matsievskiysv has quit [Ping timeout: 268 seconds]
matsievskiysv has joined #openocd
<matsievskiysv>
Sorry, my IRC client keeps kicking me out
shoragan[m] has joined #openocd
ormaaj has joined #openocd
LinuxHackerman has joined #openocd
losfair[m] has joined #openocd
barath has joined #openocd
<PaulFertser>
matsievskiysv: ok, so a regular bluepill plus ft2232h breakout. Are you aware that when you connect CMOS output to a high-Z CMOS input you often need some line impedance matching? JTAG adapters usually use series termination (47--100 Ohm) on all data lines for that.
<PaulFertser>
It might or might not matter, depends mostly on the target.
<PaulFertser>
If you have an oscilloscope you can check signal edges to make sure there's no "ringing" on transitions.
<matsievskiysv>
No, I didn't include any impedance
<matsievskiysv>
but openocd finds TAPs just fine
<PaulFertser>
matsievskiysv: do you probably have a regular JTAG adapter for comparison?
<matsievskiysv>
I have ft2232 and stlinkv2 clone
<matsievskiysv>
I'll try to include impedance and lower jtag clock speed
<matsievskiysv>
I'll be back in an hour with that
<PaulFertser>
Can stlinkv2 access this same bluepill board with SWD (and without reset) with default upstream OpenOCD configs?
<matsievskiysv>
Yes, swd works fine
<PaulFertser>
matsievskiysv: can you "mass_erase" it with stlink prior to trying jtag just in case?
matsievskiysv has quit [Quit: Client closed]
kraiskil has quit [Ping timeout: 246 seconds]
zkrx has quit []
zkrx has joined #openocd
kraiskil has joined #openocd
matsievskiysv has joined #openocd
matsievskiysv_ has joined #openocd
matsievskiysv_ has quit [Client Quit]
merethan has quit [Remote host closed the connection]
merethan has joined #openocd
<matsievskiysv>
Adding serial termination and lowering clock to 10k did not help
<matsievskiysv>
I'll try substituting ft2232 with stlink2
<PaulFertser>
matsievskiysv: lowering the clock can't help with ringing, if it happens, it happens on each transition (edge).
<PaulFertser>
stlink2 in swd mode, there's no reason to try jtag with stlink.
<matsievskiysv>
I will look at signal edges tomorrow at work
<PaulFertser>
I suggest you try mass_erase over swd to make sure it's not firmware disabling pins.
<matsievskiysv>
I previously programmed it baremetal without HAL, so I'm pretty sure that it does not mess with the pin assignment.
<matsievskiysv>
But I'll do this anyway
<PaulFertser>
matsievskiysv: have you already tried using e.g. 0.9.0 release in JTAG mode with this target?
<matsievskiysv>
Just Debian stock which is 0.11
<PaulFertser>
It's old enough, a bug like that would have been noticed.
<matsievskiysv>
I don't really trust my `ftdi layout_init` settings
<matsievskiysv>
We need to set TDI as input and others as output, right?
<matsievskiysv>
And the reset values to HIGH
<matsievskiysv>
BTW, if I set nTRST and nSRST in `layout_init` to LOW, the board does not answer
<matsievskiysv>
Even though these signals are defined with `layout_signal`
<PaulFertser>
matsievskiysv: I suggest you do not connect nTRST and nSRST at all. And use upstream config for ftdi breakout board.
<PaulFertser>
interface/ftdi/minimodule.cfg
<matsievskiysv>
Well, it's a progress
<matsievskiysv>
Now I have Invalid ACK (7) instead of (4)