Kernel Omnislash (Unofficial) - Aprendendo a voar sem segredos!!!

Stivekx · 14 de Janeiro de 2011, 22:50

Obrigaodo pela dica, vou recompilar de novo agora ;p

Desativar typos de arquivos com JFS que eu não uso pode ter alguma relação? É tudo ext4 aqui ;s

galactus · 14 de Janeiro de 2011, 22:58

Citação de: Stivekx online 14 de Janeiro de 2011, 22:50
Obrigaodo pela dica, vou recompilar de novo agora ;p

Desativar typos de arquivos com JFS que eu não uso pode ter alguma relação? É tudo ext4 aqui ;s

Não!

Só se você esqueceu de colocar o ext4 dentro do kernel!

Talvez você tenha "tirado" demais do seu kernel!

4d4c47 · 15 de Janeiro de 2011, 00:39

aqui vai uma comparação de cflags, faça um teste e um bom estudo delas:

codigo-fonte = teste.c

Código Selecionar


#include <stdio.h>

main ()
{
  printf("Hello world\n");
  return 0;
}

-------------------------------------------------------------------------------------------------------------------------------------

gcc -fverbose-asm -S teste.c -o a.s

cat a.s:

Código Selecionar



	.file	"teste.c"
# GNU C (Ubuntu 4.4.3-4ubuntu5) version 4.4.3 (x86_64-linux-gnu)
#	compiled by GNU C version 4.4.3, GMP version 4.3.2, MPFR version 2.4.2-p1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -mtune=generic
# -auxbase-strip a.s -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg -fcommon
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -ffunction-cse -fgcse-lm -fident -finline-functions-called-once
# -fira-share-save-slots -fira-share-spill-slots -fivopts
# -fkeep-static-consts -fleading-underscore -fmath-errno
# -fmerge-debug-strings -fmove-loop-invariants -fpeephole
# -freg-struct-return -fsched-interblock -fsched-spec
# -fsched-stalled-insns-dep -fsigned-zeros -fsplit-ivs-in-unroller
# -fstack-protector -ftrapping-math -ftree-cselim -ftree-loop-im
# -ftree-loop-ivcanon -ftree-loop-optimize -ftree-parallelize-loops=
# -ftree-reassoc -ftree-scev-cprop -ftree-switch-conversion
# -ftree-vect-loop-version -funit-at-a-time -funwind-tables
# -fvect-cost-model -fverbose-asm -fzero-initialized-in-bss
# -m128bit-long-double -m64 -m80387 -maccumulate-outgoing-args
# -malign-stringops -mfancy-math-387 -mfp-ret-in-387 -mfused-madd -mglibc
# -mieee-fp -mmmx -mno-sse4 -mpush-args -mred-zone -msse -msse2
# -mtls-direct-seg-refs

# Compiler executable checksum: 2129e1a56226bd6e8f7af5e0a3ff467d

	.section	.rodata
.LC0:
	.string	"Hello world"
	.text
.globl main
	.type	main, @function
main:
.LFB0:
	.cfi_startproc
	pushq	%rbp	#
	.cfi_def_cfa_offset 16
	movq	%rsp, %rbp	#,
	.cfi_offset 6, -16
	.cfi_def_cfa_register 6
	movl	$.LC0, %edi	#,
	call	puts	#
	movl	$0, %eax	#, D.2045
	leave
	ret
	.cfi_endproc
.LFE0:
	.size	main, .-main
	.ident	"GCC: (Ubuntu 4.4.3-4ubuntu5) 4.4.3"
	.section	.note.GNU-stack,"",@progbits

---------------------------------------------------------------------------------------------------------------------------------------------------------------------

gcc -O2 -march=k8 -fverbose-asm -S teste.c -o march_k8.s

cat march_k8.s:

Código Selecionar


	.file	"teste.c"
# GNU C (Ubuntu 4.4.3-4ubuntu5) version 4.4.3 (x86_64-linux-gnu)
#	compiled by GNU C version 4.4.3, GMP version 4.3.2, MPFR version 2.4.2-p1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -march=k8 -auxbase-strip
# march_k8.s -O2 -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-interblock -fsched-spec -fsched-stalled-insns-dep
# -fschedule-insns2 -fsigned-zeros -fsplit-ivs-in-unroller
# -fsplit-wide-types -fstack-protector -fstrict-aliasing -fstrict-overflow
# -fthread-jumps -ftoplevel-reorder -ftrapping-math -ftree-builtin-call-dce
# -ftree-ccp -ftree-ch -ftree-copy-prop -ftree-copyrename -ftree-cselim
# -ftree-dce -ftree-dominator-opts -ftree-dse -ftree-fre -ftree-loop-im
# -ftree-loop-ivcanon -ftree-loop-optimize -ftree-parallelize-loops=
# -ftree-pre -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-sra
# -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version -ftree-vrp
# -funit-at-a-time -funwind-tables -fvect-cost-model -fverbose-asm
# -fzero-initialized-in-bss -m128bit-long-double -m3dnow -m64 -m80387
# -maccumulate-outgoing-args -malign-stringops -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4
# -mpush-args -mred-zone -msse -msse2 -mtls-direct-seg-refs

# Compiler executable checksum: 2129e1a56226bd6e8f7af5e0a3ff467d

	.section	.rodata.str1.1,"aMS",@progbits,1
.LC0:
	.string	"Hello world\n"
	.text
	.p2align 4,,15
.globl main
	.type	main, @function
main:
.LFB22:
	.cfi_startproc
	subq	$8, %rsp	#,
	.cfi_def_cfa_offset 16
	movl	$.LC0, %esi	#,
	movl	$1, %edi	#,
	xorl	%eax, %eax	#
	call	__printf_chk	#
	xorl	%eax, %eax	#
	addq	$8, %rsp	#,
	ret
	.cfi_endproc
.LFE22:
	.size	main, .-main
	.ident	"GCC: (Ubuntu 4.4.3-4ubuntu5) 4.4.3"
	.section	.note.GNU-stack,"",@progbits

----------------------------------------------------------------------------------------------------------------------------------------------------------------------

gcc -O2 -march=native -fverbose-asm -S teste.c -o march_native.s

cat march_native.s:

Código Selecionar


	.file	"teste.c"
# GNU C (Ubuntu 4.4.3-4ubuntu5) version 4.4.3 (x86_64-linux-gnu)
#	compiled by GNU C version 4.4.3, GMP version 4.3.2, MPFR version 2.4.2-p1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -march=amdfam10 -mcx16
# -msahf -mpopcnt --param l1-cache-size=64 --param l1-cache-line-size=64
# --param l2-cache-size=512 -mtune=amdfam10 -auxbase-strip march_native.s
# -O2 -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-interblock -fsched-spec -fsched-stalled-insns-dep
# -fschedule-insns2 -fsigned-zeros -fsplit-ivs-in-unroller
# -fsplit-wide-types -fstack-protector -fstrict-aliasing -fstrict-overflow
# -fthread-jumps -ftoplevel-reorder -ftrapping-math -ftree-builtin-call-dce
# -ftree-ccp -ftree-ch -ftree-copy-prop -ftree-copyrename -ftree-cselim
# -ftree-dce -ftree-dominator-opts -ftree-dse -ftree-fre -ftree-loop-im
# -ftree-loop-ivcanon -ftree-loop-optimize -ftree-parallelize-loops=
# -ftree-pre -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-sra
# -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version -ftree-vrp
# -funit-at-a-time -funwind-tables -fvect-cost-model -fverbose-asm
# -fzero-initialized-in-bss -m128bit-long-double -m3dnow -m64 -m80387 -mabm
# -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4 -mpopcnt
# -mpush-args -mred-zone -msahf -msse -msse2 -msse3 -msse4a
# -mtls-direct-seg-refs

# Compiler executable checksum: 2129e1a56226bd6e8f7af5e0a3ff467d

	.section	.rodata.str1.1,"aMS",@progbits,1
.LC0:
	.string	"Hello world\n"
	.text
	.p2align 5,,31
.globl main
	.type	main, @function
main:
.LFB22:
	.cfi_startproc
	subq	$8, %rsp	#,
	.cfi_def_cfa_offset 16
	movl	$.LC0, %esi	#,
	movl	$1, %edi	#,
	xorl	%eax, %eax	#
	call	__printf_chk	#
	xorl	%eax, %eax	#
	addq	$8, %rsp	#,
	ret
	.cfi_endproc
.LFE22:
	.size	main, .-main
	.ident	"GCC: (Ubuntu 4.4.3-4ubuntu5) 4.4.3"
	.section	.note.GNU-stack,"",@progbits

dá pra ver as diferenças das instruções que vai na cflag, pelo que vi o march=native é o mais completo,

tenho aqui um AMD athlon II x2 M300

o gcc é a versão 4.4.3 do ubuntu 10.04 64 bits

vale a pena testar.

vampire_thunder · 15 de Janeiro de 2011, 01:23

Cara, que maneiro isso aí. Mas dá até medo

Pra q serve o teste.c? Só para chamar o gcc?

Aqui apareceram umas informações interessantes:

Código Selecionar

	.file	"teste.c"
# GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu)
#	compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7.
# warning: MPC header version 0.8.2 differs from library version 0.8.1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -march=amdfam10 -mcx16 -msahf -mpopcnt -mabm
# --param l1-cache-size=64 --param l1-cache-line-size=64 --param

Parece que tem umas libs desatualizadas.

galactus · 15 de Janeiro de 2011, 01:38

Saída do cat a.s:

Código Selecionar

galactus@galactus-desktop:~$ cat a.s
        .file   "teste.c"
# GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu)
#       compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -mtune=generic -march=x86-64
# -auxbase-strip a.s -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg -fcommon
# -fdelete-null-pointer-checks -fdwarf2-cfi-asm -fearly-inlining
# -feliminate-unused-debug-types -ffunction-cse -fgcse-lm -fident
# -finline-functions-called-once -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-debug-strings
# -fmove-loop-invariants -fpeephole -freg-struct-return
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fstack-protector -ftrapping-math -ftree-cselim
# -ftree-forwprop -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize
# -ftree-parallelize-loops= -ftree-phiprop -ftree-pta -ftree-reassoc
# -ftree-scev-cprop -ftree-slp-vectorize -ftree-vect-loop-version
# -funit-at-a-time -funwind-tables -fvect-cost-model -fverbose-asm
# -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387
# -maccumulate-outgoing-args -malign-stringops -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4
# -mpush-args -mred-zone -msse -msse2 -mtls-direct-seg-refs

# Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9

        .section        .rodata
.LC0:
        .string "Hello world"                                                                                                                                                                                                                                                  
        .text                                                                                                                                                                                                                                                                  
.globl main                                                                                                                                                                                                                                                                    
        .type   main, @function                                                                                                                                                                                                                                                
main:                                                                                                                                                                                                                                                                          
.LFB0:                                                                                                                                                                                                                                                                         
        .cfi_startproc                                                                                                                                                                                                                                                         
        pushq   %rbp    #                                                                                                                                                                                                                                                      
        .cfi_def_cfa_offset 16                                                                                                                                                                                                                                                 
        movq    %rsp, %rbp      #,                                                                                                                                                                                                                                             
        .cfi_offset 6, -16                                                                                                                                                                                                                                                     
        .cfi_def_cfa_register 6                                                                                                                                                                                                                                                
        movl    $.LC0, %edi     #,                                                                                                                                                                                                                                             
        call    puts    #                                                                                                                                                                                                                                                      
        movl    $0, %eax        #, D.3278                                                                                                                                                                                                                                      
        leave                                                                                                                                                                                                                                                                  
        .cfi_def_cfa 7, 8                                                                                                                                                                                                                                                      
        ret                                                                                                                                                                                                                                                                    
        .cfi_endproc                                                                                                                                                                                                                                                           
.LFE0:                                                                                                                                                                                                                                                                         
        .size   main, .-main                                                                                                                                                                                                                                                   
        .ident  "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1"                                                                                                                                                                                                                    
        .section        .note.GNU-stack,"",@progbits

Saída do core2:

Código Selecionar

galactus@galactus-desktop:~$ cat march_core2.s
        .file   "teste.c"
# GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu)
#       compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -march=core2 -auxbase-strip
# march_core2.s -O2 -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fsplit-wide-types -fstack-protector
# -fstrict-aliasing -fstrict-overflow -fthread-jumps -ftoplevel-reorder
# -ftrapping-math -ftree-builtin-call-dce -ftree-ccp -ftree-ch
# -ftree-copy-prop -ftree-copyrename -ftree-cselim -ftree-dce
# -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre
# -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize
# -ftree-parallelize-loops= -ftree-phiprop -ftree-pre -ftree-pta
# -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-slp-vectorize
# -ftree-sra -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version
# -ftree-vrp -funit-at-a-time -funwind-tables -fvect-cost-model
# -fverbose-asm -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387
# -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4
# -mpush-args -mred-zone -msahf -msse -msse2 -msse3 -mssse3
# -mtls-direct-seg-refs

# Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9

        .section        .rodata.str1.1,"aMS",@progbits,1
.LC0:
        .string "Hello world"
        .text
        .p2align 4,,15
.globl main
        .type   main, @function
main:
.LFB22:
        .cfi_startproc
        subq    $8, %rsp        #,
        .cfi_def_cfa_offset 16
        movl    $.LC0, %edi     #,
        call    puts    #
        xorl    %eax, %eax      #
        addq    $8, %rsp        #,
        .cfi_def_cfa_offset 8
        ret
        .cfi_endproc
.LFE22:
        .size   main, .-main
        .ident  "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1"
        .section        .note.GNU-stack,"",@progbits

Saída do native:

Código Selecionar

galactus@galactus-desktop:~$ cat march_native.s
        .file   "teste.c"
# GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu)
#       compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -D_FORTIFY_SOURCE=2 -march=core2 -mcx16 -msahf
# -mpopcnt -msse4.2 --param l1-cache-size=32 --param l1-cache-line-size=64
# --param l2-cache-size=8192 -mtune=core2 -auxbase-strip march_native.s -O2
# -fverbose-asm -fstack-protector
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fsplit-wide-types -fstack-protector
# -fstrict-aliasing -fstrict-overflow -fthread-jumps -ftoplevel-reorder
# -ftrapping-math -ftree-builtin-call-dce -ftree-ccp -ftree-ch
# -ftree-copy-prop -ftree-copyrename -ftree-cselim -ftree-dce
# -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre
# -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize
# -ftree-parallelize-loops= -ftree-phiprop -ftree-pre -ftree-pta
# -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-slp-vectorize
# -ftree-sra -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version
# -ftree-vrp -funit-at-a-time -funwind-tables -fvect-cost-model
# -fverbose-asm -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387
# -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mpopcnt -mpush-args
# -mred-zone -msahf -msse -msse2 -msse3 -msse4 -msse4.1 -msse4.2 -mssse3
# -mtls-direct-seg-refs

# Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9

        .section        .rodata.str1.1,"aMS",@progbits,1
.LC0:
        .string "Hello world"
        .text
        .p2align 4,,15
.globl main
        .type   main, @function
main:
.LFB22:
        .cfi_startproc
        subq    $8, %rsp        #,
        .cfi_def_cfa_offset 16
        movl    $.LC0, %edi     #,
        call    puts    #
        xorl    %eax, %eax      #
        addq    $8, %rsp        #,
        .cfi_def_cfa_offset 8
        ret
        .cfi_endproc
.LFE22:
        .size   main, .-main
        .ident  "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1"
        .section        .note.GNU-stack,"",@progbits

E agora 4d4c47? Eu coloco o quê na linha do HOSTCFLAGS do MakeFile do Kernel?

A opção -march=native foi a mais completa, sem dúvida! Ela especifica os três níveis de Caches do meu processador e todas as instruções SSE suportadas por ele! A opção Core2 não especifica os caches e não coloca todas as instruções do processador! A genérica é muito pior!

4d4c47 · 15 de Janeiro de 2011, 01:59

Citação de: galactus online 15 de Janeiro de 2011, 01:38
Saída do cat a.s:

Código Selecionar Expandir
galactus@galactus-desktop:~$ cat a.s .file "teste.c" # GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu) # compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2 # GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 # options passed: teste.c -D_FORTIFY_SOURCE=2 -mtune=generic -march=x86-64 # -auxbase-strip a.s -fverbose-asm -fstack-protector # options enabled: -falign-loops -fargument-alias # -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg -fcommon # -fdelete-null-pointer-checks -fdwarf2-cfi-asm -fearly-inlining # -feliminate-unused-debug-types -ffunction-cse -fgcse-lm -fident # -finline-functions-called-once -fira-share-save-slots # -fira-share-spill-slots -fivopts -fkeep-static-consts # -fleading-underscore -fmath-errno -fmerge-debug-strings # -fmove-loop-invariants -fpeephole -freg-struct-return # -fsched-critical-path-heuristic -fsched-dep-count-heuristic # -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic # -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic # -fsched-stalled-insns-dep -fshow-column -fsigned-zeros # -fsplit-ivs-in-unroller -fstack-protector -ftrapping-math -ftree-cselim # -ftree-forwprop -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize # -ftree-parallelize-loops= -ftree-phiprop -ftree-pta -ftree-reassoc # -ftree-scev-cprop -ftree-slp-vectorize -ftree-vect-loop-version # -funit-at-a-time -funwind-tables -fvect-cost-model -fverbose-asm # -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387 # -maccumulate-outgoing-args -malign-stringops -mfancy-math-387 # -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4 # -mpush-args -mred-zone -msse -msse2 -mtls-direct-seg-refs # Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9 .section .rodata .LC0: .string "Hello world" .text .globl main .type main, @function main: .LFB0: .cfi_startproc pushq %rbp # .cfi_def_cfa_offset 16 movq %rsp, %rbp #, .cfi_offset 6, -16 .cfi_def_cfa_register 6 movl $.LC0, %edi #, call puts # movl $0, %eax #, D.3278 leave .cfi_def_cfa 7, 8 ret .cfi_endproc .LFE0: .size main, .-main .ident "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1" .section .note.GNU-stack,"",@progbits

Saída do core2:

Código Selecionar Expandir
galactus@galactus-desktop:~$ cat march_core2.s .file "teste.c" # GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu) # compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2 # GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 # options passed: teste.c -D_FORTIFY_SOURCE=2 -march=core2 -auxbase-strip # march_core2.s -O2 -fverbose-asm -fstack-protector # options enabled: -falign-loops -fargument-alias # -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg # -fcaller-saves -fcommon -fcprop-registers -fcrossjumping # -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks # -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types # -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse # -fgcse-lm -fguess-branch-probability -fident -fif-conversion # -fif-conversion2 -findirect-inlining -finline # -finline-functions-called-once -finline-small-functions -fipa-cp # -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots # -fira-share-spill-slots -fivopts -fkeep-static-consts # -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings # -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move # -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return # -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop # -fsched-critical-path-heuristic -fsched-dep-count-heuristic # -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic # -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic # -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros # -fsplit-ivs-in-unroller -fsplit-wide-types -fstack-protector # -fstrict-aliasing -fstrict-overflow -fthread-jumps -ftoplevel-reorder # -ftrapping-math -ftree-builtin-call-dce -ftree-ccp -ftree-ch # -ftree-copy-prop -ftree-copyrename -ftree-cselim -ftree-dce # -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre # -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize # -ftree-parallelize-loops= -ftree-phiprop -ftree-pre -ftree-pta # -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-slp-vectorize # -ftree-sra -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version # -ftree-vrp -funit-at-a-time -funwind-tables -fvect-cost-model # -fverbose-asm -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387 # -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387 # -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4 # -mpush-args -mred-zone -msahf -msse -msse2 -msse3 -mssse3 # -mtls-direct-seg-refs # Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9 .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Hello world" .text .p2align 4,,15 .globl main .type main, @function main: .LFB22: .cfi_startproc subq $8, %rsp #, .cfi_def_cfa_offset 16 movl $.LC0, %edi #, call puts # xorl %eax, %eax # addq $8, %rsp #, .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE22: .size main, .-main .ident "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1" .section .note.GNU-stack,"",@progbits

Saída do native:

Código Selecionar Expandir
galactus@galactus-desktop:~$ cat march_native.s .file "teste.c" # GNU C (Ubuntu/Linaro 4.5.1-7ubuntu2) version 4.5.1 (x86_64-linux-gnu) # compiled by GNU C version 4.5.1, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2 # GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 # options passed: teste.c -D_FORTIFY_SOURCE=2 -march=core2 -mcx16 -msahf # -mpopcnt -msse4.2 --param l1-cache-size=32 --param l1-cache-line-size=64 # --param l2-cache-size=8192 -mtune=core2 -auxbase-strip march_native.s -O2 # -fverbose-asm -fstack-protector # options enabled: -falign-loops -fargument-alias # -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg # -fcaller-saves -fcommon -fcprop-registers -fcrossjumping # -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks # -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types # -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse # -fgcse-lm -fguess-branch-probability -fident -fif-conversion # -fif-conversion2 -findirect-inlining -finline # -finline-functions-called-once -finline-small-functions -fipa-cp # -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots # -fira-share-spill-slots -fivopts -fkeep-static-consts # -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings # -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move # -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return # -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop # -fsched-critical-path-heuristic -fsched-dep-count-heuristic # -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic # -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic # -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros # -fsplit-ivs-in-unroller -fsplit-wide-types -fstack-protector # -fstrict-aliasing -fstrict-overflow -fthread-jumps -ftoplevel-reorder # -ftrapping-math -ftree-builtin-call-dce -ftree-ccp -ftree-ch # -ftree-copy-prop -ftree-copyrename -ftree-cselim -ftree-dce # -ftree-dominator-opts -ftree-dse -ftree-forwprop -ftree-fre # -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize # -ftree-parallelize-loops= -ftree-phiprop -ftree-pre -ftree-pta # -ftree-reassoc -ftree-scev-cprop -ftree-sink -ftree-slp-vectorize # -ftree-sra -ftree-switch-conversion -ftree-ter -ftree-vect-loop-version # -ftree-vrp -funit-at-a-time -funwind-tables -fvect-cost-model # -fverbose-asm -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387 # -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387 # -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mpopcnt -mpush-args # -mred-zone -msahf -msse -msse2 -msse3 -msse4 -msse4.1 -msse4.2 -mssse3 # -mtls-direct-seg-refs # Compiler executable checksum: 55ab920cb876e6d7d91c14d105b791d9 .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Hello world" .text .p2align 4,,15 .globl main .type main, @function main: .LFB22: .cfi_startproc subq $8, %rsp #, .cfi_def_cfa_offset 16 movl $.LC0, %edi #, call puts # xorl %eax, %eax # addq $8, %rsp #, .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE22: .size main, .-main .ident "GCC: (Ubuntu/Linaro 4.5.1-7ubuntu2) 4.5.1" .section .note.GNU-stack,"",@progbits

E agora 4d4c47? Eu coloco o quê na linha do HOSTCFLAGS do MakeFile do Kernel?

A opção -march=native foi a mais completa, sem dúvida! Ela especifica os três níveis de Caches do meu processador e todas as instruções SSE suportadas por ele! A opção Core2 não especifica os caches e não coloca todas as instruções do processador! A genérica é muito pior!

@galactus, na boa, vai de march=native mesmo, faça um teste mais completo aé e passa o resultado para nois, beleza ?

vou continuar testando aqui também, se houver mesmo muita diferença retorno mais informação.

valeu.

Citação de: vampire_thunder online 15 de Janeiro de 2011, 01:23
Cara, que maneiro isso aí. Mas dá até medo

Pra q serve o teste.c? Só para chamar o gcc?

Aqui apareceram umas informações interessantes:

Código Selecionar Expandir
.file "teste.c" # GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu) # compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2 # warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7. # warning: MPC header version 0.8.2 differs from library version 0.8.1. # GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 # options passed: teste.c -march=amdfam10 -mcx16 -msahf -mpopcnt -mabm # --param l1-cache-size=64 --param l1-cache-line-size=64 --param

Parece que tem umas libs desatualizadas.

@vampire_thunder, teste.c é o codigo-fonte pare ser compilado pelo gcc

procure mais informação sobre linguagem C na net, é um pouco complicado, mas com força de vontade aprenderá facil facil

valeu.

galactus · 15 de Janeiro de 2011, 02:10

Tá bom 4d4c47, vou tentar apenas com o march=native pra ver o que rola!

Hqxriven · 15 de Janeiro de 2011, 02:17

Uma coisa eu não sabia vendo esse teste -msse4 -msse4.1 -msse4.2 -mssse3

Mesmo colocando core 2 ele não engloba os novos cores i3 a i7...

Então sempre que não usarmos native e nosso processador for mais novo que os listados no site usaremos instruções "incompletas"

galactus · 15 de Janeiro de 2011, 09:17

Citação de: Hqxriven online 15 de Janeiro de 2011, 02:17
Uma coisa eu não sabia vendo esse teste -msse4 -msse4.1 -msse4.2 -mssse3

Mesmo colocando core 2 ele não engloba os novos cores i3 a i7...

Então sempre que não usarmos native e nosso processador for mais novo que os listados no site usaremos instruções "incompletas"

Hqx, você não tinha um config com a opção de usar os CFLAGS no menuconfig?

O kernel do DiamondMax tinha, mas nenhum link dele está funcionando no momento!

Algum outro lugar para pegar uma config que ativa o CFLAGS?

vampire_thunder · 15 de Janeiro de 2011, 09:53

Citação de: 4d4c47 online 15 de Janeiro de 2011, 01:59

@vampire_thunder, teste.c é o codigo-fonte pare ser compilado pelo gcc

procure mais informação sobre linguagem C na net, é um pouco complicado, mas com força de vontade aprenderá facil facil

valeu.

Já desisti. Acho que esse tipo de coisa não dá para aprender sem um professor. Já faço minhas compilações desde que usava Windows, compilando o emulador Mame, lá para os meados da década passada, e realmente queria mesmo aprender C.

Por exemplo, eu nem sei onde vou colocar essa "native"

Lá vai meu teste completo:

a.s

Código Selecionar

	.file	"teste.c"
# GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu)
#	compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7.
# warning: MPC header version 0.8.2 differs from library version 0.8.1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -mtune=generic -march=x86-64 -auxbase-strip a.s
# -fverbose-asm
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg -fcommon
# -fdelete-null-pointer-checks -fdwarf2-cfi-asm -fearly-inlining
# -feliminate-unused-debug-types -ffunction-cse -fgcse-lm -fident
# -finline-functions-called-once -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-debug-strings
# -fmove-loop-invariants -fpeephole -freg-struct-return
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -ftrapping-math -ftree-cselim -ftree-forwprop
# -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-optimize
# -ftree-parallelize-loops= -ftree-phiprop -ftree-pta -ftree-reassoc
# -ftree-scev-cprop -ftree-slp-vectorize -ftree-vect-loop-version
# -funit-at-a-time -funwind-tables -fvect-cost-model -fverbose-asm
# -fzero-initialized-in-bss -m128bit-long-double -m64 -m80387
# -maccumulate-outgoing-args -malign-stringops -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4
# -mpush-args -mred-zone -msse -msse2 -mtls-direct-seg-refs

# Compiler executable checksum: a968648126f8c105e89f7b7d56fd651a

	.section	.rodata
.LC0:
	.string	"Hello world"
	.text
.globl main
	.type	main, @function
main:
.LFB0:
	.cfi_startproc
	pushq	%rbp	#
	.cfi_def_cfa_offset 16
	movq	%rsp, %rbp	#,
	.cfi_offset 6, -16
	.cfi_def_cfa_register 6
	movl	$.LC0, %edi	#,
	call	puts	#
	movl	$0, %eax	#, D.3275
	leave
	.cfi_def_cfa 7, 8
	ret
	.cfi_endproc
.LFE0:
	.size	main, .-main
	.ident	"GCC: (Debian 4.5.2-1) 4.5.2"
	.section	.note.GNU-stack,"",@progbits

k8:

Código Selecionar

	.file	"teste.c"
# GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu)
#	compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7.
# warning: MPC header version 0.8.2 differs from library version 0.8.1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -march=k8 -auxbase-strip march_k8.s -O2
# -fverbose-asm
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fsplit-wide-types -fstrict-aliasing
# -fstrict-overflow -fthread-jumps -ftoplevel-reorder -ftrapping-math
# -ftree-builtin-call-dce -ftree-ccp -ftree-ch -ftree-copy-prop
# -ftree-copyrename -ftree-cselim -ftree-dce -ftree-dominator-opts
# -ftree-dse -ftree-forwprop -ftree-fre -ftree-loop-im -ftree-loop-ivcanon
# -ftree-loop-optimize -ftree-parallelize-loops= -ftree-phiprop -ftree-pre
# -ftree-pta -ftree-reassoc -ftree-scev-cprop -ftree-sink
# -ftree-slp-vectorize -ftree-sra -ftree-switch-conversion -ftree-ter
# -ftree-vect-loop-version -ftree-vrp -funit-at-a-time -funwind-tables
# -fvect-cost-model -fverbose-asm -fzero-initialized-in-bss
# -m128bit-long-double -m3dnow -m64 -m80387 -maccumulate-outgoing-args
# -malign-stringops -mfancy-math-387 -mfp-ret-in-387 -mfused-madd -mglibc
# -mieee-fp -mmmx -mno-sse4 -mpush-args -mred-zone -msse -msse2
# -mtls-direct-seg-refs

# Compiler executable checksum: a968648126f8c105e89f7b7d56fd651a

	.section	.rodata.str1.1,"aMS",@progbits,1
.LC0:
	.string	"Hello world"
	.text
	.p2align 4,,15
.globl main
	.type	main, @function
main:
.LFB11:
	.cfi_startproc
	subq	$8, %rsp	#,
	.cfi_def_cfa_offset 16
	movl	$.LC0, %edi	#,
	call	puts	#
	xorl	%eax, %eax	#
	addq	$8, %rsp	#,
	.cfi_def_cfa_offset 8
	ret
	.cfi_endproc
.LFE11:
	.size	main, .-main
	.ident	"GCC: (Debian 4.5.2-1) 4.5.2"
	.section	.note.GNU-stack,"",@progbits

core2:

Código Selecionar

	.file	"teste.c"
# GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu)
#	compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7.
# warning: MPC header version 0.8.2 differs from library version 0.8.1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -march=core2 -auxbase-strip march_core2.s -O2
# -fverbose-asm
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fsplit-wide-types -fstrict-aliasing
# -fstrict-overflow -fthread-jumps -ftoplevel-reorder -ftrapping-math
# -ftree-builtin-call-dce -ftree-ccp -ftree-ch -ftree-copy-prop
# -ftree-copyrename -ftree-cselim -ftree-dce -ftree-dominator-opts
# -ftree-dse -ftree-forwprop -ftree-fre -ftree-loop-im -ftree-loop-ivcanon
# -ftree-loop-optimize -ftree-parallelize-loops= -ftree-phiprop -ftree-pre
# -ftree-pta -ftree-reassoc -ftree-scev-cprop -ftree-sink
# -ftree-slp-vectorize -ftree-sra -ftree-switch-conversion -ftree-ter
# -ftree-vect-loop-version -ftree-vrp -funit-at-a-time -funwind-tables
# -fvect-cost-model -fverbose-asm -fzero-initialized-in-bss
# -m128bit-long-double -m64 -m80387 -maccumulate-outgoing-args
# -malign-stringops -mcx16 -mfancy-math-387 -mfp-ret-in-387 -mfused-madd
# -mglibc -mieee-fp -mmmx -mno-sse4 -mpush-args -mred-zone -msahf -msse
# -msse2 -msse3 -mssse3 -mtls-direct-seg-refs

# Compiler executable checksum: a968648126f8c105e89f7b7d56fd651a

	.section	.rodata.str1.1,"aMS",@progbits,1
.LC0:
	.string	"Hello world"
	.text
	.p2align 4,,15
.globl main
	.type	main, @function
main:
.LFB11:
	.cfi_startproc
	subq	$8, %rsp	#,
	.cfi_def_cfa_offset 16
	movl	$.LC0, %edi	#,
	call	puts	#
	xorl	%eax, %eax	#
	addq	$8, %rsp	#,
	.cfi_def_cfa_offset 8
	ret
	.cfi_endproc
.LFE11:
	.size	main, .-main
	.ident	"GCC: (Debian 4.5.2-1) 4.5.2"
	.section	.note.GNU-stack,"",@progbits

Native:

Código Selecionar

	.file	"teste.c"
# GNU C (Debian 4.5.2-1) version 4.5.2 (x86_64-linux-gnu)
#	compiled by GNU C version 4.5.2, GMP version 4.3.2, MPFR version 3.0.0-p3, MPC version 0.8.2
# warning: MPFR header version 3.0.0-p3 differs from library version 3.0.0-p7.
# warning: MPC header version 0.8.2 differs from library version 0.8.1.
# GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072
# options passed:  teste.c -march=amdfam10 -mcx16 -msahf -mpopcnt -mabm
# --param l1-cache-size=64 --param l1-cache-line-size=64 --param
# l2-cache-size=512 -mtune=amdfam10 -auxbase-strip march_native.s -O2
# -fverbose-asm
# options enabled:  -falign-loops -fargument-alias
# -fasynchronous-unwind-tables -fauto-inc-dec -fbranch-count-reg
# -fcaller-saves -fcommon -fcprop-registers -fcrossjumping
# -fcse-follow-jumps -fdefer-pop -fdelete-null-pointer-checks
# -fdwarf2-cfi-asm -fearly-inlining -feliminate-unused-debug-types
# -fexpensive-optimizations -fforward-propagate -ffunction-cse -fgcse
# -fgcse-lm -fguess-branch-probability -fident -fif-conversion
# -fif-conversion2 -findirect-inlining -finline
# -finline-functions-called-once -finline-small-functions -fipa-cp
# -fipa-pure-const -fipa-reference -fipa-sra -fira-share-save-slots
# -fira-share-spill-slots -fivopts -fkeep-static-consts
# -fleading-underscore -fmath-errno -fmerge-constants -fmerge-debug-strings
# -fmove-loop-invariants -fomit-frame-pointer -foptimize-register-move
# -foptimize-sibling-calls -fpeephole -fpeephole2 -freg-struct-return
# -fregmove -freorder-blocks -freorder-functions -frerun-cse-after-loop
# -fsched-critical-path-heuristic -fsched-dep-count-heuristic
# -fsched-group-heuristic -fsched-interblock -fsched-last-insn-heuristic
# -fsched-rank-heuristic -fsched-spec -fsched-spec-insn-heuristic
# -fsched-stalled-insns-dep -fschedule-insns2 -fshow-column -fsigned-zeros
# -fsplit-ivs-in-unroller -fsplit-wide-types -fstrict-aliasing
# -fstrict-overflow -fthread-jumps -ftoplevel-reorder -ftrapping-math
# -ftree-builtin-call-dce -ftree-ccp -ftree-ch -ftree-copy-prop
# -ftree-copyrename -ftree-cselim -ftree-dce -ftree-dominator-opts
# -ftree-dse -ftree-forwprop -ftree-fre -ftree-loop-im -ftree-loop-ivcanon
# -ftree-loop-optimize -ftree-parallelize-loops= -ftree-phiprop -ftree-pre
# -ftree-pta -ftree-reassoc -ftree-scev-cprop -ftree-sink
# -ftree-slp-vectorize -ftree-sra -ftree-switch-conversion -ftree-ter
# -ftree-vect-loop-version -ftree-vrp -funit-at-a-time -funwind-tables
# -fvect-cost-model -fverbose-asm -fzero-initialized-in-bss
# -m128bit-long-double -m3dnow -m64 -m80387 -mabm
# -maccumulate-outgoing-args -malign-stringops -mcx16 -mfancy-math-387
# -mfp-ret-in-387 -mfused-madd -mglibc -mieee-fp -mmmx -mno-sse4 -mpopcnt
# -mpush-args -mred-zone -msahf -msse -msse2 -msse3 -msse4a
# -mtls-direct-seg-refs

# Compiler executable checksum: a968648126f8c105e89f7b7d56fd651a

	.section	.rodata.str1.1,"aMS",@progbits,1
.LC0:
	.string	"Hello world"
	.text
	.p2align 5,,31
.globl main
	.type	main, @function
main:
.LFB11:
	.cfi_startproc
	subq	$8, %rsp	#,
	.cfi_def_cfa_offset 16
	movl	$.LC0, %edi	#,
	call	puts	#
	xorl	%eax, %eax	#
	addq	$8, %rsp	#,
	.cfi_def_cfa_offset 8
	ret
	.cfi_endproc
.LFE11:
	.size	main, .-main
	.ident	"GCC: (Debian 4.5.2-1) 4.5.2"
	.section	.note.GNU-stack,"",@progbits

Meu processador é um AMD Phenom II N830, com 3 núcleos.

galactus · 15 de Janeiro de 2011, 10:08

vampire_thunder, o Native vai dentro do MakeFile do Kernel!

Descompacta o source do kernel vanilla que você usa, abre o Makefile que está lá dentro, mais ou menos na página 4 tem isso aqui:

# SHELL used by kbuild
CONFIG_SHELL := $(shell if [ -x "$$BASH" ]; then echo $$BASH; \
else if [ -x /bin/bash ]; then echo /bin/bash; \
else echo sh; fi ; fi)

HOSTCC = gcc
HOSTCXX = g++
HOSTCFLAGS = -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -march=native -fomit-frame-pointer
HOSTCXXFLAGS = "${CFLAGS}"

Ali na linha do HOSTCFLAGS você coloca o -march=native!

Infelizmente não é só aí que tem que mudar, mas já muda alguma coisa!

maiconramones · 15 de Janeiro de 2011, 13:12

Pessoal percebi no meu kernel novo o driver da placa de rede sem fio não está funcionando tem que reinstalar alguma coisa ou tem que marcar alguma coisa nas opções de compilação?

Sobre o que falaram do trio 1000 + low latency +performance.. Eu confesso que não marquei essa possibilidade de performance... aonde que eu habilito isso?

Stivekx · 15 de Janeiro de 2011, 17:41

Citação de: maiconramones online 15 de Janeiro de 2011, 13:12
Pessoal percebi no meu kernel novo o driver da placa de rede sem fio não está funcionando tem que reinstalar alguma coisa ou tem que marcar alguma coisa nas opções de compilação?

Sobre o que falaram do trio 1000 + low latency +performance.. Eu confesso que não marquei essa possibilidade de performance... aonde que eu habilito isso?

se voce usou o kernelcheck, é na hora que aparece uma janela cheia de coisas pra configurar..
Se não é quando tu digita make config

Stivekx · 15 de Janeiro de 2011, 17:43

Achei esse tutorial explicando várias coisas especialmente da configuração do kernel, como otimizar pro seu processador, dicas de segurança...
Vou ler ele certinho na prox compilação.

http://www.faqs.org/docs/securing/secopt-kernel.html

Edit.:
Mais tutoriais legais sobre otimização nesse livro:
http://www.faqs.org/docs/securing/
To gostando das dicas (Y)

4d4c47 · 16 de Janeiro de 2011, 03:39

Citação de: galactus online 15 de Janeiro de 2011, 10:08
vampire_thunder, o Native vai dentro do MakeFile do Kernel!

Descompacta o source do kernel vanilla que você usa, abre o Makefile que está lá dentro, mais ou menos na página 4 tem isso aqui:

# SHELL used by kbuild
CONFIG_SHELL := $(shell if [ -x "$$BASH" ]; then echo $$BASH; \
else if [ -x /bin/bash ]; then echo /bin/bash; \
else echo sh; fi ; fi)

HOSTCC = gcc
HOSTCXX = g++
HOSTCFLAGS = -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -march=native -fomit-frame-pointer
HOSTCXXFLAGS = "${CFLAGS}"

Ali na linha do HOSTCFLAGS você coloca o -march=native!

Infelizmente não é só aí que tem que mudar, mas já muda alguma coisa!

como diria o grande @galactus:

Senhoras e Senhores,

CONSEGUIR compilar o kernel todo com a cflag -march=native

aí vai um simples tutorial:

1 - baixa o kernel linux-2.6.36.tar.bz2

2 - descompacta em /usr/src

3 - baixa o patch do kernel zen, exatamente este http://downloads.zen-kernel.org/2.6.36/2.6.36-zen1.patch.lzma

4 - descompacta o patch zen na pasta /usr/src, depois aplique o patch zen no kernel linux-2.6.36

5 - pegue o .config do omnislash e copie para /usr/src/linux-2.6.36

6 - executa um make menuconfig e marque o seu processador exato ( não pentium4 ou outra coisa, marque a sua arquitetura exata)

7 - abaixo da escolha do processador, tem uma opção de -march=native ( isso é só de enfeite, não serve pra nada, mas marque ela mesmo assim)

8 - salve tudo e feche o menuconfig, agora vamos pra parte que interessa...

9 - edit o arquivo Makefile:

Citação de: galactus online 15 de Janeiro de 2011, 10:08
... mais ou menos na página 4 tem isso aqui:

# SHELL used by kbuild
CONFIG_SHELL := $(shell if [ -x "$$BASH" ]; then echo $$BASH; \
else if [ -x /bin/bash ]; then echo /bin/bash; \
else echo sh; fi ; fi)

HOSTCC = gcc
HOSTCXX = g++
HOSTCFLAGS = -Wall -Wmissing-prototypes -Wstrict-prototypes -O2 -march=native -fomit-frame-pointer
HOSTCXXFLAGS = "${CFLAGS}"

Ali na linha do HOSTCFLAGS você coloca o -march=native!

10 - agora começa a ***** toda, execute: cd /usr/src/linux-2.6.36/arch/x86/

11 - dentro dessa pasta tem todos os codigo-fonte da parte principal do kernel, dentro dessa pasta existe varias outras pastas (e varias outras dentro delas), e o que voce vai fazer é adentrar uma por uma e procurar o arquivo Makefile dentro delas, e edita-lo: onde tiver -march=XXXX ou -mtune=XXXX ou -mcpu=XXXX, vc vai substituir por -march=native (obs: alguns Makefile não tem cflag nenhuma, então não mexa neles, só modifique os que tem -march=XXXX ou -mtune=XXXX ou -mcpu=XXXX), modifique tambem os arquivos: Kbuild, Kconfig, Kconfig.cpu, Kconfig.debug, Makefile_32.cpu( esse por exemplo está lotado de -march=XXXX e -mtune=XXXX, subistitua tudo sem medo de ser feliz)

12 - depois de tudo isso, comesse a compilação ( pra ver se tá compilando com march=native mesmo, abra um outro terminal e execute após 2 minutos de compilação: ps ax | grep gcc | grep march )

13 - pronto, taí o kernel exato pro seu processador

rapaz, o troço ficou bem mais rápido mesmo (nossa senhora, não sei se é só impressão minha só sei que a diferença é grande, :O )

espero ter ajudado, agradeço ao grande galactus pela ideia, fico na espera do novo omnislash, valeu a todos.