http://xdaforums.com/showthread.php?t=1519448&page=21:D
once we get it ported to the D3 helloooo camcorder and adobe flash!
once we get it ported to the D3 helloooo camcorder and adobe flash!
kexec bootloader bypass coming in a few days
- Thread starter ssinisterss
- Start date
-
- Tags
- development
this loads a seperate kernel once the stock locked one loads so we can run a custom kernel for our roms.
If this works for us, im kicking Bar Refaeli out of my house. This would be better than anything she could provide.
Well, we won't necessarily have it in a few days. It will take a lot of work to port it to our devices because of hardware differences, and we would need someone very good at coding these types of things.
it wont be as hard as it seems, we have the same internals as the razr minus the 1gb ram. i dont know if the kexec is the same (it should be for the omap 4430) but the only coding that would be different is getting the kexec to take the modded kernel. once its released for the razr within the week im sure hash is gonna have some fun with this.
Last edited:
There was some discussion of this over at the Razr forum in that thread. The differences between the GSM and CDMA versions of the Razr are alone enough to cause major pains in porting it between the two devices, and it was originally developed for the GSM version.
i would think that since its a nice sized topic in our razr forums its gonna work for CDMA too. the dev of the work around won a developer contest and recieved a CDMA razr along with kejar.
Pardon my ignorance on the topic, but this allows us to use another kernel on top of the stock one, but not replace it right? Is the only downside a longer boot time? Also will this be relatively easy to port due to similar hardware or will that another challenge in itself.
so far the longer boot time is the only downside but its not a bad thing considering we reboot only to flash roms or refresh the phone a couple times a week and since it will let us have a fully working rom. this lets the phone boot with the gingerbread kernel behind the bootloader, then after the M splash screen it will boot into the ICS or modified GB kernel that is loaded.
we would have to have hashcode or anyone that has any kind of knowledge of how this really works tell us the time frame. as it stands we dont know anything other than it works on GSM razrs, and the dev has a CDMA razr he is porting it to and its days from release. we need to see the details on how to approach porting this to the D3/D4/bionic. it may not be that hard since its designed to do only 1 thing and hardware is mostly the same.
OP you are being way too optimistic in this thread. Chances of this being ported for D3 is very low as aboves has stated. It won't be anywhere as easy as changing the string "RAZR" to "Droid 3" in the kernels
like i said we have a few guys in here that have experience with coding. if it could be done with the cdma razr it can be done for the rest of them.
It's taking quite a long time to be ported from the GSM Razr to the CDMA one, though. There aren't many hardware differences between those two devices, but for those differences it's taking a large amount of time. Now imagine how long it will take, and how complicated it will be to port, with all of the hardware differences between the Droid 3 and the Razr (and not to mention the different models of the D3; I assume a kexec module for the XT862 would not work on the XT860 or XT883 without changes). You are being a bit too optimistic here, especially with your thread title. It's less a bootloader bypass and more a "we're just gonna load our own kernel after the bootloader loaded the default one, but we still can't change anything about the boot process."
I hate to be the bearer of bad news and the one to shoot down everyone's hopes and dreams, but at the same time I don't want people to get too worked up over this and then get incredibly disappointed when it doesn't come or takes ungodly forever for it to come to use.
Well, you went to the other extreme and assumed it's impossible to port and that even if it could be done it would suck or something like that.
I'd like to take a more neutral stance on this and
a) be exited that this approach is being taken on locked devices.
b) be exited that a relatively similar device to the D3 is getting this kind of attention. because it could make things easier for D3 devs to port it to this device.
c) wait patiently, with no ETA expectations and offer to help with whatever I can to the devs.
I think it's much better than waiting for someone to crack moto's keys.
It's at least a chance.
Hope devs can ( and want to ) do something with this : )
I'd like to take a more neutral stance on this and
a) be exited that this approach is being taken on locked devices.
b) be exited that a relatively similar device to the D3 is getting this kind of attention. because it could make things easier for D3 devs to port it to this device.
c) wait patiently, with no ETA expectations and offer to help with whatever I can to the devs.
I think it's much better than waiting for someone to crack moto's keys.
It's at least a chance.
Hope devs can ( and want to ) do something with this : )
Similar threads
Top Liked Posts
-
There are no posts matching your filters.
-
24Right now the new kernel has a display issue that both me and jonpry are working through.
Situation is like this:
- We now has a new 3.0 kernel. So it's old PowerVR drivers are out of date and can't be used.
- Can't use any of the PowerVR drivers on RAZR, Droid 4 or Bionic due to memory allocation changes in the kernel.
- Texas Instruments provides "stock" PowerVR drivers as binaries and kernel module source for developers working on their "stock" 3.0 kernel.
Solution:
- We're altering the Motorola 3.0 kernel to be more like the stock TI kernel so that these binaries will work.
Takes time. :/
And obviously w/ JB out, I'm splitting duties.
Thanks for being patient!22From my end, I have a very set focus at the moment.
#1 for me is getting the new kernel merged up to a point where it can use all of the goodies I have for it.
This is a fairly long process starting with a grueling merge of code that I'm on day #4 of. Then, I get to debug the thing for days/weeks until most the kernels are resolved,. This includes turning back on the rpmsg system and debugging memory allocation issues that we had previously.
Only after that, do we finally get to see how effective this new ducati runs on the D3, and if it will work with our camera hardware, etc.
So sound issues aside, I still have several days/weeks before I can see a new bootup of the kernel in what I consider a really usable state.
#2 Once the new kernel is functional, I can choose 1 of 2 paths:
A) I could work on kexec binaries to use the new kernel on the other phones. This would be needed before we could test it on the RAZR and compare audio notes.
B) I could continue to troubleshoot the D3's audio w/o a reference point like the RAZR and see where it goes.
I'm not sure I would call that "back burner"
Sent from my XT894 using Tapatalk 219Quick status update:
- I have successfully tested video decoding using the new ducati on the Kindle Fire (doesn't have a camera)
- And I'm now working on the new Moto kernel bringing it up-to-date so it can use the same drivers and ducati as the KFire, and once done it will be easier to update straight from TI. The new kernel version is 4AI.1.6 (3.0.31 mainline) for those that are curious.
Reading this thread and thinking about it some more, I will continue to do ICS builds as it's ready for the new codecs w/o any additional changes (once the kernel is done). Be nice to see what's going work and what isn't.
Once ICS is running well, I'll start on JB. Which will need audio changes and a bunch of optimizations.
Thoughts?
Sent from my XT894 using Tapatalk 214Starting the "big merge" to add the functionality to this kernel that the D3 would need to fully operate under ICS:
- Stuff like updated PowerVR binaries
- Update WLAN drivers and firmware
- Handling for 512MB codecs (may only be HD codecs not camera-- testing will tell)
- Etc
You can see the new kernel developing here:
http://gitorious.org/droid3-kexec/droid3-kexec-kernel/commits/3.0.8-dev
Right now I'm debugging the rpmsg sub-system which loads the HD codec binary (ducati).
