**** BEGIN LOGGING AT Thu Nov 13 03:00:00 2014 Nov 13 23:28:16 hi lumag_ Nov 13 23:30:22 pls tell me how to adapt http://lists.infradead.org/pipermail/linux-arm-kernel/2010-March/012183.html Nov 14 00:01:21 You mean to my patches? Nov 14 00:01:48 In drivers/mfd/locomo.c there is pair of functions locomo_suspend/locomo_resume. Nov 14 00:01:57 hello Nov 14 00:05:19 ant__, ^^ Nov 14 00:05:54 ok but the driver differs Nov 14 00:06:17 Yes. Nov 14 00:06:53 the proper line would be regmap_write(lchip->regmap, LOCMO_DAC, LOCOMO_DAC_SCLOEB | LOCOMO_DAC_SDAOEB); Nov 14 00:08:30 I'm already scared by the change writel -> writew Nov 14 00:10:36 That is a crazy bit. Nov 14 00:10:49 Usually, *writel beans write long (32bit). Nov 14 00:10:59 However locomo_writel always wrote 16 bits. Nov 14 00:11:10 I tried to follow that playing with mtd but got lost ;) Nov 14 00:11:12 Because locomo_writel was defined as writew Nov 14 00:11:32 It is important with registers. Nov 14 00:12:02 ah..locomo_writel was writew... Nov 14 00:12:18 Because if one uses incorrect width of a register write, either some bits will get lost, or another register can be overwritten. Nov 14 00:12:20 Yes. Nov 14 00:12:52 If you are adapting to pre-regmap version of patches, it would be Nov 14 00:13:40 writew(LOCOMO_DAC_SCLOEB | LOCOMO_DAC_SDAOEB, lchip->base + LOCOMO_DAC); Nov 14 00:14:28 ok, I'll add to 3.18-rc4 Nov 14 00:14:33 :) Nov 14 00:14:46 I've built it for collie right now Nov 14 00:15:02 with all 16 patches + gpio-charger Nov 14 00:17:06 ok. Nov 14 00:17:13 Would be great to test. Nov 14 00:17:56 I'm still puzzled about side-effects of gpio-keys. Nov 14 00:19:18 I hope I will have time to look on that tomorrow. Nov 14 00:21:04 logging boot right now Nov 14 00:21:21 there is an SD inside... Nov 14 00:21:24 Also few patches for pxa are pending to be done and submitted. Before Robert merges clock patches Nov 14 00:21:30 argh.. Nov 14 00:22:31 ? Nov 14 00:22:43 http://pastebin.com/xk3kYrCG Nov 14 00:24:44 is at detection Nov 14 00:25:26 I'll take it out for the boot Nov 14 00:26:33 looks better http://pastebin.com/tbu1KF1x Nov 14 00:26:58 is vanilla, not yocto (just the defconfig) Nov 14 00:27:16 Yes, one of those -38 cards :( Nov 14 00:27:32 Does that card work with earlier 2.6 kernels? Nov 14 00:28:02 I'll sort it out tomorrow Nov 14 00:28:19 the same card is ok on poodle Nov 14 00:28:42 ah...lcd is white Nov 14 00:28:47 poodle uses pxa mmc controller (=> native mode). Nov 14 00:28:59 sa1100-no-cpufreq is missing Nov 14 00:31:48 root@collie:~# mmc0: new SD card on SPI Nov 14 00:31:48 mmcblk0: mmc0:0000 SD01G 968 MiB Nov 14 00:31:48 mmcblk0: p1 Nov 14 00:32:04 then error -38 Nov 14 00:32:25 root@collie:~# mmc0: SPI card removed Nov 14 00:32:28 on eject Nov 14 00:32:50 I'll try more cards Nov 14 00:34:44 I have a bunch of sd cards with different symptoms. I will try to check what is the problem, if I have time Nov 14 00:35:03 dunno if lcd or fb is borked here... Nov 14 00:36:20 probably missing cpu clock patches Nov 14 00:36:36 https://ez.analog.com/message/42705 Nov 14 00:38:34 no, I have all the patches apart the two locomo 'fixes' Nov 14 00:39:08 maybe I borked the defconfig... Nov 14 00:39:21 I'll push my meta-handheld Nov 14 00:40:27 "read the spec and find what the error is" Nov 14 00:41:36 SD spec (at least partial) is public now, so we should be able to check. Nov 14 00:45:21 64M SD card -> -38 Nov 14 00:47:31 these were sandisk Nov 14 00:47:36 see that Nov 14 00:47:37 mmcblk0: mmc0:0000 LEXAR 952 MiB Nov 14 00:47:37 mmcblk0: p1 p2 Nov 14 00:47:37 mmcblk0: error -14 sending stop command, original cmd response 0x0, card status 0x0 Nov 14 00:48:07 that should be normal. Nov 14 00:48:10 Try reading that card. Nov 14 00:48:47 root@collie:~# mount /dev/mmcblk0p1 /mnt/.psplash/ Nov 14 00:48:47 EXT4-fs (mmcblk0p1): warning: mounting unchecked fs, running e2fsck is recommended Nov 14 00:48:47 EXT4-fs (mmcblk0p1): mounted filesystem without journal. Opts: (null) Nov 14 00:48:56 great :) Nov 14 00:49:06 root@collie:/mnt/.psplash# ls Nov 14 00:49:06 bin dllx home lost+found packages sl Nov 14 00:49:06 bk etc lib opt sbin usr Nov 14 00:49:06 root@collie:/mnt/.psplash# Nov 14 00:54:01 other 2 sandisk failing Nov 14 00:54:03 mmc0: new SD card on SPI Nov 14 00:54:03 mmcblk0: mmc0:0000 SD02G 1.89 GiB Nov 14 00:54:03 mmcblk0: p1 Nov 14 00:54:03 mmcblk0: error -38 sending status command, retrying Nov 14 00:54:22 :( Nov 14 00:54:53 mmc0: SPI card removed Nov 14 00:54:53 So sandisk cards do not work. Looks like some misbehaviour common to all of them. Nov 14 00:54:53 mmc0: new SDHC card on SPI Nov 14 00:54:53 mmcblk0: mmc0:0000 SD08G 7.60 GiB Nov 14 00:54:53 mmcblk0: p1 Nov 14 00:54:53 mmcblk0: error -38 sending status command, retrying Nov 14 00:55:08 ^ sandisk sdhc Nov 14 00:55:19 38 is -ENOSYS Nov 14 00:56:02 well, before it reads nr of partitions Nov 14 00:57:19 I'll look around, maybe I find some other brand aroud Nov 14 00:57:29 have to go now Nov 14 00:57:32 zZZZzz Nov 14 00:57:34 gn Nov 14 00:57:54 gn **** BEGIN LOGGING AT Fri Nov 14 01:14:24 2014 **** ENDING LOGGING AT Fri Nov 14 03:00:00 2014