firmware tracing

[Frank Ch. Eigler]

zhiwei.ying wrote:

> Has anyone thought about using systemtap to trace firmware code? Any
> information?

To the extent the firmware is of the form that the host kernel
interprets, like ACPI, that's straightforward.  One'd probe key points
in the interpreter.  If the firmware is binary code for some embedded
microcontroller, and is run without the participation of the host CPU,
then I can't think of simple uses of systemtap to trace the firmware.
Tracing interactions (host code that deals with interrupts, PCI
transactions, DMA) should be doable though.

- FChE