**** BEGIN LOGGING AT Wed Jul 13 02:59:58 2016 Jul 13 07:07:00 Wormo: there ? Jul 13 07:07:23 nerdboy: ? Jul 13 07:07:59 how can I get all levels of kernel logs in dmesg ? Jul 13 07:08:23 I tried 'echo "7" > /proc/sys/kernel/printk Jul 13 07:09:02 but that didnt work. Jul 13 07:09:11 ds2: ^ Jul 13 07:33:10 Remember facing something similar in my GSoC days but I simply logged everything as info and didn't do much with the DEBUG level. Jul 13 07:48:28 Yeah .. its really 'weird' .. Jul 13 07:48:51 printk(KERN_DEBUG "printk - debug\n"); Jul 13 07:48:52 259 pr_debug("debug\n"); Jul 13 07:48:52 260 pr_err("error - printing message\n"); Jul 13 07:48:52 261 pr_debug("debug\n"); Jul 13 07:48:52 262 printk(KERN_DEBUG "printk - debug\n"); Jul 13 07:49:17 out of this .. some are are visible in the dmesg .. while other are not Jul 13 07:50:08 even when cat /proc/sys/kernel/printk say : 7 4 1 7 Jul 13 07:50:22 dont know what does it need .. Jul 13 07:50:54 I was trying to go with the 'suggested' way .. Jul 13 07:51:14 Abhishek_: do you know where is the uBoot boot command ? I mean .. how can i edit it ? Jul 13 07:51:39 Look at /boot/uEnv.txt Jul 13 07:51:55 tried .. didnt work .. Jul 13 07:52:37 can you check by doing 'cat /proc/cmdline' to see if your modifications did make it to boot? Jul 13 07:53:17 yes .. thats how I am telling that it didnt work .. Jul 13 07:53:36 So you can't see the flags you added to the kernel command line? Jul 13 07:53:54 nope Jul 13 07:53:58 loglevel=7 Jul 13 07:54:07 thats what I want to add Jul 13 07:54:53 is the boot partition already mounted ? .. Jul 13 07:55:08 on the newer images, it's just one partition Jul 13 07:55:22 Just look in /boot/uEnv.txt Jul 13 07:59:19 adding loglevel=7 to /boot/uEnv.txt is not working Jul 13 08:02:11 oh .. i remember that grub has a command to update the settings .. dont know about uboot .. is there any such command ? Jul 13 08:02:13 Abhishek_: ^ Jul 13 08:02:51 After you add loglevel=7 into uEnv.txt, can you see it under /proc/cmdline after the system is booted? Jul 13 08:03:00 nope Jul 13 08:03:42 hmm, can you post your uEnv.txt here? Jul 13 08:05:02 Abhishek_: https://tmate.io/t/R506ULb0ngBYHj1yypE7w1fj4 Jul 13 08:05:11 Abhishek_: you there ? Jul 13 08:10:15 can you see the logs now? Jul 13 08:12:42 printk message - yes Jul 13 08:12:48 pr_debug messages - no Jul 13 08:13:13 Abhishek_: ^ Jul 13 08:15:11 but its wrong .. that loglevel thing should work when written in a new line too ;) Jul 13 08:19:17 * Abhishek_ liked tmate. Jul 13 08:20:57 thank you ... :P Jul 13 08:20:59 hahaha Jul 13 12:57:21 hey bradfa ,are you there? Jul 13 12:57:38 hi chanakya_vc Jul 13 12:58:18 bradfa, I posted quite a long mail yesterday explaining the current situation. Did you get a chance to read it? Jul 13 12:58:27 chanakya_vc: yes Jul 13 12:58:38 chanakya_vc: you are going to work on getting your logic analyzer plots Jul 13 13:00:01 bradfa, Yup I am. I am currently working on writing the driver using the SPI subsysytem Jul 13 13:00:11 Been struggling actually :P Jul 13 13:01:35 But getting the char driver to insmod correctly was such a huge relief. I am thinking of documenting the entire process so that people don't have to struggle the amount I did. Jul 13 13:03:47 chanakya_vc: why not try to get your first logic analyzer plots just using the char driver? Jul 13 13:04:55 Okay I can work on that right now bradfa ,if you would want that. I just thought I should be working on the SPI-system now.' Jul 13 13:06:31 Will you be online bradfa in the coming days? I haven't been able to reach out to mdp. Jul 13 13:06:31 chanakya_vc: I think it would be easier to debug your pru to driver interface using the char driver interface, as you can send down known bytes from userspace in linux and see them on the spi bus Jul 13 13:07:24 chanakya_vc: I will be online as best I can, until my wife goes into labor (baby will be coming some unknown time in the near future). mdp is in Japan right now for a conference, I suspect he'll be back to his "office" next week Jul 13 13:08:55 Okay bradfa , I will try to get most of the work done while you are here : ). I will start working on the plots immediately. Jul 13 13:11:18 and bradfa , just an unrelated question, I am thinking of buying a new machine, with better RAM and processor. Any suggestions? Jul 13 13:13:09 I mean which laptop are you using? Jul 13 13:44:06 Looks like I won't be able to make it to today's meeting. Jul 13 14:20:26 chanakya_vc: at work I have a dell precision 5810 workstation (http://www.dell.com/us/business/p/precision-t5810-workstation/pd) and I've just bought myself a refurbished Dell Latitude E6220 (http://www.dell.com/support/home/us/en/04/product-support/product/latitude-e6220/diagnose) Jul 13 14:20:48 chanakya_vc: I will let you know what I think of the E6220 when it arrives (FedEx says tomorrow) Jul 13 14:25:13 chanakya_vc: if you are looking for a laptop with fast processor/ram, I think there are a few models out there with quad core i7 processors Jul 13 14:25:32 chanakya_vc: but they are much more rare than dual core i7 or i5 processors in laptops, at least Jul 13 14:26:24 chanakya_vc: I'm personally a fan of Dell Latitude and Lenovo ThinkPad brands for laptops Jul 13 14:27:25 chanakya_vc: if you do buy a new Dell computer, it may be worth shopping from a 3rd party for RAM, often Dell's RAM prices are quite a bit higher than quality suppliers (like Crucial) Jul 13 14:55:01 alexhiam: there? Jul 13 14:55:08 alexhiam: lot of questions to ask!! Jul 13 15:00:21 bradfa, My hear was set on a Dell xps 13. But the hardware specs don't seem to justify the high price. I mean for 1200 Dollars, I would expect atleast 12 Gb od RAM :P Jul 13 15:01:26 *heatr Jul 13 15:01:30 *heart Jul 13 15:01:49 chanakya_vc: the current revision of the xps13 has some (or has recently had some) issues with Linux and getting proper performance out of it Jul 13 15:02:24 chanakya_vc: I believe there are fixes happening, but any of the skylake based laptops might be worth avoiding for a little longer, but you can probably find a previous generation xps13 and have a good experience Jul 13 15:02:47 Ohh bradfa, that's sad. Many suggested that for that price I could buy a Mac Jul 13 15:03:08 chanakya_vc: there are many price points and many laptops to choose from :) Jul 13 15:03:36 Yeah bradfa . But companies like ASUS offer a better deal I guess. Jul 13 15:03:41 And there are many under the Apple price point for arguably more value Jul 13 15:04:36 chanakya_vc: over the past 10 years, I've only ever bought Lenovo ThinkPads and Dell Latitudes and I've not had any problems. I can't really speak to the quality/cost/etc for other brands Jul 13 15:05:00 foreverska, I don't want to shift to mac. Companies like ASUS are giving a better deal with hardware. What do you suggest? Jul 13 15:05:37 chanakya_vc: just a point (glanced over all the reports) .. LOCALVERSION is not necessary to compile and generate the right lkm. You can get your lkm insmoded even when you dont specify LOCALVERSION, it was the compiler that was not right . Jul 13 15:05:37 [alexanderhiam] kiran4399: sorry, things have been crazy here. I'll be back to my desk by the meeting... Jul 13 15:05:51 Okay bradfa I will most probably go for Dell. Even I have only used Dell and Toshiba. Currently have a Dell Inspiron n5010 Jul 13 15:06:09 I'm a big Asus fanboy personally but I have a soft spot for Lenovo and I really have come to like my work Dell Latitude. Jul 13 15:06:24 chanakya_vc: have a look at the Dell Latitude E6xxx series laptops Jul 13 15:06:31 ZeekHuge, I could have by using the -force-vermagic option. Jul 13 15:06:50 chanakya_vc: the Latitude line is more "business" than "consumer" while the Inspiron line is much more focused on low cost and "consumer" grade Jul 13 15:07:15 chanakya_vc: the right way I mean, and --force-* is not the right way. Jul 13 15:07:32 anyway .. not much relevant though ... Jul 13 15:08:01 alexhiam: ok.. waiting for you Jul 13 15:08:03 Okay ZeekHuge I took it into my head to get the kernel compiled correctly :P Jul 13 15:08:04 but you should use the new downloaded cross-compiler and not the one that gets downloaded using apt Jul 13 15:08:19 chanakya_vc: I correct myself, apparently the E6xxx series Latitudes are no more :( The 5000 or 7000 series Latitudes are worth looking at now Jul 13 15:08:56 Okay bradfa I will look through those. I hope they ship them in India : ) Jul 13 15:09:36 chanakya_vc: http://www.dell.com/in/business/p/latitude-laptops?~ck=bt Jul 13 15:10:01 ZeekHuge, I guess you are right. It might have been because the fstack thing was not supported by my compiler. Jul 13 15:10:13 bradfa, That's great!! :D Jul 13 15:10:41 chanakya_vc: the E6440 uses 4th generation Intel Core processors, should have very good Linux support Jul 13 15:11:21 Okay so it is best to avoid 6th gen processors as of now bradfa ? Jul 13 15:12:17 chanakya_vc: last I knew, yes, 6th gen Intel Core processors (skylake) were having issues. Fixes are happening, but if you want to run an older kernel those fixes might take a while to get to you (ie: if you want to run Ubuntu 14.04, you may have issues with skylake parts) Jul 13 15:12:39 chanakya_vc: https://mjg59.dreamwidth.org/41713.html Jul 13 15:13:28 chanakya_vc: https://askubuntu.com/questions/691216/no-version-of-ubuntu-can-be-installed-with-any-skylake-6th-generation-intel-proc Jul 13 15:14:13 chanakya_vc: fixes are happening to fix these kinds of issues for skylake parts, but the 4th gen Core processors are not much slower than 6th gen processors, so my advice is to avoid skylake (6th gen) for another few months, if you can Jul 13 15:14:22 how about thinkpads ? they have a classy look and work quite well ! Jul 13 15:14:30 Ohh..So either wait for some time or go with the latitude series of dell Jul 13 15:14:35 I actually like them Jul 13 15:14:36 ZeekHuge: I'm a Thinkpad fan, too :) Jul 13 15:14:53 ZeekHuge: I've had a few different T series ThinkPads, they were all good machines Jul 13 15:15:25 bradfa: T430 .. really good one that i have Jul 13 15:15:38 ZeekHuge: my home family computer is a T530 :) Jul 13 15:15:50 with all those sim card and fingerprint reader ;) Jul 13 15:18:43 though it has some issues with the NVIDIA gpu. Jul 13 15:18:49 when using ubuntu Jul 13 15:18:57 ZeekHuge: due to the optimus switching? Jul 13 15:19:01 but, I have just switched it of Jul 13 15:19:02 yeah Jul 13 15:19:20 ubuntu 16 makes things worse Jul 13 15:19:25 * bradfa T530 does not have discrete GPU Jul 13 15:20:25 though i didnt try harder, but for 16.04 , i was not able to turn of the NVIDIA when i am not using it Jul 13 15:20:33 and therefore came back to 14 Jul 13 15:20:55 *turn off Jul 13 15:39:03 bradfa, Where the dell site's does not mention ubuntu as an operating system, it means it can't be run. Jul 13 15:39:25 Like this seems to be a nice deal:http://www.dell.com/in/p/inspiron-15-7548-laptop/pd?oc=y548502hin8&model_id=inspiron-15-7548-laptop Jul 13 15:39:45 But it has 5 gen processor. Jul 13 15:41:46 bradfa, ZeekHuge This has some smart student's offer also :P Jul 13 15:48:04 chanakya_vc: any computer can have ubuntu (or derivatives ) as its is OS. Jul 13 15:48:18 even the one with 512MBs of RAM Jul 13 15:49:06 they have probably just mentioned the os that comes pre-installed. Jul 13 15:49:30 You know ZeekHuge Abhishek_ bought a lenovo yoga laptop and he told me he couldn't run ubuntu on it as its BIOS seemed to be locked Jul 13 15:50:05 So it would be pretty sad to discover something of this sort after you spend 90k rupees on it :P Jul 13 15:50:06 *with some efforets Jul 13 15:50:09 *efforts Jul 13 15:50:26 even a Mac can have ubuntu OS Jul 13 15:51:38 But what do you think of this dell machine? Jul 13 15:52:01 chanakya_vc: I believe all implementations of UEFI firmware now allow you to disable secure boot and/or insert your own keys. Likely if people have a hard time booting Linux, it's due to having Secure Boot enabled and trying to run an unsigned bootloader. Jul 13 15:52:12 I have a different taste :) Jul 13 15:52:41 chanakya_vc: my Dell machines and my Lenovo all allow disabling secure boot in the UEFI BIOS setup screens Jul 13 15:52:58 chanakya_vc: but specs are good Jul 13 15:53:47 chanakya_vc: that seems like a nice laptop :) Jul 13 15:54:05 chanakya_vc: no Ethernet port, though, if that matters to you (it does to me) Jul 13 15:54:49 kiran4399: ok, back in the office. Any quick questions before the meeting starts? Jul 13 15:55:11 chanakya_vc: and the offers valid only till 31st .. Jul 13 15:55:27 alexhiam; yeah.. Jul 13 15:55:36 does the prussdrv for 4.4.. Jul 13 15:55:44 accept .bin as the firmware? Jul 13 15:55:49 or does it support only elf? Jul 13 15:55:59 no prussdrv, we want to use remoteproc Jul 13 15:56:32 alexhiam: oh sorry. Jul 13 15:56:39 alexhiam: i mean to say remoteproc Jul 13 15:57:16 kiran4399: remoteproc .. only elf Jul 13 15:57:42 needs the memory stuff that the elf has Jul 13 15:58:01 alexhiam: actually for servo and encoders.. the firmware is in .bin.. Jul 13 15:58:09 alexhiam: so wondering how to load it.. Jul 13 15:58:23 just need to change the compile step to get an elf instead Jul 13 15:58:42 though that current firmware is for prussdrv, so it's gonna take some rewriting Jul 13 15:58:49 yes .. so as to be able to place data in different memory sections Jul 13 15:59:17 bradfa, Yes I did not notice that. But there are rj45 to usb converters available I guess. I have never used them though. Jul 13 15:59:19 and I see no reason not to totally rewrite it in C, they're both pretty basic pieces of firmware Jul 13 16:01:20 servos only need a 50Hz signal, so there's not a need for assembly Jul 13 16:01:58 for the encoder firmware (which I believe we decided would be a stretch goal) the existing assembly could just be used inline as well Jul 13 16:03:08 alexhiam: stretch goal? Jul 13 16:03:12 for encoders? Jul 13 16:03:36 for the 4th encoder input Jul 13 16:03:47 the other 3 use eqep Jul 13 16:03:51 alexhiam; oh.. Jul 13 16:03:52 meeting today? Jul 13 16:03:55 yeah.. remember Jul 13 16:04:03 <_av500_> yes Jul 13 16:04:06 <_av500_> in min Jul 13 16:04:09 <_av500_> in 1 min Jul 13 16:05:43 alexhiam: do you have some example piece of code for pru firmware which uses remoteproc? Jul 13 16:05:55 kiran4399_: the other PRU projects ;) Jul 13 16:06:11 alexhiam; ok.. I'll ask them ;) Jul 13 16:06:19 chanakya_vc and ZeekHuge are both working with it a lot Jul 13 16:06:25 lots of examples in latest pru support package... Jul 13 16:06:52 <_av500_> ok Jul 13 16:06:55 <_av500_> 0) WELCOME Jul 13 16:07:00 <_av500_> hello everybody Jul 13 16:07:01 * nerdboy still massaging libadancurses to build multilib correctly... Jul 13 16:07:21 hello _av500_ Jul 13 16:07:25 hello Jul 13 16:07:25 hi all Jul 13 16:07:54 <_av500_> lets do a count Jul 13 16:07:55 <_av500_> 3 Jul 13 16:08:06 <_av500_> kiran4399_ is 4 Jul 13 16:08:12 4 Jul 13 16:08:15 hi all Jul 13 16:09:20 ZeehHuge was just here... Jul 13 16:09:26 ZeekHuge that is Jul 13 16:09:44 5 Jul 13 16:10:00 chanakya_vc, Visaoni: wake up! Jul 13 16:10:33 * _av500_ rings the mission bell Jul 13 16:10:39 Namastey everyone Jul 13 16:10:41 Sorry Jul 13 16:10:42 <_av500_> 6 Jul 13 16:11:22 <_av500_> no Visaoni ? Jul 13 16:12:01 all "pokes" are equal on irc... Jul 13 16:12:40 last sighting was last night: charlie5: I'll give it a shot after I get sleep. might need a reminder tomorrow though... close to waking up with keymarks on my face Jul 13 16:12:53 <_av500_> ic Jul 13 16:13:12 <_av500_> ok Jul 13 16:13:15 <_av500_> lwts begin Jul 13 16:13:17 <_av500_> 1 Jul 13 16:13:22 <_av500_> arhgh Jul 13 16:13:24 <_av500_> bad laptop Jul 13 16:13:36 <_av500_> 1) general issues Jul 13 16:13:45 <_av500_> do we still have hw issues? Jul 13 16:13:57 only sw;-) Jul 13 16:14:12 <_av500_> thats ok :) Jul 13 16:14:14 waiting for shipment to ^^ no word yet Jul 13 16:14:27 <_av500_> nerdboy: so you shipped it out? Jul 13 16:14:46 jason was sending to wormo i think Jul 13 16:15:11 * nerdboy not in the distribution chain this time Jul 13 16:15:56 <_av500_> ok Jul 13 16:16:12 <_av500_> but some student had a non working board? Jul 13 16:16:15 <_av500_> who was that? Jul 13 16:16:19 <_av500_> Visaoni? Jul 13 16:16:23 yup Jul 13 16:16:26 <_av500_> ok Jul 13 16:16:34 <_av500_> so its being shipped? Jul 13 16:16:52 jason asked about address, that was last i saw Jul 13 16:16:56 <_av500_> ok Jul 13 16:17:08 <_av500_> assuming I wont have to ask again next week :) Jul 13 16:17:10 <_av500_> lets go on Jul 13 16:17:27 <_av500_> any other urgent issues that have no been covered in reports or with your mentors? Jul 13 16:17:52 nope Jul 13 16:19:08 <_av500_> ok Jul 13 16:19:14 <_av500_> evervbody on track then Jul 13 16:19:42 <_av500_> then we can start with reports earyl Jul 13 16:19:44 <_av500_> early Jul 13 16:20:36 <_av500_> who was on for this week? Jul 13 16:21:05 me Jul 13 16:21:08 keyface i think Jul 13 16:21:09 <_av500_> ok Jul 13 16:21:13 <_av500_> then go on Jul 13 16:21:18 okay .. Jul 13 16:21:29 Stand up report : Jul 13 16:21:33 Hi everyone Jul 13 16:22:01 So my project, Beaglescope can be broken down broadly into 4 parts : Jul 13 16:22:32 firmware .. parallel_bus_driver .... parallel subsystem ..... ADC iio device driver Jul 13 16:23:06 So we are aiming to get highest possible sampling rate , and yet make it configurable Jul 13 16:23:35 this will be done by using 2 different functions that are defined using PRU assembly Jul 13 16:24:33 one of them will be aiming to get the highest sampling rate Jul 13 16:24:33 probably 40MHz Jul 13 16:24:33 and the other will be configurable in nature , but for lower sampling rates .. Jul 13 16:24:48 I am almost done with that 'configurable sampling fw' Jul 13 16:24:59 it can sample at about 28Mhz Jul 13 16:25:10 <_av500_> nice Jul 13 16:25:29 Though some clock duty cycle issues remain... but that will be soon resolves Jul 13 16:25:35 *resolved Jul 13 16:26:05 I theory ... the changes that i will be doing in the firmware in next few says Jul 13 16:26:36 will allow to have any kind of duty cycle ... and any sampling frequency .. Jul 13 16:26:49 even as low as 1Hz Jul 13 16:27:02 and yet configurable from the userspace Jul 13 16:27:26 okay so there are some issues regarding high sampling frequency tooo .. Jul 13 16:27:43 its related to the virtio bus Jul 13 16:28:16 basically .... we have only 512 bytes of buffer and each buffer needs one kick (interrupt to ARM) to send it to the ARM Jul 13 16:28:49 at higher sampling frequency ... this may result in freezing up the ARM .. or slowing it down Jul 13 16:29:04 more on this later ... Jul 13 16:29:04 <_av500_> cant the pru see SDRAM? Jul 13 16:29:25 it can ... infact .. that is how the PRU writes to the ARM Jul 13 16:30:05 but ... the virtio implementation .. that is virtio_rpmsg_bus allows it to write at max 512bytes of data Jul 13 16:30:13 and then .. an interrupt to ARM Jul 13 16:30:21 to tell it that the data has arrived Jul 13 16:30:27 <_av500_> right Jul 13 16:30:38 thats the standard implementation.. Jul 13 16:30:41 <_av500_> so you have to use vurtio for capture data? Jul 13 16:31:02 <_av500_> why not set aside some memory and write directly to it? Jul 13 16:31:03 virtio -- to transfer data from pru to arm .. Jul 13 16:31:19 well .. Jul 13 16:31:28 the 'STANDARD approach" Jul 13 16:31:56 is to use rpmsg ... rather than hacking it to write on its own Jul 13 16:32:02 later on this .. Jul 13 16:32:02 <_av500_> ok Jul 13 16:32:52 For now ... I am going to leave the fw ... only will improve what we have ... Jul 13 16:33:04 and later develop that 'MAX sampling firmware' Jul 13 16:33:16 so now comes the linux side of the project Jul 13 16:33:53 So i will first work on to make a single lkm ... that will talk to the PRUs and will register itself to IIO Jul 13 16:34:13 actually ... I already have this .. and its working ... Jul 13 16:34:23 but with limited functionality .. Jul 13 16:34:36 that is ... it just reads a single value from the PRUs Jul 13 16:35:12 I havent pushed it up... as soon as i got it working ... I took some sleep and need to do some cleaning stuff and will push it Jul 13 16:35:20 <_av500_> ok Jul 13 16:35:37 <_av500_> sounds good Jul 13 16:35:50 * nerdboy guesses the answer is probably dma/interrupt/no-rpmsg Jul 13 16:35:50 So the plan is .. as i said to get this complete lkm ready in the next week Jul 13 16:36:12 yeah .. so whats for that issue .. Jul 13 16:36:35 Now .. as far as i have understood the working of virtio is : Jul 13 16:36:59 virtio framework allows to make virtual devices Jul 13 16:37:13 please correct me if i am wrong at any point Jul 13 16:37:23 so .. virtio_rpmsg_bus Jul 13 16:37:46 acts like it is a bus ... and the rpmsg_channel data structure acts like it is a device Jul 13 16:37:56 ah .. logical device Jul 13 16:38:24 so .... a client driver uses this rpmsg_channel to talk to the prus Jul 13 16:38:50 there is a client driver in my repo, that i have written .. and this ^ is how it works Jul 13 16:39:30 but ... the parallel_bus we have been talking about ... should be actually a virtual bus ... just like virtion_rpmsg_bus us Jul 13 16:39:31 *is Jul 13 16:40:21 so ... I think that for that ... we can use the code from virtio_rpmsg_bus and use it to develop the parallel_bus driver Jul 13 16:40:51 the devices .. would then fit onto this parallel_bus ... and we can probably change the size of the buffer too . Jul 13 16:40:59 <_av500_> ZeekHuge: we have to postpone that :) Jul 13 16:41:08 <_av500_> thanks for the report Jul 13 16:41:16 infcat there is a scope of having configurable buffers Jul 13 16:41:20 okay. Jul 13 16:41:39 <_av500_> thanks again Jul 13 16:41:43 :) Jul 13 16:41:47 <_av500_> who was also for this week? Jul 13 16:42:15 http://git.ti.com/pru-software-support-package/pru-software-support-package/trees/master/examples/am335x/PRU_edmaConfig Jul 13 16:42:42 just this link : http://lxr.free-electrons.com/source/drivers/rpmsg/virtio_rpmsg_bus.c#L108 Jul 13 16:42:54 <_av500_> nerdboy: after the meeting :) Jul 13 16:43:01 <_av500_> who was next? Jul 13 16:43:35 Visaoni: report? Jul 13 16:44:09 <_av500_> im sure we had two volunteers last week Jul 13 16:44:19 <_av500_> nobody remember, so long ago :) Jul 13 16:44:25 yeah, wasn't it him? Jul 13 16:44:41 <_av500_> nerdboy: oh ok Jul 13 16:44:54 <_av500_> well, anybode else for a quick report? Jul 13 16:46:12 <_av500_> who is for next week? Jul 13 16:46:31 <_av500_> amr_ragaey: chanakya_vc pmezydlo kiran4399_ henrix_ ? Jul 13 16:46:38 <_av500_> who did not do round two? Jul 13 16:47:03 me.. Jul 13 16:47:06 <_av500_> ok Jul 13 16:47:09 kind of hiding here :( Jul 13 16:47:12 :-P Jul 13 16:47:18 <_av500_> you are on next wek Jul 13 16:47:21 I did it the first day itself av500 Jul 13 16:47:25 <_av500_> ok Jul 13 16:47:31 <_av500_> pmezydlo: ? Jul 13 16:47:34 ok Jul 13 16:47:37 no problem Jul 13 16:47:38 <_av500_> ok Jul 13 16:47:43 <_av500_> next week pmezydlo and kiran4399_ Jul 13 16:47:54 oops, I gotta run Jul 13 16:48:01 <_av500_> ok Jul 13 16:48:02 <_av500_> thanks Jul 13 16:48:07 I thought pmezydlo did round 2 already Jul 13 16:48:07 * alexhiam will be back in a bit Jul 13 16:48:11 <_av500_> hmm Jul 13 16:48:16 <_av500_> I will check logs Jul 13 16:48:23 <_av500_> and point somebody out by mail :) Jul 13 16:48:35 <_av500_> cant right now, not at real computer Jul 13 16:48:55 av500, amr_ragaey also did it the first day with me i guess Jul 13 16:48:59 <_av500_> ok Jul 13 16:49:03 * Wormo hates browsers in imaginary computers... Jul 13 16:49:13 <_av500_> as said, I'll check Jul 13 16:49:19 _av500_: using virtio for a virtual computer ? :P Jul 13 16:49:19 <_av500_> than you all Jul 13 16:49:21 yes I did it with you chanakya_vc Jul 13 16:49:24 <_av500_> ok Jul 13 16:49:31 Because alexhiam chose us alphabetically I remember that :P Jul 13 16:49:39 <_av500_> you get 10 minutes early this week Jul 13 16:49:46 <_av500_> use them wisely Jul 13 16:49:51 :) Jul 13 16:49:54 I was in last week Jul 13 16:49:58 <_av500_> ok Jul 13 16:49:59 : ) Jul 13 16:50:43 <_av500_> see you all next week Jul 13 16:50:45 what is an imaginary computer ? actually ? mobile phone ? Jul 13 16:50:56 thanks _av500_ Jul 13 16:51:08 can't be that, phones are real pocket computers ;) Jul 13 16:51:10 Well a smartphone is a kind of a computer Jul 13 16:51:20 True that Wormo :P Jul 13 16:51:34 then what ? Jul 13 16:51:42 <_av500_> my kids netbook Jul 13 16:51:49 <_av500_> .) Jul 13 16:51:54 <_av500_> bbl Jul 13 16:52:19 What would you mean by a computer that is imaginary? Something that is in development perhaps:P Jul 13 16:52:51 thanks _av500_ Jul 13 16:53:19 or have an 'i' (iota) before it name :P Jul 13 16:53:34 hahah ZeekHuge :D Jul 13 16:53:48 Maybe the i series by apple? :P Jul 13 16:54:02 heh yes Jul 13 16:54:25 i've still got the dsp firmware in connection with the soundcard driver. can anybody explain me how to get memory areas of device drivers in runtime to check overlapping? Jul 13 16:55:04 henrix_ you need to focus on finding what physical memory range your dsp wants to use Jul 13 16:55:16 then you make a "hole" as you mentioned in report Jul 13 16:55:54 ok, can you give me some hint about finding the physical memory range in runtime? Jul 13 16:56:02 or can i get it from device tree? Jul 13 16:56:38 yob, hat es Jul 13 16:56:55 iomem ? Jul 13 16:56:56 ups, wrong channel;-) Jul 13 16:56:59 yes Jul 13 16:57:08 cat /proc/iomem Jul 13 16:57:17 ok, i checked that already Jul 13 16:57:19 not sure .. Wormo would know. Jul 13 16:57:27 there is only the same information i get from device tree Jul 13 16:57:32 there used to be some debug tool from TI that might do it, but other than that it's checking the source for creating dsp file Jul 13 16:57:56 ok Jul 13 16:58:01 do you have that yet, the source? Jul 13 16:58:20 not of dsp firmware. of remoteproc, yes Jul 13 16:58:22 ds2: hi ! there ? Jul 13 16:59:05 dsp firmware source you need, it might ship some binary object files but the wrapper part will deal with resources such as memory Jul 13 16:59:29 ok, then I will search the TI stuff;-) Jul 13 17:00:01 what's the name of the dsp pkg you use? Jul 13 17:00:13 I'll see if I can find it quickly Jul 13 17:00:20 dra7-dsp1-fw.xe66 Jul 13 17:00:24 dra7-dsp2-fw.xe66 Jul 13 17:00:30 thanks Jul 13 17:01:02 what .deb file does it belong to, or do you install it separately (if so where does it get downloaded from) Jul 13 17:01:56 it is included in image Jul 13 17:02:15 https://debian.beagleboard.org/images/bbx15-debian-8.4-lxqt-4gb-armhf-2016-05-13-4gb.img.xz Jul 13 17:02:30 Ok Jul 13 17:30:31 laters all Jul 13 17:37:27 Wormo, anything found? Jul 13 17:39:27 henrix_: I don't have a beagleboard x15, would you run 'dpkg -S dra7-dsp1-fw.xe66' for me since I'm not seeing the packages I expected in RCN debian repo Jul 13 17:40:04 ZeekHuge, still about? Jul 13 17:41:05 Meh, I am sure you will read logs later. Jul 13 17:41:46 Had very busy few weeks so not been keeping up as well as I would like. Jul 13 17:41:59 Couple of really quick points. Jul 13 17:42:06 jic23b: Hi ! Jul 13 17:42:14 yes please Jul 13 17:42:25 1. Nothing stops a bus master sitting on another bus. Jul 13 17:42:53 Hence your parallel bus master can be on the rpmsg buffer. Jul 13 17:42:56 Se Jul 13 17:43:00 Cond Jul 13 17:43:03 Ly Jul 13 17:43:56 2. That virtio driver is full of notes that they want to handle different sized buffers. Jul 13 17:44:12 yes .. ? Jul 13 17:44:28 I would initially just patch it to have fewer larger buffers. Jul 13 17:44:53 yeah .. that is true .. and they have even said that we can just change that number and everything would work fine. Jul 13 17:44:55 okay Jul 13 17:45:08 Say 16 buffers of 512 x32 bytes. Jul 13 17:45:11 henrix_: wait I might have overlooked it... Jul 13 17:45:15 ah Jul 13 17:45:16 firmware-am57xx-opencl-monitor: /lib/firmware/dra7-dsp1-fw.xe66 Jul 13 17:45:22 ok Jul 13 17:46:09 Later can work out how to do buffers provided elsewhere or variable sized buffers. Jul 13 17:46:32 Might ultimately come all the way from userspace. Jul 13 17:47:24 Keep those changes as small and simple as possible for now. Jul 13 17:47:44 okay. Jul 13 17:49:25 3. Don't worry to much at this point about parallel bus subsystem stuff... get a monolithic module up as that will act as an example of what is actually being supported. May make sense to spend remaining time in gsoc on other things. Jul 13 17:49:47 Can always do that bit afterwards... Jul 13 17:49:53 yep . Jul 13 17:51:12 Otherwise sounds like standard engineering fun. Always takes twice as long as you expect :). Jul 13 17:53:04 jic23b: it should ? or is that I am doing it slow ? Jul 13 17:53:37 Nah more time was always needed than schedule allowed! Jul 13 17:53:59 Also perfectly normal real world engineering :) Jul 13 17:54:28 Far as can tell you are getting along fine. Jul 13 17:55:23 okay :) Jul 13 17:55:33 Classic Linux kernel issues of discovering stuff you thought would just work is half written... Jul 13 17:56:54 Kind of how I ended up with a subsystem :) Jul 13 17:57:23 anyhow out of time. Keep up the good work. Bye for now. Jul 13 17:57:52 jic23b: thank you . Jul 13 17:59:26 You are welcome. Bye Jul 13 18:01:41 henrix_: as far as I can tell the firmware is not actually being built in debian but rather using TI's layer for openembedded build system Jul 13 18:01:49 https://github.com/rcn-ee/repos/tree/master/ti-firmware-am57xx-opencl-monitor Jul 13 18:02:09 ok, so there if find the wrapper code? Jul 13 18:02:11 no source package in the debian pool Jul 13 18:02:15 http://repos.rcn-ee.net/debian/pool/main/f/firmware-am57xx-opencl-monitor/ Jul 13 18:03:07 ahh, ok, so it's not open sourced Jul 13 18:03:09 hmm Jul 13 18:03:29 no I believe it is Jul 13 18:03:55 ok, I see there's a recipe for bitbake Jul 13 18:04:00 just no convenient way to build on debian, so RCN used the officially supported method with bitbake Jul 13 18:04:07 ok Jul 13 18:04:25 then I'll look for the wrapper code in the ti repos Jul 13 18:04:49 thanks a lot! Jul 13 18:04:58 http://downloads.ti.com/mctools/esd/docs/opencl/memory/ddr-partition.html#am57 Jul 13 18:05:24 That might be the answer right there, hopefully it matches with what got built by bitbake... Jul 13 18:05:42 oh nice Jul 13 18:06:05 so look for that in your devicetree Jul 13 18:06:09 argh, looked so often in this document at the beginning of gsoc-.- Jul 13 18:06:10 ok Jul 13 18:06:17 thanks Jul 13 18:06:55 yw, have fun Jul 13 18:07:01 sure;-) Jul 13 21:35:09 alexhiam, There? Jul 13 21:36:46 m_w, There? I am loading the pru firmware but looking at the dmesg I am confused whether is it actually loading or not. Can you help me out if free? Jul 13 21:37:46 send a pastebin of the commands ran and the dmesg output Jul 13 21:38:52 Okay m_w . Jul 13 21:44:32 http://pastebin.com/kpXHvDYN Jul 13 21:44:34 m_w, Jul 13 21:44:59 I am only loading the stuff to pru 0. For that I just unload and then load pru_rproc Jul 13 21:45:26 The firmware is already there in /lib/firmware/am335x-pru0-fw Jul 13 21:46:28 lemme take a look Jul 13 22:04:10 this line looks bad Jul 13 22:04:12 pru-rproc: probe of 4a334000.pru0 failed with error -12 Jul 13 22:06:38 -ENOMEM Jul 13 22:22:25 hmmn what do you think is happening m_w , the firmware not being loaded correctly? Jul 13 22:23:10 Or not being loaded at all. This is weird because it was working couple of weeks back. Jul 13 22:24:36 https://e2e.ti.com/support/arm/sitara_arm/f/791/t/368069 Jul 13 22:25:44 resource_table missing? Jul 13 22:26:06 I don't think so m_w . Because it wouldn't compile otherwise. Jul 13 22:26:13 Let me check again though Jul 13 22:38:30 m_w, getting this now: Jul 13 22:38:31 booting the PRU core manually Jul 13 22:38:32 [ 1583.683670] remoteproc1: powering up 4a334000.pru0 Jul 13 22:38:32 [ 1583.683962] remoteproc1: Booting fw image am335x-pru0-fw, size 33172 Jul 13 22:38:32 [ 1583.684043] remoteproc1: remote processor 4a334000.pru0 is now up Jul 13 22:38:45 I think it is resolved now. Jul 13 22:39:13 better Jul 13 22:56:25