00:10.43 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
00:17.43 | jonpry | hrm. where is this asm/mach/mmc.h |
00:20.09 | detule | problems? |
00:24.12 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.11) |
00:27.39 | detule | hm the overwhelming majority of my wakeup wake locks are reported as wakeup wake lock: event0-1422 |
00:27.45 | detule | event0 being "h2w headset" |
00:29.51 | jonpry | they changed something with includes. fixed now |
00:31.29 | jonpry | wistilt2 made this new pm.c that supposedly worked great for him. but my power in collapse was like 240ma. i reverted it was back down to 24 |
00:32.20 | jonpry | not sure if maybe it got reupgraded |
00:32.23 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
00:32.27 | detule | it did, meaning i broke something |
00:32.49 | detule | how are you measuring it |
00:33.31 | jonpry | i did it with a multimeter. but i have not measured lately. this is just from memory |
00:33.56 | detule | oh pm.c currently is mostly stock codeaurora very few hacks in there |
00:34.15 | detule | currently in 3.2 that is |
00:34.47 | jonpry | that may not be a good thing. it seems the amss is quite involved in collapse and if you don't do it a certain way it gets unhappy |
00:36.12 | detule | i was having problems with pm.c as it was before once we introduced idle pc....i reverted my changes yesterday and i found the same probelms waiting for me again....these random wakes from susepnd, where it just stays awake for more than a minute doing absolutely nothing |
00:36.46 | jonpry | idle suspend isn't really a good idea |
00:36.52 | jonpry | er idle pc |
00:37.04 | detule | i think i am starting to agree with that |
00:37.42 | jonpry | it saves like 2ma when the screen is on using hundreds anyways |
00:38.20 | detule | i have a suspcion that these power collapses are also doing some unhealthy things to the radio |
00:38.51 | jonpry | like what? |
00:39.14 | jonpry | how old is our drivers/usb now? |
00:39.45 | detule | i think you originally got it from .38 CA, at some point i updated the udc driver to CA 3.0 |
00:39.51 | detule | i didn't touch _otg |
00:40.10 | detule | i've been cherrypicking the android gadget stuff from their tree |
00:40.34 | detule | like these reports of gsm users |
00:40.42 | detule | (btw the smd patch didn't fix it for them) |
00:40.44 | jonpry | i'm not so worried about the _otg driver. just we have been copying over the whole usb core |
00:41.14 | detule | oh yeah, we should probably try to just copy udc and otg and the android stuff this time aorund if possible |
00:41.44 | jonpry | i think we need a working kernel before that can be attempted |
00:41.54 | jonpry | maybe even upgrade the 3.2 usb to 3.2 |
00:42.27 | detule | back to the gsm stuff -> it's like the radio falls asleep at stupid times and is unresponsive to AT commands |
00:42.51 | detule | i am seeing these AT< [WCDMA] WAKEUP PDA: U(),T(),R(DS) messages in my radio log -> i have no idea why I would be getting WCDMA prefixed messages |
00:43.25 | jonpry | would it be possible for them to run the radio log via dmesg thing |
00:43.48 | jonpry | i am not convinced the radio can fail to respond |
00:45.31 | *** join/#htc-linux ALoGeNo (~alogeno@unaffiliated/alogeno) |
00:45.53 | *** join/#htc-linux |lippa| (~lippa@CPE-121-220-44-147.lnse2.win.bigpond.net.au) |
00:46.01 | detule | yeah i guess we can try that kernel dump |
00:46.15 | jonpry | also the build fingerprint is weird |
00:46.24 | jonpry | shouldn't it be a12d2827cd6f8d27d61879cd2b9e3da9c93a5feb |
00:47.52 | jonpry | for that matter i can't even find a 3c01b |
00:48.26 | detule | it is |
00:48.31 | detule | ignore the "g" i guess |
00:50.01 | detule | uh oh |
00:50.07 | detule | i see what you are saying |
00:50.38 | detule | he's on a kernel i sent him -> it had a full revert of hyc's patch |
00:51.02 | detule | (all channels not just data) |
00:51.16 | jonpry | i think my patch is more of a problem than the hyc one |
00:52.06 | jonpry | that whole ldisc short circuiting thing was kanged from a driver that only used N_PPP |
00:52.33 | detule | but here's the thing your original patch is in .27 |
00:52.33 | jonpry | i never expected it to work on the raw channels but it seemed to so i never made it channel specific |
00:52.40 | detule | and hyc says everything is peachy over there |
00:52.47 | jonpry | yeah its timing sensitive |
00:55.39 | *** join/#htc-linux AstainW00T (~AstainHel@unaffiliated/astainhellbring) |
00:56.50 | *** join/#htc-linux ychavan (ychavan@nat/redhat/x-xwmhzlhbjnzfvwwi) |
00:57.17 | *** join/#htc-linux infernix (nix@unaffiliated/infernix) |
00:57.22 | jonpry | did codahighland try the latest autobuild? |
00:58.46 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
00:58.51 | ahigerd | sneezes |
00:59.31 | detule | jonpry, meet ahigerd = codahighland |
00:59.40 | ahigerd | waves |
00:59.50 | jonpry | we were just talking about you |
00:59.53 | ahigerd | So I heard |
01:00.22 | detule | so i've been feeding ahigerd kernels but perhaps i should've just had him try the autobuild |
01:00.41 | jonpry | yeah a12d2827cd6f8d27d61879cd2b9e3da9c93a5feb is the bees knees |
01:01.57 | detule | yeah the kernel there's more of a difference between the kernel i gave you and the one in the autobuild than i originally thought |
01:02.26 | ahigerd | Whoops |
01:02.37 | ahigerd | Does that include the tasklet change? |
01:02.49 | detule | disregard that kernel for now |
01:02.54 | ahigerd | snrks |
01:02.56 | ahigerd | I just got that one booted XD |
01:03.02 | detule | kexec ftw |
01:03.04 | ahigerd | Yup |
01:03.10 | ahigerd | But I had to reboot first to make adb wake up |
01:03.18 | jonpry | tasklet? |
01:03.26 | detule | i changed the queues in tty to tasklets |
01:03.42 | detule | also changed the wakelocks to IDLE |
01:03.45 | jonpry | bold |
01:03.57 | jonpry | ps ax | grep smd? |
01:04.35 | detule | eh? |
01:04.39 | ahigerd | pppd on smd1 |
01:04.46 | ahigerd | That's it |
01:05.08 | jonpry | tasklets show up is ps no? no real purpose |
01:05.42 | detule | well i thought there was less latency |
01:05.51 | detule | also they don't sleep |
01:06.04 | jonpry | hrm |
01:06.22 | jonpry | this is for the rescheduling on slab failure? |
01:06.45 | ahigerd | He was saying something about power collapse interfering with the buffers or something like that, and thinking tasklets would prevent that |
01:06.46 | detule | the entire smd_read queue business |
01:06.55 | ahigerd | doesn't remember exactly |
01:07.09 | detule | changing the wake locks to idle had more to do with the PC business |
01:09.20 | jonpry | this whole business with writing to the tty layers internal data structures is highly risky and suspect |
01:09.47 | ahigerd | PC business? |
01:09.52 | detule | power collapse |
01:09.54 | ahigerd | Oh yeah |
01:11.33 | jonpry | can you reproduce these problems quickly? |
01:11.53 | jonpry | on my phone it was almost constant. usually could only get one phonecall in before it freaked |
01:11.54 | ahigerd | Sadly no |
01:12.18 | ahigerd | For me it's maybe one or two times a week, seems signal-related |
01:12.34 | ahigerd | Rarely happens at home, happens more often when I'm out |
01:13.43 | *** join/#htc-linux whiteE (~whiteE@66.239.12.42.ptr.us.xo.net) |
01:14.07 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
01:14.07 | jonpry | imho that autobuild kernel will fix it right up |
01:16.27 | ahigerd | orly |
01:16.58 | ahigerd | pulls it down |
01:18.08 | ahigerd | kexecs |
01:18.37 | detule | you should also probably go to the stock ril |
01:18.43 | ahigerd | k |
01:18.55 | ahigerd | I'll make that switch after it finishes booting |
01:19.13 | detule | it might help manifest any issues sooner |
01:19.19 | ahigerd | Fair enough |
01:20.50 | jonpry | what were the change to drivers/base/power/Makefile? |
01:22.16 | *** join/#htc-linux mitsutaka (~mitsutaka@219.143.36.82) |
01:22.55 | detule | http://pastebin.com/ubJUuU41 |
01:23.28 | jonpry | :p i have that. what does it do? |
01:24.26 | detule | you put it there |
01:24.41 | detule | for one i think we are not compiling clock_ops |
01:25.38 | jonpry | probably need to ask myself next time i'm around |
01:26.04 | detule | probably commenting out clock_ops is the only thing that's needed |
01:27.20 | jonpry | it sort of builds. won't link yet |
01:33.39 | *** join/#htc-linux mitsutak_ (~mitsutaka@219.143.36.82) |
01:36.04 | *** join/#htc-linux bitrot (~rajkosto@2001:470:d76b:da7a:fd2f:5462:fff6:41c1) |
01:42.59 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-muiylpvulcmcyeqk) |
01:47.39 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
02:01.39 | jonpry | hrm what to do about drivers/staging |
02:02.57 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
02:03.47 | detule | not much has changed right? |
02:04.12 | detule | i mean we should be able to use staging/android from 3.3 |
02:06.25 | jonpry | not sure haven't looked at it. there are lots of different pmem's floating about |
02:06.36 | detule | oh yeah we would need our own pmem |
02:07.59 | detule | there's some changes to their alarm driver (they moved it from driver/rtc to staging/android) but i've backported that to 3.2 so it should work |
02:13.21 | jonpry | i think if this thing had pmem it would link and run |
02:14.18 | jonpry | dur. how do i removed files i accidentally committed |
02:17.33 | jonpry | k situation rectified |
02:18.28 | jonpry | the goods are pushed |
02:19.23 | jonpry | hrm. our ashmem is in /mm |
02:23.17 | jonpry | that was easier than i thought :p |
02:25.48 | jonpry | hrm no vibe |
02:33.03 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
02:46.58 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
02:53.47 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
03:07.11 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
03:30.51 | *** join/#htc-linux DuperMa (~Duper@93-173-141-174.bb.netvision.net.il) |
04:15.25 | jonpry | hmm. seems we are not supposed to be iomapping at the end to vmalloc anymore |
04:55.12 | *** join/#htc-linux friehmaen (~fm@fm.xers.de) |
04:55.51 | *** join/#htc-linux Rob2222 (~Miranda@p4FFD3CAC.dip.t-dialin.net) |
04:59.37 | *** join/#htc-linux DuperMan (~Duper@93-173-141-174.bb.netvision.net.il) |
05:08.50 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-66-44.dynamic.mnet-online.de) |
05:09.36 | *** join/#htc-linux lamikr (lamikr@nat/nokia/x-opsunpjijbqrveuy) |
05:35.42 | *** join/#htc-linux friehmaen (~fm@fm.xers.de) |
05:51.40 | *** join/#htc-linux G4y (~heaheahah@77.106.182.232) |
05:51.44 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
05:56.51 | *** join/#htc-linux DuperMan (~Duper@93-173-13-4.bb.netvision.net.il) |
06:27.19 | *** join/#htc-linux kiozen (~kiozen@p578a42db.dip0.t-ipconnect.de) |
06:33.00 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-164-174.netcologne.de) |
06:50.16 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
07:35.31 | *** join/#htc-linux pbaxter (~pbaxter@host254-13-dynamic.244-95-r.retail.telecomitalia.it) |
07:39.00 | *** join/#htc-linux gauner1986 (~Miranda@ip-109-91-241-106.unitymediagroup.de) |
07:57.33 | *** join/#htc-linux polyrhythmic (~polyrhyth@c-71-197-239-27.hsd1.wa.comcast.net) |
08:05.05 | *** join/#htc-linux Rob2222 (~Miranda@p4FFD16E0.dip.t-dialin.net) |
08:07.17 | *** join/#htc-linux MacDrunk (~marper@201.165.119.167) |
08:15.21 | *** join/#htc-linux swc|666 (~goldbond@Aircrack-NG/Friend) |
08:15.22 | *** join/#htc-linux ychavan (ychavan@nat/redhat/x-gnjkfdesveltdmtm) |
09:39.55 | *** join/#htc-linux Ondalf (~ondalf@cable-roi-fe98dc00-121.dhcp.inet.fi) |
09:49.50 | *** join/#htc-linux MethoS- (~clemens@134.102.106.250) |
10:09.10 | *** part/#htc-linux MacDrunk (~marper@201.165.119.167) |
10:53.29 | *** join/#htc-linux ychavan (ychavan@nat/redhat/x-pebihdvefgxbshxm) |
11:08.33 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
11:10.43 | *** join/#htc-linux zmc (~zmc@lurian.net) |
11:19.36 | *** join/#htc-linux GNUtoo (~gnutoo@host149-73-dynamic.56-82-r.retail.telecomitalia.it) |
11:31.02 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
11:43.19 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.11) |
11:53.15 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
12:07.05 | *** join/#htc-linux G4y (~heaheahah@160.8.202.84.customer.cdi.no) |
12:31.41 | *** join/#htc-linux yuuki (~yuuki@dsbg-d9bb1b40.pool.mediaWays.net) |
12:47.44 | *** join/#htc-linux AstainHellbring (AstainHell@unaffiliated/astainhellbring) |
13:11.57 | *** join/#htc-linux milaq|afk (~milaq@lb-info.de) |
13:15.08 | *** join/#htc-linux ccube (ccube@nx.ccube.de) |
13:19.31 | *** join/#htc-linux balans2 (~user@82-170-217-205.ip.telfort.nl) |
13:21.58 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
13:30.08 | jonpry | hi detule |
13:30.57 | detule | hey |
13:31.06 | detule | vmalloc whoa |
13:31.39 | jonpry | yeah big trouble |
13:32.25 | jonpry | writing has been on the wall for awhile so this work can be done on 3.2 |
13:32.44 | detule | wait what have they done |
13:33.07 | jonpry | for one your not allowed to define VMALLOC_END anymore |
13:33.14 | jonpry | just hard coded to 0xff000000 |
13:33.26 | jonpry | which is right before the coherent dma segment |
13:34.12 | jonpry | beyond that i'm not totally sure of the ramifications |
13:34.29 | detule | weren't we defining it at 0xf8 or whatever |
13:34.32 | jonpry | all arm soc's now have there io area mapped at 0xe0000000 |
13:34.41 | jonpry | yes we were |
13:35.07 | jonpry | i'm not sure what region that is in, but i suspect that ioremap doesn't work at all |
13:35.21 | detule | so they are forcing us to use more vmalloc? |
13:35.36 | *** join/#htc-linux Rob2222 (~Miranda@p4FFD16E0.dip.t-dialin.net) |
13:44.31 | detule | jonpry can we do something like this commit for us "ARM: move VMALLOC_END down temporarily for shmobile" |
13:45.47 | detule | oh all that does is an undef and a define |
13:50.16 | detule | so their commit log seems to indicate they want us to iomap within the vmalloc region |
13:51.21 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
13:51.38 | detule | <detule> so their commit log seems to indicate they want us to iomap within the vmalloc region |
13:53.10 | jonpry | its worth a try. seems to be what is happening in mach-msm |
13:53.24 | jonpry | checkout arm/mm/mmu.c:iotable_init |
13:53.43 | detule | yeah that's what i am looking at |
13:53.54 | jonpry | this is how the other platforms are doing it |
13:54.00 | detule | though if that's available for vmalloc what's to say we won't grab it by mistake? |
13:54.15 | jonpry | ? |
13:55.45 | detule | nevermind i see we are setting vm_ioremap flag |
13:56.40 | detule | which i guess make sures we can't trample over that with vmalloc calls |
13:57.17 | jonpry | i can't tell if iotable_init and ioremap have the same effect |
13:57.41 | jonpry | its possible that iotable_init way notifies the allocator somehow |
13:57.51 | jonpry | like the call to early_alloc |
14:03.25 | detule | so er just take out msm_map_common_io in io.c? |
14:05.46 | jonpry | we are talking 3.2? |
14:06.50 | jonpry | i think the first thing is to set vmalloc_end to the new address and rebase all the mappings on to 0xe* |
14:07.28 | *** join/#htc-linux LargePrime (~LargePrim@70-8-143-167.pools.spcsdns.net) |
14:08.09 | jonpry | dur i didn't realize io.c called iotable_init |
14:08.21 | detule | so er help me understand....i thought we can leave all our mappings to 0xf8 and above in this new scheme....but now they will be inside the vmalloc region |
14:09.03 | detule | and instead of us mapping them in io.c all this is doing is letting the mapping be done in mmu.c inside the vmalloc region |
14:10.57 | detule | dur someone still has to call iotable_init |
14:11.24 | jonpry | 3.2 and 3.3 iotable_init are way different |
14:12.16 | jonpry | i'm not sure why they want the iomappings at the bottom of vmalloc |
14:13.06 | detule | i mean different in that all it does is creates a vm struct for each entry |
14:21.46 | jonpry | somehow i think vmalloc starts at 0xf* |
14:22.21 | jonpry | "the default is 240m." |
14:25.33 | detule | i think in 3.3 they just gave us the option of defining vmalloc_start |
14:25.38 | detule | i think the default hasn't changed |
14:27.18 | jonpry | in vanilla 3.2 they used end==0xd* with all io starting at 0xe* |
14:27.44 | jonpry | and stuff apparently still worked when end became 0xff |
14:30.34 | jonpry | i like that idea for some reason |
14:35.37 | *** join/#htc-linux AstainHellbring (~AstainHel@unaffiliated/astainhellbring) |
14:51.24 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
14:52.05 | jonpry | changing everything to 0xe didn't help |
14:53.00 | detule | do you have any ramconsole output after the failed boot |
14:53.29 | detule | i mean i am sure this business is no good but could it be failing for other reasons as well |
14:54.22 | jonpry | no ramconsole |
14:54.35 | jonpry | that has lead me to believe its a memory mapping problem |
14:55.10 | jonpry | other possibilities are atag problems or segfault somewhere |
15:00.31 | detule | well one thing we could do is to revert the vmalloc patch in 3.3 to pin down if that's the problem |
15:02.16 | jonpry | yeah. there is some interrupt stuff going on too. but i think ramconsole comes up before interrupt |
15:04.20 | jonpry | there is this strange build error about ARCH_MSM not being defined that has be worried |
15:09.22 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
15:10.21 | detule | they've also done a number on ioremap |
15:22.10 | *** join/#htc-linux rpierce99 (~rpierce99@96-42-107-19.dhcp.stcd.mn.charter.com) |
15:25.28 | *** join/#htc-linux LordDeath (~LordDeath@cable-81-173-164-174.netcologne.de) |
15:37.30 | jonpry | http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=b12e9ba59c83f7df846602b201b64e4ddf28ccee |
15:38.55 | detule | lol yeah that might be important |
15:50.45 | *** join/#htc-linux rob_w (~bob@ppp-88-217-92-252.dynamic.mnet-online.de) |
15:50.45 | *** join/#htc-linux rob_w (~bob@unaffiliated/rob-w/x-1112029) |
15:50.51 | detule | this as well? http://lists.infradead.org/pipermail/linux-arm-kernel/2012-March/088009.html |
16:05.32 | *** join/#htc-linux helicopter88 (~helicopte@95.234.35.188) |
16:10.04 | jonpry | i'm guessing those made it into 3.3 |
16:13.39 | *** join/#htc-linux jzma (5278306c@gateway/web/freenode/ip.82.120.48.108) |
16:34.58 | *** join/#htc-linux Bry8Star (~Bry8Star@gateway/tor-sasl/bry8star) |
16:35.59 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-66-44.dynamic.mnet-online.de) |
16:53.22 | *** join/#htc-linux mitsutaka (~mitsutaka@123.116.125.0) |
16:56.50 | jonpry | arch/arm/mach-msm/board-htcrhodium.c:1398:8: error: 'arch_ioremap_caller' undeclared (first use in this function) |
16:57.42 | *** join/#htc-linux CptAJ (AJ@190.74.22.95) |
17:00.59 | jonpry | guess we did need that other patch |
17:06.05 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
17:06.05 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
17:10.27 | jonpry | ahigerd any trouble with the new kernel? |
17:10.54 | ahigerd | I dunno, I've been sleeping most of the time since I installed it |
17:11.02 | ahigerd | Haven't made any calls |
17:11.19 | jonpry | ah |
17:11.20 | arrrghhh | you don't sleepcall |
17:11.26 | ahigerd | I try not to |
17:11.33 | arrrghhh | that was one of the requirements of running a kernel from jonpry |
17:11.55 | jonpry | have to put that in the readme |
17:12.29 | arrrghhh | lol |
17:18.50 | detule | yeah that patch didn't help it boot over here |
17:20.24 | jonpry | ouch |
17:20.48 | jonpry | maybe we need a more powerful debugging method |
17:24.32 | *** join/#htc-linux paulk-desktop (~paulk@lib33-1-82-233-88-171.fbx.proxad.net) |
17:44.10 | jonpry | is there a region of ebi/ebin that would work for ramconsole? |
17:52.03 | *** join/#htc-linux GNUtoo (~gnutoo@host149-73-dynamic.56-82-r.retail.telecomitalia.it) |
18:09.43 | *** join/#htc-linux whitekidney (~wkl0l@141.0.136.22) |
18:09.47 | whitekidney | damn autojoin not working |
18:22.00 | jonpry | i got it to print something after decompress |
18:25.10 | *** join/#htc-linux Rob2223 (~Miranda@p4FFD1EF3.dip.t-dialin.net) |
18:25.51 | *** join/#htc-linux Cthulhu (77eb360e@gateway/web/freenode/ip.119.235.54.14) |
18:37.15 | jonpry | detule is your assembler getting better? |
18:43.48 | *** join/#htc-linux Rob2222 (~Miranda@p4FFD3212.dip.t-dialin.net) |
18:53.39 | detule | my assembler is rudimentary at best |
18:54.03 | detule | what printed out? |
18:54.16 | detule | and what did you do to get it that far |
19:02.24 | *** join/#htc-linux Rob2223 (~Miranda@p4FFD0F2C.dip.t-dialin.net) |
19:02.50 | jonpry | http://pastebin.com/7EuZUxc5 |
19:03.03 | jonpry | i'm just writing a,b,c into ramconsole area |
19:03.32 | jonpry | c hasn't happened yet. its the tricky one |
19:04.55 | jonpry | but once it does. this could be a super ultimate tool |
19:06.31 | jonpry | hrm c worked |
19:06.35 | ahigerd | What are you guys up to? |
19:06.54 | jonpry | trying to figure out why kernel 3.3 won't boot |
19:10.57 | helicopter88 | 3.2 boots right? |
19:11.21 | arrrghhh | lol |
19:11.26 | arrrghhh | 3.2's worked for a while man |
19:11.41 | arrrghhh | probably before you got that battery sucking device |
19:12.16 | arrrghhh | (on RHOD mind you) |
19:12.33 | arrrghhh | i don't think .39+ works on anything but RHOD as far as our livery of devices goes |
19:12.49 | arrrghhh | .35 probably works on a few because of alex and michael |
19:13.07 | helicopter88 | I missed a "," |
19:13.22 | arrrghhh | whut |
19:13.29 | helicopter88 | what have they changed between 3.2 and 3.3 ? |
19:14.17 | jonpry | nothing obvious |
19:14.35 | jonpry | i think either the mmu is fubar or irq |
19:15.10 | detule | if it got as far as irq would we see anything in ramconsole? |
19:15.20 | helicopter88 | arrrghhh: it lasts 2 days with moderate usage |
19:16.59 | jonpry | detule, not sure |
19:17.00 | *** join/#htc-linux ali1234 (~ajbuxton@s15821883.onlinehome-server.info) |
19:21.04 | detule | mach-msm/irq.c hasn't changed on korg in over a year |
19:22.20 | jonpry | yeah they are pushing for this new thing that is common to all socs |
19:23.50 | jonpry | this thing i am doing with the mmu is actually pretty cool |
19:24.16 | jonpry | it could be refactored with minimal effort to give debug_ll level of functionality via smi |
19:26.17 | jonpry | sweet we have entry into c code at start_kernel() |
19:29.26 | *** join/#htc-linux furtardo (~mks@nat/yahoo/x-iawihgkcgyqzyqlo) |
19:36.04 | *** join/#htc-linux LargePrime (~LargePrim@173-108-144-110.pools.spcsdns.net) |
19:44.22 | *** join/#htc-linux swc|666 (~goldbond@Aircrack-NG/Friend) |
19:45.00 | *** join/#htc-linux detule (~detule@unaffiliated/d3tul3) |
19:54.04 | jonpry | death is in setup_arch |
19:54.45 | jonpry | or maybe not |
20:01.37 | *** join/#htc-linux leviathan (~quassel@2001:470:26:484:6ef0:49ff:fee6:8dca) |
20:13.52 | *** join/#htc-linux kiozen (~kiozen@ppp-93-104-66-44.dynamic.mnet-online.de) |
20:27.44 | *** join/#htc-linux ahigerd (~ahigerd@173.218.156.102) |
20:27.57 | *** join/#htc-linux ahigerd (~ahigerd@libqxt/developer/ahigerd) |
20:28.24 | *** join/#htc-linux jonpry (~jon@adsl-98-85-196-230.mco.bellsouth.net) |
20:36.41 | jonpry | appears to be hanging in iotable_init |
20:40.56 | detule | early_alloc_alligned? |
20:41.52 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4D79.dip.t-dialin.net) |
20:43.06 | jonpry | could be. i'm not sure how many more add traces and reboot i can take |
20:45.25 | jonpry | stupid usb cable keeps crashing my laptop |
20:48.03 | *** join/#htc-linux mastermerlin (~Adium@p4FEE4D79.dip.t-dialin.net) |
20:55.25 | detule | MSM_SHARED_RAM_BASE that's IOMEM(0xE0100000) |
20:56.14 | detule | and that last entry in msm_io_desc is at that virt |
20:56.43 | detule | is that under vmalloc_start? |
20:58.20 | jonpry | ? |
20:58.21 | jonpry | #define MSM_SHARED_RAM_BASE IOMEM(0xF8100000) |
20:59.16 | jonpry | i went back to 0xf8 btw |
21:00.05 | detule | ugh looking at the wrong branch |
21:01.15 | jonpry | interested in doing any tracing? |
21:02.30 | detule | sure, but my usb cable situation is worse than yours....at least until i get home later tonight |
21:03.16 | detule | i need to have an elaborate setup (often involves the phone hanging by the usb cable) for it to register |
21:03.36 | jonpry | lol |
21:03.48 | jonpry | http://pastebin.com/HwpHu5Fm |
21:04.17 | jonpry | once it gets going it just phys=virt for smi |
21:04.26 | jonpry | so you do for example *((unsigned char*)0x8e0025) = 'j' |
21:05.44 | CptAJ | detule, My usb cable is being held together by medical tape >_> |
21:06.28 | detule | CptAJ, did you figure out your screen issues on the whitestone |
21:07.30 | CptAJ | I used haret to watch the tssc registers and logged a couple of taps with the stylus. I'm trying to compare that to the driver source and see exactly where it is going wrong. |
21:07.31 | detule | jonpry, so how far do you get with the trace |
21:07.49 | jonpry | i know that it stops in msm_map_common_io |
21:08.02 | CptAJ | I'm slowly trying to wrap my head around it |
21:08.06 | detule | CptAJ, after you left last time someone mentioned it could be toolchain issues |
21:08.12 | detule | which toolchain are you using? |
21:08.19 | CptAJ | latest from the site |
21:08.25 | jonpry | uh oh |
21:08.28 | detule | probably need to downgrade |
21:08.30 | CptAJ | which one is the prefered one? |
21:08.33 | detule | 2010 q3 |
21:08.41 | CptAJ | I'll try that. thanks |
21:09.50 | detule | jonpry, so you see 'd' but no 'e' ? |
21:11.52 | jonpry | correct |
21:12.04 | jonpry | i did some strange modulo 4 deal |
21:12.21 | jonpry | so it prints AaaaBbbbCcccDdddEee Ff |
21:14.52 | detule | i don't see your capital A or B |
21:16.18 | jonpry | assembler |
21:16.39 | jonpry | + mov r0, #0x8e0000 |
21:16.39 | jonpry | + mov r1, #0x41 |
21:17.21 | detule | imagine that 0x41 |
21:17.49 | detule | thanks |
21:18.02 | jonpry | no problem |
21:18.25 | detule | you're right this is pretty cool |
21:19.37 | jonpry | it should be possible to change debug-macro.S to use it |
21:35.55 | detule | i gotta go, i have a feeling you'll have it figured out by the time i get a chance to run these traces...pretty cool eitherway |
23:24.20 | *** join/#htc-linux LargePrime (~LargePrim@184.244.158.129) |
23:25.44 | *** join/#htc-linux Alex[sp3dev] (~alexander@178.76.204.11) |