#318 adapter driver presto - not work
I'm trying a first time to use Asix PRESTO with OpenOCD. (this Presto is normally working with the Asix UP software for Windows)
I tried to use in Windows precompiled OpenOCD 0.11 and in Linux fresh compiled by me.
In the configuration I use "adapter driver presto", or in addition "presto_serial 013979". However, on both systems, OpenOCD cannot correctly initialize PRESTO.
==== Linux dmesg ====
[395089.615983] usb 2-2.2: new full-speed USB device number 29 using uhci_hcd
[395089.820689] usb 2-2.2: New USB device found, idVendor=0403, idProduct=f1a0, bcdDevice= 4.00
[395089.820693] usb 2-2.2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[395089.820694] usb 2-2.2: Product: ASIX PRESTO
[395089.820696] usb 2-2.2: Manufacturer: ASIX
[395089.820697] usb 2-2.2: SerialNumber: A6013979
==== OpenOCD --debug ====
Debug: 56 5 presto.c:139 presto_open_libftdi(): searching for PRESTO using libftdi
Error: 57 7 presto.c:149 presto_open_libftdi(): unable to open PRESTO: usb_open() failed
Error: 58 8 presto.c:540 presto_jtag_init(): Cannot open PRESTO
Debug: 59 8 command.c:629 run_command(): Command 'init' failed with error code -100
Hello Michal,
On Wed, Sep 08, 2021 at 07:25:38AM -0000, Michal Hucik wrote:
That's apparently a rather unusual adapter.
Can you possibly join #openwrt on Libera IRC network for an
interactive debugging session today?
Have you made sure it's not a permissions problem?
--
Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software!
mailto:fercerpav@gmail.com
Hello Paul, you're right! It seems that the whole problem is only in udev permissions.
PRESTO is now connected (as root) - unfortunately it doesn't seem to be very fast. Now I will try to connect JTAG to the neorv32 RiscV for gdb session.
Problem is solved - I apologize for the commotion.
Michal
==== OpenOCD ====
Info : PRESTO open, serial number ''
Info : setting speed to 0, max. TCK freq. is 3 MHz
Info : clock speed 3000 kHz
I'm looking for info on this old/odd AXIS Presto programmer. Unfortunately no schematics in internet.
From what I have found, it looks based on FTDI FT232BM that is a USB1.1 UART adapter.
The JTAG seems emulated through bit-banging on the UART control signals RTS/CTS/...
Shouldn't be very fast!
The permission issue is caused by the adapter not being listed in contrib/60-openocd.rules
I will send out a patch for this.
Since you own this odd device, may I ask your help in the future for testing some rework of this driver?
I have connected OpenOCD with PRESTO, however communication with the CPU is failed. If someone wants to take a look, then in the attachment I send communication log on JTAG created by logic analyzer. I have communication log with neorv32 from PRESTO and by J-Link (J-Link communication is OK).
Log is created by logic analyzer Omega - a next product from Asix :)
Here is the software for viewing attached files https://asix.tech/dbg_sigma-omega_en.html
It's no problem to contact me if you think of anything else to try or measure on the PRESTO.
Hi,
On Tue, Sep 14, 2021 at 12:19:22PM -0000, Michal Hucik wrote:
Were both performed using OpenOCD? Can you also attach -d3 output from
OpenOCD in both cases please?
Here is dbg output from OpenOCD.
On Tue, Sep 14, 2021 at 12:32:32PM -0000, Michal Hucik wrote:
Error: 81 2255 core.c:1132 jtag_examine_chain_check(): JTAG scan chain interrogation failed: all ones
Error: 82 2255 core.c:1133 jtag_examine_chain_check(): Check JTAG interface, timings, target power, etc.
So it looks like no communication happens at all. At this stage
there's nothing fancy going on, it's almost like communicating to a
regular shift register. Probably you can attach screenshots from your
LA so that one wouldn't need to install custom software to take a
look?
I have no idea what that communication should look like.
J-link:
1) TMS = 1, 8x CLK = 1.2 MHz, TMS = 0
2) 28 ms gap
3) TMS = 1, 8 x CLK = 421 kHz, TMS = 0
4) 6 ms gap
5) CLK = 1.2 MHz - first reply on TDO (LA_jlink_has_reply.png)
PRESTO (CLK is a little crazy, but when is stabilised so is 545 kHz):
1) LA_presto_1.png
2) 1500 us gap
3) LA_presto_2.png
4) 4x (7 ms gap and cca 2 us area with some CLK pules)
5) 7 ms gap, LA_presto_3.png
6) 6 ms gap, LA_presto_4.png
7) 6 ms gap, LA_presto_5.png
8) 2 ms gap, LA_presto_6.png
9) ...and silence...
LA_presto2.png and LA_jlink_has_reply.png look the same in the first cycles
up to the first reaction from the target on the TDO. Except that the PRESTO
keeps nTRST low the entire time so that's probably why nothing happens.
On Tue, Sep 14, 2021 at 5:53 PM Michal Hucik chaky@users.sourceforge.net
wrote:
Related
Tickets: #318