[nchntrman]

Has anyone else seen md5 chechsum errors when unpacking the Firmware system.img? Using the official release of both 7_24 and 7_4 extracting the system.img to a folder then checking with the system_check.md5 included in the official release. I get about 49 failed files (mostly in the /bin directory). I was wondering if anyone else had this problem as well. Trying to determine if it is my md5summer application, my .img extractor, OR is it possible that the system_check.md5 is not correct as included in the firmware zip file.

 

[clockworx]

the bin directory contains quite a few symbolic links, so it is quite possible that could be the reason. Are you doing this on Windows or Linux?

 

[nchntrman]

Windows XP

 

[clockworx]

Yeah, that's probably it. Windows will probably just drop the symlinks altogether, making the 0-length files. In Linux/Android they're almost all linked to bintoolbox, which is apparently a hodgepodge of system tools crammed into a single binary.

 

[mindseye73]

Has anyone else seen md5 chechsum errors when unpacking the Firmware system.img? Using the official release of both 7_24 and 7_4 extracting the system.img to a folder then checking with the system_check.md5 included in the official release. I get about 49 failed files (mostly in the /bin directory). I was wondering if anyone else had this problem as well. Trying to determine if it is my md5summer application my .img extractor, OR is it possible that the system_check.md5 is not correct as included in the firmware zip file.

which .img extractor are you using?take a look here ---> http://jiggawatt.org/badc0de/android/index.html

 

[clockworx]

I really don't think it is his utility. Here's an easy way to check. Look at the md5 file itself. Are there a bunch of MD5's that are exactly the same? (toolbox for sure, I think mv and ls are as well)

c86b1c1e257266a7995cf070964160af ./bin/catc86b1c1e257266a7995cf070964160af ./bin/chmodc86b1c1e257266a7995cf070964160af ./bin/chownc86b1c1e257266a7995cf070964160af ./bin/cmpc86b1c1e257266a7995cf070964160af ./bin/datec86b1c1e257266a7995cf070964160af ./bin/ddc86b1c1e257266a7995cf070964160af ./bin/dfc86b1c1e257266a7995cf070964160af ./bin/dmesgc86b1c1e257266a7995cf070964160af ./bin/geteventc86b1c1e257266a7995cf070964160af ./bin/getpropc86b1c1e257266a7995cf070964160af ./bin/hdc86b1c1e257266a7995cf070964160af ./bin/idc86b1c1e257266a7995cf070964160af ./bin/iftopc86b1c1e257266a7995cf070964160af ./bin/insmodc86b1c1e257266a7995cf070964160af ./bin/ioctlc86b1c1e257266a7995cf070964160af ./bin/killc86b1c1e257266a7995cf070964160af ./bin/lnc86b1c1e257266a7995cf070964160af ./bin/logc86b1c1e257266a7995cf070964160af ./bin/lsc86b1c1e257266a7995cf070964160af ./bin/lsmodc86b1c1e257266a7995cf070964160af ./bin/mkdirc86b1c1e257266a7995cf070964160af ./bin/mountc86b1c1e257266a7995cf070964160af ./bin/mvc86b1c1e257266a7995cf070964160af ./bin/netstatc86b1c1e257266a7995cf070964160af ./bin/newfs_msdosc86b1c1e257266a7995cf070964160af ./bin/notifyc86b1c1e257266a7995cf070964160af ./bin/printenvc86b1c1e257266a7995cf070964160af ./bin/psc86b1c1e257266a7995cf070964160af ./bin/rebootc86b1c1e257266a7995cf070964160af ./bin/renicec86b1c1e257266a7995cf070964160af ./bin/rmc86b1c1e257266a7995cf070964160af ./bin/rmdirc86b1c1e257266a7995cf070964160af ./bin/rmmodc86b1c1e257266a7995cf070964160af ./bin/routec86b1c1e257266a7995cf070964160af ./bin/schedtopc86b1c1e257266a7995cf070964160af ./bin/sendeventc86b1c1e257266a7995cf070964160af ./bin/setconsolec86b1c1e257266a7995cf070964160af ./bin/setpropc86b1c1e257266a7995cf070964160af ./bin/sleepc86b1c1e257266a7995cf070964160af ./bin/smdc86b1c1e257266a7995cf070964160af ./bin/startc86b1c1e257266a7995cf070964160af ./bin/stopc86b1c1e257266a7995cf070964160af ./bin/syncc86b1c1e257266a7995cf070964160af ./bin/toolboxc86b1c1e257266a7995cf070964160af ./bin/topc86b1c1e257266a7995cf070964160af ./bin/umountc86b1c1e257266a7995cf070964160af ./bin/vmstatc86b1c1e257266a7995cf070964160af ./bin/watchpropsc86b1c1e257266a7995cf070964160af ./bin/wipe

See how they're all exactly the same? That's because in Linux they're all linked to toolbox, the "real" one. In windows, the symlinks dump as empty files, so that's why he gets a mismatch. If he went through the trouble to copy "toolbox" to all those filenames , they would be correct again.

 

[nchntrman]

Well then that might take windows out of the equation? I presume the yaffs packer will build an incorrect image unless done in an environment that supports symlinks.

 

[clockworx]

In terms of building roms, yeah. The other possibility is to script in symlinking after the flash, which I believe one of the example "updater-script" that I linked to has in it. It's not really that much hassle.

 

[stoic]

Ive been trying to extract system.img from several roms but almost never succeed, sometimes a partial, sometimes broken image. In using Ubuntu 10 an unyaffs, and the perl scripts, but no luck. Any one have suggestions?Also Ive found that the SMDK6410 it the samsund mobile development kit. It also appears that the SmartQ5 has a newer android based rom available and the hw was created using the same kit so I supect that there would be some usable parts, apps and kernel.

 

[roebeet]

I'm using Lucid x64 and no issues with unyaffs (yet). have you used it on a PDN system.img, just as a baseline so you know it's working? Maybe the other ROM's are corrupt / unusable with the tool.Anything useful in the kit, like a base system.img?

 

[stoic]

I havent found the images(sdk) released with the smdk6410, but the SmartQ5 android is supposed to be either 2.1 or 2.2. Its a single file in a rar. I ran the perl script on it and it completed. if I then run unyaffs on the img it fails. (the smdk6410 appears to be the develpment platform used for the PDN)

 

[clockworx]

Which perl script? I downloaded a SmartQ5 image and just got that big file too. Had no idea what to do with it.

 

[stoic]

I think someone else posted this too but here is the how to and scripts: HOWTO: Unpack, Edit, and Re-Pack Boot Images - Android Wiki

 

[roebeet]

What's the SMARTQ5 link? I'll take a stab at it.

 

[clockworx]

I couldn't find 2.1 for SmartQ 5, only for SmartQ V5 (which I believe is a different platform, sadly)

 

[stoic]

I think the one on thier site is linux based. (after using thier extraction tool)(and another AndroidHere is a 2.1 (DLing now) JT's Digital World ? Android 2.1 on SmartQ V5The Smartq5 is a 6410 based device and appears to be based off of the same smdk6410 development board. I also found thier extraction util(jp). -special extractor? so far i cant get in to the good parts out of it and hex editor shows hhtech where it shouldnt be- might be a dead end unless we can extract the image.--ok well there are several android firmware versions for the SmartQ5 but they appear to be in a special format. Even with thier extractor I get .tar.gz files that wont extract.move on....:confused: