Hello, Thanks for all of your hard work. Inspired by the recent progress of Michal in perhaps fixing memwrite burst for arm1136, I decided to try the latest revision (2727).
It does not work at all with my Signalyzer and i.MX35. I have triangulated back to the last revision that lets me load *AND RUN* code successfully. That is revision 2046. The output at openocd startup (r2727) is as follows, though the error messages vary with repeated invocations: ----------------- Open On-Chip Debugger 0.3.0-in-development (2009-09-18-03:02) svn:2727 $URL: svn://svn.berlios.de/openocd/trunk/src/openocd.c $ For bug reports, read http://svn.berlios.de/svnroot/repos/openocd/trunk/BUGS 100 kHz jtag_nsrst_delay: 200 jtag_ntrst_delay: 200 Info : clock speed 100 kHz Error: couldn't read the requested number of bytes from FT2232 device (76 < 81) Error: couldn't read from FT2232 Error: number of discovered devices in JTAG chain (0) does not match (enabled) configuration (4), total taps: 4 Error: check the config file and ensure proper JTAG communication (connections, speed, ...) Error: trying to validate configured JTAG chain anyway... Info : found ARM1136 -------------------- Here's my config: --------------------- telnet_port 4000 gdb_port 2000 source [find interface/signalyzer.cfg] source [find target/imx35.cfg] jtag_khz 100 reset_config srst_only srst_pulls_trst jtag_nsrst_delay 200 jtag_ntrst_delay 200 ----------------------- If I try to load any code (e.g. in gdb), I get immediate errors: ------------------- 'arm11 target' JTAG communication error SCREG SCAN OUT 0x1f (expected 0x10) examine-fails: -4 Runtime error, file "command.c", line 473: 'arm11 target' JTAG communication error SCREG SCAN OUT 0x1f (expected 0x10) -------------------- As mentioned above, revision 2046 works properly (but memwrite burst must be disabled). Revision 2047 (and many above that) still allow code loading, but running does nothing (in GDB, continuing stays on the first instruction always; single stepping has never worked for this platform). This could be a problem with RAM initialization caused by other errors. Looking at svn diff -c 2047, I see several changes in jtag/jtag.c that replace global variable 'cmd_queue_end_state' with function argument 'state'. Any suggestions? Thanks, Ethan Eade _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
