May 2008 - Darwin-dev - lists.apple.com

access time changes vs. kevent observer?
by Jim Correia 14 May '08

14 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Which is the correct case? Jim _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I have a file that I am observing with kevents. I am observing with the NOTE_ATTRIB flag set. (For this test case, I am using an HFS+ disk on Leopard.) Most things which bump the access time of the file do not cause my kevent observer to fire. (For example, `cat filename` bumps the access time, doesn't fire the kevent.) Some things, however, do. Any operation which causes the custom icon to be read from the resource fork (by opening the file in any Mac OS X GUI application.) bumps the access date and fires a kevent. I'm interested in many of the changes that come along with NOTE_ATTRIB, but I don't really care about the access date. Any advice for ignoring events for which only the access date change? This email sent to site_archiver(a)lists.apple.com

1 0

Problem reading struct file on Mac 10.3.0
by Manish Chaturvedi 14 May '08

14 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: Aci1pdjR9ZkApo6QSRC+RG1sBEXHJw== _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Section Alignment in mach-o files
by MT 14 May '08

14 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thanks for any insight Mark _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I have a question about the section alignment specified in the load command in a mach-o file. Is the alignment only relevant for the virtual address, or also for the file offset? Some tests with otool seem to indicate that both the virtual address and the file offset seem to respect that alignment normally, but does it need to be this way? This email sent to site_archiver(a)lists.apple.com

1 0

Re: Details on: "malloc: *** mmap(size=16883712) failed (error code=12)"
by Michael Smith 14 May '08

14 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 12, 2008, at 12:04 PM, darwin-dev-request(a)lists.apple.com wrote: Sometimes I see messages like: malloc: *** mmap(size=16883712) failed (error code=12) *** error: can't allocate region *** set a breakpoint in malloc_error_break to debug = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I'm trying to find more details about this message. Specifically, what is error code 12? The default allocator is trying to allocate a ~16MiB chunk of virtual space, but there are no holes left in the task address map into which that will fit. Eric is correct; you would want to start with vmmap to try to work out what is causing the task address space to become fragmented. If you aren't seeing significant paging activity, you may have allocations being made but not used and thus possibly not cleaned up correctly. This email sent to site_archiver(a)lists.apple.com

1 0

Re: 25 Year old BSD bug
by Terry Lambert 13 May '08

13 May '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Hi, -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On May 13, 2008, at 8:30 AM, Andre-John Mas wrote: I came across this article this morning, by way of Gizmodo, and though it might be of interest for the list, since it also apparently effects MacOS X: http://osnews.com/story/19731/The-25-Year-Old-UNIX-Bug The code was valid until BSD 4.4, which attempted to abstract the struct dirent and struct direct to abstract UFS dependence. It remained valid in BSD 4.3 because of the intentional "coincidence" in the structures, which was done for performance reasons. The issue is deletion of the directory entry in a directory entry block in UFS internalizing an d_ino value of zero as an implementation detail, and expecting it to have been externalized as a copy of the on disk structure. Technically, the fix that was published is incorrect. The correct fix is to not externalize the entry outside of UFS, and to remove the workaround code, since deleted files at the start of some arbitrary block boundary are not going to result in 0 d_ino values on other file systems. Also, technically, there is no way to reasonable solve the problem here, as directories are not snap-shotted nor is a mandatory write lock asserted when a DIR * is returned, and held until the DIR * is disposed. Any interleaved operations which modify the directory under UFS can still trigger block compaction, entry coelescing, or directory truncation (if the end could end up with one or more trailing empty blocks). This includes not only deletes and creates, but also renames, if the new name will not fit in the old entry record. Using telldir() values across the underlying getdirentries() block boundaries from which readdir() gets its data is simply not safe for most file systems. This email sent to site_archiver(a)lists.apple.com

1 0

25 Year old BSD bug
by Andre-John Mas 13 May '08

13 May '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Hi, I came across this article this morning, by way of Gizmodo, and though it might be of interest for the list, since it also apparently effects MacOS X: http://osnews.com/story/19731/The-25-Year-Old-UNIX-Bug Andre _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

source for applesingle command?
by James Walker 12 May '08

12 May '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com User-agent: Thunderbird 2.0.0.14 (Macintosh/20080421) The man page for the applesingle tool says "Darwin" at the bottom, but I can't find the source code, either by looking at project lists or by searching the Darwin site. Any hints? -- James W. Walker, Innoventive Software LLC <http://www.frameforge3d.com/> _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

kexec to/from darwin
by Scott D. Davilla 12 May '08

12 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com It's a two stage operation. Thanks Scott _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I'm pursuing an interesting sub-project that involved porting kexec to darwin. For background, kexec is a Linux tool that is used to boot another kernel/initrd from a Linux OS. kexec usage in the Linux world has been quite interesting, from bypassing slow PC BIOS for a much faster reboot, to a boot into a updated kernel without a reboot cycle, to a jump into a crash dump kernel that can recover from a kernel panic. It's potential benefits under a darwin kernel are quite obvious once one understand the kexec operation. 1st stage involves setting up a "bootloader" code block to load the kernel and then storing the new kernel/initrd/params into a reserved section of ram and setting a reboot hook. This obviously involves a kext to do the heavy lifting in kernel land. 2nd step issues a reboot and when almost to the point of reboot, the kernel hook jumps to the code block that loads the new kernel in that same way as it would be loaded from a real bootloader. The critical point here is to be able to intercept the reboot/shutdown at the last possible safe moment and jump off into the kexec created "bootloader" code block. A kexec from darwin to darwin should be easier as the correct boot params can be easily extracted from the running darwin kernel. A kexec from darwin to Linux is the next easiest as this involves converting darwin boot params into Linux boot params and I already have existing code that does this. A kexec from Linux to darwin is more difficult as the darwin boot params will have to be created from scratch using available Linux resources. The EFI boot params should be easy, creating a proper flat device tree is the complex part. I'm not asking for help on this (yet) but just putting it out for comments and suggestions, pitfalls, etc. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Details on: "malloc: *** mmap(size=16883712) failed (error code=12)"
by Eric Gouriou 12 May '08

12 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Sometimes I see messages like: malloc: *** mmap(size=16883712) failed (error code=12) *** error: can't allocate region *** set a breakpoint in malloc_error_break to debug /usr/include/sys/errno.h : #define ENOMEM 12 /* Cannot allocate memory */ I'd give vmmap a try and look for a VM fragmentation issue. Eric _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… On May 9, 2008, at 4:16 PM, Sean McBride wrote: I'm trying to find more details about this message. Specifically, what is error code 12? In my case, it is a 32 bit process using 60 MB of 'real memory' and 1.2 GB of 'virtual memory' (says Activity Monitor). For a mmap() the "real" (aka resident) memory is unlikely to be relevant. I'm trying to determine if 1) the process' virtual memory is just really fragmented and it really can't allocate 16.8 MB, or 2) I have some kind of memory smasher or other bug. This email sent to site_archiver(a)lists.apple.com

1 0

iPhone SDK Beta 5 Related Sources
by William Siegrist 10 May '08

10 May '08

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com http://www.opensource.apple.com/darwinsource/iPhoneSDKBeta5/ ---- William Siegrist Darwin Team _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com The latest beta of the iPhone SDK included new versions of gcc and gdb. You can find their sources at the usual location: smime.p7s

1 0