| --- Log opened Fri Jul 05 00:00:13 2013 | ||
| -!- Netsplit *.net <-> *.split quits: jonibo | 01:38 | |
| -!- trevorman_ is now known as trevorman | 01:49 | |
| -!- Netsplit *.net <-> *.split quits: Logxen | 01:59 | |
| -!- Netsplit over, joins: Logxen | 02:01 | |
| -!- trevorman is now known as Guest45696 | 04:27 | |
| -!- stekern_ is now known as stekern | 04:35 | |
| stekern | is there really no-one before me that has tried to investigate the usb-blaster ii protocols... | 05:29 |
|---|---|---|
| stekern | google doesn't pop anything on the subject of reverse engineering it | 05:29 |
| stekern | the jtag part looks like it acts as an emulated ftdi device | 07:53 |
| stekern | but the endpoints aren't right, so libftdi doesn't work with it | 07:54 |
| stekern | "aren't right" = differs from a real ftdi device | 07:54 |
| stekern | ah, but you can access those in the ftdi context struct, let's see what happens if you set them right | 08:11 |
| stekern | hmm, no... that's not it, since it fails when it tries to set the baudrate through the control end point | 08:17 |
| stekern | maybe that's not implemented in the emulation (if I'm right about the emulation part) | 08:17 |
| -!- Netsplit *.net <-> *.split quits: Amadiro | 08:28 | |
| -!- Netsplit *.net <-> *.split quits: andresjk | 12:47 | |
| -!- Netsplit *.net <-> *.split quits: poke53282, olofk_ | 12:52 | |
| -!- Netsplit *.net <-> *.split quits: jeremybennett | 12:53 | |
| -!- Netsplit *.net <-> *.split quits: Amadiro | 13:08 | |
| -!- Netsplit over, joins: Amadiro | 13:08 | |
| -!- Netsplit *.net <-> *.split quits: heroux, stekern, mboehnert, Logxen | 13:14 | |
| -!- Netsplit over, joins: Logxen, stekern, heroux, mboehnert | 13:15 | |
| -!- Netsplit *.net <-> *.split quits: Guest45696, LoneTech, enghong, blueCmd, forkG_, hno, O01eg, _franck_, jonmasters, hle_ | 13:19 | |
| -!- Netsplit over, joins: O01eg, hle_ | 13:19 | |
| -!- Netsplit over, joins: Guest45696, LoneTech, hno | 13:19 | |
| hno | stekern, do you have usb data traces to stare at? | 14:16 |
| stekern | hno: yes | 15:31 |
| stekern | but only comparison between a usb blaster and the usb blaster 2 when configuring the fpga so far | 15:34 |
| stekern | olofk_: did you manage to run the bsp-editor and build u-boot_spl during the workshop? | 17:28 |
| stekern | it generates crap in sdram_config.h for me | 17:29 |
| poke53282 | stekern: Your native port of gcc works. Could compile a hello world. | 19:22 |
| poke53282 | Nice work | 19:22 |
| poke53282 | For some reason I have to add "-I/usr/include" to my command line to include stdio.h | 19:23 |
| poke53282 | I thought it is hardcoded. | 19:24 |
| stekern | hmm, you shouldn't need that | 19:50 |
| stekern | how did you configure it? | 19:52 |
| stekern | http://pastie.org/8113694 | 19:53 |
| stekern | that's from some note I've made | 19:53 |
| poke53282 | http://pastie.org/8113768 | 20:25 |
| poke53282 | I think I have found it. I shouldn't change the sysroot | 20:26 |
| poke53282 | "--with-sysroot=$SYSROOT" is wrong | 20:26 |
| poke53282 | --enable-threads=single is a good point | 20:33 |
| stekern | ah, well, it was from a first test | 20:34 |
| poke53282 | quadmath is included in my build :) Not that I want to do a lot of numerics with this emulator | 20:35 |
| stekern | but, yes, you shouldn't have --with-sysroot | 20:35 |
| stekern | I think the rest of your options are fine | 20:36 |
| poke53282 | I am not sure about --without-headers | 20:36 |
| poke53282 | If it changes anything or not | 20:36 |
| poke53282 | The MAKEINFO line which is added to the Makefile is necessary on my system. | 20:37 |
| stekern | I think most of the things in my line is from building a first stage gcc | 20:37 |
| stekern | so yours is better | 20:37 |
| stekern | I've built with similar to your options, but that wasn't in my notes | 20:38 |
| stekern | hmm, what does it (MAKEINFO) do? I can't recall seeing that | 20:38 |
| poke53282 | The new texinfo is no longer compatible with the gcc help files | 20:38 |
| poke53282 | But there is no option not to compile them. | 20:39 |
| stekern | ok, maybe I have an old texinfo then | 20:39 |
| poke53282 | I use archlinux which has always very new packages. | 20:39 |
| poke53282 | So in a few month or so you will have the same problem. | 20:39 |
| stekern | looking forward to it ;) | 20:40 |
| poke53282 | This line with MAKEINFO prevents compiling the info files. | 20:40 |
| stekern | ah, I see | 20:40 |
| poke53282 | Update the repository would help. | 20:40 |
| stekern | from gcc upstream you mean? | 20:41 |
| stekern | yeah, we should do that | 20:41 |
| poke53282 | Yes, in github.com/openrisc | 20:41 |
| poke53282 | I have found the bug descriptions and the patches somewhere online. Can't remeber and can't find it right now. But this was the easiest solution if you don't want to apply the patch. | 20:42 |
| --- Log closed Sat Jul 06 00:00:14 2013 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!