**** BEGIN LOGGING AT Fri Mar 01 02:59:58 2013 Mar 01 03:44:07 trelane, no Mar 01 03:56:46 what did i miss while spinning the wheels of commerce. Mar 01 04:27:24 i'm getting a feeling this pru interrupt documentation isn't right. Mar 01 04:28:26 the CMRn registers don't seem to do anything. they're supposed to map an event to a channel, but you can set them to anything Mar 01 04:28:42 the HMRn registers seem to work Mar 01 04:29:00 but, it looks like the channel mapping is fixed Mar 01 04:30:04 and the descriptions contradict what some of the text says Mar 01 04:32:52 i think what's most confusing is that the uio_pruss driver makes it look like there are 8 events you can access...but from what i can tell, there's only two physical interrupts to the arm. you can't access the others. Mar 01 04:33:07 (through the pruss_uio uio) Mar 01 04:38:33 prpplague: nice nick Mar 01 04:38:39 er PRU_EVTOUT_2 Mar 01 04:38:44 damn tab complete Mar 01 04:38:59 Ahh, I see what happened. they use the term "system interrupt" for the arm host and the pru system side. AWESOME. Mar 01 04:39:31 I saw that in the description of the register, but assumed it was a typo. Mar 01 04:39:57 not a typo at all, just absolutely ambiguous Mar 01 04:42:12 definitely going to write a tutorial after this. Mar 01 05:31:59 i give up. i can't figure out how to trigger interrupt 22 from the pru. Mar 01 05:33:10 heh Mar 01 05:33:22 the documentation is contradictory, the pru interrupt numbers seem to be different than the system side (with no reference table), and i'm dumb. Mar 01 05:34:17 or, could be nothing works because the kernel i have doesn't define those interrupts or something. Mar 01 05:34:28 need a evil TRM? :) Mar 01 05:36:04 is it bound in human flesh, washed with the tears of crying children, and written with an inconsistent ballpoint pen? Mar 01 05:36:25 has secret sections Mar 01 05:37:06 that would be awesome. I'm reading spruhf8. Mar 01 05:37:11 lesson never delete a TI datasheet ever :) Mar 01 05:37:36 emeb_mac: collect 'em all! :) Mar 01 05:38:17 mranostay: be the first in your neighborhood! Mar 01 05:39:41 good thing i don't/can't work for TI i'd be fired at this point :P Mar 01 05:44:11 so all it takes is them hiring you as a janitor to hush you? :D Mar 01 05:44:36 unless that janitor position pays 100k it is unlikely :) Mar 01 05:46:06 i clearly can be bought just the terms change :P Mar 01 05:48:21 stick you in the analog groups ;) Mar 01 05:50:06 analog janitors? Mar 01 05:50:15 sure Mar 01 05:50:19 got to handle the curves Mar 01 05:52:49 dirt isn't discrete! Mar 01 05:53:15 heh Mar 01 06:03:17 Evil data sheet wins! The interrupt table between the two is totally different. Looks like the pru uses the evil table mapping. Mar 01 06:03:39 ugh, I'm not insane. Mar 01 06:03:39 * mranostay whistles Mar 01 06:03:51 wonder who emailed that Mar 01 06:03:54 * mranostay whistles Mar 01 06:04:53 that info is suppose to be available as an addendum Mar 01 06:04:56 look for it Mar 01 06:05:04 it was weird. i had notepad open, and i dropped my keyboard. as it hit the ground, it typed out a bunch of stuff in notepad. i saved it as a pdf, and bam! it was the evil datasheet! Mar 01 06:05:21 ok, time to go home and have some wine. Mar 01 06:05:33 thanks mranostay! Mar 01 06:08:50 i have no clue what PRU_EVTOUT_2 is talking about Mar 01 06:08:52 :) Mar 01 06:28:16 heh - just watched the xiph.org videos. monty did a good job with the explanations. Mar 01 06:30:33 * mranostay drags _av500_ in Mar 01 06:30:56 as if Mar 01 06:41:17 dont' worry folks, the PRU has even toasted some of the trolls today. Mar 01 06:43:55 what is the difference between using cdev.h or miscdevices.h in char drivers Mar 01 06:44:50 clear me in what kind of char driver we have to proceed withe cdev.h and miscdevices.h ? Mar 01 06:45:40 just been hunting for trolls down the river near moordoor... with Jack... Mar 01 06:45:42 https://fbcdn-sphotos-g-a.akamaihd.net/hphotos-ak-prn1/549320_10200761938877811_1868997087_n.jpg Mar 01 06:48:19 clear me the difference between char drivers and misc driver( miscdevices.h) > Mar 01 07:17:47 * mranostay clears chainz Mar 01 07:59:21 moin Mar 01 07:59:38 mru: work is progressing...slowly Mar 01 07:59:50 mru: but i could give you one of my photon mortars Mar 01 08:23:15 all quiet Mar 01 08:23:22 thats what happens just before the trolls come out Mar 01 08:35:23 http://asset.soup.io/asset/1664/5653_5c93.jpeg - someone had too much time to fiddle... Mar 01 08:44:07 OMFG Mar 01 08:50:18 I suspect it was the xbox hack or something like that Mar 01 08:50:55 i wonder whether it still works... prolly not Mar 01 09:40:49 why can't people just solder stuff like normal people Mar 01 09:41:09 I'm thinking there'd be issues with inductance Mar 01 09:46:13 like "normal people" ? with giant drops of lead on the PCB and so on ? Mar 01 09:46:27 like reball and reflow Mar 01 09:49:41 best I can tell, this is the source Mar 01 09:49:43 http://www.reballing.es/viewtopic.php?p=4176 Mar 01 10:06:45 Russ: some people are too cheap Mar 01 10:06:54 Russ: i wouldnt even reball, just get a new chip and solder it on Mar 01 11:19:33 * mdp senses that mranostay has been trafficing illicit PRU-enabled TRMs to users all night Mar 01 11:21:29 koenkooi: if I'm using oe-core + meta-beagleboard the u-boot in oe-core is too old, where should it come from and can it be ported to update oe-cores u-boot? Mar 01 11:48:56 mdp: he bitbanged them Mar 01 12:36:34 troll troll troll Mar 01 12:44:05 * ogra_ hands mru a bridge to hide under Mar 01 12:47:54 ogra_: what if he is a moutain troll? Mar 01 12:48:13 then he gets a mountain bridge indeed :) Mar 01 12:57:40 .o0(friday afternoon) Mar 01 12:57:45 .o0(the end is neigh) Mar 01 13:03:03 friday afternoon and no trolling! wtf?! Mar 01 13:03:19 AND CURSE THE DAMN CCS INSTALLER Mar 01 13:03:30 * KotH is busy trying to figure out how to beat this *censored* IAR compiler to get his stuff compiling Mar 01 13:03:50 hehe Mar 01 13:04:06 it's not _that_ hard Mar 01 13:04:12 I managed to build libav with it Mar 01 13:04:44 yeah.. but register definitions are missing, some are named in interesting ways etc pp Mar 01 13:04:53 what target? Mar 01 13:04:54 and the editor sucks Mar 01 13:05:00 msp430frXXX Mar 01 13:05:02 why would you use their editor? Mar 01 13:05:08 use the command line compiler Mar 01 13:05:14 it even runs under wine Mar 01 13:06:01 I'm assuming they didn't suddenly grow a brain and release a linux version Mar 01 13:06:41 i'm still trying to get a first, just pin toggling application working Mar 01 13:07:17 and doesn't TI have a compiler for that? Mar 01 13:07:33 * mru would program it in assembly of course Mar 01 13:08:59 ti used to support iar to build a compiler for them... it seems they patched gcc until it supports msp430 and dropped iar like a hot potato Mar 01 13:09:09 on the bright side of things, someone or something at the tax office has gone mad Mar 01 13:09:15 they keep sending me money Mar 01 13:09:32 the thing is, people have used the iar stuff here and there is some knowhow.. i leverage on that until i get my stuff working, then i can switch to anything that is suitable for me Mar 01 13:09:33 the balance is still in their favour, of course Mar 01 13:10:51 my fear is they'll suddenly change their mind and demand it back Mar 01 13:11:47 the money you owe them? Mar 01 13:12:05 right now they say I don't owe them anything Mar 01 13:12:20 as I said, they just deposited a large amount in my bank account Mar 01 13:12:41 its a trap Mar 01 13:12:47 yes, most likely Mar 01 13:13:00 send it off to your swiss bank account Mar 01 13:13:17 make them believe you have not gotten anything and that it's a computer mistake on their side Mar 01 13:13:18 well, they know they sent it to me Mar 01 13:13:30 and my bank has records of it Mar 01 13:13:37 which they can probably get a copy of Mar 01 13:13:37 I once declared and paid too much taxes, corrected it after 3ys and got 6% on the amount, not bad.... Mar 01 13:14:06 _I_ did not submit any corrections Mar 01 13:23:48 maybe you will get corrected Mar 01 13:23:57 and they are just preparing Mar 01 14:02:45 * KotH whistles Mar 01 14:13:37 hey, let's troll alan_o Mar 01 14:14:51 haudi alan_o! Mar 01 14:15:15 hey alan_o Mar 01 14:16:28 looks like he is still asleep Mar 01 14:18:52 helo trolls Mar 01 14:22:37 helo is old, it's ehlo these days Mar 01 14:23:38 there are old trolls Mar 01 14:23:46 ehlo trolls Mar 01 14:23:55 it's alive! Mar 01 14:24:09 so what'd I miss? Mar 01 14:24:28 depends on when you left Mar 01 14:24:33 hehe Mar 01 14:24:58 After ELC, I went up to Seattle, then to Orcas Island. Got back last night Mar 01 14:25:29 Did some day hikes with the wife Mar 01 14:25:41 gm all Mar 01 14:25:47 btw, what's the significance of a psu I got from TI being marked 'shannon'? Mar 01 14:26:03 hola jkridner. Mar 01 14:26:16 Looking around the office this morning, it's obviously I left in a mad rush. Mar 01 14:26:17 is it saying something about the information bandwith available, or is some poor sod desperately searching for his missing psu? Mar 01 14:26:19 * dm8tbr only knows shannon VOLMET Mar 01 14:26:44 alan_o: mad rush is the only way to leave Mar 01 14:27:14 like the time I was going to CA and was woken by the cab outside the door, and I hadn't yet packed Mar 01 14:29:53 koenkooi: I got an LCD 4, but when I boot it off of the 3.2.28 kernel, while it says it is initialized the screen is blank. Mar 01 14:31:13 alan_o: missed nothing...the beaglebone is still zipped up. Mar 01 14:34:13 mru: woken? That's madder of a rush than I've ever left in. Mar 01 14:34:38 I'm a bit weird like that Mar 01 14:35:01 On the van ride back to the airport after ELC 2010, we had to circle back to this guy's apartment so he could get his phone. There were these two drunk British girls that harassed him non-stop. So at least you weren't woken by the shared ride van :) Mar 01 14:35:26 I try to avoid those Mar 01 14:35:46 shared vans and drunk brits Mar 01 14:37:50 drunk brits in shared van Mar 01 14:39:01 I really like the new dell 'linux developer' laptop Mar 01 14:39:33 small, sleek, fast, decent resolution and cheaper than the Windows variant Mar 01 14:40:33 shame I already bought a new laptop not so long ago :( Mar 01 14:40:57 * av500 thought it was for "ubuntu developers" Mar 01 14:41:27 yeah it's Ubuntu powered, but in this case I'll let it slide as they seem to have done a good jon Mar 01 14:41:29 I'd rather have one with nothing preinstalled Mar 01 14:41:34 s/jon/job Mar 01 14:42:11 I don't think you're going to get Dell doing clean laptops; but you can buy clean laptops elsewhere Mar 01 14:42:15 you could just put lennartOS^Wfedora on it Mar 01 14:42:26 systemDOS Mar 01 14:42:48 System Denial of Service? Mar 01 14:42:58 that's an accurate description Mar 01 14:43:15 jackmitchell: what the laptop comes with is far down my list of priorities Mar 01 14:43:31 as long as it's easy to install whatever I want Mar 01 14:44:32 mru, +1 Mar 01 14:44:44 mru: yes, but I'm pertaining to the fact that you don't pay for the Windows licence and it actually shows in the price; the fact that you know at least _some_ form of driver support is already their for all the hardware is added bonus Mar 01 14:45:08 my laptops tend to have mostly intel hardware Mar 01 14:45:24 it's the sensibility of it all that has inspired me Mar 01 14:45:24 never had a problem with that Mar 01 14:45:25 jackmitchell: their what? Mar 01 14:45:52 and the cheapest windows licence is usually minor compared to the total cost anyway Mar 01 14:46:14 a few % Mar 01 14:46:16 well this is £1,019 vs £899 Mar 01 14:46:34 that's a very expensive windows licence Mar 01 14:46:51 go for the 'home basic' or whatever it's called Mar 01 14:46:58 should be about £20 Mar 01 14:47:19 not that they'll tell you Mar 01 14:47:45 Sure, if you never want to do anything ever on your computer. Mar 01 14:47:45 but if you choose one of the 'pro' etc versions, the price goes up quite a bit Mar 01 14:47:47 I don't know the ins and outs of it, but regardless a cheaper laptop with Linux is a step forward Mar 01 14:47:58 agmlego: I want to wipe out the windows installation Mar 01 14:47:58 Like change your desktop background. Mar 01 14:48:02 so it doesn't matter what it was Mar 01 14:48:22 the last laptop I bought, I never even booted it Mar 01 14:48:29 jackmitchell: I'll add a u-boot recipe next week when I get back Mar 01 14:48:58 well if it was osx you could guarantee it would be 20% more expensive, so it does't kind of matter what it is Mar 01 14:49:02 jackmitchell: are you in the States? Mar 01 14:49:04 jkridner: you know how to debug stuff, right? Mar 01 14:49:13 s/does't/does Mar 01 14:49:17 jkridner: find the bug, send a patch to fix it, done Mar 01 14:49:19 ka6sox-farfarawa: no, UK Mar 01 14:49:26 okay Mar 01 14:49:28 I'm simply keeping you informed. Mar 01 14:49:48 There seems to be some fixes on the beaglebone-3.2 branch, so I'm trying those out. Mar 01 14:49:52 koenkooi: cheers koen, I've posted something about the u-boot recipes to oe-core if you want to stick your oar in, I don't know how far it will go though Mar 01 14:49:54 jackmitchell: yes, cheaper is better Mar 01 14:50:35 and it's good that someone is selling windowless laptops Mar 01 14:50:37 mru: Ah, in that case, Home Basic is perfect for your needs. Mar 01 14:50:40 mru: that's what I was really getting at; no monopoly -> cheaper hardware Mar 01 14:50:41 jackmitchell: I don't use the u-boot in oe-core, it's so useless it isn't even funny Mar 01 14:51:06 jackmitchell: and real users of u-boot can't get it fixed, so everyone uses their own copy Mar 01 14:51:08 but since the OS is such a small part of the total price, I'm still looking at hw specs first Mar 01 14:51:11 koenkooi: hah, maybe I'm barking up the wrong tree then Mar 01 14:51:11 mru: Sadly, I need to have a functional Windows installation on bare metal for work, so i spend money on a full Pro licence and dual-boot. Mar 01 14:51:12 zareason sells windowsless laptops Mar 01 14:51:22 once I've decided on the hw, I pick the cheapest sw option Mar 01 14:51:26 http://www.h-online.com/open/news/item/Dell-Sputnik-now-being-sold-in-Europe-1814767.html Mar 01 14:51:33 that's the comparison if anyone is interested Mar 01 14:51:58 russ: ping Mar 01 14:52:03 jkridner: anything more about audio that I need to know before I spend more hours on this? Mar 01 14:52:10 My hardware selection goes like this: "Sliders to the right on everything but RAM and SSD" Mar 01 14:52:39 jackmitchell: those are reasonable specs Mar 01 14:52:41 a bit heavy Mar 01 14:52:43 amlego always max ram out if you can afford it. Mar 01 14:52:47 ka6sox-farfarawa: It is a series of sinusoidal waveforms summed together causing pressure waves in a transmission medium. Mar 01 14:52:52 not that I know. Mar 01 14:52:56 ka6sox-farfarawa: Not from the vendor. That is just stupid. Mar 01 14:53:06 I can get RAM *way* cheaper from Newegg. Mar 01 14:53:11 agmlego: my algorithm is "hw right, sw left" Mar 01 14:53:20 agmlego: not if it's soldered Mar 01 14:53:25 jkridner: okay thanks. Mar 01 14:53:38 mru: Works for me, except the aforementioned RAM and SSD--those are rarely cheaper from the manufacturer than from other sources. Mar 01 14:53:48 mru: Fuck that shit, that is not even an option. Mar 01 14:54:00 it is if you value weight Mar 01 14:54:02 also some configurations get a different mainboard. I've seen Thinkpads that are prepared for 4 modules but the board only had 2 slots Mar 01 14:54:08 and slimness Mar 01 14:54:14 which I do Mar 01 14:54:19 mru: My netbook weighs basically nothing and has slots. Mar 01 14:54:25 I do not buy that argument at all. Mar 01 14:55:45 how thick is it? Mar 01 14:55:52 my laptop is ~16mm with lid closed Mar 01 14:56:34 I think thickness is the main contributor to soldered ram Mar 01 14:56:42 with lock-in coming a close second Mar 01 14:56:49 mru:You measure weight in mm? You are weird, sir. Mar 01 14:57:05 as in mm of Beer? Mar 01 14:57:06 jackmitchell: thickness and price Mar 01 14:57:13 sockets cost money Mar 01 14:57:32 do they solder it directly to the motherboard, or is it a stick that just has solder connects? Mar 01 14:57:38 directly Mar 01 14:57:39 it came with 8GB, I'm unlikely to want to upgrade that before I replace it outright Mar 01 14:58:15 getting the ram from the manufacturer also means it's guaranteed compatible Mar 01 14:58:31 I've had enough issues with ram/mb combos not quite working as specified Mar 01 14:59:11 and I'm not interested in spending a week tweaking voltages and timings to get it stable Mar 01 14:59:52 same here Mar 01 14:59:56 dont have a steady hand? Mar 01 15:00:17 single channel ram and Intel Graphics...not a "gaming laptop" Mar 01 15:00:38 mru: I have never, ever had an issue with RAM and laptops. Mar 01 15:00:40 Ever. Mar 01 15:00:57 nor me, because I buy them preconfigured Mar 01 15:01:01 desktops.... Mar 01 15:01:03 And I know of no one except you now who has. Mar 01 15:01:19 mru: We were not talking about desktops though. Mar 01 15:01:32 what difference does it make? Mar 01 15:01:34 And honestly, unless I am overclocking? Never had an issue with desktop RAM either. Mar 01 15:01:43 other than laptop bios not providing the tweaking options in the first place Mar 01 15:01:50 mru: Exactly. Mar 01 15:01:51 mru: more variety and cheaper RAM. Mar 01 15:02:02 I _never_ buy the cheapest Mar 01 15:02:14 * agmlego buys reputable brands. Mar 01 15:02:15 that's guaranteed not to work Mar 01 15:02:26 there are reputable brands? Mar 01 15:02:42 Currently, Crucial, Mushkin, G.Skill, Corsair. Mar 01 15:02:44 Yes. Mar 01 15:02:45 I've put new wireless cards in laptops for them to be 'rejected' as not being manufacturer specified, I think I've only ever upped laptop ram once and that was without issue Mar 01 15:02:58 corsair in my desktop currently Mar 01 15:02:59 jackmitchell: RAM and wireless cards are different. Mar 01 15:03:07 will not work reliably at advertised clock rate Mar 01 15:03:21 Interesting. Mar 01 15:03:23 DDR? Mar 01 15:03:26 ddr3 Mar 01 15:03:29 I know they had issues back then. Mar 01 15:03:32 agmlego: I know, I was just mentioning items I've had success with and items that have failed Mar 01 15:04:12 I've had ram in desktop systems that wouldn't run at rated timings unless they had a voltage kick Mar 01 15:13:32 * mru attemps to install yet another version of ccs Mar 01 15:13:58 at least they all fail differently Mar 01 15:14:39 mru: there is always something to be said about being unique even if it is being a unique failure Mar 01 15:16:01 and of course the error messages contain absolutely NO useful information Mar 01 15:17:16 mru: "ya brek it reel guud" Mar 01 15:18:05 sometimes the installer actually finishes Mar 01 15:18:11 don't ya just luv things that break silently :P Mar 01 15:18:24 this is anything but silent Mar 01 15:18:32 it's just not informative Mar 01 15:19:27 org.mozilla.xpcom.XPCOMException: Failed to register JavaXPCOM methods (0x80004005) Mar 01 15:19:37 helpful, right? Mar 01 15:20:02 followed by a few pages of java stack trace Mar 01 15:20:07 then "SEVERE: Can not connect to DebugServer. null" Mar 01 15:21:11 that's with the version specified in the exact steps Mar 01 15:21:32 mru: 6678? Mar 01 15:21:38 ack Mar 01 15:21:44 libav on it? Mar 01 15:21:53 _anything_ on it would be a start Mar 01 15:22:08 put bacon on it Mar 01 15:22:33 no use until it's running something to generate heat Mar 01 15:23:25 and all you get from google is ph0rum posts with people having the same problem and no solution Mar 01 15:23:38 fucking stupid shit Mar 01 15:23:45 and to think they _sell_ this Mar 01 15:26:05 someone needs to be tarred, feathered, strangled, hung, shot, drawn, and quartered Mar 01 15:26:15 mru: roger maybe? Mar 01 15:26:28 why did you not try yesterday, I had him cornered... Mar 01 15:26:28 mru, rack'em and stack'em Mar 01 15:27:04 well, I usually give things a fair try before whining for exact steps Mar 01 15:27:10 but this has gone too far Mar 01 15:31:25 laters peeps and trolls Mar 01 15:44:26 hey all Mar 01 15:44:32 hey punk Mar 01 15:45:00 mranostay: hey, why'd you split the net? Mar 01 15:49:16 not everyone can walk on water Mar 01 15:49:30 I can, if it's frozen Mar 01 15:49:38 depends on the inflatability of your shoes Mar 01 15:49:49 or what mru said Mar 01 15:50:28 mru, I am with the CCS team, for the CCS issues you are experiencing if you go to our forum http://e2e.ti.com/support/development_tools/code_composer_studio/default.aspx we would be more than happy to help you get past them Mar 01 15:50:50 the ph0rums Mar 01 15:51:13 CCS: first of all, which version of ccs should I be using with a 6678 evm? Mar 01 15:53:05 all I want to do (for a start) is flash the prebuilt linux image Mar 01 15:53:23 Honestly it depends. If you are using the MCSDK it may want you to use a specific version but in general I would recommend 5.3.0. Mar 01 15:53:25 following the steps in the included user guide gets me nothing but bizarre error messages Mar 01 15:53:50 5.3 seems to get a bit further than other versions before bailing out Mar 01 15:54:01 but the error messages are anything but helpful Mar 01 15:54:22 SEVERE: Can't generate board data file /opt/ti/.TI/605260831/0/0/BrdDat/ccBoard0.dat Mar 01 15:54:25 An invalid processor ID has been found. Mar 01 15:54:48 mru: why are you using CCS? Mar 01 15:54:57 mranostay: because they give no other option Mar 01 15:55:30 BTW what Linux distro are you running on your PC? Mar 01 15:55:36 that doesn't matter Mar 01 15:56:34 heh Mar 01 15:56:39 * mranostay gets popcorn Mar 01 15:59:53 helpful chap Mar 01 16:00:31 mru, browser malfunction Mar 01 16:01:26 can anyone explain why for spi0 pinmux sclk has to be set to INPUT_PULLUP - isn't SCLK an output like cs0? it's confusing me greatly Mar 01 16:01:28 mru, having no clue about the multiprocessor ccs support, kinda pinged for help. the team is more comfortable on the furums Mar 01 16:01:37 well, I'm not Mar 01 16:01:39 s/furums/forums/ Mar 01 16:01:55 mru, neither am I :( Mar 01 16:01:57 and when they ask what distro, I know to give up Mar 01 16:02:23 cmicali: that's just the way it is Mar 01 16:02:58 I think I'll just go and bang my head against a brick wall Mar 01 16:03:00 mru, yep, I understand. CCS team needs to track failures and improve support in future releases etc, hence the forum obsession Mar 01 16:03:01 much more fun Mar 01 16:03:01 jackmitchell: +1 Mar 01 16:03:02 cmicali: I expect something generates the clock on the chip, and then the spi block takes that clock as it's sclk Mar 01 16:03:17 NishanthMenon: the ph0rums Mar 01 16:03:21 NishanthMenon: the same problem has been posted there dozens of times Mar 01 16:03:29 mru: what's the intent behind 6678? Mar 01 16:03:36 cmicali: hence the sclk is an input on the spi block; that would be my (un)educated guess anyway Mar 01 16:03:42 av500: something new to tinker with Mar 01 16:03:46 mru, if you have a few handy links, I can point the team to them Mar 01 16:03:55 jackmitchel: ahh - that would make sense - i was wondering if it would be possible to supply an external clock to drive the mcspi controller, and i bet if that's true and external clock is a no-go Mar 01 16:04:31 cmicali: of course, it supports slave mode as well Mar 01 16:04:41 by definition it can be driven externally Mar 01 16:04:42 NishanthMenon: https://www.google.com/search?hl=en&safe=off&q=%22An+invalid+processor+ID+has+been+found%22 Mar 01 16:05:00 mru, thanks. will point them over Mar 01 16:05:06 first page is full of hits Mar 01 16:05:14 mdp: damn, now I have more STM32 than LM4 hardware here.... Mar 01 16:05:19 mdp: that's what i read, but wasn't sure if i could pinmux the right pins out to do that Mar 01 16:05:26 (at least on the beaglbone) Mar 01 16:05:36 av500: I need to collect some stm32..missed one deal a while back Mar 01 16:05:48 argh. the beaglebone-3.2 kernel doesn't build: Error: unrecognized option -mauto-it Mar 01 16:06:00 jkridner: old gcc Mar 01 16:06:00 some of the parts i want to use need other clock speeds (like 40mhz), but i'm starting to think I should just use the PRU instead of trying to make that happen Mar 01 16:06:21 k, where is a good gcc? Mar 01 16:06:22 jkridner: I've got an STM32 sitting here, I haven't had a play with it yet; are they fun? Mar 01 16:06:35 sorry mdp^ Mar 01 16:06:57 jkridner: try latest linaro Mar 01 16:07:03 that's usually a good bet Mar 01 16:07:07 jackmitchel: i noticed in a thread you said you were wiring up an fpga via SPI.. what kind of data rates did you eventually get with that? Mar 01 16:07:10 mru, http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/p/194610/695498.aspx is an better indication? Mar 01 16:07:27 cmicali: I don't crank it hard enough to do any high datarates Mar 01 16:07:56 NishanthMenon: do you see a solution there? Mar 01 16:08:03 and my spi protocol is weird, I use gpio for read/write latches and it's not a single spi read so my results are a bit moot anyway Mar 01 16:08:07 mru, nope Mar 01 16:08:07 ah Mar 01 16:08:32 mru: ppa:linaro-maintainers/toolchain ? Mar 01 16:08:32 mru, i need to send something specific to ccs guys to better match up with the symptom, so was searching Mar 01 16:08:36 I've been trying to get 1-2msps out of an a/d (16-bit words), i'm getting close but i'm running into trouble getting a nice regular/consistent stream of data Mar 01 16:08:50 cmicali: in or out of kernel? Mar 01 16:09:00 in kernel Mar 01 16:09:17 driver that's doing the transfers using chained spi_asyncs into a ring buffer Mar 01 16:10:00 jackmitchell: as emeb points out on g+ the fpu is a nice bonus for some apps on stm32..I think the discovery board layout for the expansion pins is nice...I've only done some basic toying with mine though Mar 01 16:10:19 cmicali: sounds like a sensible attempt, all I can suggest for a more regular stream is maybe waiting for 3.8-rt and see if playing with that helps... Mar 01 16:10:21 mru, I have passed on the link (google and the above e2e) to the team. Mar 01 16:11:02 this one looks very much like what I'm trying: http://e2e.ti.com/support/dsp/c6000_multi-core_dsps/f/639/t/160413.aspx Mar 01 16:11:04 jackmitchel: i've seen references to that but don't know much about it. i was reading about the PRU last night and, although code will be tedious, it looks like that might be the best way to go for now Mar 01 16:11:28 jackmitchell: https://plus.google.com/112266164281670850856/posts/S8eFF9JzAzm see comment by emeb Mar 01 16:11:29 especially since it should be almost no load on the main cpu Mar 01 16:12:15 mru, i suppose the ccxml is not known Mar 01 16:12:27 I don't even know what that means Mar 01 16:12:38 I think I have a not so fun one; STM32 Value Line Discovery Mar 01 16:13:13 there are some suggestions that only ccs 5.0 works Mar 01 16:13:15 jackmitchell: the VL parts are pretty good Mar 01 16:13:18 but that fails even quicker Mar 01 16:13:27 sigh Mar 01 16:13:37 emeb: ok; I'll try and get round to playing with it someday Mar 01 16:13:43 mru, can you update on http://e2e.ti.com/support/dsp/c6000_multi-core_dsps/f/639/t/160413.aspx ? Mar 01 16:13:44 mru: 4.7? Mar 01 16:13:56 jkridner: 4.7 is usually good Mar 01 16:14:21 they're very inexpensive (< $3), have lots of handy peripherals on-chip and they're still a full-up 32-bit CPU. Mar 01 16:14:43 yeah I got this one free from a distributor some time ago if I remember correctly Mar 01 16:14:48 emeb: can you program them from linux? Mar 01 16:14:55 av500: absolutely! Mar 01 16:15:00 this one http://jeremyherbert.net/get/stm32f4_getting_started cant :( Mar 01 16:15:03 cmicali: jackmitchell f3 or f4 Mar 01 16:15:13 I use GCC ARM Embedded + makefiles to code for STM32 Mar 01 16:15:29 and to upload? Mar 01 16:16:18 av500: the latest versions of openocd support communication with the ST-LINK programmer that's included on the Discovery boards. Mar 01 16:16:38 plus there's another FOSS toolchain that provides a GDB server. Mar 01 16:16:49 but what do you use? Mar 01 16:17:46 av500: I use an open hardware 3rd-party download gizmo. Built my own version here: http://ebrombaugh.studionebula.com/embedded/bmp2/index.html Mar 01 16:17:53 https://www.olimex.com/Products/ARM/ST/STM32-H152/resources/stm32l.cfg Mar 01 16:18:01 openocd works fine Mar 01 16:18:12 but the original is available for purchase: http://www.blacksphere.co.nz/main/blackmagic Mar 01 16:18:19 I also used openocd on f3 discovery Mar 01 16:18:32