**** BEGIN LOGGING AT Sun Mar 12 03:00:01 2017 Mar 12 03:06:28 zmatt: i'm hoping to get away with 4 layers Mar 12 03:06:36 but i realize that might be crazy talk Mar 12 03:06:44 why Mar 12 03:06:59 meh just going off the datasheet Mar 12 03:07:32 when you have a ground layer with a double sided board its hard to design it for start point Mar 12 03:07:46 my first design had interleaving ground and digital ground planes Mar 12 03:08:21 but i had digital and analog components on both sides Mar 12 03:09:17 i dont know if thats normal or not Mar 12 03:09:36 are you designing a custom beagle? Mar 12 05:07:23 diycookies i'm not designing a custom beagle Mar 12 05:07:30 i don't know if i'm gooing to run linux, i don't want to Mar 12 05:08:24 but i might have to Mar 12 05:20:18 I know the feeling Mar 12 05:21:39 though, apparently my oven contains an am335x with baremetal software that uses the gpu interestingly, even though TI does not offer any way to do so Mar 12 05:22:50 both StarterWare and PowerVR are mentioned in their listing of open source components used (bsd/mit, no gpl) Mar 12 06:01:56 Hi Mar 12 06:02:40 how to read i2c device register data in linux?? Mar 12 06:02:46 which function to use? Mar 12 06:02:51 any idea? Mar 12 06:22:37 lol Mar 12 07:32:21 hey whats up Mar 12 17:24:04 ugh zmatt, the gpu- excited to dive into it, but i think you know, getting support from TI... well I might have to call them Mar 12 17:24:18 if it's not available online, I'd sign an NDA tbh Mar 12 17:25:57 ayjay_t: well I got the latest 1.14 drivers working on debian stretch Mar 12 17:26:02 without support ;) Mar 12 17:26:32 still need to try wayland, but kmscube and Qt5 eglfs_kms are working Mar 12 17:28:25 in theory what's available suffices to get the gpu working in a baremetal environment... if you're very masochistic that is Mar 12 17:29:24 it just requires porting/rewriting the (open source) kernel component and whatever the shared libs need to work Mar 12 17:30:16 imgtec made it relatively easy since the same calls are used in userspace as in kernel space Mar 12 17:35:47 that's good to note Mar 12 17:36:11 i'm actually applying to ycomb on a startup idea Mar 12 17:36:26 i have two clients who i need to wrap up before i do that and if i can get some funding i'm going all in on the product Mar 12 22:33:55 Hi all, where can I find the latest and greatest kernal for beaglebone black? Mar 12 22:37:43 KK_: just check with apt Mar 12 22:39:47 there are a few options, such as 4.9.13-ti-r24 (latest -ti kernel), 4.9.14-bone4 (latest LTS -bone kernel), 4.11-rc1-bone0 (latest -bone kernel) Mar 12 22:40:47 and rt versions of course Mar 12 22:42:32 zmatt: I think you were right about the mpu clock Mar 12 22:43:14 zmatt: the problem is that the mpu clock might not gated, so the PRCM doesn't generate the interrupt to the M3 Mar 12 22:43:35 zmatt: currently browsing the starterware sources to find out what I'm missing Mar 12 22:48:12 zmatt: interface clock comes to mind, that's the only other part in the mpu clock domain Mar 12 22:48:32 zmatt: that's a DPLL output, I think Mar 12 22:49:55 hmmwhat? Mar 12 22:50:09 what did I say about the mpu clock? Mar 12 22:50:48 the mpu PLL is embedded in mpuss Mar 12 22:52:57 I have no idea what you're saying, sorry :) Mar 12 23:04:48 zmatt: hm, I thought you said that maybe the mpu clock is not gated and therefore the clock domain isn't idling Mar 12 23:06:50 zmatt: I'm musing about step #7 in the sleep sequence described in the TRM Mar 12 23:07:01 chapter 8.1.4.6.2 Mar 12 23:07:41 says there: After MPU power domain is clock gated PRCM will provide an interrupt to Cortex M3 (using INTR1 shown in the block diagram) Mar 12 23:08:15 I reason, since this interrupt is not given, probably the MPU power domain is not clock gated. Whatever that means Mar 13 00:25:19 under what circumstances are not getting an m3 interrupt? my observation was that it fired under more circumstances than I anticipated Mar 13 00:25:48 I didn't see you saying anything along those lines either, except for when it was accidently masked in the m3 **** ENDING LOGGING AT Mon Mar 13 03:00:01 2017