flat assembler
Message board for the users of flat assembler.
![]() Goto page Previous 1, 2 |
Author |
|
dvorakvik 01 Sep 2004, 13:13
Dear Mike,
many thanks for revisiting the SIS900.INC driver. Commenting the lines 1034, 1035 etc ;****make sure packet transmitted successfully**** ; mov esi,10 ; call delay_ms SOLVES THE SIS900.INC driver problem. Now Realtek 8029, Realtek 8139 and SiS900 based drivers behaved similar. Time delay between two UDP packets cca 85 microseconds with MIKE.ASM test program. <For others see above for download> Viktor |
|||
![]() |
|
Matrix 07 Sep 2004, 23:50
dvorakvik wrote: Using 10 msec delay ebx = 1 function 5 int 0x40 solves the problem ... hey, how can u count on RDTSC? it is not present on 486 and oldre processors, and how do you determine the divisor? - the accurate cpu frequency, i give u an advise: use pit timer directly ![]() MATRIX |
|||
![]() |
|
dvorakvik 08 Sep 2004, 18:20
Look at Intel HW manuals. Shortly newer processors all have RDTSC 64long binary couter, which counts CPU clock frequency RDTSC => EAX = LSB EDX => MSB.
This is the most precise timing at CPU. Viktor |
|||
![]() |
|
Matrix 08 Sep 2004, 23:59
dvorakvik wrote: Look at Intel HW manuals. Shortly newer processors all have RDTSC 64long binary couter, which counts CPU clock frequency RDTSC => EAX = LSB EDX => MSB. look at my other thread about timing http://board.flatassembler.net/topic.php?t=2207 MATRIX |
|||
![]() |
|
Goto page Previous 1, 2 < Last Thread | Next Thread > |
Forum Rules:
|
Copyright © 1999-2025, Tomasz Grysztar. Also on GitHub, YouTube.
Website powered by rwasa.