On Wed, 2009-04-22 at 23:52 +0800, SimonQian wrote: > Hi, > How about adding support to other microcontrollers for example AVR and > C8051F. > Although debugging protocol is not open, but programming protocol is.
I hereby propose the mission statement for the OpenOCD project to be: "OpenOCD will be an Unlimited On-Chip Debugger and Programmer." I am not great at writing these off-the-cuff, but this should convey the point required for this conversation. It is worth noting that this so called "mission statement" is only now being seen by the community at large. Maybe it will be endorsed by everyone as-is; maybe I will awake to a bag of dog poop on my doorstep. Maybe somewhere in the middle. If AVR programming support comes to exists in OpenOCD, someone might then be motivated to find a way to then support the debugging protocol. > I implement AVR_JTAG and C8051F_JTAG in my project vsprog. Assuming it's license-compatible, we could probably swap code, yah? > Actually vsprog can support many other MCUs, but OpenOCD can currently > only > handle with standard JTAG interface. > I can rewrite these codes to make it work in OpenOCD when I have time. > But I don't know is these supports out of OpenOCD's box? The mission statement that I proposed derives from the fact that I can imagine a very big box. Others may have smaller boxes in mind. Practically speaking, the implementation is a much smaller box. I think there are several factors to consider when a project like this considers a request that is far outside its existent box: 1) Is the architecture ready to push on the side that needs expanding? 2) Given that, how hard will it be to keep it from imploding over time? 3) How much interest (and by that I mean development resources) can be brought to bear on development? There are other factors that I probably missed here. I am not familiar enough with the code to make a full assessment of the first item, but I believe that completely new architectures should probably wait until we manage to clean up the existing code a bit more. I don't know about the second item, *but* the third item could moot the others. Linus believed Linux would be x86-only until it was ported to Alpha. We can probably all recount similar stories from other areas of life. Anyone that tells you that it cannot be done can't do it themselves, but you might be the person that can do it when you believe that you can. That may sound ridiculously pedantic, but I believe that many people fail to grasp how often they control their own limitations in this way. If you do decide to take a heroic leap, just be prepared to help with the eventual merge, or you might need to maintain the patches out-of-tree until we clean up any existing mess for you. Consider the statistical life expectancy of heroes before you take the plunge, and make sure to post regular reports of your progress during development. The community can keep you from unnecessarily "falling" off of a cliff. Cheers, Zach _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
