procyberian/gcc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

re PR target/81616 (Update -mtune=generic for the current Intel and AMD processors)

Browse source 
PR target/81616
	* athlon.md: Disable for generic.
	* haswell.md: Enable for generic.
	* i386.c (ix86_sched_init_global): Add core hooks for generic.
	* x86-tune-sched.c (ix86_issue_rate): Increase issue rate for generic
	to 4.
	(ix86_adjust_cost): Move generic to haswell path.

From-SVN: r255395
This commit is contained in:
Jan Hubicka 2017-12-05 00:59:11 +01:00 • committed by Jan Hubicka
parent 8cc9a08704
commit e7bf9583fa
5 changed files with 189 additions and 183 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
gcc/ChangeLog
View file

@ -1,3 +1,13 @@
2017-12-01 Jan Hubicka <hubicka@ucw.cz>
PR target/81616
* athlon.md: Disable for generic.
* haswell.md: Enable for generic.
* i386.c (ix86_sched_init_global): Add core hooks for generic.
* x86-tune-sched.c (ix86_issue_rate): Increase issue rate for generic
to 4.
(ix86_adjust_cost): Move generic to haswell path.
2017-12-04 Eric Botcazou <ebotcazou@adacore.com>
* config/sparc/sparc.c (sparc_do_work_around_errata): Use mem_ref

187
gcc/config/i386/athlon.md
View file

@ -151,11 +151,11 @@
;; Jump instructions are executed in the branch unit completely transparent to us
(define_insn_reservation "athlon_branch" 0
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "ibr"))
"athlon-direct,athlon-ieu")
(define_insn_reservation "athlon_call" 0
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "call,callv"))
"athlon-vector,athlon-ieu")
(define_insn_reservation "athlon_call_amdfam10" 0
@ -166,15 +166,15 @@
;; Latency of push operation is 3 cycles, but ESP value is available
;; earlier
(define_insn_reservation "athlon_push" 2
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "push"))
"athlon-direct,athlon-agu,athlon-store")
(define_insn_reservation "athlon_pop" 4
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "pop"))
"athlon-vector,athlon-load,athlon-ieu")
(define_insn_reservation "athlon_pop_k8" 3
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "pop"))
"athlon-double,(athlon-ieu+athlon-load)")
(define_insn_reservation "athlon_pop_amdfam10" 3
@ -186,13 +186,13 @@
(eq_attr "type" "leave"))
"athlon-vector,(athlon-ieu+athlon-load)")
(define_insn_reservation "athlon_leave_k8" 3
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(eq_attr "type" "leave"))
"athlon-double,(athlon-ieu+athlon-load)")
;; Lea executes in AGU unit with 2 cycles latency.
(define_insn_reservation "athlon_lea" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "lea"))
"athlon-direct,athlon-agu,nothing")
;; Lea executes in AGU unit with 1 cycle latency on AMDFAM10
@ -209,13 +209,13 @@
"athlon-vector,athlon-ieu0,athlon-mult,nothing,nothing,athlon-ieu0")
;; ??? Widening multiply is vector or double.
(define_insn_reservation "athlon_imul_k8_DI" 4
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "imul")
(and (eq_attr "mode" "DI")
(eq_attr "memory" "none,unknown"))))
"athlon-direct0,athlon-ieu0,athlon-mult,nothing,athlon-ieu0")
(define_insn_reservation "athlon_imul_k8" 3
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "imul")
(eq_attr "memory" "none,unknown")))
"athlon-direct0,athlon-ieu0,athlon-mult,athlon-ieu0")
@ -231,13 +231,13 @@
(eq_attr "memory" "load,both")))
"athlon-vector,athlon-load,athlon-ieu,athlon-mult,nothing,nothing,athlon-ieu")
(define_insn_reservation "athlon_imul_mem_k8_DI" 7
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "imul")
(and (eq_attr "mode" "DI")
(eq_attr "memory" "load,both"))))
"athlon-vector,athlon-load,athlon-ieu,athlon-mult,nothing,athlon-ieu")
(define_insn_reservation "athlon_imul_mem_k8" 6
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "imul")
(eq_attr "memory" "load,both")))
"athlon-vector,athlon-load,athlon-ieu,athlon-mult,athlon-ieu")
@ -251,12 +251,12 @@
;; Using the same heuristics for amdfam10 as K8 with idiv
(define_insn_reservation "athlon_idiv" 6
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "idiv")
(eq_attr "memory" "none,unknown")))
"athlon-vector,(athlon-ieu0*6+(athlon-fpsched,athlon-fvector))")
(define_insn_reservation "athlon_idiv_mem" 9
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "idiv")
(eq_attr "memory" "load,both")))
"athlon-vector,((athlon-load,athlon-ieu0*6)+(athlon-fpsched,athlon-fvector))")
@ -264,13 +264,13 @@
;; as idiv to create smaller automata. This probably does not matter much.
;; Using the same heuristics for amdfam10 as K8 with idiv
(define_insn_reservation "athlon_str" 6
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "str")
(eq_attr "memory" "load,both,store")))
"athlon-vector,athlon-load,athlon-ieu0*6")
(define_insn_reservation "athlon_idirect" 1
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "direct")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "none,unknown"))))
@ -282,7 +282,7 @@
(eq_attr "memory" "none,unknown"))))
"athlon-direct,athlon-ieu")
(define_insn_reservation "athlon_ivector" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "none,unknown"))))
@ -295,13 +295,13 @@
"athlon-vector,athlon-ieu,athlon-ieu")
(define_insn_reservation "athlon_idirect_loadmov" 3
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "imov")
(eq_attr "memory" "load")))
"athlon-direct,athlon-load")
(define_insn_reservation "athlon_idirect_load" 4
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "direct")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "load"))))
@ -313,7 +313,7 @@
(eq_attr "memory" "load"))))
"athlon-direct,athlon-load,athlon-ieu")
(define_insn_reservation "athlon_ivector_load" 6
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "load"))))
@ -326,13 +326,13 @@
"athlon-vector,athlon-load,athlon-ieu,athlon-ieu")
(define_insn_reservation "athlon_idirect_movstore" 1
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "imov")
(eq_attr "memory" "store")))
"athlon-direct,athlon-agu,athlon-store")
(define_insn_reservation "athlon_idirect_both" 4
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "direct")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "both"))))
@ -349,7 +349,7 @@
athlon-store")
(define_insn_reservation "athlon_ivector_both" 6
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "both"))))
@ -368,7 +368,7 @@
athlon-store")
(define_insn_reservation "athlon_idirect_store" 1
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "direct")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "store"))))
@ -383,7 +383,7 @@
athlon-store")
(define_insn_reservation "athlon_ivector_store" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "unit" "integer,unknown")
(eq_attr "memory" "store"))))
@ -405,7 +405,7 @@
(eq_attr "mode" "XF"))))
"athlon-vector,athlon-fpload2,athlon-fvector*9")
(define_insn_reservation "athlon_fldxf_k8" 13
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fmov")
(and (eq_attr "memory" "load")
(eq_attr "mode" "XF"))))
@ -417,7 +417,7 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fany")
(define_insn_reservation "athlon_fld_k8" 2
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fmov")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fstore")
@ -429,7 +429,7 @@
(eq_attr "mode" "XF"))))
"athlon-vector,(athlon-fpsched+athlon-agu),(athlon-store2+(athlon-fvector*7))")
(define_insn_reservation "athlon_fstxf_k8" 8
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fmov")
(and (eq_attr "memory" "store,both")
(eq_attr "mode" "XF"))))
@ -440,16 +440,16 @@
(eq_attr "memory" "store,both")))
"athlon-direct,(athlon-fpsched+athlon-agu),(athlon-fstore+athlon-store)")
(define_insn_reservation "athlon_fst_k8" 2
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fmov")
(eq_attr "memory" "store,both")))
"athlon-direct,(athlon-fpsched+athlon-agu),(athlon-fstore+athlon-store)")
(define_insn_reservation "athlon_fist" 4
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fistp,fisttp"))
"athlon-direct,(athlon-fpsched+athlon-agu),(athlon-fstore+athlon-store)")
(define_insn_reservation "athlon_fmov" 2
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fmov"))
"athlon-direct,athlon-fpsched,athlon-faddmul")
(define_insn_reservation "athlon_fadd_load" 4
@ -458,12 +458,12 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_fadd_load_k8" 6
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fop")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_fadd" 4
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fop"))
"athlon-direct,athlon-fpsched,athlon-fadd")
(define_insn_reservation "athlon_fmul_load" 4
@ -472,16 +472,16 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fmul")
(define_insn_reservation "athlon_fmul_load_k8" 6
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fmul")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fmul")
(define_insn_reservation "athlon_fmul" 4
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fmul"))
"athlon-direct,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_fsgn" 2
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fsgn"))
"athlon-direct,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_fdiv_load" 24
@ -490,7 +490,7 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fmul")
(define_insn_reservation "athlon_fdiv_load_k8" 13
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fdiv")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fmul")
@ -499,16 +499,16 @@
(eq_attr "type" "fdiv"))
"athlon-direct,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_fdiv_k8" 11
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(eq_attr "type" "fdiv"))
"athlon-direct,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_fpspc_load" 103
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "fpspc")
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload,athlon-fvector")
(define_insn_reservation "athlon_fpspc" 100
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fpspc"))
"athlon-vector,athlon-fpsched,athlon-fvector")
(define_insn_reservation "athlon_fcmov_load" 7
@ -521,12 +521,12 @@
(eq_attr "type" "fcmov"))
"athlon-vector,athlon-fpsched,athlon-fvector")
(define_insn_reservation "athlon_fcmov_load_k8" 17
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fcmov")
(eq_attr "memory" "load")))
"athlon-vector,athlon-fploadk8,athlon-fvector")
(define_insn_reservation "athlon_fcmov_k8" 15
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(eq_attr "type" "fcmov"))
"athlon-vector,athlon-fpsched,athlon-fvector")
;; fcomi is vector decoded by uses only one pipe.
@ -537,13 +537,13 @@
(eq_attr "memory" "load"))))
"athlon-vector,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_fcomi_load_k8" 5
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fcmp")
(and (eq_attr "athlon_decode" "vector")
(eq_attr "memory" "load"))))
"athlon-vector,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_fcomi" 3
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "athlon_decode" "vector")
(eq_attr "type" "fcmp")))
"athlon-vector,athlon-fpsched,athlon-fadd")
@ -553,18 +553,18 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_fcom_load_k8" 4
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "fcmp")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_fcom" 2
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(eq_attr "type" "fcmp"))
"athlon-direct,athlon-fpsched,athlon-fadd")
;; Never seen by the scheduler because we still don't do post reg-stack
;; scheduling.
;(define_insn_reservation "athlon_fxch" 2
; (and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
; (and (eq_attr "cpu" "athlon,k8,amdfam10")
; (eq_attr "type" "fxch"))
; "athlon-direct,athlon-fpsched,athlon-fany")
@ -580,13 +580,8 @@
(and (eq_attr "type" "ssemov")
(match_operand:DF 1 "memory_operand")))
"athlon-direct,athlon-fploadk8,athlon-fstore")
(define_insn_reservation "athlon_movsd_load_generic" 2
(and (eq_attr "cpu" "generic")
(and (eq_attr "type" "ssemov")
(match_operand:DF 1 "memory_operand")))
"athlon-double,athlon-fploadk8,(athlon-fstore+athlon-fmul)")
(define_insn_reservation "athlon_movaps_load_k8" 2
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssemov")
(and (eq_attr "mode" "V4SF,V2DF,TI")
(eq_attr "memory" "load"))))
@ -604,7 +599,7 @@
(eq_attr "memory" "load"))))
"athlon-vector,athlon-fpload,(athlon-fany*2)")
(define_insn_reservation "athlon_movss_load_k8" 1
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssemov")
(and (eq_attr "mode" "SF,DI")
(eq_attr "memory" "load"))))
@ -615,7 +610,7 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fany")
(define_insn_reservation "athlon_mmxsseld_k8" 2
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "mmxmov,ssemov")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fstore")
@ -637,19 +632,19 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8, athlon-fany")
(define_insn_reservation "athlon_mmxssest" 3
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "mmxmov,ssemov")
(and (eq_attr "mode" "V4SF,V2DF,TI")
(eq_attr "memory" "store,both"))))
"athlon-vector,(athlon-fpsched+athlon-agu),((athlon-fstore+athlon-store2)*2)")
(define_insn_reservation "athlon_mmxssest_k8" 3
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "mmxmov,ssemov")
(and (eq_attr "mode" "V4SF,V2DF,TI")
(eq_attr "memory" "store,both"))))
"athlon-double,(athlon-fpsched+athlon-agu),((athlon-fstore+athlon-store2)*2)")
(define_insn_reservation "athlon_mmxssest_short" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "mmxmov,ssemov")
(eq_attr "memory" "store,both")))
"athlon-direct,(athlon-fpsched+athlon-agu),(athlon-fstore+athlon-store)")
@ -673,7 +668,7 @@
(eq_attr "memory" "store,both")))
"athlon-direct,(athlon-fpsched+athlon-agu),(athlon-fstore+athlon-store)")
(define_insn_reservation "athlon_movaps_k8" 2
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssemov")
(eq_attr "mode" "V4SF,V2DF,TI")))
"athlon-double,athlon-fpsched,((athlon-faddmul+athlon-faddmul) | (athlon-faddmul, athlon-faddmul))")
@ -683,25 +678,25 @@
(eq_attr "mode" "V4SF,V2DF,TI")))
"athlon-vector,athlon-fpsched,(athlon-faddmul+athlon-faddmul)")
(define_insn_reservation "athlon_mmxssemov" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "mmxmov,ssemov"))
"athlon-direct,athlon-fpsched,athlon-faddmul")
(define_insn_reservation "athlon_mmxmul_load" 4
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "mmxmul")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-fmul")
(define_insn_reservation "athlon_mmxmul" 3
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "mmxmul"))
"athlon-direct,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_mmx_load" 3
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "unit" "mmx")
(eq_attr "memory" "load")))
"athlon-direct,athlon-fpload,athlon-faddmul")
(define_insn_reservation "athlon_mmx" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "unit" "mmx"))
"athlon-direct,athlon-fpsched,athlon-faddmul")
;; SSE operations are handled by the i387 unit as well. The latency
@ -713,7 +708,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload2,(athlon-fmul*2)")
(define_insn_reservation "athlon_sselog_load_k8" 5
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "sselog,sselog1,sseshuf,sseshuf1")
(eq_attr "memory" "load")))
"athlon-double,athlon-fpload2k8,(athlon-fmul*2)")
@ -727,7 +722,7 @@
(eq_attr "type" "sselog,sselog1,sseshuf,sseshuf1"))
"athlon-vector,athlon-fpsched,athlon-fmul*2")
(define_insn_reservation "athlon_sselog_k8" 3
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "sselog,sselog1,sseshuf,sseshuf1"))
"athlon-double,athlon-fpsched,athlon-fmul")
(define_insn_reservation "athlon_sselog_amdfam10" 2
@ -743,13 +738,13 @@
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_ssecmp_load_k8" 4
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "ssecmp")
(and (eq_attr "mode" "SF,DF,DI,TI")
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_ssecmp" 2
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "ssecmp")
(eq_attr "mode" "SF,DF,DI,TI")))
"athlon-direct,athlon-fpsched,athlon-fadd")
@ -759,7 +754,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload2,(athlon-fadd*2)")
(define_insn_reservation "athlon_ssecmpvector_load_k8" 5
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssecmp")
(eq_attr "memory" "load")))
"athlon-double,athlon-fpload2k8,(athlon-fadd*2)")
@ -773,7 +768,7 @@
(eq_attr "type" "ssecmp"))
"athlon-vector,athlon-fpsched,(athlon-fadd*2)")
(define_insn_reservation "athlon_ssecmpvector_k8" 3
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "ssecmp"))
"athlon-double,athlon-fpsched,(athlon-fadd*2)")
(define_insn_reservation "athlon_ssecmpvector_amdfam10" 2
@ -786,7 +781,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_ssecomi_load_k8" 6
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssecomi")
(eq_attr "memory" "load")))
"athlon-vector,athlon-fploadk8,athlon-fadd")
@ -796,7 +791,7 @@
(eq_attr "memory" "load")))
"athlon-direct,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_ssecomi" 4
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(eq_attr "type" "ssecomi"))
"athlon-vector,athlon-fpsched,athlon-fadd")
(define_insn_reservation "athlon_ssecomi_amdfam10" 3
@ -811,13 +806,13 @@
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fpload,athlon-fadd")
(define_insn_reservation "athlon_sseadd_load_k8" 6
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "sseadd,sseadd1")
(and (eq_attr "mode" "SF,DF,DI")
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fploadk8,athlon-fadd")
(define_insn_reservation "athlon_sseadd" 4
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "sseadd,sseadd1")
(eq_attr "mode" "SF,DF,DI")))
"athlon-direct,athlon-fpsched,athlon-fadd")
@ -827,7 +822,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload2,(athlon-fadd*2)")
(define_insn_reservation "athlon_sseaddvector_load_k8" 7
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "sseadd,sseadd1")
(eq_attr "memory" "load")))
"athlon-double,athlon-fpload2k8,(athlon-fadd*2)")
@ -841,7 +836,7 @@
(eq_attr "type" "sseadd,sseadd1"))
"athlon-vector,athlon-fpsched,(athlon-fadd*2)")
(define_insn_reservation "athlon_sseaddvector_k8" 5
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "sseadd,sseadd1"))
"athlon-double,athlon-fpsched,(athlon-fadd*2)")
(define_insn_reservation "athlon_sseaddvector_amdfam10" 4
@ -855,7 +850,7 @@
;; cvtss2sd
(define_insn_reservation "athlon_ssecvt_cvtss2sd_load_k8" 4
(and (eq_attr "cpu" "k8,athlon,generic")
(and (eq_attr "cpu" "k8,athlon")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "direct")
(and (eq_attr "mode" "DF")
@ -869,7 +864,7 @@
(eq_attr "memory" "load")))))
"athlon-double,athlon-fploadk8,(athlon-faddmul+athlon-fstore)")
(define_insn_reservation "athlon_ssecvt_cvtss2sd" 2
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "direct")
(eq_attr "mode" "DF"))))
@ -882,7 +877,7 @@
"athlon-vector,athlon-fpsched,athlon-faddmul,(athlon-fstore*2)")
;; cvtps2pd. Model same way the other double decoded FP conversions.
(define_insn_reservation "athlon_ssecvt_cvtps2pd_load_k8" 5
(and (eq_attr "cpu" "k8,athlon,generic")
(and (eq_attr "cpu" "k8,athlon")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "double")
(and (eq_attr "mode" "V2DF,V4SF,TI")
@ -896,7 +891,7 @@
(eq_attr "memory" "load")))))
"athlon-direct,athlon-fploadk8,athlon-fstore")
(define_insn_reservation "athlon_ssecvt_cvtps2pd_k8" 3
(and (eq_attr "cpu" "k8,athlon,generic")
(and (eq_attr "cpu" "k8,athlon")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "double")
(eq_attr "mode" "V2DF,V4SF,TI"))))
@ -932,7 +927,7 @@
(eq_attr "memory" "load")))))
"athlon-vector,athlon-fpload,(athlon-fstore*2)")
(define_insn_reservation "athlon_sseicvt_cvtsi2ss_load_k8" 9
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "sseicvt")
(and (eq_attr "athlon_decode" "double")
(and (eq_attr "mode" "SF,DF")
@ -947,7 +942,7 @@
"athlon-double,athlon-fploadk8,(athlon-faddmul+athlon-fstore)")
;; cvtsi2sd reg,reg is double decoded (vector on Athlon)
(define_insn_reservation "athlon_sseicvt_cvtsi2sd_k8" 11
(and (eq_attr "cpu" "k8,athlon,generic")
(and (eq_attr "cpu" "k8,athlon")
(and (eq_attr "type" "sseicvt")
(and (eq_attr "athlon_decode" "double")
(and (eq_attr "mode" "SF,DF")
@ -962,7 +957,7 @@
"athlon-vector,athlon-fploadk8,(athlon-faddmul+athlon-fstore)")
;; cvtsi2ss reg, reg is doublepath
(define_insn_reservation "athlon_sseicvt_cvtsi2ss" 14
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "sseicvt")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "mode" "SF,DF")
@ -977,7 +972,7 @@
"athlon-vector,athlon-fploadk8,(athlon-faddmul+athlon-fstore)")
;; cvtsd2ss mem,reg is doublepath, troughput unknown, latency 9
(define_insn_reservation "athlon_ssecvt_cvtsd2ss_load_k8" 9
(and (eq_attr "cpu" "k8,athlon,generic")
(and (eq_attr "cpu" "k8,athlon")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "double")
(and (eq_attr "mode" "SF")
@ -992,7 +987,7 @@
"athlon-double,athlon-fploadk8,(athlon-faddmul+athlon-fstore)")
;; cvtsd2ss reg,reg is vectorpath, troughput unknown, latency 12
(define_insn_reservation "athlon_ssecvt_cvtsd2ss" 12
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "mode" "SF")
@ -1006,7 +1001,7 @@
(eq_attr "memory" "none")))))
"athlon-vector,athlon-fpsched,athlon-faddmul,(athlon-fstore*2)")
(define_insn_reservation "athlon_ssecvt_cvtpd2ps_load_k8" 8
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "mode" "V4SF,V2DF,TI")
@ -1022,7 +1017,7 @@
;; cvtpd2ps mem,reg is vectorpath, troughput unknown, latency 10
;; ??? Why it is fater than cvtsd2ss?
(define_insn_reservation "athlon_ssecvt_cvtpd2ps" 8
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "ssecvt")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "mode" "V4SF,V2DF,TI")
@ -1037,7 +1032,7 @@
"athlon-double,athlon-fpsched,(athlon-faddmul+athlon-fstore)")
;; cvtsd2si mem,reg is doublepath, troughput 1, latency 9
(define_insn_reservation "athlon_secvt_cvtsX2si_load" 9
(and (eq_attr "cpu" "athlon,k8,generic")
(and (eq_attr "cpu" "athlon,k8")
(and (eq_attr "type" "sseicvt")
(and (eq_attr "athlon_decode" "vector")
(and (eq_attr "mode" "SI,DI")
@ -1059,7 +1054,7 @@
(eq_attr "memory" "none")))))
"athlon-vector,athlon-fpsched,athlon-fvector")
(define_insn_reservation "athlon_ssecvt_cvtsX2si_k8" 9
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "sseicvt")
(and (eq_attr "athlon_decode" "double")
(and (eq_attr "mode" "SI,DI")
@ -1097,13 +1092,13 @@
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fpload,athlon-fmul")
(define_insn_reservation "athlon_ssemul_load_k8" 6
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "ssemul")
(and (eq_attr "mode" "SF,DF")
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fploadk8,athlon-fmul")
(define_insn_reservation "athlon_ssemul" 4
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "ssemul")
(eq_attr "mode" "SF,DF")))
"athlon-direct,athlon-fpsched,athlon-fmul")
@ -1113,7 +1108,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload2,(athlon-fmul*2)")
(define_insn_reservation "athlon_ssemulvector_load_k8" 7
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssemul")
(eq_attr "memory" "load")))
"athlon-double,athlon-fpload2k8,(athlon-fmul*2)")
@ -1127,7 +1122,7 @@
(eq_attr "type" "ssemul"))
"athlon-vector,athlon-fpsched,(athlon-fmul*2)")
(define_insn_reservation "athlon_ssemulvector_k8" 5
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "ssemul"))
"athlon-double,athlon-fpsched,(athlon-fmul*2)")
(define_insn_reservation "athlon_ssemulvector_amdfam10" 4
@ -1142,13 +1137,13 @@
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fpload,athlon-fmul*17")
(define_insn_reservation "athlon_ssediv_load_k8" 22
(and (eq_attr "cpu" "k8,generic,amdfam10")
(and (eq_attr "cpu" "k8,amdfam10")
(and (eq_attr "type" "ssediv")
(and (eq_attr "mode" "SF,DF")
(eq_attr "memory" "load"))))
"athlon-direct,athlon-fploadk8,athlon-fmul*17")
(define_insn_reservation "athlon_ssediv" 20
(and (eq_attr "cpu" "athlon,k8,generic,amdfam10")
(and (eq_attr "cpu" "athlon,k8,amdfam10")
(and (eq_attr "type" "ssediv")
(eq_attr "mode" "SF,DF")))
"athlon-direct,athlon-fpsched,athlon-fmul*17")
@ -1158,7 +1153,7 @@
(eq_attr "memory" "load")))
"athlon-vector,athlon-fpload2,athlon-fmul*34")
(define_insn_reservation "athlon_ssedivvector_load_k8" 35
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(and (eq_attr "type" "ssediv")
(eq_attr "memory" "load")))
"athlon-double,athlon-fpload2k8,athlon-fmul*34")
@ -1172,7 +1167,7 @@
(eq_attr "type" "ssediv"))
"athlon-vector,athlon-fmul*34")
(define_insn_reservation "athlon_ssedivvector_k8" 39
(and (eq_attr "cpu" "k8,generic")
(and (eq_attr "cpu" "k8")
(eq_attr "type" "ssediv"))
"athlon-double,athlon-fmul*34")
(define_insn_reservation "athlon_ssedivvector_amdfam10" 20

170
gcc/config/i386/haswell.md
View file

@ -74,12 +74,12 @@
(define_reservation "hsw_p01" "hsw_p0|hsw_p1")
(define_insn_reservation "hsw_complex_insn" 6
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "other,multi,str"))
"hsw_decoder0")
(define_insn_reservation "hsw_call" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "call,callv"))
"hsw_decoder0")
@ -87,104 +87,104 @@
;; imovx always decodes to one uop, and also doesn't use the integer
;; units if it has memory operands.
(define_insn_reservation "hsw_imov" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "imov,imovx")))
"hsw_decodern,hsw_p0156")
(define_insn_reservation "hsw_imov_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "imov,imovx")))
"hsw_decodern,hsw_p23")
(define_insn_reservation "hsw_imov_store" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(eq_attr "type" "imov")))
"hsw_decodern,hsw_p4+(hsw_p2|hsw_p3|hsw_p7)")
(define_insn_reservation "hsw_icmov" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "icmov")))
"hsw_decodern,hsw_p0156,hsw_p0156")
(define_insn_reservation "hsw_icmov_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "icmov")))
"hsw_decodern,hsw_p23+hsw_p0156,hsw_p0156")
(define_insn_reservation "hsw_push_reg" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(eq_attr "type" "push")))
"hsw_decodern,hsw_p4+hsw_p237")
(define_insn_reservation "hsw_push_mem" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "both")
(eq_attr "type" "push")))
"hsw_decodern,hsw_p4+hsw_p237,hsw_p237")
;; Consider lea latency as having 2 components.
(define_insn_reservation "hsw_lea" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "lea")))
"hsw_decodern,hsw_p1|hsw_p5")
(define_insn_reservation "hsw_shift_rotate" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "ishift,ishift1,rotate,rotate1")))
"hsw_decodern,hsw_p0|hsw_p6")
(define_insn_reservation "hsw_shift_rotate_mem" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(eq_attr "type" "ishift,ishift1,rotate,rotate1")))
"hsw_decodern,(hsw_p0|hsw_p6)+hsw_p237+hsw_p4")
(define_insn_reservation "hsw_branch" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "ibr")))
"hsw_decodern,hsw_p6")
(define_insn_reservation "hsw_indirect_branch" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(eq_attr "type" "ibr")))
"hsw_decoder0,hsw_p23+hsw_p6")
(define_insn_reservation "hsw_leave" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "leave"))
"hsw_decoder0,hsw_p23+hsw_p0156,hsw_p0156")
;; imul and imulx with two/three operands only execute on port 1.
(define_insn_reservation "hsw_imul" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "imul")))
"hsw_decodern,hsw_p1")
(define_insn_reservation "hsw_imul_mem" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(eq_attr "type" "imul")))
"hsw_decodern,hsw_p23+hsw_p1")
(define_insn_reservation "hsw_imulx" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "imulx")))
"hsw_decodern,hsw_p0156,hsw_p0156")
(define_insn_reservation "hsw_imulx_mem" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(eq_attr "type" "imulx")))
"hsw_decodern,hsw_p23+hsw_p0156,(hsw_p0|hsw_p6|hsw_p6)")
@ -193,13 +193,13 @@
;; div and idiv are very similar, so we model them the same.
;; Use the same latency for all QI,HI and SI modes.
(define_insn_reservation "hsw_idiv" 23
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "idiv")))
"hsw_decoder0,(hsw_p0p1p5p6+hsw_idiv)*9")
(define_insn_reservation "hsw_idiv_load" 23
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "idiv")))
"hsw_decoder0,hsw_p23+hsw_p0+hsw_idiv,(hsw_p0p1p5p6+hsw_idiv)*9")
@ -207,103 +207,103 @@
;; x87 floating point operations.
(define_insn_reservation "hsw_fxch" 0
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "fxch"))
"hsw_decodern")
(define_insn_reservation "hsw_fop" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none,unknown")
(eq_attr "type" "fop")))
"hsw_decodern,hsw_p1")
(define_insn_reservation "hsw_fop_load" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "fop")))
"hsw_decodern,hsw_p23+hsw_p1,hsw_p1")
(define_insn_reservation "hsw_fop_store" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(eq_attr "type" "fop")))
"hsw_decodern,hsw_p0,hsw_p0,hsw_p0+hsw_p4+hsw_p3")
(define_insn_reservation "hsw_fop_both" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "both")
(eq_attr "type" "fop")))
"hsw_decodern,hsw_p2+hsw_p0,hsw_p0+hsw_p4+hsw_p3")
(define_insn_reservation "hsw_fsgn" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "fsgn"))
"hsw_decodern,hsw_p0")
(define_insn_reservation "hsw_fistp" 7
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "fistp"))
"hsw_decoder0,hsw_p1+hsw_p4+hsw_p23")
(define_insn_reservation "hsw_fcmov" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(eq_attr "type" "fcmov"))
"hsw_decoder0,hsw_p0+hsw_p5,hsw_p0")
(define_insn_reservation "hsw_fcmp" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "fcmp")))
"hsw_decodern,hsw_p1")
(define_insn_reservation "hsw_fcmp_load" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "fcmp")))
"hsw_decodern,hsw_p23+hsw_p1")
(define_insn_reservation "hsw_fmov" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "fmov")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_fmov_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "mode" "!XF")
(eq_attr "type" "fmov"))))
"hsw_decodern,hsw_p23")
(define_insn_reservation "hsw_fmov_XF_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "mode" "XF")
(eq_attr "type" "fmov"))))
"hsw_decodern,(hsw_p23+hsw_p0)*2")
(define_insn_reservation "hsw_fmov_store" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(and (eq_attr "mode" "!XF")
(eq_attr "type" "fmov"))))
"hsw_decodern,hsw_p4p7")
(define_insn_reservation "hsw_fmov_XF_store" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(and (eq_attr "mode" "XF")
(eq_attr "type" "fmov"))))
"hsw_decodern,hsw_p4p7,hsw_p4p7")
(define_insn_reservation "hsw_fmul" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "fmul")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_fmul_load" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "fmul")))
"hsw_decodern,hsw_p23+hsw_p01")
@ -314,42 +314,42 @@
;; that. Throughput is equal to latency - 1, which we model using the
;; hsw_div automaton.
(define_insn_reservation "hsw_fdiv_SF" 18
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "SF")
(eq_attr "type" "fdiv,fpspc"))))
"hsw_decodern,hsw_p0+hsw_fdiv,hsw_fdiv*16")
(define_insn_reservation "hsw_fdiv_SF_load" 19
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "mode" "SF")
(eq_attr "type" "fdiv,fpspc"))))
"hsw_decodern,hsw_p23+hsw_p0+hsw_fdiv,hsw_fdiv*16")
(define_insn_reservation "hsw_fdiv_DF" 32
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "DF")
(eq_attr "type" "fdiv,fpspc"))))
"hsw_decodern,hsw_p0+hsw_fdiv,hsw_fdiv*30")
(define_insn_reservation "hsw_fdiv_DF_load" 33
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "mode" "DF")
(eq_attr "type" "fdiv,fpspc"))))
"hsw_decodern,hsw_p23+hsw_p0+hsw_fdiv,hsw_fdiv*30")
(define_insn_reservation "hsw_fdiv_XF" 38
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "XF")
(eq_attr "type" "fdiv,fpspc"))))
"hsw_decodern,hsw_p0+hsw_fdiv,hsw_fdiv*36")
(define_insn_reservation "hsw_fdiv_XF_load" 39
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "mode" "XF")
(eq_attr "type" "fdiv,fpspc"))))
@ -358,253 +358,253 @@
;; MMX instructions.
(define_insn_reservation "hsw_mmx_add" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "mmxadd,sseiadd")))
"hsw_decodern,hsw_p1|hsw_p5")
(define_insn_reservation "hsw_mmx_add_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "mmxadd,sseiadd")))
"hsw_decodern,hsw_p23+(hsw_p1|hsw_p5)")
(define_insn_reservation "hsw_mmx_shft" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "mmxshft")))
"hsw_decodern,hsw_p0")
(define_insn_reservation "hsw_mmx_shft_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "mmxshft")))
"hsw_decodern,hsw_p23+hsw_p0")
(define_insn_reservation "hsw_mmx_sse_shft" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "type" "sseishft")
(eq_attr "length_immediate" "!0"))))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_mmx_sse_shft_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "type" "sseishft")
(eq_attr "length_immediate" "!0"))))
"hsw_decodern,hsw_p01+hsw_p23")
(define_insn_reservation "hsw_mmx_sse_shft1" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "type" "sseishft")
(eq_attr "length_immediate" "0"))))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_mmx_sse_shft1_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(and (eq_attr "type" "sseishft")
(eq_attr "length_immediate" "0"))))
"hsw_decodern,hsw_p01+hsw_p23")
(define_insn_reservation "hsw_mmx_mul" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "mmxmul,sseimul")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_mmx_mul_load" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "mmxmul,sseimul")))
"hsw_decodern,hsw_p23+hsw_p01")
(define_insn_reservation "hsw_sse_mmxcvt" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "mode" "DI")
(eq_attr "type" "mmxcvt")))
"hsw_decodern,hsw_p1")
;; (define_insn_reservation "hsw_sse_mmxshft" 2
;; (and (eq_attr "cpu" "haswell")
;; (and (eq_attr "cpu" "generic,haswell")
;; (and (eq_attr "mode" "TI")
;; (eq_attr "type" "mmxshft")))
;; "hsw_decodern,hsw_p01")
;; The sfence instruction.
(define_insn_reservation "hsw_sse_sfence" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "unknown")
(eq_attr "type" "sse")))
"hsw_decoder0,hsw_p23+hsw_p4")
(define_insn_reservation "hsw_sse_SFDF" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "mode" "SF,DF")
(eq_attr "type" "sse")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_sse_V4SF" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "mode" "V4SF")
(eq_attr "type" "sse")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_sse_V8SF" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "mode" "V8SF,V4DF")
(eq_attr "type" "sse")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_sse_addcmp" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "sseadd1,ssecmp,ssecomi")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_sse_addcmp_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "sseadd1,ssecmp,ssecomi")))
"hsw_decodern,hsw_p23+hsw_p01")
(define_insn_reservation "hsw_sse_logic" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "sselog,sselog1")))
"hsw_decodern,hsw_p015")
(define_insn_reservation "hsw_sse_logic_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "sselog,sselog1")))
"hsw_decodern,hsw_p015+hsw_p23")
(define_insn_reservation "hsw_sse_add" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "sseadd")))
"hsw_decodern,hsw_p1|hsw_p5")
(define_insn_reservation "hsw_sse_add_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "sseadd")))
"hsw_decodern,(hsw_p1|hsw_p5)+hsw_p23")
(define_insn_reservation "hsw_sse_mul" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "ssemul")))
"hsw_decodern,hsw_p0")
(define_insn_reservation "hsw_sse_mul_load" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "ssemul")))
"hsw_decodern,hsw_p0+hsw_p23")
;; Use skylake pipeline.
(define_insn_reservation "hsw_sse_muladd" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "ssemuladd")))
"hsw_decodern,hsw_p01")
(define_insn_reservation "hsw_sse_muladd_load" 5
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "ssemuladd")))
"hsw_decodern,hsw_p01+hsw_p23")
(define_insn_reservation "hsw_sse_div_SF" 18
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "SF,V4SF,V8SF")
(eq_attr "type" "ssediv"))))
"hsw_decodern,hsw_p0,hsw_ssediv*14")
(define_insn_reservation "hsw_sse_div_SF_load" 18
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "SF,V4SF,V8SF")
(eq_attr "type" "ssediv"))))
"hsw_decodern,(hsw_p23+hsw_p0),hsw_ssediv*14")
(define_insn_reservation "hsw_sse_div_DF" 28
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "DF,V2DF,V4DF")
(eq_attr "type" "ssediv"))))
"hsw_decodern,hsw_p0,hsw_ssediv*20")
(define_insn_reservation "hsw_sse_div_DF_load" 28
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "DF,V2DF,V4DF")
(eq_attr "type" "ssediv"))))
"hsw_decodern,(hsw_p23+hsw_p0),hsw_ssediv*20")
(define_insn_reservation "hsw_sse_icvt" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "sseicvt")))
"hsw_decodern,hsw_p1")
(define_insn_reservation "hsw_sse_icvt_load" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(eq_attr "type" "sseicvt")))
"hsw_decodern,hsw_p23+hsw_p1")
(define_insn_reservation "hsw_sse_icvt_SI" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(and (eq_attr "mode" "SI")
(eq_attr "type" "sseicvt"))))
"hsw_decodern,hsw_p1")
(define_insn_reservation "hsw_sse_icvt_SI_load" 3
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "!none")
(and (eq_attr "mode" "SI")
(eq_attr "type" "sseicvt"))))
"hsw_decodern,hsw_p23+hsw_p1")
(define_insn_reservation "hsw_sse_mov" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none")
(eq_attr "type" "ssemov")))
"hsw_decodern,hsw_p015")
(define_insn_reservation "hsw_sse_mov_load" 2
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "ssemov")))
"hsw_decodern,hsw_p23")
(define_insn_reservation "hsw_sse_mov_store" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(eq_attr "type" "ssemov")))
"hsw_decodern,hsw_p4p7")
(define_insn_reservation "hsw_insn" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "none,unknown")
(eq_attr "type" "alu,alu1,negnot,incdec,icmp,test,setcc,sseishft1,mmx,mmxcmp")))
"hsw_decodern,hsw_p0156")
(define_insn_reservation "hsw_insn_load" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "load")
(eq_attr "type" "alu,alu1,negnot,incdec,icmp,test,setcc,pop,sseishft1,mmx,mmxcmp")))
"hsw_decodern,hsw_p23+hsw_p0156")
(define_insn_reservation "hsw_insn_store" 1
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "store")
(eq_attr "type" "alu,alu1,negnot,incdec,icmp,test,setcc,sseishft1,mmx,mmxcmp")))
"hsw_decodern,hsw_p0156+hsw_p4p7")
@ -612,7 +612,7 @@
;; read-modify-store instructions produce 4 uops so they have to be
;; decoded on hsw_decoder0 as well.
(define_insn_reservation "hsw_insn_both" 4
(and (eq_attr "cpu" "haswell")
(and (eq_attr "cpu" "generic,haswell")
(and (eq_attr "memory" "both")
(eq_attr "type" "alu,alu1,negnot,incdec,icmp,test,setcc,pop,sseishft1,mmx,mmxcmp")))
"hsw_decodern,hsw_p23+hsw_p0156+hsw_p4p7")

1
gcc/config/i386/i386.c
View file

@ -28844,6 +28844,7 @@ ix86_sched_init_global (FILE *, int, int)
case PROCESSOR_NEHALEM:
case PROCESSOR_SANDYBRIDGE:
case PROCESSOR_HASWELL:
case PROCESSOR_GENERIC:
/* Do not perform multipass scheduling for pre-reload schedule
to save compile time. */
if (reload_completed)

4
gcc/config/i386/x86-tune-sched.c
View file

@ -54,7 +54,6 @@ ix86_issue_rate (void)
case PROCESSOR_ATHLON:
case PROCESSOR_K8:
case PROCESSOR_AMDFAM10:
case PROCESSOR_GENERIC:
case PROCESSOR_BTVER1:
return 3;
@ -67,6 +66,7 @@ ix86_issue_rate (void)
case PROCESSOR_NEHALEM:
case PROCESSOR_SANDYBRIDGE:
case PROCESSOR_HASWELL:
case PROCESSOR_GENERIC:
return 4;
default:
@ -354,7 +354,6 @@ ix86_adjust_cost (rtx_insn *insn, int dep_type, rtx_insn *dep_insn, int cost,
case PROCESSOR_BDVER4:
case PROCESSOR_BTVER1:
case PROCESSOR_BTVER2:
case PROCESSOR_GENERIC:
/* Stack engine allows to execute push&pop instructions in parall. */
if ((insn_type == TYPE_PUSH || insn_type == TYPE_POP)
&& (dep_insn_type == TYPE_PUSH || dep_insn_type == TYPE_POP))
@ -424,6 +423,7 @@ ix86_adjust_cost (rtx_insn *insn, int dep_type, rtx_insn *dep_insn, int cost,
case PROCESSOR_NEHALEM:
case PROCESSOR_SANDYBRIDGE:
case PROCESSOR_HASWELL:
case PROCESSOR_GENERIC:
/* Stack engine allows to execute push&pop instructions in parall. */
if ((insn_type == TYPE_PUSH || insn_type == TYPE_POP)
&& (dep_insn_type == TYPE_PUSH || dep_insn_type == TYPE_POP))