September 2004 - Darwin-kernel

Re: Further debugging kernel panics.
by Matt Jaffa 27 Sep '04

27 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Well I think I know what it is, but can't get the symbols to load. Any ideas? Matt On Sep 25, 2004, at 9:03 PM, Matt Jaffa wrote: Hi, Any help would be appreciated, Thanks, Matt This email sent to matt(a)knowligent.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.… When I load my kext, and do kextstat, the addresses begin there, so that is must be where it is crashing within my NKE, but I try to load my kext's symbol file generated by kextload -s, and do backtrace(bt) and still it shows the ??, I am getting this backtrace after loading the 10.3.5 symbol file, and loading in my NKE symbol file, but two lines are still undefined, any ideas on what could be causing the panic?? Is it references way out of space memory address, because there is a huge difference between the unknown and the known symbols. Sorry not very good at debugging. Don't know where to get more info on this, and what these 2 unknown lines could be. (gdb) backtrace #0 0x201d80a8 in ?? () #1 0x201d7f9c in ?? () #2 0x00116178 in ip_output (m0=0x1ce75a00, opt=0x130e004, ro=0xcdf3b00, flags=215956208, imo=0x13f12c8) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_output.c:959 #3 0x0011d284 in tcp_output (tp=0xcdf39e0) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_output.c:1053 #4 0x0011f48c in tcp_timers (tp=0x130e004, timer=32896) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_timer.c:510 #5 0x0011f098 in tcp_slowtimo () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_timer.c:279 #6 0x0022c160 in pfslowtimo (arg=0x2) at /SourceCache/xnu/xnu-517.7.21/bsd/kern/uipc_domain.c:434 #7 0x0003559c in _call_thread_continue () at /SourceCache/xnu/xnu-517.7.21/osfmk/kern/thread_call.c:1029 (gdb) _______________________________________________ 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/ matt%40knowligent.com This email sent to site_archiver(a)lists.apple.com

1 0

Further debugging kernel panics.
by Matt Jaffa 26 Sep '04

26 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, Any help would be appreciated, Thanks, Matt _______________________________________________ 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 am getting this backtrace after loading the 10.3.5 symbol file, and loading in my NKE symbol file, but two lines are still undefined, any ideas on what could be causing the panic?? Is it references way out of space memory address, because there is a huge difference between the unknown and the known symbols. Sorry not very good at debugging. Don't know where to get more info on this, and what these 2 unknown lines could be. (gdb) backtrace #0 0x201d80a8 in ?? () #1 0x201d7f9c in ?? () #2 0x00116178 in ip_output (m0=0x1ce75a00, opt=0x130e004, ro=0xcdf3b00, flags=215956208, imo=0x13f12c8) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_output.c:959 #3 0x0011d284 in tcp_output (tp=0xcdf39e0) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_output.c:1053 #4 0x0011f48c in tcp_timers (tp=0x130e004, timer=32896) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_timer.c:510 #5 0x0011f098 in tcp_slowtimo () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/tcp_timer.c:279 #6 0x0022c160 in pfslowtimo (arg=0x2) at /SourceCache/xnu/xnu-517.7.21/bsd/kern/uipc_domain.c:434 #7 0x0003559c in _call_thread_continue () at /SourceCache/xnu/xnu-517.7.21/osfmk/kern/thread_call.c:1029 (gdb) This email sent to site_archiver(a)lists.apple.com

1 0

Re: dlil_inject_if_input cause crashes
by Matt Jaffa 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Matt On Sep 22, 2004, at 3:21 PM, Matt Jaffa wrote: Matt On Sep 22, 2004, at 3:04 PM, Joshua Graessley wrote: Are you setting the receive interface (m_pkthdr.rcvif) correctly? -josh On Sep 22, 2004, at 12:27 PM, Matt Jaffa wrote: Hi, More on my recent problems with dlil_inject_if_input. Matt This email sent to matt(a)knowligent.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.… Got it, getting new results with netstat -s , and the computer isn't crashing, since I added the ifnet_ptr to the mbuf, m_pkthdr.rcvif. Thanks a bunch for that suggestion I was thinking that the rcvif was set by one of the functions in between dlil_inject_if_input and ip_input, but I guess I had to set it. I did not know that I was supposed to do that to the mbuf that I created. How would I go about that? I have my computer now doing a core dump to a remote server, and I did a back trace on the core dump. It is crashing here: (gdb) bt #0 0x0011076c in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:454 #1 0x00110634 in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:411 #2 0x0011115c in ipintr () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:967 #3 0x000ff6fc in run_netisr () at /SourceCache/xnu/xnu-517.7.21/bsd/net/netisr.c:79 #4 0x000f1d54 in dlil_input_thread_continue () at /SourceCache/xnu/xnu-517.7.21/bsd/net/dlil.c:568 (gdb) That is great that I know where it was at when it was crashing, but it seems to be crashing due to something in ip_input.c on line 454, I checked the line out, but don't know why it would crash at that line. Any help on figuring this out would be appreciated. See my previous posts with dlil_inject_if_input for my code. Is this a bug now? _______________________________________________ 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/ jgraessley%40apple.com This email sent to jgraessley(a)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/ matt%40knowligent.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: dlil_inject_if_input cause crashes
by Matt Jaffa 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com I never did this before, but I am trying this: testin->m_pkthdr.rcvif = temp->the_ifnet; See how that goes. Matt On Sep 22, 2004, at 3:04 PM, Joshua Graessley wrote: Are you setting the receive interface (m_pkthdr.rcvif) correctly? -josh On Sep 22, 2004, at 12:27 PM, Matt Jaffa wrote: Hi, More on my recent problems with dlil_inject_if_input. Matt _______________________________________________ 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.… where temp->the_ifnet is the ifnet ptr that is obtained from the outgoing filter for the dlil if filtering. I have my computer now doing a core dump to a remote server, and I did a back trace on the core dump. It is crashing here: (gdb) bt #0 0x0011076c in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:454 #1 0x00110634 in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:411 #2 0x0011115c in ipintr () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:967 #3 0x000ff6fc in run_netisr () at /SourceCache/xnu/xnu-517.7.21/bsd/net/netisr.c:79 #4 0x000f1d54 in dlil_input_thread_continue () at /SourceCache/xnu/xnu-517.7.21/bsd/net/dlil.c:568 (gdb) That is great that I know where it was at when it was crashing, but it seems to be crashing due to something in ip_input.c on line 454, I checked the line out, but don't know why it would crash at that line. Any help on figuring this out would be appreciated. See my previous posts with dlil_inject_if_input for my code. Is this a bug now? _______________________________________________ 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/ jgraessley%40apple.com This email sent to jgraessley(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: dlil_inject_if_input cause crashes
by Matt Jaffa 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Matt On Sep 22, 2004, at 3:04 PM, Joshua Graessley wrote: Are you setting the receive interface (m_pkthdr.rcvif) correctly? -josh On Sep 22, 2004, at 12:27 PM, Matt Jaffa wrote: Hi, More on my recent problems with dlil_inject_if_input. Matt _______________________________________________ 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 did not know that I was supposed to do that to the mbuf that I created. How would I go about that? I have my computer now doing a core dump to a remote server, and I did a back trace on the core dump. It is crashing here: (gdb) bt #0 0x0011076c in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:454 #1 0x00110634 in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:411 #2 0x0011115c in ipintr () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:967 #3 0x000ff6fc in run_netisr () at /SourceCache/xnu/xnu-517.7.21/bsd/net/netisr.c:79 #4 0x000f1d54 in dlil_input_thread_continue () at /SourceCache/xnu/xnu-517.7.21/bsd/net/dlil.c:568 (gdb) That is great that I know where it was at when it was crashing, but it seems to be crashing due to something in ip_input.c on line 454, I checked the line out, but don't know why it would crash at that line. Any help on figuring this out would be appreciated. See my previous posts with dlil_inject_if_input for my code. Is this a bug now? _______________________________________________ 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/ jgraessley%40apple.com This email sent to jgraessley(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: dlil_inject_if_input cause crashes
by Joshua Graessley 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Are you setting the receive interface (m_pkthdr.rcvif) correctly? -josh On Sep 22, 2004, at 12:27 PM, Matt Jaffa wrote: Hi, More on my recent problems with dlil_inject_if_input. Matt _______________________________________________ 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 I have my computer now doing a core dump to a remote server, and I did a back trace on the core dump. It is crashing here: (gdb) bt #0 0x0011076c in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:454 #1 0x00110634 in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:411 #2 0x0011115c in ipintr () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:967 #3 0x000ff6fc in run_netisr () at /SourceCache/xnu/xnu-517.7.21/bsd/net/netisr.c:79 #4 0x000f1d54 in dlil_input_thread_continue () at /SourceCache/xnu/xnu-517.7.21/bsd/net/dlil.c:568 (gdb) That is great that I know where it was at when it was crashing, but it seems to be crashing due to something in ip_input.c on line 454, I checked the line out, but don't know why it would crash at that line. Any help on figuring this out would be appreciated. See my previous posts with dlil_inject_if_input for my code. Is this a bug now? _______________________________________________ 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/jgraessley% 40apple.com This email sent to jgraessley(a)apple.com smime.p7s

1 0

dlil_inject_if_input cause crashes
by Matt Jaffa 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, More on my recent problems with dlil_inject_if_input. Matt _______________________________________________ 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 have my computer now doing a core dump to a remote server, and I did a back trace on the core dump. It is crashing here: (gdb) bt #0 0x0011076c in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:454 #1 0x00110634 in ip_input (m=0x1d571a00) at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:411 #2 0x0011115c in ipintr () at /SourceCache/xnu/xnu-517.7.21/bsd/netinet/ip_input.c:967 #3 0x000ff6fc in run_netisr () at /SourceCache/xnu/xnu-517.7.21/bsd/net/netisr.c:79 #4 0x000f1d54 in dlil_input_thread_continue () at /SourceCache/xnu/xnu-517.7.21/bsd/net/dlil.c:568 (gdb) That is great that I know where it was at when it was crashing, but it seems to be crashing due to something in ip_input.c on line 454, I checked the line out, but don't know why it would crash at that line. Any help on figuring this out would be appreciated. See my previous posts with dlil_inject_if_input for my code. Is this a bug now? This email sent to site_archiver(a)lists.apple.com

1 0

Re: debugging kernel paniced system
by Brian Bechtel 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Tue, 21 Sep 2004 22:58:09 -0600, Matt Jaffa <matt(a)knowligent.com> wrote: > I am trying this > > DTS Technote 2118 "Kernel Core Dumps" has just been published. > > > > <http://developer.apple.com/technotes/tn2004/tn2118.html> > > However it doesn't seem to work, > is it because I have a NKE and it crashes the system so the network > won't work then?? > Or maybe cause I am using the wireless, do I need to be wired in for > this to work?? You need to be wired for this to work. Neither airport driver supports kernel debugging. _______________________________________________ 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: debugging kernel paniced system
by Matt Jaffa 22 Sep '04

22 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, Quinn and the list, I am trying this DTS Technote 2118 "Kernel Core Dumps" has just been published. Thanks, Matt On Jul 27, 2004, at 6:04 AM, Quinn wrote: _______________________________________________ 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.… <http://developer.apple.com/technotes/tn2004/tn2118.html> However it doesn't seem to work, is it because I have a NKE and it crashes the system so the network won't work then?? Or maybe cause I am using the wireless, do I need to be wired in for this to work?? At 7:39 -0400 27/7/04, Andrew Gallatin wrote: Yes, that's how I found out about the crashdumps ;) Official documentation for this is coming soon. It's currently being reviewed. I expect to have it on the Apple developer web site next week. S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Disappearing Driver
by Godfrey van der Linden 21 Sep '04

21 Sep '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com G'day Brent, Godfrey On Sep 20, , at 21:28, Brent Marykuca wrote: Questions: Many thanks if you can shed any light on what's just happened here. Cheers, Brent Marykuca _______________________________________________ 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/gvdl%40apple.com This email sent to gvdl(a)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.… If a device is 'unplugged' (see later) then it is probably that your driver (kext) will be unloaded. This will automatically happen after at least 1 minute with no instances extant in the kext. I'd say that something like this happened, perhaps the system slept the bus or something. To answer question one is a bit more difficult. If you unplug and plug your device again does the driver turn up again (tty.???). If not it looks like something is happening to your kext or perhaps to the kernel state itself. 1. I thought kexts were kept in wired memory and remained loaded even if you deleted the executable. Am I wrong about this? I think my experience with previous versions of Mac OS gave me this impression. Why would kextd be running on my driver if it had been loaded for three days? 2. Where did my kext go? I certainly didn't touch it. Do the messages from the journal have any significance? This email sent to site_archiver(a)lists.apple.com

1 0