**** BEGIN LOGGING AT Sat Feb 03 02:59:59 2007 Feb 03 09:11:31 vmaster: everything works fine now, it was a hardware problem at my second SDRAM Feb 03 10:05:06 vmaster: are you there ? Feb 03 10:05:22 nm: yeah Feb 03 10:05:41 vmaster: i got openocd from trunk ? Feb 03 10:05:51 nm: the trunk doesn't contain XScale support Feb 03 10:06:13 nm: you'll need /branches/xscale Feb 03 10:06:15 vmaster: oh, so how do i get it work w/ XScale Feb 03 10:06:26 vmaster: okie, Feb 03 10:06:46 vmaster: so, what about arm9 - at91rm9200 ? is trunk worked w/ it Feb 03 10:06:55 yeah Feb 03 10:07:47 gotta get breakfast, i'll be back in about 20 minutes Feb 03 10:10:53 vmaster: okie :-) Feb 03 10:26:46 vmaster: i have problem when trying to halt arm9-at91rm9200 w/ openocd on windoze from yarto Feb 03 10:26:49 > halt Feb 03 10:26:52 requesting target halt... Feb 03 10:26:54 > resume Feb 03 10:26:56 target not halted Feb 03 10:28:05 which cfg file did you use? Feb 03 10:29:02 vmaster: i wired nTRST to nRST Feb 03 10:30:27 vmaster: it still worked, i created my arm9_wig_config from arm9_ft2232.cfg Feb 03 10:37:36 nm: connecting nTRST to nSRST isn't a problem on an ARM7/9, but you can't debug the target out of reset with that setup Feb 03 10:39:12 nm: you have to tell the OpenOCD about your broken reset setup: reset_config trst_and_srst srst_pulls_trst Feb 03 10:50:26 vmaster: i've just fix my jtag interface, http://wiki.openwrt.org/OpenWrtDocs/Customizing/Hardware/JTAG_Cable Buffered Cable, Wiggler Feb 03 10:50:45 vmaster: here is my log http://pastebin.com/874287 Feb 03 10:51:09 vmaster: and my config file http://pastebin.com/874288 Feb 03 10:51:43 vmaster: i dont'know what wrong when i try to halt on openocd ... requesting target halt... Feb 03 10:51:55 vmaster: can you give me some advice ? Feb 03 10:55:56 the config file you've posted is for an ARM7 Feb 03 10:56:09 21. target arm7tdmi little run_and_halt 0 arm7tdmi-s_r4 Feb 03 10:56:49 vmaster: oh sorry, i have mistake when posting Feb 03 10:59:25 vmaster: myconfig http://vnoss.net/p/311 Feb 03 11:00:52 nm: remove the comment in front of the 'reset_config' line Feb 03 11:01:13 nm: change the target type to "arm920t" instead of "arm9tdmi" - you can remove the last argument (arm920t), as this isn't used anymore Feb 03 11:01:41 nm: change "reset_halt" in the 'target' line to "run_and_halt" Feb 03 11:02:52 vmaster: i fixed my cable to control nTRST and nRST separately Feb 03 11:05:56 ok, then use "reset_config trst_and_srst", and "run_and_halt" in the 'target' line Feb 03 11:10:36 vmaster: http://vnoss.net/p/313 Feb 03 11:10:48 vmaster: i still can't halt the target Feb 03 11:10:59 > halt Feb 03 11:10:59 requesting target halt... Feb 03 11:10:59 > Feb 03 11:10:59 > resume Feb 03 11:10:59 target not halted Feb 03 11:17:57 ok, now give me a log from that run Feb 03 11:28:24 vmaster: http://vnoss.net/p/314 Feb 03 11:29:36 which version of the OpenOCD was that? Feb 03 11:32:24 vmaster: i'm currently run openocd on windoze (from yaroto distribution) Feb 03 11:32:27 vmaster: C:\Program Files\openocd-2006re115\bin>openocd-pp.exe -d 3 -f configs/arm9_wi Feb 03 11:32:30 Info: openocd.c:82 main(): Open On-Chip Debugger (2006-10-12 18:00 CEST) Feb 03 11:33:11 vmaster: do i need switching to openocd-trunk on linux ? Feb 03 11:33:13 get the latest one (R128 iirc), and try again Feb 03 11:33:26 r128 is available from yagarto Feb 03 11:41:08 vmaster: i'm downloading Feb 03 11:41:37 vmaster: i have some funny, but i don't know why w/ openocd and my jtag interface Feb 03 11:41:51 vmaster: when i wired nTRST and nRST, Feb 03 11:42:00 vmaster: using the config file i've posted, Feb 03 11:42:17 vmaster: then telnet to openocd, try halt , nothing happend, Feb 03 11:42:40 but if i press resset button on board, the target will halt immediately Feb 03 11:42:48 > halt Feb 03 11:42:48 requesting target halt... Feb 03 11:42:48 > Feb 03 11:42:48 > Feb 03 11:42:48 > Feb 03 11:42:50 > Feb 03 11:42:53 > Target 0 halted Feb 03 11:42:55 target halted in Thumb state due to debug request, current mode: System Feb 03 11:42:58 cpsr: 0xffffffff pc: 0x77bc4d3e Feb 03 11:43:00 MMU: disabled, D-Cache: disabled, I-Cache: disabled Feb 03 11:43:03 > Feb 03 11:43:30 you must never press the reset when button while the debugger is running Feb 03 11:43:49 there's no way for the OpenOCD to detect that, and the target and OpenOCD will get out of synch Feb 03 11:44:08 chances are that the target is still running, but the bogus information looked like it entered halt Feb 03 11:44:32 ah, and reduce the "run_and_halt_time" from 5000 to 500 Feb 03 11:44:49 and add these two lines: Feb 03 11:44:50 jtag_nsrst_delay 1000 Feb 03 11:44:50 jtag_ntrst_delay 1000 Feb 03 11:45:06 vmaster: to my current config ? Feb 03 11:45:21 yeah Feb 03 11:50:20 very slow when download, i'll switch to trunk version Feb 03 11:58:58 vmaster: here is my log run w/ the same config http://vnoss.net/p/315 Feb 03 12:04:12 you have a high level on JTAGSEL - this enables the AT91's boundary scan TAP controller Feb 03 12:04:31 for debug you have to pull JTAGSEL low, to enable the ARM920t's TAP controller Feb 03 12:11:30 vmaster: oh it works, Feb 03 12:11:33 my jumper so suck Feb 03 12:12:11 vmaster: why you know my JTAGSEL is high when reading log ? Feb 03 12:12:23 vmaster: Error: EmbeddedICE version get failed ? Feb 03 12:13:40 line 14: Debug: jtag.c:1214 jtag_examine_chain(): JTAG device found: 0x15b0203f (Manufacturer: 0x01f, Part: 0x5b02, Version: 0x1 Feb 03 12:14:25 vmaster: ah, 05b0203f when jtagsel low level Feb 03 12:14:30 right Feb 03 12:14:30 vmaster: okie, i see :-D Feb 03 12:25:29 vmaster: thx guy ;) Feb 03 12:25:33 * nm have dinner **** ENDING LOGGING AT Sun Feb 04 02:59:58 2007