[PRJ][OS]Linux and Android on HTC Diamond "NOT YET WORKING" [Archive]

View Full Version : [PRJ][OS]Linux and Android on HTC Diamond "NOT YET WORKING"

Pages : [1] 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

dcordes

21st June 2008, 03:13 PM

This thread discusses Linux kernel for HTC Diamond and HTC Raphael

The uptodate information on progress with Linux and Android and help can be found at the Raphael Linux wiki page (http://wiki.xda-developers.com/index.php?pagename=RaphaelLinux).
UPDATE
The raphael/diamond linux information in the xda wiki (rsearch, status, etc) is outdated. The core developers now use http://www.htc-linux.org/wiki/index.php?title=Raphael

The latest kernel code for Raphael and Diamond can be found in the htc-msm-2.6.27 branch (http://git.linuxtogo.org/?p=groups/mobile-linux/kernel.git;a=shortlog;h=refs/heads/htc-msm-2.6.27) of the linuxtogo.org git repository.

To clarify some android related questions: We are not trying to make android work. But we are making the linux kernel work well on our phones. Gogle Android (tm) already works.
Posts that do not contribute to progress are not useful because they make it complicated to find relevant information for people who can help.
Please read the wiki with all the FAQs before posting. We are encouraging everybody to update the wiki with their questions, answers and findings.

NEW!!!!

http://htc-linux.org/wiki now available.

_______________________________

original post:

Hi,

Would be nice if somebody could try and test if the kaiser linux kernel boots on their diamond. Here's what you need

http://linuxtogo.org/~lgorris/kaiser-bootkit/

It has a small documentation but here's what you need to do again:

Get haret.exe, default.txt, zImage and angstrom-console on your diamond into the same folder, run haret.exe and tap run.

edit: the link above points to a website with kernel builds for the kaiser (msm7200). since diamond has msm7201A, it would be better to build directly from the android-msm-2.6.25 branch which has the halibut machine type. halibut is also an msm7x00A. there are some differences between the A and non-A msms.

google git is at
git.android.com

checkout the branch android-msm-2.6.25 ( it's most uptodate)

DesktopMan

21st June 2008, 05:02 PM

I have in fact been researching this since yesterday. The vanilla Android is for 7x00A devices, and should work on Diamond without too much fuss. I'm not a Linux developer though, so I'll need some help with that part. (though I am a developer)

The Kaiser boot kit does not work. Once it's done loading the device freezes entirely. Kaiser uses 7200 though, not 7201A, and for all I know it might actually be (partly somewhat) running, but with no frame buffer.

seraph1024

21st June 2008, 05:05 PM

agreed with the freezing part.

DesktopMan

21st June 2008, 08:00 PM

From Haret:

WinCE reports memory size 134217728 (phys=118571008 store=115261440)

Where did the rest 64 MByte go?

someone1234

21st June 2008, 08:28 PM

Ok who's seriously interested in getting this to work?

Whats the progress so far? Lets try to get something packaged and up for diamond.

teddyen66

21st June 2008, 10:52 PM

I am definately going to spend some time trying to gets this to work when I get my Diamond.

DesktopMan

22nd June 2008, 12:15 AM

I am definately going to spend some time trying to gets this to work when I get my Diamond.

I haven't been able to compile the kernel, but it sounds like you'll be able to. Give it a shot and I'll test it.

seraph1024

22nd June 2008, 01:10 AM

I'll test it too.

dcordes

22nd June 2008, 04:45 PM

k, can you also try the vogue builds? http://it029000.massey.ac.nz/vogue/

http://wiki.xda-developers.com/index.php?pagename=DiamondResearch

the shared memory looks similar to the on on kaiser. So it will probably be easy to make communication with the phone work as soon as it works on kaiser. Of course you need to be able to boot first

happyguy82

22nd June 2008, 09:06 PM

Hi guys,

I've been trying to get Android running on my Diamond for a few days to no avail.

I've tried the Kaiser set of files and Vogue set of files. The furthest I've got is to make it display the 'Go Go Go...' message and then it hangs.

Has anyone actually got Android to run on a Diamond before?

Thanks.

DesktopMan

22nd June 2008, 09:33 PM

Vogue does not work, but might be because of HaRET, don't know.

7201A is different from 7200. I would not expect it to work. However, the vanilla kernel is for Diamond type chips, so if anyone could compile a vanilla kernel it's worth testing. HaRET might need a few modifications though.

Riel

22nd June 2008, 11:52 PM

Very interesting!

This Android thing, getting it to run, what would the expectations be? Functionality, speed?

Just very curious, having once installed some linux-thingy on my magician.

toastedlogic

22nd June 2008, 11:59 PM

I don't have a Diamond to test on, but I suspect the dev team for the Advantage could be of help. The build they have is running ok on Advantage, the touchscreen driver even works. No phone yet (last I checked). Maybe it will also run on Diamond?

dcordes

23rd June 2008, 01:00 PM

nopes advantage is a different platform

happyguy82

23rd June 2008, 01:54 PM

I believe it'll just be a matter of time before we get a fully functioning Android build for the HTC Diamond because I can foresee it to be a very popular platform for the Android.

If there's any modifications to the default.txt boot file that would potentially make this work I'll be happy to try it out and let you guys know the results.

nebi

9th July 2008, 11:48 PM

I believe it'll just be a matter of time before we get a fully functioning Android build for the HTC Diamond because I can foresee it to be a very popular platform for the Android.

I hope that too . I can not wait to run Android on my Diamond.

adwinp

10th July 2008, 12:36 PM

smogdragon

10th July 2008, 01:13 PM

Would this link be of any help ?

http://nemustech.blogspot.com/2007/12/android-porting-to-real-target-hw.html

I would like to get involved if someone is already working on it

Riel

10th July 2008, 01:21 PM

Darn that's running smooth ...

... hope we can get some linux running on this hardware.

Ingvarr

10th July 2008, 01:26 PM

I guess it'd be a matter of setting up a cross compile (toolchains - chroot, binutils, gcc, glibc) environment for the ARM, compiling the kernel with the correct drivers and bootstrapping for the ARM.

Am I right? Or did I miss something?

Well, kinda. Where do you intend to get "correct drivers" for Diamond ;)
Its not that easy, since HTC doesnt provide 'em ;)

Hornet331

10th July 2008, 01:26 PM

hmm looking forward to android, i am curious if the diamond speeds up and only win mob 6.x hold it back. :)

Huey85

10th July 2008, 02:01 PM

I'm quite sure the hardware on the Diamond will be able to handle running Android perfectly fine.. It's just WinMo6.x that's slow

smogdragon

10th July 2008, 02:40 PM

If it works on polaris I can't see why it should not on diamond. I've seen .cabs to install it (probably they cannot solve issues on directly booting android yet) and ppl are already able to make and accept calls. It's just a metter of time someone gets to boot into linux on diamond too.

Ingvarr

10th July 2008, 04:14 PM

Maybe thats because they dont actually access hardware like phone module, touchscreen, etc directly, but piggyback on top of Windows Mobile to do it.

In any case, Polaris has a very different hardware from Diamond, at least from driver point.

smogdragon

10th July 2008, 06:33 PM

Maybe thats because they dont actually access hardware like phone module, touchscreen, etc directly, but piggyback on top of Windows Mobile to do it.

Well, I don't know much about the topic but I dont' think so. It's a linux build after all, it surely does not lay on winmo for anything except booting. I guess it's just a bootloader issue.

In any case, Polaris has a very different hardware from Diamond, at least from driver point.

Yeah, yeah sure. but I don't believe they got drivers from hardware vendors so there will probably be someone who'll take care of diamond too ... sooner or later :D

Ingvarr

10th July 2008, 07:39 PM

Yeah, yeah sure. but I don't believe they got drivers from hardware vendors so there will probably be someone who'll take care of diamond too ... sooner or later :D
By this time you'll have a devices that run Android out of the box already ;)

skiwithpete

11th July 2008, 11:10 AM

someone1234

11th July 2008, 03:51 PM

I'm waiting for someone to get android onto a diamond, and post a youtube of it running at a better speed than winmobile, then I'm gonna run out and buy one.

I don't know how to code or I would have run out and got one to help play along with your lads already.

I would just wait. Once WM7 comes out, there'll be a ton of phones out with multitouch.

Once android comes out, there'll be a ton of phones with sleak dimensions like the diamond. It should be possible to run android with a much slower CPU, which should give you a smaller device and amazing battery life too.

LuitvD

15th July 2008, 12:43 AM

Hopefully HTC will make the smart move of releasing Android for the Diamond... It sure looks like they designed the phone to run it, as it has a VGA screen and the buttons on it are exactly what Android expects. Being a D-pad, Send/End, Home, Back, Volume and Power. See the touch-sensors in the D-pad and the stylus-sensor as bonus. :p

Ingvarr

15th July 2008, 09:01 AM

In the near time HTC will be busy with fixing the WM for Diamond, not for making Android for it ;)

dangay

15th July 2008, 02:13 PM

The below is from an article on the register.

http://www.theregister.co.uk/2008/07/08/t_mobile_android/

T-Mobile will be pushing out an Android handset in time for its US 3G launch on 1 October, according to rumours floating around the internet. But that deadline could cost the Android ideal dearly.

Google desperately wants to see an Android-based handset in 2008, having said it will happen, and will lose face if it doesn't. T-Mobile could do with some publicity for its 3G launch, which it's claiming will be national but will actually only cover major cities. The operator is also further along than most toward launching an Android handset with HTC, so the rumour is easy to credit.

Getting a handset out this year will be a challenge and Android has suffered from delays, but there are ways in which T-Mobile and Google could compromise to get that headline launch in time for October.

The first task is to dump all the network-branded applications that operators like to install. Rumour has it that porting these applications is proving more difficult than expected and removing them should help the schedule considerably. That might mean less revenue on value-added services for T-Mobile, but they can always be offered as a download later; this might even emphasise the flexibility of the Android platform. Assuming that the device launched by T-Mobile will allow downloading of applications, anyway.

Nothing in the Android licence prevents an operator from launching a locked-down handset unable to download new applications beyond Java Midlets or similar. None of the LiMo handsets so far launched are able to download native applications, despite all being based on that "open" version of Linux. Releasing an Android handset that can't download applications might go against the spirit of the Open Handset Alliance, but it would give T-Mobile that "first Android handset" headline, and most punters wouldn't care anyway.

T-Mobile and Google will move heaven and earth to get an Android-based handset into punters' hands in time for the 3G launch, even if that means pushing the Android brand rather than the dream. ®

timn2087

15th July 2008, 02:44 PM

In the near time HTC will be busy with fixing the WM for Diamond, not for making Android for it ;)

wishful thinking :p

Dosman

21st July 2008, 11:48 PM

Hi
First, sorry for my english XD

Ok. I saw those Android Movies on Youtube and i love this thing. Now, does anybody has a working version of android that works on the diamond or other 7200A devices?

skiwithpete

31st July 2008, 01:19 AM

I can't believe this discussion isn't progressing, surely someone can get something to boot on the diamond. Surely a linux kernel will run...

Right?

CutePinkSox

31st July 2008, 02:22 AM

I can't believe this discussion isn't progressing, surely someone can get something to boot on the diamond. Surely a linux kernel will run...

Right?

Thats because everyone just keeps going off topic - grrr guys dont post useless info and make comments - I have been reading this for some time but havn't commented as I don't have anything useful to say to contribute. Please do the same!

skiwithpete

3rd August 2008, 10:01 PM

How can I help?

happyguy82

3rd August 2008, 11:11 PM

Simple.... just get it working please :) HaRET to book Android on a Diamond :)

Dosman

4th August 2008, 02:58 PM

How can I help?

Get it to work on the Diamond :D

Riel

5th August 2008, 02:55 PM

http://handhelds.org/moin/moin.cgi/HpIpaqH1940StepByStep

Is this not possible on Android? (vanilla kernel)

marbalon

6th August 2008, 10:58 AM

Hi all Diamond owners,
Here is a package contain lastest haret, default.txt, and zImage build from android-git repository for msm7201A (halibut board with disabled some function, just for test). Can someone try it ? What machine type find haret ?

Ronnymes

6th August 2008, 11:35 AM

Riel

6th August 2008, 11:57 AM

Errors:

Unexpected input =1439
Expected numeric
Undefined MTYPE

Riel

6th August 2008, 12:00 PM

Remove the = from default.txt

It asks for Start of ram (RAMADDR)

Riel

6th August 2008, 12:06 PM

Tried HaRet 0.5.1 and still freezes at green 'booting linux' bar.
The newer HaRet I will try when I know how to enter the RAMADDR command for start of ram in Default.txt for Diamond.

Kisja

6th August 2008, 12:11 PM

Found machine Generic ARM v6

Remove the = from default.txt
It asks for Start of ram (RAMADDR)

Still fails to start (RAMADDR Error)
Should it look like this?
set MTYPE 1439
set KERNEL zImage
boot

marbalon

6th August 2008, 12:18 PM

Try to set RAMMSIZE and RAMADDR, these are values for Kaiser, so try to setup correct for you device, or try this one.

default.txt:
set MTYPE 1439
set KERNEL zImage
set RAMSIZE 0x7800000
set RAMADDR 0x10000000
bootlinux

Riel

6th August 2008, 12:22 PM

Freeze at booting linux ;)

marbalon

6th August 2008, 12:38 PM

I've just raplace kernel image in attachment above. Try this one.
If not booting try to add empty file earlyharetlog.txt. It enable logging for kaiser and attach it to next reply. I think haret should found generic msm machine not ARMv6. So maybe this is a problem.

Kisja

6th August 2008, 01:08 PM

Same problem here, freeze on booting Linux

marbalon

6th August 2008, 01:13 PM

Like I said. We need patch haret sources. Haret should detect msm machine. But I dont have cegcc instaled on my sysytem. Give me a little time to do that and maybe I can fix that.

Kisja

6th August 2008, 01:18 PM

Good news! Looking forward ...

Btw, its only the kernel. What about android itself? What to do when the kernel boots?

Riel

6th August 2008, 01:22 PM

Hangs.
Log:

===== HaRET pre-0.5.2-20080726_131657 =====
Setting KMode to true.
Old KMode was 0
Finished initializing output
Loading dynamically bound functions
Function 'AllocPhysMem' in library 'coredll' at 03F63EE8
Function 'FreePhysMem' in library 'coredll' at 03F63F54
Function '?GXOpenDisplay@@YAHPAUHWND__@@K@Z' in library 'gx' at 02A2C9EC
Function '?GXCloseDisplay@@YAHXZ' in library 'gx' at 02A2CDD8
Function '?GXBeginDraw@@YAPAXXZ' in library 'gx' at 02A2C2A8
Function '?GXEndDraw@@YAHXZ' in library 'gx' at 02A2C36C
Function 'LoadLibraryExW' in library 'coredll' at 03F64090
Function 'NLedSetDevice' in library 'coredll' at 03F8B590
Function 'GetSystemPowerStatusEx2' in library 'coredll' at 03F6AA14
Function 'SleepTillTick' in library 'coredll' at 03F63FB0
Function 'CreateToolhelp32Snapshot' in library 'toolhelp' at 02FC705C
Function 'CloseToolhelp32Snapshot' in library 'toolhelp' at 02FC707C
Function 'Process32First' in library 'toolhelp' at 02FC7140
Function 'Process32Next' in library 'toolhelp' at 02FC71AC
Function 'Module32First' in library 'toolhelp' at 02FC7340
Function 'Module32Next' in library 'toolhelp' at 02FC73B0
Function 'Heap32ListFirst' in library 'toolhelp' at 02FC744C
Function 'Heap32ListNext' in library 'toolhelp' at 02FC74D4
Function 'Heap32First' in library 'toolhelp' at 02FC756C
Function 'Heap32Next' in library 'toolhelp' at 02FC7620
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Detecting current machine
Trying to detect machine (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
Wince reports processor: core=MSM7201A-528MHz name=???? cat= vend=QUALCOMM
Looking at machine Alpine
LOTS OF LOOKING ATS NO DIFFERENCE - NEED TO CUT THE POST
Looking at machine Generic ARM v6
Looking at arch Generic Intel PXA27x
Looking at arch Generic Intel PXA
Looking at arch Generic Intel StrongArm
Looking at arch Generic TI OMAP
Looking at arch Generic TI OMAP15xx
Looking at arch Generic Samsung s3c24xx
Looking at arch Generic MSM7500
Looking at arch Generic Atlas
Looking at arch Generic ARM 920t
Looking at arch Generic ARM 926
Looking at arch Generic ARM v6
Detecting ram size
WinCE reports memory size 134217728 (phys=116412416 store=89456640)
Mapping mmu table
Build L1 reverse map
Found 248 uncached and 157 cached L1 mappings (ignored 108).
Not registering command IGPIO
Not registering command WG|PIO
Not registering command GPLR

Not registering command GPDR
Not registering command GAFR
Not registering command GPIO
Not registering command GPIOST
Registering command LOADLIBRARYEX
Registering command NLEDSET
Not registering command TRACE
Not registering command TRACEMASK
Not registering command TRACE2
Not registering command TRACETYPE
Not registering command TRACE2TYPE
Not registering command TRACEFORWATCH
Not registering command INSN
Not registering command INSNREENABLE
Not registering command INSNREG1
Not registering command INSNREG2
Not registering command INSN2
Not registering command INSN2REENABLE
Not registering command INSN2REG1
Not registering command INSN2REG2
Registering command KILL
Registering command PS
Registering command LSMOD
Registering command ADDR2MOD
Not registering command AC97
Not registering command ATIDBG
Initializing for machine 'Generic ARM v6'
Welcome, this is HaRET pre-0.5.2-20080726_131657 running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Generic ARM v6/generic (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
CPU is ARM ARM arch 6 stepping 2 running in system mode
Enter 'HELP' for a short command summary.

Running WSAStartup
Starting gui
In initdialog
Found machine Generic ARM v6
executing startup.txt
HaRET(1)# set MTYPE 1439
HaRET(2)# set KERNEL zImage
HaRET(3)# set RAMSIZE 0x7800000
HaRET(4)# set RAMADDR 0x10000000
HaRET(5)# bootlinux
boot KERNEL=zImage INITRD=
Opening file zImage
boot params: RAMADDR=10000000 RAMSIZE=07800000 MTYPE=1439 CMDLINE='root=/dev/ram0 ro console=tty0'
Boot FB feedback: 1
Built virtual to physical page mapping
Allocated 286 pages (tags=59100000/14514000 kernel=59101000/14513000 initrd=5921A000/143c4000 index=5921A000/143c4000)
Built kernel tags area
Built page index
Video buffer at 5126A000 sx=480 sy=640 mx=120 my=106
Video Phys FB=1686a000 Fonts=143c6064
preload=2260@5921D000/143c7000 sj=5921D000 stack=5921B000/143c5000 data=5921C000/143c6000 exec=143c7128
Reading 1150332 bytes...
Read complete
Launching to physical address 143c7010
Trampoline setup (tram=136@00024D14/28024d14/10b42d14)
MMU setup: mmu=A03F0000/103f0000
Go Go Go...

===== HaRET pre-0.5.2-20080726_131657 =====
Setting KMode to true.
Old KMode was 0
Finished initializing output
Loading dynamically bound functions
Function 'AllocPhysMem' in library 'coredll' at 03F63EE8
Function 'FreePhysMem' in library 'coredll' at 03F63F54
Function '?GXOpenDisplay@@YAHPAUHWND__@@K@Z' in library 'gx' at 02A2C9EC
Function '?GXCloseDisplay@@YAHXZ' in library 'gx' at 02A2CDD8
Function '?GXBeginDraw@@YAPAXXZ' in library 'gx' at 02A2C2A8
Function '?GXEndDraw@@YAHXZ' in library 'gx' at 02A2C36C
Function 'LoadLibraryExW' in library 'coredll' at 03F64090
Function 'NLedSetDevice' in library 'coredll' at 03F8B590
Function 'GetSystemPowerStatusEx2' in library 'coredll' at 03F6AA14
Function 'SleepTillTick' in library 'coredll' at 03F63FB0
Function 'CreateToolhelp32Snapshot' in library 'toolhelp' at 02FC705C
Function 'CloseToolhelp32Snapshot' in library 'toolhelp' at 02FC707C
Function 'Process32First' in library 'toolhelp' at 02FC7140
Function 'Process32Next' in library 'toolhelp' at 02FC71AC
Function 'Module32First' in library 'toolhelp' at 02FC7340
Function 'Module32Next' in library 'toolhelp' at 02FC73B0
Function 'Heap32ListFirst' in library 'toolhelp' at 02FC744C
Function 'Heap32ListNext' in library 'toolhelp' at 02FC74D4
Function 'Heap32First' in library 'toolhelp' at 02FC756C
Function 'Heap32Next' in library 'toolhelp' at 02FC7620
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Detecting current machine
Trying to detect machine (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
Wince reports processor: core=MSM7201A-528MHz name=???? cat= vend=QUALCOMM
Looking at machine Alpine
Looking at machine Apache
Looking at machine AximX50
SAME HERE
Looking at machine Generic ARM v6
Looking at arch Generic Intel PXA27x
Looking at arch Generic Intel PXA
Looking at arch Generic Intel StrongArm
Looking at arch Generic TI OMAP
Looking at arch Generic TI OMAP15xx
Looking at arch Generic Samsung s3c24xx
Looking at arch Generic MSM7500
Looking at arch Generic Atlas
Looking at arch Generic ARM 920t
Looking at arch Generic ARM 926
Looking at arch Generic ARM v6
Detecting ram size
WinCE reports memory size 134217728 (phys=116412416 store=89456640)
Mapping mmu table
Build L1 reverse map
Found 248 uncached and 157 cached L1 mappings (ignored 108).
Not registering command IGPIO
Not registering command WG|PIO
Not registering command GPLR
Not registering command GPDR
Not registering command GAFR
Not registering command GPIO
Not registering command GPIOST
Registering command LOADLIBRARYEX
Registering command NLEDSET
Not registering command TRACE
Not registering command TRACEMASK
Not registering command TRACE2
Not registering command TRACETYPE
Not registering command TRACE2TYPE
Not registering command TRACEFORWATCH
Not registering command INSN
Not registering command INSNREENABLE
Not registering command INSNREG1
Not registering command INSNREG2
Not registering command INSN2
Not registering command INSN2REENABLE
Not registering command INSN2REG1
Not registering command INSN2REG2
Registering command KILL
Registering command PS
Registering command LSMOD
Registering command ADDR2MOD
Not registering command AC97
Not registering command ATIDBG
Initializing for machine 'Generic ARM v6'
Welcome, this is HaRET pre-0.5.2-20080726_131657 running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Generic ARM v6/generic (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
CPU is ARM ARM arch 6 stepping 2 running in system mode
Enter 'HELP' for a short command summary.

Running WSAStartup
Starting gui
In initdialog
Found machine Generic ARM v6
executing startup.txt
Calling WSACleanup
Shutting down

Riel

6th August 2008, 01:24 PM

I think ROOT = /dev/

needs to be where Internal Storage is at.

CMDLINE='root=/dev/ram0 ro console=tty0'

I read that somewhere on HaRet site?

teddyen66

6th August 2008, 01:25 PM

marbalon

6th August 2008, 01:28 PM

Mine shows: Found machine Generic MSM7500

Try to run it and attach logs like others.

teddyen66

6th August 2008, 01:35 PM

Try to run it and attach logs like others.

It freezes on boot.
Here is the log file.

marbalon

6th August 2008, 01:38 PM

Put empty file "earlyharetlog.txt" in the same directory where haret.

Riel

6th August 2008, 01:45 PM

Mine shows: Found machine Generic MSM7500

Version 0.5.1 says that indeed.

pof

6th August 2008, 01:48 PM

Download newest haret from here:
http://handhelds.org/~koconnor/haret/

Latest one is 2008-07-26, version 0.5.1 is from 2007.

marbalon

6th August 2008, 01:58 PM

Haret in my package is http://handhelds.org/~koconnor/haret/haret-20080726f.exe so it is newest.

Riel

6th August 2008, 02:17 PM

But teddyenn isn't ;)

teddyen66

6th August 2008, 02:27 PM

But teddyenn isn't ;)

ups, sorry :-)

Yeah now mine says the same: Found machine Generic ARM v6

teddyen66

6th August 2008, 02:48 PM

I just tried the Vouge version and it got further.
It stops at: Go Go Go
And partly blacks screen.

ihatethis

7th August 2008, 12:56 AM

Ok, here is my log with HaRET pre-0.5.2-20080726f

It detects Generic ARM v6, freezes on booting linux and log shows Go Go Go...

Thanks a lot to marbalon for his help with that :D

Araldwenn

7th August 2008, 01:26 PM

I'm very interested in this post, as it would be syply great to make Android works on Diamond...

Just one question (i'm not an expert in PDA...)
As Android, if i'v understood erverything, is based on a linux kernel, is it the same as on home computer ? I mean, does the file system has to be changed in order to make it work ?

espenfjo

7th August 2008, 02:01 PM

Ok, a few basic questions from me.
I gather that you need hardSPL for this to work?
And it is safe to flash these ROMs, i can just flash again with one of the cooked ROMs on this site to get my device "normal" again?

Kisja

7th August 2008, 02:24 PM

Hi,
As I understand this, Android/Linux IS NOT A ROM. You copy the Rootfs,Kernel and Haret to the Storage and run it. No flashing at all. Haret unloads Windows and starts Linux. The internal storage is mounted under Linux as a normal partition. After a SoftReset it returns to WM automatically. So no flashing and no formatting with different filesystem.

correct me if I'm wrong. On REAL Android devices (shipped with android and not with windows) it will be something like a ROM.

ihatethis

7th August 2008, 11:50 PM

Good night, something new about this?

I would like to help patching haret for running it on Diamond

I have cegcc installed on my linux and compiled haret, tried it and it runs well, so marbalon, I can help you if you want compiling haret and trying the changes you want to do. By the way, I'm programmer but never code nothing for handled devices, but maybe I can help you patching it if you guide me a little where to start, I have been viewing code under src/mach, I will read something about haret and how it runs if I have time tomorrow

PD. sorry for my poor english!

ihatethis

8th August 2008, 12:56 AM

Hi again, which machine it have to detect? I did a small hack on haret to force machine without autodetection

There is no Generic MSM7200 machine so I tried with Generic MSM7500 machine and Kaiser (MSM7200) but I got the same result as with Generic ARM v6

log shows this machines and architectures:

Looking at machine Alpine
Looking at machine Apache
Looking at machine AximX50
Looking at machine AximX5
Looking at machine Beetles
Looking at machine Blueangel
Looking at machine Himalaya
Looking at machine Magician
Looking at machine Universal
Looking at machine H1910
Looking at machine H1940
Looking at machine H2200
Looking at machine H3600b
Looking at machine H3700
Looking at machine H3800
Looking at machine H3900
Looking at machine H4000
Looking at machine H4300
Looking at machine H5000
Looking at machine H6340
Looking at machine HX2000
Looking at machine HX4700
Looking at machine Sable
Looking at machine Wizard
Looking at machine Hermes
Looking at machine Trinity
Looking at machine Athena
Looking at machine G500
Looking at machine Artemis
Looking at machine Herald
Looking at machine Prophet
Looking at machine RX3000
Looking at machine Treo700wx
Looking at machine e310
Looking at machine e740
Looking at machine Acer_n30
Looking at machine Mio_P550
Looking at machine Kaiser
Looking at machine Loox5xx
Looking at machine Loox400
Looking at machine MioA701
Looking at machine Jornada820
Looking at machine H3100
Looking at machine H3600a
Looking at machine Tornado
Looking at machine Libra
Looking at machine Generic Intel PXA27x
Looking at machine Generic Intel PXA
Looking at machine Generic Intel StrongArm
Looking at machine Generic TI OMAP
Looking at machine Generic TI OMAP15xx
Looking at machine Generic Samsung s3c24xx
Looking at machine Generic MSM7500
Looking at machine Generic Atlas
Looking at machine Generic ARM 920t
Looking at machine Generic ARM 926
Looking at machine Generic ARM v6
Looking at arch Generic Intel PXA27x
Looking at arch Generic Intel PXA
Looking at arch Generic Intel StrongArm
Looking at arch Generic TI OMAP
Looking at arch Generic TI OMAP15xx
Looking at arch Generic Samsung s3c24xx
Looking at arch Generic MSM7500
Looking at arch Generic Atlas
Looking at arch Generic ARM 920t
Looking at arch Generic ARM 926
Looking at arch Generic ARM v6

sensorama

8th August 2008, 10:10 AM

This link below should supposedly contain some drivers for MSM7201...
I dont know what to make of it but i hope you guys can sort some thing out.

http://git.android.com/?p=linux-msm.git;a=tree;f=arch/arm/mach-msm

espenfjo

8th August 2008, 11:27 AM

Ok, since this will preserve WM on the main storage one should simply add hell loads of debuging information in the kernel and make it write to a file (ok.. I know that writing to files directly from the kernel isnt that easy, but somehow it should be doable). It might just be the framebuffer not working properly.

ihatethis

8th August 2008, 03:05 PM

Hi, maybe yes, I had been reading other threads about android & kaiser & polaris and some people has the same problem and it was the framebuffer driver.

I'm at work now, this weekend i will try to compile the latest snapshot for android msm kernel and will make more test

Cheers

ihatethis

8th August 2008, 08:19 PM

I'm downloading now kernel from git.android.com, I really hope we'll be able to run it in diamond soon, at google io 08 in May google presented the latest version of android under device with MSM7201A and I think (and hope) they really release all source code of kernel and use the same kernel for it

I would like to help with this task but I need help too! I'm totally noob developing for handled devices but I like to collaborate with the community

sensorama

8th August 2008, 08:45 PM

Well, report any success or failures.

Good luck.

agjell

10th August 2008, 03:05 PM

Atse

10th August 2008, 04:35 PM

Any progress?

BTW I read in this article that HTC will be shipping the Diamond with Android by Q4 this year:
http://phonereport.info/2008/06/15/htc-google-android-q4-windows-mobile-7-q1-2009/

Old news maybe?

Edit: Found a newer article that confirms it:
http://www.digitimes.com/news/a20080801PD216.html

Neither of those articles confirm what you are saying. They both confirm that HTC is going to ship a device with Android, but nothing is said about this device being the Diamond. I would like that though...

agjell

10th August 2008, 04:53 PM

Neither of those articles confirm what you are saying. They both confirm that HTC is going to ship a device with Android, but nothing is said about this device being the Diamond. I would like that though...

Damn, I must have been too excited... Sorry to be off topic then, not intended.

sensorama

12th August 2008, 05:12 PM

Any progress on this subject?

espenfjo

13th August 2008, 12:43 AM

I have now tried with various kernels (self compiled and from the kaiser), and i get the same results as everyone else here..
For me it seems like Haret isnt able to boot the linux at all, so maybe its the memory settings that is wrong? RAMSIZE RAMADDR?
Anyways, ill start looking into the kernel itself to see if i can see anything obvious there (cant say that im neither a kernel expert, nor any good in C).

kState

13th August 2008, 12:43 PM

hi,
i too tested kernel from kaiser, from this thread and selfcompiled kernel from htc/kaiser git.. and result is same as have many people in this thread...
log end with Go Go Go... (btw: GoGoGo="now i switch to linux")
is here anyone who know that value in haret conf is right for diamond ? model, ram address etc... ?
kState

espenfjo

14th August 2008, 01:33 AM

Ok..
After doing some debugging i can confirm that Haret is not the problem (although it have some problems with the FB).
Haret goes all the way to this point

// Disable MMU and launch linux.
mmu_trampoline(physAddrTram, virtAddrMmu, bm->physExec, Mach->flushCache);

void mmu_trampoline(uint32 phys, uint8 *mmu, uint32 code, void (*)(void))
@ In: r0 = Physical address of this function
@ r1 = Virtual address of MMU in non-cached ram.
@ r2 = Physical address of code to jump to
@ r3 = cpu flush cache function

The odd thing is that i cant find the value in r2(uint32 code) any place in the logs.

Anyways..

Ill start debugging the linux kernel now, add debug functions to the early boot there, see if i can get any info that the kernel is loading at all (im no developer, so the only thing i can do is this work :)

EDIT:
Ill attach my log aswell.. Should be funny for most of you to read my debuging info :) Guess you have to read the code to understand most of them ;)
And maybe someone can point out something i dont understand.

===== HaRET pre-0.5.2-20080814_012449 =====
Setting KMode to true.
Old KMode was 0
Finished initializing output
Loading dynamically bound functions
Function 'AllocPhysMem' in library 'coredll' at 03F62EE8
Function 'FreePhysMem' in library 'coredll' at 03F62F54
Function '?GXOpenDisplay@@YAHPAUHWND__@@K@Z' in library 'gx' at 029929EC
Function '?GXCloseDisplay@@YAHXZ' in library 'gx' at 02992DD8
Function '?GXBeginDraw@@YAPAXXZ' in library 'gx' at 029922A8
Function '?GXEndDraw@@YAHXZ' in library 'gx' at 0299236C
Function 'LoadLibraryExW' in library 'coredll' at 03F63090
Function 'NLedSetDevice' in library 'coredll' at 03F8A590
Function 'GetSystemPowerStatusEx2' in library 'coredll' at 03F69A14
Function 'SleepTillTick' in library 'coredll' at 03F62FB0
Function 'CreateToolhelp32Snapshot' in library 'toolhelp' at 02FC605C
Function 'CloseToolhelp32Snapshot' in library 'toolhelp' at 02FC607C
Function 'Process32First' in library 'toolhelp' at 02FC6140
Function 'Process32Next' in library 'toolhelp' at 02FC61AC
Function 'Module32First' in library 'toolhelp' at 02FC6340
Function 'Module32Next' in library 'toolhelp' at 02FC63B0
Function 'Heap32ListFirst' in library 'toolhelp' at 02FC644C
Function 'Heap32ListNext' in library 'toolhelp' at 02FC64D4
Function 'Heap32First' in library 'toolhelp' at 02FC656C
Function 'Heap32Next' in library 'toolhelp' at 02FC6620
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Unable to load library 'ace_ddi'
Detecting current machine
Trying to detect machine (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
Wince reports processor: core=MSM7201A-528MHz name=?^?? cat= vend=QUALCOMM
Looking at machine Alpine
Looking at machine Apache
Looking at machine AximX50
Looking at machine AximX5
Looking at machine Beetles
Looking at machine Blueangel
Looking at machine Himalaya
Looking at machine Magician
Looking at machine Universal
Looking at machine H1910
Looking at machine H1940
Looking at machine H2200
Looking at machine H3600b
Looking at machine H3700
Looking at machine H3800
Looking at machine H3900
Looking at machine H4000
Looking at machine H4300
Looking at machine H5000
Looking at machine H6340
Looking at machine HX2000
Looking at machine HX4700
Looking at machine Sable
Looking at machine Wizard
Looking at machine Hermes
Looking at machine Trinity
Looking at machine Athena
Looking at machine G500
Looking at machine Artemis
Looking at machine Herald
Looking at machine Prophet
Looking at machine RX3000
Looking at machine Treo700wx
Looking at machine e310
Looking at machine e740
Looking at machine Acer_n30
Looking at machine Mio_P550
Looking at machine Kaiser
Looking at machine Loox5xx
Looking at machine Loox400
Looking at machine MioA701
Looking at machine Wallaby
Looking at machine Jornada820
Looking at machine H3100
Looking at machine H3600a
Looking at machine Tornado
Looking at machine Libra
Looking at machine Generic Intel PXA27x
Looking at machine Generic Intel PXA
Looking at machine Generic Intel StrongArm
Looking at machine Generic TI OMAP
Looking at machine Generic TI OMAP15xx
Looking at machine Generic Samsung s3c24xx
Looking at machine Generic MSM7500
Looking at machine Generic Atlas
Looking at machine Generic ARM 920t
Looking at machine Generic ARM 926
Looking at machine Generic ARM v6
Looking at arch Generic Intel PXA27x
Looking at arch Generic Intel PXA
Looking at arch Generic Intel StrongArm
Looking at arch Generic TI OMAP
Looking at arch Generic TI OMAP15xx
Looking at arch Generic Samsung s3c24xx
Looking at arch Generic MSM7500
Looking at arch Generic Atlas
Looking at arch Generic ARM 920t
Looking at arch Generic ARM 926
Looking at arch Generic ARM v6
Detecting ram size
WinCE reports memory size 134217728 (phys=120619008 store=119324672)
Mapping mmu table
Build L1 reverse map
Found 248 uncached and 157 cached L1 mappings (ignored 108).
Not registering command IGPIO
Not registering command WG|PIO
Not registering command GPLR
Not registering command GPDR
Not registering command GAFR
Not registering command GPIO
Not registering command GPIOST
Registering command LOADLIBRARYEX
Registering command NLEDSET
Not registering command TRACE
Not registering command TRACEMASK
Not registering command TRACE2
Not registering command TRACETYPE
Not registering command TRACE2TYPE
Not registering command TRACEFORWATCH
Not registering command INSN
Not registering command INSNREENABLE
Not registering command INSNREG1
Not registering command INSNREG2
Not registering command INSN2
Not registering command INSN2REENABLE
Not registering command INSN2REG1
Not registering command INSN2REG2
Registering command KILL
Registering command PS
Registering command LSMOD
Registering command ADDR2MOD
Not registering command AC97
Not registering command ATIDBG
Initializing for machine 'Generic ARM v6'
Welcome, this is HaRET pre-0.5.2-20080814_012449 running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Generic ARM v6/generic (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
CPU is ARM ARM arch 6 stepping 2 running in system mode

Enter 'HELP' for a short command summary.
Running WSAStartup
Starting gui
In initdialog
Found machine Generic ARM v6
executing startup.txt
Setting KMode to true.
Old KMode was 0
Incoming connection from 10.0.0.1:37725
Welcome, this is HaRET pre-0.5.2-20080814_012449 running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Generic ARM v6/generic (Plat='PocketPC' OEM='HTC Touch Diamond P3700')
CPU is ARM ARM arch 6 stepping 2 running in system mode
Enter 'HELP' for a short command summary.

HaRET(1)# runscript default.txt
HaRET(1)# set MTYPE 1805
HaRET(3)# set KERNEL zImage.bin
HaRET(5)# set RAMADDR 0x10000000
HaRET(6)# Set initrd android-espen.bin
HaRET(7)# Set fbduringboot 1
HaRET(2)# bootlinux
boot KERNEL=zImage.bin INITRD=android-espen.bin
Opening file zImage.bin
Opening file android-espen.bin
boot params: RAMADDR=10000000 RAMSIZE=08000000 MTYPE=1805 CMDLINE='root=/dev/ram0 ro console=tty0'
Boot FB feedback: 1
Built virtual to physical page mapping
Allocated 5387 pages (tags=5C000000/14bdb000 kernel=5C001000/14bdc000 initrd=5C12A000/1497e000 index=5D502000/135a4000)
Built kernel tags area
Built page index
Video buffer at 5116A000 sx=480 sy=640 mx=120 my=106
Video Phys FB=1686a000 Fonts=1359d064
preload=2260@5D50A000/1359c000 sj=5D50A000 stack=5D508000/1359e000 data=5D509000/1359d000 exec=1359c128
Reading 1213396 bytes...
Read complete
Reading 20805304 bytes...
Read complete
Launching to physical address 1359c010
Launching kernel
Setting up the trampoline
Trampoline setup (tram=136@00024D84/20024d84/10953d84)
MMU setup: mmu=A03F0000/103f0000
Go Go Go...
Did the stuff //Disabling interrupts
Called the per arch things? //Dunno what this does
324648976 //r2 uint32 code kernel address

espenfjo

14th August 2008, 02:03 AM

Oh.. ok..
Sorry for double posting, but this is new information, and the previous post was long enough already.

It seems like there is a section after the mmu_trampoline aswell, the preloader.

The FB shows that it jumps to the PSR-part (Program Status Register, the thing that holds the flag information about programs and instructions) in the preloader.

However, it doesnt seem like it will continue after this(the fb doesnt show anything else atleast), and only come this far only one out of thousand tries (well.. haven tried that many times yet tho..).
Tho.. The logging often surpass the FB information.

So this seems the furthest i can go without having any means to stabilize the FB output, or run an external debugger like gdb on haret.

Global_Inferno

14th August 2008, 09:32 AM

Has anyone had a look at Ubuntu Mobile Edition??

http://www.ubuntu.com/products/mobile

skiwithpete

14th August 2008, 10:46 AM

Has anyone had a look at Ubuntu Mobile Edition??

http://www.ubuntu.com/products/mobile

Ubuntu MID is aimed more at computer/netbook devices like the EEEpc, MSI Wind, etc.

On the Ubuntu MID site it confirms that a MID device has:

MIDs typically have the following features and attributes:

* Small size/form factor
* 4 to 7 inch touch screen
* Physical and/or virtual keyboard
* Wi-Fi, 3G, Bluetooth, GPS, WiMAX
* 2GB to 8GB Flash or disk storage, 256MB+ memory/512MB+ recommended
* OpenGL 3D
* USB, camera, head phone jack, speakers, microphone

Note that while it says 3G this is implied for data transfer and it doesn't say Phone functionality. For mobile phone based Linux, the bleeding edge is really Android.

Global_Inferno

14th August 2008, 10:47 AM

Oh ok... Just was curious as I dont know much about the whole Linux Mobile OS development at the moment, in what sort of state is Android at the moment then? Usable?

sensorama

14th August 2008, 10:50 AM

Oh ok... Just was curious as I dont know much about the whole Linux Mobile OS development at the moment, in what sort of state is Android at the moment then? Usable?

Have you even read this thread? :confused:

sama

14th August 2008, 11:07 AM

I hate it when ppl do that, they see the title of the thread and are too lazy to read the whole thing. They expect someone to just summarise the entire thread for them... and worst of all, is that this realisation will come as a shock to them and they'll get offended by it, as that's not what the meant... grrrr.....

not my best morning today either

Global_Inferno

14th August 2008, 11:11 AM

Um I have read the thread, but as it goes off topic about every 1 in 3 posts, it's kinda hard to see where people have got too... From the looks of it, noone has even managed to get Android to start up, let alone be usable? That sound about right?

Next question, what can people do to help?

sensorama

14th August 2008, 11:23 AM

Global_Inferno

14th August 2008, 11:40 AM

Spot on as you can see. :D

Im asking if I can help and you flame me.... right then.. **** the lot of you then.

skiwithpete

14th August 2008, 01:40 PM

Global>

Here's some stuff you can help with:

1)
http://www.gnome.org/mobile/
What is it? Is it just another MID interface?
What does it need to run on top of? Android or not?
Can you find instances of mobile devices running it?

2)
Compile some youtube videos of other phones running Android, to add some motivation.

espenfjo

14th August 2008, 01:49 PM

What about keeping this thread as on topic as possible?

At the moment my problem is still harets unstable FB. The FB address should be right, but it might be a size problem or something?

Riel

14th August 2008, 02:41 PM

Im asking if I can help and you flame me.... right then.. **** the lot of you then.

Just was curious as I dont know much about the whole Linux Mobile OS

Not much about help. Just lazy question IMHO.

That Linux-booting thing must just be a small glitch. I just feel not very much people are trying it :) I really do not like WM. I am still waiting for linux to go onto or mobiles.

I have bought a Motorole Z6 just because it's running linux :)

Global_Inferno

14th August 2008, 02:42 PM

1) Looks to me as though it's an entirely new branch of GNOME development, based off the foundations of X, GLib and the D-Bus. Using normal GTK+ UI toolsets and offering API's in C, C++ and Python. (Although there really, isn't a lot of information on it available at the moment!)

So what is it? An completely new Mobile OS, based on the existing GNOME Desktop and Embedded builds, with API support for developers.

Licence? It's distributed under the terms of the LGPL which, in there own words, "allows for royalty-free use in proprietary software products, but encourages contribution of platform code changes back to the community."

What does it need to run on top of? Nothing, as far as I can see.. Other than a Linux Kernel.

Android or not? Not.

Any devices running it? Not currently, all the information relating to the devices which will be supported are all shown running their own proprietry software currently.

Global_Inferno

14th August 2008, 02:48 PM

Demo:
http://www.youtube.com/watch?v=1FJHYqE0RDg

http://www.youtube.com/watch?v=arXolJrLVEg
http://www.youtube.com/watch?v=4PRfVKzuUJ4
http://www.youtube.com/watch?v=DfruA4RzzpQ
http://www.youtube.com/watch?v=iBHx5jHsqiU

Development:

http://www.youtube.com/watch?v=QBGfUs9mQYY
http://www.youtube.com/watch?v=fL6gSd4ugSI
http://www.youtube.com/watch?v=MPukbH6D-lY

http://www.youtube.com/watch?v=I6ObTqIiYfE

espenfjo

14th August 2008, 04:45 PM

I have a question.

The Diamond utilizes an ati chipset, or an ati portion of the msm7201, right?

the ati driver is the ace_ddi?
But.. I cant seem to find an ace_ddi on my diamond, and haret wont utilize the ati fb portion either..
So.. Are we just missing a driver? or is it something else?

EDIT: Ok.. found out that some ati functions are in ahi2dati.dll, edited haret to use that instead of ace_ddi, hacked som more on it to not include missing functions in the ahi2dati, got this info;

HaRET(2)# atidbg
ATI asic* =0x50F00000 / 0x02b8f000 ret=0

ATI ChipId = 0x51301002
ATI Revision = 0x00000000
ATI RevisionId = 0x00000000
ATI BusInterfaceMode = 0x00000002
ATI InternalMemSize = 0x00040000
ATI ExternalMemSize = 0x00000000
ATI Caps1 = 0x00000031
ATI Caps2 = 0x00000001
ATI Caps3 = 0x00000000
ATI Caps4 = 0x00000000
ATI ctx* =0x50F12888 / 0x02b7d888 ret=0

ATI mmPLL_REF_FB_DIV = 0x08924924
margins: xrest=0 yrest=0 left=0/0 right=0/0 up=0/0 down=0/0

pll: ref_div(M)=4 div_int(N_int)=9 div_frac(N_frac)=2 lock_time=8 Fout=2*xtal

sclk: sysclk_src=0 sysclk_div: fast=0 slow=0

pclk: pixclk_src=0 pixclk_div=0

Hope someone can get something out of this

sensorama

14th August 2008, 06:25 PM

skiwithpete

15th August 2008, 02:26 AM

ace_ddi.dll
http://forum.xda-developers.com/showthread.php?t=296241
last page for the download

sensorama

15th August 2008, 10:44 AM

ace_ddi.dll
http://forum.xda-developers.com/showthread.php?t=296241
last page for the download

What do you mean?
The diamond, for what i can se, doesnt use ace_ddi.dll.

skiwithpete

15th August 2008, 03:55 PM

Sens,

Doesn't have or doesn't make use of?

I thought what was meant was doesn't have - as in 'file was missing.'

As espenfjo said...

But.. I cant seem to find an ace_ddi

But I'm still trying to figure out what your Qualcomm Brew link posting has to do with anything...

sensorama

15th August 2008, 05:43 PM

Sorry for my quick post... was and still is at work so its quite hard to follow the thread at times...

Doesnt have... wich was perfectly mentioned in espens post.

carry on. :o

espenfjo

15th August 2008, 11:35 PM

Hehe, this is funny.

GETRAWFRAMEBUFFER method gives me offset 0x1686a000 for the FB
The GAPI method gives 0x16844800
And GETGXINFO method gives me the same as GAPI.
http://wiki.xda-developers.com/index.php?pagename=DiamondMemoryMap gives 0x1666a000.

Anyone can shed some light on this?

EDIT: Sadly none of the offsets fixes my previous problems with the FB.

premy

16th August 2008, 02:47 PM

Espenfjo,

I'm facing the same issues as you on my Diamond.

However, when you do a pdump 0x1666a000 0x96000, and then feed the output into gimp, you definitely get what's on the screen.

So I would tend to think that 0x1666a000 is correct.

flibblecx

17th August 2008, 09:20 PM

So I would tend to think that 0x1666a000 is correct.

Indeed, if you move the code that clears the frame buffer to before the 'Go Go Go' part it does indeed clear the FB and print to it.

This is probably not much help.

I will continue to do some debugging and find out what is causing it to apparently crash before clearing the FB during a normal run.

espenfjo

18th August 2008, 10:07 PM

parazitu22

18th August 2008, 11:58 PM

New here to the forum, but i've been reading for a while. I don't have a lot of programming skills, but I wish to thank all of the members here who work so hard on this.

Looks like google has released a new SDK v.0.9

http://code.google.com/android/download_list.html

dave94o

19th August 2008, 04:25 AM

Hey guys if you need some help from the people who made the hackint0sh(maybe it will help you) then I can get some information from them,because i know these people and i can talk to them at anytime,so if you need anything just tell me

Global_Inferno

19th August 2008, 09:36 AM

mikesch261

19th August 2008, 11:58 AM

by the way who want Linux or that android Google spyware Crap
is there a Navigation Software for it !?
don't think so

every Software from google got tons of Spyware and Backdoors so why would anyone with a little bit of brain install a crao like this !?

liamhere

19th August 2008, 12:26 PM

gebruikershaes

19th August 2008, 12:34 PM

the release notes say there is no bluetooth added to that release. wich is a shame :(

Smokuevo

19th August 2008, 04:18 PM

someone please make it happen!!!

Global_Inferno

19th August 2008, 07:30 PM

Maybe once the first HTC Android-based phone comes out, we can rip the OS from that? Or HTC will offer an alternative? Never know..

parazitu22

19th August 2008, 08:38 PM

i am a total noob, but I was just wondering why it is more difficult to get android working on the diamond then on the vogue. is it just because it doesn't have an SD card?

espenfjo

19th August 2008, 09:22 PM

I dont know why, but there could be several reasons.
Very different hardware with less publically information about it.
The core developers not moving over to the Diamond yet.
Time.
Etc..

I have as you might have noticed looked into making this happen, but im no expert neither with C, hardware or Linux programming.

I just want to help the persons capable of such things.
But it doesnt seem like they are here? :\

Im really stuck now, no FB in the preloader, and no FB when/if Linux starts, tried using VSFB from athena but no luck.

NikPraskaton

19th August 2008, 09:46 PM

every Software from google got tons of Spyware and Backdoors...

I think the same thing about Microsoft (windows mobile) too ;) lol

Global_Inferno

19th August 2008, 10:41 PM

Well I am a C++ programmer by trade, but the problem is, while I know a fair bit about Linux... The Diamond's software and hardware are completely :O to me..

rjpearce

20th August 2008, 01:15 AM

Hi chaps,

I've read the thread and would like to help. (I'd love to see android on the diamond).

I sent an email to the lists.arm.linux mailgroup about support the the MSM7201A (the chipset for the diamond)..

I got the following response from Brian Swetland:

> > Can anyone confirm if the Qualcomm MSM7201A chipset is supported in
> > kernel v2.6.25 (on the latest GIT repository) ?
> >
> > I notice from here:
> > http://lists.arm.linux.org.uk/lurker/message/20071108.075502.4ff93ee7.html
> > That MSM7200A was supported from Nov 2007. Does this include all MSM7K
> > models?

There is quite a bit of MSM7201A support in the patches available in the
git tree from git.android.com (check out the msm-2.6.25 branch, for
example). The first couple patches are in the mainline 2.6.25, but
that's about it -- we need to clean up stuff and start submitting more
of this for review to the lakml, and hopefully see more complete msm7k
support in later mainline kernels.

Expect to see some new patchsets sent to the lakml in the near future.

I know there are folks working on 7225 and 8k as well.

Brian

---

I dont know if this is any use for your chaps, but is it worth compiling with the latest kernel from google? and then creating a new boot image, just an idea (not my area of expertise!). Happy to help out anyway i can ;)

-Rich

andrewloh

20th August 2008, 04:14 AM

http://www.youtube.com/watch?v=quwXY3w26wg

http://www.ismodel.net/forum/richedit/upload/2k02d337c4cc.jpg http://www.ismodel.net/forum/richedit/upload/2kfa849b14d9.jpg http://www.ismodel.net/forum/richedit/upload/2k76c477a9d9.jpg http://www.ismodel.net/forum/richedit/upload/2kd681f4b9c4.jpg

The Andorid 0.9 SDK r1 Beta is quite impressive. If anyone would port it to HTC Diamond, that would be excellent.

Global_Inferno

20th August 2008, 09:42 AM

It's shouldn't need "porting", simply intergrating with the Touch Diamonds drivers and hardware.

rjpearce

20th August 2008, 11:08 AM

Another email i got from Brian after i told him it was for diamond..

==

Trickiest issue is likely to be the radio firmware. I don't think the
AMSS version people use for winmo and the stuff we're using for Linux
are converged yet, so unless you have some QCT or HTC contacts helping
you, there's probably going to be a good bit of reverse engineering of
the shared memory interface to make everything go.

Good luck!

Brian

Global_Inferno

20th August 2008, 11:50 AM

Owch... Sounds painfull.

andrewloh

20th August 2008, 02:36 PM

There is quite a bit of MSM7201A support in the patches available in the
git tree from git.android.com (check out the msm-2.6.25 branch, for
example). The first couple patches are in the mainline 2.6.25, but
that's about it -- we need to clean up stuff and start submitting more
of this for review to the lakml, and hopefully see more complete msm7k
support in later mainline kernels.

Expect to see some new patchsets sent to the lakml in the near future.

I know there are folks working on 7225 and 8k as well.

In such stage, I think still have to wait for 1-2Q before a workable complied Android CODE will be there for Diamond... :o

skiwithpete

20th August 2008, 04:34 PM

http://www.engadget.com/2008/08/20/htcs-touch-diamond-pro-are-multi-touch-gods-just-not-how-you/

This little beaut is multi-touch, lets hope that Android will take that to the next level.

Delusionz

20th August 2008, 08:29 PM

I'm missing something here,

This thing boots Android from the existing WM6.1 OS? it's not a ROM you install and it wipes the OS? it doesnt flash the ROM right?

If while trying this the phone stucks, a soft-reset will solve it?

Android looks so cool, hope it would work smoother on a diamond than the damn slow WM6.1

cant wait for the final rls and hopefully a rom for the diamond!

gebruikershaes

20th August 2008, 10:25 PM

Is is me or does the "phone" on the video that andrewloh posted look exaclty like the diamond? Could it be that the Dream phone from HTC will look like that?

NikPraskaton

20th August 2008, 10:40 PM

I will definitively try to install Android on my Diamond once someone get it to work...

sensorama

21st August 2008, 11:06 AM

Is is me or does the "phone" on the video that andrewloh posted look exaclty like the diamond? Could it be that the Dream phone from HTC will look like that?

Its just a dummy interface that comes with the SDK.

flibblecx

21st August 2008, 09:54 PM

Ok, i cant seem to make any progress on harets FB. Tried about everything :)
Someone should start looking at linux instead, and just hope that haret actually runs through all the preloader code and executes linux.

From the debugging I have done I am sure that it is haret giving the current issues.

If you get it to clear the FB earlier you will see that it will crash a different points (fractions of seconds apart), sometimes writing small amounts of garbage to the FB.

This leads me to think that it isn't taking control of the hardware correctly and something else tramples on it before it trampolines

I haven't done any more work on it since my last post but wanted to make sure people working on it are aware that we haven't got to the stage where we need to work on the linux zImage yet. I have built both patched kiser and vogue kernels and complete checkouts of the android (git) kernel. Hopefully one will show promising results when we get to that stage.

p65174

21st August 2008, 10:15 PM

From the debugging I have done I am sure that it is haret giving the current issues.

If you get it to clear the FB earlier you will see that it will crash a different points (fractions of seconds apart), sometimes writing small amounts of garbage to the FB.

This leads me to think that it isn't taking control of the hardware correctly and something else tramples on it before it trampolines

I haven't done any more work on it since my last post but wanted to make sure people working on it are aware that we haven't got to the stage where we need to work on the linux zImage yet. I have built both patched kiser and vogue kernels and complete checkouts of the android (git) kernel. Hopefully one will show promising results when we get to that stage.

yes, i agree. From my tests, the problem seems to be in mmu_trampoline (armstuff.S). Some guys on #htc-linux think that it can be sram (??) or physical<=>virtual ram mapping problem. So i can't do better tests on it because i dont really know how it work :( :(

espenfjo

21st August 2008, 10:58 PM

dzo

26th August 2008, 07:34 AM

Ill do some FB tests with haret tonight i think, just to see how stable the FB is before the trampoline.

Its just so odd that i _once_ saw the PSR output from the preloader function called by the asm code (trampoline).
And sometimes i get some garbage just after the trampoline is called, just some pixels drawn where the PSR should be.

EDIT: Ok... FB works perfect before the trampoline, so the next step for me should have been debugging the trampoline itself, the only problem is that i dont understand ASM, and not that good with memory management/MMU stuff, and the other stuff done in the trampoline.

I wrote the MSM7X00 port for the Vogue and the problem is that these devices can't do any I/O with the MMU off (including MDP DMA for the FB). What I did was boot with the MMU on until I could get the VIC and MDP code working but you shouldn't need to do that. All you really need to do is rip out all the clock code and use the android VIC code. For debugging, there is a simple mdp driver (ttyMSM) in my board-htcvogue.c. My patches are at http://cs-alb-pc3.massey.ac.nz/vogue but these won't work completely because the MSM7X01A VIC is different.

Hope this helps.

tughi

26th August 2008, 07:52 PM

HTC Dream comes around October... I don't believe it will be very different in hardware from the Touch Diamond and that would mean that the Dream ROM could be adapted for Touch Diamond :cool:
I really hope this is the case, because WM is driving me crazy :(

NutsyUK

27th August 2008, 01:16 AM

The real question i dont think anyone has asked yet is...

Will HTC or any of the devs here be building a full touchflow3d interface for the Android diamond operating system?

Im under the impression Android is allot less hoggy than wm6.1/7, wouldn't this mean it would be much more suited for customizing and flashyness?

While android is a great alternative to WM its still ugly as f***, so some system wide touchflow 3d would be great :D

skiwithpete

27th August 2008, 10:31 PM

skiwithpete

29th August 2008, 11:01 AM

Results of the developer challenge are in. Have a look at the link below to see the winners and the great new apps coming to the diamond (once Android has been... you know)

http://code.google.com/android/adc_gallery/

premy

30th August 2008, 10:24 PM

I wrote the MSM7X00 port for the Vogue and the problem is that these devices can't do any I/O with the MMU off (including MDP DMA for the FB). What I did was boot with the MMU on until I could get the VIC and MDP code working but you shouldn't need to do that. All you really need to do is rip out all the clock code and use the android VIC code. For debugging, there is a simple mdp driver (ttyMSM) in my board-htcvogue.c. My patches are at http://cs-alb-pc3.massey.ac.nz/vogue but these won't work completely because the MSM7X01A VIC is different.

Hope this helps.

Thanks for the guidance. I understand that you don't think that our problem comes from Haret and the trampoline function, but rather from the kernel. Correct ?

dzo

30th August 2008, 11:28 PM

Thanks for the guidance. I understand that you don't think that our problem comes from Haret and the trampoline function, but rather from the kernel. Correct ?

Yes, it looks like haret because as soon as the MMU is off you can't do any i/o but if you fix the VIC code then it should come to life when the kernel switches the MMU back on. I guess another problem could be the GPU being on since the kernel only knows about MDP.

NikPraskaton

30th August 2008, 11:31 PM

Wish I knew what the f you guys are talking about.. but I sure do hope Ill see Android runnin on my PDA soon :)

skiwithpete

30th August 2008, 11:46 PM

Wish I knew what the f you guys are talking about.. but I sure do hope Ill see Android runnin on my PDA soon :)

Same.

P

skiwithpete

1st September 2008, 01:30 PM

I'd like to help, have been trying to learn little bits (like what a trampoline is).

I have no programming experience (except HTML) but everyone needs to start somewhere right?

Can someone point me in the right direction to learn how to help this progress. Whether that be a beginners coding site, or an android site.

Cheers,

P

doopsie

1st September 2008, 02:26 PM

I know this post is hardly of any use, but i really am super excited to see linux on my phone. :) All the best for the progress! Bless those working on this project, Thank you for taking a time out of your busy schedules just to share your knowledge and make the world a better place to live in. :)

NikPraskaton

1st September 2008, 02:43 PM

I would also like to help.. but sincerely I dont know ANYthing about linux :( Just webdesign..

dave94o

6th September 2008, 08:09 PM

so this is great,no new progress,great

sensorama

7th September 2008, 10:58 AM

yea. Come on. *bump*

sibbor

8th September 2008, 12:20 AM

If you make any progress with Linux/Android on the HTC Touch Diamond, I'm willing to spare a few dollars for donation.

NikPraskaton

8th September 2008, 12:41 AM

If you make any progress with Linux/Android on the HTC Touch Diamond, I'm willing to spare a few dollars for donation.

Me too :) Just to see it runnin...

GurneyHalleck

8th September 2008, 03:00 AM

Perhaps something to bear in mind... it's only a couple of months (-ish) untill the HTC Dream comes out with Android ready to rock. It'll be so much easier to figure out a 'port' to the Diamond with a hands on working version of Android to refer to, that I imagine most devs will wait that couple of months, rather than flailing around in the dark, so to speak.

Besides, who knows - we might even see an official version from Google or HTC?

Global_Inferno

8th September 2008, 11:05 AM

I'd love to think that HTC will bring out a "port" for Touch Diamond users, but the reality is that they haven't really got any reason too... It's a retail product already. Although you could argue it might spark a renewed interest in the Diamond if they did?

molnitza

10th September 2008, 03:03 PM

Is it possible to emulate the complete hardware of the HTC Touch Diamond. That would make it easier to devlop/compile a linux for the phone.

NikPraskaton

10th September 2008, 04:33 PM

Any help from this?

http://vivien.chappelier.free.fr/typhoon/index.html

dcordes

11th September 2008, 01:30 PM

unfortunately it will not help because it is a whole different chipset and quite old.

I sent some info to premy

eranhazout

13th September 2008, 11:52 AM

Hi

I have an HTC Diamond ...
I would like to install Android on my Diamond ... did you find any solution for this ?
you are welcome to send me any version to check on my mobile and I will post the bugs here ...

Thank you

Riel

13th September 2008, 03:24 PM

Read this topic.

sjeems

13th September 2008, 06:17 PM

Read this topic.
@riel:
Erm exactly what topic should we read?
Or were you just informing us you had read this topic? ;)

I wss hoping you have a nice topic with some more info (and maybe even some solutions) on this topic.

Grtz,
Sjeems

NetRipper

14th September 2008, 02:26 AM

Basically the latest news is, that there is no news. Linux is still not booting on Diamond*/Raphael. Even worse, currently it's stuck at the point _before_ even booting the linux kernel (i.e. haret configuration trouble).

* I myself own a Raphael, so I'm not 100% sure if there may be more progress for the Diamond. But as the platforms are almost identical, I doubt it.

skiwithpete

15th September 2008, 04:29 PM

Basically the latest news is, that there is no news. Linux is still not booting on Diamond*/Raphael. Even worse, currently it's stuck at the point _before_ even booting the linux kernel (i.e. haret configuration trouble).

I've done a search, but where can I find more information about haret? and its configuration? Which version is being tested on the diamond?

schmeichler

15th September 2008, 06:08 PM

here is what I found on HaRET (http://handhelds.org/moin/moin.cgi/HaRET)

dave94o

15th September 2008, 06:27 PM

well that is just dendy

dcordes

16th September 2008, 12:01 AM

here's a snippet of irclog which might contain a useful hint for solving the diamond boot issue. btw: does booting the msm7* kernels on raphael show the same symptoms?

<dcordes> dzo gave some interesting hints on the diamond boot problem:
http://forum.xda-developers.com/showthread.php?t=402002&page=14
<swetland> dcordes: actually they can do IO with the mmu off
<dcordes> Espenfjo, if you didn't see it yet, you might look at it.
unlike you were assuming, he says it is more likely to be a problem in
the kernel code
<swetland> but you have to setup the periperhal window correctly
<dcordes> hm what's that?
<swetland> arm11 has sort of an alternate bus for peripherals
<swetland> which can be accessed by setting the right bits in the PTEs
(with MMU on)
<swetland> or configuring a range of address space for it (with MMU off)
<swetland> periph_2gb_open: /* open the peripheral port 2gb */ /* see
1136trm 3-164 */ ldr r0, =0x80000016 mcr p15, 0, r0, c15, c2, 4 bx lr
* swetland kicks irssi
<dcordes> no clue what are PTEs either. maybe I should put that info
in the forum
<swetland> imagine there are some linebreaks in there ^^
<swetland> page table entries
<Espenfjo> hmm, yeah, ok
<Espenfjo> Sadly the MMU stuff is way out of my league :\
<dcordes> Espenfjo, you understood what dzo said further? (because I
have no clue :'( )maybe we should pass that info to premy and dzo
<dcordes> I mean what swetland just said.. sorry
<Espenfjo> No, not really :)
<Espenfjo> or.. well.. somewhat.. but no clue how to address those issues
<swetland> feel free to repost. if their issue is peripheral access
when the mmu is off, the above snippet should do the trick.
<swetland> of course the kernel should be turning the mmu on first thing

skiwithpete

17th September 2008, 11:55 AM

Addicteddddd

17th September 2008, 12:41 PM

let's make an IRC channel and point everyone in that direction, id suggest somewhere big like EFNet

JustANotherMe

17th September 2008, 01:26 PM

QN maybe ?

NetRipper

17th September 2008, 03:22 PM

#htc-linux @ freenode? It's the channel advertised at xda wiki pages for development. If there are some people with lowlevel experience that'd be great.

Addicteddddd

18th September 2008, 12:48 PM

#htc-linux @ freenode? It's the channel advertised at xda wiki pages for development. If there are some people with lowlevel experience that'd be great.

cya there :)

xcesspda

18th September 2008, 10:13 PM

dcordes

18th September 2008, 10:18 PM

They must provide the kernel code they run on the dream. Indeed it might be interesting and also might help with the diamond boot problem. To my knowledge there is no real htc specific code in google git at this point.

lemagi

20th September 2008, 06:15 PM

you can run Linux on htc diamond? how?

Addicteddddd

20th September 2008, 06:55 PM

you can run Linux on htc diamond? how?

good work on not reading any of the posts in this thread

lemagi

20th September 2008, 07:14 PM

I do not understand English well! I might say if you can do and how?

gebruikershaes

20th September 2008, 08:18 PM

You can't..

kebe

23rd September 2008, 08:18 PM

Today G1 was announced. It seems that hardware is the same as Diamond, Touch Pro and HD. There is a light in this tunnel :)

TripleQ

23rd September 2008, 09:35 PM

Yeah!
It looks like they both have the same processor!
a Qualcomm® MSM7201A™, 528 MHz

G1 with Android specifications (http://www.htc.com/www/product/g1/specification.html)
Touch pro/Android specifications (http://www.htc.com/www/product.aspx?id=49524)

What are the experts saying about this?

skiwithpete

23rd September 2008, 10:48 PM

Well, if the hardware is the same, where is the code?!?

Fraction

23rd September 2008, 11:49 PM

3. You'll have tons of hardware options. Android's designed to be versatile, so lots of manufacturers will be putting it on lots of phones—ones with QWERTY keyboards, touchscreens, T9, outrageously spec'd out phones, as well as ones that are kinda crappy in the spec department, actually. But this also provides a common platform for developers, making it easy to put their apps on millions of phones. It's the benefit of any OS that runs on a lot of hardware—like Windows or Linux, etc. Of course, this is also the Windows Mobile argument against all of the other proprietary OSes like Palm and BlackBerry.

" Jkr's bold comment:
3) It doesn't matter who carries the android. Once the OS is ported to your device, you can install it. We at XDAdevelopers have been doing this for years w/ windows mobile. It doesn't even mater if the OS is open or closed, we just use the OS as modular. We interchange the radio stack, and hardware profiles. Since Android is open, we can just make the conversion that much more seamless. What's more, most devices will be ported already by the carrier, since most there is a lot of overlap in what devices carriers provide. And lastly, unlocked versions of almost all device specific winmo get's leaked, android will be even more so. <....>"

source: Five reasons to be super psyched about android (and five not to be) (http://gizmodo.com/5052576/five-reasons-to-be-super-psyched-about-android-and-five-not-to-be)

boolean22

23rd September 2008, 11:56 PM

Code should be available later in the year...

I wonder if would be tricky to get this to work on our little diamond?

thevery

24th September 2008, 01:18 AM

Well, if the hardware is the same, where is the code?!?

well, some code is pblished here:
http://code.google.com/p/android/downloads/list

and it seems that newer files are uploaded right now

benko286

24th September 2008, 10:50 AM

so still imposible that android works on our diamond

:(

MasterMerlin

24th September 2008, 11:36 AM

As mentioned so often .. "patience is a virtue"

Wait what time will bring. I can rememeber when it was impossible to run WM6 or WM6.1 on BlueAngel ...

commodore65

24th September 2008, 11:37 AM

come on guys lets get this to work, we can shutup those iphone users for good. any idea of time scales what are the stumbling blocks for this at the moment

Riel

24th September 2008, 11:50 AM

Android looks really impressive !

Am thinkin about buying a G1 :p

But maybe the diamond has better hardware?

is it running that smooth or also just Windows Mobilish? Anyone got a hand on it?

faethon

24th September 2008, 11:59 AM

come on guys lets get this to work, we can shutup those iphone users for good. any idea of time scales what are the stumbling blocks for this at the moment

The Android OS source code is not yet available I think. According to this post on the Android blog: announcing-android-10-sdk-release (http://android-developers.blogspot.com/2008/09/announcing-android-10-sdk-release-1.html) we can expect the Android source code to be available Q4 this year. That might mean tomorrow, but could also be 31st of December... IF it does not slip into 2009. Untill then we have to wait for it to be ported to the Diamond.

faethon

24th September 2008, 12:42 PM

Android looks really impressive !

Am thinkin about buying a G1 :p

But maybe the diamond has better hardware?

is it running that smooth or also just Windows Mobilish? Anyone got a hand on it?

The G1 has the same processor as the Diamond (Qualcomm MSM7201A). It's screen is smaller (320x480 vs 480x640) than the Diamond though. Amount of memory is slightly different as well.

So, from the hardware point of view, Android should be able to run on the Diamond without any problems.

schmeichler

24th September 2008, 01:03 PM

The G1 has the same processor as the Diamond (Qualcomm MSM7201A). It's screen is smaller (320x480 vs 480x640) than the Diamond though. Amount of memory is slightly different as well.

just to complete this

no d-pad / the complete bottom plate
no front cam
no focus on main cam
no internal storage
G1 has a sd-slot
G1 has a hardware keyboard
G1 has a digital compass!

but I am also very confident that we will see Android on our Diamonds ;)

Fraction

24th September 2008, 01:18 PM

Deleted...

Reason, dobbel post.

Fraction

24th September 2008, 01:19 PM

just to complete this

no d-pad / the complete bottom plate
no front cam
no focus on main cam
no internal storage
G1 has a sd-slot
G1 has a hardware keyboard
G1 has a digital compass!

but I am also very confident that we will see Android on our Diamonds ;)

So my Touch Pro, is like x2 better then G1.?

like it has all that the Diamond and G1 got, in hardware :P

*wiii*

thevery

24th September 2008, 02:43 PM

The Android OS source code is not yet available I think. According to this post on the Android blog: announcing-android-10-sdk-release (http://android-developers.blogspot.com/2008/09/announcing-android-10-sdk-release-1.html) we can expect the Android source code to be available Q4 this year. That might mean tomorrow, but could also be 31st of December... IF it does not slip into 2009. Untill then we have to wait for it to be ported to the Diamond.

IC that it is already available:
http://code.google.com/p/android/downloads/list

NetRipper

24th September 2008, 03:03 PM

IC that it is already available:
http://code.google.com/p/android/downloads/list
Now we just need the kernel and the initrd of the HTC Dream / G1. And then we also need to get HaRET working for the Diamond (and Raphael), before we can even attempt booting the linux kernel.

Still a few steps to go... HaRET being the most important one.

skiwithpete

24th September 2008, 03:21 PM

Now we just need the kernel and the initrd of the HTC Dream / G1. And then we also need to get HaRET working for the Diamond (and Raphael), before we can even attempt booting the linux kernel.

Still a few steps to go... HaRET being the most important one.

Can you post screenshots or information about HaRET failing?

I don't know much about the subject, but can research specific problems for you.

NetRipper

24th September 2008, 03:44 PM

Can you post screenshots or information about HaRET failing?

I don't know much about the subject, but can research specific problems for you.
Erm, no I can't. It's a bit more low level.

Basically, as I understood (I'm not an expert either), we need to find GPIOs and IRQs for the MSM7201A processor (used in Diamond/Raphael). Perhaps specific support has to be coded into HaRET as well for this processor.

Currently HaRET hangs on the "Go Go Go" message, which is the moment it really goes low-level by printing to the raw framebuffer, disabling hardware (including MMU) and attempting to boot linux. It's hanging at one of those points. The printing to raw framebuffer does not work, so it's hard to see where exactly it goes wrong.

faethon

24th September 2008, 04:09 PM

Erm, no I can't. It's a bit more low level.

Basically, as I understood (I'm not an expert either), we need to find GPIOs and IRQs for the MSM7201A processor (used in Diamond/Raphael). Perhaps specific support has to be coded into HaRET as well for this processor.

Currently HaRET hangs on the "Go Go Go" message, which is the moment it really goes low-level by printing to the raw framebuffer, disabling hardware (including MMU) and attempting to boot linux. It's hanging at one of those points. The printing to raw framebuffer does not work, so it's hard to see where exactly it goes wrong.

So how does Android do it? Is it a similar system or is there a bootloader in ROM somewhere that loads the Android OS directly? Does that mean we need to change the SPL to bootstrap the Linux/Android OS as well? (also bit of a newbie, but willing to look into this)

NetRipper

24th September 2008, 04:26 PM

So how does Android do it? Is it a similar system or is there a bootloader in ROM somewhere that loads the Android OS directly? Does that mean we need to change the SPL to bootstrap the Linux/Android OS as well? (also bit of a newbie, but willing to look into this)
On the G1 the kernel will most likely be on the ROM, booted by the SPL (instead of booting windows mobile).

I'm not sure how the SPL finds and initiates the OS (either windows or linux). If it's OS specific, we'll need a modified SPL. If it's not, then we probably don't. Once the G1 is out, the ROM will be checked out and we'll probably find out.

skiwithpete

24th September 2008, 04:27 PM

Tiny bit of info about the Qualcomm MSM7201A RISC Chipset.

http://pdadb.net/index.php?m=cpu&id=a7201a

It has a list of other devices that use the same chip which might be the most useful thing there

Other linux chat about the chip here

http://groups.google.com/group/linux-msm-kernel/browse_thread/thread/590f2db32981569f

remi.rez

24th September 2008, 06:18 PM

lexi1986

24th September 2008, 06:46 PM

With this .cab you can try to let Android start but if you run it you will get the screen Haret and then youre phone will hang and you must soft reset the phone. So this .cab is not working for our phone.

Dj!pOner

24th September 2008, 06:47 PM

Wow!! Really!!
Does it work on Diamond?

remi.rez

24th September 2008, 07:28 PM

Wow!! Really!!
Does it work on Diamond?

xda-developers > Diamond > Diamond ROM development

So this .cab is not working for our phone

:o

lexi1986

24th September 2008, 08:10 PM

xda-developers > Diamond > Diamond ROM development

:o

Some people can't read :p. I think we must wait till someone get's the G1 version of Android. The hardware is almost the same as the Diamond so i think it will work on the Diamond verry soon.

exclaim

24th September 2008, 11:53 PM

The G1 has a slide out QWERTY keyboard for text entry and I think I read somewhere that the version of Android on the G1 solely relies on that for text entry. I know the G1 has a touch screen and it seems to be used to great effect although no on screen keyboard.

If indeed Android from G1 can be ported over to Diamond, how would the issue of text entry be resolved?

vanaedium

25th September 2008, 02:41 AM

Hi
I,ve got a lot of experience with linux on pc, I don't use windoze since 5 years and I'm very happy about it and I can say that:
Android will be loaded (on htc diamond, hd, pro) as OS by a linux native SPL, that now we call GRUB or LILO (pc world of course), Haret for what I understand on this 20 pages is a way to interface SPL bootloader with linux os, the problem is that in linux world linux can see "win" hardware but win hd can't see linux.
So the solution will be rip only the linux native bootloader from G1 and the rest will come itself!!!! otherwise contact the sourceforce team section grub to ask them to develop an easy bootloader (they've already done with ipod, nintendo ds and so many different devices!).
My "sensation" of haret failing is connected to vga visualization....
Who's scary about text message input? C'mon the best thing in android is the open source solution and the modularity, everyone will be supported by a super massive community and tons of applications for touchscreens, I'm pretty sure we will see also a swype touch keyboard in android.
Just wait until t mobile will sell G1 in US, can be late october.
Android will for sure "kill" th f*****g, lazy, heavy WM forever!! ;)
Cheers guys

Fruzion

25th September 2008, 03:24 AM

insaino

25th September 2008, 07:04 AM

I so hope there is a console that i have to type crap like "sh dialler-answer" no wait lets make it more confusing, "sh Dlr-anr 100.1 /?#$"...
üBER!

Howell

25th September 2008, 11:06 AM

Has anyone tried this cab file?
It is spposed to install Android, but it is not said on whch device.
It seems a bit dubious though as a cab file implies that whatever's installed will run on top of WinMo.
Or did I misunderstand and it's an install file for the SDK????
Weird

http://www.myhtcphone.com/downloads/14-files/3-file-for-automatic-and-easy-installation

I tried this CAB on my Diamond (attached is haretlog(1).txt)...

My Diamond froze and I had to soft-reset as it has already been stated...

But I noticed that a "Storage Card" folder is created at the root of my Diamond instead of using "Internal Storage".

I tried to move the content of that folder (data.gz and system.gz) to internal storage but had no more luck... (attached is haretlog(2).txt)

skiwithpete

25th September 2008, 11:47 AM

I tried this CAB on my Diamond (attached is haretlog(1).txt)...

My Diamond froze and I had to soft-reset as it has already been stated...

But I noticed that a "Storage Card" folder is created at the root of my Diamond instead of using "Internal Storage".

I tried to move the content of that folder (data.gz and system.gz) to internal storage but had no more luck... (attached is haretlog(2).txt)

I looked at your text outputs, and noticed ace_ddi problems that had been noted before.

it might be useful: I posted a link to ace_ddi on page 10 of this forum. Not sure if that will help anything, but you might want to give it a try.

Best,

P

khazm2k

25th September 2008, 03:04 PM

Tried to install that Android... No go on my phone since there are no SD card slot it doesn't even want to install. I just get cannot install Android... WHY ANDY WHY???

NetRipper

25th September 2008, 05:03 PM

Android will be loaded (on htc diamond, hd, pro) as OS by a linux native SPL, that now we call GRUB or LILO (pc world of course), Haret for what I understand on this 20 pages is a way to interface SPL bootloader with linux os, the problem is that in linux world linux can see "win" hardware but win hd can't see linux.
So the solution will be rip only the linux native bootloader from G1 and the rest will come itself!!!!
I'd like to see SPL as some sort of BIOS when compared to PC. And I'd see HaRET as some kind of advanced bootloader as it also shuts down the hardware to start freshly from within windows mobile. It just depends on how much of a BIOS the SPL acts like. If it simply looks on a specific memory location for a kernel, it could easily be replaced by a linux kernel.

I looked at your text outputs, and noticed ace_ddi problems that had been noted before.
The ace_ddi should not be neccesary. I also read elsewhere that it was neccesary for the framebuffer in some situations, but i have seen the framebuffer work with a dummy test, so that should not be the case.

I just get cannot install Android... WHY ANDY WHY???
Patience... devs are working hard on it and when it's operational, it'll be posted on this forum and thread.

rerven

25th September 2008, 06:49 PM

hope u guys can make that android platform run on diamond ...

waiting for good news.

clarkemcrobb

25th September 2008, 08:29 PM

charly_mx

25th September 2008, 09:02 PM

Ive been a member of this site for a good few months now and I spend most of my nights browsing and learning, and this is my first post, ever! :)
This is something I am very excited about and I'm sorry this is off topic but all you really smart guys need plenty of support,
I don't have any particularly useful skills or knowledge but I do have an income!

I'm sure there are lots of people like me who are willing to donate plenty in order to get someone properly motivated.

Somebody must be close to figuring something out? I'd like to show my support by donating!

Could we all start chipping in to a pot for one of the developers to buy a G1 device as soon as it is available? Would that help us figure out how to get it onto our Diamonds?

Its great that you are here. This forum is great. Try to read a lot because there is a lot on info about many things. There are many developers and ROM cooks (they call themselves that way because they prepare the ROMS by their own taste!!) and the will love the donations. All they work they do they do it just for fun and are always willing to help everyone. I have not been able to donate :( i dont have that great income, however if you can support this guys it will be great. One cup of coffe and they will be happy :)

Howell

25th September 2008, 09:37 PM

1.0 SDK is out with a new build and a CAB is available :
http://www.myhtcphone.com/downloads/14-files/3-file-for-automatic-and-easy-installation
or
http://code.google.com/android/download_list.html

Update on September 25, 2008:

With the release of the Android SDK 1.0, a quick update of the port has also brought it to 1.0. This is by no means a stable release as I think Google rushed it to market to match the announcement of Tmobile's G1. This is a milestone however. Look for Google to add features and improve stability in the near future.

Martin's (dzo) release notes for this version:

"Updated to the newest SDK (1.0), Audio should work well now. SMS Works using the Google app."

Update on September 23, 2008:

Martin's (dzo) release notes for this version:

"Audio is working now, phone will ring and play music, SMS works using my old testing app, now has full screen keyboard. The volume up and down keys are now mapped to back and menu. This is all with the 0.9 SDK, get the latest android-new.zip to try it. There are still problems with suspend, sometimes it doesn't wake up properly when you press the power button. If this happens press the red end call button, wait till the led goes red and try again. Sometimes the phone will reboot after an audio call, I'm looking into this. "

Same symptoms, here is the new log ;)

Fruzion

25th September 2008, 10:10 PM

The more i think about it, the Touch pro (with android) would work the exact same as the Dream, and would lose no functionality, except maybe the flash on the camera... We could just wait for that, it's less than a month here in the US...

Quick Question: How fast do you think the Android ROM will be ripped of the Dream, once it comes out?

onex77

25th September 2008, 10:22 PM

I looked at your text outputs, and noticed ace_ddi problems that had been noted before.

it might be useful: I posted a link to ace_ddi on page 10 of this forum. Not sure if that will help anything, but you might want to give it a try.

Best,

P

First things first: i'm a total noob to this. This got me thinking though: certain ROM's change the name of "Internal Storage" to "Storage Card". Is that defined in the ROM or SPL? If it can be defined before installing Android, maybe it will find the "Internal Storage/Storage Card"... just thinking out loud...

nummer20

26th September 2008, 09:18 AM

The more i think about it, the Touch pro (with android) would work the exact same as the Dream, and would lose no functionality, except maybe the flash on the camera...

And the digital compass.

cr2

26th September 2008, 05:02 PM

The more i think about it, the Touch pro (with android) would work the exact same as the Dream, and would lose no functionality, except maybe the flash on the camera...

My raphael has flashlight, and it's hooked to a gpio58:
http://wiki.xda-developers.com/index.php?pagename=Raphael_GPIO

How fast do you think the Android ROM will be ripped of the Dream, once it comes out?
Let's hope that androids will publish the _full_ kernel source.
Dumping ROM may be difficult. Linux is a much more secure system than, hm, wince ;) So it depends on what t-mob/htc will use.

cr2

26th September 2008, 05:03 PM

And the digital compass.

Raphael does not have a compass ?

cr2

26th September 2008, 05:05 PM

touch pro ftw! that would fix text input too, and with the same driver and all too, so no additional modding (i would think)

G1 has a different (half-vga vs vga) LCD panel, so modding will be necessary.

skiwithpete

27th September 2008, 03:23 PM

just imagine a world where XDA devs port Android to the diamond before the G1 is released...

itsallsubliminal

27th September 2008, 11:01 PM

just imagine a world where XDA devs port Android to the diamond before the G1 is released...

1 month to go... :D i hope so too.

yataglanzy

27th September 2008, 11:53 PM

Some links about android porting on htc titan.
... I hope this can help even a little

http://www.talkandroid.com/136-android-htc-titan/
http://wiki.xda-developers.com/index.php?pagename=HTC_Titan

cristian974

28th September 2008, 11:13 AM

Hello...

http://www.handhelds.org/~koconnor/haret/haretconsole-20080927b.tar.gz

"Found machine Generic MSM7xxxA"

but HaRET don't start Android...

Freeze on "Go Go Go..."

cr2

28th September 2008, 01:13 PM

"Found machine Generic MSM7xxxA"
but HaRET don't start Android...

"Android OS" is just a marketing term innovated by google.
Haret loads and starts the linux kernel image (zImage), which
is device-specific. HTC creates so much device/driver variations,
that you can't expect a zImage for a different device working for you.
AFAIK, nobody has created a zImage for Diamond yet.

mligor

29th September 2008, 09:39 AM

"Android OS" is just a marketing term innovated by google.
Haret loads and starts the linux kernel image (zImage), which
is device-specific. HTC creates so much device/driver variations,
that you can't expect a zImage for a different device working for you.
AFAIK, nobody has created a zImage for Diamond yet.

That is true, but zImage is just Linux kernel, you do not have different kernels for different computer configurations (it is mostly just CPU related). I have tried yesterday, and seams that it doesn't even come so far to load zImage... and my opinion is that if we get Haret to work, we will able to boot some zImage (maybe without some drivers, or maybe will crash during loading, but that will be one step forward anyway).

kState

29th September 2008, 12:59 PM

hi, in haret cvs before some hour is changes relating to Raphael :)
i try compile it, now say: "Found machine Raphael", but still freeze :(

sibbor

29th September 2008, 01:30 PM

hi, in haret cvs before some hour is changes relating to Raphael :)
i try compile it, now say: "Found machine Raphael", but still freeze :(Oh, nice! Hopefully we'll see some more progress soon. Keep us updated!

TripleQ

29th September 2008, 03:31 PM

Yeah I also notificed that...

I just compiled it with the latest CVS source..

For people who are interested..see attachment for that version of haret.

Juggalo_X

29th September 2008, 03:58 PM

Yeah I also notificed that...

I just compiled it with the latest CVS source..

For people who are interested..see attachment for that version of haret.
tried it for giggles as i am really interested in android.

with your haret, it doesnt even attempt to load i get "Please se start of ram (RAMADDR)"
then it errors out when it opens

TripleQ

29th September 2008, 04:12 PM

Hmm thats strange..

I just worked here..

These are the release changes for this version.
20080928 0.5.2 <kevin@koconnor.net>, <ipaqlinux@oliford.co.uk>, <pmiscml@gmail.com>

* Add support for "Centrality" arm cpus.

* Better support for MSM7xxx and MSM7xxxA machines.

* Support arm6 MMU extensions. The arm6 mmu no longer supports 1K
"tiny" pages. It also adds several new page flags.

* Scan for all 1Meg mappings suitable for virtToPhys translation at
haret startup.

* Check that haret is running in "system" mode at startup - warn if
not.

* Enhance "mmutrace" to support "stm", "ldm", and "swp"
instructions.

* Add new "mmumerge" feature to "wirq". The "mmumerge" tool allows
one to detect if wince is changing the mmu mappings during a
"wirq" session.

* The high frequency polling system (the "TRACES" variable) will now
also report the approximate PC location. This can help find what
code is changing memory when "mmutrace" isn't available.

* Support "wirq" even on older machines that don't have
"allocPhysMem".

* Support GETGXINFO method of finding the video frame buffer.

* Only dynamically bind to NLedSetDevice(). This fixes startup
problems on some wince devices that don't have this function.

Juggalo_X

29th September 2008, 04:20 PM

Hmm thats strange..

I just worked here..

These are the release changes for this version.
Ill try again
maybe it was a fluke

Edit. i got it working. still locks up on "Booting Linkux..."

Howell

29th September 2008, 05:17 PM

Tried with Android092408 and latest Haret (given a few posts before by TripleQ) and the result is the same...

Here is haretlog.txt :

TripleQ

29th September 2008, 05:28 PM

I see some differents in this new version I compiled...

The post above this one was also made by the new 0.5.2_TripleQ version

and his haretlog says:

Welcome, this is HaRET 0.5.2_TripleQ running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Generic MSM7xxxA/MSM7xxxA (Plat='PocketPC' OEM='HTC Touch Diamond P3700')

and I have a Touch pro (rapheal)
and it says

Welcome, this is HaRET 0.5.2_TripleQ running on WindowsCE v5.2
Minimal virtual address: 00010000, maximal virtual address: 7FFFFFFF
Detected machine Raphael/MSM7xxxA (Plat='PocketPC' OEM='HTC Touch Pro T7272')

So this new version is only for the touch pro?
But is also freeze the phone when booting linux...

kState

29th September 2008, 05:47 PM

[...]
So this new version is only for the touch pro?
[...]

i look in source and diamond is not in machine list, but some changes maybe help on diamond too, i don't know, anyway Diamond&TouchPro still have some problem as say some people in this tread, mmu ? fb ?

keerttttt

29th September 2008, 06:19 PM

Will we be needing HaReT for the android that will probably be 'ripped' from the T-mobile G1 / HTC Dream?

cr2

29th September 2008, 06:33 PM

I
So this new version is only for the touch pro?

It's only the name that is registered. diamond and raphael are
mostly identical in hardware, so wait a bit for the raphael zImage.
There are 3 LCD panels used by HTC on these devices (toppoly, hitachi and
sharp), and the framebuffer driver needs to detect them and initialize properly.
It seems that current raphael phones use hitachi LCD, so it'll be the first supported by linux. Unless you'll do it faster for other panels :)

cr2

29th September 2008, 06:36 PM

Will we be needing HaReT for the android that will probably be 'ripped' from the T-mobile G1 / HTC Dream?

G1 will boot directly from NAND flash, and will not need haret.
Don't forget that G1 has a different LCD, so you can't just use G1 zImage
for diamond/raphael. Let's hope that G1 will use the same TI1251 wifi chipset.

edwinzhang

29th September 2008, 10:41 PM

Any progress now?

keerttttt

30th September 2008, 02:51 PM

Nothing new yet?

nummer20

30th September 2008, 03:24 PM

Be patient.

ballex

30th September 2008, 03:53 PM

I'm following this thread intensive right now. To me it's very exciting if it will be able to run android on the device. You go guys!

Dits_s

30th September 2008, 04:14 PM

Good luck to everyone working on this! Cant wait for the results :D

Riel

30th September 2008, 04:45 PM

Well, please cut it out making posts like 'nothing new yet'. I also want to know, and I see a new post every time beeing just SPAM.

I think we will know it when something happens in this area, please stop ASKING for progress, sjeez.

remi.rez

30th September 2008, 05:17 PM

I know nada when it comes to Linus and such.
I have kind of understtod that it's necessary to be able to boot HaRET to be able to launch Android and it's not possible at the moment because of unknown IRQs and GPIOS (just spitting out what I've read a few posts ago).

I've done a little bit of reading on HaRET and there seem to be a GPIOs polling function included which you launch from a desktop PC remotely connected to the device.

Has anyone tried it?

Here is the link : http://handhelds.org/moin/moin.cgi/HaRET_20Documentation

Section exploring a device.

My 2 cents and sorry if it's common knowledge already :-)

remi.rez

1st October 2008, 03:16 PM

Did I kill the thread?
Sorry lol

NetRipper

1st October 2008, 04:15 PM

Did I kill the thread?
Sorry lol
IRQs and GPIOs are being investigated and documented on wiki.

http://wiki.xda-developers.com/index.php?pagename=RaphaelResearch

HaRET doesn't seem to boot the kernel yet, or, the kernel hangs in a very early stage. I'm still trying to figure that one out.

I'll definitely post something here when I know something is running ;-)

mishulica21

2nd October 2008, 12:10 PM

nummer20

2nd October 2008, 12:33 PM

They have it running on a Touch, and looked at pdadb.net to list other devices with the same chipset. This version was already available but doesn't work on the Diamond if I recall correctly.

gebruikershaes

2nd October 2008, 12:33 PM

I don't get it. How can they run it on the diamond when the diamond has no SD card. I say: bullshit.

TripleQ

2nd October 2008, 12:34 PM

They are using the same cab file as already discussed in this thread...
Its not working for our kind of processor..

remi.rez

2nd October 2008, 12:36 PM

I'm gonna give a shot at his package, see if it really works on the Diamond.
I have a cooked rom though so might run into trouble.
Will report soon.

NetRipper

2nd October 2008, 12:40 PM

here is a page where some russians say they managed to run android on several HTC devices including Diamond . The page is in russian so the translation is not perfectly correct.

http://translate.google.com/translate?u=http://grigory.wordpress.com/2008/10/01/android-sdk-1-on-htc-touch/&hl=en&ie=UTF-8&sl=ru&tl=en
They mention the Diamond, but they mean the HTC Touch (non-Diamond). The one on the videos is the HTC Touch. The Touch has an already supported processor.

The vogue kit does not work on the Diamond or the Raphael, it supports the MSM7xxx processor, not the MSM7xxxA.

remi.rez

2nd October 2008, 12:52 PM

As expected, it hangs on the Go Go Go message.
No luck!