blob: e2deed6db042f2afe1a2c5dc125f1af2368b7f66 [file] [log] [blame]
; Options for the Adapteva EPIPHANY port of the compiler
;
; Copyright (C) 2005-2015 Free Software Foundation, Inc.
; Contributed by Embecosm on behalf of Adapteva, Inc.
;
; 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 3, 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 COPYING3. If not see
; <http://www.gnu.org/licenses/>.
mhalf-reg-file
Target Mask(HALF_REG_FILE)
Don't use any of r32..r63.
mprefer-short-insn-regs
Target Mask(PREFER_SHORT_INSN_REGS)
preferentially allocate registers that allow short instruction generation.
mbranch-cost=
Target RejectNegative Joined UInteger Var(epiphany_branch_cost) Init(3)
Set branch cost
mcmove
Target Mask(CMOVE)
enable conditional move instruction usage.
mnops=
Target RejectNegative Joined UInteger Var(epiphany_n_nops) Init(0)
set number of nops to emit before each insn pattern
; Problems with using the flags from fsub for comparison are:
; - Because of underflow (lack of subnormal numbers), different small numbers
; can compare as equal.
; - the set of comparisons is limited, and reversing comparisons doesn't work
; in the presence of NaNs.
; The latter problem might be tolerated with -ffinite-math-only , but nothing
; in -funsafe-math-optimizations says different small numbers may be considered
; equal.
msoft-cmpsf
Target Mask(SOFT_CMPSF)
Use software floating point comparisons
msplit-lohi
Target Mask(SPLIT_LOHI)
Enable split of 32 bit immediate loads into low / high part
mpost-inc
Target Mask(POST_INC)
Enable use of POST_INC / POST_DEC
mpost-modify
Target Mask(POST_MODIFY)
Enable use of POST_MODIFY
mstack-offset=
Target RejectNegative Joined UInteger Var(epiphany_stack_offset) Init(EPIPHANY_STACK_OFFSET)
Set number of bytes on the stack preallocated for use by the callee.
mround-nearest
target Mask(ROUND_NEAREST)
Assume round to nearest is selected for purposes of scheduling.
mlong-calls
Target Mask(LONG_CALLS)
Generate call insns as indirect calls
mshort-calls
Target Mask(SHORT_CALLS)
Generate call insns as direct calls
msmall16
Target Mask(SMALL16)
Assume labels and symbols can be addressed using 16 bit absolute addresses.
mfp-mode=
Target RejectNegative Joined Var(epiphany_normal_fp_mode) Enum(attr_fp_mode) Init(FP_MODE_CALLER)
; The values are from enum attr_fp_mode, but using that enum would bring
; problems with enum forward declarations.
Enum
Name(attr_fp_mode) Type(int)
EnumValue
Enum(attr_fp_mode) String(caller) Value(FP_MODE_CALLER)
EnumValue
Enum(attr_fp_mode) String(round-nearest) Value(FP_MODE_ROUND_NEAREST)
EnumValue
Enum(attr_fp_mode) String(truncate) Value(FP_MODE_ROUND_TRUNC)
EnumValue
Enum(attr_fp_mode) String(int) Value(FP_MODE_INT)
may-round-for-trunc
Target Mask(MAY_ROUND_FOR_TRUNC)
A floatig point to integer truncation may be replaced with rounding to save mode switching
mvect-double
Target Mask(VECT_DOUBLE)
Vectorize for double-word operations.
max-vect-align=
Target RejectNegative Joined Var(epiphany_vect_align) Enum(vect_align) Init(8)
Enum
Name(vect_align) Type(int)
EnumValue
Enum(vect_align) String(4) Value(4)
EnumValue
Enum(vect_align) String(8) Value(8)
msplit-vecmove-early
Target Mask(SPLIT_VECMOVE_EARLY)
Split unaligned 8 byte vector moves before post-modify address generation.
mfp-iarith
Target Mask(FP_IARITH)
Use the floating point unit for integer add/subtract.
m1reg-
Target RejectNegative Joined Var(epiphany_m1reg) Enum(m1reg) Init(-1)
Set register to hold -1.
Enum
Name(m1reg) Type(int)
EnumValue
Enum(m1reg) String(none) Value(-1)
EnumValue
Enum(m1reg) String(r43) Value(43)
EnumValue
Enum(m1reg) String(r63) Value(63)