June 2007 - Darwin-kernel - lists.apple.com

Re: Kauth stability? (kernel panic)
by Greg 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Jun 21, 2007, at 8:09 AM, Terry Lambert wrote: Load the kgmacros and debug from another machine (showallstacks). (gdb) source kgmacros kgmacros:28: Error in sourced command file: No struct type named mach_header. - Greg _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… The most likely cause of this is a kernel stack overflow, either because you've declared some huge data object on the stack, or you accidently caused recursion. You can tell by looking at the "stacktop" "stackbottom", and the stack pointer at the time, to see if it's in the range or out of it. -- Terry I'm having difficulty using the kgmacros script, I searched google for for it and found a google-cached version of it from xnu-792.6.61. Whenever I try to load it in gdb I get this error: I tried doing that after I added symbols from the various kpis that my KEXT uses and I still get that error. On the other hand I think I've determined that the problem is elsewhere (not in the kauth stuff), either it's some sort of threading issue or it's the kernel control stuff that I'm using. This email sent to site_archiver(a)lists.apple.com

1 0

routing socket issue.
by Andreas Fink 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: CGDisplayBaseAddress at kernel?
by Terry Lambert 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com The short answer you are going to get is "no; never". -- Terry On 6/21/07, Michael Smith < drivers(a)mu.org> wrote: On Jun 21, 2007, at 8:35 AM, rohit dhamija wrote: > Just an update to my previous post > Can we have access to these api's at kernel level ?? I require > CGGetActiveDisplayList and CGDisplayBaseAddress in specific. > > Actually i am accessing accessing frambuffer from my user > application. Now i need to do the same from my kernel extension. I think you are making a number of flawed architectural assumptions here, but at this point you have very clearly passed beyond the level of help you can expect from this list. You should take this matter up with DTS directly. I would encourage you to follow their advice. _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… The longer answer you will get is "whatever you are trying to do should be done in user space; you are trying to do something in the kernel that does not belong in the kernel". This will likely involve one of several kernel/user IPC mechanisms, all of which are documented with sample code at <http://developer.apple.com/>. The people on this list capable of discussing them will likely only be willing to discuss them if you have read the documentation already, and need clarification on minor issues, after having written your code. On Jun 21, 2007, at 10:52 AM, rohit dhamija wrote: Apple people charge money for this service, it would be great if I can get any hints before taking that route. Rohit = Mike This email sent to site_archiver(a)lists.apple.com

1 0

Re: CGDisplayBaseAddress at kernel?
by rohit dhamija 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=AG200/ouOPTCxAy5ooPgDtPiBKk3cyAXuAaGrckWCrl1IX4rFh92XhFZ1mNrlHUAMGXWqUmwWkdqOzXU1+01mD2du0tuTeBc11xXKKslpmuGcAHt818kjeVQ1qPlhjfFKTDJ+kmZq1vQDNwUBEnYsdTrF1wtOWnyt/rs5zMGD0M= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=Aqc9d4TYLpT2RT4k7AErASFBbw1gKqlhNXCMSZMz5gLhL7BzHy4x4jLNF5BymRXzCrwyTxzY3SUGQXuwd/YO8DXJFf7vN7uTQaQvgu0cEsmFf9AvnpM0+B64GsTfagawHqJtkiRKqTEWrE7/JaFahzUsNYXuefQ8ufk80Mi9jfw= _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: CGDisplayBaseAddress at kernel?
by Michael Smith 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Jun 21, 2007, at 8:35 AM, rohit dhamija wrote: = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Just an update to my previous post Can we have access to these api's at kernel level ?? I require CGGetActiveDisplayList and CGDisplayBaseAddress in specific. Actually i am accessing accessing frambuffer from my user application. Now i need to do the same from my kernel extension. I think you are making a number of flawed architectural assumptions here, but at this point you have very clearly passed beyond the level of help you can expect from this list. You should take this matter up with DTS directly. I would encourage you to follow their advice. This email sent to site_archiver(a)lists.apple.com

1 0

Re: CGDisplayBaseAddress at kernel?
by rohit dhamija 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=XL+2C6hrkQXIUelwQjsw/kRS9/f9y/5zwgeIpgqv5d/SLotGIpV701FiTpnlorndE9gI6DQiw2XKjm3rBr1TevfP3A2z98H+XbfwsDcTW8jCO9Wn4yMmYTIdJupof3zdDgyca4DcOrReEiXfzzP+FofTB2WvA6gkc+vFBvkHfnU= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=dyNtyVlOC4azDd4RdQzRw28NLO5K6uRF1QHYqF8HVYxPIRZx5TopvNd7vJy7xUL9Qtcg7IkqCMtIHvkqi/giKd0EoTd5gycO/J6/rPdFIgPI7h3jtUexcVmW08aAaAjHoAWUuuljIQwAbWRmBZuY4/YL0hTexooB4gMUeo5X14g= _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Kauth stability? (kernel panic)
by Terry Lambert 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Load the kgmacros and debug from another machine (showallstacks). -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Jun 20, 2007, at 9:51 PM, Greg wrote: I'm experiencing a strange kernel panic and I can't figure out why it's happening. This is what happened: I have my vnode listener set to deny access to a certain file. I created a shell script that attempted to copy this file to another location many times (each time it would get a "permission denied" error because the listener was set to reject it). After running this script several times rather quickly, I then clicked on another terminal window and tried to launch 'top' to see how much CPU the kernel was taking up. The second I attempted to run top the system kernel panicked and on my development machine (which was attached during this entire time) just gave me this backtrace in gdb: Program received signal SIGTRAP, Trace/breakpoint trap. 0x002ff4ac in pe_run_clock_test () (gdb) bt #0 0x002ff4ac in pe_run_clock_test () #1 0x002ff498 in pe_run_clock_test () Previous frame identical to this frame (gdb could not unwind past this frame) Any clues as to why this is happening? I'm not really doing much in the vnode listener right now and generally that code runs fine, just sometimes in intense situations like this it kernel panics. Is this my fault or could there be a stability issue in 10.4.9's kauth stuff (on PPC)? The most likely cause of this is a kernel stack overflow, either because you've declared some huge data object on the stack, or you accidently caused recursion. You can tell by looking at the "stacktop" "stackbottom", and the stack pointer at the time, to see if it's in the range or out of it. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Build Procedure for x86
by Joel Beaudoin 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:mime-version:in-reply-to:references:content-type:message-id:content-transfer-encoding:from:subject:date:to:x-mailer; b=JmKXVoseiNBLbkZUmlvE8MTVjuGxBkSOHAPWCidfD+G0N2TqGbJCkUTrQJCa5Ufr5xZA6kyCXEDRTcvHpE9Ku44QMiT6YGdqimbQflZY/HKsJ0pYf6HFyLWvjjMDcB/u3B4nO3BqhbD4wFPNKkmiPFM7fWt1ALQProW/T/eBALw= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:mime-version:in-reply-to:references:content-type:message-id:content-transfer-encoding:from:subject:date:to:x-mailer; b=e+vYe7GDnxCtOqEahuba1WYNYpI5FxIhfxO0+kPkg0aUyJgROfVv0n1qhHiZcunS9MUYk5Abj5l4e4HXAkmDihAysgbGAHSFzxSKm2GOCWaK7HR3bXdPu9gLuueFaaDqzGi3K7nFFIWDnIgHnhmfjo7IFKSSt3o/eXP1NqoMyHA= cc -I ../cctools-622.5/include kextsymboltool.c -o kextsymboltool Joel On Jun 20, 2007, at 10:55 PM, Joel Beaudoin wrote: Hi, _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Well, with the exception of needing to add an include path when compiling kextsymboltool, the procedure did produce a working kernel: I am trying to build an x86 kernel following the procedure documented on ADC. The first problem that I ran into was the lack of gcc 3.3 for x86. A search indicated that 4.x is used for x86 builds. Does an x86 build of the kernel work with 4.x? Is there any documentation for x86 build of the kernel or does the ADC info just work using gcc 4.0? Joel This email sent to site_archiver(a)lists.apple.com

1 0

Re: Kauth stability? (kernel panic)
by Michael Smith 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Jun 20, 2007, at 9:51 PM, Greg wrote: = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Any clues as to why this is happening? I'm not really doing much in the vnode listener right now and generally that code runs fine, just sometimes in intense situations like this it kernel panics. Is this my fault or could there be a stability issue in 10.4.9's kauth stuff (on PPC)? The default file permissions handling in MacOS 10.4.x is implemented using kauth, so the basic robustness of the infrastructure, when used correctly, seems reasonable. For all that, it's possible that you have found a problem; it would help immensely if you would file a bug (http://bugreporter.apple.com) attaching a copy of your kauth listener, test script and instructions for reproducing the panic. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Getting vnode from filename?
by Michael Smith 21 Jun '07

21 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Jun 20, 2007, at 8:09 PM, Greg wrote: In what context? You can find many examples of its use throughout the xnu sources. = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Is there a way to get a vnode_t object from a filename? There seems to be zero documentation on this, and the header file (vnode.h) is not very helpful. There is an interesting sounding function called "vnode_lookup" that I found in there, but I can't find any documentation on how to use it. The reason I want to do this is because I'd like to deal with vnodes instead of filenames so that I can handle symbolic links to the file as well. I'm not sure how one would help you with the other. In what context are you trying to do this? What do you mean by "handle symbolic links"? This email sent to site_archiver(a)lists.apple.com

1 0