| ??? 02/14/01 09:06 Read: times |
#9291 - RE: use of multiply in large data arrays |
Peter,
MUL A,B will cost 24 instruction cycles with the hardware accelerator. Beneath I have paste the assembler code generated by th compiler. As you see it takes a lot of time to just clear one byte of an member. MOV B,#011H MOV A,_47 MUL AB MOV R2,A ADD A,# LOW (_PCcommBuffer) MOV DPL,A CLR A ADDC A,# HIGH (_PCcommBuffer) MOV DPH,A CLR A MOVX @DPTR,A Patrick |



