hi, it will do, linux also uses flat memory model anyway, what u do is write a loadable kernel module, load the interrupt table(u must be knowing sidt %0) etc,used GNU inline assembly for that, change the interrupt stub's address in yr idt,i.e original address in idt to yr stub, this means whenver hardware interrupts occurs it will invoke yr stub ,and in that stub u do the work, but probably hardware interrupts are very time critical so for using yr work and not blocking the system, u have to used tasklets, and give back control to original ISR, the same mechanism is used in RTAI-RTHAL (Real time hardware abstarction layer ) for extending linux ro real time linux,ofcourse much more things are there in RTAI,they had also make the abstarction of every call to cli/sti etc. if u dont sure how to do,see RTAI site documenation. Prasanna --- Nilanjan Roychowdhury <nilanjan@genesis-microchip.com> wrote: > > I have a simple driver in Vxworks ( lets say in a > flat memory RTOS) > What I do is whenever my ISR detects any H/W intr it > posts a message to a task T1. T 1 does the > processing and then notifies the application task > T2. > > How do I do the same in linux?? Is it possible at > all. > Regards > Nilanjan > > -- > Kernelnewbies: Help each other learn about the Linux > kernel. > Archive: > http://mail.nl.linux.org/kernelnewbies/ > FAQ: http://kernelnewbies.org/faq/ > __________________________________ Do you Yahoo!? Yahoo! Photos: High-quality 4x6 digital prints for 25¢ http://photos.yahoo.com/ph/print_splash -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/