1 ; Options for the SH port of the compiler.
3 ; Copyright (C) 2005, 2006, 2007, 2008, 2009, 2010, 2011
4 ; Free Software Foundation, Inc.
6 ; This file is part of GCC.
8 ; GCC is free software; you can redistribute it and/or modify it under
9 ; the terms of the GNU General Public License as published by the Free
10 ; Software Foundation; either version 3, or (at your option) any later
13 ; GCC is distributed in the hope that it will be useful, but WITHOUT ANY
14 ; WARRANTY; without even the implied warranty of MERCHANTABILITY or
15 ; FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
18 ; You should have received a copy of the GNU General Public License
19 ; along with GCC; see the file COPYING3. If not see
20 ; <http://www.gnu.org/licenses/>.
22 ;; Used for various architecture options.
25 ;; Set if the default precision of th FPU is single.
28 ;; Set if we should generate code using type 2A insns.
31 ;; Set if we should generate code using type 2A DF insns.
32 Mask(HARD_SH2A_DOUBLE)
34 ;; Set if compiling for SH4 hardware (to be used for insn costs etc.)
37 ;; Set if we should generate code for a SH5 CPU (either ISA).
40 ;; Set if we should save all target registers.
41 Mask(SAVE_ALL_TARGET_REGS)
44 Target RejectNegative Mask(SH1) Condition(SUPPORT_SH1)
48 Target RejectNegative Mask(SH2) Condition(SUPPORT_SH2)
52 Target RejectNegative Condition(SUPPORT_SH2A)
53 Generate default double-precision SH2a-FPU code
56 Target RejectNegative Condition(SUPPORT_SH2A_NOFPU)
57 Generate SH2a FPU-less code
60 Target RejectNegative Condition(SUPPORT_SH2A_SINGLE)
61 Generate default single-precision SH2a-FPU code
64 Target RejectNegative Condition(SUPPORT_SH2A_SINGLE_ONLY)
65 Generate only single-precision SH2a-FPU code
68 Target RejectNegative Condition(SUPPORT_SH2E)
72 Target RejectNegative Mask(SH3) Condition(SUPPORT_SH3)
76 Target RejectNegative Condition(SUPPORT_SH3E)
80 Target RejectNegative Mask(SH4) Condition(SUPPORT_SH4)
84 Target RejectNegative Condition(SUPPORT_SH4)
88 Target RejectNegative Condition(SUPPORT_SH4)
91 ;; TARGET_SH4_300 indicates if we have the ST40-300 instruction set and
92 ;; pipeline - irrespective of ABI.
94 Target RejectNegative Condition(SUPPORT_SH4) Var(TARGET_SH4_300)
98 Target RejectNegative Condition(SUPPORT_SH4_NOFPU)
99 Generate SH4 FPU-less code
102 Target RejectNegative Condition(SUPPORT_SH4_NOFPU)
103 Generate SH4-100 FPU-less code
106 Target RejectNegative Condition(SUPPORT_SH4_NOFPU)
107 Generate SH4-200 FPU-less code
110 Target RejectNegative Condition(SUPPORT_SH4_NOFPU) Var(TARGET_SH4_300)
111 Generate SH4-300 FPU-less code
114 Target RejectNegative Condition(SUPPORT_SH4_NOFPU) Var(TARGET_SH4_300)
115 Generate code for SH4 340 series (MMU/FPU-less)
116 ;; passes -isa=sh4-nommu-nofpu to the assembler.
119 Target RejectNegative Condition(SUPPORT_SH4_NOFPU)
120 Generate code for SH4 400 series (MMU/FPU-less)
121 ;; passes -isa=sh4-nommu-nofpu to the assembler.
124 Target RejectNegative Condition(SUPPORT_SH4_NOFPU)
125 Generate code for SH4 500 series (FPU-less).
126 ;; passes -isa=sh4-nofpu to the assembler.
129 Target RejectNegative Condition(SUPPORT_SH4_SINGLE)
130 Generate default single-precision SH4 code
133 Target RejectNegative Condition(SUPPORT_SH4_SINGLE)
134 Generate default single-precision SH4-100 code
137 Target RejectNegative Condition(SUPPORT_SH4_SINGLE)
138 Generate default single-precision SH4-200 code
141 Target RejectNegative Condition(SUPPORT_SH4_SINGLE) Var(TARGET_SH4_300)
142 Generate default single-precision SH4-300 code
145 Target RejectNegative Condition(SUPPORT_SH4_SINGLE_ONLY)
146 Generate only single-precision SH4 code
149 Target RejectNegative Condition(SUPPORT_SH4_SINGLE_ONLY)
150 Generate only single-precision SH4-100 code
153 Target RejectNegative Condition(SUPPORT_SH4_SINGLE_ONLY)
154 Generate only single-precision SH4-200 code
157 Target RejectNegative Condition(SUPPORT_SH4_SINGLE_ONLY) Var(TARGET_SH4_300)
158 Generate only single-precision SH4-300 code
161 Target RejectNegative Mask(SH4A) Condition(SUPPORT_SH4A)
165 Target RejectNegative Condition(SUPPORT_SH4A_NOFPU)
166 Generate SH4a FPU-less code
169 Target RejectNegative Condition(SUPPORT_SH4A_SINGLE)
170 Generate default single-precision SH4a code
173 Target RejectNegative Condition(SUPPORT_SH4A_SINGLE_ONLY)
174 Generate only single-precision SH4a code
177 Target RejectNegative Condition(SUPPORT_SH4AL)
178 Generate SH4al-dsp code
181 Target RejectNegative Condition(SUPPORT_SH5_32MEDIA)
182 Generate 32-bit SHmedia code
185 Target RejectNegative Condition(SUPPORT_SH5_32MEDIA_NOFPU)
186 Generate 32-bit FPU-less SHmedia code
189 Target RejectNegative Condition(SUPPORT_SH5_64MEDIA)
190 Generate 64-bit SHmedia code
193 Target RejectNegative Condition(SUPPORT_SH5_64MEDIA_NOFPU)
194 Generate 64-bit FPU-less SHmedia code
197 Target RejectNegative Condition(SUPPORT_SH5_32MEDIA)
198 Generate SHcompact code
201 Target RejectNegative Condition(SUPPORT_SH5_32MEDIA_NOFPU)
202 Generate FPU-less SHcompact code
204 maccumulate-outgoing-args
205 Target Report Var(TARGET_ACCUMULATE_OUTGOING_ARGS) Init(1)
206 Reserve space for outgoing arguments in the function prologue
209 Target Report Mask(ADJUST_UNROLL) Condition(SUPPORT_ANY_SH5)
210 Throttle unrolling to avoid thrashing target registers unless the unroll benefit outweighs this
213 Target Report RejectNegative InverseMask(LITTLE_ENDIAN)
214 Generate code in big endian mode
217 Target Report RejectNegative Mask(BIGTABLE)
218 Generate 32-bit offsets in switch tables
221 Target Report RejectNegative Mask(BITOPS)
222 Generate bit instructions
225 Target RejectNegative Joined UInteger Var(sh_branch_cost) Init(-1)
226 Cost to assume for a branch insn
229 Target Var(TARGET_CBRANCHDI4)
230 Enable cbranchdi4 pattern
233 Target Var(TARGET_CMPEQDI_T)
234 Emit cmpeqdi_t pattern even when -mcbranchdi is in effect.
237 Target RejectNegative Var(TARGET_SH5_CUT2_WORKAROUND)
238 Enable SH5 cut2 workaround
241 Target Report RejectNegative Mask(ALIGN_DOUBLE)
242 Align doubles at 64-bit boundaries
245 Target RejectNegative Joined Var(sh_div_str) Init("")
246 Division strategy, one of: call, call2, fp, inv, inv:minlat, inv20u, inv20l, inv:call, inv:call2, inv:fp, call-div1, call-fp, call-table
249 Target RejectNegative Joined Var(sh_divsi3_libfunc) Init("")
250 Specify name for 32 bit signed division function
253 Target RejectNegative Mask(FMOVD)
254 Enable the use of 64-bit floating point registers in fmov instructions. See -mdalign if 64-bit alignment is required.
257 Target RejectNegative Joined Var(sh_fixed_range_str)
258 Specify range of registers to make fixed
261 Target Var(TARGET_FMAC)
262 Enable the use of the fused floating point multiply-accumulate operation
265 Target RejectNegative Joined UInteger Var(sh_gettrcost) Init(-1)
266 Cost to assume for gettr insn
269 Target Report RejectNegative Mask(HITACHI)
270 Follow Renesas (formerly Hitachi) / SuperH calling conventions
273 Target Var(TARGET_IEEE)
274 Increase the IEEE compliance for floating-point comparisons
277 Target Report Mask(ALLOW_INDEXED_ADDRESS) Condition(SUPPORT_ANY_SH5_32MEDIA)
278 Enable the use of the indexed addressing mode for SHmedia32/SHcompact
280 minline-ic_invalidate
281 Target Report Var(TARGET_INLINE_IC_INVALIDATE)
282 inline code to invalidate instruction cache entries after setting up nested function trampolines
285 Target Report Mask(INVALID_SYMBOLS) Condition(SUPPORT_ANY_SH5)
286 Assume symbols might be invalid
289 Target Report RejectNegative Mask(DUMPISIZE)
290 Annotate assembler instructions with estimated addresses
293 Target Report RejectNegative Mask(LITTLE_ENDIAN)
294 Generate code in little endian mode
297 Target Report RejectNegative Mask(NOMACSAVE)
298 Mark MAC register as call-clobbered
300 ;; ??? This option is not useful, but is retained in case there are people
301 ;; who are still relying on it. It may be deleted in the future.
303 Target Report RejectNegative Mask(PADSTRUCT)
304 Make structs a multiple of 4 bytes (warning: ABI altered)
307 Target Report RejectNegative Mask(PREFERGOT)
308 Emit function-calls using global offset table when generating PIC
311 Target Report Mask(PT_FIXED) Condition(SUPPORT_ANY_SH5)
312 Assume pt* instructions won't trap
315 Target Report RejectNegative Mask(RELAX)
316 Shorten address references during linking
319 Target Mask(HITACHI) MaskExists
320 Follow Renesas (formerly Hitachi) / SuperH calling conventions
323 Target Report Mask(SOFT_ATOMIC)
324 Use software atomic sequences supported by kernel
327 Target RejectNegative Alias(Os)
328 Deprecated. Use -Os instead
331 Target RejectNegative Joined UInteger Var(sh_multcost) Init(-1)
332 Cost to assume for a multiply insn
335 Target Report RejectNegative Mask(USERMODE)
336 Don't generate privileged-mode only code; implies -mno-inline-ic_invalidate if the inline code would not work in user mode.
338 ;; We might want to enable this by default for TARGET_HARD_SH4, because
339 ;; zero-offset branches have zero latency. Needs some benchmarking.
341 Target Var(TARGET_PRETEND_CMOVE)
342 Pretend a branch-around-a-move is a conditional move.