| 07:23.59 | *** join/#openjtag rd_ (n=redragon@212.12.163.101) |
| 08:47.34 | *** join/#openjtag xela (n=xela@p549C9AB1.dip0.t-ipconnect.de) |
| 08:47.49 | xela | Hey all, Hey drath |
| 08:55.24 | drath | hey xela |
| 08:56.05 | xela | i am back from my vacation... |
| 08:56.16 | xela | trying out the amontec jtag-tiny |
| 08:56.25 | xela | works really fine and so cheap! |
| 09:06.14 | drath | xela: heh, yeah, it's just "cute" |
| 09:06.59 | drath | xela: only problem with the tiny is the limited voltage range - i learned that lesson the hard way, when I recommended the Tiny to someone who wanted to debug an XScale based phone, which unfortunately had 1.8V JTAG levels |
| 09:07.16 | xela | good to know |
| 09:07.23 | xela | i noticed (doesnt have to do with the tiny) that gdb still seems to get unexpected data |
| 09:09.53 | drath | hum, I lost my backlog when I rebooted the server about a week ago - don't think I remember this |
| 09:10.44 | xela | is there a way to stop gdb from fetching data of 0xffffffff? |
| 09:11.03 | xela | think we talked about gdb examining the stack |
| 09:15.13 | xela | (gdb) s |
| 09:15.15 | xela | Cannot access memory at address 0xffffffff |
| 09:15.16 | xela | Debugger exited abnormally with code 5 |
| 09:51.50 | xela | any idea? |
| 09:52.24 | xela | or maybe we can define memory areas in OpenOCD so no Data Trap or alike occurs? |
| 09:56.28 | drath | the concept of "memory areas" would certainly be a good thing |
| 09:56.48 | drath | we'll also need something like that to implement transparent flash support |
| 09:57.03 | drath | but GDB chose to use XML to describe this kind of things |
| 09:57.23 | drath | and I nenver felt comfortable usng XML, so I ignored this for now |
| 09:57.38 | xela | huh |
| 09:57.38 | xela | ? |
| 09:57.38 | xela | where does it use XML? |
| 09:58.27 | drath | http://sourceware.org/gdb/current/onlinedocs/gdb_33.html#SEC708 |
| 09:59.01 | drath | GDB expects the remote server (i.e. OpenOCD) to supply a XML file describing memory maps, so GDB knows when to use memory commands, and when to use flash commands |
| 09:59.05 | xela | so, one could use gdb to program directly? |
| 09:59.09 | drath | yeah |
| 10:00.27 | xela | ah this sounds good |
| 10:00.27 | xela | but right now it returns only FOO? |
| 10:00.27 | drath | of course that wouldn't be necessary for your particular idea, but if openocd is going to implement memory maps, then GDB's way is certainly the way to go |
| 10:00.27 | drath | it doesn't support any of those query commands atm, yeah |
| 10:01.21 | xela | ah i see |
| 10:01.30 | xela | (well, have to run, will be back tomorrow) |
| 10:01.34 | xela | bye until then |
| 11:48.08 | *** join/#openjtag p2-mate (n=p2@cable-87-244-173-194.upc.chello.be) |
| 11:51.54 | *** join/#openjtag Chocobo (n=swinchen@strongbad.eece.maine.edu) |
| 13:14.58 | *** join/#openjtag prpplague^2 (n=dave@mail.americanmicrosystems.com) |
| 14:40.34 | *** join/#openjtag nm (n=hongtd@58.187.131.77) |
| 17:47.54 | *** join/#openjtag p2-mate (n=p2@cable-87-244-173-194.upc.chello.be) |
| 22:28.03 | *** join/#openjtag prpplague (n=dave@mail.americanmicrosystems.com) |
| 22:31.22 | *** join/#openjtag drath_ (i=vmaster@p5B07EB89.dip.t-dialin.net) |
| 23:47.40 | *** join/#openjtag p2-mate (n=p2@cable-87-244-173-194.upc.chello.be) |