November 2008 - Darwin-kernel

Re: ACL handling for NFSv4
by Michael Smith 28 Nov '08

28 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Nov 27, 2008, at 9:06 AM, Rick Macklem 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.… >From looking at kern_credential.c, all I can think of is doing kauth_cred_guid2gid() first and assuming it is a group, if it succeeds. (Which won't work if a given guid_t represents both a gid and uid.) Any suggestions on how to handle this? GUIDs are globally unique IDs; it is a directory service administration error to have a GUID that maps to more than one entity. Group first is a sensible policy; it's reasonable to expect that ACL entries will tend to indirect through role groups than directly reference users. This email sent to site_archiver(a)lists.apple.com

1 0

Re: ACL handling for NFSv4
by Rick Macklem 27 Nov '08

27 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Thu, 20 Nov 2008, Terry Lambert wrote: Ok, thanks to the help you guys have provided, I've been able to code most of it. I have run into one more glitch w.r.t. handling ace_applicable when setting an ACL. If kauth_wellknown_guid() returns KAUTH_WKG_NOT, I have an ace_applicable that I need to turn into: name(a)dns.domain and to do that, I need to know: - is it a user or group? kauth_cred_guid2gid() first and assuming it is a group, if it succeeds. (Which won't work if a given guid_t represents both a gid and uid.) Any suggestions on how to handle this? Thanks in advance for any help, rick _______________________________________________ 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.… [good stuff snipped] So if I have an ACE I am trying to set, since the storage format *happens* to be GUID, then we translate the integer for the gid and uid of the current credential into a GUID for the purposes of storage. For the purposes of comparison on lookup, we then convert the current credential into a GUID (in general, the uid, since that is the primary identity in a credential), and compare the GUID to the GUID in the ACE. This is handled in xnu in bsd/kern/kern_authorization.c:kauth_acl_evaluate(). Although there are support functions for translating the other direction, e.g. kauth_cred_guid2uid(), and kauth_cred_guid2gid() no one uses them, and I would, in fact, like to keep things that way. The only exception to this is kauth_cred_guid2gid() is used by kauth_cred_ismember_guid() when attempting to set a group owner on a file. This use is poorly supported by the rest of the code and should be avoided. Once I know that, I can use kauth_cred_guid2uid() or kauth_cred_guid2gid(). (I know you say you'd rather they not be used, but I don't see another way.) >From looking at kern_credential.c, all I can think of is doing This email sent to site_archiver(a)lists.apple.com

1 0

Re: Panics on a new MacPro
by Ed Wynne 25 Nov '08

25 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Nov 25, 2008, at 4:20 PM, Brian Bechtel wrote: On Tue, Nov 25, 2008 at 11:53 AM, George Plymale <george(a)southernohio.net> wrote: The two panics are pretty different. CPU 3 held off interrupts too long (the cpus_to_respond is a mask) This panic is a message from the core which noticed the problem, so the backtrace is useless. 10.5.5 has some new code to report the unresponsive processor, but not all cases can be covered. I would tend to think this is a software issue, not a hardware issue. -Ed _______________________________________________ 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 just purchased a refurbished MacPro and am quite concerned and was hopeful for perhaps some additional enlightenment on the following panics that seem fairly disjointed. Obviously I'm rather concerned with the issue as I fear perhaps a processor may be at fault. Googling the first panic has lead me to wonder if it is tied to Parallels and the second is fairly nebulous to me at the moment. Please advise as to what you may see is going on here: Sat Nov 22 13:22:49 2008 panic(cpu 7 caller 0x00193077): "pmap_flush_tlbs() timeout: " "cpu(s) failing to respond to interrupts, pmap=0x536500 cpus_to_respond=0x4"@/SourceCache/xnu/xnu-1228.7.58/osfmk/i386/ pmap.c:4580 Actually, this can be a hardware problem... but it usually implies a firmwarey / software misconfiguration of the hardware. I recently worked on a product that had issues responding to PCI-e read requests fast enough to prevent the issuing CPU from occasionally panic'ing this way. smime.p7s

1 0

Re: Panics on a new MacPro
by Brian Bechtel 25 Nov '08

25 Nov '08

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:received:received:message-id:date:from:to :subject:cc:in-reply-to:mime-version:content-type :content-transfer-encoding:content-disposition:references; bh=h4uhANwcWSmceoYHdXWJG1zvRXSmVsiyp0CzgnQM+w8=; b=B2TvYsKZxVGZcva41GWc1BaK62VqrIsafXWFB1C6h7alQTO17wJ1yorunezXBz8WDs GwdH+dVO2NqryfqxcCl4hxLSqseRckJzXdqLHQsUaJemQn9rKl2VKNwYbTjsmdXzxGfk wARm0NpcvQhAYhSPCMl63CWEPe2qQ6dSO+Ysk= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=wtkhd1F/ovyfiJKzT1JO+4lmevvpgSCPPHWHIy599Zb3n+oI05eAB8XbWehg20DXvn Hl6zMIYsvOlsMdN4eAYBHAVwW/dt2gE7tFiSa5NUjY7B/fJmG6J8Qa+e9LTAJxy2lVdg s5ASZWy9lr9NAfOVADAZ96Bg9q4McnV1kY37k= On Tue, Nov 25, 2008 at 11:53 AM, George Plymale <george(a)southernohio.net> wrote: > I just purchased a refurbished MacPro and am quite concerned and was hopeful > for perhaps some additional enlightenment on the following panics that seem > fairly disjointed. The two panics are pretty different. > Obviously I'm rather concerned with the issue as I fear > perhaps a processor may be at fault. Googling the first panic has lead me > to wonder if it is tied to Parallels and the second is fairly nebulous to me > at the moment. Please advise as to what you may see is going on here: > > Sat Nov 22 13:22:49 2008 > panic(cpu 7 caller 0x00193077): "pmap_flush_tlbs() timeout: " "cpu(s) > failing to respond to interrupts, pmap=0x536500 > cpus_to_respond=0x4"@/SourceCache/xnu/xnu-1228.7.58/osfmk/i386/pmap.c:4580 CPU 3 held off interrupts too long (the cpus_to_respond is a mask) This panic is a message from the core which noticed the problem, so the backtrace is useless. 10.5.5 has some new code to report the unresponsive processor, but not all cases can be covered. I would tend to think this is a software issue, not a hardware issue. > Tue Nov 25 13:16:03 2008 > panic(cpu 0 caller 0x001A8CEC): Kernel trap at 0x001924c7, type 14=page > fault, registers: > CR0: 0x8001003b, CR2: 0x767bc000, CR3: 0x00d44000, CR4: 0x00000660 > EAX: 0x767bc000, EBX: 0x91d8d000, ECX: 0x1114832c, EDX: 0x00000000 > CR2: 0x767bc000, EBP: 0x8825bdd8, ESI: 0x1114832c, EDI: 0x1114832c > EFL: 0x00010206, EIP: 0x001924c7, CS: 0x00000008, DS: 0x00000010 > Error code: 0x00000000 > > Backtrace (CPU 0), Frame : Return Address (4 potential args on stack) > 0x8825bbc8 : 0x12b0fa (0x459234 0x8825bbfc 0x133243 0x0) > 0x8825bc18 : 0x1a8cec (0x4627a0 0x1924c7 0xe 0x461f50) > 0x8825bcf8 : 0x19eed5 (0x8825bd10 0xa 0x8825bdd8 0x1924c7) > 0x8825bd08 : 0x1924c7 (0xe 0x48 0x10 0x10) > 0x8825bdd8 : 0x19273b (0x1114832c 0x91d8d000 0x0 0x193796) > 0x8825be38 : 0x193796 (0x1114832c 0x91d8d000 0x0 0x0) > 0x8825be98 : 0x193b91 (0x1114832c 0x91d8d000 0x0 0x53c88d4) > 0x8825bed8 : 0x193c04 (0x46e20 0x60 0x0 0x3d065ac) > 0x8825bef8 : 0x17c2c0 (0x46e20 0x8825bf7c 0x12c 0xf4240) > 0x8825bf98 : 0x17c67d (0x0 0x0 0x8825bfc8 0x12fa2410) > 0x8825bfc8 : 0x19eccc (0x0 0x0 0x1a20b5 0x23bb79e0) > Backtrace terminated-invalid frame pointer 0 This panic has the following backtrace: 0x12b0fa <panic+422> 0x1a8cec <kernel_trap+1450> 0x19eed5 <trap_from_kernel+26> 0x1924c7 <pmap64_pdpt+92> 0x19273b <pmap64_pde+79> 0x193796 <pmap_pte+82> 0x193b91 <phys_attribute_test+193> 0x193c04 <pmap_get_refmod+25> 0x17c2c0 <vm_pageout_scan+3578> 0x17c67d <vm_pageout_continue+58> 0x19eccc <call_continuation+28> a corruption of kernel data structures is the simple way to describe this. The pmap data structures are damaged somehow. This is most commonly software, but I suppose it could conceivably be hardware... First suspect Parallels. It's a common reason. Upgrade to the latest version available, or remove it altogether and see if the panics stop. Next suspect other hack-ish software which tries to modify system behavior. It's remotely possible it's the hardware. Use the hardware diagnostics which came with the machine, or get them from http://www.info.apple.com/discimages/index.html and if you find a problem, contact Apple and ask for a replacement. _______________________________________________ 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

Panics on a new MacPro
by George Plymale 25 Nov '08

25 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com BSD process name corresponding to current thread: DirectoryService Mac OS version: 9F33 Backtrace (CPU 0), Frame : Return Address (4 potential args on stack) 0x8825bbc8 : 0x12b0fa (0x459234 0x8825bbfc 0x133243 0x0) 0x8825bc18 : 0x1a8cec (0x4627a0 0x1924c7 0xe 0x461f50) 0x8825bcf8 : 0x19eed5 (0x8825bd10 0xa 0x8825bdd8 0x1924c7) 0x8825bd08 : 0x1924c7 (0xe 0x48 0x10 0x10) 0x8825bdd8 : 0x19273b (0x1114832c 0x91d8d000 0x0 0x193796) 0x8825be38 : 0x193796 (0x1114832c 0x91d8d000 0x0 0x0) 0x8825be98 : 0x193b91 (0x1114832c 0x91d8d000 0x0 0x53c88d4) 0x8825bed8 : 0x193c04 (0x46e20 0x60 0x0 0x3d065ac) 0x8825bef8 : 0x17c2c0 (0x46e20 0x8825bf7c 0x12c 0xf4240) 0x8825bf98 : 0x17c67d (0x0 0x0 0x8825bfc8 0x12fa2410) 0x8825bfc8 : 0x19eccc (0x0 0x0 0x1a20b5 0x23bb79e0) Backtrace terminated-invalid frame pointer 0 BSD process name corresponding to current thread: kernel_task Mac OS version: 9F33 _______________________________________________ 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 just purchased a refurbished MacPro and am quite concerned and was hopeful for perhaps some additional enlightenment on the following panics that seem fairly disjointed. Obviously I'm rather concerned with the issue as I fear perhaps a processor may be at fault. Googling the first panic has lead me to wonder if it is tied to Parallels and the second is fairly nebulous to me at the moment. Please advise as to what you may see is going on here: Sat Nov 22 13:22:49 2008 panic(cpu 7 caller 0x00193077): "pmap_flush_tlbs() timeout: " "cpu(s) failing to respond to interrupts, pmap=0x536500 cpus_to_respond=0x4"@/ SourceCache/xnu/xnu-1228.7.58/osfmk/i386/pmap.c:4580 Backtrace (CPU 7), Frame : Return Address (4 potential args on stack) 0x883c3a88 : 0x12b0fa (0x459234 0x883c3abc 0x133243 0x0) 0x883c3ad8 : 0x193077 (0x45f3a4 0x536500 0x4 0x19d8f8) 0x883c3b48 : 0x19339d (0xe452d60 0x1 0x0 0x10d2cc94) 0x883c3bf8 : 0x195806 (0x536500 0x87329000 0x0 0xfec39948) 0x883c3c58 : 0x169e6f (0x536500 0x87329000 0x0 0x8732d000) 0x883c3d78 : 0x16a437 (0x8732d000 0x0 0x1 0x0) 0x883c3db8 : 0x164699 (0x1634f78 0x87329000 0x0 0x8732d000) 0x883c3df8 : 0x38f8eb (0x1634f78 0x87329000 0x4000 0x199ad328) 0x883c3e18 : 0x38fa8c (0x23d3084 0x1b8b2020 0x883c3e48 0x1f1eaf) 0x883c3e48 : 0x38faee (0xe7e8808 0x202 0x883c3e78 0x3e8c57) 0x883c3e78 : 0x36751b (0x1e75f040 0x883c3ec8 0x200000 0xe14dde4) 0x883c3ee8 : 0x36790c (0x110b4e00 0x1e75f040 0xe7e87b0 0x1a7efc) 0x883c3f28 : 0x3679c8 (0x0 0x17 0x883c3f5c 0x1) 0x883c3f78 : 0x3ddd6e (0xe7e87b0 0x115bdd00 0x115bdd44 0x1fd8c6c4) 0x883c3fc8 : 0x19f3b3 (0x1f59e160 0x0 0x1a20b5 0x202f5a78) 0xb0490eb8 : 0xff202528 (0xff1f2428 0xff1f2428 0xff202528 0xff212528) Backtrace continues... Kernel version: Darwin Kernel Version 9.5.0: Wed Sep 3 11:29:43 PDT 2008; root:xnu-1228.7.58~1/RELEASE_I386 System model name: MacPro3,1 (Mac-F42C88C8) Tue Nov 25 13:16:03 2008 panic(cpu 0 caller 0x001A8CEC): Kernel trap at 0x001924c7, type 14=page fault, registers: CR0: 0x8001003b, CR2: 0x767bc000, CR3: 0x00d44000, CR4: 0x00000660 EAX: 0x767bc000, EBX: 0x91d8d000, ECX: 0x1114832c, EDX: 0x00000000 CR2: 0x767bc000, EBP: 0x8825bdd8, ESI: 0x1114832c, EDI: 0x1114832c EFL: 0x00010206, EIP: 0x001924c7, CS: 0x00000008, DS: 0x00000010 Error code: 0x00000000 Kernel version: Darwin Kernel Version 9.5.0: Wed Sep 3 11:29:43 PDT 2008; root:xnu-1228.7.58~1/RELEASE_I386 System model name: MacPro3,1 (Mac-F42C88C8) This email sent to site_archiver(a)lists.apple.com

1 0

Re: ACL handling for NFSv4
by Rick Macklem 25 Nov '08

25 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Thu, 20 Nov 2008, Terry Lambert wrote: Great, sounds like I don't have to worry about filesec_t's, just the ACLs. Hmm, I seem to be stuck using these. For a setattr, what I have to put on the wire is an ACL with "who" entries in the ACEs that look like: @OWNER, @GROUP, ... or name(a)dns.domain (where "name" is a user or group name) The only way I see I can do this are the kauth_cred_guid2* calls. Correct? But, if these fail for any reason, it's really not my problem. All I do is return the error on the VNOP_SETATTR() call. (ie. If they fail, it just limits what ACLs can be set on the server.) _______________________________________________ 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.… [good stuff snipped] No, the operations are symmetric. If vnode_setattr() returns without having claimed to have supported the ACL set, the fallback path will attempt to save the ACL using an extended attribute. Likewise if the ACL is requested and vnode_getattr() returns without claiming to have supported the request, the fallback will look for an EA. [more good stuff snipped] So if I have an ACE I am trying to set, since the storage format *happens* to be GUID, then we translate the integer for the gid and uid of the current credential into a GUID for the purposes of storage. For the purposes of comparison on lookup, we then convert the current credential into a GUID (in general, the uid, since that is the primary identity in a credential), and compare the GUID to the GUID in the ACE. This is handled in xnu in bsd/kern/kern_authorization.c:kauth_acl_evaluate(). Although there are support functions for translating the other direction, e.g. kauth_cred_guid2uid(), and kauth_cred_guid2gid() no one uses them, and I would, in fact, like to keep things that way. The only exception to this is kauth_cred_guid2gid() is used by kauth_cred_ismember_guid() when attempting to set a group owner on a file. This use is poorly supported by the rest of the code and should be avoided. Now, it seems that I have to get the "who" from the ace_applicable field of the ace in the ACL passed down through VNOP_SETATTR(). For @OWNER, @GROUP,... the kauth_wellknown_guid() should handle it, but when it returns KAUTH_WKG_NOT, I have an ace_applicable field that I need to turn into a uid or gid. (Once I have a uid or gid, I already have a cache plus upcall to a userland daemon to turn that into "name(a)dns.domain".) Basically, I'd like to thank Mike and Terry for their great responses. I think I know what I need to do now, rick. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Porting Darwin to PowerPC BookE
by mm w 24 Nov '08

24 Nov '08

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:received:received:message-id:date:from:to :subject:cc:in-reply-to:mime-version:content-type :content-transfer-encoding:content-disposition:references; bh=6eBUK7lZeajUHc/dcYZ/RoxwCEAbkGcQvlCIZwMvvZQ=; b=xXI5vcVOvq7N2B22C8APMC0X4tyh+EgiP4OVpe02ulhGNVhvO1Ly3CCysWxkiWPKT8 qOkQW33iOtHgKqNgd0KSInErSH0Djh7TELS8RiJ07kAe3+0InTeLk2wktI3DdV3tj/pJ d4WiQhe8/w2xYT0LAf/291Q+OwuppnQT/TOxc= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=KPEVXBDT+KJ/Y4iwdq3XI3Q60HDysTeUw6nRFZnSUxY/3dddv+E1zLWr6GtQ+ftW7X o35p5eYVtA6Id2vQhQbUZ87f6qyC6AORqAjRh1t3cM1hw8xnc5zLm7+hMrouS75UOpXE Ub0IzTCvdLdnJm8aN6v1Nv0JB/6G+zl8c9Jn8= have you notice that the powerpc support should be dropped? if you take an old version of darwin 6.6 ... Cheers! On Mon, Nov 24, 2008 at 12:06 AM, feng kan <fengkan(a)yahoo.com> wrote: > Hi All: > I apologize in advance if this email is incorrect for the group. I am > looking to port the Darwin OS to > PowerPC Book E architecture. I am seeking to know if anyone has already done > such porting? If such > porting is already done, can someone point me to the person or the parties > involved in it? I am interested > in obtaining such port. > > Much Thanks > Feng Kan > > > > _______________________________________________ > 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/openspecies%40gmail.com > > This email sent to openspecies(a)gmail.com > -- -mmw _______________________________________________ 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

Porting Darwin to PowerPC BookE
by feng kan 24 Nov '08

24 Nov '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Mailer:Date:From:Subject:To:MIME-Version:Content-Type:Message-ID; b=S27tE8kDbLxFOYkRKyOgnJjiSNIVzuy5/OpFV/AWk2Krtt8yUzfJ9w4b2RZAn3k2kpkMUxIxCqwV16Lv2yO/TY+aqfqDShd0u+OR5FLryOcjAldNAkr7rsi7FKO9YlGDob0gL5Vz42ALN+dmzRjg5WQvwfah+Eib2GcJ0jbfkXU=; _______________________________________________ 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: Safely reading Intel MSRs?
by mm w 21 Nov '08

21 Nov '08

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:received:received:message-id:date:from:to :subject:cc:in-reply-to:mime-version:content-type :content-transfer-encoding:content-disposition:references; bh=K9l41EcrMS8aQ7f+L20+Jv1/Ige0awywLoclEpjcotA=; b=rI6IJ95HnLkoqhHn6SV7Qp4/fDEfzD4HOOoA9mWo4RgWgz9eXDGiueWNAkTmR5J8YJ 5PzafeRBHb4Yutwi93x4VaFTaUr7f4YlpO/XrfLHWWROaYadkTS+aoAWoJh+fss91ZqO 6tZFX+yK9SSpi2ok6ClmrS5kV4NjsgEpUvDuI= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=ksaIb9OoCgHA67+lNZpzWlzPI/SCjjgteFhDzWQt/uLCiFp+4gwntbO+HNK0eRJRoD 4KLs4146eo55FpdSugVMABlqW81Ssh6IPSTCaX0sdZSFyv5iGORAyn/G3sh3wP+o1P28 6S6ucW9lC4fyxq9koYyM7PlBgjsXPVofPs/v4= Hi 2 , I forgot, if you could choose a proper design and not to read these values from your kext it should be great for the safety of your customers Cheers! On Fri, Nov 21, 2008 at 2:03 PM, mm w <openspecies(a)gmail.com> wrote: > Hi nop (and yes include what you need), > > osfmk/i386/proc_reg.h > so can use the defined rdmsr, I guess , instead of doing it by hand, > if I did "well" understand, > > unsigned hi, lo, msr; > > __asm( > movl %%edx, % > ... > ); > > Cheers! > > On Fri, Nov 21, 2008 at 8:58 AM, Brent Burton <brentpburton(a)gmail.com> wrote: >> Hi all, >> >> I am writing some internal system software including a kernel >> extension that has the ability to read MSR values. >> While I have a runtime check for processor family and model, the >> mapping to valid MSR registers isn't finished (and of course, may have >> errors when it is finished). If an invalid MSR is read or written, it >> causes a kernel panic (via GPF), and I'd like to protect against this >> for all cases. >> >> Are there kernel functions, accessible from a kext, that provide safe >> reading of Intel MSRs? >> >> If not, one alternative is to (temporarily) redirect GPF handling. I >> don't particularly like this path, but I want to get more information >> on it. Where is documentation or headers that cover this, if any? >> >> thanks, >> -Brent >> _______________________________________________ >> 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/openspecies%40gmail.com >> >> This email sent to openspecies(a)gmail.com >> > > > > -- > -mmw > -- -mmw _______________________________________________ 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: Safely reading Intel MSRs?
by mm w 21 Nov '08

21 Nov '08

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:received:received:message-id:date:from:to :subject:cc:in-reply-to:mime-version:content-type :content-transfer-encoding:content-disposition:references; bh=UHvwac5XCqjP7OtYIKgj/8q1qfa9Ck0nzD9zCJQcscM=; b=gY42x5BnjLd4Lt0RMMnTZM3ZZGeYOVqlMg188joACzUrP7gOqIFvP7B4AOsn/ZBIi/ JSxccVngjxUqjqPGsGN+erZehtaZqXQ5l9MOuV6eEgVpFLRvcI9zCWDAnHC5ENYg5RVZ dnidFAeeNvWWLvscd/C3n+O0Yic7ODu/376jk= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=iyEuBioCCkQmFK5iZjldeuqLMVxL2ixoCxbimmoVNS3HdjENjM/BSmJmRyFWd1clvv 7kxkusy0SdeILy5a1kowrhLtYyFlXWBMvTbOZgJiOVGNcb8LB1maNLIHBF/bEcEIO2rz SE7DbcoXosi48e6vmLtswaIZkWwAAb4SJWZNs= Hi nop (and yes include what you need), osfmk/i386/proc_reg.h so can use the defined rdmsr, I guess , instead of doing it by hand, if I did "well" understand, unsigned hi, lo, msr; __asm( movl %%edx, % ... ); Cheers! On Fri, Nov 21, 2008 at 8:58 AM, Brent Burton <brentpburton(a)gmail.com> wrote: > Hi all, > > I am writing some internal system software including a kernel > extension that has the ability to read MSR values. > While I have a runtime check for processor family and model, the > mapping to valid MSR registers isn't finished (and of course, may have > errors when it is finished). If an invalid MSR is read or written, it > causes a kernel panic (via GPF), and I'd like to protect against this > for all cases. > > Are there kernel functions, accessible from a kext, that provide safe > reading of Intel MSRs? > > If not, one alternative is to (temporarily) redirect GPF handling. I > don't particularly like this path, but I want to get more information > on it. Where is documentation or headers that cover this, if any? > > thanks, > -Brent > _______________________________________________ > 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/openspecies%40gmail.com > > This email sent to openspecies(a)gmail.com > -- -mmw _______________________________________________ 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