M051如何计算出一条指令的运行时间？谢谢！

只看该作者 · 2016-3-14 16:51

1万 · 2016-3-14 19:01

这个是Cortex-M0的核心，你要考察的是这个核心的运行效率。

1万 · 2016-3-14 19:03

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

Cortex-M0指令和它们的周期数

续表

操作	描述	汇编指令	周期
	Any to Any	ADD Rd,Rd,Rm	1
	Any to PC	ADD PC,PC,Rm	3
	8-bit immediate	ADDS Rd,Rd,#<imm>	1
	With carry	ADCS Rd,Rd,Rm	1
	Immediate to SP	ADD SP,SP,#<imm>	1
	Form address from SP	ADD Rd,SP ,#<imm>	1
	Form address from PC	ADR Rd,<label>	1
Subtract	Lo to Lo	SUBS Rd,Rn,Rm	1
	3-bit immediate	SUBS Rd,Rn,#<imm>	1
	8-bit immediate	SUBS Rd,Rd,#<imm>	1
	With carry	SBCS Rd,Rd,Rm	1
	Immediate from SP	SUB SP,SP,#<imm>	1
	Negate	RSBS Rd,Rn,#0	1
Multiply	Multiply	MULS Rd,Rm,Rd	1或32
Compare	Compare	CMP Rn,Rm	1
	Negative	CMP Rn,Rm	1
	Immediate	CMP Rn,#<imm>	1
Logical	AND	ANDS Rd, Rd, Rm	1
	Exclusive OR	EORS Rd, Rd, Rm	1
	OR	ORRS Rd, Rd, Rm	1
	Bit clear	BICS Rd, Rd, Rm	1
	Move NOT	MVNS Rd, Rm	1
	AND test	TST Rn, Rm	1
Shift	Logical shift left by immediate	LSLS Rd, Rm, #<shift>	1
	Logical shift left by register	LSLS Rd, Rd, Rs	1
	Logical shift right by immediate	LSRS Rd, Rm, #<shift>	1
	Logical shift right by register	LSRS Rd, Rd, Rs	1
	Arithmetic shift right	ASRS Rd, Rm, #<shift>	1
	Arithmetic shift right by register	ASRS Rd, Rd, Rs	1
Rotate	Rotate right by register	RORS Rd, Rd, Rs	1
Load	Word, immediate offset	LDR Rd, [Rn, #<imm>]	2
	Halfword, immediate offset	LDRH Rd, [Rn, #<imm>]	2
	Byte, immediate offset	LDRB Rd, [Rn, #<imm>]	2
	Word, register offset	LDR Rd, [Rn, Rm]	2

续表

操作	描述	汇编指令	周期
	Halfword, register offset	LDRH Rd, [Rn, Rm]	2
	Signed halfword, register offset	LDRSH Rd, [Rn, Rm]	2
	Byte, register offset	LDRB Rd, [Rn, Rm]	2
	Signed byte, register offset	LDRSB Rd, [Rn, Rm]	2
	PC-relative	LDR Rd, <label>	2
	SP-relative	LDR Rd, [SP, #<imm>]	2
	Multiple, excluding base	LDM Rn!, {<loreglist>}	1+N
	Multiple, including base	LDM Rn, {<loreglist>}	1+N
Store	Word, immediate offset	STR Rd, [Rn, #<imm>]	2
	Halfword, immediate offset	STRH Rd, [Rn, #<imm>]	2
	Byte, immediate offset	STRB Rd, [Rn, #<imm>]	2
	Word, register offset	STR Rd, [Rn, Rm]	2
	Halfword, register offset	STRH Rd, [Rn, Rm]	2
	Byte, register offset	STRB Rd, [Rn, Rm]	2
	SP-relative	STR Rd, [SP, #<imm>]	2
	Multiple	STM Rn!, {<loreglist>}	1+N
Push	Push	PUSH {<loreglist>}	1+N
Push	Push with link register	PUSH {<loreglist>, LR}	1+N
Pop	Pop	POP {<loreglist>}	1+N
Pop	Pop and return	POP {<loreglist>, PC}	4+N
Branch	Conditional	B<cc> <label>	1或3
	Unconditional	B<label>	3
	With link	BL<label>	4
	With exchange	BX Rm	3
	With link and exchange	BLX Rm	3
Extend	Signed halfword to word	SXTH Rd, Rm	1
	Signed byte to word	SXTB Rd, Rm	1
	Unsigned halfword	UXTH Rd, Rm	1
	Unsigned byte	UXTB Rd, Rm	1
Reverse	Bytes in word	REV Rd, Rm	1
	Bytes in both halfwords	REV16 Rd, Rm	1
	Signed bottom half word	REVSH Rd, Rm	1
State change	Supervisor Call	SVC #<imm>	-
State change	Disable interrupts	CPSID i	1

操作	描述	汇编指令	周期
	Enable interrupts	CPSIE i	1
	Read special register	MRS Rd, <specreg>	4
	Write special register	MSR <specreg>, Rn	4
	Breakpoint	BKPT #<imm>	—
Hint	Send event	SEV	1
	Wait for event	WFE	2
	Wait for interrupt	WFI	2
	Yield	YIELD	1
Hint	No operation	NOP	1
Barriers	Instruction synchronization	ISB	4
	Data memory	DMB	4
	Data synchronization	DSB	4

ISO/IEC的C代码不能直接地获取一些Cortex-M0的指令。表7-23列举了CMSIS的C编译器中提供的部分内部函数用于产生这些指令。如果一个C编译器不支持恰当的内部函数，则需要用内嵌汇编来获取有关指令。
表7-23 CMSIS内部函数

CMSIS也提供了几个函数用于获取特殊的寄存器，如表7-24所示。
表7-24 CMSIS提供的用于获取特殊寄存器的函数

特殊寄存器	获取方式	CMSIS函数
PRIMASK	读	uint32_t __get_PRIMASK(void)
PRIMASK	写	void __set_PRIMASK(void)
CONTROL	读	Uint32_t __get_CONTROL(uint32_t value)
CONTROL	写	Void __set_CONTROL(uint32_t value)
MSP	读	Uint32_t __get_MSP(void)
MSP	写	Void __set_MSP(uint32_t TopOfMainStack)
PSP	读	Uint32_t __get_PSP(void)
PSP	写	Void __set_PSP(uint32_t TopOfMainStack)

只看该作者 · 2016-3-15 09:37

哦，如果周期是一的话，是不是就是1/主频，要考虑分频不？谢谢！

只看该作者 · 2016-3-15 14:54

每个芯片的指令运行时间是固定的吧，比如这个是cortex的核，arm的datasheet里面应该讲到

只看该作者 · 2016-3-15 20:27

你说的指令是汇编指令还是C语言的语句？

只看该作者 · 2016-3-16 09:48

这个得看你设置的时钟频率吧，不同指令运行时间不同