June 2007 - Darwin-kernel - lists.apple.com

Re: IOKit KEXT Won't Load If Start Code Tries to Attach and Register?
by Godfrey van der Linden 13 Jun '07

13 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Good luck Godfrey On 12/06/2007, at 16:01 , Chase Colman wrote: Declarations: IOMedia* _childMedia; UInt32 _partitionCount; Code: bool com_ChangeLater_driver_TouchDrives::start(IOService* provider) { IOMedia* media = OSDynamicCast(IOMedia, provider); assert(media); if (super::start(provider)==false) return false; IOMedia* childMedia = new IOMedia; if (childMedia) { if (childMedia->init(0, media->getSize(), media->getPreferredBlockSize(), media->isEjectable(), false, false //Don't allow any write access. )) { _partitionCount++; char name[24]; sprintf(name,"ChangeLater_External%ld",_partitionCount); childMedia->setName(name); char location[12]; sprintf(location,"%ld",_partitionCount); childMedia->setLocation(location); _childMedia=childMedia; childMedia->attach(this); if (media->isWritable()==false) { DEBUG_LOG("about to attach"); (void) attachMediaObjectToDeviceTree(childMedia); DEBUG_LOG("attached, about to register"); childMedia->registerService(); } DEBUG_LOG("passed attach&reg"); return true; } else { childMedia->release(); childMedia=0; } } return false; } Here is the IOKitPersonalities: <key>CFBundleIdentifier</key> <string>com.ChangeLater.driver.TouchDrives</string> <key>IOClass</key> <string>com_ChangeLater_driver_TouchDrives</string> <key>IOProviderClass</key> <string>IOMedia</string> <key>Physical Interconnect</key> <string>USB</string> <key>Physical Interconnect Location</key> <string>External</string> <key>IOMatchCategory</key> <string>IOStorage</string> Thanks, Chase _______________________________________________ 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%40mac.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.… As Mike surmised you may be tripping over a recursive match. You are probably going to match on your self as you are publishing an IOMedia object and matching on them too, I can't see anyway of breaking out of the recursion (except for the Physical.* tags in the matching dictionary, but I'm not familiar with IOMedia matching so an expert will need to comment). A way to prove that you have gone into an infinite loop is to print a short log message and go to sleep for 10 milliseconds on entry to your start routine. This will give the kernel logger enough runtime to send data to the system log. I am working on an IOKit KEXT that blocks external USB drives that are not writable. When I try to load the KEXT my computer locks up with everything except for mouse movement, but this is only when it has the code that attaches the media object and registers service. When I remove the code that attaches the media and registers the service the KEXT loads fine, but blocks all drives including the read-only ones that I want. The code I will give is heavily based off of the SampleFilterScheme example. The attachMediaObjectToDeviceTree function is exactly the same as the one from the SampleFilterScheme example. If anyone could help me figure out why it is locking up my system instead of attaching and registering, I would be very appreciative. A fix would be good too. This email sent to gvdl(a)mac.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: lsof
by Justin Ko 12 Jun '07

12 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:reply-to:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=BL6ReSIGt2kZSujwo7v50Z4sT2/gKSjQx9B2DVX0mJsghh+qJzQztwdtZdsQg56UvW7M2Kafs5wL1ge1PGhCcOtRWLXH+SahP/ambEnf5TXCMzGZfRPOVKWYNbsTNZE3SzGXV18Wekm7NaJfcVSLTe5V5a8/ICzODjLhRkHp/zg= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:reply-to:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=l7ASYl7WdAw6j8nQo7h91605RIvmH37hqIKxNeqpO0OFx7lInFeGCUvX0HPvefvzyQ6nytlnTkgzy72lSx/GzOoz8vInwq1g+F5F41snfz0bVq4EhPyv36XpCsUfntuenbJHDmKl1pyBDN6VAlGGcRSVbVtTieORLFysfnw0fMk= You could also search the boot args (there is an argument that will reenable /dev/kmem for local hacks), but I can't remember what it is off the top of my head. The trick is to enter the following command at the command line: sudo nvram boot-arg="kmem=1" -justin _______________________________________________ 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 6/12/07, Terry Lambert <tlambert(a)apple.com> wrote: This email sent to site_archiver(a)lists.apple.com

1 0

Re: IOKit KEXT Won't Load If Start Code Tries to Attach and Register?
by Garth Cummings 12 Jun '07

12 Jun '07

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: vmmap source?
by Quinn 12 Jun '07

12 Jun '07

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: vmmap source?
by Terry Lambert 12 Jun '07

12 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com The source is not published. -- Terry _______________________________________________ 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 Jun 11, 2007, at 5:03 PM, Yount, Chuck wrote: I can't seem to find the source for the vmmap utility in the darwin source. Can anyone point me to it? Alternatively, does anyone have some sample code that enumerates the virtual memory regions for a process? The normal way you would get this information programmatically is to popen vmmap, and read its output (the process is setgid procview, so it requires privileges that are not available to ordinary users). This email sent to site_archiver(a)lists.apple.com

1 0

Re: lsof
by Terry Lambert 12 Jun '07

12 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Did you build with the 3.3 compiler or the 4.0 compiler? -- Terry _______________________________________________ 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 Jun 11, 2007, at 1:37 PM, cristian pancià wrote: After rebuiding the kernel in debug version xnu-792.17.14/BUILD/obj/ RELEASE_PPC lsof is not giving me any result,just got lsof: can't read process table Anyone can pointng' me to the light? If you built from Intel sources for PPC so that you could use the newer compiler, then you need to recompile lsof. Specifically, /dev/kmem is disabled by default on Intel and later builds, and lsof has to use a library that lets it get kernel information through a procedural rather than a data interface. This incidently reduces lsof overhead, and also makes the output more like a snapshot, so that on a per process basis, it's more self-consistent. You could also search the boot args (there is an argument that will reenable /dev/kmem for local hacks), but I can't remember what it is off the top of my head. This email sent to site_archiver(a)lists.apple.com

1 0

Re: IOKit KEXT Won't Load If Start Code Tries to Attach and Register?
by Michael Smith 12 Jun '07

12 Jun '07

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

IOKit KEXT Won't Load If Start Code Tries to Attach and Register?
by Chase Colman 12 Jun '07

12 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:mime-version:content-type; b=X3DmiuBVJAKjf0IwqtpKO/NSoeE/VpxLHrZaqSuvR7Ett2m4mfAnSWZv+m1haGpqESsgIVYjQmQk575phazUlaaw876HegoZPU7+gZB5VRWqH8Vqzee71EyIrWFive1tQeuNdfAoVvm8ufrxJ0gLhW43zakg8A1ZayPh81EK6vw= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:mime-version:content-type; b=ipT2NTpimvQa2KHfhAAPX+TfCp118Hx92wLNM/lFjCQCQDIs0Jj8tq5MAxRFZNMUda92yFJ81Y+jZxUctJrQvF18Po/NsnNemfMnQaAr2291KkUVnz24IYUTZI8L6A/qK1W5H1k+iO1H925HGIMPCutyv6x7JIlCXnDngKaCWX8= _______________________________________________ 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

vmmap source?
by Yount, Chuck 12 Jun '07

12 Jun '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Thread-index: AceshRdIipDouR2qQKaw3j44f95OZw== Thread-topic: vmmap source? _______________________________________________ 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: lsof
by Dan Shoop 11 Jun '07

11 Jun '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Permissions/ownerships issue? Does your's look like this: $ ls -als /usr/sbin/lsof 224 -rwxr-sr-x 1 root kmem 111380 May 26 2006 /usr/sbin/lsof -- -dhan ------------------------------------------------------------------------ Dan Shoop AIM: iWiring Systems & Networks Architect http://www.ustsvs.com/ shoop(a)iwiring.net http://www.iwiring.net/ 1-714-363-1174 _______________________________________________ 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.… At 10:37 PM +0200 6/11/07, cristian pancià wrote: After rebuiding the kernel in debug version xnu-792.17.14/BUILD/obj/RELEASE_PPC lsof is not giving me any result,just got lsof: can't read process table Anyone can pointng' me to the light? This email sent to site_archiver(a)lists.apple.com

1 0