May 2005 - Darwin-kernel - lists.apple.com

Re: Tiger: ipfw2 and sysctl variables
by Josh Graessley 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Please file a bug to request this support. Thanks, -josh On May 25, 2005, at 2:50 PM, Brian Warsing wrote: On Wed, May 25, 2005 at 11:24:34AM -0700, Vincent Lubet wrote: The explanation is simple: the man page is incorrect. Darwin does not support ipfw layer 2 filtering Okay. Maybe this is the wrong forum for this discussion? 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/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com A philosophical discussion on when and why Apple implements something is beyond the scope of this mailing list. Can you tell me why Apple would choose to include ipfw2 without compiling the kernel as to support one of its most basic features? -- B. _______________________________________________ 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 smime.p7s

1 0

Re: kpi socket filter: Safe to block in callbacks?
by Josh Graessley 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com That should be safe. -josh On May 25, 2005, at 2:38 PM, Brian Bergstrand wrote: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Is it safe to block on a R/W lock inside a socket filter callback? I'm specifically interested in sf_notify_func(). I doubt lock contention will be an issue for us at the point the notify callback happens, but it's always possible. Thanks. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.0.1 (Build 2130) 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/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com iQA/AwUBQpTwTXnR2Fu2x7aiEQLNgwCglmmXfGDO2+bs5Qs0n+GB5lN1ORUAn2zN 6Z4+Pzx1v65Y7efnWDgw8ShM =nVmj -----END PGP SIGNATURE----- _______________________________________________ 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 smime.p7s

1 0

Re: Tiger: ipfw2 and sysctl variables
by Brian Warsing 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com User-agent: Mutt/1.5.6i On Wed, May 25, 2005 at 11:24:34AM -0700, Vincent Lubet wrote: > The explanation is simple: the man page is incorrect. > > Darwin does not support ipfw layer 2 filtering Okay. Maybe this is the wrong forum for this discussion? Can you tell me why Apple would choose to include ipfw2 without compiling the kernel as to support one of its most basic features? -- B. _______________________________________________ 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

kpi socket filter: Safe to block in callbacks?
by Brian Bergstrand 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Is it safe to block on a R/W lock inside a socket filter callback? I'm specifically interested in sf_notify_func(). I doubt lock contention will be an issue for us at the point the notify callback happens, but it's always possible. Thanks. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.0.1 (Build 2130) iQA/AwUBQpTwTXnR2Fu2x7aiEQLNgwCglmmXfGDO2+bs5Qs0n+GB5lN1ORUAn2zN 6Z4+Pzx1v65Y7efnWDgw8ShM =nVmj -----END PGP SIGNATURE----- _______________________________________________ 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

async io (aio_read, aio_write) in 64-bit mode
by Steve Olson 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Does anyone know of any issues with using the aio routines (aio_read, aio_write, aio_error, aio_return) with -arch ppc64 on Tiger? I'm seeing some strange behavior that suggests what's in my write buffer is not what actually gets written to disk. When i read the buffer back, all I get is zeroes (all bits off). This works great for me in 32-bit mode on Tiger, but compiling in 64-bit mode seems to be problematic. Of course, it could be me, but I've been wrestling with this for several days, and thought maybe 2 or more heads are better than one (or none!!) Regards, 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

Re: mulckPatch_eieio
by Shawn Erickson 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On May 25, 2005, at 11:59 AM, Eric Long wrote: Hi, The panic log shows: panic(cpu 0): 0x300 - Data access Latest stack backtrace for cpu 0: Backtrace: 0x000836E4 0x00083BC8 0x0001EDA4 0x00090C60 0x0009406C Proceeding back via exception chain: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) If you want to decode the panic yourself review... <http://developer.apple.com/technotes/tn2002/tn2063.html> -Shawn _______________________________________________ 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.… Can anyone explain what mulckPatch_eieio is and what might cause a panic to occur within it? [snip] Unresolved kernel trap(cpu 0): 0x300 - Data access DAR=0x0000000000000011 PC=0x00000000002302EC Latest crash info for cpu 0: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Proceeding back via exception chain: Exception state (sv=0x22095C80) previously dumped as "Latest" state. skipping... Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) Thoughts? Best to list the version of the kernel involved when you supply a panic log. It would allow someone to possibly decode it better by knowing which symbols to load. Other then that I ... I don't know this area of things so I am out of here. This email sent to site_archiver(a)lists.apple.com

1 0

Re: mulckPatch_eieio
by Godfrey van der Linden 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Looks like a double panic to me. The first panic is significant not the second one. Godfrey On 05/25/2005, at 11:59 , Eric Long wrote: Hi, The backtrace shows this: #0 0x0008da08 in mulckPatch_eieio () #1 0x00017cec in ipc_port_release_sonce () #2 0x00236c50 in connect () #3 0x00246d84 in unix_syscall () #4 0x000941c0 in shandler () #5 0x000392f0 in zone_page_alloc () Cannot access memory at address 0xbfffd770 The panic log shows: panic(cpu 0): 0x300 - Data access Latest stack backtrace for cpu 0: Backtrace: 0x000836E4 0x00083BC8 0x0001EDA4 0x00090C60 0x0009406C Proceeding back via exception chain: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) Thoughts? Thanks in advance, Eric Long Allume Systems, Inc. http://www.allume.com 245 Westridge Drive, Watsonville, CA 95076 (831) 761-6200 x628 N 36° 54.768 W 121° 46.990 Matt. 11:28 _______________________________________________ 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.… Check out whatever function '0x002302EC' is in this kernel, as it is being passed in a NULL pointer. Can anyone explain what mulckPatch_eieio is and what might cause a panic to occur within it? We have a socket layer NKE that exhibits a problem related to it. On the first run of a particular third party application, providing it's never been run before on the machine, on 10.3.x systems, we see a panic. It's difficult to reproduce and only seems to occur in this one obscure instance. Our NKE intercepts the connect() and puts it on hold temporarily, It then allows the connect to proceed normally. But then things go boom. It never happens again with the app. We have never seen it happen anywhere else. Unresolved kernel trap(cpu 0): 0x300 - Data access DAR=0x0000000000000011 PC=0x00000000002302EC Latest crash info for cpu 0: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Proceeding back via exception chain: Exception state (sv=0x22095C80) previously dumped as "Latest" state. skipping... Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) ______________________________________________________________________ _____ Macintosh Software Engineer This email sent to site_archiver(a)lists.apple.com

1 0

mulckPatch_eieio
by Eric Long 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com User-agent: Microsoft-Entourage/11.1.0.040913 Hi, Can anyone explain what mulckPatch_eieio is and what might cause a panic to occur within it? We have a socket layer NKE that exhibits a problem related to it. On the first run of a particular third party application, providing it's never been run before on the machine, on 10.3.x systems, we see a panic. The backtrace shows this: #0 0x0008da08 in mulckPatch_eieio () #1 0x00017cec in ipc_port_release_sonce () #2 0x00236c50 in connect () #3 0x00246d84 in unix_syscall () #4 0x000941c0 in shandler () #5 0x000392f0 in zone_page_alloc () Cannot access memory at address 0xbfffd770 It's difficult to reproduce and only seems to occur in this one obscure instance. Our NKE intercepts the connect() and puts it on hold temporarily, It then allows the connect to proceed normally. But then things go boom. It never happens again with the app. We have never seen it happen anywhere else. The panic log shows: Unresolved kernel trap(cpu 0): 0x300 - Data access DAR=0x0000000000000011 PC=0x00000000002302EC Latest crash info for cpu 0: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Proceeding back via exception chain: Exception state (sv=0x22095C80) previously dumped as "Latest" state. skipping... Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) panic(cpu 0): 0x300 - Data access Latest stack backtrace for cpu 0: Backtrace: 0x000836E4 0x00083BC8 0x0001EDA4 0x00090C60 0x0009406C Proceeding back via exception chain: Exception state (sv=0x22095C80) PC=0x002302EC; MSR=0x00009030; DAR=0x00000011; DSISR=0x40000000; LR=0x0023030C; R1=0x11FEBE30; XCP=0x0000000C (0x300 - Data access) Backtrace: 0x0023030C 0x0023518C 0x002452B4 0x00094200 0x00090009 Exception state (sv=0x24A30780) PC=0x90031C6C; MSR=0x0200F030; DAR=0x0089A008; DSISR=0x42000000; LR=0x30844AFC; R1=0xBFFFD750; XCP=0x00000030 (0xC00 - System call) Thoughts? Thanks in advance, Eric Long ___________________________________________________________________________ Macintosh Software Engineer Allume Systems, Inc. http://www.allume.com 245 Westridge Drive, Watsonville, CA 95076 (831) 761-6200 x628 N 36° 54.768 W 121° 46.990 Matt. 11:28 _______________________________________________ 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: Tiger: ipfw2 and sysctl variables
by Vincent Lubet 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com The explanation is simple: the man page is incorrect. Darwin does not support ipfw layer 2 filtering Vincent On May 24, 2005, at 9:28 AM, Brian Warsing wrote: I know the man page is written for a FreeBSD, but why is this control and several other important ones, missing in the Darwin kernel? _______________________________________________ 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.… Reading the man page for ipfw, it suggests that in order for layer 2 filtering to occur, net.link.ether.ipfw must be enabled, but I don't see this variable is available. Can anyone explain? This email sent to site_archiver(a)lists.apple.com

1 0

kld(): Undefined symbols: sysctl
by Alexey Manannikov 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hello! kextload: sending 1 personality to the kernel kld(): Undefined symbols: _sysctl__hw_children _sysctl_register_oid _sysctl_unregister_oid kextload: kld_load_from_memory() failed for module What I need to make it work on Tiger? Alexey _______________________________________________ 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.… After upgrading to Tiger/Xcode 2.0 I receive this errors with loading my kext: This email sent to site_archiver(a)lists.apple.com

1 0