#milestone-modding | Logs for 2012-10-16

Back
[00:06:34] -!- rcf [rcf!~rcf@178-117-158-72.access.telenet.be] has joined #milestone-modding
[00:26:48] -!- sfahadzb_ [sfahadzb_!~sfahadzb@2.91.152.41] has joined #milestone-modding
[00:28:08] -!- sfahadzb has quit [Ping timeout: 245 seconds]
[00:31:30] sfahadzb_ is now known as sfahadzb
[00:32:49] -!- rcf has quit [Quit: This war is mine]
[00:54:39] -!- eXtremo has quit [Ping timeout: 276 seconds]
[01:00:22] -!- eXtremo [eXtremo!~eXtremo@unaffiliated/extremo] has joined #milestone-modding
[01:15:54] Skrilax_CZ is now known as Skrilax_CZ|Away
[01:27:42] -!- sfahadzb has quit [Ping timeout: 272 seconds]
[01:32:54] -!- sfahadzb_ [sfahadzb_!~sfahadzb@2.91.152.41] has joined #milestone-modding
[01:36:17] sfahadzb_ is now known as sfahadzb
[05:07:05] -!- cubi_ [cubi_!~mcd@p4FFE1527.dip.t-dialin.net] has joined #milestone-modding
[05:10:07] -!- cubi has quit [Ping timeout: 245 seconds]
[05:10:07] cubi_ is now known as cubi
[05:11:03] -!- mrvek [mrvek!~mrvek@109.60.107.150] has joined #milestone-modding
[05:17:15] -!- mrvek has quit [Ping timeout: 260 seconds]
[05:19:22] -!- mrvek [mrvek!~mrvek@109.60.107.150] has joined #milestone-modding
[05:23:49] -!- mrvek has quit [Ping timeout: 255 seconds]
[05:24:06] -!- mrvek [mrvek!~mrvek@109.60.107.150] has joined #milestone-modding
[05:28:11] -!- mrvek [mrvek!~mrvek@109.60.107.150] has parted #milestone-modding
[05:28:11] -!- mrvek [mrvek!~mrvek@109.60.107.150] has joined #milestone-modding
[05:28:11] -!- mrvek has quit [Client Quit]
[05:28:53] -!- mrvek [mrvek!~mrvek@109.60.107.150] has joined #milestone-modding
[05:33:41] -!- pontomedon^ [pontomedon^!~poseidon@chello062178135187.2.14.univie.teleweb.at] has joined #milestone-modding
[05:33:42] -!- mode/#milestone-modding [+o pontomedon^] by ChanServ
[05:37:36] -!- pontomedon has quit [Ping timeout: 246 seconds]
[06:11:49] -!- endstille [endstille!~endstille@port-87-193-224-26.static.qsc.de] has joined #milestone-modding
[06:28:26] -!- endstille has quit [Read error: Connection reset by peer]
[06:35:06] -!- mifritscher [mifritscher!~michi@p57B2E2B7.dip.t-dialin.net] has joined #milestone-modding
[06:35:33] -!- endstille [endstille!~endstille@port-87-193-224-26.static.qsc.de] has joined #milestone-modding
[06:43:34] -!- pontomedon [pontomedon!~poseidon@chello062178135187.2.14.univie.teleweb.at] has joined #milestone-modding
[06:43:34] -!- mode/#milestone-modding [+o pontomedon] by ChanServ
[06:47:45] -!- pontomedon^ has quit [Ping timeout: 265 seconds]
[06:50:43] -!- pontomedon has quit [Quit: Bin nur schnell in Austin, TX]
[08:10:06] <mrvek> and my sdcard speed is the same as in running system (2ndboot CM10). though it was tested about 2 days ago
[08:41:29] -!- mifritscher has quit [Quit: Leaving.]
[09:11:47] -!- Quarx [Quarx!~Quarx@94.137.58.58] has joined #milestone-modding
[09:21:08] -!- mifritscher [mifritscher!~michi@zft-server-1.telematik-zentrum.de] has joined #milestone-modding
[09:47:27] -!- pontomedon [pontomedon!~poseidon@chello062178135187.2.14.univie.teleweb.at] has joined #milestone-modding
[09:47:27] -!- mode/#milestone-modding [+o pontomedon] by ChanServ
[09:47:36] -!- K4sperle [K4sperle!~lutz@0x0badc0.de] has joined #milestone-modding
[09:48:57] -!- RCFwork [RCFwork!~klaas@85.255.197.126] has joined #milestone-modding
[09:50:57] -!- endstille has quit [Read error: Connection reset by peer]
[10:00:50] -!- k1l_ [k1l_!~k1l@ubuntu/member/k1l] has joined #milestone-modding
[10:04:04] -!- k1l has quit [Ping timeout: 244 seconds]
[10:04:50] K4sperle is now known as Kasperle
[10:20:28] -!- stlifey_ [stlifey_!~stlifey@183.46.13.190] has joined #milestone-modding
[10:23:24] -!- stlifey has quit [Ping timeout: 246 seconds]
[10:27:55] -!- mrvek_ [mrvek_!~mrvek@m77-219-35-180.cust.tele2.hr] has joined #milestone-modding
[10:32:33] -!- mrvek_ has quit [Ping timeout: 260 seconds]
[10:34:43] -!- mrvek_ [mrvek_!~mrvek@m77-219-57-78.cust.tele2.hr] has joined #milestone-modding
[10:34:53] <mrvek_> I haven't noticed slowdowns in 2ndbootOR. Copying ~200mb files finished rather quickly. Nandroid backup (excluding sd-ext) also finished fast.
[10:37:22] -!- sfrancis [sfrancis!sfrancis@nat/redhat/x-pneodryvkagbwrsd] has joined #milestone-modding
[10:37:34] -!- endstille [endstille!~endstille@80.154.119.181] has joined #milestone-modding
[10:39:13] Skrilax_CZ|Away is now known as Skrilax_CZ
[10:43:52] -!- mrvek_ has quit [Ping timeout: 252 seconds]
[11:07:38] <Skrilax_CZ> mrvek_: I don't think the slowdowns in recovery will be noticeable, but clock setup is different too
[11:09:38] <Skrilax_CZ> It looks like something causes the clock setup to be different afterall
[11:10:31] <Skrilax_CZ> as when I patched hboot to fix the setup on MPU and IVA2, it still ended up on different values
[11:15:13] <Skrilax_CZ> damn, forgot on overclock :D
[12:13:34] <Skrilax_CZ> nadlabak: mrvek_: well, try forcing 250 MHz after healing suspend / resume; it is like the sluggish mode, isn't it?
[12:19:06] <Quarx> Skrilax_CZ: yes, looks like sluggish mode
[12:19:10] <Quarx> 300mhz on defy
[12:19:51] <Skrilax_CZ> Quarx: I compared CPU usage, and in sluggish mode, it's x4 for everything
[12:20:26] <Quarx> Yes, I noticed this too.
[12:20:57] <Skrilax_CZ> but I can't see where it comes from yet (dumping registers from TRM related to clocking) :|
[12:43:00] <nadlabak> Skrilax_CZ, Quarx: although it looks the same, it's definitely not the same
[12:43:23] <nadlabak> quadrant CPU scores:
[12:43:33] <nadlabak> sluggish - 1112
[12:43:53] <Skrilax_CZ> nadlabak: yeah I know (MPU opps in sluggish mode is for 1 GHz, while when forcing it is for 250 MHz)
[12:43:54] <nadlabak> 250 mhz in healed state - 312
[13:01:31] -!- TobiX [TobiX!tobias@zoidberg.org] has joined #milestone-modding
[14:09:03] -!- magerquark [magerquark!~magerquar@host-88-217-184-252.customer.m-online.net] has joined #milestone-modding
[14:11:11] -!- Bavilo [Bavilo!~Bavilo@ip-95-223-22-136.unitymediagroup.de] has joined #milestone-modding
[14:52:10] -!- endstille has quit [Quit: endstille]
[15:21:22] -!- endstille [endstille!~endstille@dslb-188-109-010-045.pools.arcor-ip.net] has joined #milestone-modding
[15:41:59] -!- fireglow [fireglow!~fireglow@irssi.eu] has parted #milestone-modding
[15:58:19] <zewelor> im wondering, maybe try to add some sleep before 2ndboot
[15:58:45] <zewelor> if i recall correctly executing something in sh hijack it before some kernel stuff thats still goin on
[15:59:34] <zewelor> not sure how much long BP can wait before enumeration
[16:00:14] <zewelor> no it can wait 30 s as its being enumerated in first kernel ( ? )
[16:01:21] <zewelor> 30 seconds after being enumerated on usb and before netmux / ipc modules are loaded
[17:03:12] -!- CopyLiu_ has quit [Read error: Connection reset by peer]
[17:03:24] -!- CopyLiu [CopyLiu!~copyliu@61.142.73.165] has joined #milestone-modding
[17:06:13] DAGr8 is now known as random1
[17:06:15] random1 is now known as random11
[17:06:18] random11 is now known as random112
[17:06:24] random112 is now known as random3
[17:06:29] random3 is now known as DAGr8
[18:24:07] -!- RCFwork has quit [Quit: Leaving.]
[18:30:27] -!- jausk [jausk!~quassel@44.85-85-66.dynamic.clientes.euskaltel.es] has joined #milestone-modding
[18:39:09] <Skrilax_CZ> zewelor: it's not a real problem to "fake suspend (= suspend all domains, then instead of waiting for an interrupt wait several cycles then resume)" on the phone and then reinit everything in hboot
[18:40:18] <Skrilax_CZ> complicated a bit, however, I think it's better to figure out where the problem is exactly; noone says that suspend / resume guarantees "healing"
[18:41:37] <zewelor> Skrilax_CZ: oh i didnt mean to fake suspend, by sleep i mean like "sleep 10" before invoking 2ndboot
[18:42:23] <Skrilax_CZ> zewelor: ah you don't understand
[18:43:12] <Skrilax_CZ> there is no "magic" sleep function - you have to do the suspend / resume manually
[18:43:48] <zewelor> i mean to delay 2ndboot during first kernel boot
[18:43:52] <zewelor> by few seconds
[18:44:14] <Skrilax_CZ> for what?
[18:45:50] <zewelor> because sh init get executed somehow fast and some kernel stuff is still going, like we could insert module to break usb enumeration in the middle so MAYBE there is something going on which gets interrupted by 2ndbooting and if it would settle down it would work
[18:47:01] <zewelor> well you got much more knowledge about that topic than me so you will knew better if its something possible or not
[18:49:16] <Skrilax_CZ> sh hijack get's executed quite late actually (cca 28 seconds after kernel start)
[18:50:12] <Skrilax_CZ> or well, that' sin the log I just opened (simply right after yaffs filesystems are mounted)
[18:50:22] <TheSpiritof69> Additionally, I have the feeling the sluggishness isn't only restricted to first sleep/suspend
[18:50:39] <TheSpiritof69> Sometimes, while attached to AC or USB the device starts lagging again
[18:50:44] <TheSpiritof69> Until unplugged and sleep
[18:51:19] <TheSpiritof69> But I'm not entirely sure if this is caused by the same problem or just device having too much trash installed
[18:51:58] <Skrilax_CZ> TheSpiritof69: perhaps you had in ac / usb since when it booted (no sleep when USB / AC connected)?
[18:53:01] <nadlabak> Skrilax_CZ: it's actually 14s
[18:53:07] <TheSpiritof69> No, it happened to when I attached to charge phone after work
[18:53:11] <TheSpiritof69> too*
[18:53:50] <TheSpiritof69> Being out the house I guess there were a few sleep states during 8hrs
[18:54:17] <Skrilax_CZ> perhaps it rebooted while in the charger :P
[18:54:46] <TheSpiritof69> Well that may be possible, but I actually don't suspect it to have rebooted
[18:55:07] <TheSpiritof69> Unless it's always the weird no pin reboot thing
[18:55:52] <Skrilax_CZ> nadlabak: the timer is doubled?
[18:58:08] <nadlabak> Skrilax_CZ: dmesg just before 2ndboot - http://pastebin.com/iUxNf6Ff
[18:58:35] <TheSpiritof69> But I'll try to get logfiles next time it happens
[18:58:52] <TheSpiritof69> Maybe that'll help or make it clear
[19:00:34] <Skrilax_CZ> nadlabak: ah, well I just read some non-2ndboot dmesg I had, in which yaffs2 got mounted on 28s
[19:03:43] <zewelor> so i was thinking maybe 2ndboot interrupts something, which should be finished in 1st kernel and cause slugginess in 2ndboot kernel, but probably its nonsense
[19:06:05] <Skrilax_CZ> zewelor: it's theoretically possible (you can tryout & error that if you want); but normally you want to 2nd-boot to child kernel ASAP
[19:07:32] -!- jausk has quit [Remote host closed the connection]
[19:07:44] <nadlabak> zewelor: sh_hijack at 14s is still 5s late after the usb modem enumeration already happened, btw.
[19:13:52] <zewelor> nadlabak: but its not finished, only maybe here its finished, we were just trying with inserting module to overwrite some kernel function during enumeration and it was being inserted somewhere during enumeration i dont remmeber which function now, i dont have any logs now i deleted all :(
[19:17:13] <nadlabak> zewelor: well, the timing of enumeration was probably different under 2.6.29 kernel
[19:18:06] <zewelor> oh they are in original ko-hook sources from nothize, some functions we overwrite to dump some BP usb info during enumeration
[19:18:11] <zewelor> yes it was for sure
[19:18:38] <zewelor> ye now i rembmer it was too late in .32 to alter it so we didnt do anything there
[19:19:08] <zewelor> maybe that timing differences between 2 kernel versions matters in context of sluginess
[19:19:25] <zewelor> as .29 -> .29 i dont recall any sluginess
[19:20:38] <nadlabak> anyway, just to be sure, I tried 10s sleep before 2ndboot now and it does not help
[19:24:15] <Skrilax_CZ> zewelor: well, you can flash back .29 kernel and try 2nd-booting from there (needs to update cpcap firmware in the child .32 as we don't reset it, see recovery branch)
[19:24:34] <Skrilax_CZ> *2ndboot_recovery branch in kernel repo
[19:28:37] <nadlabak> the current 2ndboot sources need to be patched to be compatible with 2.6.29
[19:29:00] <mrvek> recovery_2ndboot branch?
[19:29:47] <Skrilax_CZ> well you actually could reboot from 2nd-boot recovery?
[19:31:12] <nadlabak> the hbootmod - plat vs. mach includes, FA000000 vs. D8000000 base
[19:32:17] <Skrilax_CZ> FA000000 is manually remapped here: https://github.com/nadlabak/2ndboot/blob/master/hbootmod/hboot.c#L181
[19:35:32] <nadlabak> I see, thanks
[19:53:20] -!- Quarx has quit []
[20:06:14] -!- sfahadzb_ [sfahadzb_!~sfahadzb@2.91.173.39] has joined #milestone-modding
[20:06:15] -!- sfahadzb has quit [Ping timeout: 255 seconds]
[20:10:31] -!- Bavilo has quit [Read error: Connection reset by peer]
[20:13:02] <nadlabak> as expected, there's the same sluggishness after 2ndboot from 2.6.29 (recovery) to 2.6.32
[20:14:22] <nadlabak> so maybe only 2.6.29 to 2.6.29 is not affected, but that is not real option for us
[20:15:09] <DAGr8> https://www.youtube.com/watch?v=iX50dkXXKlY&feature=player_embedded
[20:18:48] -!- sfahadzb_ has quit [Ping timeout: 245 seconds]
[20:21:53] -!- mifritscher has quit [Quit: Leaving.]
[20:25:30] -!- sfahadzb_ [sfahadzb_!~sfahadzb@2.91.173.39] has joined #milestone-modding
[20:25:47] sfahadzb_ is now known as sfahadzb
[20:43:33] -!- rcf [rcf!~rcf@178-117-158-72.access.telenet.be] has joined #milestone-modding
[20:44:39] <mrvek> that sluggishness is there only after 2ndboot? eng ms is not affected?
[20:51:43] -!- pontomedon has quit [Quit: I'll be back.]
[21:03:55] <Skrilax_CZ> mrvek: yes
[21:04:40] <mrvek> wow, a spontaneous reboot
[21:04:46] <mrvek> thanks
[21:07:43] -!- endstille has quit [Quit: endstille]
[21:10:45] -!- mifritscher [mifritscher!~michi@p57B2E2B7.dip.t-dialin.net] has joined #milestone-modding
[21:11:03] -!- pontomedon [pontomedon!~poseidon@chello062178135187.2.14.univie.teleweb.at] has joined #milestone-modding
[21:11:03] -!- mode/#milestone-modding [+o pontomedon] by ChanServ
[21:21:49] <Skrilax_CZ> nadlabak: thanks for clearing that out
[22:06:29] -!- pontomedon has quit [Quit: I'll be back.]
[22:06:33] -!- k1l [k1l!~k1l@ubuntu/member/k1l] has joined #milestone-modding
[22:09:40] -!- k1l_ has quit [Ping timeout: 245 seconds]
[22:11:48] -!- sfrancis has quit [Quit: Leaving]
[22:25:38] -!- mrvek has quit [Ping timeout: 240 seconds]
[22:39:57] -!- mifritscher has quit [Ping timeout: 252 seconds]
[22:39:57] -!- mifritscher1 [mifritscher1!~michi@p57B2E2B7.dip.t-dialin.net] has joined #milestone-modding
[22:39:58] mifritscher1 is now known as mifritscher
[23:05:11] -!- mifritscher has quit [Ping timeout: 246 seconds]
[23:29:21] <Skrilax_CZ> btw. in here both SDRC and SMS are softreset: https://github.com/nadlabak/kernel/blob/master/arch/arm/mach-omap2/sdrc.c#L136
[23:30:59] <Skrilax_CZ> it's however interesting such things are not made from code running in SRAM
[23:42:09] <nadlabak> hmm, yes...
[23:53:23] -!- magerquark has quit [Remote host closed the connection]