December 2007 - Darwin-kernel

Re: msleep() with PDROP
by Terry Lambert 11 Dec '07

11 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com In various places in the xnu kernel I see code like the following: #1 msleep(<event>, my_mutex, PDROP, "mysleep", 0); lck_mtx_lock(my_mutex); So, why do the above instead of: #2 msleep(<event>, my_mutex, 0, "mysleep", 0); My understanding was that msleep() would release the mutex and then re-acquire it, if PDROP isn't specified. Is there any fundamental difference between door #1 and door #2? Does this make sense (ie. #1 is used to give other threads a chance to lock the mutex even if the msleep() doesn't actually sleep)? -- 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 Dec 9, 2007, at 1:30 PM, Rick Macklem wrote: I feel a bit silly responding to my own posting, but I think I figured it out and thought I'd see if I got it right: #1 - always releases/acquires the mutex, allowing other threads wanting the mutex to get a turn #2 - holds the mutex unless it sleeps Still seems a little weird to call msleep() unless you need to sleep. If you specify PCATCH as well, then it's possible for a pending signal to interrupt you before you actually get to sleep. If you are getting a lot of signals, then it's likely that the PDROP would be needed to allow another thread that also needs the mutex to make any progress at all. In general, you're going to want to specify a PCATCH so that you don't end up being non-interruptible, if you are making the call on behalf of a user space process that may decide that it doesn't want to block forever. If it's a kernel thread, you won't want PCATCH in most cases, unless you call the sleep directly with a continuation function. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Sparse files on Mac
by Terry Lambert 11 Dec '07

11 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com He could use this same technique. -- 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 Dec 8, 2007, at 2:24 AM, Anton Altaparmakov wrote: Hi, But note as things stand at present Mac OS X does not allow booting from a kext based file system so your users would not be able to have their boot file system be your file system. Actually, Apple RAID boots out of the kext cache to get the raid support using a "boot root" partition. See the third paragraph after ""BootX, boot.efi, and System Initialization" here: <http://developer.apple.com/documentation/MacOSX/Conceptual/BPSystemStartup/… > This email sent to site_archiver(a)lists.apple.com

1 0

Re: Sparse files on Mac
by Terry Lambert 11 Dec '07

11 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, On 8 Dec 2007, at 09:33, Damir Dezeljin wrote: Hi folks, Correct. UFS, ZFS, and others, however, do. -- 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 Dec 8, 2007, at 1:51 AM, Anton Altaparmakov wrote: I have an non strictly kernel related question; however, I guess you'll be able to point me to the right resources. I developed a KEXT and I though of using sparse files functionality in user space for implementing some internal logic. I tried creating a sparse file using the same set of commands I usually use on Linux, but I found out I just create a normal file: ---- ftruncate(fd, 0); ftruncate(fd, 4*1024*1024*1024); ---- Now I'm wondering if sparse files are supported on the default file system used on OS 10.4 desktop (as I know, the HFS is used)? If yes, what I'm doing wrong? HFS/HFS+ do not support sparse files. Additionally I'm wondering if is it possible to get some kind of 'holes table' as it's on Windows -> so, let suppose the sparse file has some data on an offset of 1 GB, some on 1.7 GB and so on. As far as I know, this isn't possible on Linux, but I'm sure it is on Windows. As there is no sparse support this is not relevant. You can get allocation information using the fcntl F_LOG2PHYS (see the fcntl man page) but that is quite a cumbersome interface. Also note that it will (at least on Leopard have not checked Tiger) return RANDOM JUNK for file systems supporting sparse files. Radar is down for maintenance at the moment or I would file a bug report immediately! Which FS are you talking about? For UFS, I see zero filled pages being returned, which is per the spec.. Are you sure this is not a bug specific to the FS you were testing on? The FS is more or less required to fake this up for any sparse areas, and if it doesn't, it's broken (not to mention a security issue, if it's returning random data from the disk). This email sent to site_archiver(a)lists.apple.com

1 0

Re: clear_wait() using thread address
by Jim Magee 10 Dec '07

10 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Dec 7, 2007, at 6:09 AM, Vishal Shetye wrote: I've written a KEXT, wherein a thread, sleeping on wait_queue is invoked by another thread using clear_wait(). For this I keep a global thread safe structure for storing thread's address. What I want to know is, * Is there a possibility that my thread reference may become invalid, because of kernel's internal memory movement (such as to defragment, I'm just guessing)? A thread's kernel address is valid for as long as someone holds a reference to it. Valid, executing threads always hold references on themselves. But there's the rub. How are you assuring that the thread wasn't already awoken and already terminated itself (in case of a process termination, system shutdown, etc...). Are you taking a reference on the thread? More importantly, are you really intending to wake the thread from any wait it might be performing - even ones unrelated to the condition that (you think) the thread might be waiting on? It is always better to post the specific condition than to try and determine and then blast a particular thread. * Is there a way of invoking selected process from wait_queue by its pid (like wake_up_process() in case of linux). In Darwin/Mac OS X, processes can have many threads. Each waits independently. There is no wake_up_process(). I'm sure even Linux isn't happy with that legacy API given its current trend towards native threads as well. --Jim _______________________________________________ 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: Locking of thread
by David Rose 10 Dec '07

10 Dec '07

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:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding:Message-ID; b=gfE7A57ocBtAXMWWuUMK1d05gD3IJWOjYfNmC9Yobtfz7e4LwEPRs9npoHHzmXDCPCeDFtZYQjqwe/UICnbgyUEoYjK20lVPTeUqRIlmtl+d85HGf+3HkERiNKGmq2Am47INdvYGxCTDUIlJNZlefRvDoPFT6MG8RTVVERVEeoE=; --- Rakesh Singhal <rakesh.singhal(a)gmail.com> wrote: > Thanks Mike. But here, problem is that I am not > using workloop. I want to > lock thread in one function and unlock it from > another function. > > Regards We are using lck_mtx_lock and lck_mtx_unlock. You may have to include the com.apple.kpi.libkern bundled library in your plist. Thanks, David Rose ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs _______________________________________________ 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: Page Fault Kernel Panic on Leopard
by Brian Bechtel 10 Dec '07

10 Dec '07

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=Q49rNilyypobKeGYgE55fCHZ2xCO6gG+1XzWaHtmuto=; b=CDV79mshUhfKNIJhot3eG9Mq1VSsOOccEaiWf1WWS2+DiB9hDnkYy1E0UKrCCYHpPfOWqmRWxuJCc4fKMAJoln9KKKWhMNhP4EU+N1b8kupxyXovscvs2XZr78+958yDZAs8yJc/J2nGn5/Y+6VYO2A2zSlHp88jSeEw3uaSCWA= 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=qcG9diww2aoh+yXjPfJgYeUspM4jZjZYleyrLOwlJUMF8s045wYgU2/+hkC6/s4eGAGDSBY56IBjxmeGWAMf3rVqq1KRV/E0uWpcjj91c8aBboFiZhMp3tg4DCqbdyck/nWBkjguEqV4EfT6aPuWqKw0MHLOIvQyGbB9Vk3N5oY= On Dec 7, 2007 10:25 PM, JanakiRam <johnyatforums(a)gmail.com> wrote: > Hi All, > > My Software runs into new problems after upgrading to Leopard. > > My Customers/Users are complaining that my kext gets > crashed in a specific flow in an inconsistent way , which didn't panic > on Tiger :). > > I'll setup a Core-Dump server to analyze the exact panic > information. Mean while any one can provide the information on the > page fault panic. > > > panic(cpu 0 caller 0x001A7BED): Kernel trap > at 0x346e9f78, type 14=page fault, registers: > > When does a Page Fault panic appear and any reasons why it > panics a system. A page fault panic occurs when you try to access memory which is not valid. Why not post the rest of the panic log? That's the information that tells you what really happened. In particular, on Intel, the CR2 register contains the address which caused the panic. You can draw certain conclusions from the type of value found in CR2. There are too many potential causes to classify this further based on insufficient evidence. _______________________________________________ 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: msleep() with PDROP
by Rick Macklem 09 Dec '07

09 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com I feel a bit silly responding to my own posting, but I think I figured it out and thought I'd see if I got it right: In various places in the xnu kernel I see code like the following: #1 msleep(<event>, my_mutex, PDROP, "mysleep", 0); lck_mtx_lock(my_mutex); So, why do the above instead of: #2 msleep(<event>, my_mutex, 0, "mysleep", 0); My understanding was that msleep() would release the mutex and then re-acquire it, if PDROP isn't specified. #1 - always releases/acquires the mutex, allowing other threads wanting the mutex to get a turn #2 - holds the mutex unless it sleeps Does this make sense (ie. #1 is used to give other threads a chance to lock the mutex even if the msleep() doesn't actually sleep)? Still seems a little weird to call msleep() unless you need to sleep. Again, thanks in advance for any info, 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.… Is there any fundamental difference between door #1 and door #2? This email sent to site_archiver(a)lists.apple.com

1 0

msleep() with PDROP
by Rick Macklem 09 Dec '07

09 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com In various places in the xnu kernel I see code like the following: #1 msleep(<event>, my_mutex, PDROP, "mysleep", 0); lck_mtx_lock(my_mutex); So, why do the above instead of: #2 msleep(<event>, my_mutex, 0, "mysleep", 0); My understanding was that msleep() would release the mutex and then re-acquire it, if PDROP isn't specified. Is there any fundamental difference between door #1 and door #2? Thanks in advance for any info, 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.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Help Building Kernel for PowerMac G5 Quad
by mm w 09 Dec '07

09 Dec '07

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=ZYcDgSZVZtGVedZqvGtETVqwVz7fRNkNf2Dy12L5H3M=; b=NR6t+ZnTQYHLSfoWVcm1MnW5yFggRZgyPAOvzZ6FBh1dRaeCjsSThqFd+DTpuK6oLADXILuxCK9qtP2NSnPXqSWqv2MKzPLsGVvnVeV8B6W8sMj/pJC438TKT2HHB00VUIKsgLisZ9RC5CVSUioLeND5Wm0Aq1AMRyOwTzzsZCk= 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=qBMDmHJ+2B3knNZQehxw1Rh5a6j+0TDKtNM2x7zhOXgP9Ibmk19cXHmWtHH1X3IdXrfUvF1g5btXMWgKynm26pVHC8IANivSRwOQj9DIIOLSDaEDYYCxtnI0mS9UWAVghDEKjJKhioq0QPGpx81Og2/W4txlAPKSC1klUrsSyk0= Hi Rand, you should post on darwinbuild-dev and/or a search in the history!!!! and this is the "weekend", so just wait, or try to catch someone by IRC http://trac.macosforge.org/projects/darwinbuild/ Cheers! On Dec 8, 2007 7:18 PM, Rand Childs <randchilds(a)mac.com> wrote: > An explanation. > > The reason I was trying to build a kernel for the PowerMac G5 QUAD is > that I have Oracle server installed on that machine and when started > with it's initialization parameter of "processes" greater than about > 50, the 10.5 kernel panics, whereas 10.4 kernel did not. > > The reason for the kernel panic is a bug in the semaphore code in the > kernel which I patched before building the 10.5 kernel. I was trying > to make sure that the kernel would no longer panic and Oracle would > start. Since I was not able to figure out how to configure the build > to properly build a kernel that would boot from the multi-processor > machine, I moved Oracle to the TI PowerBook G4 and booted the patched > kernel there. Oracle now starts without a kernel panic so I think > that I've properly identified and have a solution to the kernel bug. > > I've reported this using Apple's Bug Reporter. If there is anyone > from Apple that reads this list and is interested in getting the > kernel patched for the next release, the bug report id is 5574916. > > Although I no longer need to build a kernel that will boot on the > PowerMac G5 QUAD, I would still be interested to know how to configure > the build so that the kernel will properly boot on this multi- > processor machine. So if anyone could help with this, I'm still > interested. > > Thanks > > Rand > > > On Dec 8, 2007, at 12:49 PM, Rand Childs wrote: > > > I have followed Apple's instructions for building the 10.5 kernel > > and have successfully built mach_kernel. To install it I did the > > following: > > > > cp xnu-1228/BUILD/obj/RELEASE_PPC/mach_kernel /mach_kernel.new > > cp /mach_kernel /mach_kernel.orig > > cp /mach_kernel.new /mach_kernel > > > > I've done this on a portable firewire drive so that I don't screw up > > by machine. > > > > The problem is that I can boot from this kernel on my TI PowerBook > > G4 which means that the kernel will boot, however, I am unable to > > boot my PowerMac G5 QUAD. I get the white screen with the apple > > logo, and then eventually the fans ramp up to full speed. I never > > see the spinning circle and holding down "command v" doesn't do > > anything. > > > > So, the kernel that I build does work on the TI PowerBook G4, but > > not on the PowerMac G5 QUAD. It was built on the PowerMac G5 QUAD > > using Apple's instructions. > > > > Does anyone know what I need to do to build a kernel that will boot > > a PowerMac G5 QUAD? > > > > Any help would be appreciated. I couldn't find anything in the list > > archives. > > > > Thanks > > > > Rand > > > > > > > > _______________________________________________ > > 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/randchilds%40mac.com > > > > This email sent to randchilds(a)mac.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/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

Re: Help Building Kernel for PowerMac G5 Quad
by Rand Childs 09 Dec '07

09 Dec '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com An explanation. Thanks Rand On Dec 8, 2007, at 12:49 PM, Rand Childs wrote: cp xnu-1228/BUILD/obj/RELEASE_PPC/mach_kernel /mach_kernel.new cp /mach_kernel /mach_kernel.orig cp /mach_kernel.new /mach_kernel Thanks Rand _______________________________________________ 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/randchilds%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.… The reason I was trying to build a kernel for the PowerMac G5 QUAD is that I have Oracle server installed on that machine and when started with it's initialization parameter of "processes" greater than about 50, the 10.5 kernel panics, whereas 10.4 kernel did not. The reason for the kernel panic is a bug in the semaphore code in the kernel which I patched before building the 10.5 kernel. I was trying to make sure that the kernel would no longer panic and Oracle would start. Since I was not able to figure out how to configure the build to properly build a kernel that would boot from the multi-processor machine, I moved Oracle to the TI PowerBook G4 and booted the patched kernel there. Oracle now starts without a kernel panic so I think that I've properly identified and have a solution to the kernel bug. I've reported this using Apple's Bug Reporter. If there is anyone from Apple that reads this list and is interested in getting the kernel patched for the next release, the bug report id is 5574916. Although I no longer need to build a kernel that will boot on the PowerMac G5 QUAD, I would still be interested to know how to configure the build so that the kernel will properly boot on this multi- processor machine. So if anyone could help with this, I'm still interested. I have followed Apple's instructions for building the 10.5 kernel and have successfully built mach_kernel. To install it I did the following: I've done this on a portable firewire drive so that I don't screw up by machine. The problem is that I can boot from this kernel on my TI PowerBook G4 which means that the kernel will boot, however, I am unable to boot my PowerMac G5 QUAD. I get the white screen with the apple logo, and then eventually the fans ramp up to full speed. I never see the spinning circle and holding down "command v" doesn't do anything. So, the kernel that I build does work on the TI PowerBook G4, but not on the PowerMac G5 QUAD. It was built on the PowerMac G5 QUAD using Apple's instructions. Does anyone know what I need to do to build a kernel that will boot a PowerMac G5 QUAD? Any help would be appreciated. I couldn't find anything in the list archives. This email sent to randchilds(a)mac.com This email sent to site_archiver(a)lists.apple.com

1 0