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

Re: Trylocks in the kernel
by Ryan McGann 14 Dec '05

14 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Dec 13, 2005, at 9:21 AM, Brian Bergstrand wrote: On Dec 13, 2005, at 12:23 AM, Ryan McGann wrote: Ryan _______________________________________________ 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… The new Tiger locking KPIs seem to lack a trylock variant for any of the lock types (mutex, spinlock, read/write). A mutex try variant is available in the Unsupported KPI. It's probably safe to use this for 8.0, but may break in the next major release. All of the try variants are in the kernel proper, so one has to wonder why they are not exported in the KPI. Well they also have a way to promote read-write locks to exclusive from shared exported in the kernel, but not in the KPIs (lck_rw_exclusive_to_shared and lck_rw_shared_to_exclusive). That bug was closed :-( It would be really nice to have those KPIs back in the supported list, if anybody is listening... This email sent to site_archiver(a)lists.apple.com

1 0

Re: Trylocks in the kernel
by Ryan McGann 14 Dec '05

14 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Dec 13, 2005, at 9:44 AM, Josh Graessley wrote: Thanks, Ryan On Dec 12, 2005, at 10:23 PM, Ryan McGann wrote: This email sent to jgraessley(a)apple.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… It's not that you can't block in your data in filter. You want to be careful about what you block on and how long it is likely to be blocked. For example, if you were to take a lock where you enqueue data and then drop it, that would be fine as long as the only other code taking the lock is something that takes the lock to dequeue the data so it can process it after dropping the lock. Thanks for this clarrification Josh. I don't feel so bad now taking the lock, since it should never be locked for more than a couple milliseconds (worse case). It still would be nice to have a trylock though. An example of a bad place to block would be writing the data out another socket to a user process and waiting for a response. Another bad example would be attempting to take a lock that some other piece of code holds while calling in to the network stack to perform an operation that may block. You need to evaluate the other places where the lock you will be taking is held and make sure that lock isn't held while some other long term or interdependent operation is performed. Doing so may lead to deadlocks. I take the lock inside of dataIn and dataOut to protect access to my queue of buffered data. I take it also when I need to reinject the code. As long as sock_reinject_* doesn't block, all of the statements you made about my code should hold. Basically I need to do the same thing as the tcplognke sample, and I see that sample takes a mutex inside of dataIn as well. I don't feel bad now :-) The new Tiger locking KPIs seem to lack a trylock variant for any of the lock types (mutex, spinlock, read/write). I'm trying to figure out how to get around this. I need to take a read-write lock in a socket filter dataIn callback, but of course you cannot block inside dataIn for any length of time. So I _was_ going to use a trylock, and if I failed to get the lock, just bail out. But it looks like the only way to accomplish what I want to do is to use lck_rw_sleep_deadline. However I'm not exactly sure of the semantics of this API. If I give it a deadline of "now", will it still perform a context switch regardless of whether the lock is taken or not? I don't want the expense of a context switch if the lock isn't even going to be taken. Is there any way of duplicating the behavior of a trylock using the Tiger KPIs? Or am I stuck using lck_rw_sleep_deadline? And can that API context switch on me without taking the lock? Thanks, Ryan rmcgann(a)mac.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/jgraessley% 40apple.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: Using [uw]tmp
by Jason Townsend 14 Dec '05

14 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Graham, If you change to use sizeof(struct utmp) then this works as expected. -Jason #import <Foundation/Foundation.h> #import <utmp.h> #import <stdlib.h> #import <stdio.h> extern int errno; int main (int argc, const char * argv[]) { NSAutoreleasePool *pool=[[NSAutoreleasePool alloc] init]; FILE *utmp_file; struct utmp *utmp_store; _______________________________________________ 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 Dec 13, 2005, at 7:38 AM, Graham J Lee wrote: I want to get login information for users (basically my goal is a Bonjour-enabled equivalent to the old GNU or ICSI distributed finger server[*], where any systems on the local network register themselves with the server so I can find out if a user has logged in anywhere on the local network) and as neither of the distributed fingers come anywhere near close to compiling on Darwin (8.3.0) have tried to start from scratch. Having observed that /var/run/ utmpx is empty, and that there's no getutent on Darwin, I'm doing it the same way that programs like w(1) do and reading struct utmps from the appropriate file. However, what I get back doesn't seem consistent with the idea that utmp(5) [or indeed wtmp(5)] is just a bunch of struct utmps stored back to back. Here's my code: It looks like your problem is that you're using sizeof on a pointer, which is going to be 4, rather than the correct size, so you're looking at the file four bytes at a time. errno=0; if((utmp_file=fopen(_PATH_UTMP,"rb"))==NULL) { perror("csalutd: couldn't open utmp"); exit(EXIT_FAILURE); } if((utmp_store=(struct utmp *)malloc(sizeof(utmp_store)))==NULL) { perror("csalutd: couldn't malloc struct"); exit(EXIT_FAILURE); } while(fread(utmp_store,sizeof(utmp_store),1,utmp_file)==1) { printf("That fool %s, over %s way, was responsible.\nIt was at % d on %s.\n",utmp_store->ut_name,utmp_store->ut_host,utmp_store- >ut_time,utmp_store->ut_line); } fclose(utmp_file); [pool release]; return 0; } This email sent to site_archiver(a)lists.apple.com

1 0

smartcard info
by Dave Corbett 14 Dec '05

14 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AcYAS1od0Hg+B4N0Rj6EnYaj3sRxjg== _______________________________________________ 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: Trylocks in the kernel
by Josh Graessley 13 Dec '05

13 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -josh On Dec 12, 2005, at 10:23 PM, Ryan McGann wrote: This email sent to jgraessley(a)apple.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 It's not that you can't block in your data in filter. You want to be careful about what you block on and how long it is likely to be blocked. For example, if you were to take a lock where you enqueue data and then drop it, that would be fine as long as the only other code taking the lock is something that takes the lock to dequeue the data so it can process it after dropping the lock. An example of a bad place to block would be writing the data out another socket to a user process and waiting for a response. Another bad example would be attempting to take a lock that some other piece of code holds while calling in to the network stack to perform an operation that may block. You need to evaluate the other places where the lock you will be taking is held and make sure that lock isn't held while some other long term or interdependent operation is performed. Doing so may lead to deadlocks. The new Tiger locking KPIs seem to lack a trylock variant for any of the lock types (mutex, spinlock, read/write). I'm trying to figure out how to get around this. I need to take a read-write lock in a socket filter dataIn callback, but of course you cannot block inside dataIn for any length of time. So I _was_ going to use a trylock, and if I failed to get the lock, just bail out. But it looks like the only way to accomplish what I want to do is to use lck_rw_sleep_deadline. However I'm not exactly sure of the semantics of this API. If I give it a deadline of "now", will it still perform a context switch regardless of whether the lock is taken or not? I don't want the expense of a context switch if the lock isn't even going to be taken. Is there any way of duplicating the behavior of a trylock using the Tiger KPIs? Or am I stuck using lck_rw_sleep_deadline? And can that API context switch on me without taking the lock? Thanks, Ryan rmcgann(a)mac.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/jgraessley% 40apple.com smime.p7s

1 0

Re: Trylocks in the kernel
by Brian Bergstrand 13 Dec '05

13 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Dec 13, 2005, at 12:23 AM, Ryan McGann wrote: I've had a radar open on this issue for 6 months now: rdar://4136250 Haven't used this, so I can't comment. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (Darwin) iD8DBQFDnwMledHYW7bHtqIRAiD8AKCLJVeDumv1SmQ6uiShPvnfZekZwgCg8y6E AnSgcvbfzyx1qh0Ol84mYfU= =uHtB -----END PGP SIGNATURE----- _______________________________________________ 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… The new Tiger locking KPIs seem to lack a trylock variant for any of the lock types (mutex, spinlock, read/write). A mutex try variant is available in the Unsupported KPI. It's probably safe to use this for 8.0, but may break in the next major release. All of the try variants are in the kernel proper, so one has to wonder why they are not exported in the KPI. I believe Sam V. from SGI opened one too. You should open your own too. It will be marked as duplicate, but in the backwards world of non-public access bugs, more duplicates means more attention. I'm trying to figure out how to get around this. I need to take a read-write lock in a socket filter dataIn callback, but of course you cannot block inside dataIn for any length of time. So I _was_ going to use a trylock, and if I failed to get the lock, just bail out. But it looks like the only way to accomplish what I want to do is to use lck_rw_sleep_deadline. This email sent to site_archiver(a)lists.apple.com

1 0

Using [uw]tmp
by Graham J Lee 13 Dec '05

13 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi, #import <Foundation/Foundation.h> #import <utmp.h> #import <stdlib.h> #import <stdio.h> extern int errno; int main (int argc, const char * argv[]) { NSAutoreleasePool *pool=[[NSAutoreleasePool alloc] init]; FILE *utmp_file; struct utmp *utmp_store; [Session started at 2005-12-13 15:32:15 +0000.] That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on cons. That fool , over way, was responsible. It was at 0 on ole. That fool , over way, was responsible. It was at 0 on leeg. That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on . That fool , over way, was responsible. It was at 0 on \242\213\305X. How should I read struct utmp records from the utmp file? Thanks, Graham. -- Graham Lee GPG Key ID: 01D5B9D8 UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 01865 273450 _______________________________________________ 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 want to get login information for users (basically my goal is a Bonjour-enabled equivalent to the old GNU or ICSI distributed finger server[*], where any systems on the local network register themselves with the server so I can find out if a user has logged in anywhere on the local network) and as neither of the distributed fingers come anywhere near close to compiling on Darwin (8.3.0) have tried to start from scratch. Having observed that /var/run/utmpx is empty, and that there's no getutent on Darwin, I'm doing it the same way that programs like w(1) do and reading struct utmps from the appropriate file. However, what I get back doesn't seem consistent with the idea that utmp(5) [or indeed wtmp(5)] is just a bunch of struct utmps stored back to back. Here's my code: errno=0; if((utmp_file=fopen(_PATH_UTMP,"rb"))==NULL) { perror("csalutd: couldn't open utmp"); exit(EXIT_FAILURE); } if((utmp_store=(struct utmp *)malloc(sizeof(utmp_store)))==NULL) { perror("csalutd: couldn't malloc struct"); exit(EXIT_FAILURE); } while(fread(utmp_store,sizeof(utmp_store),1,utmp_file)==1) { printf("That fool %s, over %s way, was responsible.\nIt was at %d on %s.\n",utmp_store->ut_name,utmp_store->ut_host,utmp_store- >ut_time,utmp_store->ut_line); } fclose(utmp_file); [pool release]; return 0; } The first few lines of output: So there's a lot of 0 going on, and what look like valid data (e.g. leeg is currently logged in on console). I get similar output if I use _PATH_WTMP, by the way. But the data don't seem aligned in the same way that <utmp.h> suggests they would be, especially as there seem to be so many \0 characters all over the place. [*]It's nice of you to be concerned about the security implications of a distributed finger service, I've taken those into consideration and still need such a service kthx. This email sent to site_archiver(a)lists.apple.com

1 0

General Guidelines on bulk data through Posix Sockets
by Mark Gilbert 13 Dec '05

13 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Folks. I have an app which communicates to another remote app via a posix socket. Any particular tricks ? pitfalls ? Thanks Mark Gilbert. -- mark(a)gallery.co.uk Tel: +44 208 340 5677 fax: +44 870 055 7790 http://www.gallery.co.uk New Product ! - MetaCorder Location Sound Field Recorder http://www.metacorder.info _______________________________________________ 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… Previously the communication has been very small messages, up to 10 bytes or so, back and forth. Now one app is requesting a piece of larger data from the other app (30k Bytes), and I have noticed that the communication is a less smooth. In particular, whereas previously I could read 1 byte blocking, then continue reading until 'read' returned-1 (which would be the end of the reply), I now see a return of -1 part way through the 30k reply, and it picks up again later. Before I rework this code to deal with this, I wanted to get any general guidelines on the best way to handle exchange of larger packets of data via sockets. This email sent to site_archiver(a)lists.apple.com

1 0

Trylocks in the kernel
by Ryan McGann 13 Dec '05

13 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thanks, Ryan rmcgann(a)mac.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… The new Tiger locking KPIs seem to lack a trylock variant for any of the lock types (mutex, spinlock, read/write). I'm trying to figure out how to get around this. I need to take a read-write lock in a socket filter dataIn callback, but of course you cannot block inside dataIn for any length of time. So I _was_ going to use a trylock, and if I failed to get the lock, just bail out. But it looks like the only way to accomplish what I want to do is to use lck_rw_sleep_deadline. However I'm not exactly sure of the semantics of this API. If I give it a deadline of "now", will it still perform a context switch regardless of whether the lock is taken or not? I don't want the expense of a context switch if the lock isn't even going to be taken. Is there any way of duplicating the behavior of a trylock using the Tiger KPIs? Or am I stuck using lck_rw_sleep_deadline? And can that API context switch on me without taking the lock? This email sent to site_archiver(a)lists.apple.com

1 0

Re: preprocessor definition identifying Mac OS X version
by David Fang 12 Dec '05

12 Dec '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com > > Is there somewhere (presumably among the headers in /usr/include) a > > preprocessor definition I can use to know what version of Mac OS X > > some C source code is being compiled on ? > > > > Thanks, > > Aurelien Hi, You could run 'uname -r -s' and save the result to a header file, like: echo #define OS \"`uname -r -s`\" > os.h There's always the autoconf option, but this is quick and simple. David Fang _______________________________________________ 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