November 2007 - Darwin-dev - lists.apple.com

Re: A change with the -n option for cp in Mac OS X 10.5
by Kevin Van Vechten 06 Nov '07

06 Nov '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com On Nov 5, 2007, at 11:39 AM, Eric Gorr wrote: Yes. I was just wondering if: 1. 10.4.10 had a bug 2. 10.5 has a bug 3. the definition of how the -n option is supposed to work has changed - Kevin [1] <http://www.opengroup.org/onlinepubs/009695399/utilities/cp.html> _______________________________________________ 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… Is this the right place for this question? Under 10.4.10, when executing cp -n, cp would return an exit status of 0 if the file you told it to copy existed or not. Under 10.5, cp will return an exit status of 1 if the file already exists and one uses the -n option. It seems strange that cp would return an exit status of 1 now if the file already exists. After all, the command will do exactly what it is supposed to do and not copy the file. I suppose it was #3. It looks like the exit status was changed due to feedback that returning a zero exit status gives no way to determine whether or not a file was actually copied (<rdar://problem/3624563>). This is a divergence from the FreeBSD behavior in 10.4.10 (FreeBSD introduced the non-standard -n flag). Because -n is non-standard[1], I'd recommend against using it. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Personality?
by Stéphane Sudre 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On lundi, novembre 5, 2007, at 07:59 PM, Nik Gervae wrote: I'm curious why there's a complaint in one case and none in the other. <dict> <key>com.apple.kpi.bsd</key> <string>8.0.0</string> <key>com.apple.kpi.libkern</key> <string>8.0.0</string> <key>com.apple.kpi.mach</key> <string>8.0.0</string> </dict> _______________________________________________ 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 02 Nov, 2007, at 11:07, Stéphane Sudre wrote: Ok, long version of the story. I'm looking at an application that uses 3 kernel extensions: - 2 KPI Kext (one has a dependency on the IOKit kext below) - 1 IOKit/KPI kext. The KPI kexts are loaded through a launchd script with the kexload -v argument. The Info.plist files of both kexts are globally the same (names differ). On Leopard, when I look at system.log, one of KPI is said to not have any personality while there's no complaint for the other. >From the information I gather, I think it would be better if I remove the IOKit kext and merge it into the kext that depends on it. First thing I'll note is that if kextload is run with -v, it's going to print/log basic warnings and items of note. Explicitly loading an IOKit kext that has no personalities is essentially a one-minute no-op, since that kext will normally not match any devices and be automatically unloaded after that time period. Hence, kextload points out the lack of personalities. If one of the "KPI kexts" (I'm not sure what you mean by that term) A Kernel Extension relying only on the KPIs and not on any unsupported API. One that should not face any trouble in theory when the OS is updated and whose OSBundleLibraries value looks like this: This email sent to site_archiver(a)lists.apple.com

1 0

A change with the -n option for cp in Mac OS X 10.5
by Eric Gorr 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Is this the right place for this question? I was just wondering if: 1. 10.4.10 had a bug 2. 10.5 has a bug 3. the definition of how the -n option is supposed to work has changed Under 10.4.10, when executing cp -n, cp would return an exit status of 0 if the file you told it to copy existed or not. Under 10.5, cp will return an exit status of 1 if the file already exists and one uses the -n option. It seems strange that cp would return an exit status of 1 now if the file already exists. After all, the command will do exactly what it is supposed to do and not copy the file. _______________________________________________ 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

Re: Personality?
by Nik Gervae 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com I'm curious why there's a complaint in one case and none in the other. -- Nik Gervae kext management engineer Apple, Inc. _______________________________________________ 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 02 Nov, 2007, at 11:07, Stéphane Sudre wrote: On vendredi, novembre 2, 2007, at 04:29 PM, Shawn Erickson wrote: On Nov 2, 2007, at 5:29 AM, Stéphane Sudre wrote: On jeudi, novembre 1, 2007, at 11:41 PM, Dean Reece wrote: On Nov 1, 2007, at 3:20 PM, Stéphane Sudre wrote: What's a personality supposed to be for a StartupItem or Kernel Extension? OK, but then why does Leopard tells me a kext that is not related to IOKit has no personalities? If the future please provide more information in your posts about the problem or question you have. Your first post was a vague question that left many of guessing at what you wanted to know and now you give us little more to work with on what type kext you are seeing this issue with and what that kext has in its plist, etc. So more information please ... it will get you a better answer and likely more quickly. Ok, long version of the story. I'm looking at an application that uses 3 kernel extensions: - 2 KPI Kext (one has a dependency on the IOKit kext below) - 1 IOKit/KPI kext. The KPI kexts are loaded through a launchd script with the kexload - v argument. The Info.plist files of both kexts are globally the same (names differ). On Leopard, when I look at system.log, one of KPI is said to not have any personality while there's no complaint for the other. >From the information I gather, I think it would be better if I remove the IOKit kext and merge it into the kext that depends on it. First thing I'll note is that if kextload is run with -v, it's going to print/log basic warnings and items of note. Explicitly loading an IOKit kext that has no personalities is essentially a one-minute no- op, since that kext will normally not match any devices and be automatically unloaded after that time period. Hence, kextload points out the lack of personalities. If one of the "KPI kexts" (I'm not sure what you mean by that term) has a dependency on the IOKit kext, that IOKit kext will be loaded as a library with no concern for personalities, and the warning shouldn't occur in that scenario; the explicit load of the IOKit kext is unnecessary. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Window-server access bug in parallel programming ... ?
by Terry Lambert 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com First, compile and link -g. -- 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 Nov 4, 2007, at 9:47 PM, Eric Schlegel <ericsc(a)apple.com> wrote: On Nov 4, 2007, at 6:12 PM, Jay Reynolds Freeman wrote: What I see is a failure in thread zero of my app -- that's the one where GUI I/O is done, always deep inside a display or displayIfNeeded of my main window. The actual crash is an EXC_BAD_ACCESS with code KERN_PROTECTION_FAILURE in a function called szone_free, in libSystem.B.dylib. There have been many good suggestions in this thread which might isolate your problem. But in mybexperience, when you hear hoofbeats, it's best to look for horses. Rather than add to the list of places where you should look, though, why don't you just ask the code? Second, before doing anything in the program, right at the top of main (), use setrlimit() to set the core dump size to unlimited for cur. Then when the program crashes, run gdb vs. The -g binary and the core file, type bt, and it will list the calls leading up to the crash, including the source code. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Window-server access bug in parallel programming ... ?
by Eric Schlegel 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Nov 4, 2007, at 6:12 PM, Jay Reynolds Freeman wrote: What I see is a failure in thread zero of my app -- that's the one where GUI I/O is done, always deep inside a display or displayIfNeeded of my main window. The actual crash is an EXC_BAD_ACCESS with code KERN_PROTECTION_FAILURE in a function called szone_free, in libSystem.B.dylib. -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… This usually indicates heap corruption of some kind, which then causes the malloc code (containing szone_free) to crash later. You should probably try standard approaches for tracking down heap corruption - MallocDebug, libgmalloc, etc. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Window-server access bug in parallel programming ... ?
by Eric Gouriou 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Nov 4, 2007, at 6:12 PM, Jay Reynolds Freeman wrote: [...] What I see is a failure in thread zero of my app -- that's the one where GUI I/O is done, always deep inside a display or displayIfNeeded of my main window. The actual crash is an EXC_BAD_ACCESS with code KERN_PROTECTION_FAILURE in a function called szone_free, in libSystem.B.dylib. So I am wondering if failure to open this Mach port (I don't know what it is actually for, by the way) is in some way causing this extremely rare failure. It's quite unlikely. I'd look for the usual malloc debugging tools (Guard Malloc, various malloc options found in man malloc). Run your code with those debugging help enabled. 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… Crashes in szone_free() tend to be due to memory corruption. Occam's razor says there is a memory smasher somewhere in your code, a malloc/free usage error and/or a race leading to one of those. [...] If anyone is still reading, do you have a sense of whether I am on the right track? Any ideas for a fix? Any ideas for how to instrument and test to see if I can better understand what is going on? [Off topic for the list] Launching multiple copies of an OS X application is rarely a good idea. If possible, I'd recommend one UI process and multiple UI-less processes. But that's not relevant to Darwin-Dev. This email sent to site_archiver(a)lists.apple.com

1 0

Window-server access bug in parallel programming ... ?
by Jay Reynolds Freeman 05 Nov '07

05 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com I am not sure whether this is the best group for my problem; if not, please accept my apologies and, er, tell me where to go ... I have been working on a parallel Lisp implementation (actually, Scheme, which is a dialect of Lisp), in which separate instances of the *same* Macintosh application run at the same time, and share a good deal of memory, both for Scheme object storage and for interprocess communication. If N copies of my application are to be launched, the first one is launched in some normal way, e.g., by mousing on it, and it launches the other N-1 via in essence system("<path to my application>/Contents/MacOS/<my app's name> &"); (Actually, the call uses argv[0] to get the full path, and adds a few flags and operands that have to do with identifying which of the N processes is being started, setting up the mmap, and the like.) To my complete and utter astonishment, this works like gangbusters -- I really do get N complete instances of my application, each with a its own main window, menu bar, and so on. (And by the way, much thanks to Terry Lambert, a few months ago, for hints on why and how to use mmap.) After a time chasing down deadlocks, critical-section violations, and other untoward consequences of parallel processing, I have gotten things to where I begin to see errors that are not obviously the consequence of my own obtuseness and inadequate coding skills. One of them is my subject for tonight. It occurs rarely -- roughly, once in 10000 runs of my application (I am doing lots of regression testing) -- actually, that's once in about 2000 runs of 5 parallel copies of my application at a time; that's about once a day with my Mac running regression tests nearly full-time. I get a crash with a crash log, and I won't bother you with too many details. What I see is a failure in thread zero of my app -- that's the one where GUI I/O is done, always deep inside a display or displayIfNeeded of my main window. The actual crash is an EXC_BAD_ACCESS with code KERN_PROTECTION_FAILURE in a function called szone_free, in libSystem.B.dylib. This is nested some fifteen or twenty function calls inside of anything I myself wrote. (And I should say that my GUI code is written in Cocoa -- my app is model/view/controller, with the view and controller in cocoa and the model a separate thread of straight C++.) (I should also say that I am still working in Tiger, running XCode 2.4.1, Mac OS X 10.4.10, on a 2006 model 13-inch Macbook with an Intel Core-Duo). While contemplating this problem, I remembered something I had read in Dalrymple and Hillegass's "Advanced Mac OS X Programming": "... Mach ports are used for a lot of interprocess communications, particularly to the window server, and are very important to Cocoa." (p. 366) I then noticed that every time I launched N parallel copies of my application, I would get N-1 error messages in the console log, each of the form Perusal of archives suggested that each instance of my application is trying to open a Mach port based on the CFBundleIdentifier in my app's Info.plist, and only the first of the N is able to do so -- the others all find the port name they are looking for in use. The archives I found suggested that the message was harmless, which blindsided me about it for a while, perhaps. Furthermore, the "first of N" of my parallel processes is always the one that does the system calls to open the others, and the processes that have crashed have always been one of the other N-1; that is, one of the ones that did not get the Mach port it wanted. So I am wondering if failure to open this Mach port (I don't know what it is actually for, by the way) is in some way causing this extremely rare failure. If anyone is still reading, do you have a sense of whether I am on the right track? Any ideas for a fix? Any ideas for how to instrument and test to see if I can better understand what is going on? (For the terminally curious, the application in question is "Wraith Scheme", described on my web site, whose URL is in my .sig below, but what's on the web site is *not* the parallel version, that last is still under development.) Thanks much! -- Jay Reynolds Freeman --------------------- Jay_Reynolds_Freeman(a)mac.com http://web.mac.com/jay_reynolds_freeman (personal web site) _______________________________________________ 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… "... CFLog (99): CFMessagePortCreateLocal(): failed to name Mach port ..." This email sent to site_archiver(a)lists.apple.com

1 0

Clamshell notify
by Andrew James 04 Nov '07

04 Nov '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:mime-version:content-type; bh=wZ5blqdfVtnmd4HM65ms8wl7TFfYSfu3rLzWl/EScfQ=; b=AuAtM/Yf1WLh4jKJiwTw6e6rhegyjN6pAqwZsxHTlQOjbsCs2S7yqS0OG5teOZo1NpJnUeTotkteKd5eYTubHfXwtgasdxuK0mzbkU8VWP7tCPbi9/QHJfya//JCGkcu6zxMAzEQBeG0iexB37Lqw0EslafS4UQZApQpMhr5vgo= 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=NATe2YDOzGlCReKLb41v/EKbyGAC+Rde84OKNE4aRYZUWHFA0Q096lTQ9twrET+I5vPXSIKdXzgGkNb79B8ik01E+yasN/Sa1xsML00RTXUoa1chXfuaf2s1UPLyZ6GEqShXXls1Rr1QlA4jgl3CPNMsQ/i8RBusQX9H43nn6pA= _______________________________________________ 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

Re: fcntl questions
by Terry Lambert 04 Nov '07

04 Nov '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Your code should: #include <unistd.h> ... -- 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 Nov 4, 2007, at 9:30 AM, Steve Checkoway <s(a)pahtak.org> wrote: On Nov 4, 2007, at 7:32 AM, Thierry Passeron wrote: The reason why they use SIGRTMIN + x in the linux example is that they use the possibility to get some si_fd extra member in the siginfo_t struct passed as argument to the signal handler installed with sigaction(). And so it is possible to know which descriptor became ready when the signal was sent (no need to use select() in the signal handler) I have not found this si_fd member in the Darwin siginfo_t structure. So how can I proceed ? I know I could use the default signals SIGIO/ SIGURG and use a select() in the signal handler to check which descriptor are available for reading now. I can't answer all of your questions but the only functions you should call in a signal handler are those listed in the sigaction man page. select(2) is not among them. If I were doing this and I really needed to not block in select, I'd either catch the SIGIO signal and set a flag (which is about all you should ever do in a signal handler) which I would check each time through a loop and when it is set, I'd use select with a zero timeout to get the pipes that need to be read from and then read from them. Of course, I don't know how often one gets that signal. I would think it would be easier to either use poll(2) or select(2) each time through the loop to be sure I was catching all of the data that was ready to read. Use a separate thread, if you are unwilling to mux all your events through a workloop or a kqueue or a workqueue. MacOS X does not support POSIX real time signals due to binary compatibility issues, and due to not a lot of perceived gain for our developers other than easing porting of the small amount of code that depends on them. #if defined(_POSIX_REALTIME_SIGNALS) && (_POSIX_REALTIME_SIGNALS >= 200112L) /* my code relying on the new extension here */ ... #else /* my portable code here */ #endif This email sent to site_archiver(a)lists.apple.com

1 0