Talk:Kernel

XNU?

Isn't it XNU Kernel, not Darwin? Darwin was the name of the Operating System the last I knew of. --Jacob 22:57, 3 September 2011 (MDT)

Yep. Fixed. beej 15:51, 7 September 2011 (MDT)

ASLR

Because of Apple's implementation of he ASLR in 4.3, there is no chance the kernel is still vulnerable to Ramdisk Hack. Itaiyz97 12:17, 4 December 2011 (MST)

Mod

Added the kextstat listing, fixed obsolete mention of 0xC00... (now 0x8000.. instead) --Morpheus 07:27, 10 February 2012 (MST)

Boot-arg address

I'm looking at the iPod 4G Kernel in IDA. Isn't the address for PE_parse_boot_argn at 0x80241314 --Haifisch 21:47, 5 July 2012 (MDT)

XNU Versions

I find it weird how Apple lists the XNU Version for OS X, but not iOS over at their Open Source Listing page. AFAIK, that isn't legal. --5urd 19:01, 17 November 2012 (MST)

Version List

AFAIK, the kernel version used in the iPad 4 (3G) and iPad mini (3G) use a different version **number** (and build date), although I may be wrong. Because of this, I think we should split it up by device. If we can't do that, I would like to do something about the variations in application processor. Maybe splitting up by revisions (S5L8720, S5L8920, etc.) or replacing it with something like S5L89xx (S5L8xxx for ones that the n72ap use). I personally don't like the second option as the x's aren't really in the build string. For the first option, maybe something along the lines of:

Version	Build	S5L8900	S5L8720
2.0	5A345	herp derp 3G	N/A
	5A347	herp derp	herp derp

(obviously split up by major releases). This might work because, for example, with S5L890, that's 3 devices put into one column, and 5 devices for A4 in one, as opposed to VFDecrypt Keys where each device has it own column. Any thoughts? --5urd 15:53, 22 November 2012 (MST)

This is the Kernel page. I don't see any reason to introduce device-specific lists here unless absolutely necessary. Even if some identifier string is different, that's not reason enough. --http 06:15, 23 November 2012 (MST)

I am not necessarly against device specific lists, but the version list is far from being 'complete' yet. So I think this shall be done first. --M2m 19:32, 23 November 2012 (MST)

Finding the version of kernel

How do you find the kernel version? --adaminsull (talk) 11:49, 2 February 2013 (UTC)

Either "uname -a" in terminal of a JB'ed device. Or as part of the Diagnostics & Usage Data (Latest Crash, LowMemory, etc).--M2m (talk) 12:11, 2 February 2013 (UTC)

Ah ok thanks. --adaminsull (talk) 13:22, 2 February 2013 (UTC)

boot-args

The boot-args list is quite long. I think that should go to a separate page (with "Main-Article: ..." here). I found this page: Boot-args (iBoot variable). Is that the same? If yes, then it belongs there and we should link that article to here. --http (talk) 23:10, 8 February 2013 (UTC)

RELEASE_ARM

I was wondering, could we remove the RELEASE_ARM as it does vary by processor and can be vey confusing? If so I am willing to remove it. --adaminsull (talk) 17:25, 14 February 2013 (UTC)

I think the complete string as it appears should be listed. We don't need a link on the processor though. --http (talk) 00:02, 15 February 2013 (UTC)

So do you mean delete processor but keep RELEASE_ARM at the end? --adaminsull (talk) 00:17, 15 February 2013 (UTC)

No, just the link, not any text. --http (talk) 01:20, 15 February 2013 (UTC)

What do you mean? Example please. --adaminsull (talk) 09:57, 15 February 2013 (UTC)

I see. Is the processor really needed though as it just makes it harder to detect? --adaminsull (talk) 23:59, 18 February 2013 (UTC)

I believe an earlier discussion covered this somewhat. I'm against removing it as then the string is not the exact string that would be produced by the device. For someone writing a tool to parse the strings, it should be pretty obvious that the RELEASE_ARM part varies between processors. --5urd (talk) 00:10, 19 February 2013 (UTC)

Undo revision argument

I found the address at 0x80241314 actually --Haifisch (talk) 02:33, 5 April 2013 (UTC)