**** BEGIN LOGGING AT Fri Feb 17 02:59:57 2012 Feb 17 04:07:40 Is anyone up here... got a simple question (for you perhaps)... confusing for me... Feb 17 05:31:32 AtomSoft?? Feb 17 07:59:17 jkridner: heh, after 3 years Feb 17 07:59:28 :) Feb 17 08:15:17 jkridner: it worked out? Feb 17 08:19:17 * etheretic gasps at the unbridled galaxy-conquering power unleashed in #beagle Feb 17 08:21:31 shame rich people's imagination can be put in a matchbox without removing the matches. Feb 17 08:22:25 (re. financing certain very interesting, userfriendly and potentially lucrative projects) Feb 17 08:42:47 oh look, got a nice reply from PixelQi re. my extraordinarily fantastic bb-based navs device :-D Feb 17 09:22:16 http://free-electrons.com/blog/ Feb 17 09:22:23 nice views on yocto Feb 17 09:29:02 http://free-electrons.com/graphics/gregory-clement.png Feb 17 09:29:14 all I can imagine that guy saying is "je suis une jeune fille" Feb 17 09:31:35 hi all, I just built my angstrom image using http://narcissus.angstrom-distribution.org/ . Now when I boot my BeagleBoard Xm rev C, I get the log in screen and also the console on the serial port. But my mouse is not detected.. Feb 17 09:31:50 I even tried using a self powered hub..even that does not work Feb 17 09:32:12 the usb port is detected and even displayed when I write lsusb Feb 17 09:32:25 but still the mouse does not work Feb 17 09:32:32 so please help me out Feb 17 09:33:26 * av500 wonders why Linux software needs to be "ported" from e.g. Fedora to Ubuntu..... Feb 17 09:44:26 anyone know a solution to my problem?? Feb 17 09:45:09 you need a hub inbetween Feb 17 09:45:14 ah no Feb 17 09:45:17 XM has a hub Feb 17 09:45:20 it should work Feb 17 09:45:49 pastebin full boot log and dmegs when you plug the mouse Feb 17 09:46:10 * LetoThe2nd notes that building a cross gcc on the bone results in OOM striking hard and fast :) Feb 17 09:47:03 This is the output I get on minicom when the board boots up: Te�as Instruments X-L ader 1.5.0 (Jun 14 2011 - 22:04:07) Beagle �M U�� boot sect r L ading u-boot.bin fr m mmc U-B ot 2011.03-rc1-00000-g9a3cc57-dirt~B �� 01 2011 - 17:41:42) _MAP36XX?37XX-GP ES2.1, CPU-OPP2, L3-165MH? j��CPU Cl ck 1 Gh _MAP3 Beagle board + LPDDR?NAND I2C: �� DRAM: 512 MiB Feb 17 09:47:30 this is not the full log..it is too long. Feb 17 09:47:35 lokks like a confused uart or rs232 adapter. Feb 17 09:47:46 but i also get these strange symbols on minicom Feb 17 09:51:04 i didn't say that i mean the one on the beagle ;) Feb 17 09:51:50 Jayneil: use pastebin Feb 17 09:51:54 but when I type lsusb in the console, it gives me a valid output such as : Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 rot hub Feb 17 09:52:08 pastebin full boot log and dmegs when you plug the mouse Feb 17 09:52:14 dmesg Feb 17 10:02:03 This is the link to my paste bin : http://pastebin.com/mV4pFgWJ Feb 17 10:02:11 This is what I get from dmseg: http://pastebin.com/uA9h6S7u Feb 17 10:03:54 JAYNEIL: fix your serial connection Feb 17 10:05:20 yes, I think I will have to buy a new serial to USB cable . This is the only working one I have right now.. Feb 17 10:06:25 what does lsusb say? Feb 17 10:06:29 (pastebi) Feb 17 10:08:18 http://pastebin.com/KDWWN8em Feb 17 10:08:35 ok, so your hub is not detected Feb 17 10:08:50 and thus your mouse is neither Feb 17 10:09:59 so what should I do? Feb 17 10:10:07 use something newer than 2.6.32 Feb 17 10:10:35 http://www.angstrom-distribution.org/demo/beagleboard/ Feb 17 10:12:31 I see. But when I built my image from http://narcissus.angstrom-distribution.org/, I selected the latest release from 2011. So is this the wrong approach? Feb 17 10:13:43 no Feb 17 10:15:15 then why does this approach not work? Feb 17 10:15:32 Is there anything I can do to make this approach work? Feb 17 10:18:32 I dont know why it does not work Feb 17 10:19:05 ok. thanks a lot for your help. Feb 17 10:22:54 there was something about the GPIO polarity to powe the hub changing Feb 17 10:23:04 so newer kernels/uboots should have that Feb 17 10:24:00 JAYNEIL: I would try "next" image Feb 17 10:29:36 I am now downloading the Angstrom-Beagleboard-demo-image-glibc-ipk-2011.1-beagleboard.rootfs.tar.bz2 file..I hope that it works Feb 17 10:32:27 JAYNEIL: no Feb 17 10:32:34 thats 2011.1 Feb 17 10:32:38 use some of the 2012 stuff Feb 17 10:32:48 Angstrom-TI-GNOME-image-eglibc-ipk-v2012.01-core-beagleboard-2012.01.11.img.gz Feb 17 10:32:53 MLO-beagleboard-2011.12 Feb 17 10:33:05 uImage-3.0.17-r115c-beagleboard-20120125152700.bin Feb 17 10:33:12 u-boot-beagleboard-2011.12-r4.img Feb 17 10:41:14 ok. I see.. btw are you klaas ? Feb 17 10:44:16 is that a trick question? Feb 17 10:44:34 but no :) Feb 17 10:46:05 actually I am going to make a video series on this soon. so would like to acknowledge you for the help..thats all.. and av500 is not someone's name. Feb 17 10:48:34 got it! Vladimir, thanks once again! Feb 17 10:53:55 JAYNEIL: try to keep my name out of your video series... Feb 17 10:58:45 ok.. will only mention av500.. Feb 17 11:48:29 av500: my G9 has the same or similar issue dm8tbr reported at Fosdem: spurious touchscreen events at the lower left corner. reproducible only after charging, goes away after touching screen for ~20 seconds. Feb 17 11:49:37 hmm Feb 17 13:13:09 can someone help me about jtag debugging with XDS100v2 emulator ? Feb 17 13:14:01 question impossible to answer Feb 17 13:15:01 www.crystal-balls-on-sale.com Feb 17 14:42:52 hello can someone help me Feb 17 14:43:30 I would like to use the UART (console) of my beagleboard-xm for serial cummunication. Feb 17 14:44:25 i havn't found any good infos on the web yet Feb 17 14:44:45 did anybody tried this yet Feb 17 14:47:00 its a standard linux UART Feb 17 14:49:10 well what i did is remove console=... from uEnv.txt and delete ttyO2.conf Feb 17 14:49:24 but i still cannt make use of it Feb 17 14:50:18 define "can'" Feb 17 14:50:20 define "can't" Feb 17 16:08:57 chaosss: http://catb.org/~esr/faqs/smart-questions.html Feb 17 16:09:43 chaosss: better yet, http://beagleboard.org/chat and follow the link to http://www.mahdiyusuf.com/post/6176918750/survival-guide-irc-1 Feb 17 16:17:36 can some one help me about beagleboard c4? i just want to learn basic things Feb 17 16:18:17 basic things can be googled Feb 17 16:19:57 anyone? Feb 17 16:20:46 try to actually ask a question... Feb 17 16:26:29 there is a new trend of asking to ask questions lately Feb 17 16:30:06 jsabeaudry: new it is not Feb 17 16:30:11 hey guys, i was thinking about asking a question about asking a question. Would that be alright? .... Feb 17 16:30:46 Could someone answer a question about BlInK311's question? Feb 17 16:31:10 anyone? Feb 17 16:31:55 If you cant answer my inteded question right away. ill keep asking. without any more detail on what I want. Feb 17 16:32:28 intended* Feb 17 16:50:47 Hi, I am newbie. Before, i have used Texas Instruments C5505 type DSPs with code composer studio. To investigate omap platform i have ordered beagleboard and XDS100v2 emulator.I don't want to use OS. I can connect omap3530 and see registers etc via emulator. Which steps should i follow to compile and flash assembly codes in Code composer studio v5? Feb 17 16:59:17 chaoss i see you have been asking at the ti forum too :) sorry i cant help Feb 17 17:02:34 chaosss: and you have code to boot the omap3 and setup SDRAM etc? Feb 17 17:07:32 morning Feb 17 17:08:26 yes i have asked in ti forum i hope i get an answer Feb 17 18:03:15 I'm beginner in embedded linux, need some help.can someone suggest me whether to buy beaglebone or beagleboard ? Feb 17 18:03:33 Which one is easier to start with Feb 17 18:04:13 please helpp..... Feb 17 18:04:14 aj2903, My guess is that beagleboard should be easier beacause it has been around longer: thus more documentation, blog posts, etc. Feb 17 18:04:56 so shall i buy beagleboard or beagleboard-xm Feb 17 18:05:08 aj2903 depends on your needs Feb 17 18:05:54 thanks for help Feb 17 18:06:15 o.O Feb 17 18:06:34 does beagleboard has CAN support ??? Feb 17 18:07:30 i'mean CAN protocol Feb 17 18:07:33 not like bone Feb 17 18:08:02 y? Feb 17 18:08:22 the bone soc has more hardware support for industrial protocols Feb 17 18:09:22 none of them has can you need an adapter Feb 17 18:09:48 isn't their a CAN cape? Feb 17 18:09:52 *there Feb 17 18:10:49 CAN cape ???? i haven't seen yet. Feb 17 18:10:55 is there any link Feb 17 18:11:26 http://www.towertech.it/en/products/hardware/tt3201-can-cape/ Feb 17 18:13:45 thanks @emeb, but CAN cape is for beaglebone. can it be used for beagleboard or beagleboard-xm Feb 17 18:14:41 probably not with level shifting Feb 17 18:14:45 *without Feb 17 18:17:27 but pinout may also be different... Feb 17 18:18:23 _will_ be different - I/O connectors on 'bone & BB/BBXM are not the same. Feb 17 18:18:30 is there any other add-on board for beagleboard with CAN protocol ???? Feb 17 18:18:54 support for CAN protocol ???* Feb 17 18:18:56 not to mention the I/O voltage Feb 17 18:19:01 Plus, CAN cape for 'bone uses CAN controller on the AM335x SoC. OMAP335x doesn't have CAN controller. Feb 17 18:19:25 emeb ah good to know Feb 17 18:19:35 aj2903 so buy a bone Feb 17 18:19:36 aj2903: any reason you can't get a bone? Feb 17 18:19:43 <_av500_> mranostay: just do a 50% pwm on the 3.3v gpios to get 1.8V Feb 17 18:19:54 hi av500 Feb 17 18:20:00 av500 do you need CAN? Feb 17 18:20:01 thanks emeb .... Feb 17 18:20:10 np Feb 17 18:20:24 <_av500_> woglinde: I can! Feb 17 18:20:38 _av500_: hmmm didn't think of that Feb 17 18:20:46 emeb: omap3517 has can (cranboard) Feb 17 18:20:50 +e Feb 17 18:20:52 probably since it has been a long time since i done any backlight work Feb 17 18:20:57 cranberry? Feb 17 18:21:00 *i've Feb 17 18:21:01 but bone doesn't has nand/nor flash memory... this can be little problem. Feb 17 18:21:12 aj2903: why? :) Feb 17 18:21:14 aj2903: can't use the SD? Feb 17 18:21:18 <_av500_> nand nor flash Feb 17 18:21:19 MMC SD ftw Feb 17 18:21:28 <_av500_> emmc ftw Feb 17 18:21:33 * mranostay kicks _av500_ Feb 17 18:21:40 _av500_: there was an eemc talk yesterday Feb 17 18:21:46 clay tablets FTW Feb 17 18:21:51 koen: how was that one? Feb 17 18:21:55 mranostay: dunno Feb 17 18:21:58 didn't attend Feb 17 18:21:59 but sd card is not good as compared to nand/nor flash memories Feb 17 18:22:02 ok same here Feb 17 18:22:20 aj2903???? how you exchnage nand? Feb 17 18:22:31 woglinde: screwdriver Feb 17 18:22:34 aj2903: what's your metric for "as good"? Feb 17 18:22:45 aj2903: ahem SD is nand :) Feb 17 18:22:48 speed / endurance / cost / etc? Feb 17 18:23:00 just with another controller interface Feb 17 18:23:01 all is nand but you dont see it Feb 17 18:23:39 koen still in us? Feb 17 18:23:48 u mean sd-card performance is as good as nand/nor flash memories ??? Feb 17 18:24:13 woglinde: yes Feb 17 18:24:19 aj2903 o.O what you have to write out on internal mem? Feb 17 18:24:24 woglinde: sitting next to Tartarus and mranostay Feb 17 18:24:44 av500: one of the conference rooms is called 'champagne' Feb 17 18:24:45 * Tartarus waves as pidgin sprung up :) Feb 17 18:24:57 hi tom Feb 17 18:26:26 continuos operation on sd-card can damage it ???? Feb 17 18:26:42 is it true ??? Feb 17 18:26:49 of course Feb 17 18:26:55 you didn't know? Feb 17 18:27:00 hey woglinde Feb 17 18:27:15 Holy crap, starting a kernel from localmodconfig helps, but distro stuff still has tons of shit on Feb 17 18:27:19 * Tartarus will do a few passes at this, heh Feb 17 18:28:06 Tartarus: ok i never heard of localmodconfig until yesterday Feb 17 18:28:23 that is so helpful though Feb 17 18:29:28 what is localmodconfig o_o Feb 17 18:29:31 if linux is booted from sd-card,and some application is continuos running, will this damage sd-card Feb 17 18:29:46 usually smoke appears Feb 17 18:30:37 aj2903: wear leveling and bad blocks should prevent that Feb 17 18:30:56 and who cares really. SD/MMC is cheap Feb 17 18:31:19 mranostay hm localmodconfig was discussed some time ago Feb 17 18:31:23 mranostay: In some cases, yeah. I wouldn't use it on my box where I don't know what random USB shit i'll need Feb 17 18:31:38 or network (due to bridging, etc, for kvm quests) Feb 17 18:31:38 that's true but if application shutdown, it can be major problem Feb 17 18:31:40 guests Feb 17 18:31:48 aj2903, damage after a few thousand writes Feb 17 18:32:03 panto depensd on the quality Feb 17 18:32:06 under normal load it won't happen Feb 17 18:32:09 agreed Feb 17 18:32:10 it can break sooner or later Feb 17 18:32:16 you never know Feb 17 18:32:23 but most SD cards are MLC now Feb 17 18:32:24 but than you exchange it Feb 17 18:32:54 so expect less reliability Feb 17 18:32:55 aj2903 I think your canbus stuff can reside in ram only so no write cycles needed on sd Feb 17 18:33:16 oh no panto is here too Feb 17 18:33:16 aj2903, write your temp data in a tmpfs partition Feb 17 18:33:25 only write to the device when you need to Feb 17 18:33:58 but i also will have qt application running on the board, in future Feb 17 18:34:17 koen: your famous now :P Feb 17 18:34:28 aj2903 and? need this to write something to the sd? Feb 17 18:34:35 yes Feb 17 18:34:42 what? Feb 17 18:34:46 I really wonder Feb 17 18:36:22 no,i mean qt application would be running for 24 hrs, so it will be running from sd-card. is my assumption correct ?? Feb 17 18:36:32 usb gadget guy showing off beagle right now Feb 17 18:36:53 sounds like a dog show. Feb 17 18:37:00 aj2903 depends you can load initrd from sdcard and run all in ram Feb 17 18:37:06 bone has plenty of ram Feb 17 18:37:12 as long as it isn't a donkey one :) Feb 17 18:37:19 cows Feb 17 18:37:31 * emeb can't read the screen on koen's g+ pix Feb 17 18:37:41 thanks woglinde Feb 17 18:37:54 sure try this Feb 17 18:38:01 sure Feb 17 18:38:04 alternative Feb 17 18:38:06 but first buy it Feb 17 18:38:19 hm I should hire at ti as salesman Feb 17 18:38:38 no that's my problem i will sort out Feb 17 18:38:43 ah no bone is toy too Feb 17 18:38:44 woglinde: do you have "salesman hair"? Feb 17 18:38:52 emeb I can make it so Feb 17 18:38:54 I think Feb 17 18:39:04 with enough product, anything is possible. Feb 17 18:39:18 not in all cases Feb 17 18:40:13 guys,thanks for help Feb 17 18:40:18 sure Feb 17 18:40:30 buy me a beer once Feb 17 18:40:36 not twice? Feb 17 18:40:56 maybee when he needs an advice again Feb 17 18:41:13 * woglinde backports some meta-java stuff Feb 17 18:42:03 woglinde,next time i will keep box of beer for u. :) Feb 17 18:47:37 hm Feb 17 18:47:39 https://blogs.oracle.com/jtc/entry/comparing_jvms_on_arm_linux Feb 17 18:48:52 oh no he tested on debian 6.0.4 Feb 17 18:49:00 which means armv code Feb 17 18:49:03 ups armv4 Feb 17 18:52:12 woglinde, You beleive the results would be much different on armv7a? Feb 17 18:52:25 woglinde, Thanks for the link btw, very interesting Feb 17 18:52:38 yes xerces comment it Feb 17 18:54:31 hi crofton Feb 17 18:54:37 so has anyone had luck getting ALSA to work out of the box/ Feb 17 18:54:38 ? Feb 17 18:56:16 gm Feb 17 18:56:28 bbl Feb 17 18:56:45 hmm, a hit and run . . . Feb 17 18:57:16 TheAlphaNerd: probably everyone Feb 17 18:57:27 mranostay: I havn't Feb 17 18:57:41 on the version of ubuntu from the wiki audio out works Feb 17 18:57:42 woglinde, Do you know if angstrom uses the same optimizations as in Debian wheeze armhf? (In other words, would my JVM application run faster if I switch to Debian wheezy armhf ? Feb 17 18:57:43 but not in Feb 17 18:57:51 angstrom both works Feb 17 18:57:55 linaro nothing works Feb 17 18:58:15 linaro? Feb 17 18:58:19 you mean ubuntu Feb 17 18:58:26 linaro makes only tools Feb 17 18:58:29 not distro Feb 17 18:58:48 the linaro distributed ubuntu Feb 17 18:58:50 is what I mea Feb 17 18:59:48 you mean ubuntu with the included linaro toolchain Feb 17 19:00:01 but there isnt much support here for it Feb 17 19:00:14 because of using older kernels which often dont have all features Feb 17 19:04:25 word Feb 17 19:04:55 I have to run for now Feb 17 19:05:02 but perhaps someone can help me in a bit Feb 17 19:09:43 if someone asks again about can, there are usb-can-devices with linux-drivers available Feb 17 19:09:57 aholler thats what I thought Feb 17 19:10:13 TheAlphaNerd yes use angstroem Feb 17 19:11:14 woglinde: problem with angstrome is that I couldn't get puredata to compile properly for it :( Feb 17 19:11:39 and they don't have packages for a bunch of stuff I need Feb 17 19:16:16 TheAlphaNerd: sometimes pkgs are obscurely renamed. Did you try the angstrom pkg browser? Feb 17 19:17:49 TheAlphaNerd than make the recipes it isnt that hard Feb 17 19:19:46 fraviofii: no pm pls Feb 17 19:39:33 so i'm doing some poking around Feb 17 19:39:36 i mean like literally: Feb 17 19:39:44 root@bone:~# devmem2 0x44e00000 w Feb 17 19:39:44 /dev/mem opened. Feb 17 19:39:44 Memory mapped at address 0x4005e000. Feb 17 19:39:44 Read at address 0x44E00000 (0x4005e000): 0x00024502 Feb 17 19:40:38 but some stuff just simply doesn't work. Like, I should be able to access I2C2's registers through L4, but when I try, I get a bus error: Feb 17 19:40:44 root@bone:~# devmem2 0x4819c000 b Feb 17 19:40:44 /dev/mem opened. Feb 17 19:40:44 Memory mapped at address 0x402f7000. Feb 17 19:40:44 Bus error Feb 17 19:40:51 koen's session, "[am335x has] more features than balls" Feb 17 19:41:22 [ 1031.557724] Unhandled fault: external abort on non-linefetch (0x1018) at 0x4022d000 Feb 17 19:41:27 that's what dmesg says. What does this mean? Feb 17 19:43:39 Could it be because the PRUSS is just powered off ? Feb 17 19:44:45 yes Feb 17 19:45:09 That'd make it make a bus error? interesting Feb 17 19:45:19 mdp: he is only .eu'ian that seems to get the Champagne Room connotation :) Feb 17 19:45:32 hehe Feb 17 19:46:10 ok so Feb 17 19:46:17 i can leave it in reset though and i should be able to read its registers right Feb 17 19:47:43 koen explains MUX? Feb 17 19:49:21 so I want the PER domain. So I want PM_PER_PWRSTCTRL, gotta find that. Feb 17 19:49:40 hmmmm Feb 17 19:50:51 ds2: you at koen's talk? Feb 17 19:52:41 DT for world peace Feb 17 19:53:03 hi silicagel Feb 17 19:53:17 hi there. I am up to no good. Feb 17 19:54:12 koen he must miss something like a storm flood. today there were millions of netherlands on our autobahn. I've really thought they must escape from something Feb 17 19:54:37 silicalgel already read -> http://labb.zafena.se/ ? Feb 17 19:56:12 nope, i'm checking it out now Feb 17 19:57:56 OK. PRM_PER lives at 0x44E00C00. The offfset to PM_PER_PWRSTCTRL is 0x0C Feb 17 19:58:12 SilicaGel: you're doing PRU bringup? Feb 17 19:58:19 Attempting to! Feb 17 19:58:25 gogo Feb 17 19:58:35 cool Feb 17 19:58:43 nope dammit Feb 17 19:58:46 you get the assembler? Feb 17 19:58:48 I have something wrong. Feb 17 19:59:00 yeah but I'm not even there yet Feb 17 19:59:14 I'm using devmem2 to try to read/write the PRU memory Feb 17 19:59:20 butI have to turn said memory on first Feb 17 19:59:23 I seem to be stuck at that part :) Feb 17 19:59:37 oh, getting pru clocks up? Feb 17 20:00:10 no. Should I need to get clocks up to read their RAM? Feb 17 20:00:22 I was hoping I could just turn the pru on in PER_PWRSTCTRL Feb 17 20:00:28 the whole PRUSS needed a few clocks Feb 17 20:00:32 what else should make that ram work? Feb 17 20:00:33 iclk, fclk Feb 17 20:00:42 a couple more iirc Feb 17 20:01:13 maybe I need to get it out of reset Feb 17 20:01:14 you were talking about clocks yesterday, so i figured you did that part Feb 17 20:01:18 no I don't think so Feb 17 20:01:21 I think I can leave it in reset Feb 17 20:02:02 ok. ok then. Feb 17 20:02:09 if the PRUSS is not on the L3/L4 bus it won't work at all Feb 17 20:02:27 and the clocks are required to get it up on L3/L4 ? Feb 17 20:02:29 you'll get data abort when you try to access their PA Feb 17 20:02:31 yes Feb 17 20:02:38 ah ok Feb 17 20:02:43 ok I found CM_PER_PRUSS_CLKCTRL Feb 17 20:05:20 hmmm CLKSEL_PRUSS_OCP_CLK. I dno't knwo what an OCP is. Feb 17 20:05:36 open core protocol Feb 17 20:05:48 OCP = L3 Feb 17 20:08:13 ok. CM_PER = 0x44E00000. CM_PER_PRUSS_CLKCTRL is at offset 0xE8. And I have to write 0x04 to wake it up it looks like Feb 17 20:08:40 goddammit bus error Feb 17 20:09:19 there's usually 4-5 hoops you have to jump through to bring these things up Feb 17 20:10:04 yeah. Feb 17 20:10:18 well I see in 8.1.1.3.1 what you are talking about, with there being interface clocks and functional clocks Feb 17 20:10:32 hopefully if I keep reading it will tell me how to set them up Feb 17 20:10:58 there's usually an enable register Feb 17 20:11:00 CLKEN Feb 17 20:12:26 There was a register named CM_PER_PRUSS_CLKCTRL which has a bit STBYRST (i set it to zero), and MODULEMODE I set to 2 - ENABLE: module is explicitly enabled. Interface clock may be gated according to the clock domain state." :( Feb 17 20:12:53 ok so i must be missing an enable like you said Feb 17 20:13:15 * mdp heckles koen Feb 17 20:13:30 the best way to heckle him is to remove yoru pants Feb 17 20:14:21 check PM_PER_PWRSTCTRL Feb 17 20:15:23 I did set a bit on in that register already, there's bits 5 and 6 which are named pruss_mem_ONState. I set them to both bits on Feb 17 20:17:23 so PRUSS has 3 clocks required Feb 17 20:17:40 CORE_CLKOUTM4, PER_CLKOUTM2 and CORE_CLKOUTM4 Feb 17 20:17:48 you need to make sure all those are up and running Feb 17 20:18:03 see 4.2.2 Feb 17 20:18:34 ok Feb 17 20:18:40 l3_clk is almost certainly running if your board is up Feb 17 20:18:40 i probably don't REALLY need the uart clock to be on right Feb 17 20:18:45 uart_clk and iep_clk dunno Feb 17 20:19:04 it may not work with one of them off Feb 17 20:19:18 you gotta try it a bunch of times to see what it wants Feb 17 20:20:03 but I still need to probably set that "connetion to OCP clock" but on too right, in CM_PER_PRUSS_CLKCTRL ? Feb 17 20:21:48 STBYST should be 0 Feb 17 20:21:52 IDLEST 0 Feb 17 20:22:00 MODULEMODE 2 Feb 17 20:25:33 I see all kinds of references to CORE_CLKOUTM4 but not a bit or somethign that says to turn it on Feb 17 20:26:15 that's the L3_CLK Feb 17 20:26:24 you wouldn't be up and running if that weren't on Feb 17 20:26:41 oh oh i see it just needs to be *on* Feb 17 20:26:43 and it probably already is Feb 17 20:26:45 or I'd be hosed Feb 17 20:27:29 but i'm not really that familiar with the am335x Feb 17 20:27:41 it looks a little different Feb 17 20:28:21 the clock tree is a complicated thing that can be really frustrating Feb 17 20:28:39 well the PLLs must already be configured Feb 17 20:28:52 they must just be gated, is my thinking Feb 17 20:28:52 some of them obviously, maybe not everything Feb 17 20:29:08 gated either because they're switched off manually or the module is in idle state Feb 17 20:30:20 can you read the PRUSS REVID register? Feb 17 20:30:33 If it's in the PRU block theN I doubt it. Let me try. Feb 17 20:34:00 still looking for it Feb 17 20:36:37 man, that TRM sucks… they don't list the PA of the register with the register description Feb 17 20:41:00 ok i have to get organized here Feb 17 20:41:04 I"m writin ga shell script, God help me Feb 17 20:41:21 Step 1: enable clock to the PRU in CM_PER_PRUSS_CLKCTRL Feb 17 21:11:04 looking at the init code from the kernel and it only does 2 things: enable the clock, and take the module out of reset Feb 17 21:18:50 woglinde: do you know where there is online documnetation on how to make recipes? Feb 17 21:20:33 SilicaGel: can you read the REVID yet? Feb 17 21:23:44 TheAlphaNerd read other recipes and you will learn it quickly Feb 17 21:23:52 word Feb 17 21:23:53 haha no I dont' even know where the hell REVID is Feb 17 21:23:55 I tried! Feb 17 21:24:01 jay6981: I'll show you waht I'm doing in one second Feb 17 21:24:37 aplhanerd read http://www.openembedded.org/wiki/How_to_create_a_bitbake_recipe_for_dummies Feb 17 21:26:49 alphanerd this good too http://www.yoctoproject.org/docs/current/dev-manual/dev-manual.html Feb 17 21:28:12 alphanerd and read other recipes and ask on #oe when you are stucked somewhere Feb 17 21:29:51 jayu here is my attempt to enable the PRU: http://pastebin.com/LHVtguqG Feb 17 21:30:44 what's 401af000? Feb 17 21:31:54 oh that's to be ignored. devmem2 shows you the address of the local mapping point for some reason, it's ... Feb 17 21:32:15 Geez, I'm stuck. I've got a custom camera hooked up to the ccdc. First frame captures fine, but further frames are coming back 0x55 (yavta initialized value). VD0 is firing ok. I suspected WEN problem, but EXWEN bit in CCDC_SYN_MODE is cleared. Ring any bells with anyone? Feb 17 21:32:22 it's the point that it's mapped into userspace Feb 17 21:32:23 what does `devmem2 0x4a320000 w` do? Feb 17 21:32:25 it's an artifact of devmem2 Feb 17 21:32:30 ok so Feb 17 21:32:34