February 2006 - Darwin-kernel

Re: The mistery of the dlil_filter_detach call
by Vincent Lubet 21 Feb '06

21 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Feb 21, 2006, at 11:00 AM, Stephane Sudre wrote: Vincent _______________________________________________ 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.… My question is more related to the fact that, for any 10.4.x version, it seems that the path of the backtrace makes no sense after the address of dlil_detach_filter. That is if you look at the source code of dlil_detach_filter. My point was the backtrace will most often not make sense if you don't use the correct version of the kernel symbol file. This email sent to site_archiver(a)lists.apple.com

1 0

Re: The mistery of the dlil_filter_detach call
by Stephane Sudre 21 Feb '06

21 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On 21 févr. 06, at 18:30, Vincent Lubet wrote: Stéphane, Vincent On Feb 21, 2006, at 8:12 AM, Stephane wrote: Let's say we have the following KP log: The first two calls are: Call_continuation and dlil_detach_filter and then it goes into ip_input, etc... Problems: - there's been no interface filter detached as far as I know. Funny part: Question: _______________________________________________ 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.… Actually, my question is not related to this particular kernel panic. It's "ok" it's kernel-panicing since this is a project in dedvelopment. My question is more related to the fact that, for any 10.4.x version, it seems that the path of the backtrace makes no sense after the address of dlil_detach_filter. That is if you look at the source code of dlil_detach_filter. xnu-790.1 is a pre-released version of Tiger (may be a seed) for which we don't have a symbol version of the mach_kernel. Please try to reproduce on an official release version of Tiger (xnu version 792 or higher) and we'll be able to help. I'm beginning to wonder if I'm not missing one point with the panic logs. Tue Feb 21 16:51:12 2006 panic(cpu 0 caller 0x002917A8): sbappendaddr Latest stack backtrace for cpu 0: Backtrace: 0x00095564 0x00095A7C 0x00026838 0x002917A8 0x00291994 0x001614D0 0x0014E6CC 0x0014F3A8 0x00139FFC 0x0013A110 0x0011C2A8 0x000A9654 Proceeding back via exception chain: Exception state (sv=0x44553280) PC=0x00000000; MSR=0x0000D030; DAR=0x00000000; DSISR=0x00000000; LR=0x00000000; R1=0x00000000; XCP=0x00000000 (Unknown) Kernel version: Darwin Kernel Version 8.0.0: Thu Mar 24 19:14:36 PST 2005; root:xnu-790.1.obj~1/RELEASE_PPC - I don't see ifmaddr_ifnet being called in the source code of dlil_detach_filter. This Kernel Panic happens with a kext in its early stage of development and when double-clicking on a .pkg (Installer.app is launched and then bang!) Is it possible to trust a KP log on 10.4.x? I don't see any reason why dlil_detach_filter would be called (no interface filter is detached). So why can I see it in the log? The trace does not look perfectly correct as usual with the KP I see involving dlil_detach_filter. This email sent to site_archiver(a)lists.apple.com

1 0

Re: The mistery of the dlil_filter_detach call
by Vincent Lubet 21 Feb '06

21 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Stéphane, Vincent On Feb 21, 2006, at 8:12 AM, Stephane wrote: Let's say we have the following KP log: The first two calls are: Call_continuation and dlil_detach_filter and then it goes into ip_input, etc... Problems: - there's been no interface filter detached as far as I know. Funny part: Question: _______________________________________________ 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.… xnu-790.1 is a pre-released version of Tiger (may be a seed) for which we don't have a symbol version of the mach_kernel. Please try to reproduce on an official release version of Tiger (xnu version 792 or higher) and we'll be able to help. I'm beginning to wonder if I'm not missing one point with the panic logs. Tue Feb 21 16:51:12 2006 panic(cpu 0 caller 0x002917A8): sbappendaddr Latest stack backtrace for cpu 0: Backtrace: 0x00095564 0x00095A7C 0x00026838 0x002917A8 0x00291994 0x001614D0 0x0014E6CC 0x0014F3A8 0x00139FFC 0x0013A110 0x0011C2A8 0x000A9654 Proceeding back via exception chain: Exception state (sv=0x44553280) PC=0x00000000; MSR=0x0000D030; DAR=0x00000000; DSISR=0x00000000; LR=0x00000000; R1=0x00000000; XCP=0x00000000 (Unknown) Kernel version: Darwin Kernel Version 8.0.0: Thu Mar 24 19:14:36 PST 2005; root:xnu-790.1.obj~1/RELEASE_PPC - I don't see ifmaddr_ifnet being called in the source code of dlil_detach_filter. This Kernel Panic happens with a kext in its early stage of development and when double-clicking on a .pkg (Installer.app is launched and then bang!) Is it possible to trust a KP log on 10.4.x? I don't see any reason why dlil_detach_filter would be called (no interface filter is detached). So why can I see it in the log? The trace does not look perfectly correct as usual with the KP I see involving dlil_detach_filter. This email sent to site_archiver(a)lists.apple.com

1 0

The mistery of the dlil_filter_detach call
by Stephane 21 Feb '06

21 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Let's say we have the following KP log: The first two calls are: Call_continuation and dlil_detach_filter and then it goes into ip_input, etc... Problems: - there's been no interface filter detached as far as I know. Funny part: Question: _______________________________________________ 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.… I'm beginning to wonder if I'm not missing one point with the panic logs. Tue Feb 21 16:51:12 2006 panic(cpu 0 caller 0x002917A8): sbappendaddr Latest stack backtrace for cpu 0: Backtrace: 0x00095564 0x00095A7C 0x00026838 0x002917A8 0x00291994 0x001614D0 0x0014E6CC 0x0014F3A8 0x00139FFC 0x0013A110 0x0011C2A8 0x000A9654 Proceeding back via exception chain: Exception state (sv=0x44553280) PC=0x00000000; MSR=0x0000D030; DAR=0x00000000; DSISR=0x00000000; LR=0x00000000; R1=0x00000000; XCP=0x00000000 (Unknown) Kernel version: Darwin Kernel Version 8.0.0: Thu Mar 24 19:14:36 PST 2005; root:xnu-790.1.obj~1/RELEASE_PPC - I don't see ifmaddr_ifnet being called in the source code of dlil_detach_filter. This Kernel Panic happens with a kext in its early stage of development and when double-clicking on a .pkg (Installer.app is launched and then bang!) Is it possible to trust a KP log on 10.4.x? I don't see any reason why dlil_detach_filter would be called (no interface filter is detached). So why can I see it in the log? The trace does not look perfectly correct as usual with the KP I see involving dlil_detach_filter. This email sent to site_archiver(a)lists.apple.com

1 0

Re: DPC's with MacOS X
by Mike Smith 18 Feb '06

18 Feb '06

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

DPC's with MacOS X
by Brad Post 17 Feb '06

17 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Thread-index: AcY0FYRd3+Gdiy8wTDmcIG5sfaXfeA== Thread-topic: DPC's with MacOS X _______________________________________________ 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: ucontext
by Steve Dekorte 16 Feb '06

16 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Am 16.02.2006 um 11:50 schrieb Steve Dekorte: This is a obsolete question for two reasons: 1) Apple never talks about future plans. -- Steve _______________________________________________ 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 16-Feb-06, at AM 03:37, Markus Hitter wrote: Are there any plans to support ucontext on OSX? 2) If the sole purpose of your programming task is making use of "ucontext" (whatever this is), then you better choose a platform which supports it. Probably you have some higher goal in mind, so asking how to solve this might get you there more quickly. The purpose is coroutines. I'm currently use setjmp for this, but it's an ugly solution. It looks like all other major platforms provide ucontext/fiber APIs for this except osx and openbsd. This email sent to site_archiver(a)lists.apple.com

1 0

Re: ucontext
by Jonas Maebe 16 Feb '06

16 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On 16 feb 2006, at 11:50, Steve Dekorte wrote: Jonas _______________________________________________ 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.… Are there any plans to support ucontext on OSX? There's at least already an ucontext available for signal handlers, but I'm not sure about which ucontext you are talking. This email sent to site_archiver(a)lists.apple.com

1 0

Re: ucontext
by Markus Hitter 16 Feb '06

16 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Am 16.02.2006 um 11:50 schrieb Steve Dekorte: This is a obsolete question for two reasons: 1) Apple never talks about future plans. Markus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ _______________________________________________ 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.… Are there any plans to support ucontext on OSX? 2) If the sole purpose of your programming task is making use of "ucontext" (whatever this is), then you better choose a platform which supports it. Probably you have some higher goal in mind, so asking how to solve this might get you there more quickly. This email sent to site_archiver(a)lists.apple.com

1 0

ucontext
by Steve Dekorte 16 Feb '06

16 Feb '06

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Are there any plans to support ucontext on OSX? -- Steve _______________________________________________ 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