May 2005 - Darwin-dev - lists.apple.com

Re: Bizarre out of memory bug
by Justin Walker 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Wow. On May 25, 2005, at 3:40, Tim Seufert wrote: [mightygorilla:~] tas% ssh yobgorgle Password: Welcome to Darwin! [yobgorgle:~] tas% ls tcsh: Out of memory. Connection to yobgorgle closed. [mightygorilla:~] tas% Truly strange. Have you tried changing the hostname? :-} A couple of thoughts: - did you also check the console log? - have you tried logging into an account with 'bash' or 'sh' as the default shell? Me too. [snip] Cheers, Justin -- Justin C. Walker, Curmudgeon at Large Institute for General Semantics ----------- My wife 'n kids 'n dog are gone, I can't get Jesus on the phone, But Ol' Milwaukee's Best is my best friend. ----------- _______________________________________________ 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… So I'm on my Powerbook (10.4.1), and SSH across the good old domestic WLAN to my G5 (Also 10.4.1). I get behavior like this: This tugs at a few loose memory strands. I have a vague recollection of similar behavior, perhaps with tcsh, having to do with something in the "rc" files in users' home directories. Bizarre thing #1: the G5 named "Yobgorgle" had roughly 2GB free RAM at the time of this event. Physical memory certainly isn't a problem. The only justification for the error in question is if some piece of code tried to allocate more address space than the kernel was willing to give a single process. Bizarre thing #2: So, "yobgorgle" is really "yobgorgle.local" since I have "local" in my search domains list. In other words, it's a Ren^h^h^h Bonjour name. On a whim, I decided to SSH to the actual IP address: [snip] Wow. I just don't know what to say. Anybody have any suggestions before I file a bug? Only to try to remove tcsh, or its startup files, from the picture. This is a beaut. This email sent to site_archiver(a)lists.apple.com

1 0

Re: GCC 4.0 version
by Justin Walker 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 25, 2005, at 1:04, Joaquín Mª López Muñoz wrote: (Sorry if this is off-topic; if so, could you please redirect me to the appropriate forum?) You have the right list. Inspecting the source code of the latest Darwin release (8.1), I see that the GCC 4.0 version shipped is 4.0.0 20041026 (Apple Computer, Inc. build 4061) Are there plans to upgrade this? The official GNU version 4.0 includes bugfixes commited after 2004/10/26. I'm asking because Darwin 8.1 is 6 months newer than the GCC code it ships with. Cheers, Justin -- Justin C. Walker, Curmudgeon-At-Large Institute for General Semantics -------- Men are from Earth. Women are from Earth. Deal with it. -------- _______________________________________________ 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 has been observed on numerous occasions. You should be aware of a few things about the list though: - this is *not* an official channel to Apple - Apple guys to contribute to these lists, but essentially as individuals who have knowledge that may be useful to the rest of us, not as representatives of Apple - bug reports go to <http://bugreporter.apple.com> - specific pieces of technology may have mail addresses (not discussion lists) for feedback to Apple. Often, this is for use during pre-release testing, where the technology is under NDA - Apple rarely announces product update content and schedules, and almost never on these lists. This email sent to site_archiver(a)lists.apple.com

1 0

fsevents oddities (was Re: EVFILT_VNODES?)
by Hamish Allan 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi Mike, firstshell$ mdfind -live "kMDItemFSName=='hamishtest*'" [Type ctrl-C to exit] ........Then in another shell I type........ secondshell$ touch hamishtest1 ........I get an update in the first shell........ Query update: 1 matches ........Then if I type........ secondshell$ ln hamishtest1 hamishtest2 ^C firstshell$ mdfind -live "kMDItemFSName=='hamishtest*'" /Users/hamish/hamishtest1 /Users/hamish/hamishtest1 [Type ctrl-C to exit] firstshell$ mdfind -live "kMDItemFSName=='hamishtest*'" /Users/hamish/hamishtest1 [Type ctrl-C to exit] firstshell$ sudo fslogger-1.0-beta secondshell$ touch foo secondshell$ ln foo bar secondshell$ rm foo secondshell$ rm bar [firstshell stdout, edited for brevity:] # Event type = CREATE FILE VNODE 18 path = /Users/hamish/foo # Event type = CREATE FILE STRING 17 string = /Users/hamish/bar # Event type = DELETE STRING 18 string = /Users/hamish/foo # Event type = DELETE STRING 18 string = /Users/hamish/foo Thanks and best wishes, Hamish On 25 May 2005, at 07:41, Mike Smith wrote: Can anyone tell me how it works? Or could anyone recommend an efficient method for determining when a file in a given directory or any of its subdirectories changes, for a potentially large number of subdirectories? As a general rule, there isn't one. You're making a common mistake in associating "file" with "location". = Mike This email sent to hamish.allan(a)gold.ac.uk _______________________________________________ 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… Thank you for your detailed reply. You are right that I have a problem, but Spotlight has it too! For instance, if I initiate the following query........ ........I get nothing in the first shell, but if I restart the query I get........ ........If I then use the second shell to remove hamishtest1, no update, but if I run it again........ ........So now I have a file called hamishtest2, no file called hamishtest1, and a query which returns hamishtest1! You can see a related behaviour using Amit Singh's fslogger utility (thank you to Sam for pointing this out), which uses the new fsevents API: It looks like fsevents is performing caching, as rebooting between "rm foo" and "rm bar" causes the second DELETE to have the correct string = /Users/hamish/bar. It looks as though the first CREATE FILE might be setting this up, as it seems to create a relationship between VNODE and path (whereas the hard link event just has STRING / string). But I don't really understand why this is the case -- surely a DELETE refers not to a vnode, but to an index entry, in which case the correct string ought to be able to be reported? Presumably there is an inverted index (from vnode to path) for every vnode on the filesystem. This must be populated at startup time, as Spotlight returns hamishtest2 for my mdfind query after a reboot. But the behaviour would suggest that although it can be added to for a fresh vnode, the entry for an existing vnode cannot be updated. And the entry is only cleared when the link count reaches zero. I think that if the inverted index allowed multiple paths to be stored for a vnode, the correct behaviour could be achieved. Of course, some of my assumptions may not be correct. Any thoughts? This email is quite long already, but I'll give you a brief outline of my application. You may remember that I posted to this list recently about writing a Spotlight filesystem (http://lists.apple.com/ archives/darwin-dev/2005/May/msg00229.html). I thought it might be possible to get similar functionality without writing a VFS by having a daemon monitor the creation of new directories (either in a particular place, or with Spotlight- searchable attributes) and fill them with hard links to Spotlight search results. To preserve the pathname of the files found, I would create the full directory structure within the given folder; I therefore wanted a way to monitor whether that structure had changed, in order to halt any live updating of results and turn it into a 'normal' folder I should be fine with the existing fsevents API for this, because my daemon process would always create new directories rather than link to existing ones. File paths are indices used to locate files. There is no guarantee that there is exactly one index for a file, nor is there any way to efficiently enumerate all of the indices for a file. Once a file is open, there is a direct relationship between the descriptor and the vnode that does not require the existence of an index; this is how you can delete open files. This means that when a file is modified, there's no guarantee that you know "where" the file may be looked up; nor where it was looked up by whomever is modifying it; nor where the user "thinks" it is. These may be three different locations. Spotlight makes a number of specific assumptions about the notifications that it receives, and it is interested in a view of files that sounds very different from yours. Without knowing a lot more about you application, I can't help you much more; perhaps you could describe what you're trying to achieve? _______________________________________________ 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/hamish.allan% 40gold.ac.uk This email sent to site_archiver(a)lists.apple.com

1 0

Re: (nevermind) which source is ioctl() implemented in?
by Philip George 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com nevermind found what i needed elsewhere - philip _______________________________________________ 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

which source is ioctl() implemented in?
by Philip George 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com which source is ioctl() implemented in? thanks. - philip _______________________________________________ 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

Bizarre out of memory bug
by Tim Seufert 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com [mightygorilla:~] tas% ssh yobgorgle Password: Welcome to Darwin! [yobgorgle:~] tas% ls tcsh: Out of memory. Connection to yobgorgle closed. [mightygorilla:~] tas% Wow. I just don't know what to say. Nothing appears in system.log or /cores, by the way. Anybody have any suggestions before I file a bug? _______________________________________________ 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… So I'm on my Powerbook (10.4.1), and SSH across the good old domestic WLAN to my G5 (Also 10.4.1). I get behavior like this: Some shell commands do succeed, but most fail identically to this example with ls. Bizarre thing #1: the G5 named "Yobgorgle" had roughly 2GB free RAM at the time of this event. Physical memory certainly isn't a problem. The only justification for the error in question is if some piece of code tried to allocate more address space than the kernel was willing to give a single process. Bizarre thing #2: So, "yobgorgle" is really "yobgorgle.local" since I have "local" in my search domains list. In other words, it's a Ren^h^h^h Bonjour name. On a whim, I decided to SSH to the actual IP address: [mightygorilla:~] tas% ssh 192.168.1.40 Password: Welcome to Darwin! [yobgorgle:~] tas% ls Applications/ Documents/ Movies/ Public/ bin/ Desktop/ Library/ Pictures/ Temporary Items/ [yobgorgle:~] tas% Bizarre thing #3: Everything seems to work normally when I SSH in the opposite direction, regardless of whether I use the Bonjour name or the IP address of the PowerBook. The only remotely relevant difference in configuration which I can think of is that the PowerBook gets its IP address via DHCP whereas the G5 has a static IP. My brain hurts. Why should the shell (or alternately the programs it's invoking) care how the network name lookup for setup of the SSH session it's communicating through took place? And how can this behavior happen on one machine and not the other? This email sent to site_archiver(a)lists.apple.com

1 0

GCC 4.0 version
by Joaquín Mª López Muñoz 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com (Sorry if this is off-topic; if so, could you please redirect me to the appropriate forum?) Inspecting the source code of the latest Darwin release (8.1), I see that the GCC 4.0 version shipped is 4.0.0 20041026 (Apple Computer, Inc. build 4061) Are there plans to upgrade this? The official GNU version 4.0 includes bugfixes commited after 2004/10/26. I'm asking because Darwin 8.1 is 6 months newer than the GCC code it ships with. Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo _______________________________________________ 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

OpenAfs on Darwin 8.0.0
by Ido Levy 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi All I am trying to install OpenAfs on Darwin 8.0.0 OS which reside on mac Power PC platform. If someone has experience with this issue, I would appreciate if you share this information with me. Thanks in advance Ido Levy Linux Support and Services Servers and Network Group IBM Haifa Research Lab, Israel Office +972 4 8281082 Internet e-mail : idol(a)il.ibm.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

Re: EVFILT_VNODES?
by Mike Smith 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Can anyone tell me how it works? Or could anyone recommend an efficient method for determining when a file in a given directory or any of its subdirectories changes, for a potentially large number of subdirectories? As a general rule, there isn't one. You're making a common mistake in associating "file" with "location". = 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… File paths are indices used to locate files. There is no guarantee that there is exactly one index for a file, nor is there any way to efficiently enumerate all of the indices for a file. Once a file is open, there is a direct relationship between the descriptor and the vnode that does not require the existence of an index; this is how you can delete open files. This means that when a file is modified, there's no guarantee that you know "where" the file may be looked up; nor where it was looked up by whomever is modifying it; nor where the user "thinks" it is. These may be three different locations. Spotlight makes a number of specific assumptions about the notifications that it receives, and it is interested in a view of files that sounds very different from yours. Without knowing a lot more about you application, I can't help you much more; perhaps you could describe what you're trying to achieve? This email sent to site_archiver(a)lists.apple.com

1 0

Re: EVFILT_VNODES?
by Sam Vaughan 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Also see <http://www.kernelthread.com/software/fslogger/>. Sam On 25/05/2005, at 10:25 AM, Sam Vaughan wrote: Hi Hamish, Sam On 25/05/2005, at 5:04 AM, Hamish Allan wrote: Hi all, As I understand it, for the contents of a directory to be monitored using a kqueue, it must be open (at least O_EVTONLY). Can anyone tell me how it works? Or could anyone recommend an efficient method for determining when a file in a given directory or any of its subdirectories changes, for a potentially large number of subdirectories? Thanks, Hamish _______________________________________________ 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 think you can receive notifications on vnode operations using the kauth API, which is new in Tiger. I just searched developer.apple.com for "kauth" and didn't turn up much, but there looks to be some meat in sys/kauth.h. Therefore, given that the mds process detects whenever any file on the entire filesystem is written, it seems doubtful that it uses kqueue. This email sent to site_archiver(a)lists.apple.com

1 0