January 2009 - Darwin-dev - lists.apple.com

GetTickCount on the mac
by Christopher Khaljah Xu 30 Jan '09

30 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi there! I'm a rookie on a mac. I've been trying to port some windows programs that was written in C/C++ onto the mac. I need to rewrite a timer that measures CPU ticks between the start and end of an operation. I wonder if there is something in the mac api similar as GetTickCount or GetTickCount64 ? Any help will be appreciated. _______________________________________________ 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: Max ops per semop call
by Terry Lambert 30 Jan '09

30 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -- 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 Jan 29, 2009, at 10:57 AM, Rohitash Panda <rohitash.panda(a)oracle.com> wrote: Folks, Is there a sysctl parameter that can be changed , to set the limit on the number of operations in a single semop() call ? Is it configurable ? No. The value is hard-coded to MAX_SOPS, which is currently defined in sem.h as 5; also note the comment in that file which indicates the value is informational and subject to change in the future. If you need more than that, you will need to make multiple calls. Note that even if you limit yourself, the operations you group will only be individually atomic, not group idempotent (POSIX does not guarantee it), so there's no benefit other than reduced number of system calls to grouping a larger number of them together. Given that semaphore operations are themselves pretty heavyweight, the amortized overhead costs of extra calls for a larger number of them is proportionally small. This email sent to site_archiver(a)lists.apple.com

1 0

Re: sysctl semantics for KERN_PROC_PID
by Steve Checkoway 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 29, 2009, at 11:04 AM, A.M. wrote: if(kill(somePid,0)==0) { //process for pid was running //who knows if it still is Comment corrected above. <http://en.wikipedia.org/wiki/TOCTTOU> -- Steve Checkoway "Anyone who says that the solution is to educate the users hasn't ever met an actual user." -- Bruce Schneier _______________________________________________ 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 } smime.p7s

1 0

Re: sysctl semantics for KERN_PROC_PID
by A.M. 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Jan 29, 2009, at 1:50 PM, Rohitash Panda wrote: sysctl(mib, 4, &kp, &len, NULL, 0); Hello, Just use if(kill(somePid,0)==0) { //process for pid is running } You have a man 2 kill Cheers, M _______________________________________________ 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… Hi, I am trying to check the existence of a process through sysctl ( CTL_KERN ) . I am using the below construct : mib[0] = CTL_KERN; mib[1] = KERN_PROC; mib[2] = KERN_PROC_PID; mib[3] = pid; In the case that the pid is not around , what's the documented return value for sysctl , -1 or 0 , for Mac OS X 10.5 ? Is the behaviour same for Mac OS X 10.4 too ? This email sent to site_archiver(a)lists.apple.com

1 0

Max ops per semop call
by Rohitash Panda 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: Oracle Corporation User-agent: Thunderbird 2.0.0.19 (Windows/20081209) Thanx. Rohitash _______________________________________________ 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… Folks, Is there a sysctl parameter that can be changed , to set the limit on the number of operations in a single semop() call ? Is it configurable ? This email sent to site_archiver(a)lists.apple.com

1 0

sysctl semantics for KERN_PROC_PID
by Rohitash Panda 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: Oracle Corporation User-agent: Thunderbird 2.0.0.19 (Windows/20081209) mib[0] = CTL_KERN; mib[1] = KERN_PROC; mib[2] = KERN_PROC_PID; mib[3] = pid; sysctl(mib, 4, &kp, &len, NULL, 0); Thanks. Rohitash _______________________________________________ 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… Hi, I am trying to check the existence of a process through sysctl ( CTL_KERN ) . I am using the below construct : In the case that the pid is not around , what's the documented return value for sysctl , -1 or 0 , for Mac OS X 10.5 ? Is the behaviour same for Mac OS X 10.4 too ? This email sent to site_archiver(a)lists.apple.com

1 0

Documentation for values of Previous Sleep and Shutdown Cause
by Jeff Johnson 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com kernel[0]: Previous Sleep Cause: 0 kernel[0]: Previous Shutdown Cause: 5 -Jeff _______________________________________________ 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 there somewhere in the Darwin source that I can find documentation for the values of these (from system.log)? This email sent to site_archiver(a)lists.apple.com

1 0

Preparing an IOMemoryDescriptor: driver or user client?
by Joel Reymont 29 Jan '09

29 Jan '09

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=gamma; h=domainkey-signature:received:received:message-id:from:to :content-type:content-transfer-encoding:mime-version:subject:date:cc :x-mailer; bh=tjHoryCnQqhHjm4fzjT5Wn3HLdd0DPyKSZnMnfFcaLc=; b=TOI+eIKcpP1PxdNKVi7eDkD5+X8cr4ssl/8j7QoPxhBbQfgvsLk7wxWb1rbWwvRJ80 NGE5Mx5XFi3gEWtWncju+T2oOqiSR9LUwJcKgC0NI7Z0TYmL++sg8cWbRXsWXCtWgmi7 R2a+Mba3SylxS0yKdlepLEG5fGPSCLmTUpXPI= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:from:to:content-type:content-transfer-encoding :mime-version:subject:date:cc:x-mailer; b=I/ZQs91KSDYvEgsFuyi3MdiRFPOlcMY4IO2aahc/Qa9m6inUY7aG+wdozb6ZyZbWKq 1nPCcE25IeUrfkDrF6zmwC7u03GQahhXVool22BNBZXERxvZjF3SYWEzokxo49xAuBIZ rV8iXRD6bkGGiSQEfXKUXnd1d++Dxu30GuCVY= Where am I supposed to prepare/complete an IOMD? Thanks, Joel -- http://wagerlabs.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 is in the context of passing a buffer from user land into a driver via a user client. This email sent to site_archiver(a)lists.apple.com

1 0

RE: Easiest way to get function name/offset from a hex address?
by Karan, Cem (Civ, ARL/CISD) 29 Jan '09

29 Jan '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AcmBkKyHb4EqD9+PTTu05qaW/Mq9DgAjGxxg Thread-topic: Easiest way to get function name/offset from a hex address? Terry Lambert wrote on Wednesday, January 28, 2009 4:37 PM > On Jan 28, 2009, at 11:15 AM, Karan, Cem (Civ, ARL/CISD) wrote: > > I'm using -finstrument-functions to figure out what the stream of > > calls is that causes my program to crash. Since -finstrument- > > functions instruments everything with calls to > > __cyg_profile_func_enter() & __cyg_profile_func_enter(), I have the > > hex address of the function that is being called, and the > site where > > it is being called from. I'd like to convert that information to > > printable strings so I know the name of the function and the offset > > from the start of the function (or, better yet, the line > number, but > > I'll live with the offset). Is there a simple way of doing this > > lookup? I looked into nlist(), and it appears to have some of the > > information I need, but I don't see how to get the function > names from > > the information it gives me. > > > > Any help is appreciated. > > > > Within the program itself: > > man dladdr > > Outside the program itself: > > man atos > > The debugger, gdb, also has multiple modes in which it > operates, two of which can be programmatically controlled (to > permit things like GUIs that still use gdb internally). You > could open the binary with gdb in one of these modes and > interactively query gdb for the information. Can you tell me what these modes are? Or at least what to look for in the manual? > If the intention is from within the instrumentation, dladdr > will be your best bet. Funnily enough, that was one of the things I tried. Earlier, I couldn't get my program to link when I had it in there. Now it linked. I'm starting to think someone is trying to drive me crazy... > If the intention is post-mortem analysys, your best bet is > going to be atos. > > If the intention is interactive external analysis (for > example, if you are recording non-statistical profiling arcs > in shared memory and analyzing them on the fly), your best > bet is going to be gdb. My first choice is GDB, if I can get it to launch GUI apps correctly. However, for alpha/beta testing, the dladdr trick is probably what I'll need. Crash reports just don't give enough info on how I got into the state I'm currently in. Thanks, Cem Karan _______________________________________________ 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: kernel debug kit 10.5.6
by Joel Reymont 29 Jan '09

29 Jan '09

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=gamma; h=domainkey-signature:received:received:references:message-id:from:to :in-reply-to:content-type:content-transfer-encoding:mime-version :subject:date:cc:x-mailer; bh=03P/hzVgFy/qJjv7cp3bmtWDNm7NP5UQsubHAxqYBPU=; b=dsdptCk0ef0TCNUTKHzj/M1vdAZseMTUnRW/YCZ+tKJtQBZ3IXhTANKtdmvhz24Z1O 5fwR88/ko6pxMti5gFh0ILsLreGRYRFJ1hRvvIkTd3f71KgLnxbiVvxG0cvgino8iZ+u 8yBQqV84klkKSsDErCfWgV480pkcv6/x4tybU= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=references:message-id:from:to:in-reply-to:content-type :content-transfer-encoding:mime-version:subject:date:cc:x-mailer; b=D2cIEEUbv9HTySPaSyhAE0bkQCliH64JS7fbp29aVUBobpVVbXU4d1osskXow1UY/G NG3HsThba8pOU0TlVo6lL4FXshLlHUn9Xoof35P4JDP3RAiGDz25lkGbIRHIViP86pz+ BfcCovyIFk6dg2reiQAlBA2UWypf/EUxpagyM= On Jan 29, 2009, at 1:03 PM, Alexander von Below wrote: Mea culpa. I came upon the dsk directory from a google search. Got the proper debug kit now, thanks! -- http://wagerlabs.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… I do not know where you are looking, but the Kernel Debug Kit for 10.5.6 is right at http://connect.apple.com for me This email sent to site_archiver(a)lists.apple.com

1 0