**** BEGIN LOGGING AT Mon Jul 08 02:59:59 2013 Jul 08 04:53:41 ah Jul 08 18:04:23 ping jj2baile Jul 08 21:00:23 bradfa: pong Jul 08 21:23:59 hiya jj2baile Jul 08 21:26:18 ka6sox: hi. Jul 08 21:26:33 Why is device tree the only way to set pinmux :( Jul 08 21:29:38 Oh perhaps the precondition of that statement is false. Jul 08 21:43:53 jj2baile, I have to run but just wanted to ping you on the blog and video posts, where are you at on those as we need to make sure we don't forget them Jul 08 21:44:15 jj2baile, I'll be back in the morning EDT and maybe we can talk then (no internet at home right now) Jul 08 21:44:22 but ttyl Jul 08 21:48:08 bradfa: Ok. Jul 08 21:49:14 Im starting to write another blogpost, however haven't done anything with regards to the video yet. For the video am I just to explain my project basically? Jul 08 21:53:56 because people are idiots Jul 08 22:24:22 hiya jj2baile sorry, work called again Jul 08 22:24:33 jj2baile, yes, we need to explain the project Jul 08 22:27:11 * vvu has no mood for video but needs to be done! Jul 08 23:06:22 mluckham, you around? Jul 08 23:06:32 ka6sox: I See. What mechanism of the kernel prevents this from happening? Jul 08 23:07:00 jj2baile, the kernel owns those registers Jul 08 23:07:17 and with DT the device tree manages them for you Jul 08 23:07:29 Aha. So absolutely no way around. got it. Jul 08 23:07:36 you can call exclusive on pins you need Jul 08 23:07:44 via a DTC Jul 08 23:07:51 and you can compile it on the device as well Jul 08 23:08:00 (and then load it via the capemanager. Jul 08 23:08:01 ) Jul 08 23:08:35 the latest kernel code has the DT file for doing some testing using the PRU pins Jul 08 23:08:40 yeah. I'd been trying with an example dt fragment from some code online, but still didn't get it to do what i'd wanted Jul 08 23:09:06 ka6sox: what is the file called? Jul 08 23:09:14 let me look...brb Jul 08 23:11:24 look at BB-BONE-PRU* in /firmare/capes Jul 08 23:11:43 hipefully the dts files are there and not just compiled Jul 08 23:12:03 also you can use the dtc compiler to translate between compiled and source Jul 08 23:16:24 Ok Jul 08 23:24:17 bb10m..running the the mailbox Jul 08 23:38:40 i'm here now Jul 08 23:42:09 met oo Jul 08 23:42:17 (okay that was 20 minutes) Jul 08 23:42:27 no timestamp on mIRC Jul 08 23:42:35 (maybe there's an option) Jul 08 23:43:26 ya, probably Jul 08 23:43:55 today I looked at OpenOCD internals, and the OpenJTAG driver patch Jul 08 23:44:25 excellent Jul 08 23:44:29 thoughts? Jul 08 23:44:39 looking less like black magic Jul 08 23:44:56 someone else is using a Beagle GSoC 2010 project - http://blog.gimx.fr/?p=604 Jul 08 23:45:22 one of the OpenOCD moderators didn't like the OpenJTAG idea, but only because it uses an FTDI part and there may be a similar function Jul 08 23:45:45 who was that? Jul 08 23:45:52 CareBear\? Jul 08 23:46:10 (I saw that USB SNiffer project - looked interesting too) Jul 08 23:46:20 Nicolas Jul 08 23:46:30 Andreas Fritiofson Jul 08 23:47:12 ok Jul 08 23:47:23 it was on the patch 1406, which was accepted Jun 5 Jul 08 23:48:12 anyway it looks like a way to accelerate throughput, if needed Jul 08 23:48:51 so we have OpenOCD (new driver) -> vring -> PRU -> CPLD ?? Jul 08 23:49:44 I don't know enough to judge whether 'standard' JTAG signalling, or OpenJTAG-type signalling, should be used ... Jul 08 23:50:25 is vring bidirectional? does it need to be? (JTAG events?) Jul 08 23:50:55 mluckham, thats wht I'm thinking Jul 08 23:50:59 jj2baile, you here? Jul 08 23:53:56 Yes Jul 08 23:55:10 did either of those dts make sense? Jul 08 23:55:19 01 or 02? Jul 08 23:57:10 Yeah. Although building it and loading it didnt seem to do anything. Jul 09 00:03:39 jj2baile, what's supposed to happen ... do you have a PRU demo that should toggle pins? Jul 09 00:07:56 mluckham: yeah. im trying to set the pin to the correct mode. Jul 09 00:08:20 Something i just tried might have worked, or i might have to reflash the bone now... unsure which yet. Jul 09 00:08:49 sounds drastic Jul 09 00:09:01 reflash? Jul 09 00:10:53 the emmc, i tried editing the am335x-boneblack dtb. Seems like the bone isn't fully booting anymore. probably because of conflict on pin with the hdmi cape/firmware? Jul 09 00:23:16 the hdmi cape would just not load Jul 09 00:23:30 if it interfered with the eMMC then it wouldnt' boot Jul 09 00:23:43 which pin number are you working with? Jul 09 00:24:26 can you boot from SD image, edit the emmc? Is there a log? Jul 09 00:24:42 it was gpio1_12 Jul 09 00:25:33 I am looking for an microsd->sd adapter to load a build unto the sd. I should have one somewhere around. Jul 09 00:27:18 is emmc available as a flash drive (under Windows, Linux)? Jul 09 00:29:19 ah, no. Jul 09 00:29:28 but basically from the serial console, it complains/stops at: Jul 09 00:29:36 [ 0.478059] pinctrl-single 44e10800.pinmux: could not request pin 12 on device pinctrl-single Jul 09 00:29:39 [ 60.501575] bone-capemgr bone_capemgr.8: failed to load firmware 'cape-boneblack-hdmin-00A0.dtbo' Jul 09 00:29:46 perhaps i should give it more time Jul 09 00:34:05 does serial console provide a way to control the boot process? like a 'safe' mode, that doesn't use the HDMI cape? Jul 09 00:34:14 sorry for asking basic questions, my BBB isn't here yet Jul 09 00:35:15 well you can access uboot, but I couldn't get it to load the load the nand using the ext2ls or ext4ls commands. (i was trying to see if I could replace the original through u-boot) Jul 09 00:35:27 Im not sure which fs the beagle was using by default Jul 09 00:35:37 or can you SSH to the Ethernet port? Jul 09 00:36:50 Well, I don't think that would do very much additional good, considering linux hasn't even loaded. Although connecting it to ethernet would offer the opportunity of netbooting, but SD card would probably be easier anyway Jul 09 00:37:01 jj2baile, hmmm... Jul 09 00:37:03 that should work Jul 09 00:37:12 the point is to be able to un-edit the dtb file, somehow Jul 09 00:37:16 Yeah Jul 09 00:39:28 what pin number is that on the P8/P9 scheme? Jul 09 00:39:35 ka6sox: P8.12 Jul 09 00:40:35 so 0x030 0x46? Jul 09 00:40:49 or 0x06? Jul 09 00:41:11 0x06 Jul 09 00:43:35 ?? https://groups.google.com/forum/?fromgroups#!category-topic/beagleboard/beaglebone-black/ai9hFRsLcwc Jul 09 00:44:29 be back in a bit, I have to go get an adapter. Jul 09 00:49:17 mluckham, what issues do the openOCD folks forsee? Jul 09 00:50:59 I have no info from them (didn't ask) but I did sent an email to the OpenJTAG patch guy (Ryan Corbin) Jul 09 00:51:39 there is already demo code for vring, by panto? Jul 09 00:52:01 a bytestream, I noticed Jul 09 00:54:24 I'm still looking at OpenOCD, to see what a 'normal' JTAG driver does Jul 09 00:54:32 that's for tomorrow Jul 09 00:54:47 for I/O, I mean Jul 09 00:56:32 OpenOCD says it is not intended for boundary scan, do you need that? there is goJTAG project ... Jul 09 00:59:49 another project that might be simpler to understand is https://github.com/gsmcmullin/blackmagic Jul 09 01:02:25 mluckham: OpenOCD does have a couple of very limited boundary scan commands, btw Jul 09 01:02:48 wmat, I was hoping to find those Jul 09 01:03:33 mluckham: just SVF and XSVF Jul 09 01:03:42 http://openocd.sourceforge.net/doc/html/Boundary-Scan-Commands.html Jul 09 01:03:54 tgh, i have no idea if they work Jul 09 01:04:03 s/tgh/tbh Jul 09 01:04:46 tx, I will look at them tomorrow Jul 09 01:06:34 they run scripts, I don't know enuf about JTAG yet for that. Blackmagic has a boundary scan function() jtag_scan() ... Jul 09 01:07:29 it looks generic, just for discovery Jul 09 01:12:22 ka6sox, should I ping the OpenOCD guys? Jul 09 01:19:29 mluckham, probably not yet, let me ask someone about this Jul 09 01:19:54 if I don't get an answer tonight I'll tell you in the morning Jul 09 01:24:29 ok, I'm done for today Jul 09 02:06:17 oh, right, 4gb sdcard necessary, not 2. ill get one tomorrow. **** ENDING LOGGING AT Tue Jul 09 03:00:00 2013