00:00.09 | *** join/#htc-linux Dan617 (n=Dan@24.83.213.216) |
00:00.51 | *** part/#htc-linux Dan617 (n=Dan@24.83.213.216) |
00:01.04 | Kevin2 | cr2: I guess p2ve(0xABC) would be the output of "dump mmu 2 0xABC" |
00:01.21 | Kevin2 | NetRipper: No IDE - I just use emacs. |
00:03.59 | bryan_ | hey Kevin2 hows it going ? |
00:04.41 | Kevin2 | bryan_: Hi |
00:05.07 | bryan_ | i managed to get my haretlog.txt |
00:05.44 | bryan_ | looks like its detecting my msm7500 core wrong .. it says the manufacturer is samsung. |
00:06.14 | Kevin2 | bryan_: Can you pastebin the log? |
00:06.47 | bryan_ | im sorry pastebin? |
00:07.29 | Kevin2 | Go to pastebin.ca and copy and paste your log into the window. Then send the resulting url back to this channel. |
00:08.11 | bryan_ | ohh ok .. sorry never heard of that service b4 .. k one sec .. hehe |
00:10.27 | bryan_ | http://pastebin.ca/1211922 |
00:10.36 | bryan_ | cool |
00:12.35 | Kevin2 | Hrmm. That's from haret 0.5.1 - can you grab a more recent version from: http://www.handhelds.org/~koconnor/haret/haret-20080726f.exe |
00:17.59 | cr2 | Kevin2: hi |
00:18.51 | Kevin2 | Hi cr2! |
00:18.52 | cr2 | Kevin2: how can i teach regs_foo.py about multiple virtual mappings. if they are dynamic |
00:19.19 | Kevin2 | Hrmm. Dynamic in which way? Per machine, per session? |
00:19.50 | cr2 | i hope that per user machine |
00:20.18 | cr2 | my mapping for the interrupt controller is different from NetRipper |
00:20.25 | Kevin2 | So, hermes might have one map, and athena has another? Or do you mean tom's phone has one map and bob's has another? |
00:20.35 | cr2 | no. |
00:20.41 | cr2 | we have the same phone |
00:20.53 | cr2 | raphael aka touch pro |
00:21.06 | Kevin2 | How will you know which map to use? |
00:21.47 | cr2 | looking at dump mmu, adding all mappings to 'irqs' and looking at the output |
00:22.01 | bryan_ | kevin2: heres with the version you mentioned http://pastebin.ca/1211930 |
00:23.55 | Kevin2 | cr2: Does both of your machines have the same ids in the haretlog.txt file? |
00:24.36 | cr2 | yes |
00:25.20 | cr2 | i'm using his patched version of haret |
00:26.00 | cr2 | <PROTECTED> |
00:27.12 | Kevin2 | cr2: I was thinking of the wince ids - something like: |
00:27.12 | Kevin2 | Trying to detect machine (Plat='SmartPhone' OEM='S640') |
00:28.08 | Kevin2 | bryan_: Looks like your phone is not unlocked - I've gotten a few reports of that. haret shouldn't of crashed though. I'll fix that up. |
00:28.26 | cr2 | Kevin2: (Plat='PocketPC' OEM='HTC Touch Pro T7272') |
00:28.44 | Kevin2 | And same processor report too? |
00:29.00 | cr2 | Detected machine Raphael/MSM7201A (Plat='PocketPC' OEM='HTC Touch Pro T7272') |
00:29.10 | cr2 | yes |
00:30.14 | Kevin2 | cr2: Best I can think of is to write some code in a new machine class that tries to figure out which model it is. Then have it report Raphael-a or Raphael-b -- then have the python stuff tie the mappings to one of the two phone types. |
00:30.43 | cr2 | Kevin2: ok, obviously we need to ask other people to do dump mmu. |
00:31.22 | Kevin2 | Alternatively, we could extend haret so that one could force a specific machine type. Then you wouldn't have to code up a model detection algorithm. |
00:31.38 | cr2 | Kevin2: the static mapping is the same, but the 'ioremap()' values happen to be different. |
00:32.23 | cr2 | yes, we need some more user-friendly way of updating machtypes.txt |
00:32.26 | Marajin | I really must stop trying to kill myself... |
00:32.30 | cr2 | without recompiling haret |
00:32.35 | bryan_ | kevin2: locked as in winmo Reg is locked or the hardware is locked .. anyidea if qpst is able to do this unlocking ? |
00:32.37 | cr2 | Marajin: hehe. |
00:33.17 | cr2 | bryan_: "Old KMode was 0" |
00:33.24 | Marajin | cr2: I just walked twice as far as I should walk... and I did so at midnight, while it's quite cold, for what turned out to be a wasted journey, and while ill |
00:33.29 | Kevin2 | cr2: Yeah - that would be nice. |
00:33.41 | Marajin | cr2: I'm fairly sure this counts as technically trying to kill myself. |
00:33.49 | cr2 | bryan_: on Plat='SmartPhone' you need to edit registry |
00:34.13 | Kevin2 | cr2: Hrmm. the python stuff only cares about physical addresses. |
00:34.33 | cr2 | Kevin2: strange |
00:34.54 | bryan_ | cr2: edit reg to what ? PPC |
00:34.55 | cr2 | then why do i get such messages: |
00:34.59 | cr2 | 029997: mem IRQS(2) 5e302000=00000400 (00000400)M |
00:35.11 | cr2 | 5e302000 is remapped 0xc0000000 |
00:35.14 | cr2 | = irq controller |
00:36.00 | cr2 | bryan_: yes, if it's not forbidden by wince setup. |
00:36.26 | bryan_ | so my winmo Reg then |
00:36.48 | bryan_ | so ill be tricking haret into thinking its a PPC |
00:36.56 | bryan_ | ? |
00:37.18 | cr2 | bryan_: no. otherwise you need to sign haret with a trusted key. |
00:37.24 | *** join/#htc-linux surgex (i=surge@pool-98-118-165-224.bflony.fios.verizon.net) |
00:37.34 | cr2 | which we obviously don't have. |
00:37.46 | bryan_ | im confused . |
00:37.54 | Kevin2 | bryan_: Can you try: http://www.handhelds.org/~koconnor/haret/haret-20080926.exe |
00:38.04 | bryan_ | sure thing .:) |
00:38.19 | cr2 | the hack is to disable key check, or installing your own trusted cert and signing haret with it. == pita |
00:38.24 | Kevin2 | bryan_: It wont fix your problem - but hopefully it will give a more meaningful error message. |
00:38.58 | cr2 | Kevin2: please apply this haret patch too http://pastebin.ca/1211934 |
00:39.17 | Kevin2 | cr2: That message wasn't processed by haretconsole. What does it look like after the console massaging? |
00:39.29 | cr2 | the loox cpld gpios are write-only, traacing them does not make sense. |
00:40.15 | cr2 | Kevin2: 029.997 IRQS 5e302000(74)=400 |
00:40.46 | bryan_ | kevin2: its actually running ... Mtype is wrong ... |
00:41.03 | cr2 | Kevin2: for 029997: mem IRQS(2) 5e302000=00000400 (00000400)M |
00:41.52 | Kevin2 | cr2: Can you post the full haret log? My guess is the phone just wasn't detected as an msm7500. It should have translated that to the name. |
00:42.11 | cr2 | Kevin2: the gpio names are parsed properly |
00:42.33 | cr2 | Regs_msm7201A = { |
00:42.33 | cr2 | <PROTECTED> |
00:42.35 | cr2 | ... |
00:42.41 | cr2 | memalias.RegsList['ARCH:MSM7201A'] = Regs_msm7201A |
00:42.51 | bryan_ | http://pastebin.ca/1211944 heres the new one . |
00:42.52 | Kevin2 | bryan_: Okay - cool. You probably wont be able to do much with it though - needs unlocking to do anything really interesting. BTW, did you get a warning during haret startup? |
00:42.57 | cr2 | <PROTECTED> |
00:42.58 | cr2 | <PROTECTED> |
00:43.16 | bryan_ | kevin2: yes it cant run in system mode |
00:43.26 | *** join/#htc-linux ludoviclange (n=chatzill@c66-235-4-183.sea2.cablespeed.com) |
00:43.44 | cr2 | bryan_: unsigned haret is not trusted to switch into system mode |
00:43.50 | ludoviclange | join #handhelds |
00:44.10 | cr2 | bryan_: it's some rudimentary security on "smartphone" |
00:44.37 | Kevin2 | bryan_: Okay - thanks. You'll need to figure out how to unlock the phone. cr2 - isn't there a signed register editor out there somewhere? |
00:44.57 | cr2 | Kevin2: don#t think so |
00:45.08 | bryan_ | so its the winmo reg that is locked as is my understanding . |
00:45.14 | Kevin2 | cr2: Can you post the full log? The setup to the watch/wirq will show what haret thinks the mappings are. |
00:45.32 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
00:45.44 | cr2 | bryan_: try to read xda-developers.com forum for working with this problem. |
00:45.59 | ludoviclange | Hello, do anybody know how to mount a DiskOnChip device ? I'd like to access the DOC of my HTC Typhoon Smartphone, but don't know the address of the chip. |
00:46.05 | bryan_ | ok thanks guys . |
00:46.11 | ludoviclange | (And it is not autodetected) |
00:46.48 | cr2 | Kevin2: it's too big |
00:47.07 | cr2 | ludoviclange: i think the DOC position on typhoon is known |
00:47.14 | cr2 | Kevin2: here is the header |
00:47.18 | cr2 | Beginning memory tracing.M |
00:47.18 | cr2 | Watching IRQS(00): Addr 95300000(@c0000000)M |
00:47.18 | cr2 | Watching IRQS(01): Addr 95300004(@c0000004)M |
00:47.18 | cr2 | Watching IRQS(02): Addr 5e302000(@252d0000)M |
00:47.18 | cr2 | 00: Mapping 92300000(@a9900000) accesses to e1100000 (tbl a9902402)M |
00:47.48 | ludoviclange | cr2: good news ! Where can I search more information about it ? |
00:47.49 | cr2 | Kevin2: hmm |
00:47.57 | cr2 | <PROTECTED> |
00:48.08 | cr2 | dump mmu provides different data... |
00:48.13 | Kevin2 | Yeah - it doesn't think it is c0000000. |
00:48.34 | cr2 | ludoviclange: http://vivien.chappelier.free.fr/typhoon/index.html |
00:49.03 | cr2 | Kevin2: can you apply his 7201A patch ? maybe in modified form. |
00:49.34 | cr2 | bryan_: btw, you should read this page too |
00:49.42 | cr2 | bryan_: Make sure the following keys are set properly in your registry, using for example this signed freeware registry editor:... |
00:49.49 | ludoviclange | cr2: thanks, I know this website - actually I running linux with vivien's work. However I don't seem to find information about the DOC on his pages. |
00:50.13 | Kevin2 | cr2: It did two things - aliased msm7201a to msm7500, and added 8x8 font support. The font stuff was incomplete (he didn't diff all the files). |
00:50.18 | cr2 | ludoviclange: hmm. we even have picked the m-doc g3 driver from this page. |
00:50.36 | Kevin2 | What parts do you want me to commit? |
00:50.59 | cr2 | Kevin2: 7201A detection as msm cpu. |
00:50.59 | Kevin2 | Oh - it also added a sleep(100) after fbwrites in linboot code. |
00:51.14 | cr2 | it's a hack. |
00:51.35 | cr2 | i guess something is going wrong with disabling interrupts on msm |
00:52.28 | Kevin2 | cr2: I checked your Loox changes in. |
00:52.30 | ludoviclange | cr2 : I'll have another look there and come back. I may be blind... |
00:52.33 | cr2 | thanks. |
00:52.59 | cr2 | ludoviclange: checkout his git, i don't remember where is the driver. |
00:53.00 | Kevin2 | cr2: Sure - I'll commit the detection stuff. Can you post your haretlog.txt - I want to double check the cpu id part. |
00:53.03 | bryan_ | editing the registry to run haret. i see it now . |
00:53.33 | cr2 | Kevin2: ok, i'll redo it now, don't remember where i've put it. |
00:56.06 | Kevin2 | cr2: can you post the "dump mmu" too? I'm curious how it got the 5e302000 mapping different. |
00:56.11 | *** join/#htc-linux olesis (n=olesis@ool-4351a418.dyn.optonline.net) |
00:56.43 | cr2 | Kevin2: http://pastebin.ca/1211949 |
00:56.59 | cr2 | Kevin2: grepping |
00:57.59 | cr2 | <PROTECTED> |
00:58.03 | cr2 | <PROTECTED> |
00:58.07 | cr2 | <PROTECTED> |
00:58.08 | cr2 | <PROTECTED> |
00:58.08 | cr2 | <PROTECTED> |
00:58.08 | cr2 | 95300000 | c0000000 | 1MB section | AP=1 |
00:58.08 | cr2 | b5300000 | c0000000 | 1MB section | AP=1 |
00:58.34 | cr2 | Watching IRQS(02): Addr 5e302000(@252d0000)M |
00:58.43 | cr2 | 5e302000 | c0000000 | Extended (4K) | AP=3 |
00:59.20 | cr2 | fgrep 5e302000 haretlog-20080920_230634.logRAPH100 |
00:59.20 | cr2 | <PROTECTED> |
01:01.03 | Kevin2 | cr2: Could 5e302000 be a temporary mapping? One that can change? |
01:02.16 | Kevin2 | cr2: What does v2p(0x5e302000) report? |
01:02.35 | cr2 | Kevin2: i'll check. |
01:02.51 | bryan_ | Hey I was missing one of those reg entrys . |
01:02.59 | bryan_ | 0000101b |
01:07.27 | *** join/#htc-linux dzo (n=dzo@121.98.128.127) |
01:07.36 | cr2 | HaRET(2)# print "%x" v2p(0x5e302000) |
01:07.36 | cr2 | 25b5f000 |
01:08.18 | Kevin2 | cr2: Yeah - looks like that was just a temporary mapping. |
01:08.18 | cr2 | hmm. |
01:08.18 | cr2 | dumping mmu again |
01:08.21 | Kevin2 | What were the commands you used to add to gpios/irqs/wirq/watch. |
01:09.13 | cr2 | hmm. you are right |
01:09.15 | Kevin2 | cr2: Can you try http://www.handhelds.org/~koconnor/haret/haret-20080926b.exe -- this has just a minimalistic msm7201a / raphael detection system. |
01:09.21 | cr2 | dump mmu 2 0xc0000000 |
01:09.29 | cr2 | <PROTECTED> |
01:09.33 | cr2 | <PROTECTED> |
01:09.33 | cr2 | <PROTECTED> |
01:09.34 | cr2 | <PROTECTED> |
01:09.34 | cr2 | <PROTECTED> |
01:09.34 | cr2 | 95300000 | c0000000 | 1MB section | AP=1 |
01:09.34 | cr2 | b5300000 | c0000000 | 1MB section | AP=1 |
01:09.51 | cr2 | ok, downloading |
01:11.21 | cr2 | Detected machine Raphael/MSM7500 (Plat='PocketPC' OEM='HTC Touch Pro T7272') |
01:11.21 | cr2 | CPU is ARM ARM arch 6 stepping 2 running in system mode |
01:12.03 | cr2 | Kevin2: the problem is that 720xA has one more gpio bank5, compared to 7x00. and bank4 has one more gpio. |
01:12.37 | cr2 | it's like pxa2[56]x vs pxa27x |
01:12.57 | Kevin2 | cr2: I don't object to breaking it out to its own class. I can just do this for you while we wait for a full patch. |
01:13.10 | cr2 | ok. |
01:14.36 | bryan_ | I am no longer getting the system mode error ... but now my mtype needs to be set . |
01:15.12 | cr2 | Kevin2: ok, gpios (bank0-4) are traced and parsed properly |
01:15.49 | cr2 | Kevin2: but 'wi' is stying silent with static mapping |
01:16.43 | Kevin2 | cr2: What's the output of "wi"? |
01:18.12 | cr2 | Finished restoring windows exception handlers. |
01:18.12 | cr2 | Handled 256 irq, 2361 abort, 4166 prefetch, 0 lost, 0 errors |
01:18.26 | cr2 | i have added all mappings of 0xc0000000 now. |
01:19.21 | cr2 | ok kaiser we saw at least the lcd refresh irq |
01:19.22 | Kevin2 | cr2: What's the header of "wi" report? |
01:19.35 | cr2 | HaRET(23)# wi 10 |
01:19.35 | cr2 | irq:8001c060@A03F43F8=811420ec abort:8000104c@A03F43F0=81142110 prefetch:8001bbd0@A03F43EC=8114213c data=81101000 sizes=c:000014e4,t:000454e4 |
01:19.35 | cr2 | Beginning memory tracing. |
01:20.17 | Kevin2 | What are the lines after that? |
01:20.29 | cr2 | Watching IRQS(00): Addr 95300000(@c0000000) |
01:20.33 | cr2 | Watching IRQS(01): Addr 95300004(@c0000004) |
01:20.37 | cr2 | Watching IRQS(02): Addr 50203000(@c0000000) |
01:20.41 | cr2 | Watching IRQS(03): Addr 50305000(@c0000000) |
01:20.45 | cr2 | Watching IRQS(04): Addr 55702000(@c0000000) |
01:20.49 | cr2 | Watching IRQS(05): Addr 59402000(@c0000000) |
01:20.50 | cr2 | Watching IRQS(06): Addr 5b802000(@c0000000) |
01:20.50 | cr2 | MMU table merging disabled |
01:20.50 | cr2 | Replacing windows exception handlers... |
01:20.50 | cr2 | Finished installing exception handlers. |
01:21.02 | cr2 | looks good |
01:21.28 | Kevin2 | cr2: You don't need to have all the mappings for the IRQS variable. It's just like GPIOS - you only need one mapping. |
01:21.53 | Kevin2 | You only need to specify all the virtual mappings for mmutrace and pxa trace. |
01:22.08 | cr2 | Kevin2: hm. ok. but i don't see any changes... |
01:22.16 | cr2 | ok. that's good to know. |
01:22.27 | Kevin2 | Maybe the 7201a has the irq registers at a different phys address? |
01:22.44 | cr2 | no, this data comes from android source. |
01:23.29 | cr2 | http://git.android.com/?p=kernel.git;a=blob;f=arch/arm/mach-msm/irq.c;h=66901baf8c8e33d240c32d43af23fd485a1f5110;hb=3b55b35ea0edc29e41ac22ac0f391a5444899684 |
01:25.05 | cr2 | #define MSM_VIC_PHYS 0xC0000000 |
01:25.21 | cr2 | http://git.android.com/?p=kernel.git;a=blob;f=arch/arm/mach-msm/include/mach/msm_iomap.h;h=e6d7c9e74a93b76dd39468b1f587c881dbccb9c9;hb=3b55b35ea0edc29e41ac22ac0f391a5444899684 |
01:27.17 | cr2 | Kevin2: hmm, we are tracing these 2 registers -> |
01:27.20 | cr2 | #define VIC_INT_SELECT0 VIC_REG(0x0000) /* 1: FIQ, 0: IRQ */ |
01:27.20 | cr2 | <PROTECTED> |
01:27.29 | *** join/#htc-linux dzo (n=dzo@121.98.128.127) |
01:28.10 | Kevin2 | cr2: Yeah - don't you want STATUS0 and 1? |
01:28.32 | cr2 | yes, looks much better to me ;) |
01:28.43 | Kevin2 | (It's just a guess.) |
01:28.56 | cr2 | Kevin2: i think when we have added these registers, we just made a trace on kaiser |
01:29.00 | Kevin2 | I'm not sure anyone has ever really checked out the msm stuff that's in haret. I think it was just guesses before. |
01:29.16 | cr2 | there were no docs/code at that time. |
01:29.19 | cr2 | yes. |
01:33.03 | cr2 | ok, now it makes more sense :) |
01:33.13 | cr2 | 000.000 IRQS b5300080(232)=100 |
01:33.13 | cr2 | 000.000 IRQS b5300084(257)=2 |
01:33.28 | cr2 | time for an updated version... |
01:34.06 | cr2 | but i'm not sure it'll be the same for 7x00 |
01:35.12 | Kevin2 | Does the android source support both? |
01:35.34 | cr2 | 000.000 IRQS IRQ: IRQ8(232)=1 |
01:35.35 | cr2 | 000.000 IRQS IRQ2: IRQ2-1(257)=1 |
01:35.47 | cr2 | no, android is the 720xA |
01:36.14 | cr2 | 0x08 8 INT_DEBUG_TIMER_EXP |
01:37.11 | cr2 | 2-1 should be 0x21 33 INT_GPIO_GROUP2 |
01:37.54 | cr2 | 002.836 IRQS IRQ: IRQ19(243)=1 |
01:37.54 | cr2 | 002.840 IRQS IRQ: IRQ16(240)=1 |
01:37.57 | Kevin2 | Yeah - change regOneBits("IRQ2-") to regOneBits("IRQ", 32) |
01:38.11 | cr2 | 0x10 16 INT_MDDI_PRI + |
01:38.23 | cr2 | <PROTECTED> |
01:38.34 | cr2 | ok, this the LCD refresh |
01:38.44 | cr2 | patching. |
01:39.54 | cr2 | excellent. |
01:40.06 | cr2 | 004.926 IRQS IRQ: IRQ33(257)=1 |
01:40.06 | cr2 | Restoring windows exception handlers... |
01:40.06 | cr2 | Finished restoring windows exception handlers. |
01:41.44 | cr2 | Kevin2: it does not backtrack the gpio irqs |
01:41.58 | cr2 | because they are hidden behind irq32 and irq33 |
01:42.15 | Kevin2 | Yeah - no one knew what the gpio irq registers were. |
01:42.25 | Kevin2 | They also need to be added to IRQS. |
01:42.45 | cr2 | i think they are documented |
01:43.26 | cr2 | eh, not the status registers. |
01:46.19 | cr2 | http://git.android.com/?p=kernel.git;a=blob;f=arch/arm/mach-msm/gpio_hw.h;h=e5fc229eababb19a682a849aaa6e45cda297602b;hb=ea2e4fc53fec9008be4572fe5fa887181ddf3fef |
01:47.25 | cr2 | yes, the description is missing in wiki. but the registers are there. |
01:48.53 | cr2 | ok, it's about 4:00 am |
01:48.57 | cr2 | good night :) |
01:50.29 | Kevin2 | cr2: Should I commit the patch that does minimal detection? |
01:52.25 | *** join/#htc-linux bryan (n=bryan@206-248-131-251.dsl.teksavvy.com) |
02:13.28 | *** join/#htc-linux dzo_ (n=dzo@121.98.128.127) |
02:35.24 | *** join/#htc-linux dzo_ (n=dzo@121.98.128.127) |
02:36.48 | Kevin2 | Hi dzo. |
02:37.39 | Kevin2 | Do you know if the irq status registers for msm7x00 are at 0xc0000000? On msm7x00a it looks like they're at 0xc0000080 & 84. |
03:13.27 | *** join/#htc-linux dzo_ (n=dzo@121.98.128.127) |
04:06.04 | *** join/#htc-linux zdooder (n=troy@c-67-160-50-188.hsd1.wa.comcast.net) |
04:24.54 | *** join/#htc-linux Othello (i=Othello@gateway/tor/x-3747aff126a08ad6) |
04:38.36 | *** join/#htc-linux kikllki (n=kikllki@195-132-110-207.rev.numericable.fr) |
04:38.40 | kikllki | Hi all :) |
04:50.23 | *** join/#htc-linux snakefreak (n=s@dyndsl-091-096-105-169.ewe-ip-backbone.de) |
04:53.05 | *** join/#htc-linux dzo_ (n=dzo@121.98.128.127) |
05:45.14 | *** join/#htc-linux goxboxlive (n=goxboxli@195.159.97.196) |
07:08.00 | *** join/#htc-linux kiozen (n=oeichler@rgnb-5d87d3a1.pool.einsundeins.de) |
07:38.20 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
08:10.52 | *** join/#htc-linux rob_w (n=bob@Mb644.m.pppool.de) |
08:29.27 | *** join/#htc-linux tsdogs (n=tsdogs@net70-17.metalit.net) |
09:14.36 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
09:20.21 | *** join/#htc-linux miknix (n=miknix@bl4-92-205.dsl.telepac.pt) |
09:29.08 | *** join/#htc-linux LunohoD_ (n=alex@e180071018.adsl.alicedsl.de) |
09:38.20 | *** join/#htc-linux diogene31_ (n=rj@mur31-2-82-243-122-54.fbx.proxad.net) |
09:47.55 | *** join/#htc-linux romario (n=romario@ip-89-102-143-87.karneval.cz) |
10:02.18 | *** join/#htc-linux lkcl_ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
10:21.42 | cr2 | hi |
10:22.13 | cr2 | cmonex reported the NAND WP bit on hermes. we are moving forward :D |
10:22.21 | cr2 | bbl |
10:22.30 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
10:33.06 | *** join/#htc-linux Derek (n=opera@82-168-116-233.ip.telfort.nl) |
10:34.03 | *** join/#htc-linux pleemans (n=toi@d54C2AAB7.access.telenet.be) |
10:35.43 | Derek | A good morning/afternoon/evening or even night... Can someone explain me how to checkout the android-msm-2.6.25 branch... (of the google git) |
10:35.43 | Derek | the only thing I came up was "git checkout -b android-msm-2.6.25" but that doesn't work :) |
10:39.06 | Dubbeld00 | (Wait a sec, feeling so stupid... didn't change dir to kernel..) |
10:41.37 | *** part/#htc-linux Dubbeld00 (n=opera@82-168-116-233.ip.telfort.nl) |
10:46.02 | *** join/#htc-linux lkcl_ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
10:48.47 | *** join/#htc-linux ProstheticS (n=Prosthet@58.175.185.160) |
10:53.01 | ProstheticS | is android the only options for us kaiser folks? |
10:53.06 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
11:07.56 | *** join/#htc-linux BabelO (n=Fabrice@unaffiliated/babelo) |
11:15.50 | BabelO | http://www.project10tothe100.com/intl/EN_GB/index.html |
11:15.53 | BabelO | hi |
11:17.57 | tmzt_ | ProstheticS: ask dcordes what works now, I know we are working on X and touchscreen |
11:24.50 | ProstheticS | android already has the x and touchscreen workin, could it not be leeched from there, seeing as android is meant to be open source isnt it? |
11:24.52 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
11:28.33 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-77-107-57.dynamic.dsl.as9105.com) |
11:30.33 | tmzt_ | android is not X, and the source is not released yet for android software, only changes to the linux kernel for qualcomm test board |
11:31.34 | ProstheticS | oh |
11:31.38 | ProstheticS | right right |
11:31.43 | ProstheticS | :) i didnt claim to be smart |
11:32.15 | *** join/#htc-linux loxo1 (n=loxo8787@lns-bzn-45-82-65-191-33.adsl.proxad.net) |
11:32.24 | loxo1 | hi all :) |
11:37.36 | ProstheticS | hey loxo |
11:40.05 | loxo1 | how are u Posthetics :) |
11:40.10 | ProstheticS | im pretty good |
11:40.34 | ProstheticS | just checkin on this , hoping someone gets a good setup happening for kaisers soon :) |
11:40.35 | ProstheticS | hehehe |
11:45.44 | *** join/#htc-linux lkcl_ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
11:49.39 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
12:04.00 | NetRipper | can someone tell me what the acronym VIC stands for? |
12:04.07 | NetRipper | hard to google for VIC |
12:06.15 | Miek | voice interface card maybe |
12:07.11 | NetRipper | i think it's a lower level device, something that controls interrupts |
12:07.18 | NetRipper | within a phone |
12:07.26 | NetRipper | perhaps even part of the cpu |
12:10.36 | *** join/#htc-linux Dinde (n=kayser@sur-internet.net) |
12:11.06 | *** join/#htc-linux lkcl_ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
12:13.55 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
12:14.04 | dcordes | hi all |
12:14.50 | dcordes | ProstheticS: what tmzt_ said. In X we have two problems on the kaiser: there seem to be problems with the framebuffer and the touchscreen. |
12:16.01 | ProstheticS | oh thx dcordes , how did you know that ide asked ? lol |
12:16.07 | *** join/#htc-linux TeringTu1y (n=maarten@195-241-125-243.ip.telfort.nl) |
12:16.28 | dcordes | ProstheticS: when I have a very horrible hangover, I can go back in time |
12:17.02 | ProstheticS | can you not disable frame bufferring. |
12:17.24 | dcordes | I don't think so |
12:17.30 | dcordes | can I? |
12:17.55 | dcordes | how does android ax the fb anyway? there it works fine |
12:18.00 | dcordes | s/fb/display/ |
12:18.20 | ProstheticS | i have no idea, whatso ever |
12:18.32 | ProstheticS | ive only seen that on certain apps, frmae buffering is an option |
12:18.35 | ProstheticS | i dont code |
12:18.40 | ProstheticS | well, not that type of code anyhow |
12:18.41 | dcordes | I don't either |
12:19.03 | dcordes | trial and error / copy and paste is all I can do |
12:19.09 | dcordes | try |
12:19.12 | ProstheticS | ahh very good, we're at the same level :) |
12:19.38 | dcordes | so you also have the kaiser? |
12:19.47 | NetRipper | android also uses framebuffer afaik |
12:20.05 | ProstheticS | yah dcords i do |
12:20.11 | NetRipper | you tried the fb_dev module on X? |
12:20.11 | dcordes | hi NetRipper, groundbreaking news? |
12:20.31 | dcordes | ProstheticS: I'm trying to make the kaiser wifi work |
12:20.36 | NetRipper | dcordes, cr2 and Kevin2 made some progress yesterday, but i dont know details on what it means to getting linux to work on raph |
12:21.03 | NetRipper | yesterday being last night 4am :s |
12:21.07 | dcordes | yeaI have seen a lot of rocket science in the log |
12:21.08 | NetRipper | cet time tha tis |
12:22.10 | ProstheticS | does keyboard work yet dcordes? |
12:22.16 | NetRipper | it's not my area of expertise either, so can't draw any conclusions :) |
12:22.19 | dcordes | ProstheticS: perfectly |
12:22.26 | ProstheticS | mmmm |
12:22.41 | ProstheticS | can i get an updated cab anywhere, or everything i need to boot to console? |
12:23.12 | dcordes | http://linuxtogo.org/~lgorris/index.php?path=kaiser-bootkit/ it has a console initrd |
12:23.20 | ProstheticS | cool , ive got harett |
12:23.43 | ProstheticS | hold, r u androiding , or normal linux'n |
12:23.44 | ProstheticS | ? |
12:23.57 | dcordes | but there are advantages in using the root filesystem on sd |
12:23.58 | *** join/#htc-linux snakefreak (n=s@dyndsl-091-096-101-125.ewe-ip-backbone.de) |
12:24.05 | dcordes | angstrom.bin is linuxing |
12:24.14 | ProstheticS | right, beautiful |
12:24.17 | dcordes | I'm linux'n mostly |
12:24.34 | dcordes | http://linuxtogo.org/~lgorris/tgreer-special-sd-rootfs.tar.gz |
12:24.43 | dcordes | that's an example of a filesystem you can use on the sd card |
12:24.51 | dcordes | it's much better to do this |
12:25.01 | ProstheticS | im using root |
12:25.30 | dcordes | format your sd card that way: 1st primary partition vfat with haret.exe, default.txt, zImage, second partition ext2 with the rootfs extracted |
12:25.45 | dcordes | note the tarball has a subfolder |
12:26.28 | ProstheticS | does linux have a problem with HC SD? |
12:26.29 | dcordes | then you must rmeove the INITRD line from default.txt and set CMDLINE = "root=/dev/mmcblk0p2" |
12:26.33 | dcordes | no problem |
12:26.45 | ProstheticS | beautiful, i heard android didnt like em |
12:26.48 | dcordes | but there is a general problem with the sd on kaiser/polaris/other msm7200 |
12:27.17 | dcordes | well from the driver level it's all the same, because the two use the same kernel |
12:27.32 | dcordes | (android vs. normal userspace) |
12:28.04 | ProstheticS | yah |
12:29.09 | loxo1 | and for Diamond any news??? |
12:29.28 | dcordes | still not bootable |
12:29.46 | dcordes | diamond progress = raphael progress |
12:30.09 | loxo1 | lol i know... but Diamond = Raphael = G1 same hardware |
12:30.37 | loxo1 | the difference hare the Screen |
12:30.40 | dcordes | that's why we google should give us the g1 code soon |
12:31.11 | dcordes | -we also blackstone(touch HD) is same hardware |
12:31.15 | dcordes | all msm7201A |
12:31.20 | loxo1 | yes |
12:31.58 | loxo1 | but i the new SDK 1.0 we don't have anything to boot the kernel with msm7201a? |
12:32.01 | loxo1 | in* |
12:32.15 | dcordes | cr2: any hints on how to find out acx100 kaiser base? I could try out some regs but I don't know which. the way BabelO suggested with that atheros driver thing seems to complicated. |
12:32.41 | dcordes | loxo1: sdk* is not of interest |
12:32.44 | dcordes | when you can't boot |
12:32.53 | loxo1 | ok |
12:33.27 | ProstheticS | dcordes, is this not missing initrd for lin, or can they use same initrd? |
12:35.16 | ProstheticS | infact i think its missing initrd all together |
12:35.19 | dcordes | ProstheticS: missing initrd? sorry don't get it |
12:35.29 | dcordes | what are you trying to do? |
12:35.38 | ProstheticS | lol, dw , im stupid |
12:35.59 | dcordes | you tried to boot from the sd? |
12:36.09 | ProstheticS | from root not sd |
12:36.24 | dcordes | with the bootkit files? |
12:36.32 | ProstheticS | yeh |
12:36.56 | dcordes | if you boot the angstrom.bin you must change default.txt |
12:39.03 | dcordes | SanMehat: swetland: happy birthday hrhrhrhfr |
12:43.34 | ProstheticS | image too large? |
12:43.57 | dcordes | ProstheticS: haret says that when you boot? |
12:44.22 | ProstheticS | yar |
12:44.46 | dcordes | booting angstrom.bin? |
12:44.53 | ProstheticS | yeh |
12:45.24 | dcordes | maybe your ram is full |
12:45.36 | ProstheticS | dw found the problem |
12:45.36 | ProstheticS | ::) |
12:45.41 | dcordes | haret must upload ~17M to it |
12:51.20 | ProstheticS | ok we're all good and booting |
12:51.28 | *** join/#htc-linux dzo_ (n=dzo@121.98.128.127) |
12:52.00 | ProstheticS | damn, the whole screen sideways thing , while keyboard is open |
12:52.08 | ProstheticS | kinda uncool |
12:52.34 | dcordes | in console? |
12:52.53 | ProstheticS | yah |
12:53.02 | ProstheticS | i just managed to lock it up too :/ |
12:53.03 | ProstheticS | lol |
12:53.14 | dcordes | you can rotate the fbcon |
12:53.37 | dcordes | on boottime using cmdline or using procfs |
12:55.11 | ProstheticS | bbl |
13:00.42 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
13:26.12 | *** join/#htc-linux pronto (n=pronto@pool-71-127-145-124.bltmmd.fios.verizon.net) |
13:27.12 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
13:30.07 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
13:34.46 | *** join/#htc-linux shoragan (n=shoragan@debian/developer/shoragan) |
13:58.59 | *** join/#htc-linux amiral_ (n=amiral@gob75-2-81-56-65-10.fbx.proxad.net) |
13:59.11 | amiral_ | hello, someone know how to calculate pixel clock for lcd driver ? |
14:02.52 | *** join/#htc-linux fpletz (n=fpletz@2001:470:c041:feed:dead:beef:cafe:42) |
14:13.23 | *** join/#htc-linux LunohoD_ (n=alex@e180077146.adsl.alicedsl.de) |
14:14.44 | *** join/#htc-linux EA2 (n=chipper@cpe-024-074-138-191.carolina.res.rr.com) |
14:24.53 | *** join/#htc-linux sikkle (n=sikkle@MTLXPQAK-1176053010.sdsl.bell.ca) |
14:35.06 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
14:47.46 | *** join/#htc-linux chab7 (n=kvirc@212.92.4.114) |
14:55.21 | *** join/#htc-linux lkcl_ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
15:37.43 | kiozen | BabelO: any news? |
16:02.02 | cr2 | hi |
16:02.35 | cr2 | NetRipper: the interrupts can be traced now. |
16:02.40 | *** join/#htc-linux amd64lover (n=miknix@bl5-240-155.dsl.telepac.pt) |
16:10.55 | *** join/#htc-linux goxboxlive (n=goxboxli@139.84-48-212.nextgentel.com) |
16:12.47 | *** join/#htc-linux LunohoD_ (n=alex@e180075021.adsl.alicedsl.de) |
16:18.10 | *** join/#htc-linux BabelO_ (n=fcr@unaffiliated/babelo) |
16:23.26 | *** join/#htc-linux romak (n=romario@89-185-235-7.static.masterinter.net) |
16:45.56 | NetRipper | cr2, i read the conversation between you and Kevin2 last night, but didn't make much sense of it |
16:46.10 | NetRipper | cr2, you have other interrupt addresses now or was haret patched to make it work? |
16:47.36 | NetRipper | cr2, and if there are new irq addresses, should haret properly disable them when attempting to boot? |
16:48.12 | NetRipper | because in that code i couldn't really notice that it refers to the "irqs" variable |
16:48.15 | NetRipper | although i might have missed it |
16:48.17 | *** join/#htc-linux oxman (n=bouh@sentinel.bouh.org) |
16:48.20 | oxman | hi :) |
16:49.11 | oxman | how can i help you with the hermes ? |
16:49.50 | *** join/#htc-linux oohnoitz (i=prinny@adsl-065-005-208-086.sip.msy.bellsouth.net) |
16:50.36 | NetRipper | oxman, if you have kernel hacking experience you'll probably want to talk with Kevin2 or cr2, they'll know if you can help :) |
16:50.50 | NetRipper | both of them afk though |
16:51.12 | NetRipper | i gtg, later |
16:54.13 | oxman | ok :) |
16:54.14 | oxman | thanks |
16:54.26 | cr2 | NetRipper: the irq addresses should be changed, and the .py adjusted |
16:55.15 | cr2 | oxman: you can dump the NAND reg, and the UART regs. the we will try to decode their meaning |
16:55.30 | cr2 | oxman: with IR, BT and phone switched on. |
16:56.34 | oxman | and how can id do that ? |
16:56.49 | *** join/#htc-linux tgreer (n=tgreer@titus.tsone.info) |
16:56.56 | tgreer | geeksoc |
16:56.58 | tgreer | damn this irc client |
16:57.00 | tgreer | lol hey all |
16:57.49 | oxman | -id+i |
17:00.13 | cr2 | oxman: run haret and telnet to it |
17:02.37 | oxman | ok |
17:03.15 | *** join/#htc-linux dcordes (n=dcordes@unaffiliated/dcordes) |
17:03.22 | tgreer | hey dcordes |
17:03.34 | *** join/#htc-linux SmallR2002 (n=SmallR20@79-77-60-225.dynamic.dsl.as9105.com) |
17:04.07 | *** join/#htc-linux LunohoD_ (n=alex@e180073169.adsl.alicedsl.de) |
17:06.16 | oxman | hum, haret hang the phone :) |
17:07.07 | cr2 | which haret are you using ? |
17:07.40 | oxman | the last, haret-20080926b.exe |
17:08.54 | dcordes | one blue cable, one yellow cable: which one has +, which mass? |
17:11.02 | cr2 | oxman: strange. do you have earlyharetlog.txt in the same directory ? |
17:11.27 | oxman | no one |
17:11.38 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
17:11.45 | oxman | hum nevermind |
17:11.50 | oxman | i have the zImage of the kaiser |
17:11.55 | oxman | I download the zImage of the hermes :) |
17:12.33 | *** join/#htc-linux nebi (n=nebi@c-498fe253.02-145-7570701.cust.bredbandsbolaget.se) |
17:15.26 | oxman | hum, what default.txt should contain ? |
17:19.36 | oxman | maybe is better to use herm-linux instead of haret no ? |
17:21.30 | oxman | hum its better now (still with haret) |
17:22.16 | oxman | But now : Kernel panic -not syncing: VFS: Unable to mount root fs on unknown-block(1,0) |
17:22.17 | oxman | :D |
17:22.52 | tgreer | win |
17:24.30 | *** join/#htc-linux Dinde (n=kayser@sur-internet.net) |
17:24.52 | cr2 | oxman: i mean to run haret in wince, not booting linux |
17:36.31 | oxman | yep, but haret ask a default.txt ;) |
17:37.44 | *** join/#htc-linux TripleDES (n=sergio@89.128.236.151) |
17:38.01 | *** join/#htc-linux TripleDES (n=sergio@89.128.236.151) |
17:41.10 | *** part/#htc-linux tgreer (n=tgreer@titus.tsone.info) |
17:45.49 | *** join/#htc-linux bryan__ (n=bryan@206-248-131-251.dsl.teksavvy.com) |
17:59.45 | Marajin | yoes |
18:08.37 | *** join/#htc-linux ali1234 (n=al@62.24.214.38) |
18:14.10 | *** join/#htc-linux pleemans (n=toi@d54C2AAB7.access.telenet.be) |
18:22.36 | *** join/#htc-linux sikkle (n=sikkle@modemcable132.57-201-24.mc.videotron.ca) |
18:31.06 | *** join/#htc-linux amd64lover (n=miknix@bl5-240-155.dsl.telepac.pt) |
18:38.03 | *** join/#htc-linux JackSenechal (n=jack@204.14.59.190) |
18:43.17 | JackSenechal | I've managed to get Qtopia up and running on my BlueAngel phone, but when the "today" screen loads up, nothing works. I can press the 4 icons at the bottom and they light up, but nothing happens. I can hold down the power button and it brings up the shutdown menu, and from there I can browse help, but nothing else useful. Anyone know what's going on? |
18:52.58 | JackSenechal | here's a better question... has anyone got any version of linux running and usable on the blueangle? qtopia is totally busted when I boot it, GPE is so buggy it's unusable, and I can't find an Opie dist... |
18:54.48 | BabelO | JackSenechal: it is normal, because modem is setup and you have no sim card inside ... it take a litle timeout to get full access to qtopia interface in this case |
18:55.31 | BabelO | JackSenechal: for GPE try AlGe image available on hos servern, look at the xda thread |
18:55.34 | JackSenechal | so... I just need to boot up and wait for a long time? |
18:55.48 | BabelO | JackSenechal: yes if you have no sim |
18:55.58 | BabelO | not so long, aroubd 3 or 4 mn |
18:57.19 | JackSenechal | hmmm.... I waited that long, for sure. The only other thing that happens is after a while and pressing some buttons, the dial screen comes up. But it's way too big for the screen so I can only see a few buttons. |
19:01.24 | JackSenechal | I believe the GPE image I used is the one from AlGe. It boots up and runs, but whenever I start to mess with the programs it crashes and I've got to hard-reset |
19:01.24 | BabelO | JackSenechal: else edit qpe script in /etc/init.d and set comment out line MODEM for ba |
19:01.28 | BabelO | i have to go |
19:01.41 | BabelO | bbl |
19:01.41 | JackSenechal | ok, I'll try that |
19:01.43 | JackSenechal | thx |
19:11.14 | NetRipper | oxman, when you start haret you can choose "listen for connection" |
19:11.22 | NetRipper | when you click that, you can telnet to port 9999 |
19:11.25 | NetRipper | that's what cr2 meant |
19:11.26 | NetRipper | :) |
19:11.28 | *** join/#htc-linux Sundial (n=Sundial@unaffiliated/sundial) |
19:11.35 | oxman | Oh, well, thanks :) |
19:12.10 | JackSenechal | regarding my blueangel qtopia issues, I don't see a MODEM line in /etc/init.d/gpe, anyone know if there's something I should add there or what? |
19:13.47 | dcordes | cr2: are you still around? |
19:17.57 | NetRipper | cr2, what are the new irq addresses? i don't see them on wiki yet (unless im blind) |
19:18.26 | Kevin2 | NetRipper: 0xc0000080 and 0xc0000084 |
19:18.36 | NetRipper | Kevin2, ok |
19:18.49 | NetRipper | Kevin2, when i update the irqs list, does the take_control() use those values then? |
19:18.55 | NetRipper | in linboot.cpp |
19:19.11 | NetRipper | or is that list only for tracing? |
19:19.32 | Kevin2 | take_control() turns off irqs at the cpu level - it works regardless of the platform you're using. |
19:20.08 | NetRipper | Kevin2, well it doesn't work on raphael, i've seen my volume icon redrawn and Sleep() works after take_control(), which shouldn't if irqs are disabled |
19:20.16 | NetRipper | unless there's another explanation for that |
19:20.58 | Kevin2 | NetRipper: Calling into windows after disabling interrupts is a bad thing to do. It doesn't mean windows will hang though - Windows might just reenable them. |
19:21.48 | NetRipper | if i dont place the Sleep() the framebuffer is not drawn, the screen is never blanked |
19:22.08 | NetRipper | it just hangs then on the haret wince screen saying go go go |
19:23.13 | Kevin2 | Yeah - the msm has some framebuffer optimization thing. You need to tell the vga chip to reread the frame buffer. That's different from most other chips - they typically always reread the framebuffer and auto display any changes. |
19:23.27 | NetRipper | ah |
19:24.31 | NetRipper | so the fbwrite.cpp should be adjusted sometime to include that |
19:24.36 | NetRipper | if possible |
19:24.48 | Kevin2 | It's definitely not valid to call Sleep() when interrupts are disabled - windows will probably reenable them - and then you'll get race conditions during the boot. |
19:25.15 | NetRipper | ok, i know it wasn't valid to call Sleep() but i really wanted to see where it hung |
19:25.23 | Kevin2 | NetRipper: Yes - the command to tell the msm to refresh the FB needs to be found. Then fb_printf() should make sure to do that. |
19:25.24 | NetRipper | i can see it gets to at least the "in preloader" message |
19:25.47 | NetRipper | would i be looking for a gpio for that or something else? |
19:26.19 | Kevin2 | NetRipper: I would guess it is in the msm lcd registers - but I don't really know. |
19:26.26 | NetRipper | ok |
19:27.39 | NetRipper | i'll see if i can find something about that in the kernel code tomorrow... don't have much experience but i sure do love this low level stuff :P |
19:28.01 | NetRipper | i gtg now, later all |
19:28.05 | NetRipper | thanks for the info Kevin2 |
19:28.23 | Kevin2 | no problem - later. |
19:39.14 | *** join/#htc-linux tgreer (n=tgreer@cpc1-broo2-0-0-cust698.renf.cable.ntl.com) |
19:46.05 | *** part/#htc-linux NAiL (i=repvik@nslu2-linux/pdpc.active.NAiL) |
20:04.04 | cr2 | Kevin2: hi |
20:04.28 | cr2 | Kevin2: use gx.dll api ? i think it's deprecated... |
20:05.15 | cr2 | directx needs wince5 afair |
20:08.03 | cr2 | http://msdn.microsoft.com/en-us/library/aa454897.aspx |
20:09.32 | cr2 | hmm |
20:09.34 | cr2 | #include <d3dm.h> |
20:09.55 | cr2 | i'm not quite sure that cegcc supports it |
20:10.54 | cr2 | lol |
20:10.57 | cr2 | Figure 2. A Direct3D Mobile device is created with a blue screen |
20:15.45 | Kevin2 | Hi cr2. |
20:15.52 | Kevin2 | Not sure what you mean by gx.dll api. |
20:16.30 | cr2 | Kevin2: we can override fb_print() with a directx or gapi version. if it's available. |
20:16.57 | Kevin2 | No - fb_printf must be able to run from the preloader context. |
20:17.19 | cr2 | writing into framebuffer ram will not be always successful |
20:17.39 | cr2 | old weird configs like jornada820 will fail |
20:17.39 | Kevin2 | Agreed. However, it's a useful tool when it does work. |
20:18.14 | cr2 | and the new MDDI high-speed serial panels need CPU/irq support for the framebuffer |
20:18.24 | cr2 | to work |
20:19.07 | *** join/#htc-linux ltxda (n=anon@unaffiliated/ltxda) |
20:19.48 | cr2 | maybe it's not a good idea to use framebuffer output on such devices |
20:20.19 | cr2 | the problem is that there is nothing but flashlight on raphael which can be used ;) |
20:20.23 | Kevin2 | The last message we write when Wince is still around is "Go Go Go..." - then we kill windows. The fb_printf stuff is all about sending diagnostic info after we've killed wince. It wont work on all platforms, but it is useful on platforms where it does work. |
20:20.39 | cr2 | and on the 'touch hd' there is no flashlight... |
20:20.59 | Kevin2 | No vibrator? :-) |
20:21.14 | cr2 | vibrator is an rpc/ipc call to arm9 |
20:21.26 | cr2 | it's not a gpio |
20:21.45 | cr2 | so it's as complex as playing sound. |
20:21.53 | cr2 | or even more |
20:22.21 | dcordes | cr2: Im on linux eeepc 1000 |
20:22.23 | cr2 | there is no seria l port |
20:22.30 | cr2 | and there are no leds |
20:22.33 | Kevin2 | Same thing with IR? |
20:22.51 | cr2 | IR is not used anymore |
20:23.01 | cr2 | since 2 years or so |
20:23.06 | dcordes | cr2: it comes with 32gb flash disk same price as windows version with big hdd |
20:23.45 | cr2 | and the BT port is such an ipc PITA on msm7x0xx that even androids do not make it run completely on G1 |
20:24.16 | cr2 | dcordes: hmm., maybe i'd just take 900A |
20:24.47 | cr2 | Kevin2: 4mbit BT became as complex as an usb port now. |
20:25.15 | Kevin2 | Sounds like they made their hardware with an application processor and an IO processor. I guess you can't do any real IO from the application processor. |
20:25.16 | cr2 | because it uses dma |
20:25.31 | Kevin2 | Might need to reverse engineer the "IO processor" then too. :-( |
20:25.40 | cr2 | the flashlight is still on the arm11 gpio. |
20:26.01 | cr2 | Kevin2: it's an elf image with 14 sections |
20:26.11 | Kevin2 | And so is the LCD? It may just require more sophisticated debugging code. |
20:26.12 | cr2 | 2 sections are thumb code |
20:26.37 | cr2 | the LCD uses MDDI |
20:26.53 | cr2 | it's like having an SATA controller for the LCD ;-) |
20:27.29 | Kevin2 | Well, I don't know about MDDI, but I do know you can drive a SATA controller in PIO mode. :-) |
20:27.41 | dcordes | cr2: the keys are too tiny\ |
20:27.46 | cr2 | mddi driver in haret ? |
20:27.54 | cr2 | dcordes: in eeepc ? |
20:28.29 | cr2 | Kevin2: the MDDI code in android kernel is _very_ simple. compared to what i can see in the SPL |
20:28.35 | dcordes | yes. 1000 series is ok. but I think on the smaller models they are too small |
20:29.04 | Kevin2 | cr2: haret diagnostics aren't that important. Much more important would be to get info from Linux during its early bootup. |
20:29.51 | cr2 | Kevin2: but i'll certainly document it. it's still a bit easier than the ATI W228x setup. |
20:30.33 | cr2 | Kevin2: i have no idea how we will get any debug response from 'touch hd' |
20:30.54 | dcordes | cr2: ginge has a serial debug console on kaiser using the keyboard led |
20:31.23 | cr2 | the only led on raphael (keyboard backlight) is controlled over i2c. like the i2c-cpld on many omap850 htc phones |
20:31.31 | Kevin2 | cr2: is there any good documentation on mddi out there? |
20:32.05 | cr2 | Kevin2: i have not found it. But Qualcomm submitted it to VESA as on open standard. |
20:32.18 | cr2 | so it must be published |
20:32.30 | cr2 | i#m sure it's like OBD-2 |
20:33.04 | cr2 | where you have some "generic" part, and "vendor-specific" part, which one needs to RE anyway ;-) |
20:33.35 | cr2 | OBD-2 must be also a SAE/ISO standard |
20:34.08 | *** join/#htc-linux rmoravcik (n=rmoravci@ip-89-102-255-171.karneval.cz) |
20:34.35 | cr2 | http://www.qctconnect.com/products/mddi.html |
20:36.19 | dcordes | 2.6.21.4-eeepc |
20:36.21 | dcordes | lol |
20:36.40 | cr2 | http://news.ecoustics.com/bbs/messages/10381/198227.html |
20:38.27 | Kevin2 | cr2: http://www.vesa.org/Standards/summary/2008_7.htm -- Looks like $350 to buy the "standard". |
20:39.26 | cr2 | yeah, it's like NMEA for GPS |
20:40.57 | cr2 | i think this controller is used: http://www1.toshiba.com/taec/Catalog/Line.do?familyid=15&lineid=900069 |
20:42.02 | cr2 | it comes combined with an SPI and I2C controller for configuring the panels |
20:42.13 | cr2 | depending on the phone SPI or i2c may be used |
20:42.32 | cr2 | i think titan and kaiser use spi, and raphael uses i2c |
20:44.43 | Kevin2 | cr2: Do you have a link to the android driver for mddi? |
20:46.31 | cr2 | looking |
20:47.14 | cr2 | mddi -> http://git.android.com/?p=kernel.git;a=blob;f=drivers/video/msm/mddi.c;h=7bbcb285731f0990a0ab43845ea7947504d83660;hb=android-msm |
20:47.42 | cr2 | and mddi_client_toshiba |
20:47.44 | cr2 | http://git.android.com/?p=kernel.git;a=blob;f=drivers/video/msm/mddi_client_toshiba.c;h=268ed76bf0a0256fcbd8795138421538fbd11e4f;hb=android-msm |
20:48.29 | cr2 | the toshiba driver is too simple. SPL sets much more more registers |
20:49.56 | *** join/#htc-linux Zoolooc (n=fredsiba@nrbg-4dbfcc65.pool.einsundeins.de) |
20:52.53 | *** join/#htc-linux lpotter (n=ljp@CPE-124-191-144-181.vic.bigpond.net.au) |
20:54.10 | Kevin2 | Hrmm. How does the above interface with the linux framebuffer code? (Or does it?) |
20:54.49 | Kevin2 | I have to run though - bbl. |
20:54.59 | dcordes | Kevin2: msm video driver is drivers/video/msm/msm_fb.c in the same tree |
20:55.36 | dcordes | \ |
20:56.55 | cr2 | qtopian.net :) |
20:57.52 | *** join/#htc-linux sikkle (n=sikkle@modemcable132.57-201-24.mc.videotron.ca) |
20:59.49 | cr2 | hehe. nice code. |
20:59.54 | cr2 | <PROTECTED> |
20:59.54 | cr2 | <PROTECTED> |
20:59.54 | cr2 | <PROTECTED> |
20:59.54 | cr2 | <PROTECTED> |
21:00.38 | cr2 | dcordes: and you try to use that on kaiser ... |
21:01.06 | *** join/#htc-linux marmotta (n=skodde@unaffiliated/skodde) |
21:01.58 | dcordes | cr2: huh |
21:02.19 | cr2 | what is gpio97 on kaiser ? |
21:03.00 | cr2 | 97 0x61 1 2 I always active; with setlcd 3 silent |
21:03.14 | cr2 | may be vsync :) |
21:03.45 | dcordes | ... |
21:04.37 | cr2 | on raphael |
21:04.40 | cr2 | 97 0x61 2 stream |
21:04.55 | cr2 | need to check setlcd 3 |
21:05.30 | dcordes | what does stream signify? you see it all the time in haret? |
21:05.32 | cr2 | it's also important to know if it's an ALT gpio func |
21:05.43 | cr2 | yes, data flood |
21:06.25 | cr2 | dcordes: can you change the VIC tracing regs for kaiser ? |
21:06.37 | *** join/#htc-linux sikkle (n=sikkle@modemcable132.57-201-24.mc.videotron.ca) |
21:06.47 | dcordes | VIC? |
21:07.10 | cr2 | dcordes: they do not work on raphael, and the 'kaiser' settings do not make sense , looking at android kernel |
21:07.15 | cr2 | interrupt controller |
21:07.27 | dcordes | in that py file where the additional irq regs are set? |
21:07.34 | cr2 | 0xc0000000 and 0xc0000004 |
21:07.39 | cr2 | are now |
21:07.44 | cr2 | but should be |
21:07.51 | cr2 | 0xc0000080 and 0xc0000084 |
21:07.59 | cr2 | in .py and also in .cpp |
21:08.20 | cr2 | at tleast for A cpu it is so. probably it's the same for non-A cpu too |
21:09.57 | *** join/#htc-linux zdooder (n=troy@c-67-160-50-188.hsd1.wa.comcast.net) |
21:10.36 | dcordes | <PROTECTED> |
21:10.44 | dcordes | need to change localhost bbiab |
21:10.49 | cr2 | interesting. in android it's vsync irq. |
21:10.51 | cr2 | ok |
21:11.47 | SanMehat | dcordes: hey :) |
21:12.52 | cr2 | SanMehat: i'm looking at the LCD code, and it looks too simple in your kernel :) |
21:13.39 | SanMehat | cr2: oh? :) |
21:15.10 | cr2 | <PROTECTED> |
21:15.10 | cr2 | <PROTECTED> |
21:15.43 | cr2 | unfortunately (?) we need special initialization code... |
21:20.59 | cr2 | Kevin2: http://www.semicon.toshiba.com.cn/event/exhibition/panel_pdf/vegamagiq.pdf |
21:22.43 | *** join/#htc-linux ali1234 (n=al@62.24.214.38) |
21:24.57 | *** join/#htc-linux sikkle (n=sikkle@modemcable132.57-201-24.mc.videotron.ca) |
21:30.29 | dcordes | cr2: which haret shall I use? the latest git? |
21:32.00 | dcordes | hm it would be nice to have a recipe for haret and haretconsole git in OE. so one could ship them with the images |
21:35.53 | dcordes | cr2: can't find cpp files |
21:37.06 | dcordes | c./mach/arch-msm.cpp: "addlist irqs p2v(0xc0000004) 0 32 0\n" |
21:37.15 | cr2 | yes |
21:37.29 | cr2 | should be 84 here, and 80 above |
21:37.42 | cr2 | on A |
21:37.52 | cr2 | please check if it's the same on non-A |
21:37.58 | dcordes | ok |
21:38.18 | dcordes | the registers are commented with questions |
21:39.40 | dcordes | like that? |
21:39.41 | dcordes | <PROTECTED> |
21:39.41 | dcordes | <PROTECTED> |
21:40.24 | cr2 | yes |
21:40.30 | cr2 | and change the .py too |
21:40.48 | cr2 | then you need to change the IRQ2- string in .py |
21:42.02 | cr2 | [03:37] <Kevin2> Yeah - change regOneBits("IRQ2-") to regOneBits("IRQ", 32) |
21:42.13 | dcordes | Regs_msm7500 should better be Regs_msm7x00 |
21:42.32 | cr2 | msm7x00x |
21:42.52 | cr2 | msm7x0xx is even better :) |
21:43.12 | dcordes | aren't many registers different on msm7x00A vs msm7x00? |
21:43.56 | cr2 | yes, but it's like pxa26x and pxa27x |
21:44.41 | dcordes | the gist remains the same? |
21:45.47 | dcordes | what about the IRQ2- string? |
21:45.54 | dcordes | <PROTECTED> |
21:45.54 | dcordes | <PROTECTED> |
21:47.55 | cr2 | <PROTECTED> |
21:50.21 | dcordes | cr2: only that line? |
21:50.27 | cr2 | yes |
21:50.32 | dcordes | ok done? |
21:50.38 | cr2 | we will add gpio irqs later |
21:50.49 | cr2 | run it, and check 'wi' |
21:50.56 | dcordes | ok |
21:51.04 | cr2 | if wi traces the interupts and mmutrace |
21:51.18 | dcordes | I never compiled haret, where can I get the required toolchain? |
21:54.11 | *** join/#htc-linux lkcl__ (n=lkcl@87.113.7.225.plusnet.pte-ag1.dyn.plus.net) |
21:54.45 | cr2 | cegcc.sf.net |
21:55.01 | cr2 | dcordes: you may pick the version of Kevin2 anyway |
21:56.32 | cr2 | but there is no new version of Kevin2 |
21:56.54 | dcordes | I will use the marbalon version ^^ |
22:01.45 | cr2 | you need your own |
22:01.55 | cr2 | or i can upload mine. |
22:02.21 | cr2 | hm. i used the version of Kevin2 |
22:02.45 | cr2 | haret-20080926b.exe |
22:03.23 | dcordes | http://baliniak.pl/android/cegcc.tar.gz |
22:03.38 | dcordes | google returned it for cegcc |
22:04.02 | cr2 | ok |
22:04.18 | cr2 | if you use Kevin2's version, you need to add the 80 and 84 manually. |
22:04.23 | cr2 | they are not compiled in |
22:04.38 | dcordes | huh? I thought I'm building my own haret |
22:07.20 | cr2 | yes |
22:07.26 | cr2 | it's better |
22:11.53 | dcordes | cegcc is huge |
22:14.57 | dcordes | cr2: I have no clue what to do now |
22:17.14 | cr2 | run make in the haret dir |
22:20.22 | dcordes | it needs arm-wince-mingw32ce-g++ where do I get it? I can't see it in the cegcc tree |
22:24.50 | tmzt_ | dcordes: do you have /opt/mingw32ce/bin/arm-wince-mingw32ce-gcc ? |
22:26.40 | cr2 | tmzt_: btw, i could not compile cegcc svn |
22:26.53 | cr2 | so i have an old version |
22:27.01 | cr2 | and need to comment out sprintf() |
22:28.37 | tmzt_ | I think I had the old version from sf.net, I just have that /opt tree now |
22:29.05 | *** part/#htc-linux JackSenechal (n=jack@204.14.59.190) |
22:31.02 | dcordes | tmzt_: no, where can I get it? |
22:35.53 | cr2 | tmzt_: where did you get your ? |
22:36.06 | tmzt_ | search google "cegcc heanet" |
22:36.11 | cr2 | dcordes: then get Kevin2's version, and add the regs manually |
22:36.18 | tmzt_ | I don't know, it seems to have disappeared |
22:45.49 | dcordes | cr2: which build? have a link? |
22:49.49 | tmzt_ | cegcc-mingw32ce-0.50-1.i586.rpm, use alien -i |
22:56.54 | cr2 | dcordes: hm. it'll see both. |
22:57.04 | cr2 | tmzt_: can you compile a new haret version ? |
22:57.31 | tmzt_ | I did a month ago, I can try it if you need me to |
22:59.46 | *** join/#htc-linux Tinyboom (n=nahh@ti131310a340-1463.bb.online.no) |
23:04.36 | *** join/#htc-linux oohnoitz (i=prinny@adsl-065-005-208-086.sip.msy.bellsouth.net) |
23:07.45 | cr2 | 169.254.53.135 |
23:17.55 | bryan__ | Hey all .. |
23:27.19 | Kevin2 | cr2: Should I check the changes I made into cvs? |
23:30.25 | cr2 | Kevin2: yes |
23:30.34 | cr2 | Kevin2: then we can test them |
23:30.53 | cr2 | dcordes: still here ? |
23:32.24 | cr2 | trying to rebuild cegcc again |
23:34.52 | dcordes | yep |
23:35.46 | cr2 | dcordes: Kevin2 will patch haret |
23:35.55 | Kevin2 | http://www.handhelds.org/~koconnor/haret/haret-20080927.exe and http://www.handhelds.org/~koconnor/haret/haretconsole-20080927.tar.gz |
23:35.58 | dcordes | ok |
23:36.15 | Kevin2 | I've checked in the alias of 7201 to 7500, but I have not checked in the change to add 0x80. |
23:36.50 | cr2 | Kevin2: can you compile a test version with 0x80 for dcordes ? |
23:36.59 | cr2 | Kevin2: for me it worked. |
23:37.26 | cr2 | Kevin2: how do add the gpio irqs ? the status registers are know. |
23:37.43 | Kevin2 | cr2: the links above have 0x80 added. |
23:37.47 | dcordes | cr2: btw how cvan I proceed with the acx100 register? |
23:37.55 | Kevin2 | I just didn't commit it to cvs. |
23:38.51 | dcordes | gtreat |
23:39.13 | Kevin2 | cr2: Are the status registers the usual thing - a bit turns on when an irq is pending for the gpio? If so, just add the addresses to the IRQS list. |
23:40.13 | cr2 | Kevin2: i think so. |
23:40.26 | cr2 | Kevin2: and regOneBit for decode ? |
23:41.01 | Kevin2 | cr2: Yes. |
23:41.57 | cr2 | ok |
23:44.48 | cr2 | dcordes: can you trace interrupts and do mmutrace with the new version ? |
23:49.57 | cr2 | <PROTECTED> |
23:51.34 | cr2 | gcc-4.3 does not like some macro |
23:52.16 | dcordes | cr2: if it's the path to world domination, ok |
23:53.32 | tmzt_ | cr2: did you try that package? |
23:55.22 | tmzt_ | http://ftp.heanet.ie/disk1/sourceforge/c/ce/cegcc/cegcc-mingw32ce-0.50-1.i586.rpm |
23:56.24 | cr2 | 0.50 seem to be old ? |
23:56.54 | cr2 | ./gt-arm.h:40:1: error: unterminated argument list invoking macro "XEXP" |
23:57.36 | tmzt_ | I have gcc version 4.1.0 |
23:58.05 | Kevin2 | I use cegcc-mingw32ce-0.51.0-1.i586 |
23:58.18 | tmzt_ | where did you find that? |
23:59.07 | dcordes | 000.000 IRQS 94e00080(0 1)=3 |
23:59.08 | dcordes | 000.000 IRQS 94e00084(32 33)=3 |