**** BEGIN LOGGING AT Sun Jan 01 03:00:00 2017 Jan 01 03:04:09 hildon-home acting up with a borked desktop widget is not exactly unheard of Jan 01 03:11:25 Well, I do not have widgets. Just lots of shortcuts. Used to have bookmarks, but removed them, just in case they overwhelm hildon-home. Jan 01 09:02:21 Thanks. I have a lot to learn. "The spirit is willing but the flesh is weak". Jan 01 09:17:07 allah is doing Jan 01 17:34:26 DocScrutinizer05: in N900 RX-51 Full Schemantics v 1.0 are missing gpios: eci0 (61), eci1 (62), eci-sw1 (178) Jan 01 17:34:39 do you have some information how are connected? Jan 01 17:34:53 err, they are not missing? Jan 01 17:35:48 http://wstaw.org/m/2017/01/01/plasma-desktopa17764.png http://wstaw.org/m/2017/01/01/plasma-desktopn17764.png Jan 01 17:36:20 tvout_en is 40 Jan 01 17:36:47 and eci-sw2 is 182 Jan 01 17:37:21 but there are no gpio 61 or gpio 178 Jan 01 17:37:37 sorry I don't get it Jan 01 17:38:01 where is connected gpio 61? Jan 01 17:38:36 I do not see it Jan 01 17:38:56 eci0: N4007 -> ECI0 -> OMAP:U3 GPMC_nBE1 Jan 01 17:40:08 eci1: N4008 -> ECI1 -> OMAP:H1 GPMC_nWP Jan 01 17:40:46 I don't know what's eci_sw1 Jan 01 17:41:29 eci_sw1 --> #define RX51_ECI_SWITCH_1_GPIO 178 Jan 01 17:42:21 gpio_request(RX51_ECI_SWITCH_1_GPIO, "ECI switch 1"); gpio_direction_output(RX51_ECI_SWITCH_1_GPIO, 0); Jan 01 17:42:58 I don't know. In schematics I only could guess that this is ECI5 which is connected to OMAP:V3 McSPI2_CS1 and (possibly incorrectly) labeled "GPIO_182" Jan 01 17:43:16 #define RX51_ECI_SWITCH_2_GPIO 182 Jan 01 17:43:32 gpio 182 is eci_sw2, not eci_sw1 Jan 01 17:43:49 gpio_request(RX51_ECI_SWITCH_2_GPIO, "ECI switch 2"); gpio_direction_output(RX51_ECI_SWITCH_2_GPIO, 1); Jan 01 17:43:49 I don't know what's eci_sw1 then Jan 01 17:44:32 no faintest clue Jan 01 17:45:05 in snd rx51.c is: static void rx51_set_eci_switches(int mode) { mode == 3: /* ECI RX/TX connected to mic/bias line */ gpio_set_value(RX51_ECI_SWITCH_1_GPIO, 1); Jan 01 17:45:10 haven't seen anything in schematics that might correspond to a third ECI switch apart from TV_EN and what you call eci_sw2 Jan 01 17:45:50 sorry I can't analyze schematics from sourcecode Jan 01 17:46:02 give me a pin Jan 01 17:46:17 there are two gpio switches in kernel code which looks like set how is mic/bias lines connected Jan 01 17:46:39 and I was not able to find that eci_sw1 in schemantics... Jan 01 17:47:46 there is TV_EN and an eci switch with no name Jan 01 17:47:55 no other but those two switches Jan 01 17:48:07 tv_en is different Jan 01 17:48:27 in schematics it's not Jan 01 17:48:44 yes, I do not see it too Jan 01 17:48:58 reason why I'm asking... Jan 01 17:49:29 kernel snd code is setting that gpio and I cannot understand what it is doing... Jan 01 17:50:17 TVOUT_EN = N4022:A1 1SEL, ECI5 = N4022:A3 2SEL Jan 01 17:50:40 no other swites in whole audio which might be faintly related to ECI Jan 01 17:51:37 ECI "bus" consists of ECI0 ECI1 and ECI5, 2,3,4 are nonexistent Jan 01 17:53:18 I *guess* there are fuzzy bits in kernel sourcecode comment bits and/or naming of vars Jan 01 17:54:44 NB there's also only eci0 and eci1 in /sys/ Jan 01 17:55:35 /* ECI RX/TX connected to mic/bias line */ sounds pretty bogus Jan 01 17:56:04 there's no such thing like "ECI RX/TX" in N900 Jan 01 17:56:30 RX ok maybe, but for sure no TC Jan 01 17:56:36 TX Jan 01 17:57:20 DocScrutinizer05: run on n900: grep ECI /sys/kernel/debug/gpio Jan 01 17:57:22 lemme put it bold: I think the sourcecode is shit Jan 01 17:57:27 there is ECI switch 1 Jan 01 17:57:55 DUH!! Jan 01 17:58:59 so can't you find out which pin number GPIO178 is supposed to be on? Jan 01 17:59:46 I don't know if I can find the right table in those 3 * 3000 manual pages Jan 01 18:00:16 you're an EE ... you're supposed to :p Jan 01 18:00:16 grep -i -E 'tv|eci' /sys/kernel/debug/gpio Jan 01 18:00:32 bencoh: too little RAM Jan 01 18:00:43 this shows: tvout_sel, eci0, eci1, ecisw1, ecisw2 Jan 01 18:00:52 DocScrutinizer05: in brain or in computer? Jan 01 18:00:56 (or both?) Jan 01 18:00:58 latter Jan 01 18:01:03 huhu Jan 01 18:01:21 tvout_sel, eci0, eci1 and ecisw2 are already identified, just that ecisw1 is unknown... Jan 01 18:01:34 maybe it was just in some n900 prototype? Jan 01 18:01:50 and nokia forgot to remove dead code? Jan 01 18:03:58 GPIO_178 aka mcspi2_clk Jan 01 18:05:07 aka hsusb2_data7 Jan 01 18:06:45 aka ball AA3 Jan 01 18:08:20 neither of all that shows up in schematics Jan 01 18:11:56 http://wstaw.org/m/2017/01/01/plasma-desktopp17764.png Jan 01 18:15:15 GPIO_177 is jack insert detection (HEADPH_IND) and according to that plasma-desktopp17764.png connected to mcspi1_cs3 Jan 01 18:15:30 and in rx51 schemantics HEADPH_IND is connected to mcspi1_cs3 Jan 01 18:15:44 so seems that plasma-desktopp17764.png is correct for n900 Jan 01 18:16:05 GPIO_178 is connected to mcspi2_clk Jan 01 18:16:14 yes, checked similar plausibility tests Jan 01 18:16:51 and I searched schematics for all mcspi and found no mcspi2_clk Jan 01 18:17:32 I searched for all AA3 and even W7, I searched for hsusb and found no hsusb2 Jan 01 18:18:48 so I dare to say the schematics have no GPIO_178 Jan 01 18:18:48 maybe it is missing in rx51 schemantics? Jan 01 18:18:58 how could I tell? Jan 01 18:19:06 or was it just for n900 prototype? Jan 01 18:19:16 and later removed? Jan 01 18:19:19 seems possible for me Jan 01 18:19:38 there *is* stuff missing in RX51 schem, at very least all about UART3 aka debug serial on testpoints Jan 01 22:37:50 hi **** ENDING LOGGING AT Mon Jan 02 03:00:00 2017