On Mon, May 21, 2012 at 11:09 PM, Tomek CEDRO <[email protected]> wrote: > On Mon, May 21, 2012 at 11:01 PM, Andreas Fritiofson wrote: >>> Could you please tell what is the difference between new driver and >>> ft2232_bitbang and ft2232_transfer that I use for SWD as part of >>> LibSWD - maybe I could improve its speed too? :-) >> >> I wish I had more time to look into and discuss your and others' work >> on SWD but I haven't right now. But I'm sure that it would be >> extremely simple to write an SWD driver using the same MPSSE >> abstraction as my new ftdi driver is using. I wrote an SWD driver over >> raw MPSSE (inside the old ft2232 driver) in less than a weekend, three >> years ago. And it would have been less work if I had this MPSSE >> abstraction. And faster too, even if my previous SWD work was on par >> or better than JTAG for the same target/interface, as far as I >> remember. >> >> I won't write another SWD driver though, until the proper framework is >> in place. The difficulty with SWD is not in the driver, but in core >> OpenOCD. I'm not really satisfied with the current transport >> abstraction. My main motivation for rewriting the driver was the need >> for a solid base for experimentation with the core interfaces in >> OpenOCD. And I'll probably continue with that, as time allows. >> >> /Andreas > > Hello Andreas :-) Thanks for quick response :-) The LibSWD is the > "core" of swd transport for OpenOCD, drivers are also there, I was > wondering what else was there except MPSSE, but I can see this relates > more to OpenOCD internals and JTAG framework. > > I would suggest to create "src/interface/drivers" folder and put all > drivers there (each driver in its own file or folder), so we know > these are part of the interface framework (not jtag as it is now). I > have put some generic interface transport and target independent in > there, please take a look at > http://repo.or.cz/w/openocd/libswd.git/tree/HEAD:/src/interface and > see it this is good solution..?
Just did a quick check and this looks mildly like what I implemented inside the new ftdi driver to be able to configure reset signal layout from Tcl to replace the layout specific mess that clutters the ft2232 driver. I still don't get what this has to do with SWD. I think a SWD driver's interface to the OpenOCD side should be as high level as possible, not twiddling bits. But I would have to dig deeper to make sense of it and give any useful feedback and I haven't got time for that now. /Andreas ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ OpenOCD-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/openocd-devel
