**** BEGIN LOGGING AT Sun Aug 03 03:00:00 2014 Aug 03 05:15:31 Abhishek_ : yes, I frankly lost track of time :p normally don't even remember what day it is. sunday ? monday ? blah! :)) Aug 03 05:15:59 * karki will remember days from tomorrow. college starts :( Aug 03 05:27:15 karki: BTW, how was your interview? Aug 03 05:27:35 which one? Aug 03 05:41:16 Abhishek_ : the ebay one sucked. half the questions were C++ based, and I didn't know C++ :') Aug 03 05:42:07 I suppose they would have asked templates and stuff? Aug 03 05:42:21 oh! and I need some help with the modularization of my driver. I'm done with most of the coding part, just need steps to figure out what to do next Aug 03 05:42:33 yeah, even virtual functions etc Aug 03 05:42:43 I didn't even know C++ syntx Aug 03 05:53:11 hmm, your modularization did not even require a function pointer table like in my case :) Aug 03 06:37:03 well yeah, I just use the existing functions as is. since my case was different from yours, and I don't need any glue, I kept it simple :p Aug 03 06:58:24 sending Buffer objects over Socket.IO is *gross* . My decision to use base64 turns out justified Aug 03 07:29:13 Abhishek_ : how do i go about the compile and test of my kernel? I'm kinda stuck :x Aug 03 07:31:10 replace zImage in the partition with your compiled zImage and copy all the modules to a new directory Aug 03 07:35:02 Abhishek_ : so I have to download the kernel source, make the changes there, recompile. i didn't get the "copy all the modules to a new directory" Aug 03 07:36:26 you have to install all the kernel modules in a /lib/mpdules/3.8.13-xxxxxxx/ directory on the Bone Aug 03 07:36:33 I assume you are going to cross-compile the kernel Aug 03 07:37:08 yes Aug 03 07:38:03 Abhishek_ : is this the kernel to clone - modify - compile? Aug 03 07:38:05 https://github.com/beagleboard/linux/blob/3.8 Aug 03 07:38:15 or is there another one? Aug 03 07:39:58 hmm, seems so Aug 03 07:40:27 however: http://github.com/RobertCNelson/bb-kernel Aug 03 07:40:38 https://github.com/RobertCNelson/bb-kernel/tree/am33x-v3.8 Aug 03 07:40:52 is the one you would want your sources to go into Aug 03 07:41:04 okay Aug 03 07:41:25 s/sources/patches Aug 03 07:45:29 wait, is it fine if I just recompile the kernel with pantos driver being made a loadable module? cause I would be planning to make some modifications to pru_rproc as I go along, don't want to recompile all the time :p Aug 03 07:45:34 Abhishek_ Aug 03 07:46:19 hmm, it should be, provided you keep a different name otherwise it might conflict with the original remoteproc Aug 03 07:46:50 I followed this approach until I felt the changes were ready to be sent as patches Aug 03 07:47:48 cool! Aug 03 07:48:00 ah wait Aug 03 07:48:12 I compiled my kernel without "remoteproc" for this to work Aug 03 07:48:51 look at the pru_rproc_bl branch in BeagleLogic, that module is archived for "historical" reasons Aug 03 07:50:12 (oops, it got rebased it seems) Aug 03 07:54:29 I realized that I should've done all this on a different branch after like 2 commits towards new kernel module. :( then I let it be and continued commiting. any idea how to move the previous x commits to a different branch? Aug 03 07:55:26 git branch newbranch and then checkout master and hard-reset it to HEAD~2 Aug 03 07:56:14 not a good idea though to do forced updates on your principal GitHub repo as it may break forks Aug 03 08:10:41 Abhishek_ :another tiny doubt Aug 03 08:10:50 the zImg is that compiled image of the kernel right? Aug 03 08:11:03 where exactly does debian stuff come in? Aug 03 08:11:04 yeah Aug 03 08:11:21 OS specific stuff that is Aug 03 08:12:00 the devicetree files, the firmware, though ideally should be a part of the kernel are shipped separately Aug 03 08:12:30 the device tree is bundled for sure, but the firmware in my case was bundled in a separate repo Aug 03 08:14:02 the web frontend that I will be bundling will form a part of the web interface Aug 03 08:14:25 er, I mean, the web frontend will now be distro-specific Aug 03 08:15:04 * karki_ patiently waits while the kernel sources are cloned. Aug 03 08:15:07 :'( Aug 03 08:21:24 Abhishek_ : btw how do I use the function exported in one module in another? is there any thing specific I should do in header files, compiling etc? Aug 03 08:37:31 just use it, dependencies will be taken care of by the build system Aug 03 10:46:28 karki_: http://2net.co.uk/tutorial/android-4.4-beaglebone Aug 03 10:46:37 you asked, right? Aug 03 10:50:44 vvu : :D thanks! Aug 03 10:50:58 doesn't look like you are taking a break :p Aug 03 10:57:16 eh, just a bit of skimming around Aug 03 10:57:35 no laptop battery in the last 3 days so break Aug 03 12:11:19 Abhishek_ : latest commits here - https://github.com/deepakkarki/pruspeak/tree/pru_rproc_ps/src/driver and the export feature is not working https://gist.github.com/deepakkarki/1710920887869f4d2b98 Aug 03 12:11:23 :( Aug 03 12:11:48 the rproc module compiles with my changes, pru_speak does not! Aug 03 14:33:28 karki_: Did you try make clean? Aug 03 14:33:59 I mean, you had a copy of the kernel sources on the Bone I suppose Aug 03 14:34:43 yeah Aug 03 14:36:27 Abhishek_ : I did not recompile the kernel. I just created these two modules. and edited my Makefile Aug 03 14:37:05 I renamed pru_rproc.c to pru_rproc2.c Aug 03 14:37:13 so no clashes Aug 03 14:37:39 but anyway I'm trying to figure why it isin't compiling! Aug 03 14:38:35 karki_: try adding both these functions as "extern" prototypes in pru speak Aug 03 14:38:53 will try! Aug 03 14:40:13 Abhishek_ : while writing the code, pru_rproc.c was a module for you right? and not part of the kernel. Aug 03 14:40:52 yeah Aug 03 14:41:31 I was wondering whether this compile failed as pru_rproc2 is a module and not part of the kernel (but I dont think that could be a problem) Aug 03 14:41:53 hmm... so module worked for you. it should work for me too then Aug 03 14:58:51 Abhishek_ : I get a warning now instead of an error - /root/pruspeak/src/driver/pru_speak.h:42:12: warning: ‘pruspeak_downcall_idx’ used but never defined [enabled by default] Aug 04 01:16:16 that is a interesting warning **** ENDING LOGGING AT Mon Aug 04 02:59:59 2014