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

Re: pthreads and CF runloops (possibly slightly OT)
by appledisclose＠batchmaker.de 12 Nov '05

12 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: handmade ! User-agent: MacSOUP/D-2.7 (Mac OS X version 10.3.9) Garth Cummings <gcummings(a)apple.com> wrote: > I can guess that they're trying to point out that the Cocoa threading > model, for example, can explicitly run any code on the main thread > even if initiated from another thread. If you are using naked > pthreads and want to do UI, you'd have to work out your own > rendezvous mechanism with your main thread to have it run the UI. Hoi, I want to jump in here. One thing (apparently more in Aqua ;-) is that I have in mind only to draw from the main thread (aka AppRunLoop). Do I understand the above right in the way that a secondary thread could append some source to the main loop and let the main loop then execute some drawing (which it is not allowed in its own thread). For Philip, I experimented with the runloops some time ago. My "feel easy with pthreads" skeleton goes a simple way. In the pthread parameter the thread gets some preallocated storage (don't alloc persistent from the thread this brings trouble on dp machines vs. cpu cache). In the threads funtion I get the master pointer, the threads runloop and some other vals I want to have in any thread. Since the threads storage is prealloced I now can dig that storage and install sources in any runloop the apps has. And of course dealloc them. Like this you don't need bitflags in your global storage. Especially toggle menu items (checked or not) can be handled safely in the main loop with iterating installed handlers in a secondary runloop. Oh yes, and on onet hing you can rely also: pthread_cancel them from the main loop and the secondary runloops will automaticalley be stopped tidyly. You just need to zero your stored references. -- Hado Hein, Berlin, Fed.Rep. of Germany sip: ++49/30/916488488 http://www.batchmaker.de (software authoring, including stage lighting) _______________________________________________ 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: pthreads and CF runloops (possibly slightly OT)
by Philip Lukidis 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.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

1 0

Re: pthreads and CF runloops (possibly slightly OT)
by Garth Cummings 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.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

1 0

RE: pthreads and CF runloops (possibly slightly OT)
by Philip Lukidis 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi Shawn, thanks for your reply. I'll answer inline. Philip Lukidis > -----Original Message----- > From: Shawn Erickson [mailto:shawn@freetimesw.com] > Sent: Thursday, November 10, 2005 12:35 PM > To: Philip Lukidis > Cc: darwin-dev(a)lists.apple.com > Subject: Re: pthreads and CF runloops (possibly slightly OT) > > > > On Nov 10, 2005, at 9:07 AM, Philip Lukidis wrote: > > > Is this true? > > To be honest I am not sure I understand that statement in the > developer documentation... It may make sense to file a doc defect to > get that statement better clarified, it is a little lose in > terminology (I assume "your application's run loop" means your > applications main thread's runloop). > > > When I create a pthread, is a new run loop created, and is > > the pthread associated with it? > > Likely no because I believe they are created lazily but that is an > implementation detail (one that I am guessing at myself). > Lazy creation is OK, as long as it is created when I ask for it via CFRunLoopGetCurrent() (and it seems to be). > > I ask this because I've done a simple (naive) test of outputting > > the address of my main runloop, and then > > outputting the address of a pthread runloop, and they are different. > > They should be different since they are different threads. Each > thread can get its own runloop which is basically a thread > local data > structure. > Yes, I was hoping for them to be different. My initial understanding was that no runloop would ever be created for pthreads, but thankfully this is quite false. > > I've also modified my simple daemon to receive client messages in > > the created > > pthread, which seems to work. > > Can you better explain how you did this so we can help make sure we > (and you) understand things? I ask given the following question. > Well, instead of waiting for messages from main(), I did so from my pthread, using CFRunLoopGetCurrent() in order to get the runloop. This worked OK, and is what I was hoping for. > > So the question is, are pthreads associated with a runloop, > and, if > > not, how > > can I create a runloop and associate a pthread with it? > > In this case when running under your pthread simply use > CFRunLoopGetCurrent() but if in Cocoa/Carbon use their API for > runloops. Of course that only gets you a valid reference for the > runloop associated with the current thread (possibly > "creating" it if > needed). To actually have the runloop do what it does you > need to run > it. > Yes, this is what my simple server app showed me. > <http://developer.apple.com/documentation/CoreFoundation/Conceptual/ > CFRunLoops/Tasks/Running.html> > > -Shawn > Thanks for that link, I'll read up more about runloops. Philip Lukidis _______________________________________________ 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: pthreads and CF runloops (possibly slightly OT)
by Philip Lukidis 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.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

1 0

Re: pthreads and CF runloops (possibly slightly OT)
by Shawn Erickson 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Nov 10, 2005, at 9:07 AM, Philip Lukidis wrote: When I create a pthread, is a new run loop created, and is the pthread associated with it? -Shawn _______________________________________________ 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 true? To be honest I am not sure I understand that statement in the developer documentation... It may make sense to file a doc defect to get that statement better clarified, it is a little lose in terminology (I assume "your application's run loop" means your applications main thread's runloop). Likely no because I believe they are created lazily but that is an implementation detail (one that I am guessing at myself). I ask this because I've done a simple (naive) test of outputting the address of my main runloop, and then outputting the address of a pthread runloop, and they are different. They should be different since they are different threads. Each thread can get its own runloop which is basically a thread local data structure. I've also modified my simple daemon to receive client messages in the created pthread, which seems to work. Can you better explain how you did this so we can help make sure we (and you) understand things? I ask given the following question. So the question is, are pthreads associated with a runloop, and, if not, how can I create a runloop and associate a pthread with it? In this case when running under your pthread simply use CFRunLoopGetCurrent() but if in Cocoa/Carbon use their API for runloops. Of course that only gets you a valid reference for the runloop associated with the current thread (possibly "creating" it if needed). To actually have the runloop do what it does you need to run it. <http://developer.apple.com/documentation/CoreFoundation/Conceptual/ CFRunLoops/Tasks/Running.html> This email sent to site_archiver(a)lists.apple.com

1 0

Re: pthreads and CF runloops (possibly slightly OT)
by Garth Cummings 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.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

1 0

pthreads and CF runloops (possibly slightly OT)
by Philip Lukidis 10 Nov '05

10 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Sorry if this is slightly OT, but I could not find a better list. Hi, I have starting to think about the architecture of a server, which will be a daemon (it will enumerate all connected instances my hardware plugged into USB, among other things). Currently, a single threaded quick prototype server receives requests from clients through CFMessagePort. Now my background is Windows drivers (and usermode to some degree), so forgive the naivety of my questions. I've been reading up on thread implementations, and I saw something related to pthreads which alarmed me. I thought that each time a thread was created it had a runloop associated with it. My server program requires this, I believe, because runloops are used with CFMessagePort. However, I've read that a "POSIX thread does not automatically associate itself with your application's run loop" (see link below, sorry for the wrapping). http://developer.apple.com/documentation/Cocoa/Conceptual/Multithreading/ind ex.html?http://developer.apple.com/documentation/Cocoa/Conceptual/Multithrea ding/Multithreading.html Is this true? When I create a pthread, is a new run loop created, and is the pthread associated with it? I ask this because I've done a simple (naive) test of outputting the address of my main runloop, and then outputting the address of a pthread runloop, and they are different. I've also modified my simple daemon to receive client messages in the created pthread, which seems to work. So the question is, are pthreads associated with a runloop, and, if not, how can I create a runloop and associate a pthread with it? Sorry if this is OT for this group... thanks, Philip Lukidis _______________________________________________ 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: Active DHCP interfaces
by Dan Bernstein 09 Nov '05

09 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=WnwCmiqkVtXCBGOTm7mYFO7iYP/LaY1kK4279sh93Jua/NVA76ZfHPyrtysBhG0Df4hnihXTEfTYIPBU5JutYC7DrQgcUCTCavx6EDjLYFRCk6T29ciNyBlSF/evmHfGKetMnFTUYsNWQ99YtqGccXtc2+Ox4940q9oN+b1LiJ8= If you mean "which interfaces are currently configured with DHCP" then I think you can use the System Configuration framework. See <http://developer.apple.com/documentation/Networking/Reference/SysConfig/SCD…> On 11/7/05, matt jaffa <mjaffa(a)gmail.com> wrote: > Hi, > > Is there any way to find out the active interfaces using DHCP > programatically? > > Thanks, > Matt > _______________________________________________ > 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/dan.bernstein%40gmail.com > > This email sent to dan.bernstein(a)gmail.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: 64-bit question
by Mike Fischer 09 Nov '05

09 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 08.11.2005 um 21:03 schrieb Sam Hart <samh(a)granada-learning.com>: I think it is a bug in the 32bit handling too, albeit less dramatic. The process b0rked with a bus error rather than an "Out Of Memory". Here is a more correct version of said code: #include <stdio.h> #include <string.h> #include <stdlib.h> #define BUFFER_SIZE (1024*1024) int main(int argc, const char * argv[]) { long long int i = 0, total = 0; while (++i) { char *bp = malloc(BUFFER_SIZE * sizeof(char)); return 0; } Does this version crash with a bus error as well? Mike -- Mike Fischer Softwareentwicklung, EDV-Beratung Schulung, Vertrieb Address: Bundesstrasse 9, D-20146 Hamburg, Germany Phone: +49 (0)40/45038886, Fax: +49 (0)40/45038887 _______________________________________________ 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… Given the source snippet Bob originally posted on this list I think it might simply be one of the following bugs in his code: - The stack gets overwritten when i > ARRAY_SIZE. - It tries to copy from NULL after malloc failed. - The strcpy function will copy an unspecified amount of data because buf is not a \0 terminated C string. Depending what comes after buf on the stack this will overwrite more or less dynamic memory which might also cause a crash. strncpy or memset would have been better. Oh and some other non crashing bugs or oddities in that code are: - The display of the allocated memory dosn't include the initial allocation of bigbuf so it is off by about 8MB (for 64-bit, or 4MB for 32-bit). - The choice of 1000000 is somewhat arbitrary, given that the page size of the VM system is 4096 bytes. I'd use 1MB or 1024*1024. if (bp == NULL) { printf("Internal malloc failed!\n"); break; // When this happens then the situation will not get better. Give up right away. } else { memset(bp, 'A', BUFFER_SIZE * sizeof(char)); // Access the memory so the VM system will actually allocate // physical memory, not just logical address space. // Note: it would be enough to touch one byte of every memory // page but the speed difference will not justify the more // complicated code as the VM thrashing will hide any // small perfomance gains made here. total = i; // Keep track of the total. if (i % 1000 == 0) // Don't print too much. { printf("MB allocated: %lld\n",i); } } } printf("Done! Total MB allocated: %lld\n", total); while (1) { } // So it doesn't exit! Really bad style to use an infinite loop but ok for this test! This email sent to site_archiver(a)lists.apple.com

1 0