How to boost the OS X boot process...

[cparm]

No wonder cparm made this comment: "in exec_kernel(), it may be needed to disable LoadDrivers(”/”), for lion it will not work for now anyways".

 

LOL All your sources belong to ME

 

sorry, but the exact reason why i advised to disable this line is actually because i wasn't aware of this modification in load.c before that netkas pointed me to this modification in his own sources

[cparm]

I hope she comes back as the progress made here has been eye opening and good for the hackintosh scene.

 

i also think that he is doing a great work, but when he say that i have steal his work, it's not true,

the revision 636 is the last version of revolution that i have downloaded..

 

btw, his account is still valid, and he can come back if he want

[FKA]

There is also Applejack for similar purposes.

I'd installed it just in case I might need it.

 

I use applejack!

When I say I've cleared caches, I've run applejack in single user mode!

 

And Oh dear - here we go again!!

[JBraddock]

I use applejack!

When I say I've cleared caches, I've run applejack in single user mode!

 

And Oh dear - here we go again!!

I just wanted to add an alternative to what dutchhockeypro suggested as it might be useful for others. Not that I suggested you hadn't done something properly. Those exclamation marks forced me to make this explanation

My apologies.

[dgsga]

Though surely the possibility will always exist that more than one person will discover the same thing? somehow.

 

 

I think this is quite likely especially as the same issues will be encountered by people working towards the same goal i.e coding a working bootloader for 10.7. The truth is known between everyone involved.

[FKA]

I just wanted to add an alternative to what dutchhockeypro suggested as it might be useful for others. Not that I suggested you hadn't done something properly. Those exclamation marks forced me to make this explanation

My apologies.

 

No dude the "here we go again!!' comment was in reference to DHP departing the scene yet again!

Nothing at all to do with you suggestion.

 

Sorry for the confusion.

 

D

 

I think this is quite likely especially as the same issues will be encountered by people working towards the same goal i.e coding a working bootloader for 10.7. The truth is known between everyone involved.

 

Either way the subject could have been broached differently.

 

EDIT - round 2 and I still didn't manage to get the bugger to work ... 3rd time lucky?

[FKA]

I'm better at reading characters than code Dutch. Whatever the weather I like yours and I appreciate your work, else I wouldn't have spent hours following you (That sounds creepy, I mean 'so to speak' obviously.).

 

Please don't let this incident stop you. Whatever has or hasn't occurred I think you know your worth.

 

Take care

D

[cparm]

I think that I could be that kind of beginner, I'm not a dev at all, but don't understand why are you so sure that DHP and MC are the same? And if it is, what's wrong with that?

 

nobody can be 100% sure of anything, it's the magic of life, but more seriously I have access to certain tools and informations that you can not see, and it is an accumulation of evidence and the warning of some other members that cause I do not believe it

 

And if it is, what's wrong with that?

for me until he wrongly accused me, nothing (even if double account is forbidden, i had made this choice for the good of the community),

 

for me only, i'm not the only staff member at insanelymac ...

[cparm]

@cparm

In one of the earlier downloads, before SHE started using Release, is a path name to private_data.h.

That pathname is NOT a mans name .Unless HE is A SHE.

Thats why in my posts I have referred to DHP as SHE.

 

At the end of the day, we are all here for the same purpose, how about dropping it and let us get on with the job at hand.OR is there another reason for all this hoohaa.???

 

@DHP keep your chin up ***, we can beat this.

 

patch name is not enough to load the lion cache properly since the format of the cache also changed, you need to thin it before, otherwise you will get a Mach-o error

 

plus the cache patch of lion and snow is exactly the same ( /System/Library/Caches/com.apple.kext.caches/Startup/kernelcache) , for lion there is nothing to do to find the cache,

 

for snow ( and only snow) u need to add at the end '_i386(or _x86_64).THEADLER32VALUE' to find it, and i already admit that my adler32 patch is inspired from revolution, but it only work with snow, lion cache don't need this adler32 fix

[cparm]

Hey dude, DID you even READ what I put???.

Who's talking about cache path!!!

In this down load

http://www.insanelymac.com/forum/index.php...st&id=81771

 

is this

/*

* Copyright 2009 by Master Chief.

*

* Refactored (dynamic and static SMBIOS data gathering) by DHP in 2010.

*/

 

#ifndef __LIBSAIO_SMBIOS_STATIC_DATA_H

#define __LIBSAIO_SMBIOS_STATIC_DATA_H

 

#include "libsaio.h"

#include "essentials.h"

#include "/Users/***/Projects/private_data.h" // Remove me!

 

/* The following two values will be generated by smbios2struct */

#define SM_STRUCTURE_SIZE 191

#define DMI_STRUCTURE_COUNT 35

 

/* HP-SMBIOS.hex data (1614 bytes) converted with smbios2struct into little endian format. */

static uint32_t SMBIOS_Table[] =

{

PUT_YOUR_SMBIOS_DATA_HERE

};

 

#endif /* !__LIBSAIO_SMBIOS_STATIC_DATA_H */

 

 

READ the PATH NAME!!!

I think you owe *** a BIG apology.

 

 

sorry i misunderstood your post,

is this an absolute proof for you:

#include "/Users/***/Projects/private_data.h" // Remove me!

 

look at what we can do with a text file:

#include "/Users/MasterChief/Projects/private_data.h" // Remove me!

 

or

 

#include "/Users/ANYNAME/Projects/private_data.h" // Remove me!

 

sometime i just create fake user on my computer only because binary file also contain the patch where the project is stored, sometime i use the nickname jean, steve , bill .....

my current osx account is cparm, do you think cparm is my real name ??

 

edit: sorry, it seems that another team member have closed this topic while i was writing this post. and i will not re-open it .

[Alessandro17]

cparm,

 

I'm fairly unhappy with this comment. Wasn't going to reply to you, because you ignore the fact that I am a girl. Not a guy. Would you accept that from anyone? Be honest!

 

You cannot argue with a moderator, that is a basic rule.

If you are unhappy, contact an administrator.

 

GrootWitBaas, same with you.

 

Topic pruned.

[Alessandro17]

OK, I am reopening this topic.

However, if anything at all goes wrong, I'll close it forever and I might even delete it.

[scrax]

OK, I am reopening this topic.

However, if anything at all goes wrong, I'll close it forever and I might even delete it.

Thank's, good to see this topic open again.

 

SadGirl what about the svn options?

 

 

P.s I think that googlebot can't know your gender

[Time2Retire]

I am going to put all rubble behind me, and fast. Here are the first three steps for you gentlemen to do:

 

1.) Sign up for a github account and follow RevoGirl

2.) Visit gravatar.com and move your shiny avatar their place (so that github picks it up automatically).

3.) Purchase, download and install XCode 4.

 

Note: I will upload all my work when I figured out how all this github stuff works.

 

@cparm: All your source code, all your changes, belong to you. I never have and never will claim ownership (the saying 'all your source belong to me' was misunderstood) of it. And I have said this before, but you apparently missed it so here it is. Once again: Chapeau for your find. Keep up the good work.

 

Peace

[humph]

Am also glad that topic is open again, and not deleted

 

I don't know what went on behind the scenes, and will just say thank you to the Mods and those involved for reaching whatever agreements were needed to enable things to move on.

 

Especially happy this topic is still here as am now trying Revo on a P55/i5 rig with s/w RAID-0 and getting ACPI related KP, and as I still can, I will be going back over the earlier posts to get some insights on how to resolve this.

 

EDIT: Seems same problem as FKA. Even "ignoring" RAID (by using a single disk with OSX) does not change the main issue of ACPI related KP.

 

EDIT2: Also, along same lines as FKA, if I disable ACPI patch and just use MB's native ACPI tables, I get to boot - well sort of. Get a KP due to missing HPET and CPU PM kext issue (AFAIK standard DSDT needs HPET edit on this Gigabyte board). So, no boot irrespective of whether static or dynamic DSDT, boot (could be) OK if no ACPI patching...

[buoo]

Hi guys!

I compiled the Revolution for a test but I haven't founded boot0 and boot1h in the sym folder, doesn't the Revolution need these?

How can I create a bootable USB pendrive?

 

Thanks ^^

[scrax]

Am also glad that topic is open again, and not deleted

 

I don't know what went on behind the scenes, and will just say thank you to the Mods and those involved for reaching whatever agreements were needed to enable things to move on.

 

Especially happy this topic is still here as am now trying Revo on a P55/i5 rig with s/w RAID-0 and getting ACPI related KP, and as I still can, I will be going back over the earlier posts to get some insights on how to resolve this.

 

EDIT: Seems same problem as FKA. Even "ignoring" RAID (by using a single disk with OSX) does not change the main issue of ACPI related KP.

 

EDIT2: Also, along same lines as FKA, if I disable ACPI patch and just use MB's native ACPI tables, I get to boot - well sort of. Get a KP due to missing HPET and CPU PM kext issue (AFAIK standard DSDT needs HPET edit on this Gigabyte board). So, no boot irrespective of whether static or dynamic DSDT, boot (could be) OK if no ACPI patching...

What do you have booting with -x?

I tried 648 and with the last version of fakesmc in /Extra/Extrensions.mkext it hangs when loading the first plugin I also have in /E*/E*. With fakesmc in S*/L*/E* It boots but then I have this KP

I'll try to put some kext in my S*/L*/E* so to avoid big issue booting without DSDTand I'll try

 

NOTE: Booting with -x doesn't work either

[FKA]

Hi guys!

I compiled the Revolution for a test but I haven't founded boot0 and boot1h in the sym folder, doesn't the Revolution need these?

How can I create a bootable USB pendrive?

 

Thanks ^^

 

Hi buoo

 

I've been using boot0 and boot1h from chameleon trunk. I've mentioned this a couple of times here and nobody has corrected my so I'm assuning that's OK? - Although I have an hunch it may be the source of the issues I've had with Revolution?

 

D

[scrax]

Hi buoo

 

I've been using boot0 and boot1h from chameleon trunk. I've mentioned this a couple of times here and nobody has corrected my so I'm assuning that's OK? - Although I have an hunch it may be the source of the issues I've had with Revolution?

 

D

Yes guys, we are all using boot0 and boot1h from chameleon.

 

I've already registered and i'm following the project on github so i'll quote STLVNUB and make my reports there.

[FKA]

Yes guys, we are all using boot0 and boot1h from chameleon.

 

Good to know thanks

[humph]

Hi buoo

 

I've been using boot0 and boot1h from chameleon trunk. I've mentioned this a couple of times here and nobody has corrected my so I'm assuning that's OK? - Although I have an hunch it may be the source of the issues I've had with Revolution?

 

D

Hi FKA,

Thought we were using boot0 and boot1h from RC4 - based on info really early on in the thread - not RC5. Well I use RC4, and on other machine it works fine. Whether RC5 stuff can work, I don't know

 

As regards the boot issues on i5/P55 machine, did not try -x. The boot that fails at HPET is just using -v, since:

 

Am using fakesmc and other hardware specific kexts in /E/E, but also with mkext in /E, seems that gets loaded OK.

(mkext built with default options so as to include all kexts irrespective of OSBundle required status).

Have also the "system" mkext copied over from main /S/L/C/c/S etc into the local (Revo USB boot stick) /S/L/C/c/Startup. That also loads OK per debug output.

And of course a c.a.B.p in the USB's L/Preferences/SystemConfiguration/ folder. And standard mach-k on root of the USB stick - not yet trying prelinked kernel on the i5 machine, although had this working great on the netbook.

 

Was thinking to try netkas' HPET kext, but that's additional hassle, given that it would seem that the issue is related to something relating perhaps to format of Gigabyte DSDT and/or ACPI patching.

My DSDT's in /E/ACPI or loaded staticly into Revo at compile - makes no difference to outcome far as I can see.

For now, I'm gonna focus on basic single disk boot and not RAID boot, as I don't want to mess up RAID config as works fine with Chameleon and it's related stuff in the helper partitions.

[buoo]

Thanks guys

I'll try it!

I've another question: Are C & P States Generation supported by Revolution bootloader or shall I add SS code in my DSDT ?

[FKA]

Thanks guys

I'll try it!

I've another question: Are C & P States Generation supported by Revolution bootloader or shall I add SS code in my DSDT ?

 

you can add to DSDT but if you extract and convert to little endian, ACPI tables with blackosx and SVNUB's project revolution from a working chameleon install all you need will be there. Also same for any EFI data. For example EFi data is extracted from my chameleon install using Kabyl's bootloader for my ati5xxx GPU.

 

i.e no need to load ACPI tables from Extra/ACPI

 

D

[buoo]

you can add to DSDT but if you extract and convert to little endian, ACPI tables with blackosx and SVNUB's project revolution from a working chameleon install all you need will be there. Also same for any EFI data. For example EFi data is extracted from my chameleon install using Kabyl's bootloader for my ati5xxx GPU.

 

i.e no need to load ACPI tables from Extra/ACPI

 

D

 

 

Are you saying that if I want to put a patched DSDT in Extra folder I must convert it in little endian?

And add the Efi data for a GeForce 9600 from a generic Chameleon and add this one in the Revolution before to compile it?

If I understood this bootloader is the shortest possible and it has to be adjusted for what you need

 

Thanks

[FKA]

Are you saying that if I want to put a patched DSDT in Extra folder I must convert it in little endian?

 

no you can use the "#define LOAD_DSDT_TABLE_FROM_EXTRA_ACPI" option and load a standard DSDT.aml from Extra/ACPI but for me this seamed to miss some of the edits I'd made to my DSDT, for example a device id added to have AppleLPC load. So I used static ACPI data in little endian format.

 

And add the Efi data for a GeForce 9600 from a generic Chameleon and add this one in the Revolution before to compile it?

If I understood this bootloader is the shortest possible and it has to be adjusted for what you need

 

Thanks

 

Well a 9600 could have injection from either EFI or DSDT. For me to get my ati5770 recognised properly with DSDT (without suitable chameleon booter and graphicsenabler) I'd have to add a 4MB bin image to DSDT, which works! But I'm guessing that with Revolution my gpu is working because I have used projectRevolution on my chameleon install that has extracted EFI data courtesy of Kabyls code!

 

In short - (and no dis DHP /LG/ whatever) you currently need a working install to get Revolution (easily) up and running for your system. I'm sure with the way it's going with dynamic solutions, Revolution could detect more hardware, like GPU, automatically. But surely this would have to rely on work already done by other devs.

 

Which (without wanting to p!ss on anyone's fire) leads me to - what the hell was this argument about. Is this project (OSx86) not meant to be a collaborative (open source) effort? Everybody's work leading to the same goal?

Yes people should be credited accordingly for their work - but why the need for the ego trip/s -'it's my ball and you can't play with it!'

 

Surely credit/ kudos isn't really worth the bun fight.

 

Sure I'll get some {censored} for these comments but sod it, I've lost a bit of faith in this "community" recently!

 

D