September 2009 - Darwin-kernel

Re: Problem Calling namei
by William Kucharski 30 Sep '09

30 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com User-agent: Thunderbird 2.0.0.23 (X11/20090910) Look here for the xnu used for Mac OS X 10.6.1: <http://www.opensource.apple.com/source/xnu/xnu-1456.1.26/> -- William Kucharski _______________________________________________ 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: Problem Calling namei
by Terry Lambert 30 Sep '09

30 Sep '09

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: Problem Calling namei
by mominabrar 30 Sep '09

30 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:cc:content-type; bh=uiHz1UEYY93ZGsDibWOmxUVJ2+IVG1rh3qDuqTyxVt0=; b=NnadkxQbHMKpqoRBC6LO8C1oVvyf5h84y5ux8XYZzOHRWnQE9oEYBzjQsEMhcZqpnk 1QXnJbDfPQpNhJ6DykiRG+6ny4s32MUmeKoVVJPiMSvfywkGH68Chz6oExMIPlce6w3I C6PYOaLHv33bLS3Q0M3Rgoy1Lxd98jNPKJzDA= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; b=xxPirbn43ULk0p806Xeur6wZLZrZFCb37giaLHpcZIbUHYyJC0no/0633ZcATZ+46m 9U+3CfzDIAg0YwhnbiLkcsuSDtl+MNHNf1AZ4ZyU9Z6Kj/jGBZF3fkC/DdXCfanp+ZT3 +9rqVOxyye3cB7GpYjjK17c7DBGSEmIPiEk74= _______________________________________________ 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: Kernel Panic in IOMemoryDescriptor
by Michael Smith 30 Sep '09

30 Sep '09

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: Kernel Panic in IOMemoryDescriptor
by Brian Bechtel 30 Sep '09

30 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:cc:content-type :content-transfer-encoding; bh=+qf3GJwO/teC70kx4OheiXQ6VV8emD77UDB9PY8cbPI=; b=OIb4eHgfaELmixucSZfuYcbk7J0VWqPZAkQpZFVt2PJhX+iqilND9qBXregpgjxY+2 Cw0UYuCW21xQzDQxPd70CKwMOm4IOJ9yldeN6Pp30ilnLKN+TYKdruORqnfFWkNbE8X8 lVFJoG8dXCDjugjJcCNo0RasXQxy9OkZVk/CA= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; b=nDT9vgd+8am5o6jfeHLt0Qdm/sVM8rjMzmD/KSo4q5xEIsYTaLPEpSYqNyi669ZS1k xp7KIW/w3bWijuJ3i/rK/XiEw/vxjz/z1rpSTTEq5a4qXJf7ilJyWWNXsCNXeBUzP6qf HMhj84YcIGQ/Hzu02PV2mf7HCmfmcByjKgi9M= The address in CR2, 0x3e75b8c0, is not valid. Given that you can't derive that value from the instruction which is executing at the time of the panic (0x0043145a is "mov (%edi,%ebx,1),%esi" and none of those values add up to give the CR2 value) I would assume a misprogrammed DMA transaction somehow. But I don't know. It's just a guess based on what you've posted. Someone else may post a much more brilliant and useful analysis. I hope so. Good luck. On Tue, Sep 29, 2009 at 5:46 PM, Duane Murphy <duanemurphy(a)mac.com> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA256 > > Hi Folks, > I'm getting an intermittent kernel panic calling readBytes in an > IOMemoryDescriptor. The underlying descriptor is an > IOBufferMemoryDescriptor. > > The kernel panic register dump and decoded stack trace are at the end > of the message. > > What would be the cause of a kernel panic like this? > > The kext I am working with is a disk driver (subclass of > IOBlockStorageDriver). This panic happens more often when the > underlying disk is a Soft RAID disk. > > The buffer that is failing is allocated on the IO path. I am not > getting any errors for any allocations. I am working on changing this > so that there should be less allocations on the IO path. > > Could allocating this buffer on the IO path be the underlying cause of > this panic? > > Thanks for your insight to this difficult to decipher problem. > > ...Duane > > - --- > Mon Sep 28 15:13:08 2009 > panic(cpu 0 caller 0x001AB0FE): Kernel trap at 0x0043145a, type > 14=page fault, registers: > CR0: 0x8001003b, CR2: 0x3e75b8c0, CR3: 0x01107000, CR4: 0x00000660 > EAX: 0x00000280, EBX: 0x007cbe78, ECX: 0x079b6a88, EDX: 0x000f97cf > CR2: 0x3e75b8c0, EBP: 0x5697fd88, ESI: 0x08dd5280, EDI: 0x3df8fa48 > EFL: 0x00010202, EIP: 0x0043145a, CS: 0x00000008, DS: 0x01f80010 > Error code: 0x00000000 > > 0x12b4c6 <panic+422>: mov 0x4dd280,%eax > 0x1ab0fe <kernel_trap+1450>: add $0xcc,%esp > 0x1a1713 <trap_from_kernel+26>: mov %edi,%esp > 0x43145a <IOGeneralMemoryDescriptor::dmaCommandOperation(unsigned > long, void*, unsigned int) const+1260>: mov (%edi,%ebx,1),%esi > 0x4315a4 <IOGeneralMemoryDescriptor::getPhysicalSegment64(unsigned > long, unsigned long*)+160>: test %eax,%eax > 0x430e56 <IOMemoryDescriptor::readBytes(unsigned long, void*, > unsigned long)+76>: mov %edx,%ecx > > > > ...Duane > > > > > -----BEGIN PGP SIGNATURE----- > Version: PGP Desktop 9.10.0 (Build 500) > Charset: US-ASCII > > wsBVAwUBSsKqXUrg9acQ4r2CAQhU/ggAoeeWKreh7MA5WQBzfnCogYccgmHVuEKR > E2dy63TG1QFj1BsuoneAKMy3z8Gm+JwvJafui5YBWNBNSMG2FOsidK5FWmSKePuW > 1ptw9FXqszuviCkNQ5GO7ThhFA+6bzfxvQr7FeBWSmea+/R6BH8o8wydeeNNzGM8 > fqzrY7rz38xWqt2ZvofC4Y7cOj9MJX52W+7ll16rBy3TZvY+DisqcCUONdfOHuOy > wxPaPGra6oIbJBoliHcOUKF3LkMw7RFAWtkh+vXSvoJ4k88pPZm8bAn1Q6wmyMju > +mrGyc46pvBronykfxIHOXDaOdgsM6rXk6jOsJZHVWnfaoAz3o1W4w== > =9to1 > -----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/brian.bechtel%40gmail.… > > This email sent to brian.bechtel(a)gmail.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: Kernel Panic in IOMemoryDescriptor
by Duane Murphy 30 Sep '09

30 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 I need to make a correction. The buffer being used for readBytes is an IO buffer passed in from the upper level drivers. However, the void* bytes parameters used to receive the bytes is from a buffer that is allocated on the IO path. On Sep 29, 2009, at 5:46 PM, Duane Murphy wrote: > Hi Folks, > I'm getting an intermittent kernel panic calling readBytes in an > IOMemoryDescriptor. The underlying descriptor is an > IOBufferMemoryDescriptor. > > The kernel panic register dump and decoded stack trace are at the > end of the message. > > What would be the cause of a kernel panic like this? > > The kext I am working with is a disk driver (subclass of > IOBlockStorageDriver). This panic happens more often when the > underlying disk is a Soft RAID disk. > > The buffer that is failing is allocated on the IO path. I am not > getting any errors for any allocations. I am working on changing > this so that there should be less allocations on the IO path. > > Could allocating this buffer on the IO path be the underlying cause > of this panic? > > Thanks for your insight to this difficult to decipher problem. > > ...Duane > > --- > Mon Sep 28 15:13:08 2009 > panic(cpu 0 caller 0x001AB0FE): Kernel trap at 0x0043145a, type > 14=page fault, registers: > CR0: 0x8001003b, CR2: 0x3e75b8c0, CR3: 0x01107000, CR4: 0x00000660 > EAX: 0x00000280, EBX: 0x007cbe78, ECX: 0x079b6a88, EDX: 0x000f97cf > CR2: 0x3e75b8c0, EBP: 0x5697fd88, ESI: 0x08dd5280, EDI: 0x3df8fa48 > EFL: 0x00010202, EIP: 0x0043145a, CS: 0x00000008, DS: 0x01f80010 > Error code: 0x00000000 > > 0x12b4c6 <panic+422>: mov 0x4dd280,%eax > 0x1ab0fe <kernel_trap+1450>: add $0xcc,%esp > 0x1a1713 <trap_from_kernel+26>: mov %edi,%esp > 0x43145a <IOGeneralMemoryDescriptor::dmaCommandOperation(unsigned > long, void*, unsigned int) const+1260>: mov (%edi,%ebx,1),%esi > 0x4315a4 <IOGeneralMemoryDescriptor::getPhysicalSegment64(unsigned > long, unsigned long*)+160>: test %eax,%eax > 0x430e56 <IOMemoryDescriptor::readBytes(unsigned long, void*, > unsigned long)+76>: mov %edx,%ecx > > > > ...Duane > > > ...Duane -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.10.0 (Build 500) Charset: US-ASCII wsBVAwUBSsKsNUrg9acQ4r2CAQj1+Af+JxU1ncV9R3V5U2nrNZ8JWME5RjctQwjB NOEVf1LNDrMkGuBG6e1ciKIIGqBDxYqQSVjDAouqnIy0/fsqHvIUFCpD5fZX3AlF 1z49XVyuGmiQR72VoN8DbyTwl4TtyI4vFvtR5FFBw8JEFVc3p+oOkxQo5dlvmbcH QxEc4/nSR2lLSDH7G2R32k2X21niiW4/O+erKvZl/M4Ivbk14Q/z6cHzbJYGLhNu u53Ikx4Lv3p9FIGCQWxh5PQhqSUDt6uKjrzzNX1kP8/LctneUDvrqHYCb0rol+4c DrvCLqduRWiRBRwenNiq4Oz4b7UiC6I95PN9aJ5hhSsbbJxtLX2ckw== =2BGr -----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

Kernel Panic in IOMemoryDescriptor
by Duane Murphy 30 Sep '09

30 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 Hi Folks, I'm getting an intermittent kernel panic calling readBytes in an IOMemoryDescriptor. The underlying descriptor is an IOBufferMemoryDescriptor. The kernel panic register dump and decoded stack trace are at the end of the message. What would be the cause of a kernel panic like this? The kext I am working with is a disk driver (subclass of IOBlockStorageDriver). This panic happens more often when the underlying disk is a Soft RAID disk. The buffer that is failing is allocated on the IO path. I am not getting any errors for any allocations. I am working on changing this so that there should be less allocations on the IO path. Could allocating this buffer on the IO path be the underlying cause of this panic? Thanks for your insight to this difficult to decipher problem. ...Duane - --- Mon Sep 28 15:13:08 2009 panic(cpu 0 caller 0x001AB0FE): Kernel trap at 0x0043145a, type 14=page fault, registers: CR0: 0x8001003b, CR2: 0x3e75b8c0, CR3: 0x01107000, CR4: 0x00000660 EAX: 0x00000280, EBX: 0x007cbe78, ECX: 0x079b6a88, EDX: 0x000f97cf CR2: 0x3e75b8c0, EBP: 0x5697fd88, ESI: 0x08dd5280, EDI: 0x3df8fa48 EFL: 0x00010202, EIP: 0x0043145a, CS: 0x00000008, DS: 0x01f80010 Error code: 0x00000000 0x12b4c6 <panic+422>: mov 0x4dd280,%eax 0x1ab0fe <kernel_trap+1450>: add $0xcc,%esp 0x1a1713 <trap_from_kernel+26>: mov %edi,%esp 0x43145a <IOGeneralMemoryDescriptor::dmaCommandOperation(unsigned long, void*, unsigned int) const+1260>: mov (%edi,%ebx,1),%esi 0x4315a4 <IOGeneralMemoryDescriptor::getPhysicalSegment64(unsigned long, unsigned long*)+160>: test %eax,%eax 0x430e56 <IOMemoryDescriptor::readBytes(unsigned long, void*, unsigned long)+76>: mov %edx,%ecx ...Duane -----BEGIN PGP SIGNATURE----- Version: PGP Desktop 9.10.0 (Build 500) Charset: US-ASCII wsBVAwUBSsKqXUrg9acQ4r2CAQhU/ggAoeeWKreh7MA5WQBzfnCogYccgmHVuEKR E2dy63TG1QFj1BsuoneAKMy3z8Gm+JwvJafui5YBWNBNSMG2FOsidK5FWmSKePuW 1ptw9FXqszuviCkNQ5GO7ThhFA+6bzfxvQr7FeBWSmea+/R6BH8o8wydeeNNzGM8 fqzrY7rz38xWqt2ZvofC4Y7cOj9MJX52W+7ll16rBy3TZvY+DisqcCUONdfOHuOy wxPaPGra6oIbJBoliHcOUKF3LkMw7RFAWtkh+vXSvoJ4k88pPZm8bAn1Q6wmyMju +mrGyc46pvBronykfxIHOXDaOdgsM6rXk6jOsJZHVWnfaoAz3o1W4w== =9to1 -----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

Re: Problem Calling namei
by Terry Lambert 29 Sep '09

29 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -- Terry Hi, int namei(struct nameidata *ndp); void nameidone(struct nameidata *); _______________________________________________ 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 you are implementing an "ls", you are doing it un user space so what KPI is visible in the kernel should not be an issue for you. For most things the correct API is getattrlist. For private communication between your "ls" and a specific FS implementation (presumably your own), fsctl/ioctl/fcntl are the interfaces to use. If you use this approach, make sure your argument structure is size invariant between 32 and 64 bit so you don't have to carry around 64/32 conversion functions, which can otherwise get hairy when taking into account 32/64 user on 32/64 kernel. For example, if you need to pass a pointer, pass it in a uint64_t/unsigned long long (you should try to keep away from passing pointers and stay in the argument size bounds for the *ctl copy in/out to avoid multiple u/k boundary crossings, if at all possible). On Sep 29, 2009, at 7:17 AM, mominabrar <abrar.momin(a)gmail.com> wrote: I am trying to implement my filesystem specific ls (to give more file info specific to my fs) The issue is my kernel seems to be missing the namei family of functions, namely: Obviously, also the header files lack definition for struct nameidata. (bsd/sys/namei.h) Any pointers on why they are missing or how i could achieve the same functionality are appreciated. #uname -a Darwin macmini1 9.6.1 Darwin Kernel Version 9.6.1: Wed Dec 10 10:38:33 PST 2008; root:xnu-1228.9.75~3/RELEASE_I386 i386 Calling VNOP_LOOKUP in not an option, as all i have is a relative path name. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Problem Calling namei
by anubhav rakshit 29 Sep '09

29 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:cc:content-type; bh=5Ys8eEvMR8/AEJrfbYxOU8EdA+fy33uxf7PGfie78ko=; b=rHpYJDj5oTsgluAT5sNYTkK7PH3mprlMELrbHAXpjdD+cS+U2bZLcPkCcLOBqqbtav wrX7bQq3jx3mjsJ4HRPpnCr5XeNbGG33i1xjZlfAICPyBfAYJUMiHu4svqEZUDspvnKX 7D9jA74t6Mm23D/pQSScoXvUactwVywcHVFCY= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; b=VY0L1y9qyzW+rNGVoFkyFcQEpauEcHNH5Jkzr33mbd2FdYDV9f1Vbc9CGeK0CzWQXX 21RRCfgfLoQyQpupkIirlW0738TIVlQNsk6KqPZYyn5BcJFYgF6sEFG/4mccV9N6q/Jf CTI6aV6BXAZvgz+VYmxc4ek1VB5nwLEJRFano= _______________________________________________ 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

Problem Calling namei
by mominabrar 29 Sep '09

29 Sep '09

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:content-type; bh=XKghPekghngar3+vOFZqk4mSAWVcKcPAnE9vkZOPdLI=; b=fufwQJlx7lD2fPe5yG5cPd7zpr462QrJcmF37scNyXJ5pXHyvF8FdpHaqHG9uWO1+j xe/R1cgMBlQY8xaJZerBp/MFr46yzxhR92vso6n+7nhMkpx57k8wYUbBG5HtlOQQxNhZ 4E6hEeNx2zWOAxC0lSMpVBKGRgW2ezq6hHo2I= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; b=VOEjsDdNdAZJMZjIG6OTJ1VSC9fyQPHLA9SAG0aJUbmt9GtN1//JXGPGQ5i9pRP2uL brpuOg1ECCzjl8fhciT4NSH70K2UB3MgI/FF3b1sRYAYZzAMlmT6uKHQS6yhr/QqL3Qu HOXrgJlM2nJ90wi8+tpIjbz/+TLklEc6n+UC8= _______________________________________________ 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