| 02:06.04 | *** join/#htc-linux elysion_ (n=elysion@k203c.lanhovi.tontut.fi) |
| 02:54.10 | *** join/#htc-linux cyrill62_ (n=cyrill62@128.250.146.195.dynamic.adsl.abo.nordnet.fr) |
| 03:26.46 | *** join/#htc-linux nebula (n=ircap8@86.63.201.129) |
| 03:27.01 | nebula | helo |
| 03:27.18 | nebula | how can boot in wm5 the linux ? |
| 03:27.31 | nebula | i have xanadus |
| 03:27.41 | nebula | and blueangel |
| 04:22.47 | *** join/#htc-linux tudenbart (n=willi@xdsl-213-196-250-155.netcologne.de) |
| 04:36.22 | *** join/#htc-linux xia (n=abel@213.87.72.117) |
| 05:54.01 | *** join/#htc-linux amenbrk (n=mac@68-189-78-144.dhcp.wscr.ca.charter.com) |
| 05:55.39 | amenbrk | i am running familiar 2.6.15 with GPE on an hp 4700, trying to get the wifi device working |
| 05:56.17 | amenbrk | i added some BIN files to the firmeware directory and reflashed the kernel, i am also told there is a kernel module i need to load (i guess it is not loaded automatically) |
| 05:56.42 | amenbrk | so i used modprobe to load it, i can see the device if i use iwconfig, but i cant get it to associate with anything, and the GPE wireless utulity is no help |
| 06:18.24 | *** join/#htc-linux goxboxlive (n=goxboxli@206.80-202-161.nextgentel.com) |
| 06:32.16 | *** join/#htc-linux RoEn_PC (n=roen@p54A65197.dip.t-dialin.net) |
| 06:32.47 | goxboxlive | good morning everyone |
| 06:32.57 | RoEn_PC | moin |
| 06:35.08 | *** join/#htc-linux psokolovsky_ (n=psokolov@favoured-food.volia.net) |
| 06:59.25 | *** join/#htc-linux pH5 (n=ph5@e178231156.adsl.alicedsl.de) |
| 07:34.01 | goxboxlive | hi pH5 |
| 07:34.13 | pH5 | hey goxboxlive |
| 07:34.19 | goxboxlive | i amlooking at your gsmd patches |
| 07:34.41 | goxboxlive | Universal doesnt support at%cpi |
| 07:36.22 | goxboxlive | hmm maybe it does, it just hangs on "at%cpi=? |
| 07:36.37 | pH5 | at%cpi enables unsolicited "call progress indicator" messages - "SYNC", "CONNECTING", "DISCONNECTING", etc. |
| 07:36.53 | pH5 | we need a custom vendor plugin for the universal modem. |
| 07:36.56 | goxboxlive | Maybe that is why i allways get my answer machine when i try to call my self |
| 07:37.06 | goxboxlive | right |
| 07:37.16 | goxboxlive | i'll check out cpi with uucp |
| 07:37.53 | pH5 | goxboxlive: that'd be great. also, you could run gsmd from another console to see if there are problems with any of the other init commands that gsmd issues. |
| 07:39.06 | goxboxlive | normaly just to have phone up and go with uusp i use only cfun and cops |
| 07:40.27 | pH5 | I think openmoko use the %cpi messages display some info in the UI. |
| 07:40.35 | pH5 | can you issue ATPI? |
| 07:40.46 | pH5 | s/ATPI?/AT%CPI?/ |
| 07:42.35 | goxboxlive | i dont think so, i'll try now |
| 07:42.46 | goxboxlive | When i use it under gsmd it just hangs |
| 07:43.50 | pH5 | goxboxlive: that could well be a bug in gsmd. let's try AT%CPI=? and ATPI? directly. |
| 07:44.04 | pH5 | it shouldn't hang the modem. |
| 07:48.09 | goxboxlive | atpi works |
| 07:48.24 | goxboxlive | but not atcpi=? |
| 07:48.42 | pH5 | uh, ugly. |
| 07:48.44 | goxboxlive | or any combination with cpi |
| 07:48.52 | pH5 | what values can yous set it to? |
| 07:49.02 | *** join/#htc-linux rmoravcik (n=rmoravci@adsl-dyn37.91-127-27.t-com.sk) |
| 07:49.39 | pH5 | harald just wanted to set cpi to the highest possible value, thus the at%cpi=? check. |
| 07:49.55 | pH5 | the neo modem supports up to 3, the magician modem only up to 2 |
| 07:50.07 | goxboxlive | ok |
| 07:50.18 | goxboxlive | and uni dosent support it at all |
| 07:50.31 | goxboxlive | it might use something else |
| 07:50.47 | goxboxlive | I'll searc for a substitute |
| 07:52.12 | pH5 | does AT%CPI=? return ERROR? |
| 07:52.39 | goxboxlive | Yes, not error but in cu it gives me "4" |
| 07:52.49 | goxboxlive | and 4=error |
| 07:53.53 | pH5 | ah ok, you're just not in verbose mode. for some reason gsmd issues ATE0V1 first and then parses verbose strings instead of the V0 numbers. |
| 07:54.27 | goxboxlive | ?? |
| 07:56.10 | pH5 | goxboxlive: if you issue ATV1 and then AT%CPI=?, does it return ERROR instead of 4? |
| 07:56.21 | goxboxlive | hold on |
| 07:58.12 | goxboxlive | at%cpi=? |
| 07:58.19 | goxboxlive | ERROR |
| 07:58.55 | pH5 | ok, this works like expected. gsmd currently 'hangs' (stops to issue commands) if there is an error in the vendor initialization sequence. |
| 07:59.18 | goxboxlive | ok |
| 07:59.34 | goxboxlive | so that's why it hangs |
| 07:59.45 | pH5 | we need to fix that, and even better write a vendor plugin for the universal. |
| 07:59.55 | goxboxlive | But shouldt also universal hve a call in progress at command (cpi)? |
| 07:59.56 | pH5 | does universal support AT%CSQ |
| 08:00.44 | goxboxlive | at+csq=? |
| 08:00.44 | goxboxlive | +CSQ: (0-31,99),(99) |
| 08:00.46 | goxboxlive | OK |
| 08:00.52 | goxboxlive | not % but with + |
| 08:02.26 | pH5 | yeah. but %csq (or %htccsq) is unsolicited. at+csq only replies when you ask it. |
| 08:02.51 | pH5 | what is the xda-dev wiki page with all the htc at commands? |
| 08:03.10 | pH5 | ah, HTC_GSM |
| 08:03.20 | goxboxlive | yes |
| 08:04.32 | goxboxlive | at%csq=? |
| 08:04.33 | goxboxlive | ERROR |
| 08:04.59 | pH5 | at%htccsq=? |
| 08:05.10 | pH5 | hey, wait |
| 08:05.15 | pH5 | HTC_GSM has the following: |
| 08:05.19 | pH5 | AT@HTCCSQ=1 |
| 08:05.54 | goxboxlive | at%htccsq=? |
| 08:05.54 | goxboxlive | ERROR |
| 08:06.12 | pH5 | perhaps AT@HTCCSQ=?, AT@CPI=? or AT@HTCCPI=? does something? |
| 08:06.30 | goxboxlive | AT@HTCCSQ=1 |
| 08:06.30 | goxboxlive | [WCDMA] 1.14.10 |
| 08:06.31 | goxboxlive | OK |
| 08:06.31 | goxboxlive | @HTCCSQ: 4 |
| 08:07.04 | pH5 | very good. but that still leaves the cpi issue. |
| 08:07.11 | pH5 | cr2: what is AT@HTCDIS=1 ? |
| 08:08.52 | goxboxlive | seems like @ replace the %. But still no combination of cpi is working |
| 08:13.27 | pH5 | yes. seems like the qualcomm modem doesn't support cpi |
| 08:31.05 | goxboxlive | pH5: I see that you have created 3 pathes under gsmd in oe. Should i only use two of them and try to make them work with universal? (interpreter-ready.patch and vendor-tihtc.patch) |
| 08:32.06 | pH5 | goxboxlive: is the universal modem turned on only when the tty is opened? I assume yes, so you need interpreter-ready. |
| 08:32.31 | goxboxlive | yes |
| 08:33.03 | goxboxlive | ok, i'll try to make thoose two fit universal |
| 08:33.09 | pH5 | instead of vendor-tihtc you need something that drops the AT%CPI=? command and uses @HTCCSQ instead of %HTCCSQ |
| 08:33.38 | pH5 | I wonder how machine detection should be done for now - parse /proc/cpuinfo? |
| 08:34.06 | pH5 | Should I add a patch for vendor_qc for universal? |
| 08:35.38 | goxboxlive | Yes i think so |
| 08:38.00 | pH5 | goxboxlive: I added the interpreter patch for universal in oe.dev |
| 08:38.10 | pH5 | I'll look at machine detection now. |
| 08:38.26 | goxboxlive | good thx |
| 08:40.53 | pH5 | does universal support AT+CTZR ? |
| 08:45.58 | goxboxlive | let me check |
| 08:47.10 | goxboxlive | pH5 no at+ctzr isent supported |
| 08:47.41 | pH5 | goxboxlive: ok. same on magician. |
| 08:48.07 | goxboxlive | ok |
| 08:48.26 | pH5 | does ATUNS=0 work? |
| 08:48.29 | pH5 | AT%CUNS=0 |
| 08:49.44 | pH5 | and AT%CPRI ? |
| 08:49.56 | pH5 | (or the same with @) |
| 08:51.41 | goxboxlive | no non of them |
| 08:51.47 | pH5 | ok |
| 08:52.44 | goxboxlive | I did some debugging once and am looking thrue the logs when i shut the phone on i wince and cpi, cpri ctzr is not listed |
| 08:52.55 | goxboxlive | is ztzr timezone? |
| 08:53.08 | goxboxlive | s/ztzr/ctzr/ |
| 08:53.39 | goxboxlive | We use this AT@SYSTEMTIME |
| 08:54.20 | pH5 | ctzr is for automated time zone changes (reported by the gsm network) |
| 08:55.10 | goxboxlive | i c |
| 08:55.32 | goxboxlive | I have searced thrue the logs and i cant fin any % signs, only @ and + |
| 09:13.57 | *** join/#htc-linux TripleDES (n=sergio@89.129.158.218) |
| 09:29.53 | *** join/#htc-linux Disco (n=incognit@ip72-208-166-4.ph.ph.cox.net) |
| 09:39.36 | *** join/#htc-linux TripleDE1 (n=sergio@89.129.187.133) |
| 10:00.53 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 10:21.20 | cr2 | hi |
| 10:22.32 | cr2 | pH5: i think HTC_GSM lists all the "user" commands. there are some other vendor-specific ones, but i don't know their parameters |
| 10:22.59 | cr2 | and the commands set depends on the ROM version, so it is necessary to account for it. |
| 10:23.18 | cr2 | the CTZR command is supported only on athena. |
| 10:23.41 | cr2 | hm. maybe i should add the athena commands to wiki. |
| 10:23.52 | pH5 | cr2: that's interesting. I'm currently playing around with the gsmd machine/vendor plugin framework. |
| 10:24.05 | pH5 | cr2: yeah, please do that. |
| 10:24.58 | cr2 | uuh. hermes has a lot of evil commands. |
| 10:26.06 | cr2 | +MTVTDISC +HTCCTZV STAT +CHZ +RADIOVER |
| 10:26.19 | cr2 | +RADIOVER is good. available on athena too. |
| 10:26.42 | cr2 | +ECALLMODE=1 |
| 10:26.45 | *** join/#htc-linux rob_w (n=bob@p213.54.12.68.tisdip.tiscali.de) |
| 10:26.53 | cr2 | finally. |
| 10:26.55 | cr2 | +CREG=2 |
| 10:26.59 | pH5 | :) |
| 10:28.01 | cr2 | yes, they use +CREG=1 on universal. |
| 10:28.36 | cr2 | although even himalaya supports =2 |
| 10:28.54 | pH5 | cr2: the qualcomm modem doesn't use any strange mux on htc devices, right? |
| 10:29.16 | cr2 | no. plain AT. |
| 10:29.20 | pH5 | very good |
| 10:29.35 | Speedy2 | Just curious, how unhelpful has HTC been? |
| 10:29.38 | cr2 | we reset the whole device on open(). maybe it's not the best way to do it. |
| 10:30.15 | pH5 | cr2: that's what I do, too. on the neo powerup/down is separate from the tty. |
| 10:30.19 | cr2 | Speedy2: they are an m$ shop. but at least they never complain, which is good. like apple&ko. |
| 10:30.43 | Speedy2 | KO? |
| 10:30.46 | cr2 | com. |
| 10:30.50 | cr2 | co. |
| 10:30.53 | Speedy2 | ah. |
| 10:31.49 | cr2 | i've never heard they were unhappy with xda-dev for example. it was MS that requested to take down the ROM archive. |
| 10:31.50 | *** join/#htc-linux alsuren (n=alsuren@host86-148-18-74.range86-148.btcentralplus.com) |
| 10:32.36 | Speedy2 | Bastards. I take it the ROM archive is archived somewhere ? :) |
| 10:33.11 | *** join/#htc-linux rmoravci1 (n=rmoravci@adsl-d211.87-197-139.t-com.sk) |
| 10:33.14 | cr2 | don't know. |
| 10:34.04 | cr2 | pH5: @APDU ? |
| 10:34.32 | cr2 | oh. that's something new : @AudioTune= |
| 10:34.38 | pH5 | no idea. pdu sounds like sms. |
| 10:35.30 | cr2 | sms encoding. |
| 10:35.36 | cr2 | AT+TTYMODE=%u |
| 10:35.39 | cr2 | interesting. |
| 10:36.21 | cr2 | AT@AudioTuneSave AT@AudioTune?%u,%u |
| 10:36.35 | cr2 | and many UMTS things. |
| 10:36.38 | cr2 | AT+HSDPA=%u |
| 10:36.48 | cr2 | AT$3G_RESEL_STATUS |
| 10:37.13 | *** join/#htc-linux crashmatrix (n=crashmat@s5590785f.adsl.wanadoo.nl) |
| 10:37.50 | cr2 | pH5: there is PI on hermes. |
| 10:37.58 | cr2 | pH5: there is \PI on hermes. |
| 10:38.01 | cr2 | hehe. |
| 10:38.05 | pH5 | %CPI? |
| 10:38.08 | cr2 | yes. |
| 10:38.09 | pH5 | :) |
| 10:38.17 | cr2 | lol |
| 10:38.29 | cr2 | as response. |
| 10:38.49 | pH5 | does the hermes still hang early during boot? |
| 10:38.49 | cr2 | @HTCSAP=1 |
| 10:38.53 | cr2 | yes. |
| 10:39.25 | cr2 | not surprising. with our s3c24xx expertise ;) |
| 10:39.40 | cr2 | i guess pxa2xx would have been different. |
| 10:41.09 | cr2 | AT+CSIM |
| 10:41.27 | cr2 | ATATE |
| 10:41.37 | cr2 | why that? |
| 10:41.43 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 10:42.00 | pH5 | what? |
| 10:42.07 | pH5 | AT\rATE |
| 10:42.13 | cr2 | ooo. even more UMTS debug commands. |
| 10:42.58 | cr2 | AT$3G_NEIGHBOR_ST_2? AT$3G_NEIGHBOR_ST? |
| 10:43.59 | cr2 | AT@SIMON AT@SIMOFF |
| 10:50.30 | Speedy2 | cr2: What Qualcomm chipsets do you need documentation on? |
| 10:59.52 | cr2 | Speedy2: i doubt that we need it. the device-specific wiring is done by HTC anyway. |
| 11:00.17 | cr2 | and we know the command set. |
| 11:01.15 | Speedy2 | Ok. If you need datasheets on Qualcomm stuff, let me know |
| 11:02.58 | Speedy2 | That's a little bit more accessible for me. |
| 11:03.38 | cr2 | ok. |
| 11:17.23 | cr2 | pH5: finished with init. |
| 11:17.33 | cr2 | http://wiki.xda-developers.com/index.php?pagename=HTC_GSM |
| 11:19.30 | pH5 | ok. so we have no idea yet what HTCSAP and HTCDIS are? |
| 11:24.54 | cr2 | no. |
| 11:27.25 | cr2 | hahaha. AT+CAOC=0 |
| 11:29.55 | pH5 | what is this? cost info related? |
| 11:33.57 | cr2 | yes. |
| 11:34.44 | cr2 | AT+CPIN syntax adds %d at the end. |
| 11:37.34 | cr2 | at%ctzv added |
| 11:37.48 | cr2 | at+csim is new. |
| 11:39.17 | cr2 | at+esa new |
| 11:41.33 | cr2 | at+gevl |
| 11:42.24 | cr2 | at+gimg |
| 11:51.00 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 11:57.51 | cr2 | pH5: |@HTCDIS: |0010;003f | U+ | |
| 12:00.05 | *** join/#htc-linux rob_w (n=bob@p213.54.12.68.tisdip.tiscali.de) |
| 12:11.30 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 12:46.58 | *** join/#htc-linux sponged (n=sponged@85-211-207-92.dyn.gotadsl.co.uk) |
| 12:50.18 | sponged | cr2: I'm trying to load the HTC_GSM wiki page, but it comes up blank. The google cache looks fine, and for example the UNIVERSAL_GSM and other pages on the wiki look fine. Is HTC_GSM locked somehow? |
| 12:52.09 | cr2 | sponged: maybe the page is too big now? the wiki there is a bit buggy. some pages are locked and can't be edited. |
| 12:53.17 | sponged | Maybe it's too big, and that has exposed a bug in the wiki, yep. At any rate, it's unavailable from here - can you load it ok? |
| 12:53.56 | sponged | Would be a shame if all that hard work is disappeared! |
| 12:54.30 | cr2 | i've saved a local copy :) maybe i'll put it to hh.org ? |
| 12:55.06 | sponged | cr2: Or split it? You know better than me - just thought I'd alert you to the potential proe=blem.. |
| 12:55.36 | sponged | Nice to see all the work going into uni phone stuff, BTW. Much appreciated here! |
| 12:55.56 | sponged | (And dirty resume fixes, too! :) |
| 12:56.59 | cr2 | i think that the best will be to "backup" these pages at hh.org wiki. |
| 12:58.46 | cr2 | yeah. i need to patch the pxa-regs.h |
| 12:59.39 | sponged | Anyway, I'll go back to lurking via the iclem.net irc log. Keep going! :-) |
| 13:05.24 | cr2 | pH5: http://www.handhelds.org/moin/moin.cgi/HTC_25GSM |
| 13:16.54 | *** join/#htc-linux cyrill62_ (n=cyrill62@128.250.146.195.dynamic.adsl.abo.nordnet.fr) |
| 13:19.02 | *** join/#htc-linux pancake (n=xxx@55.pool85-52-227.static.orange.es) |
| 13:32.59 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 14:00.31 | Kevin2 | Hi |
| 14:01.13 | pH5 | hi Kevin2 |
| 14:03.30 | Kevin2 | pH5: Do you do all development on the Magician, or do you have other devices too? |
| 14:04.29 | pH5 | Kevin2: magician, hx4700 and neo1973 |
| 14:04.40 | pH5 | magician is my only htc phone |
| 14:08.13 | cr2 | hx4700 is HTC Roadster, but it's a pda. very nice pda :) |
| 14:12.10 | *** part/#htc-linux rmoravci1 (n=rmoravci@adsl-d211.87-197-139.t-com.sk) |
| 14:13.08 | *** join/#htc-linux pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 14:20.56 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 15:03.50 | *** join/#htc-linux food (n=xxx@55.pool85-52-227.static.orange.es) |
| 15:12.57 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 15:19.39 | *** join/#htc-linux psokolovsky_ (n=psokolov@favoured-food.volia.net) |
| 15:22.14 | *** join/#htc-linux pH5_ (n=ph5@e178231156.adsl.alicedsl.de) |
| 15:33.15 | *** join/#htc-linux pH5 (n=ph5@e178231156.adsl.alicedsl.de) |
| 15:33.15 | *** join/#htc-linux RoEn_PC (n=roen@p54A65197.dip.t-dialin.net) [NETSPLIT VICTIM] |
| 15:33.15 | *** join/#htc-linux elysion_ (n=elysion@k203c.lanhovi.tontut.fi) [NETSPLIT VICTIM] |
| 15:33.15 | *** join/#htc-linux Hymie (n=Hymie@l8r.net) [NETSPLIT VICTIM] |
| 15:33.15 | *** join/#htc-linux SmallR2002 (n=SmallR20@88-107-17-43.dynamic.dsl.as9105.com) [NETSPLIT VICTIM] |
| 15:33.32 | *** join/#htc-linux RoEn_PC (n=roen@p54A65197.dip.t-dialin.net) |
| 15:33.36 | Kevin2 | cr2: Were you able to get ir messages reliably? |
| 15:35.33 | *** join/#htc-linux alsuren (n=alsuren@host86-148-18-74.range86-148.btcentralplus.com) |
| 15:36.37 | cr2 | Kevin2: i don't see them at all. with irdadump -x |
| 15:37.10 | cr2 | is irdadump -x a really raw dump ? |
| 15:37.15 | Kevin2 | cr2: You can't use irdadump - you need to capture the raw serial traffic. |
| 15:37.32 | Kevin2 | cr2: No - it is like tcpdump - it will only show you packets which look like packets. |
| 15:38.02 | cr2 | well, then i can't see the messages without writing my own kernel driver. |
| 15:39.41 | *** join/#htc-linux RoEn_PC (n=roen@p54A65197.dip.t-dialin.net) [NETSPLIT VICTIM] |
| 15:39.41 | *** join/#htc-linux pH5 (n=ph5@e178231156.adsl.alicedsl.de) |
| 15:39.41 | *** join/#htc-linux elysion_ (n=elysion@k203c.lanhovi.tontut.fi) [NETSPLIT VICTIM] |
| 15:39.41 | *** join/#htc-linux Hymie (n=Hymie@l8r.net) |
| 15:39.41 | *** join/#htc-linux SmallR2002 (n=SmallR20@88-107-17-43.dynamic.dsl.as9105.com) [NETSPLIT VICTIM] |
| 15:39.56 | cr2 | the bios does not help me in 64bit mode. only in 32bit it seems. thanks asus for buggy bios. |
| 15:40.01 | cr2 | there was a time i had both 32bit and 64bit install only for running IR off /dev/ttyS1 |
| 15:40.18 | cr2 | but it died with the notebook hdd. |
| 15:41.33 | Kevin2 | cr2: You can't just "cat /dev/ttyS1"? |
| 15:43.30 | cr2 | no. it does not show anything in 64bit mode (od -x /dev/ttyS1) |
| 15:44.02 | cr2 | it worked in 32bit mode before. |
| 15:44.09 | Kevin2 | Does irdadump work in 64bit mode? |
| 15:44.12 | cr2 | yes. |
| 15:44.43 | Kevin2 | Can you run "irdadump" and "od" simultaneously? Does od still show no data? |
| 15:44.50 | cr2 | because the nsc-ircc deals with it properly. |
| 15:45.00 | cr2 | od says open() failed. |
| 15:45.21 | cr2 | or io error, i don't remember. |
| 15:45.47 | cr2 | i must do setserial /dev/ttyS1 uart none |
| 15:46.02 | cr2 | otherwise irda driver fails. |
| 15:46.26 | cr2 | which is logical. |
| 15:49.10 | Kevin2 | cr2: Did you try the systemtap hack? |
| 15:49.22 | cr2 | systemtap ? |
| 15:49.23 | Kevin2 | 15:50 Kevin2 The command was something like: stap -v -e 'probe module("irda*").function("async_unwrap_char") { printf("%d\n", $byte) }' 15:50 |
| 15:50.02 | Kevin2 | Systemtap lets you hook functions in the kernel and report variables. |
| 15:50.31 | Kevin2 | If you hook the async_unwrap_char function in the irda module, you should be able to grab all the raw characters. |
| 15:50.39 | cr2 | don't have it here. |
| 15:50.40 | cr2 | ok. |
| 15:50.41 | Kevin2 | It worked for me, but it seemed a bit unstable. |
| 15:51.38 | cr2 | found it. |
| 15:52.12 | cr2 | you can just add printk() to the kernel driver. |
| 15:52.31 | cr2 | do you know which IR chipset is used ? |
| 15:52.54 | cr2 | i have not recompiled the kernel on my notebook for years :) |
| 15:57.14 | *** join/#htc-linux RoEn_PC (n=roen@p54A65197.dip.t-dialin.net) [NETSPLIT VICTIM] |
| 15:57.14 | *** join/#htc-linux elysion_ (n=elysion@k203c.lanhovi.tontut.fi) [NETSPLIT VICTIM] |
| 15:57.14 | *** join/#htc-linux Hymie (n=Hymie@l8r.net) [NETSPLIT VICTIM] |
| 15:57.14 | *** join/#htc-linux SmallR2002 (n=SmallR20@88-107-17-43.dynamic.dsl.as9105.com) [NETSPLIT VICTIM] |
| 16:01.28 | *** join/#htc-linux elysion (n=elysion@k203c.lanhovi.tontut.fi) |
| 16:01.28 | pancake | how can I know what irda driver I need? |
| 16:03.54 | cr2 | pancake: do you see hermes with irdadump ? |
| 16:04.36 | cr2 | broadcast of hermes running wince. |
| 16:04.53 | pancake | that's what i'm trying to do :) |
| 16:04.54 | cr2 | Pocket_PC [ IrOBEX, IrCOMM ... |
| 16:05.05 | pancake | but i'm not very lucky with irda |
| 16:05.28 | cr2 | do you have a notebook or what ? |
| 16:06.01 | pancake | yes |
| 16:06.05 | pancake | D420 |
| 16:06.52 | cr2 | Latitude™ D420 ? |
| 16:07.29 | pancake | yes |
| 16:08.05 | cr2 | nothing about IR here http://www.linlap.com/tiki-index.php?page=Dell+Latitude+D420 |
| 16:08.23 | *** join/#htc-linux scatterp (n=zsirc@196.40.37.87) |
| 16:08.37 | scatterp | hey im trying to install linux on a hermes |
| 16:08.51 | scatterp | i found a harret but i dont see the other files |
| 16:08.59 | scatterp | can some one tell me the situation... |
| 16:09.59 | cr2 | it does not work. |
| 16:10.31 | pancake | hehe |
| 16:10.37 | scatterp | oh ok |
| 16:10.42 | scatterp | any idea when cr2 ? |
| 16:11.05 | cr2 | we are working ;) |
| 16:11.08 | cr2 | http://www-spht.cea.fr/en/Images/Pisp/jbouttier/d420-linux.php |
| 16:11.13 | cr2 | nothing here too. |
| 16:12.00 | cr2 | Fingerprint Scanner |
| 16:12.17 | pancake | yes :) |
| 16:12.19 | scatterp | i got a dell finger print scanner on my d620 if that helps... |
| 16:12.23 | pancake | i use't for unlocking the X |
| 16:13.04 | cr2 | lol. |
| 16:13.23 | cr2 | but i don't see anybody talking about IR. |
| 16:13.40 | cr2 | do you have windowz ? |
| 16:14.23 | Kevin2 | cr2: I ended up putting a printk in my driver (stir4200). It was a little more stable, but it seems to just "clam up" spontaneously. I'm guessing something on the USB side wants some attention. |
| 16:14.41 | pancake | yup |
| 16:14.47 | pancake | but i'll have to reboot |
| 16:17.00 | cr2 | Kevin2: i have an external USB "mouse"-like SIR device, but last time i have tried, it did not work. |
| 16:18.43 | pancake | let me try with windows |
| 16:18.51 | pancake | maybe i could get some driver info |
| 16:19.03 | scatterp | cr2 yeah but i would install linux if you needed me too |
| 16:20.04 | *** join/#htc-linux rmoravcik (n=rmoravci@pc-3s0zt5w2e4y0vzmhnrzq3a21zqajzfw.users.student.utc.sk) |
| 16:22.46 | *** join/#htc-linux Fossi (n=Fossi@e176096030.adsl.alicedsl.de) |
| 16:23.10 | *** join/#htc-linux pancake (n=Miranda@55.pool85-52-227.static.orange.es) |
| 16:23.17 | pancake | back |
| 16:24.17 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 16:25.56 | *** join/#htc-linux RoEn_PC (n=roen@p54a65197.dip.t-dialin.net) |
| 16:26.25 | *** join/#htc-linux psokolovsky__ (n=psokolov@82.193.98.4) |
| 16:40.26 | cr2 | Kevin2: i've found another old notebook where SIR was working many years ago. |
| 16:40.40 | cr2 | what stty settings do i need ? |
| 16:40.51 | pancake | uhm i've a serial irda device |
| 16:41.03 | pancake | i could try with usb->serial->serial-irda |
| 16:41.17 | pancake | no idea if usb->serial will break something :/ |
| 16:44.23 | Kevin2 | cr2: Not sure. |
| 16:44.52 | cr2 | it does not show anything. |
| 16:44.55 | Kevin2 | Does anyone recall if there are any special command line parameters needed when booting from an initramfs? |
| 16:45.21 | Kevin2 | cr2: You have the hermes set to "accept all incoming beams" right? |
| 16:45.25 | cr2 | Kevin2: afaik no. |
| 16:45.34 | psokolovsky__ | Kevin2: deoends on what you mean by booting |
| 16:45.37 | cr2 | yes, and i see the broadcast on another notebook. |
| 16:46.15 | Kevin2 | psokolovsky__: I can't even boot my apache with initramfs.. Something odd is going on. I get an error about - can't execute /init.. |
| 16:46.39 | psokolovsky__ | Kevin2: ok, use kernel cmdline param rdinit= |
| 16:46.56 | psokolovsky__ | To get what we had with initrd, rdinit=/sbin/init |
| 16:47.25 | Kevin2 | I'm using initramfs compiled into kernel and there is a file /init which used to work. |
| 16:47.31 | cr2 | wow. |
| 16:47.41 | cr2 | no irda, but ............................. |
| 16:47.53 | *** join/#htc-linux Hymie (n=Hymie@l8r.net) |
| 16:47.53 | *** join/#htc-linux SmallR2002 (n=SmallR20@88-107-17-43.dynamic.dsl.as9105.com) [NETSPLIT VICTIM] |
| 16:48.02 | cr2 | done, the kernel. |
| 16:48.08 | cr2 | here it stops. |
| 16:48.20 | cr2 | ............done, the kernel. |
| 16:48.28 | cr2 | plain hh.org CVS |
| 16:52.27 | Kevin2 | I get: Uncompressing Linux. done, booting the kernel.\r\n |
| 16:53.43 | Kevin2 | cr2: Can you try the patch at: http://pastebin.ca/419777 |
| 16:53.52 | cr2 | ok, how do we insert the debug messages. |
| 16:54.20 | Kevin2 | It doesn't work for me (but I think it is correct). But my ir stops so abruptly in so many places that I don't know if it is just the ir stuff. |
| 16:54.33 | cr2 | ok. |
| 16:55.51 | Kevin2 | Hrmm. It should probably have a "\n" on the end of the string. |
| 16:58.34 | cr2 | set ramaddr 0x30000000 |
| 16:58.34 | cr2 | set ramsize 0x04000000 |
| 16:58.34 | cr2 | set KERNEL "ziherm" |
| 16:58.34 | cr2 | set CMDLINE "mem=64M" |
| 16:58.34 | cr2 | set MTYPE 1213 |
| 16:58.58 | cr2 | btw, which printascii is used ? |
| 16:59.23 | cr2 | recompiling the kernel. |
| 16:59.35 | Kevin2 | printascii is in arch/arm/kernel/debug.S |
| 17:01.35 | Kevin2 | Much of the code is in include/asm-arm/arch-s3c2410/debug-macro.S |
| 17:02.06 | Kevin2 | Oh, and you'll need CONFIG_DEBUG_LL enabled. |
| 17:03.09 | cr2 | hm. yes. |
| 17:03.28 | cr2 | init/built-in.o: In function `start_kernel': |
| 17:03.28 | cr2 | calibrate.c:(.init.text+0x4e0): undefined reference to `printascii' |
| 17:04.52 | cr2 | Kernel debugging -> Kernel low-level debugging functions -> Kernel low-level debugging messages via S3C2410 UART |
| 17:05.11 | Kevin2 | Ughh. Latest CVS mainline haret wont boot Apache, but haret 0.4.7 will... |
| 17:05.47 | *** join/#htc-linux tudenbart (n=willi@xdsl-87-78-52-58.netcologne.de) |
| 17:06.09 | cr2 | ? |
| 17:06.25 | Kevin2 | I must have broken something in haret in latest CVS. |
| 17:06.59 | cr2 | if i see the "................." does it mean the kernel will boot ? |
| 17:07.33 | Kevin2 | cr2: No, the dots come from arch/arm/boot/compressed/misc.c |
| 17:07.40 | Kevin2 | That's just the gunzip. |
| 17:08.04 | cr2 | well, which haret binary should i use? |
| 17:09.35 | Kevin2 | I don't have a good answer.. haret047 didn't shutdown usb host on hermes. I don't know why apache isn't booting, so I can't say if it will also affect hermes. |
| 17:14.33 | cr2 | the IR here eats the letters, but the last message is 'el' |
| 17:17.17 | Kevin2 | el? As in kernel? |
| 17:18.24 | cr2 | i've modified the sting and will check now. |
| 17:18.47 | cr2 | do you know where are the OHCI buffers and counter ? |
| 17:19.20 | Kevin2 | I used the pxa docs for reference.. Right now, I just write a 1 to the control register which should reset it. |
| 17:19.40 | cr2 | yes, i get ############### |
| 17:19.48 | cr2 | string which i've added. |
| 17:19.57 | Kevin2 | cr2: Where did you add it? |
| 17:20.11 | cr2 | as you said, before init_...smp() |
| 17:20.19 | cr2 | smp_setup_processor_id(); |
| 17:20.37 | Kevin2 | You see the message from the "start_kernel()" function? |
| 17:20.40 | cr2 | yes. |
| 17:20.56 | Kevin2 | Wow, that's awesome. That means we survived the mmu enable. |
| 17:21.06 | cr2 | lol. |
| 17:21.32 | cr2 | btw, can we compile udc as a module ? |
| 17:21.53 | Kevin2 | So, one needs to add a bunch of messages to start_kernel to find where the boot is failing. |
| 17:22.17 | cr2 | printk(linux_banner); |
| 17:22.24 | cr2 | this one is not reached. |
| 17:22.30 | cr2 | obviously. |
| 17:22.42 | Kevin2 | By that point, we have played with the clocks and stuff I'm guessing. |
| 17:22.56 | cr2 | printk(KERN_NOTICE); |
| 17:22.59 | cr2 | too. |
| 17:23.10 | cr2 | boot_cpu_init(); |
| 17:23.14 | cr2 | maybe here. |
| 17:23.18 | cr2 | i'll check. |
| 17:27.18 | cr2 | does not show up after lock_kernel(); |
| 17:27.33 | cr2 | bisecting. |
| 17:30.12 | cr2 | before local_irq_disable(); |
| 17:30.16 | cr2 | also nothing. |
| 17:32.04 | Kevin2 | cr2: Are you still printing the message at the start of the function? And do you still see that message? |
| 17:34.25 | cr2 | i see it after smp_setup_processor_id(); |
| 17:34.38 | cr2 | last step. |
| 17:36.43 | cr2 | ok. does not show up |
| 17:37.02 | cr2 | so somewhere is unwind_init(); |
| 17:38.15 | cr2 | hm. it is empty ?? |
| 17:39.21 | Kevin2 | Are you adding messages, or moving the message? It is possible the code is failing randomly - so I'd add messages instead of moving them. |
| 17:39.49 | cr2 | ok. |
| 17:42.26 | *** join/#htc-linux SmallR2002 (n=SmallR20@88-107-17-43.dynamic.dsl.as9105.com) [NETSPLIT VICTIM] |
| 17:43.14 | cr2 | put both. don't see the message after unwind_init(); |
| 17:43.34 | cr2 | objdump ;) |
| 17:45.21 | cr2 | not in System.map ? |
| 17:46.40 | cr2 | c00084c4 W smp_setup_processor_id |
| 17:50.59 | cr2 | Kevin2: i think somebody is still corrupting the memory. |
| 17:51.13 | cr2 | do you turn off the device clocks ? |
| 17:51.26 | cr2 | afaik they are all turned on in wince. |
| 17:53.04 | lilliput | have you seen the news from the HTC website, they are going to finance the linux devel for the universal : http://www.europe.htc.com/press/pressreleases.html sounds great ! |
| 17:55.41 | cr2 | lol. |
| 17:56.20 | cr2 | Kevin2: it's repeatable. |
| 17:57.00 | cr2 | can you crc the rest of the memory ? maybe in 4M blocks. |
| 17:57.02 | Kevin2 | cr2: Where does it die? In unwind_init? |
| 17:57.22 | cr2 | unwind_init() is optimized out. |
| 17:57.38 | Kevin2 | So, where is it dying? |
| 17:58.01 | cr2 | so i have two printascii's . the first comes through, then second one not. |
| 17:58.22 | cr2 | if i remove the first, i don't see the second. strange. |
| 17:59.24 | cr2 | printascii("************************************************************************************************\n"); |
| 17:59.24 | cr2 | <PROTECTED> |
| 17:59.24 | cr2 | printascii("In start_kernel###############################################################################################\n"); |
| 17:59.39 | cr2 | i see only '***********************' |
| 18:12.28 | Kevin2 | Well, I'm thoroughly confused. I can see an error in the change I made to haret that could cause the boot to fail. I can't understand why the CRC checks didn't flag it though.. If the crc passes, how could the kernel not boot.. I don't get it. |
| 18:13.31 | cr2 | the stack ? some delayed dma ? |
| 18:13.50 | Kevin2 | No, it is related to cache flushing. |
| 18:14.15 | cr2 | ok. maybe. but it's a difficult area. |
| 18:16.14 | cr2 | and why does the kernel boot on other devices ? |
| 18:30.36 | *** join/#htc-linux goxboxlive (n=goxboxli@206.80-202-161.nextgentel.com) |
| 18:32.32 | Kevin2 | Well, I've checked the fix into Haret CVS head. I'm still confused on why the CRC checks didn't reveal this issue earlier. |
| 18:33.24 | goxboxlive | cr2: Are you around? |
| 18:33.29 | goxboxlive | hi Kevin2 |
| 18:33.39 | Kevin2 | Hi goxboxlive |
| 18:33.58 | goxboxlive | Kevin2: Have cr2 asked you about ext2 support in Haret? |
| 18:35.28 | Kevin2 | Not directly, no. |
| 18:35.52 | goxboxlive | i c |
| 18:36.02 | Kevin2 | That is a lot of code to add to haret. Why not just link what you need into an initramfs instead? |
| 18:36.16 | goxboxlive | is it a lot of work ti implement ext2 support in it? |
| 18:36.57 | Kevin2 | I'm not familiar with the ext2 code, but I think it would be. |
| 18:36.58 | goxboxlive | That's the reason to skip a initramfs |
| 18:37.05 | goxboxlive | i c. |
| 18:38.07 | pH5_ | Kevin2: I'd like to adapt another bootloader's minimal ext2 readonly support to load the kernel from an ext2 fs. |
| 18:38.19 | pH5_ | Unfortunately, I'm not familiar with ext2 either. |
| 18:39.33 | Kevin2 | pH5_: What is the gain over using an initramfs? |
| 18:40.03 | pH5_ | Kevin2: a more desktop-like experience, kernel and modules under distro package management. |
| 18:40.20 | cr2 | pH5_: what about loading uboot from haret ? |
| 18:41.56 | cr2 | tools/checkrelocs: line 15: /bin/arm-wince-mingw32ce-ld: File not found |
| 18:42.35 | Kevin2 | cr2: Are you sure you have the latest Makefile? |
| 18:44.03 | cr2 | hm. yes. but i changed 'opt' to /tmp' |
| 18:44.10 | *** join/#htc-linux cyrill62 (n=cyrill62@128.250.146.195.dynamic.adsl.abo.nordnet.fr) |
| 18:44.11 | Kevin2 | Do you still have: export BASE |
| 18:44.16 | cr2 | export BASE |
| 18:44.16 | cr2 | BASE ?= /tmp/mingw32ce |
| 18:44.28 | cr2 | maybe change the lines ? |
| 18:44.58 | pH5_ | cr2: maybe. I didn't look closely at u-boot yet. |
| 18:45.38 | cr2 | pH5_: if uboot will work, them it's the easiest solution. |
| 18:46.17 | pH5_ | u-boot frightens me. I guess it already has support for pxa2xx mci. but probably not for asic3. |
| 18:46.42 | Kevin2 | cr2: Yeah, try putting the export after the assignment |
| 18:48.32 | goxboxlive | cr2: instead of AT%CPRI=? what do we use on universal? |
| 18:48.46 | goxboxlive | do we use cpri at all? |
| 18:49.29 | cr2 | goxboxlive: what is the purpose of ATPRI ? |
| 18:50.11 | goxboxlive | actually i dont know, i think pH5 can answer that bether than me. Because i cant find it in the debugs log either |
| 18:50.18 | cr2 | i think it's not a very smart idea to compile AT commands into the program. |
| 18:50.37 | pancake | yeah I finally found the serial irda |
| 18:50.40 | pancake | let's try with it |
| 18:50.41 | cr2 | we need some abstraction layer. not really RIL, but something like that :) |
| 18:51.36 | pancake | cr2: : have you finally got the irda working? |
| 18:51.45 | goxboxlive | cr2: But how do we hold the AT commands outside ? Do you mean like making a at.conf file? |
| 18:51.59 | pancake | reboot to linux again |
| 18:53.00 | goxboxlive | pH5: Is that doable with the recent gsmd code (use a seperate at.conf file)? |
| 18:53.12 | cr2 | goxboxlive: maybe like a .so plugin. like m$ does with rilgsm.dll |
| 18:53.21 | goxboxlive | i c |
| 18:53.54 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 18:54.41 | goxboxlive | cr2: with a plugin doesnt we have to allmost make a new gmsd code? We cant use the openmoko source then? |
| 18:54.50 | cr2 | at least for the at% and at$ commands. |
| 18:55.10 | *** join/#htc-linux pancake (n=xxx@55.pool85-52-227.static.orange.es) |
| 18:55.40 | cr2 | we can have an empty function for the at%cpri if it is _so_ important. |
| 18:55.42 | goxboxlive | I have also noticed that the qualcomm modem doesnt use %. It seems like it uses @ instead. |
| 18:56.03 | Kevin2 | I'll be back later. |
| 18:56.11 | cr2 | yes, this is nonstandard and nonportable stuff anyway. |
| 18:56.41 | goxboxlive | right |
| 18:57.03 | cr2 | Kevin2: yes, i've passed ########### now. |
| 19:12.51 | goxboxlive | i need a spare sim card to test on universal. Sick and tired of swithing between a780 and uni |
| 19:15.03 | pH5_ | gsmd will get .so plugins for machine / vendor |
| 19:15.39 | pH5_ | the latter is chipset specific stuff like supported AT commands, the former machine specific, for example to set up the ldisc on blueangel/magician/alpine |
| 19:16.04 | pH5_ | the abstraction layer is libgsmd. |
| 19:16.38 | pH5_ | no idea how sane the libgsmd api is, but I think the gsmd design itself (a daemon to correctly serialize at commands and handle the answers) is pretty good. |
| 19:18.00 | cr2 | pH5: what does PRI do? |
| 19:18.13 | pH5 | cipher indications |
| 19:18.51 | cr2 | well. it's not very trustworthy anyway. |
| 19:19.38 | pH5 | agreed. afaik the provider can forcefully disable those in the sim and it's not end-to-end encryption, obviously. |
| 19:22.45 | *** join/#htc-linux Hymie (n=nHymie@l8r.net) |
| 19:23.27 | cr2 | pH5: i've seen that most of your pxa-regs.h problems are with MMC/SD ? |
| 19:24.52 | pH5 | cr2: what problems? |
| 19:25.28 | pH5 | I don't think I had problems with the MMC/SD #defines. SD support worked pretty much from the start. |
| 19:27.23 | cr2 | yes. ok. |
| 19:27.44 | pancake | the usb-serial-irda doesn't works :( |
| 19:29.14 | pancake | uhm..does exist an irda sniffer for symbian? |
| 19:30.20 | cr2 | pancake: it's too lowlevel. |
| 19:31.30 | cr2 | ok, idon't come through setup_arch(&command_line); |
| 19:31.36 | cr2 | sounds reasonable. |
| 19:48.57 | psokolovsky__ | Kevin2: wanna lol? haret-0.4 regressed DiskOnChip support in h4000 |
| 20:01.43 | goxboxlive | ok, i am doing some progress here. Now at least gsmd is registrating my incomming calls |
| 20:04.48 | goxboxlive | but openmoko-diale is not talking with gsmd at all. hmm |
| 20:22.48 | cr2 | Kevin2: i have set CMDLINE "mem=64M" |
| 20:22.52 | cr2 | <PROTECTED> |
| 20:22.52 | cr2 | <PROTECTED> |
| 20:22.52 | cr2 | <PROTECTED> |
| 20:23.00 | cr2 | then it crashes somewhere here. |
| 20:23.03 | *** join/#htc-linux pancake (n=xxx@55.pool85-52-227.static.orange.es) |
| 20:23.17 | cr2 | if i comment that line, it crashes later ;-) |
| 20:23.40 | cr2 | something with the tags is not quite right. |
| 20:23.44 | pancake | how the things are going? :) have you finally got the point where it hangs? |
| 20:24.33 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 20:25.19 | cr2 | yes, but it does not help very much. |
| 20:26.47 | cr2 | Kevin2: can i do something like sprintf() for printascii() ?? |
| 20:28.47 | goxboxlive | Well, now incomming calls pops up with openmoko-dialer, but it failes when i answer. |
| 20:29.07 | cr2 | never debugged the kernel on such a lowlevel ;) |
| 20:29.32 | cr2 | goxboxlive: fails ? where ? |
| 20:30.18 | goxboxlive | when i answer the phone |
| 20:30.27 | goxboxlive | just gui, gsmd still runs |
| 20:30.38 | cr2 | for printk() i need an IR console which needs the unusual serial clock setup. |
| 20:31.04 | cr2 | goxboxlive: what gsmd is supposed to do to answer ? |
| 20:31.43 | goxboxlive | no |
| 20:31.55 | goxboxlive | well actually it is |
| 20:32.08 | goxboxlive | based on commands from openmoko-dialer |
| 20:48.59 | goxboxlive | ohh yes, i have openmoko-dialer working. Both incomming and outgoing :-) |
| 20:51.04 | cr2 | hahaha. what about the voice ? |
| 20:51.22 | goxboxlive | well, i can hear in unoversal, but the mic on uni is dead |
| 20:51.47 | goxboxlive | so i can have a oneplex coversation :-) |
| 20:52.20 | cr2 | how does it work on neo ? |
| 20:52.32 | goxboxlive | no idea |
| 20:52.39 | goxboxlive | i'll check |
| 20:52.54 | cr2 | is there some alsa code in gsmd/libgsmd ? |
| 20:53.23 | goxboxlive | i dont know cr2, i cant see the difrents on alsa vs asoc |
| 20:53.45 | cr2 | asoc is alsa too. |
| 20:53.57 | goxboxlive | i have only checked the at commands and removed AT commands uni doesent support |
| 20:54.09 | *** join/#htc-linux hollo (n=hollo@3e6b6459.rev.stofanet.dk) |
| 20:54.24 | cr2 | :) |
| 20:54.33 | goxboxlive | I also think gsm signl strenght should work now, but i just cant get it starting |
| 20:54.53 | cr2 | i'm updating oe now. |
| 20:55.08 | cr2 | is ut based on CSQ ? |
| 20:55.20 | cr2 | oh. 252 revs. |
| 20:55.25 | goxboxlive | yes, i have changed it to at@htccsq |
| 20:55.49 | cr2 | from at%htccsq ? |
| 20:56.03 | goxboxlive | no, from at%csq |
| 20:56.20 | cr2 | should it be at+csq ? |
| 20:56.50 | goxboxlive | on universal it should be at@htccsq |
| 20:57.36 | cr2 | ok. but strange. what is it on magician? |
| 20:57.59 | goxboxlive | AFAIK at%htccsq |
| 20:58.14 | cr2 | ok. |
| 20:58.21 | Kevin2 | psokolovsky__: How did haret regress? |
| 20:59.09 | Kevin2 | cr2: You shouldn't need a command line at all. But, can you printascii() the command line it thinks it has? |
| 21:01.02 | cr2 | Kevin2: the fact is that i crash in parse_cmdline() |
| 21:01.10 | cr2 | hmm. which pointer... |
| 21:02.14 | cr2 | saved_command_line |
| 21:02.22 | cr2 | recompiling. |
| 21:04.50 | cr2 | yes, it prints something from the commandline |
| 21:05.25 | cr2 | can't check it to 100% because of the ususal big character loss. |
| 21:06.24 | Kevin2 | cr2: Note that paging_init() changes the mmu tables, so I don't know if you can use printascii() after it. |
| 21:07.29 | cr2 | ok. |
| 21:07.41 | Kevin2 | setup_arch() is after printk(linux_banner) -- do you get the printks? |
| 21:07.58 | cr2 | then we need an accurate serial clock setup. |
| 21:08.05 | cr2 | printk does not work. |
| 21:08.37 | cr2 | where does it write before the console is initialized ? |
| 21:10.35 | Speedy2 | goxboxlive: Is the goal to use OpenMoko going forward? |
| 21:10.59 | goxboxlive | well we are working on it |
| 21:11.15 | goxboxlive | it is still very unstable |
| 21:11.34 | cr2 | Please install following missing utilities: shasum |
| 21:11.43 | cr2 | where do i get it ? |
| 21:13.00 | Kevin2 | cr2: It should go to the serial console. (Your config will need to define the early serial console and you need to enable it on the cmdline.) |
| 21:13.29 | pH5 | <PROTECTED> |
| 21:14.06 | goxboxlive | cr2: i had the same problemi OE. I just commnted out the content in sanity.conf |
| 21:14.47 | cr2 | pH5: what's wrong with sha1sum ? |
| 21:16.00 | pH5 | cr2: I don't know, really. I'll ask zecke. |
| 21:17.12 | cr2 | ok. |
| 21:17.18 | cr2 | Kevin2: i#ll try. |
| 21:19.04 | cr2 | Kevin2: where is the early serial console option ? |
| 21:21.37 | cr2 | S3C2410 UART to use for low-level messages |
| 21:21.39 | *** join/#htc-linux lilliput_ (n=Touriste@82-47-146-8.cable.ubr11.brad.blueyonder.co.uk) |
| 21:21.49 | cr2 | it seems this option is present twice. |
| 21:22.06 | Kevin2 | drivers->character drivers->serial drivers-> |
| 21:22.29 | Kevin2 | Support for console on S3C2410 serial port |
| 21:22.48 | Kevin2 | It isn't really an "early serial thing" -- just the normal kernel serial stuff. |
| 21:24.42 | *** join/#htc-linux TeringTu1y (n=maarten@249-116-dyn.ipact.nl) |
| 21:25.39 | *** join/#htc-linux ljp (n=lpotter@203.94.178.46) |
| 21:25.42 | cr2 | yes, but then we will probably need a proper clock setup ? |
| 21:26.29 | cr2 | ah, something come into my mind. |
| 21:26.58 | cr2 | it should be console=ttySAC2,9600,8,1 |
| 21:32.12 | cr2 | Kevin2: i think it's buffered somewhere before the console is set up. |
| 21:32.36 | cr2 | and if you don't reach the set up code, then it is useless. |
| 21:33.51 | Kevin2 | cr2: Okay. |
| 21:34.32 | Kevin2 | BTW, I'm not sure paging_init() will break printascii. I just see a bunch of warnings about paging_init in head.S |
| 21:35.03 | pH5 | cr2: shasum vs. sha1sum is a different binaries on different platform issue. we'll switch to sha256sum |
| 21:35.04 | *** join/#htc-linux TeringTu2y (n=maarten@249-116-dyn.ipact.nl) |
| 21:37.05 | goxboxlive | what does this word mean? unsolicited |
| 21:37.36 | cr2 | unrequested. |
| 21:37.48 | goxboxlive | ok thx |
| 21:38.15 | cr2 | Kevin2: my tstIRsend code works on himalaya. that's good. |
| 21:38.41 | cr2 | pH5: didn't really get it. |
| 21:40.07 | pH5 | cr2: zecke used shasum because that is what was installed on his computer. I'll change OE to use sha256sum for checksums. |
| 21:41.33 | goxboxlive | pH5: Do macigian use AT+CFUN=1 . It gives me "0" but i know it is ok, but it spams my debug windows. Is it the same for you? |
| 21:42.00 | cr2 | pH5: weird. i've changed sanity.bbclass for sha1sum |
| 21:43.15 | pH5 | cr2: I just submitted a cset that makes sanity.bbclass and base.bbclass use sha256sum |
| 21:43.28 | cr2 | ok. |
| 21:43.29 | pH5 | goxboxlive: I never had problems with AT+CFUN=0/1 |
| 21:43.37 | pH5 | what kind of debug spam do you see? |
| 21:43.38 | goxboxlive | take a look here: http://pastebin.ca/420118 |
| 21:43.54 | goxboxlive | Is it a bug? |
| 21:45.04 | goxboxlive | i guess if it had reported back "OK" instead of "0" it wouldnt have spammed me |
| 21:45.25 | pH5 | goxboxlive: that's a bug in the chip. does it reset the at command interpreter when you +CFUN=1 ? |
| 21:46.10 | goxboxlive | it gives me "0" |
| 21:46.14 | pH5 | goxboxlive: you should send that log to the gsmd-devel mailing list. I think we should make gsmd understand numerical replies, too. |
| 21:46.46 | pH5 | goxboxlive: giving you "0" instead of "OK" means that the chip switched back to V0 itself |
| 21:46.53 | pH5 | (gsmd switches to V1 in the first command) |
| 21:46.53 | goxboxlive | ok |
| 21:47.01 | goxboxlive | right |
| 21:47.25 | cr2 | strange. |
| 21:47.33 | pH5 | I wonder if it also forgot the rest of the previous commands |
| 21:47.42 | cr2 | why doesn't it happen in wince ? |
| 21:47.47 | pH5 | (CRC, CREG, CMEE, CLIP, COLP, CFUN and @HTCCSQ) |
| 21:48.08 | pH5 | cr2: maybe it's just the V setting - wince always uses V0 |
| 21:48.58 | cr2 | ah. right. |
| 21:49.14 | goxboxlive | pH5: the phone works after all the AT commands in the list has been called |
| 21:49.29 | cr2 | parsing 0 is probably easier than OK. |
| 21:58.59 | cr2 | Kevin2: so my question is now: how we setup the serial clocks as they are setup on wince ? do we need to do it exactly this way ? how do we calculate if we need 12MHz or 16.something MHz ? |
| 22:00.31 | Kevin2 | <PROTECTED> |
| 22:02.47 | cr2 | yes, the way it is done on hermes is even more interesting. |
| 22:06.14 | cr2 | serial console in linux is implemented not in a very useful way. seems to be the braindead x86 vga console legacy thing. |
| 22:06.48 | *** join/#htc-linux TeringTuby (n=maarten@249-116-dyn.ipact.nl) |
| 22:07.39 | cr2 | srm_printk(). that was done right. |
| 22:23.23 | goxboxlive | cr2: how do i make the mic work when doing a phone call. Wasent that something with i2c ? |
| 22:23.57 | Kevin2 | cr2: Hrmm. paging_init() will clear the mapping to the uart. However, paging_init() will also call into mdesc->map_io() which is htchermes_map_io() |
| 22:24.59 | cr2 | goxboxlive: yes. you need to send some i2c commands. |
| 22:25.28 | goxboxlive | do know wich i2c commands? |
| 22:25.56 | cr2 | Kevin2: maybe we can remove the IR port from s3c24xx_serial_... and map is as some funny io_device ? |
| 22:26.13 | Kevin2 | cr2: Do you have //s3c24xx_init_io(htchermes_iodesc, ARRAY_SIZE(htchermes_iodesc)); |
| 22:26.28 | cr2 | Kevin2: yes. |
| 22:26.29 | Kevin2 | That is, do you have that line commented out? It looks like it should be run. |
| 22:26.37 | cr2 | heh. |
| 22:26.44 | Kevin2 | There is all sorts of cpu init stuff that it does. |
| 22:27.07 | cr2 | even if the iodesc is empty ? |
| 22:27.24 | Kevin2 | cr2: Yes. See cpu.c |
| 22:27.27 | *** join/#htc-linux food (n=xxx@55.pool85-52-227.static.orange.es) |
| 22:27.52 | cr2 | goxboxlive: it's in the ak4641 datasheet. i may write a summary in the next days. |
| 22:28.07 | goxboxlive | good |
| 22:28.15 | cr2 | Kevin2: well, uncomment it and tell me if you'll see the penguin :) |
| 22:28.21 | goxboxlive | well it's gettin late, night |
| 22:28.33 | cr2 | good night. |
| 22:28.41 | Kevin2 | cr2: Sure, but I don't have ir working. :-) |
| 22:29.59 | cr2 | ok, i'm trying it now. |
| 22:30.40 | cr2 | btw, why do you need IR ? i hope that the vsfb will work. |
| 22:33.24 | cr2 | it entered the paging_init() but then i don't see anything in the fb. |
| 22:33.51 | cr2 | the serial console does not work too. |
| 22:34.37 | Kevin2 | cr2: Can you put a printascii() after s3c24xx_init_io(..) in htchermes_map_io()? |
| 22:34.50 | cr2 | yes. |
| 22:35.22 | *** join/#htc-linux Foxdie (i=Foxdie@77-98-48-193.cable.ubr05.harb.blueyonder.co.uk) |
| 22:37.43 | cr2 | crashed in s3c24xx_init_io(htchermes_iodesc, ARRAY_SIZE(htchermes_iodesc)); |
| 22:38.03 | cr2 | the printascii() before it is on the screen. |
| 22:38.46 | Kevin2 | You put a printascii in htchermes_map_io before s3c24xx_init_io? |
| 22:38.54 | cr2 | yes. |
| 22:39.02 | cr2 | <PROTECTED> |
| 22:39.15 | cr2 | do you know what is that in mach-rx3715.c ? |
| 22:39.16 | Kevin2 | I didn't think that would work, because s3c24xx_init_io is what maps in the serial ports again. |
| 22:40.38 | cr2 | i have |
| 22:40.41 | cr2 | <PROTECTED> |
| 22:40.41 | cr2 | <PROTECTED> |
| 22:40.41 | cr2 | <PROTECTED> |
| 22:40.56 | cr2 | and see FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF on the notebook. |
| 22:42.35 | cr2 | /* ISA style IO, for each machine to sort out mappings for, if it |
| 22:42.35 | cr2 | <PROTECTED> |
| 22:42.35 | cr2 | <PROTECTED> |
| 22:42.35 | cr2 | #define S3C24XX_VA_ISA_WORD S3C2410_ADDR(0x02000000) |
| 22:42.36 | cr2 | #define S3C24XX_VA_ISA_BYTE S3C2410_ADDR(0x03000000) |
| 22:44.28 | cr2 | 0x18xxxxxxxx is free on hermes |
| 22:44.46 | cr2 | so we can copy this part from rx3715. |
| 22:45.05 | Kevin2 | I think the rx3715 just wants to allocate a mapping to CS3 (the third memory chip) - I don't think it needs to be present. |
| 22:45.16 | cr2 | but g500 does not define anything and it works. |
| 22:45.31 | cr2 | ok. |
| 22:45.44 | cr2 | so why i don't see GGGGGGGGGGGGG ? |
| 22:46.16 | *** join/#htc-linux lilliput (n=Touriste@82-47-146-8.cable.ubr11.brad.blueyonder.co.uk) |
| 22:46.27 | Kevin2 | Hrmm. I don't think you can reliably write out in mdesc->map_io() -- see paging_init in mmu.c -- it hasn't yet called the local_flush_tlb_all |
| 22:46.36 | Kevin2 | So, the TLB mappings may be totally in flux. |
| 22:47.44 | cr2 | ok. |
| 22:49.00 | Kevin2 | Hrmm. Maybe one could manually call local_flush_tlb_all(); flush_cache_all(); |
| 22:50.29 | cr2 | i guess you can, but the problem is imho somewhere else. |
| 22:50.58 | cr2 | well, decoding the clock setup is deterministic, so maybe i should think about it. |
| 22:52.01 | *** join/#htc-linux Ralith (n=ralith@166.194.224.109) |
| 22:54.00 | cr2 | ok, i'll go to bed. good night. |
| 22:54.22 | Kevin2 | Good night. I'll try to get ir working on apache. |
| 23:02.46 | pancake | is there any lightweight http server for wm? |