2007/10/11, Stanislav Meduna <[EMAIL PROTECTED]>: > > SIE wrote: > > > We have a project that will communicate with two different OS(uClinux > > and uC/OS-II) through UART. > > What speed are you using,what receive FIFO size does the S3C4510B's > UART have and what serial driver are you using? I can't read > http://www.mculand.com/sub1/mcu/arm7tdmi_device/s3c4510.htm > but if the source is correct, it seems that the FIFO is 32 bytes.
S3C4510B,a little different from the page you mentioned,does not have a fifo for uart. > However,uClinux is not a real-time OS, > > Well, even (hard) real-time only guarantees that you get > the processing in a specified time, and not that the time > is short enough to be able to handle the UART ;) I got it,thanks. With 32 byte and 115 kbps you need to service an interrupt > in 2.6 ms, which should be more than enough. Is there > any other hardware that might be blocking interrupts > for this long? Writing to flash or something like that? If everytime a byte arrives we bring up an interrupt,that may be too frequently for uClinux to process. > We think that it may be caused by that uClinux's > > speed of handling the interrupt is not fast enough to follow OS-II. > > First confirm what exactly is not up to speed - is it > the kernel that fails to service the interrupt in time, > or is it your application that is not able to take > the data? You might want to add a byte counter into > the kernel and to your application, send a burst of data > and compare them - if they match, it is the kernel, > if not, it is the app. > > > Does any good solution to reduce the packet-lost? > > Depends on the application - you need some kind of > flow control, but a software one won't work if the > problem is at the interrupt servicing level. It seems that I have to try one-byte-per-time flow control? Thanks . -- > Stano > _______________________________________________ > uClinux-dev mailing list > [email protected] > http://mailman.uclinux.org/mailman/listinfo/uclinux-dev > This message was resent by [email protected] > To unsubscribe see: > http://mailman.uclinux.org/mailman/options/uclinux-dev >
_______________________________________________ uClinux-dev mailing list [email protected] http://mailman.uclinux.org/mailman/listinfo/uclinux-dev This message was resent by [email protected] To unsubscribe see: http://mailman.uclinux.org/mailman/options/uclinux-dev
