tlwoerner has quit [Remote host closed the connection]
tlwoerner has joined #openocd
crabbedhaloablut has quit [Remote host closed the connection]
crabbedhaloablut has joined #openocd
nerozero has joined #openocd
PsySc0rpi0n has quit [Ping timeout: 248 seconds]
PsySc0rpi0n has joined #openocd
Haohmaru has joined #openocd
PsySc0rpi0n has quit [Ping timeout: 246 seconds]
PsySc0rpi0n has joined #openocd
zjason has joined #openocd
merethan_ has joined #openocd
<Mis012[m]>
how do I either make gdb ignore hw interrupts, or at least make openocd actually halt the cpu on hw interrupt
<Mis012[m]>
when a hw interrupt occurs, gdb prints Program stopped., but openocd says the program is in fact not stopped
<Mis012[m]>
normally I can just continue and it will only warn that it's trying to resume while the CPU isn't halted, but if I have a watchpoint set, gdb tries to re-set it and I get Error: can't add read watchpoint at 0x0032709d, target running
<PaulFertser>
Mis012[m]: gdb prints program stopped only if OpenOCD signals so.
<PaulFertser>
Mis012[m]: if you do "set debug remote 1" you'll see the sequence of events.
<PaulFertser>
Mis012[m]: if the CPU isn't getting halted, probably there's some misconfiguration with CTI and one core halts while another does not?
<Mis012[m]>
I only have one core described in openocd for now
<Mis012[m]>
but I sincerely doubt the bootrom uses more cores
<Mis012[m]>
might as well set up all of them I guess
<PaulFertser>
Mis012[m]: sometimes watchdog just reboots the whole target when it's sitting in halted state.
<Mis012[m]>
well, the thing is, the interrupt doesn't seem to result in the cpu getting halted
<PaulFertser>
(some kind of watchdog would be a more likely explanation if you see halted core suddenly running)
<Mis012[m]>
just gdb thinking so
<PaulFertser>
Mis012[m]: it's not just gdb.
<PaulFertser>
Run that command in gdb and you'll see that it's OpenOCD telling it.
<PaulFertser>
Also run openocd -d3 and you'll see why it does.