hi I am writing a patch to the 2.4.19 kernel to measure statistics of packets sent on the n/w interfaces. I have a function that is called when net/core/dev.c:net_rx_action() is invoked to record received packets. Now this works fine when small amounts of data are sent on the n/w. But doing a huge file transfer brings about a panic. The trace says a NULL pointer reference occured and that the interrupt handler was killed. The EIP points to an instruction in my routine (as figured from System.map). I am checking for NULL values before I access anything. This is what I think is the problem. This routine of mine takes too long to be part of an interrupt handler. So, when I am doing a massive data transfer, the device is probably sending more interrupts than can be handled and something goes bad. Is this possible? And how do I fix this. OR more generally, if I have a interrupt handler which takes long, is there some trick by which I can do less work in the handler and move the bulk of the work somewhere else. Thanks in advance Chandan -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
