| --- Log opened Mon Mar 24 00:00:34 2014 | ||
| blueCmd | tclStrToD.c also crashes | 00:04 |
|---|---|---|
| blueCmd | heh | 00:05 |
| olof | stekern: Just saw the pull request for an ISE backend. Guess what I've been doing the last couple of days :) | 09:41 |
| stekern | you need to keep more monologues about what you're doing | 09:52 |
| stekern | follow me, I'll show you how it's done ;) | 09:53 |
| olof | :) | 09:57 |
| olof | There are some nice features in your version. I'll probably have to merge our versions a bit | 09:57 |
| stekern | yeah, I don't think double effort is a waste of time in all cases | 10:00 |
| stekern | are we double efforting the atlys port too, or have you been working against another board port (or none at all) | 10:02 |
| stekern | ? | 10:02 |
| olof | No, I've been working on my lx9 microboard | 10:03 |
| olof | Decided to put a mor1kx on it | 10:04 |
| stekern | ah, ok | 10:05 |
| olof | I'm also considering using your wb_fifo approach for UART communication. It's a small FPGA and the uart16550 core is fucking huge | 10:05 |
| stekern | you can use the eco32-uart | 10:07 |
| olof | stekern: cool | 10:07 |
| stekern | it's small and there is even a Linux driver for it | 10:07 |
| olof | perfect | 10:08 |
| olof | Do you have a .core for it? | 10:08 |
| stekern | https://github.com/skristiansson/orpsoc-cores/tree/eco32/cores/eco32-uart | 10:08 |
| olof | With a wishbone IF? | 10:09 |
| stekern | (linux driver) out of tree of course: https://github.com/skristiansson/eco32-linux/blob/master/drivers/tty/serial/eco32_uart.c | 10:09 |
| stekern | ah, wishbone, right: https://github.com/skristiansson/orpsoc-cores/blob/eco32/systems/eco32-de0_nano/rtl/verilog/eco32-de0_nano.v#L491 | 10:10 |
| stekern | I started doing a bus converter, but since it was as simple as that, I let it be | 10:11 |
| olof | I can see why :) | 10:11 |
| olof | oh well. Time to get off the train | 10:16 |
| stekern | ise is not the sharpest tool in the box... it got two clock constraints mixed up since there happened to be a clock signal with the same name in two different modules | 10:22 |
| stekern | (completely unrelated clock signals, I might add) | 10:23 |
| LoneTech | ouch | 10:49 |
| analognoise | howdy all | 17:51 |
| blueCmd | stekern: wouldn't you specify the full (or partial) path of the signal? | 18:31 |
| stekern | blueCmd: I would and I do... ISE doesn't get any more clever from that | 19:57 |
| stekern | perhaps it got confused since both signals were connected to IBUFs | 19:58 |
| stekern | I just renamed one of the signals in the design | 19:59 |
| stekern | _franck__: do you have a command line to share to launch openocd with adv_debug_sys and a ft2232 jtag device? | 20:01 |
| _franck__ | does your interface have a config file in ./tcl/interface/ftdi ? | 20:03 |
| stekern | no, but I copied the orsoc-jtag.cfg from juliusb's old repo into ./tcl/interface/ | 20:03 |
| _franck__ | so just change -f ./tcl/interface/altera... to your new config file | 20:04 |
| _franck__ | and make sure you have ft2232 support enabled when you ./configure openocd | 20:05 |
| _franck__ | --enable-ft2232_libftdi | 20:05 |
| _franck__ | this driver should not be used anymore. It now uses ftdi driver and ft2232 interfaces might be converted to ftdi (it is just a matter of converting the configuration file) | 20:06 |
| stekern | ah, right... I had an old command line with the altera-dev, couldn't figure out how to substitute taht | 20:07 |
| stekern | but this is what should be used now, right? sudo ./src/openocd -s ./tcl -f ./tcl/interface/altera-usb-blaster.cfg -f ./tcl/board/or1k_generic.cfg | 20:08 |
| stekern | if you'd have a blaster, that is | 20:08 |
| _franck__ | if you want to use signal tap you can also, download your program with openocd, close openocd, run signal tap and reset your board with your reset button... | 20:08 |
| _franck__ | yes it is | 20:08 |
| stekern | I don't want to use signaltap with my spartan6 device ;) | 20:08 |
| _franck__ | ah ok :) | 20:08 |
| _franck__ | if you use the dbg_if, you have to change TAPE_TYPE in ./tcl/board/or1k_generic.cfg | 20:09 |
| _franck__ | I submitted a patch to openocd to allow this selection from the command line | 20:10 |
| stekern | how can I see which interfaces I have configured with? | 20:11 |
| _franck__ | I think you can do that from the telnet shell | 20:12 |
| stekern | and if I don't have ft2232, but the ftdi, how would that orsoc-jtag look like? | 20:12 |
| _franck__ | I think olof did convert the file | 20:13 |
| _franck__ | search on the opencores forum | 20:13 |
| _franck__ | http://openocd.sourceforge.net/doc/html/Debug-Adapter-Configuration.html | 20:14 |
| _franck__ | interface_list: List the debug adapter drivers that have been built into the running copy of OpenOCD. | 20:14 |
| _franck__ | brb | 20:16 |
| _franck__ | stekern: FYI, you can set TAP_TYPE to XILINX_BSCAN and use Xilinx virtual JTAG | 20:19 |
| stekern | yes, but there seems to be a lot of problems with that | 20:20 |
| stekern | maybe that's related to adv_debug_bridge though | 20:20 |
| -!- FreezingAlt is now known as FreezingCold | 20:20 | |
| stekern | the openocd driver is perhaps better | 20:20 |
| _franck__ | well, it's pretty new and the guy said it was working with adv_debug_if... | 20:21 |
| _franck__ | (new in openocd) | 20:21 |
| stekern | ok | 20:21 |
| stekern | I get this now: http://pastie.org/8965400 | 20:25 |
| stekern | why is it saying this? Warn : There are no enabled taps. AUTO PROBING MIGHT NOT WORK!! | 20:31 |
| _franck__ | ./tcl/boa ??? :) | 20:36 |
| _franck__ | boa like a snake ? | 20:37 |
| stekern | yes, it's like python | 20:37 |
| stekern | the new fancy scripting language, haven't you heard about it? | 20:37 |
| _franck__ | I know working with python can be anoying so why not use boa ? | 20:38 |
| stekern | the 'rd' somehow got lost when I copy pasted that | 20:38 |
| stekern | I have them in my command | 20:38 |
| _franck__ | -f is lost too ? | 20:39 |
| stekern | yeah | 20:39 |
| stekern | the strange thing is, if I use '-f ./tcl/interface/altera-usb-blaster.cfg' it says: | 20:40 |
| stekern | Info : mohor tap selected | 20:40 |
| _franck__ | what do you have in orsoc-jtag.cfg ? | 20:41 |
| stekern | oh, there's some hooks back into C functions in the old code... | 20:42 |
| stekern | ok, so I need to figure out the ftdi_layout_init and ftdi_layout_signal things | 20:44 |
| _franck__ | yes and I can't help you here | 20:45 |
| stekern | http://opencores.org/forum,OpenRISC,0,5385 | 20:47 |
| stekern | but there was never any config files posted | 20:48 |
| _franck__ | too bad that olof is not here, he could have save you some precious time | 20:55 |
| stekern | I still doesn't get that "no enabled taps" | 20:58 |
| stekern | how is the blaster cfg different in that regard? | 20:59 |
| stekern | isn't that picked by the or1k_generic.cfg? | 20:59 |
| _franck__ | newtap is created in or1k.cfg which is called by or1k_generic.cfg? | 21:00 |
| _franck__ | (remove this ? at the end) | 21:00 |
| stekern | yes | 21:01 |
| stekern | so how does it matter which interface I use? | 21:01 |
| _franck__ | AFAIK it doesn't | 21:01 |
| _franck__ | but it does somehow | 21:02 |
| _franck__ | what is you type "tap_select mohor" in the telnet console ? | 21:03 |
| stekern | invalid command name "tap_select" | 21:03 |
| _franck__ | rep -r "no enabled tap" * | 21:03 |
| _franck__ | ouaps | 21:04 |
| stekern | it's in src/jtag/core.c | 21:04 |
| stekern | ;) | 21:04 |
| _franck__ | thanks stekern_grep | 21:04 |
| stekern | stekern_grep - a friendly grep with almost human output. might add smileys to it's output. | 21:06 |
| _franck__ | :) | 21:06 |
| _franck__ | try to run openocd with -d 3 to make it more verbose | 21:07 |
| stekern | http://pastie.org/8965544 | 21:10 |
| _franck__ | this is the interesting part of a working openocd: http://pastie.org/private/5zodpdfz2rfv2sbej36w | 21:13 |
| _franck__ | don't you have something like this at the begining (you didn't paste the whole log) ? | 21:13 |
| stekern | oh, let's see | 21:14 |
| stekern | hmm, it never get there | 21:18 |
| _franck__ | this is the full log if it can help: http://pastie.org/private/xkbri8hxwwe3rnbg5jehia | 21:19 |
| stekern | what's the difference between adapter_khz and jtag_rclk? | 21:25 |
| _franck__ | don't know I only know adapter_khz | 21:25 |
| stekern | pastie.org/8965594 | 21:26 |
| stekern | that's my full log | 21:26 |
| _franck__ | for some reasons it doesn't read your *.cfg files.... | 21:29 |
| stekern | hmm, I put the -s ./tcl last and now it works ?? | 21:33 |
| _franck__ | I alwas put -s ./tcl at the end | 21:33 |
| _franck__ | :) | 21:33 |
| stekern | yeah, I saw that you had that, that's why I tried it | 21:33 |
| stekern | 'works' was maybe an overstatement, but I get to the 'Target ready...' | 21:36 |
| _franck__ | great | 21:36 |
| stekern | mdw 0x100 yields 'Burst read failed' though | 21:38 |
| stekern | reading and writing to gpio registers seems to work though | 21:54 |
| _franck__ | my ddr controller is definitely too slow to feed the vga_lcd core connected to my altera board lcd display... | 23:04 |
| _franck__ | 800x480x24bpp @60Hz | 23:05 |
| --- Log closed Tue Mar 25 00:00:35 2014 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!