or sign-extend. The value of this macro should
+ be the code that says which one of the two operations is implicitly
+ done, NIL if none. */
+#define LOAD_EXTEND_OP(MODE) ZERO_EXTEND
+
+/* Nonzero if access to memory by bytes is no faster than for words.
+ Value changed to 1 after reports of poor bit-field code with g++.
+ Indications are that code is usually as good, sometimes better. */
+
+#define SLOW_BYTE_ACCESS 1
+
+/* Define this to be nonzero if shift instructions ignore all but the low-order
+ few bits. */
+#define SHIFT_COUNT_TRUNCATED 0
+
+/* Value is 1 if truncating an integer of INPREC bits to OUTPREC bits
+ is done just by pretending it is already truncated. */
+#define TRULY_NOOP_TRUNCATION(OUTPREC, INPREC) 1
+
+/* Specify the machine mode that pointers have.
+ After generation of rtl, the compiler makes no further distinction
+ between pointers and any other objects of this machine mode. */
+#define Pmode SImode
+
+/* Specify the widest mode that BLKmode objects can be promoted to */
+#define MAX_FIXED_MODE_SIZE GET_MODE_BITSIZE (TImode)
+
+/* These global variables are used to pass information between
+ cc setter and cc user at insn emit time. */
+
+extern struct rtx_def *i960_compare_op0, *i960_compare_op1;
+
+/* Given a comparison code (EQ, NE, etc.) and the first operand of a COMPARE,
+ return the mode to be used for the comparison. For floating-point, CCFPmode
+ should be used. CC_NOOVmode should be used when the first operand is a
+ PLUS, MINUS, or NEG. CCmode should be used when no special processing is
+ needed. */
+#define SELECT_CC_MODE(OP,X,Y) select_cc_mode (OP, X)
+
+/* A function address in a call instruction is a byte address
+ (for indexing purposes) so give the MEM rtx a byte's mode. */
+#define FUNCTION_MODE SImode
+
+/* Define this if addresses of constant functions
+ shouldn't be put through pseudo regs where they can be cse'd.
+ Desirable on machines where ordinary constants are expensive
+ but a CALL with constant address is cheap. */
+#define NO_FUNCTION_CSE
+
+/* Use memcpy, etc. instead of bcopy. */
+
+#ifndef WIND_RIVER
+#define TARGET_MEM_FUNCTIONS 1
+#endif
+
+/* Control the assembler format that we output. */
+
+/* Output to assembler file text saying following lines
+ may contain character constants, extra white space, comments, etc. */
+
+#define ASM_APP_ON ""
+
+/* Output to assembler file text saying following lines
+ no longer contain unusual constructs. */
+
+#define ASM_APP_OFF ""
+
+/* Output before read-only data. */
+
+#define TEXT_SECTION_ASM_OP "\t.text"
+
+/* Output before writable data. */
+
+#define DATA_SECTION_ASM_OP "\t.data"
+
+/* How to refer to registers in assembler output.
+ This sequence is indexed by compiler's hard-register-number (see above). */
+
+#define REGISTER_NAMES { \
+ "g0", "g1", "g2", "g3", "g4", "g5", "g6", "g7", \
+ "g8", "g9", "g10", "g11", "g12", "g13", "g14", "fp", \
+ "pfp","sp", "rip", "r3", "r4", "r5", "r6", "r7", \
+ "r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15", \
+ "fp0","fp1","fp2", "fp3", "cc", "fake" }
+
+/* How to renumber registers for dbx and gdb.
+ In the 960 encoding, g0..g15 are registers 16..31. */
+
+#define DBX_REGISTER_NUMBER(REGNO) \
+ (((REGNO) < 16) ? (REGNO) + 16 \
+ : (((REGNO) > 31) ? (REGNO) : (REGNO) - 16))
+
+/* Don't emit dbx records longer than this. This is an arbitrary value. */
+#define DBX_CONTIN_LENGTH 1500
+
+/* This is how to output a note to DBX telling it the line number
+ to which the following sequence of instructions corresponds. */
+
+#define ASM_OUTPUT_SOURCE_LINE(FILE, LINE, COUNTER) \
+{ if (write_symbols == SDB_DEBUG) { \
+ fprintf ((FILE), "\t.ln %d\n", \
+ (sdb_begin_function_line \
+ ? (LINE) - sdb_begin_function_line : 1)); \
+ } else if (write_symbols == DBX_DEBUG) { \
+ fprintf((FILE),"\t.stabd 68,0,%d\n",(LINE)); \
+ } }
+
+/* Globalizing directive for a label. */
+#define GLOBAL_ASM_OP "\t.globl "
+
+/* The prefix to add to user-visible assembler symbols. */
+
+#define USER_LABEL_PREFIX "_"
+
+/* This is how to store into the string LABEL
+ the symbol_ref name of an internal numbered label where
+ PREFIX is the class of label and NUM is the number within the class.
+ This is suitable for output with `assemble_name'. */
+
+#define ASM_GENERATE_INTERNAL_LABEL(LABEL,PREFIX,NUM) \
+ sprintf (LABEL, "*%s%lu", PREFIX, (unsigned long)(NUM))
+
+#define ASM_OUTPUT_REG_PUSH(FILE,REGNO) \
+ fprintf (FILE, "\tst\t%s,(sp)\n\taddo\t4,sp,sp\n", reg_names[REGNO])
+
+/* This is how to output an insn to pop a register from the stack.
+ It need not be very fast code. */
+
+#define ASM_OUTPUT_REG_POP(FILE,REGNO) \
+ fprintf (FILE, "\tsubo\t4,sp,sp\n\tld\t(sp),%s\n", reg_names[REGNO])
+
+/* This is how to output an element of a case-vector that is absolute. */
+
+#define ASM_OUTPUT_ADDR_VEC_ELT(FILE, VALUE) \
+ fprintf (FILE, "\t.word L%d\n", VALUE)
+
+/* This is how to output an element of a case-vector that is relative. */
+
+#define ASM_OUTPUT_ADDR_DIFF_ELT(FILE, BODY, VALUE, REL) \
+ fprintf (FILE, "\t.word L%d-L%d\n", VALUE, REL)
+
+/* This is how to output an assembler line that says to advance the
+ location counter to a multiple of 2**LOG bytes. */
+
+#define ASM_OUTPUT_ALIGN(FILE,LOG) \
+ fprintf (FILE, "\t.align %d\n", (LOG))
+
+#define ASM_OUTPUT_SKIP(FILE,SIZE) \
+ fprintf (FILE, "\t.space %d\n", (int)(SIZE))
+
+/* This says how to output an assembler line
+ to define a global common symbol. */
+
+/* For common objects, output unpadded size... gld960 & lnk960 both
+ have code to align each common object at link time. Also, if size
+ is 0, treat this as a declaration, not a definition - i.e.,
+ do nothing at all. */
+
+#define ASM_OUTPUT_COMMON(FILE, NAME, SIZE, ROUNDED) \
+{ if ((SIZE) != 0) \
+ { \
+ fputs (".globl ", (FILE)), \
+ assemble_name ((FILE), (NAME)), \
+ fputs ("\n.comm ", (FILE)), \
+ assemble_name ((FILE), (NAME)), \
+ fprintf ((FILE), ",%d\n", (int)(SIZE)); \
+ } \
+}
+
+/* This says how to output an assembler line to define a local common symbol.
+ Output unpadded size, with request to linker to align as requested.
+ 0 size should not be possible here. */
+
+#define ASM_OUTPUT_ALIGNED_LOCAL(FILE, NAME, SIZE, ALIGN) \
+( fputs (".bss\t", (FILE)), \
+ assemble_name ((FILE), (NAME)), \
+ fprintf ((FILE), ",%d,%d\n", (int)(SIZE), \
+ (floor_log2 ((ALIGN) / BITS_PER_UNIT))))
+
+/* A C statement (sans semicolon) to output to the stdio stream
+ FILE the assembler definition of uninitialized global DECL named
+ NAME whose size is SIZE bytes and alignment is ALIGN bytes.
+ Try to use asm_output_aligned_bss to implement this macro. */
+
+#define ASM_OUTPUT_ALIGNED_BSS(FILE, DECL, NAME, SIZE, ALIGN) \
+ do { \
+ ASM_OUTPUT_ALIGNED_LOCAL (FILE, NAME, SIZE, ALIGN); \
+ } while (0)
+
+/* Output text for an #ident directive. */
+#define ASM_OUTPUT_IDENT(FILE, STR) fprintf(FILE, "\t# %s\n", STR);
+
+/* Align code to 8 byte boundary if TARGET_CODE_ALIGN is true. */
+
+#define LABEL_ALIGN_AFTER_BARRIER(LABEL) (TARGET_CODE_ALIGN ? 3 : 0)
+
+
+/* Print operand X (an rtx) in assembler syntax to file FILE.
+ CODE is a letter or dot (`z' in `%z0') or 0 if no letter was specified.
+ For `%' followed by punctuation, CODE is the punctuation and X is null. */
+
+#define PRINT_OPERAND(FILE, X, CODE) \
+ i960_print_operand (FILE, X, CODE);
+
+/* Print a memory address as an operand to reference that memory location. */
+
+#define PRINT_OPERAND_ADDRESS(FILE, ADDR) \
+ i960_print_operand_addr (FILE, ADDR)
+
+/* Determine which codes are valid without a following integer. These must
+ not be alphabetic (the characters are chosen so that
+ PRINT_OPERAND_PUNCT_VALID_P translates into a simple range change when
+ using ASCII). */
+
+#define PRINT_OPERAND_PUNCT_VALID_P(CODE) ((CODE) == '+')
+
+/* Output assembler code for a block containing the constant parts
+ of a trampoline, leaving space for the variable parts. */
+
+/* On the i960, the trampoline contains three instructions:
+ ldconst _function, r4
+ ldconst static addr, g12
+ jump (r4) */
+
+#define TRAMPOLINE_TEMPLATE(FILE) \
+{ \
+ assemble_aligned_integer (UNITS_PER_WORD, GEN_INT (0x8C203000)); \
+ assemble_aligned_integer (UNITS_PER_WORD, GEN_INT (0x00000000)); \
+ assemble_aligned_integer (UNITS_PER_WORD, GEN_INT (0x8CE03000)); \
+ assemble_aligned_integer (UNITS_PER_WORD, GEN_INT (0x00000000)); \
+ assemble_aligned_integer (UNITS_PER_WORD, GEN_INT (0x84212000)); \
+}
+
+/* Length in units of the trampoline for entering a nested function. */
+
+#define TRAMPOLINE_SIZE 20
+
+/* Emit RTL insns to initialize the variable parts of a trampoline.
+ FNADDR is an RTX for the address of the function's pure code.
+ CXT is an RTX for the static chain value for the function. */
+
+#define INITIALIZE_TRAMPOLINE(TRAMP, FNADDR, CXT) \
+{ \
+ emit_move_insn (gen_rtx_MEM (SImode, plus_constant (TRAMP, 4)), FNADDR); \
+ emit_move_insn (gen_rtx_MEM (SImode, plus_constant (TRAMP, 12)), CXT); \
+}
+
+/* Generate RTL to flush the register windows so as to make arbitrary frames
+ available. */
+#define SETUP_FRAME_ADDRESSES() \
+ emit_insn (gen_flush_register_windows ())
+
+#define BUILTIN_SETJMP_FRAME_VALUE hard_frame_pointer_rtx
+
+#if 0
+/* Promote char and short arguments to ints, when want compatibility with
+ the iC960 compilers. */
+
+/* ??? In order for this to work, all users would need to be changed
+ to test the value of the macro at run time. */
+#define PROMOTE_PROTOTYPES TARGET_CLEAN_LINKAGE
+/* ??? This does not exist. */
+#define PROMOTE_RETURN TARGET_CLEAN_LINKAGE
+#endif
+
+/* Instruction type definitions. Used to alternate instructions types for
+ better performance on the C series chips. */
+
+enum insn_types { I_TYPE_REG, I_TYPE_MEM, I_TYPE_CTRL };
+
+/* Holds the insn type of the last insn output to the assembly file. */
+
+extern enum insn_types i960_last_insn_type;
+
+/* Parse opcodes, and set the insn last insn type based on them. */
+
+#define ASM_OUTPUT_OPCODE(FILE, INSN) i960_scan_opcode (INSN)
+
+/* Table listing what rtl codes each predicate in i960.c will accept. */
+
+#define PREDICATE_CODES \
+ {"fpmove_src_operand", {CONST_INT, CONST_DOUBLE, CONST, SYMBOL_REF, \
+ LABEL_REF, SUBREG, REG, MEM}}, \
+ {"arith_operand", {SUBREG, REG, CONST_INT}}, \
+ {"logic_operand", {SUBREG, REG, CONST_INT}}, \
+ {"fp_arith_operand", {SUBREG, REG, CONST_DOUBLE}}, \
+ {"signed_arith_operand", {SUBREG, REG, CONST_INT}}, \
+ {"literal", {CONST_INT}}, \
+ {"fp_literal_one", {CONST_DOUBLE}}, \
+ {"fp_literal_double", {CONST_DOUBLE}}, \
+ {"fp_literal", {CONST_DOUBLE}}, \
+ {"signed_literal", {CONST_INT}}, \
+ {"symbolic_memory_operand", {SUBREG, MEM}}, \
+ {"eq_or_neq", {EQ, NE}}, \
+ {"arith32_operand", {SUBREG, REG, LABEL_REF, SYMBOL_REF, CONST_INT, \
+ CONST_DOUBLE, CONST}}, \
+ {"power2_operand", {CONST_INT}}, \
+ {"cmplpower2_operand", {CONST_INT}},
+
+/* Defined in reload.c, and used in insn-recog.c. */
+
+extern int rtx_equal_function_value_matters;
OpenPOWER on IntegriCloud