I've made some substantial changes to breakpoint handling in arm7/9. This should help clear up a lot of confusing and obscure bugs, and no known regressions are known at this point, but it needs more testing....
Testing greatly appreciated. Here are some areas to be tested: 1. feroceon needs to be retested. Feroceon only has a single hardware breakpoint, so check that software breakpoints work and that proper error messages are produced upon adding another hardware breakpoint. 2. full regression test is *always* appreciated. It's been a while 3. hardware breakpoints 4. software breakpoints 5. add two hardware breakpoints, then add a software breakpoint and verify that a helpful error message is produced 6. set a breakpoint using "bp" from telnet, connect from GDB and verify that "bp" no longer lists any breakpoints. 7. start GDB, add breakpoint, continue, terminate GDB process, enter telnet and run "bp" to verify that breakpoints are "stuck", reconnect w/gdb and verify that connecting to GDB will purge the dangling breakpoints 8. test gdb_breakpoint_override command: force software, force hardware, leave it to GDB to decide breakpoint type -- Øyvind Harboe http://www.zylin.com/zy1000.html ARM7 ARM9 XScale Cortex JTAG debugger and flash programmer _______________________________________________ Openocd-development mailing list [email protected] https://lists.berlios.de/mailman/listinfo/openocd-development
