| 00:50.12 | *** join/#openezx jonwil (n=jonwil@123-243-219-11.tpgi.com.au) |
| 00:50.40 | *** join/#openezx tombhadAC (n=tombhad@pD95DF7D1.dip.t-dialin.net) |
| 00:52.15 | *** join/#openezx Fosforo (n=Fosforo@20151094126.user.veloxzone.com.br) |
| 00:58.16 | jonwil | anyone here know anything much about kernel hacking? |
| 01:20.44 | *** join/#openezx Fosfor1 (n=Fosforo@20151006181.user.veloxzone.com.br) |
| 03:06.11 | *** join/#openezx l403 (n=l@85.132.159.239) |
| 03:34.25 | *** join/#openezx flyhorse|linux (n=flyhorse@116.18.178.179) |
| 03:48.33 | *** join/#openezx yakk (n=Miranda@77.52.91.27) |
| 04:09.24 | *** join/#openezx bartv (n=bart@thisnet.xs4all.nl) |
| 06:57.35 | *** join/#openezx pleemans (n=toi@d54C2AAB7.access.telenet.be) |
| 07:48.55 | *** join/#openezx darion76 (i=darion@94-248-75-87.dynamic.peoplenet.ua) |
| 07:51.21 | sabrod | good morning. |
| 08:28.42 | *** join/#openezx johanbr (n=j@JBrannlund2.MathStat.Dal.Ca) [NETSPLIT VICTIM] |
| 08:28.42 | *** join/#openezx Gottox (n=Gottox@s01.de) |
| 08:28.42 | *** join/#openezx methril (n=Methril@213.27.233.98) |
| 08:28.42 | *** join/#openezx uwe_ (n=uwe_@dslb-084-056-024-012.pools.arcor-ip.net) |
| 08:41.13 | *** join/#openezx l403 (n=l@85.132.159.239) |
| 09:33.22 | *** join/#openezx florian_kc (n=fuchs@217.146.132.69) |
| 09:53.06 | *** join/#openezx LoneStar (n=xinzhen@218.80.193.59) |
| 09:55.39 | *** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1) |
| 10:53.09 | *** join/#openezx LoneStar (n=xinzhen@218.80.193.59) |
| 11:16.03 | *** join/#openezx sabrod (n=mac@lns-bzn-60-82-254-212-182.adsl.proxad.net) |
| 11:45.46 | *** join/#openezx WyrM (n=wyrm@20158159221.user.veloxzone.com.br) |
| 11:46.26 | *** join/#openezx gmzhuo (i=gmzhuo@125.70.87.165) |
| 11:47.01 | WyrM | hi |
| 11:47.36 | gmzhuo | WyrM,hi |
| 11:56.18 | ao2 | hi |
| 11:59.06 | *** join/#openezx LoneStar (n=xinzhen@114.92.171.72) |
| 12:00.02 | WyrM | gmzhuo: I updated QT here to 4.4.3, but it still doesn't compile. |
| 12:00.11 | WyrM | ao2 made a patch to make it compile on linux. |
| 12:03.32 | ao2 | WyrM, what is the problem? |
| 12:04.05 | WyrM | heh, I think I just need your patch. |
| 12:04.13 | WyrM | I haven't saved it. |
| 12:04.34 | WyrM | http://pastebin.com/m78cd172f |
| 12:04.37 | ao2 | http://people.openezx.org/ao2/tmp/phonetools_linux.diff |
| 12:04.50 | *** join/#openezx jonwil (n=jonwil@123-243-219-11.tpgi.com.au) |
| 12:05.03 | ao2 | yes, that should be it |
| 12:05.14 | WyrM | thanks! |
| 12:05.18 | jonwil | hi |
| 12:05.21 | WyrM | I will commit this to svn. |
| 12:05.26 | WyrM | hi jonwil. |
| 12:06.34 | ao2 | well, my patch is linux only ATM, the #if 0 should turn to something platform dependent |
| 12:06.50 | ao2 | platform -> OS |
| 12:07.49 | WyrM | argh, commited with the .obj files. |
| 12:09.05 | ao2 | also it could be intersting to let svn handle the end-of-line, so we get the right one on the system we checkout on. But I haven't done this before, let me search soemthing about that |
| 12:13.05 | WyrM | does QT define any macro when running on windows? |
| 12:13.32 | WyrM | so we can replace the #if 0 with something like #ifndef WIN32 |
| 12:15.35 | CIA-15 | OpenEZX: 03wyrm * r2392 10/trunk/src/host/phonetools/ (50 files in 7 dirs): Initial import |
| 12:15.53 | ao2 | it should, maybe it even provides some portable snprintf equivalent but I don't know Qt |
| 12:16.07 | WyrM | me neither. |
| 12:24.11 | gmzhuo | me neither :0 |
| 12:24.17 | WyrM | lol |
| 12:24.19 | CIA-15 | OpenEZX: 03wyrm * r2393 10/trunk/src/host/phonetools/ (UsbPhone.cpp setup.h): apply ao2's patch to make it compile on linux |
| 12:24.40 | WyrM | gmzhuo: can you take a look at it? and try to make it compatible with windows again? |
| 12:25.47 | gmzhuo | let me try |
| 12:25.52 | WyrM | ao2: I have another error compiling, seems that the generated makefile is not trying to link libusb. |
| 12:26.22 | ao2 | yes, indeed I fixed that by hand and forgot about it |
| 12:27.09 | WyrM | what does qmake read to create the Makefile? |
| 12:27.15 | WyrM | ls |
| 12:27.20 | WyrM | oops |
| 12:27.22 | WyrM | :) |
| 12:27.26 | ao2 | WyrM, http://www.qtforum.org/article/2333/Operating-System-Detection.html |
| 12:29.09 | gmzhuo | Wyrm,it seems should add something in project file |
| 12:29.10 | CIA-15 | OpenEZX: 03wyrm * r2394 10/trunk/src/host/phonetools/UsbPhone.cpp: s/0/Q_WS_WIN/ |
| 12:31.23 | CIA-15 | OpenEZX: 03wyrm * r2395 10/trunk/src/host/phonetools/phonetools.pro: add libusb |
| 12:31.35 | WyrM | like r2395? |
| 12:31.53 | WyrM | now it compiles here with "qmake && make" |
| 12:34.07 | gmzhuo | Should we delete generated file\debug\release? |
| 12:34.22 | WyrM | yes. |
| 12:34.47 | WyrM | but since I don't know which files were generated, I need a list :) |
| 12:35.58 | WyrM | I did a make clean before commit, but as it seems they were generated by a windows build. |
| 12:38.48 | gmzhuo | Something wrong with my svn on windows |
| 12:51.12 | gmzhuo | Q_WS_WIN in vs2005 dosn't work |
| 12:52.51 | gmzhuo | use #ifdef WIN32 for windows |
| 12:53.49 | CIA-15 | OpenEZX: 03wyrm * r2396 10/trunk/src/host/phonetools/UsbPhone.cpp: s/Q_WS_WIN/WIN32 |
| 12:54.17 | gmzhuo | Wyrm,you coment out of the map&unmap? |
| 12:54.28 | WyrM | no |
| 12:54.51 | WyrM | is it commented out? |
| 12:54.59 | WyrM | iirc I downloaded again. |
| 12:55.18 | gmzhuo | yes |
| 12:55.30 | gmzhuo | and no arpa/inet.h in windows |
| 12:55.55 | CIA-15 | OpenEZX: 03wyrm * r2397 10/trunk/src/host/phonetools/phonetools.cpp: enable qt map/unmap again |
| 12:56.26 | CIA-15 | OpenEZX: 03wyrm * r2398 10/trunk/src/host/phonetools/UsbPhone.cpp: no arpa/inet on win |
| 12:59.19 | gmzhuo | files phonetools.ncb,phonetools.suo and files in debug\release\generatedfiles should be delete |
| 13:00.20 | CIA-15 | OpenEZX: 03wyrm * r2399 10/trunk/src/host/phonetools/ (28 files in 6 dirs): delete generated files |
| 13:02.45 | gmzhuo | every thing is ok now |
| 13:03.09 | WyrM | ui.phoneStatus->setText(tr("Not find!")); |
| 13:03.17 | WyrM | can I just change the text? |
| 13:03.30 | WyrM | or it will break translation? |
| 13:05.29 | gmzhuo | Yes,you can change the text,and regenerate the translation file if you need! |
| 13:05.48 | CIA-15 | OpenEZX: 03wyrm * r2400 10/trunk/src/host/phonetools/phonetools.cpp: correct string |
| 13:05.55 | WyrM | how do I regenerate the translation file? |
| 13:06.23 | gmzhuo | I don't know too:) |
| 13:06.32 | WyrM | heh |
| 13:06.47 | WyrM | are you using it with chinese? or english? |
| 13:08.00 | gmzhuo | English now |
| 13:10.10 | CIA-15 | OpenEZX: 03wyrm * r2401 10/trunk/src/host/phonetools/UsbPhone.cpp: warning: ?\226?\128?\152err?\226?\128?\153 may be used uninitialized in this function |
| 13:15.42 | CIA-15 | OpenEZX: 03wyrm * r2402 10/trunk/src/host/phonetools/ (UsbPhone.cpp UsbPhone.h phonetools.cpp): remove unused var |
| 13:26.32 | CIA-15 | OpenEZX: 03wyrm * r2403 10/trunk/src/host/phonetools/ (UsbPhone.cpp UsbPhone.h): fix remaining (const char* conversion) warnings. |
| 13:26.53 | WyrM | now compilation looks good on linux :) |
| 13:27.02 | WyrM | does it still compiles on windows? |
| 13:32.31 | gmzhuo | yes,still works on windows |
| 13:33.02 | WyrM | good |
| 13:33.10 | WyrM | now noobs can use openezx too. |
| 13:33.12 | WyrM | lol |
| 13:34.50 | WyrM | gmzhuo: please, put a windows .exe on your people account and announce boot_usb for windows on openezx-devel. |
| 13:39.48 | jonwil | Anyone here know of any good references on ARM assembly? |
| 13:40.08 | WyrM | yes. |
| 13:40.19 | WyrM | what do you need? |
| 13:40.36 | WyrM | what are you using to disassemble? IDA? objdump? |
| 13:40.59 | jonwil | IDA |
| 13:41.07 | jonwil | I know x86 asm inside out |
| 13:41.10 | jonwil | but I dont know ARM |
| 13:41.20 | jonwil | I want something that can teach me ARM asm |
| 13:41.34 | jonwil | not just a reference manual or something, something that teaches me ARM |
| 13:41.41 | jonwil | more specifically the right kind of ARM for motomagx |
| 13:41.51 | jonwil | including the Thumb stuff whatever that is |
| 13:43.00 | WyrM | heh |
| 13:43.16 | WyrM | the asm reference shouldn't be different for magx |
| 13:43.21 | jonwil | no true |
| 13:43.25 | jonwil | I have the ARM ARM somewhere |
| 13:43.27 | jonwil | but thats just a reference |
| 13:43.31 | jonwil | not a tutorial |
| 13:45.55 | WyrM | I learned arm asm from some reference cards. |
| 13:46.03 | WyrM | im not aware of any tutorial sorry. |
| 13:46.05 | jonwil | ok |
| 13:47.33 | jonwil | What I wish to do is to reverse engineer all steps taken by moto bootloader when booting normally (i.e. up to the bit where it calls lbl) and when booting into flash mode |
| 13:48.18 | *** join/#openezx darion76 (i=darion@94-248-75-87.dynamic.peoplenet.ua) |
| 13:48.27 | jonwil | First step is to write program which loads like ramdownloader (onto NoRSA phone of course) and then does steps to make phone boot normally |
| 13:48.37 | jonwil | i.e. it undoes any hardware settings set by flash mode codepath |
| 13:48.43 | jonwil | does hardware settings set by normal codepath |
| 13:48.50 | jonwil | and calls LBL in the same way as MBM does |
| 13:49.06 | WyrM | "undoes" its not always possible sorry. |
| 13:49.28 | WyrM | eg, on ezx we can't undo the "put BP on flash mode". |
| 13:49.28 | jonwil | well if its not possible we need to know that too so we can find workaround |
| 13:49.41 | jonwil | yeah EZX hardware is annoying |
| 13:49.50 | jonwil | magx hardware is less complex than EZX IMO |
| 13:49.58 | jonwil | one CPU plus one DSP instead of the 2-cpu of EZX |
| 13:50.25 | jonwil | and the DSP is controled directly from ARM side |
| 13:50.39 | jonwil | i.e. ARM side tells starcore what code to run and when |
| 13:50.48 | jonwil | so there is no "flash mode" for starcore AFAIK |
| 13:51.24 | WyrM | what exactly do you want to do? |
| 13:51.33 | WyrM | run a kernel like we do with boot_usb? |
| 13:51.38 | jonwil | yes |
| 13:52.12 | WyrM | does the NoRSA stuff stops the sig checking on RDL? |
| 13:52.13 | jonwil | more specifically run a kernel (e.g. one built from mot sources) and a mot rootfs (modified or otherwise, flashed on NAND or stored on MMC or whatever) and get a 100% working phone |
| 13:52.29 | jonwil | It stops all sig checking done by MBM |
| 13:52.41 | jonwil | i.e. it makes the RSA check always return true |
| 13:52.42 | WyrM | then just load the kernel and jump to it. |
| 13:52.44 | jonwil | even if signature check fails |
| 13:53.09 | jonwil | Obviously we cant just load the kernel, we need to load some variant of LBL which then loads the kernel |
| 13:53.17 | WyrM | no. |
| 13:53.41 | jonwil | that or we need to do something similar to what LBL does |
| 13:54.01 | WyrM | you can just send _one_ instruction to set the machid, and if the kernel has all the code/drivers and a builtin cmdline it will run. |
| 13:54.36 | jonwil | so yeah we need to build info normally given to kernel by LBL into kernel itself then |
| 13:54.49 | WyrM | what info? |
| 13:54.58 | WyrM | boot_usb support standard linux atags. |
| 13:55.16 | WyrM | it builds the atags on your host and uploads it, the same way it uploads the code. |
| 13:55.44 | WyrM | and it also supports setting the machid via some asm magic that is uploaded just before the kernel. |
| 13:55.45 | jonwil | ok |
| 13:56.02 | jonwil | ok, on magx, MBM passes a BOOT_INFO structure to LBL |
| 13:56.22 | WyrM | you can create this struct on host, and upload it to memory. |
| 13:56.32 | WyrM | its easier than writing another bootloader. |
| 13:56.42 | *** join/#openezx tombhadAC (n=tombhad@pD95DEB47.dip.t-dialin.net) |
| 13:56.43 | jonwil | no, we already have lbl code |
| 13:56.52 | jonwil | lbl is GPL |
| 13:56.58 | WyrM | so? |
| 13:57.06 | jonwil | ok |
| 13:57.22 | jonwil | so yeah I guess we can build kernel info differently ala boot_usb |
| 13:58.04 | WyrM | I mean, its easier to MBM -> kernel. instead of MBM -> modified lbl -> kernel. |
| 13:59.21 | jonwil | true |
| 13:59.38 | WyrM | also, you will probably not need the BOOT_INFO struct at all. |
| 13:59.52 | jonwil | I guess we just need some fu so that whatever pieces of info LBL passes to the kernel get filled in inside the kernel when compiling |
| 13:59.54 | WyrM | because it is used by LBL, not by linux kernel. |
| 14:00.03 | jonwil | or get fed to the kernel some other way |
| 14:00.07 | WyrM | yes. |
| 14:00.12 | WyrM | exactly. |
| 14:00.17 | jonwil | btw there IS some custom moto specific atags mentioned in the lbl source |
| 14:00.21 | jonwil | so we need to support all of those |
| 14:00.39 | WyrM | not really. |
| 14:00.47 | WyrM | gen2 also have some custom atags. |
| 14:00.47 | jonwil | we need to support the ones that matter |
| 14:00.52 | WyrM | we just ignore then :) |
| 14:00.54 | jonwil | ok |
| 14:01.09 | WyrM | but we don't use a motorola based kernel.. |
| 14:01.19 | WyrM | so it doesn't matter for OpenEZX kernel. |
| 14:01.49 | jonwil | thats the difference, on ezx, kernel is stock 2.6.x with a pile of changes (some using hacked up code from mot kernel, some not) |
| 14:02.14 | jonwil | on magx, custom kernel work will likely be modified mot kernel or possibly mot kernel ported to newer 2.6.x version and then modified |
| 14:02.23 | jonwil | i.e. most of the mot custom code will remain |
| 14:02.28 | WyrM | hum, |
| 14:02.34 | WyrM | this is not the way to go. |
| 14:02.47 | jonwil | y not? |
| 14:02.54 | jonwil | Is mot code not good enough? |
| 14:02.58 | WyrM | the "non-standard" foo that motorola implemented should be removed. |
| 14:03.12 | jonwil | true, some things in mot kernel should go |
| 14:03.13 | WyrM | and rewritten using standard methods. |
| 14:03.18 | jonwil | like Belcarra USB crap |
| 14:04.01 | WyrM | not only belcarra, the "wake up reason" atag is another example. |
| 14:04.10 | jonwil | ok |
| 14:04.29 | jonwil | in any case i am not a kernel devloper |
| 14:04.39 | jonwil | so I will leave all this stuff up to people who are |
| 14:05.08 | WyrM | lets get back on topic :) |
| 14:05.33 | WyrM | does your magx phone preserve ram data for how much time without battery? |
| 14:05.46 | jonwil | no clue |
| 14:05.53 | WyrM | try it. :) |
| 14:06.13 | jonwil | well I know that if I pull battery, phone shuts down immediatly |
| 14:06.19 | WyrM | yes. |
| 14:06.21 | jonwil | I somtimes have to do that if phone freezes up |
| 14:06.38 | WyrM | on ezx the ram is preserved even after some minutes without battery. |
| 14:07.08 | WyrM | I use this to debug code when I have no other debug methods. |
| 14:07.09 | jonwil | I also know that if I pull battery or mess in bootloader or flash stuff or whatever (not sure exactly what yet), the clock goes out of sync |
| 14:07.13 | jonwil | and no longer records the actual time |
| 14:07.35 | jonwil | possibly its seeing that something caused it to loose its time settings so it pulls time from network or something |
| 14:07.41 | jonwil | and network has wrong time |
| 14:07.44 | jonwil | I dont know |
| 14:09.23 | WyrM | with you jail break tools, do you need to upload RDL to read back RAM contents? |
| 14:09.38 | WyrM | or the bootloader supports reading from RAM? |
| 14:09.42 | jonwil | I have no way to read RAM |
| 14:09.48 | jonwil | only NAND flash |
| 14:11.11 | jonwil | also my linux box is still a little flakey so I cant even do that right now (but I do have a motorola firmware file matching the firmware on my phone which is what I am using as a backup in case I flash something that breaks phone |
| 14:12.32 | gmzhuo | jonwil,does the kernel you compiled enter into userspace now? |
| 14:13.21 | jonwil | the kernel I compiled (and posted on the ML about) works 100% except for one thing, USB wont work. Plug in AC charger and it picks it up and charges phone |
| 14:13.27 | jonwil | but plug in USB cable and it wont pick it up at all |
| 14:13.35 | jonwil | windows cant even find phone |
| 14:13.46 | jonwil | none of the USB kernel modules are being loaded for a reason I cant figure out |
| 14:13.58 | jonwil | although other kernel modules DO load |
| 14:14.07 | jonwil | so its not a case of "you rebuilt kernel, you need to replace all |
| 14:14.15 | jonwil | kernel modules with new rebuilt ones" |
| 14:14.21 | jonwil | at least I dont think it is |
| 14:14.36 | gmzhuo | can you make bluetooth work? |
| 14:14.51 | jonwil | unsure as I have nothing else bluetooth to test with |
| 14:14.53 | gmzhuo | then you can have a console on bluetooth |
| 14:15.19 | jonwil | well actually I think phone came with funky wierd set of stereo bluetooth earphones for music |
| 14:15.22 | jonwil | but thats about it |
| 14:15.31 | jonwil | so i have nothing for phone to talk to |
| 14:17.27 | gmzhuo | If bluetooth can work,have a console on buletooth can easy your debug,even for usb debug,sorry for my poor englisth |
| 14:17.30 | gmzhuo | If bluetooth can work,have a console on buletooth can easy your debug,even for usb debug,sorry for my poor englists |
| 14:17.36 | gmzhuo | If bluetooth can work,have a console on buletooth can easy your debug,even for usb debug,sorry for my poor english |
| 14:17.48 | gmzhuo | :( |
| 14:18.21 | gmzhuo | WyrM,I fogot one test button,jump button |
| 14:20.07 | WyrM | gmzhuo: send patch. |
| 14:20.27 | WyrM | jonwil: looking at /proc/kmsg will point you the problem with usb modules. |
| 14:20.40 | WyrM | easier than diff on kallsyms. |
| 14:20.46 | jonwil | except with no USB I have no way to see /proc/kmsg |
| 14:20.55 | WyrM | how do you see kallsyms? |
| 14:20.57 | WyrM | :) |
| 14:22.11 | WyrM | when developing for ezx, without USB, I used Bluetooth like gmzhuo pointed, but also used changed rootfs to dump stuff to the MMC. |
| 14:23.10 | WyrM | we straced the tapisrv app (handles communication with the modem) at the phone bootup by hacking the rootfs to add strace and to invoke tapisrv with strace logging to the mmc. |
| 14:24.22 | jonwil | getting a call going on magx with no mot binaries is going to be hard due to funky GSM stack |
| 14:25.04 | jonwil | although hopefully the really nasty parts (such as the proprietary encryption and the voice codec) are either done on the Starcore or can be liffed from some other oss project somewhere |
| 14:25.33 | WyrM | jonwil: how do you see kallsyms without usb? |
| 14:26.16 | *** join/#openezx tombhadAC (n=tombhad@pD95DCAAC.dip.t-dialin.net) |
| 14:26.30 | jonwil | I have hacked phone that allows shell scripts to run and was able to get a dump of it that way |
| 14:26.42 | jonwil | but the same trick didnt work for kmsg |
| 14:27.11 | WyrM | hum. |
| 14:27.15 | WyrM | it should :) |
| 14:27.36 | WyrM | have you tried to dump the insmod/modprobe output at least? |
| 14:29.28 | jonwil | no, havent tried that |
| 14:32.25 | gmzhuo | http://people.openezx.org/gmzhuo/jump.patch |
| 14:34.42 | WyrM | gmzhuo: patch failed to apply. |
| 14:35.40 | gmzhuo | ah? |
| 14:36.22 | WyrM | ^M foo |
| 14:36.52 | gmzhuo | What's wrong with it? |
| 14:37.03 | gmzhuo | I use the svn generated patch |
| 14:37.29 | WyrM | I will edit the files that it failed manually. |
| 14:37.53 | WyrM | 1 out of 1 hunk FAILED -- saving rejects to file phonetools.cpp.rej |
| 14:37.53 | WyrM | 1 out of 1 hunk FAILED -- saving rejects to file phonetools.h.rej |
| 14:38.11 | gmzhuo | Maybe I should update first |
| 14:39.06 | sabrod | i enhanced the ui , by my side ;) , if you want a static build, it can do: |
| 14:39.36 | sabrod | pastebin.com/m347ae341 |
| 14:40.30 | CIA-15 | OpenEZX: 03wyrm * r2404 10/trunk/src/host/phonetools/ (phonetools.cpp phonetools.h phonetools.ui): add jump code |
| 14:40.31 | WyrM | should I apply this pastebin too? |
| 14:40.43 | sabrod | it's some tips, and a larger windows... |
| 14:41.45 | WyrM | sabrod: please, send it as an attachment via email to me. |
| 14:41.45 | WyrM | so I can apply/test here. |
| 14:41.51 | sabrod | oh, but i diff -dru based on an older version... |
| 14:42.04 | sabrod | ok, i send that. |
| 14:42.08 | WyrM | can you rebase on current svn? |
| 14:42.32 | sabrod | yeah, and it wil looks cleaner ;) |
| 14:42.36 | sabrod | maybe... |
| 14:42.48 | sabrod | :) |
| 14:42.53 | WyrM | thanks! |
| 15:00.00 | sabrod | gmzhuo: what the jump button is for? |
| 15:00.30 | WyrM | to execute code on an arbitrary memory position. |
| 15:01.01 | WyrM | you can upload something to RAM, and then use the JUMP to execute it. |
| 15:02.00 | sabrod | ok, it ask for an address? |
| 15:02.13 | WyrM | I think you use the start address for it. |
| 15:02.23 | sabrod | can we place something like size: A000-8000 . :) |
| 15:02.24 | WyrM | since jump is always used after an upload. |
| 15:04.21 | sabrod | never upload manually to the ram... |
| 15:04.32 | sabrod | is an human! |
| 15:04.44 | sabrod | :) |
| 15:04.48 | WyrM | anything from 0xa0000000 to 0xa1ffffff-4 is valid to jump to. |
| 15:05.14 | sabrod | ok... |
| 15:05.18 | WyrM | heh, there is nothing wrong in using upload/jump. |
| 15:05.39 | WyrM | but 99% of people would use only the kernel image upload anyway :) |
| 15:07.15 | sabrod | sooner, i was asking if we could place a substraction, in the fields, can we? |
| 15:07.30 | WyrM | substraction? |
| 15:08.03 | sabrod | yes, offset-size before.. |
| 15:08.07 | WyrM | size fiels are only used for download. |
| 15:08.17 | WyrM | for uploads the file size is used. |
| 15:08.39 | sabrod | brrr! |
| 15:09.02 | sabrod | it never stops ;) |
| 15:09.33 | WyrM | also, offset-size makes no sense :) |
| 15:09.42 | WyrM | offset is the start, not the end. |
| 15:10.02 | sabrod | well, of course if you choose to write before the bootloader... |
| 15:10.37 | WyrM | ahh, you mean a security feature to avoid noobs from flashing the original bootloader. |
| 15:11.02 | sabrod | Oops |
| 15:11.27 | WyrM | invalid addresses will not work, they are checked on boot_usb/phonetools AND on gen-blob. |
| 15:11.29 | sabrod | well, no, iwas thinking that it doesn't matter to not specify the size when flashing... |
| 15:12.02 | WyrM | ahh, I haven't tested phonetools, but for boot_usb you don't need to specify the size on upload/flash |
| 15:12.09 | sabrod | As if you begin before the bootloader, only start address will brick the phone. |
| 15:13.53 | WyrM | sabrod: btw, im counting on you to test everything on windows, and to advertise phonetools on motorolafans. |
| 15:13.55 | *** join/#openezx ao2 (n=u@2001:1418:117:0:0:0:0:1) |
| 15:13.59 | WyrM | :) |
| 15:14.35 | gmzhuo | WyrM,you mean we can't flash the flash 0~0xa0000 through gen-blob? |
| 15:15.12 | sabrod | No, i you CAN! |
| 15:15.59 | *** join/#openezx methril (n=Methril@213.27.233.98) |
| 15:16.01 | gmzhuo | then it's better to protected in boot_usb & phonetools |
| 15:16.10 | sabrod | it's now a superb tool, btw. |
| 15:16.29 | gmzhuo | it's so dangerous |
| 15:16.31 | WyrM | err, my mistake. |
| 15:17.03 | sabrod | a red alert, maybe if a value between 0 and a0000 is found? |
| 15:17.15 | WyrM | its 0x0~0x03f the allowed range to flash, 0xa0~0xa1f the allowed range for upload. |
| 15:18.08 | gmzhuo | I think alert is not enouth |
| 15:18.28 | WyrM | a red alert/confirmation if a value between 0x0 and 0x20000 (128k) for gen1, or 0x40000(256k) for gen2. |
| 15:19.21 | WyrM | also, muromec is reporting that he can't flash the last 32mb on gen2. |
| 15:19.26 | WyrM | I need to confirm that. |
| 15:19.37 | WyrM | altough iirc I tested this already and it worked here. :/ |
| 15:19.51 | gmzhuo | I think should be prohibit,if some one "real" need to flash this range,He/She can recompiled the boot_usb/phonetools |
| 15:20.02 | WyrM | gmzhuo: ack |
| 15:20.22 | WyrM | but the problem is that its different on gen1/gen2 |
| 15:20.38 | WyrM | if you lock gen2 bootloader address, you will also lock gen1 KERNEL address. |
| 15:20.57 | gmzhuo | :( |
| 15:21.53 | WyrM | I think we should add a confirmation dialog after clicking the flash button. |
| 15:22.05 | WyrM | or a radio button to select the phone type. |
| 15:22.48 | WyrM | a radio to select the phone type would also be good to have a machid table. |
| 15:22.59 | WyrM | so people don't need to remember the phone's machid. :) |
| 15:23.47 | gmzhuo | I support to select phone type! |
| 15:24.00 | ao2 | gmzhuo, how did you design the gui? |
| 15:24.48 | gmzhuo | I use the Qt Visual Studio Integration |
| 15:25.10 | gmzhuo | or you can use qt designer on linux |
| 15:25.24 | ao2 | ok, thanks |
| 15:27.09 | gmzhuo | My pleasure:) |
| 15:28.30 | gmzhuo | I don't when will the gui will be ok on my E6 |
| 15:28.45 | WyrM | ? |
| 15:29.05 | gmzhuo | Just worry about it |
| 15:29.32 | WyrM | you mean X on the device? |
| 15:29.37 | gmzhuo | yes |
| 15:29.52 | darion76 | hi |
| 15:29.58 | WyrM | well, if nobody cares to work at it, it will never be ok :) |
| 15:30.06 | WyrM | hi darion76. :) |
| 15:30.15 | gmzhuo | yes |
| 15:30.19 | darion76 | I work on it |
| 15:30.21 | gmzhuo | Hi,darion |
| 15:30.31 | WyrM | darion76: status? |
| 15:31.19 | darion76 | I have 18bpp implenentation with libpixman |
| 15:31.28 | darion76 | still some bugs |
| 15:31.49 | darion76 | but colors are correct |
| 15:31.49 | WyrM | so, the issue was not on X? |
| 15:31.55 | gmzhuo | I don't know when will I replace the phone for 3G |
| 15:31.57 | darion76 | yes |
| 15:32.48 | WyrM | I pointed some time ago that plain X colors were ok, GPE had color offset problems somewhere and EFL had color+alignment problems. |
| 15:33.11 | gmzhuo | Why we not use qt? |
| 15:33.28 | darion76 | I work with gpe image |
| 15:33.43 | sabrod | pastebin.com/m6dc81bf3 |
| 15:34.40 | sabrod | i suggest these stuff, i don't know why, qtdesigner add soe DOCTYPE HTML mess... |
| 15:34.43 | darion76 | gpe problem is related to libpixman - this library does not include 18bpp implementation |
| 15:34.47 | WyrM | patch: **** Only garbage was found in the patch input. |
| 15:34.54 | WyrM | heh, please send me the patch via mail. |
| 15:35.00 | sabrod | ok. |
| 15:35.08 | WyrM | pasting it from pastebin haven't worked. |
| 15:40.00 | WyrM | gmzhuo: addr/size need to be hexa? |
| 15:40.10 | gmzhuo | yes |
| 15:40.12 | WyrM | on boot_usb they can be both hexa and decimal. |
| 15:40.23 | WyrM | why haven't you preserved this code? :) |
| 15:40.47 | sabrod | so, i need to correct my tooltip... |
| 15:40.50 | gmzhuo | I don't how QString to deal this |
| 15:41.28 | sabrod | ah, no i said 'hexa'... i was sure that i have asked that... |
| 15:44.40 | WyrM | <PROTECTED> |
| 15:45.02 | WyrM | does this mean hexa? |
| 15:45.16 | WyrM | and ,10); decimal? |
| 15:45.48 | gmzhuo | yes,but I don't know how to make it auto use hexa or decimal |
| 15:48.59 | gmzhuo | if ((sscanf(argv[3], "0x%x", &size) != 1)) |
| 15:49.00 | gmzhuo | size = atoi(argv[3]); |
| 15:49.10 | gmzhuo | should be hexa too? |
| 15:49.20 | gmzhuo | in boot_usb.c |
| 15:49.33 | WyrM | yes, 0x20000 = 131072 |
| 15:49.41 | WyrM | the 0x makes the difference :) |
| 15:49.57 | gmzhuo | Oh,I miss then next for decimal,sorry |
| 15:50.45 | gmzhuo | s/then/the |
| 15:51.32 | gmzhuo | I fogot how to use purl again |
| 15:51.56 | WyrM | missing the ending / |
| 15:52.10 | WyrM | s/missing/lacking/ |
| 15:52.19 | WyrM | :) |
| 16:12.18 | *** join/#openezx pleemans (n=toi@d54C2AAB7.access.telenet.be) |
| 16:26.08 | *** join/#openezx agd5f (n=alex@c-69-143-4-28.hsd1.va.comcast.net) |
| 16:38.16 | CIA-15 | OpenEZX: 03wyrm * r2405 10/trunk/src/host/phonetools/phonetools.ui: sabrod's interface changes |
| 16:53.35 | *** join/#openezx yakk (n=Miranda@77.52.96.159) |
| 17:42.46 | *** join/#openezx p3t3r__ (n=peter@134.245.164.105) |
| 17:45.07 | ao2 | hi p3t3r__ do you remember if A910 has 32MiB or 64MiB flash? |
| 17:45.58 | p3t3r__ | ao2: no, i don't remember it, sorry. |
| 17:46.33 | sabrod | ao2: 64mb |
| 17:47.09 | sabrod | hi peter, ao2, btw :) |
| 17:48.29 | ao2 | sabrod, good, thanks. I thought that in the end (when we have a reasonably stable system) we can put rootfs on flash instead of adding mmc over spi support to gen-blob which is a lot more work. |
| 17:49.35 | sabrod | it seems it's 64mb, as every fw i extracted gave me at least 64mb of datas/files... |
| 17:49.45 | sabrod | quite a cool idea! |
| 17:50.29 | sabrod | Every solution consiting in free-ing the a910 is a good one ;) |
| 17:50.40 | sabrod | consisting |
| 17:50.40 | WyrM | ao2: mmc over spi with gen-blob is not viable. |
| 17:50.59 | sabrod | but what about kernel on flash |
| 17:51.15 | WyrM | but I can prepare a gen-blob that reads both kernels from flash similar to the old blob2 for a780. |
| 17:51.29 | sabrod | (ie the /usrlocal/ cg, that rebuid by itself at a stock system reboot) |
| 17:52.32 | WyrM | but the rootfs can be on the mmc, as linux kernel supports mmc over spi (even with some limitations/speed issues) |
| 17:52.55 | ao2 | WyrM, yes, only kernel on flash is enough I know |
| 17:53.09 | ao2 | I just forgot about that :P |
| 17:53.17 | WyrM | sabrod: its easier to put the kernel on a separate partition, on /ezxlocal it would imply implementing the filesystem driver on gen-blob. |
| 17:53.50 | WyrM | ao2: heh ;) |
| 17:54.20 | sabrod | what if we just write in the middle of ezxlocal's cg? |
| 17:54.41 | sabrod | discretely :) |
| 17:55.22 | WyrM | sabrod: usually on the rootfs CG its easier to make some room. :) |
| 17:55.31 | WyrM | or on the /usr/language CG. |
| 17:55.56 | WyrM | and also on the kernel cg itself. |
| 17:56.18 | WyrM | we need space for gen-blob(128k) and for the openezx kernel (2mb). |
| 17:56.52 | WyrM | gen-blob and the original kernel will probably fit the original kernel partition. |
| 17:57.18 | WyrM | and the openezx kernel can be put on the language or rootfs partition. |
| 17:57.25 | sabrod | i wnted the user to be able to do it, the open way, without touching the motocrap's fw too much. |
| 17:57.28 | WyrM | but not as a file, as a separate partition. |
| 17:57.46 | sabrod | the ezxlocal can 'rebuild' by itelf... |
| 17:58.00 | WyrM | not really :) |
| 17:58.11 | sabrod | well, i need t test this rebuild ability... |
| 17:58.22 | sabrod | heh |
| 17:58.24 | WyrM | it can rebuild its _files_ |
| 17:58.44 | WyrM | but the partition offset/size is set by the kernel partition map. |
| 17:58.49 | sabrod | yes, we want a reformat! |
| 17:59.11 | WyrM | and the partition size would be changed to fit the openezx kernel. |
| 18:00.05 | sabrod | and it's unsafe for the poor openezx kernel, that can be overwritten by any contact database at each moment!! |
| 18:00.49 | WyrM | yes, if the "rebuild" code does an mkfs, it would certainly overwrite the openezx kernel. |
| 18:08.17 | tmzt | WyrM: would it be possible to use part of openmoko Qi in the bootloader to allow booting kernels from ext2, vfat filesystems on flash and sd card? |
| 18:08.24 | tmzt | how much room is there? |
| 18:08.59 | tmzt | gen-blob for bootloader |
| 18:09.03 | WyrM | tmzt: 128k, enough room for an ext driver. |
| 18:09.17 | WyrM | but we can increase this to up to the original kernel sizer. |
| 18:09.29 | WyrM | 896k on gen1 1024k on gen2. |
| 18:09.59 | WyrM | and fat is already supported by gen-blob. |
| 18:10.11 | WyrM | as is flash and mmc (on pxa mmc controller). |
| 18:11.35 | WyrM | all we need is a minimal, read only ext implementation, and we could put our kernels on /boot on the rootfs :) |
| 18:12.12 | tmzt | that's how Qi works, wiki openmoko.org Qi |
| 18:12.32 | tmzt | it reads a config file from /boot on either internal memory or sd and boots the kernel it find there that matches the device |
| 18:12.35 | WyrM | that's how most bootloaders work :) |
| 18:13.23 | WyrM | gen-blob does this, reads /menu.lst and boots the kernel according to the user decision on the boot menu, or /boot/default. |
| 18:13.50 | WyrM | only from FAT, first partition of the mem card. |
| 18:16.43 | WyrM | we can't do device detection on ezx. |
| 18:16.56 | WyrM | or better, we can, if we change flash layout. |
| 18:17.10 | WyrM | there is no such identity partition by default on ezx. |
| 18:18.10 | tmzt | can boot_usb put that into gen-blob when flashing it? |
| 18:18.31 | tmzt | you can format the old kernel partition any way you like right? |
| 18:26.21 | WyrM | yes. |
| 18:26.28 | WyrM | the issue is men power :) |
| 18:26.44 | WyrM | I don't want to maintain different gen-blob versions. |
| 18:27.36 | WyrM | I want one gen-blob that runs on all target phones and on any memory position (so it can be flashed and sent via boot_usb). |
| 18:29.22 | WyrM | that's why im reluctant on depending on flash configuration, etc.. |
| 18:30.51 | WyrM | but I may change my mind on this, maybe if somebody starts implementing a gen-blob that would work with a910 mmc then some flash configuration partition would be worth it. |
| 18:32.37 | WyrM | also, if you start porting gen-blob to motoq, it would be worth some flash partition for device identity. |
| 18:33.02 | WyrM | any plans on this? :D |
| 18:34.04 | tmzt | not yet |
| 18:34.19 | tmzt | I need to make it boot from ram first with haret |
| 18:35.00 | tmzt | I can't see what else to do without my phone mux though, I can keep improving the kernel (master is booting) but without the mux I can't use it as a phone |
| 18:35.21 | WyrM | lol |
| 18:35.30 | WyrM | I can't use openezx as a phone either ;) |
| 18:35.44 | tmzt | it's not working on any ezx phone? |
| 18:35.54 | WyrM | and its more than 3 years since the project started ;) |
| 18:36.00 | WyrM | it works. |
| 18:36.10 | WyrM | but mux working is not enough. |
| 18:36.46 | WyrM | without suspend/resume for example, the battery would not last for more than 6 hours. |
| 18:36.47 | tmzt | I know, the same is true of omap, most pxa's, etc. but linux on smartphone is moving on with commerical support (by some players) and people do expect the phone to work |
| 18:38.40 | tmzt | what else is missing on ezx? will the stuff mickey|zzZZzz is working on help? |
| 18:38.56 | WyrM | a lot of stuff. |
| 18:39.04 | WyrM | suspend/resume is broken on mainline. |
| 18:39.21 | WyrM | something related to the order on which gpio pins are reset on resume. |
| 18:39.27 | WyrM | all pxa's are broken. |
| 18:39.41 | WyrM | we need better usb support for both PCAP and EOC. |
| 18:39.46 | tmzt | yeah, is it not using the internal registers for that? |
| 18:39.57 | tmzt | otg-transceiver? |
| 18:40.02 | WyrM | support to usb host mode, and a proper transceiver driver. |
| 18:40.29 | WyrM | yes, but not as it is (was, the last time I looked at it). |
| 18:41.04 | WyrM | support on most graphics libraries (plain X is working) for the gen2 18bpp lcd. |
| 18:41.55 | tmzt | it would be nice to know where Xorg on arm is going |
| 18:42.37 | WyrM | we also need proper bus sharing with the spi subsystem. |
| 18:43.08 | WyrM | as mmc_spi.c is broken (and even restricts the sharing) if the spi bus is shared. |
| 18:43.28 | WyrM | we also need camera drivers for e680, a1200 and e6. |
| 18:43.49 | WyrM | and some rework on our sound driver for baseband audio. |
| 18:44.16 | tmzt | the pcap asoc? |
| 18:44.21 | WyrM | yes. |
| 18:44.31 | WyrM | we need to support network mode better. |
| 18:44.48 | tmzt | i2s? |
| 18:44.49 | WyrM | so we can for example record calls, or play digital audio while on a call, etc.. |
| 18:44.56 | tmzt | sorry, don't know what that is |
| 18:44.57 | WyrM | its i2s via SPI. |
| 18:45.22 | WyrM | network mode is for sharing the bus, with time slots. |
| 18:45.42 | tmzt | where's the clock? |
| 18:45.50 | tmzt | master/slave? |
| 18:46.14 | WyrM | audio on ezx is strange, SSP2, SSP3, PCAP, Neptune, and BCM20X5 all shares the same lines for digital audio. |
| 18:46.36 | WyrM | by default clock is from PCAP (pxa is slave). |
| 18:46.42 | WyrM | but this can be changed :) |
| 20:31.39 | ghane | lol |
| 20:31.55 | ghane | http://coolepochcountdown.com/ |
| 21:00.02 | ao2 | WyrM, I'll add vibrator support in leds-pcap, is that ok? |
| 21:26.10 | WyrM | ao2, sure. |
| 21:26.13 | WyrM | :) |
| 21:33.37 | *** join/#openezx florian (n=fuchs@f048236199.adsl.alicedsl.de) |
| 21:43.51 | *** join/#openezx florian (n=fuchs@e181243109.adsl.alicedsl.de) |
| 21:48.08 | *** join/#openezx florian (n=fuchs@g227189124.adsl.alicedsl.de) |
| 22:07.40 | *** join/#openezx florian (n=fuchs@f048234121.adsl.alicedsl.de) |
| 22:12.50 | *** join/#openezx flo_lap (n=fuchs@f049015253.adsl.alicedsl.de) |
| 22:14.27 | *** join/#openezx r00t|vaio (n=r00t@1.106.113.82.net.de.o2.com) |
| 22:15.40 | *** join/#openezx fuchs__ (n=fuchs@g228205010.adsl.alicedsl.de) |
| 22:24.59 | *** join/#openezx florian (n=fuchs@g228202201.adsl.alicedsl.de) |
| 22:30.23 | *** join/#openezx tombhadAC (n=tombhad@pD95DD5D9.dip.t-dialin.net) |
| 22:30.55 | *** join/#openezx gmzhuo (i=gmzhuo@125.70.87.165) |
| 22:50.35 | *** join/#openezx florian (n=fuchs@f049002176.adsl.alicedsl.de) |
| 22:59.19 | *** join/#openezx flo_lap (n=fuchs@f049169060.adsl.alicedsl.de) |
| 23:20.57 | *** join/#openezx fuchs__ (n=fuchs@f049011085.adsl.alicedsl.de) |
| 23:29.15 | *** join/#openezx jonwil (n=jonwil@123-243-219-11.tpgi.com.au) |
| 23:29.33 | jonwil | hi |
| 23:30.08 | ghane | hi |
| 23:30.28 | ghane | 1234567827 |
| 23:30.40 | ghane | funny unix-time is less than a minute away |
| 23:30.41 | ghane | :) |
| 23:31.08 | ghane | jonwil: hows it going with your kernel? |
| 23:31.19 | jonwil | nowhere yet |
| 23:31.55 | ghane | ok |
| 23:34.00 | jonwil | currently looking for a way to help the people with the Z6w disable the nasty "you can only use ringtones you buy from the phone carrier" BS :P |
| 23:35.51 | ghane | ok |
| 23:36.02 | ghane | jonwil: Z6w? is that one actually for real? |
| 23:36.09 | jonwil | yes |
| 23:36.14 | jonwil | Z6w with wifi does exist |
| 23:36.24 | jonwil | Some canadians have it |
| 23:36.43 | jonwil | So far the only hacking that we can do is getting at /ezxlocal/* and /data_resource/* |
| 23:38.21 | ghane | i've been looking for a Z6w for a while |
| 23:38.30 | ghane | havent seen it on sale anywhere |
| 23:38.42 | jonwil | Its available in Canada from various canadian carriers |
| 23:38.51 | jonwil | other than that I am unaware of it being available |
| 23:39.02 | ghane | ok |
| 23:39.30 | p3t3r__ | jonwil: any idea how wifi is connected in z6w or what chip it is? is it similar to a910? |
| 23:39.32 | ghane | well its just a chips difference, i can just get the chip and do some soldering, the pcb has a spot |
| 23:39.52 | jonwil | nope, no clue about hardware |
| 23:40.07 | jonwil | as for the software, its a newer version of motomagx ala e8 and u9 |
| 23:40.13 | jonwil | with the webkit browser instead of opera |
| 23:40.42 | ghane | p3t3r__: the only thing i know is that the russians found out that there is an empty space with solder pads on the Z6 where the wifi chip should go |
| 23:41.47 | p3t3r__ | jonwil: ok, cool. when i was about to get a910 i was thinking of z6w as well.. but it took to long (and i won't get it now, doesn't make sense as i just got a G1 recently) |
| 23:41.49 | jonwil | The Z6w is R6635 vs the Z6 which is R60 |
| 23:42.30 | ghane | jonwil: have they released the firmwares for Z6w? |
| 23:43.00 | jonwil | firstly, no, moto havent released oss bits for Z6w |
| 23:43.13 | jonwil | and no-one seems to have published any firmware flash files for Z6w either |
| 23:43.37 | *** join/#openezx florian (n=fuchs@f049005014.adsl.alicedsl.de) |
| 23:43.56 | p3t3r__ | leaves the discussion again, has to go on with macroeconomics ;-) |
| 23:44.17 | jonwil | which is probobly because its only canadian and canadian flash files for motos seem to be harder to get than others (if I had just ONE flash file from ANY canadian phone with the ringtone locking and the matching flash from another version of the same phone, I could find a solition easy) |
| 23:44.32 | ghane | and then the Z6w is almost impossible to get hold of |
| 23:45.45 | jonwil | I expect you could get one in less than an hour if you walked into a Fido or Rogers store :P |
| 23:46.53 | ghane | ? |
| 23:47.07 | jonwil | Z6w is available from Fido and Rogets in canada |
| 23:47.10 | jonwil | Rogers |
| 23:47.13 | jonwil | thats where you get one |
| 23:47.14 | ghane | ok |
| 23:47.20 | ghane | but i dont live in canada |
| 23:50.30 | jonwil | heh :) |
| 23:56.50 | ghane | gotta sleep |
| 23:56.52 | ghane | good nite |
| 23:56.54 | jonwil | cya |
| 23:58.50 | *** join/#openezx johanbr (n=j@blk-7-131-230.eastlink.ca) |