**** BEGIN LOGGING AT Mon Apr 03 03:00:02 2017 Apr 03 03:32:49 KeyError: Identifier('rishabh_bb') (file "/usr/local/lib/python2.7/dist-packages/sopel/tools/target.py", line 65, in clear_user) Apr 03 03:33:49 KeyError: Identifier('rishabh_bb') (file "/usr/local/lib/python2.7/dist-packages/sopel/tools/target.py", line 65, in clear_user) Apr 03 05:33:36 pflogsum Apr 03 05:33:53 silly mouse focus... Apr 03 07:11:44 zeekhuge : plz review my proposal http://elinux.org/BeagleBoard/GSoC/beagle_offloader ,I have updated it as you have suggested. Apr 03 08:01:46 ds2 : plz give your comments about my proposal Apr 03 08:04:01 7 hours left for the deadline Apr 03 10:36:44 hello Can we use both PRUs at a time for same peripheral offloading ? i.e some part of processing done by PRU0 then using Scratch Pad feature pass it to PRU1 then got to read new data PRU1 will do further processing !! Apr 03 10:37:52 I am talking about PRUs offload data processing project Apr 03 10:42:30 . Apr 03 14:08:09 m_w: http://elinux.org/BeagleBoard/beaglecv_stereo Apr 03 14:28:56 how to implemet hello world application on arm linux environment? Apr 03 14:41:33 kiran4399: did you submit to the gsoc page? Apr 03 14:41:41 m_w: yeh Apr 03 14:41:56 m_w: you mean on the summerofcode.google.com? Apr 03 14:42:52 kiran4399: nevermind I see your entry Apr 03 14:43:20 m_w: do you want the google docs link? Apr 03 14:44:29 kiran4399: not necessary Apr 03 14:46:22 m_w: are you looking at my wiki page? Apr 03 14:46:32 How many hours left for the deadline ? Apr 03 14:46:57 1 hr 15 mins Apr 03 14:48:04 zeekhuge : please review and give suggestions http://elinux.org/BeagleBoard/GSoC/PRUs_to_offload_processing_of_raw_data Apr 03 15:21:18 zeekhuge: please review this http://elinux.org/BeagleBoard/GSoC/beagle_offloader Apr 03 15:30:32 m_w: can you give me a quote? Apr 03 15:30:40 m_w: on my project? Apr 03 15:38:20 kiran4399: not really good at qoutes Apr 03 15:38:39 m_w: how is my proposal? Apr 03 15:38:45 can even spell it right :) Apr 03 15:38:50 can't Apr 03 15:42:30 it is okay Apr 03 15:43:10 m_w: Any changes you would like to suggest? Apr 03 15:43:48 is there a reason you are specifying the kernel version? Apr 03 15:45:23 m_w: because the SGX530 requried kernel modules are not present in any other kernel version Apr 03 15:53:39 m_w: please give your suggestions: http://elinux.org/BeagleBoard/GSoC/beagle_offloader Apr 03 16:10:12 "This has greatly improvised my understanding .." Apr 03 16:10:24 do you mean improved? Apr 03 16:14:09 Yeah.... Typo Apr 03 16:18:11 more specific offloading task descritions would be nice Apr 03 16:18:51 practical examples Apr 03 16:19:12 isn't the proposal due already though? Apr 03 16:21:44 Yes proposal submission time is finished Apr 03 16:23:06 But while checking the pdf do mentors read the final pdf or elinux page? Apr 03 16:35:42 Depends on what mood they are in probably! Apr 03 16:42:42 the decisions are supposed to be based on the pdf proposals Apr 03 16:43:55 but the wiki pages will likely be referenced Apr 03 17:10:27 Can we do changes in wiki if needed? Apr 03 17:16:13 After 5th April I have free time because my exams over on 5th so I am planning to do more work on my BBB so can I add my results or understanding on wiki? Apr 03 17:48:27 Me0w Apr 03 18:13:12 All deadlines were at 12:30 AM IST last year. Apr 03 18:13:24 though I am late even according to that .. Apr 03 18:28:33 Can anyone have opinion on my idea of using both PRUs to same peripheral I mean even distribute work between both prus ? Apr 03 18:29:30 I am vary excited about this 😀 Apr 03 18:30:41 zeekhuge: what do you say about it? Apr 03 18:30:51 what is your partitioning stategy? Apr 03 18:32:25 By using both prus for same peripheral I mean some part of processing done by PRU0 then using Scratch Pad feature pass it to PRU1 then got to read new data PRU1 will do further processing !! Apr 03 18:33:54 so define "some" and how would you manage to do that with PRUs not clashing with each other ? Apr 03 18:34:52 again, how ware you partitioning it? Apr 03 18:35:15 and what are you augmenting the peripheral with? Apr 03 18:35:40 there are very few reasons bypassing the proper kernel drivers with direct access Apr 03 18:36:09 Ok like pru0 connected to peripheral so it will fetch data Apr 03 18:36:39 how is that different from having EDMA fetch data? Apr 03 18:38:24 Now pru0 will remove header and send data for further processing to pru1 . I read on zeekhuge site that it takes one cycle to swap register between pru0 an pru1 Apr 03 18:39:27 that was between PRU0 and the scratch-pad I guess ? **** BEGIN LOGGING AT Mon Apr 03 18:42:20 2017 Apr 03 18:44:23 Ok , image on your block shows direct connection between pru0 and pru1 Apr 03 18:45:54 So I guess I need to read more in details abut that. Apr 03 18:48:13 But even if pru0 or pru1 can only swap register between bank not direct than intrupt can inform pru1 that data is written in scratch pad so now read and process it. Right? Apr 03 18:49:05 *sorry typo blog ( not block) Apr 03 22:24:32 hi? Apr 04 01:14:31 Hello ds2? Apr 04 01:19:04 Is anyone online? Apr 04 01:24:43 Apr 04 01:26:43 Hi ds2? Apr 04 01:27:01 Can I discuss with you? Apr 04 01:28:25 what would you like to discuss? Apr 04 01:28:44 Regarding synchronus collection Apr 04 01:28:52 'Even if it is not selected for gsoc Apr 04 01:29:00 I want to work on it Apr 04 01:29:06 'k Apr 04 01:29:09 I have submitted the proposal Apr 04 01:29:16 'k Apr 04 01:29:28 Could you review it once Apr 04 01:29:29 http://elinux.org/BeagleBoard/GSoC/SynchronousDataCollectionPRU Apr 04 01:30:36 let me look Apr 04 01:32:32 phase 3 and 4 seems to be quite a stretch Apr 04 01:33:03 Sigrok and UI? Apr 04 01:33:21 I wasn't planning on them to be critical Apr 04 01:33:51 Besides I assume it would be reusing Abhishek's code Apr 04 01:34:15 Could you clarify what are the issues in 3 and 4 Apr 04 01:35:11 yes Apr 04 01:35:24 I don't think u would have time to get to that part Apr 04 01:35:57 So my phase 1 and 2 timeline have been underestimated? Apr 04 01:36:30 Any improvements or suggestions? Apr 04 01:37:39 debugging and getting the timing right Apr 04 01:38:18 IMO, might rather you getting the timing as high as possible before mod'ing the UI Apr 04 01:38:39 I thought about it too Apr 04 01:38:43 in otherwords, get it working by making the rest of the code think it is the async stuff then update it later (maybe post GSoC) Apr 04 01:38:55 Ok Apr 04 01:39:04 That is fine Apr 04 01:39:16 maybe you work very fast... I am basing this on what I have seen from most people Apr 04 01:39:25 Have I got this one right? Apr 04 01:39:49 the clock signal is captured async Apr 04 01:41:28 ? Apr 04 01:45:35 Also are there any ways to keep in contact with you. Like any specific times Apr 04 01:45:46 It looks like you are very busy these days Apr 04 01:54:36 i just got back from traveling Apr 04 01:54:45 ok Apr 04 01:54:56 should be more available in the next few days once I am caught up Apr 04 01:55:22 so the clock signal is captured async Apr 04 01:55:22 so the clock signal is captured async Apr 04 01:55:22 the clk signal will be captured async right? Apr 04 01:58:42 yes Apr 04 01:58:57 you have to unless there is HW support (there isn't AFAICT) Apr 04 01:59:18 Yea Apr 04 01:59:38 so have I got the idea for the kernel drivers and PRU core right? Apr 04 02:00:14 yes Apr 04 02:00:51 Hmm regarding the finer details for sampling Apr 04 02:01:08 I get how to capture the clock signal Apr 04 02:01:49 But what bothers me is what approach I should be taking to leverage the captured clock to catch the external signal? Apr 04 02:05:41 Would the path to take depend whether I choose a rising edge trigger or a falling edge trigger? Apr 04 02:35:06 ds2? Apr 04 02:45:43 If you see this do respnd thanks **** ENDING LOGGING AT Tue Apr 04 03:00:01 2017