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

Re: mach in signal handler
by Steve Checkoway 26 Jan '07

26 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5.0.9 (X11/20061206) -- Steve Checkoway _______________________________________________ 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… Dave Zarzycki wrote: I don't remember if Mach based APIs are interruptible by signals, but if they are, then the answer is definitely no. Why? The reply port can get out of synchronization do to the unexpected reentrancy on the same thread stack. If it matters, I would only be calling it from within the signal handler which I don't allow to be reentrant. (If another signal happens while I'm in the handler, I give up and just die.) Personally speaking, I try and handle as many signals as I can via the kqueue API. The only signals that can't be handled in a kqueue are instruction stream or memory access related faults (SIGBUS, SIGSEGV, SIGILL, and SIGFPE). If the aforementioned signals are what you are trying to handle, I'd suggest setting up a dedicated thread to handle signals (man sigsuspend). All other threads should have their signal mask setup to block delivery. What I'm actually doing is attempting to get a back trace on x86 under crash conditions. It's fairly easy on ppc. For x86, I'd like to walk along the stack looking for what might be return addresses, presumably they'd be in executable, readable, but not writable memory and would be proceeded by one of the call instructions. If the code hasn't been compiled with -fomit-frame-pointer, then I'd also like to be able to find valid stack frames by checking that the stored %esp actually points up the stack. I was hoping that vm_region would let me both check what the valid range of memory for a thread's stack and also check if values on the stack point to potential call sites (well, right after the call site) including the protection. I have zero preferences on which api to use to get the job done, so if you can suggest a more suitable one, I'd be more than happy to use it. This email sent to site_archiver(a)lists.apple.com

1 0

Re: mach in signal handler
by Dave Zarzycki 26 Jan '07

26 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com davez On Jan 25, 2007, at 5:38 PM, Steve Checkoway wrote: -- Steve Checkoway _______________________________________________ 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/zarzycki%40apple.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 don't remember if Mach based APIs are interruptible by signals, but if they are, then the answer is definitely no. Why? The reply port can get out of synchronization do to the unexpected reentrancy on the same thread stack. Personally speaking, I try and handle as many signals as I can via the kqueue API. The only signals that can't be handled in a kqueue are instruction stream or memory access related faults (SIGBUS, SIGSEGV, SIGILL, and SIGFPE). If the aforementioned signals are what you are trying to handle, I'd suggest setting up a dedicated thread to handle signals (man sigsuspend). All other threads should have their signal mask setup to block delivery. Is it safe to use vm_region in a signal handler? I know there are a small list of functions that are considered safe <http://developer.apple.com/documentation/Darwin/Reference/ManPages/man2/sig… > and vm_region is not among them. I tried looking for vm_region in the darwin sources but I can't seem to figure out how to search the sources (clearly I'm missing something obvious here) and I can't find it looking where I expected it to be. Some man page leads me to believe that it's a syscall, yet I can't find it in any lists of syscalls in libc. This email sent to zarzycki(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

mach in signal handler
by Steve Checkoway 26 Jan '07

26 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5.0.9 (X11/20061206) -- Steve Checkoway _______________________________________________ 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 it safe to use vm_region in a signal handler? I know there are a small list of functions that are considered safe <http://developer.apple.com/documentation/Darwin/Reference/ManPages/man2/sig…> and vm_region is not among them. I tried looking for vm_region in the darwin sources but I can't seem to figure out how to search the sources (clearly I'm missing something obvious here) and I can't find it looking where I expected it to be. Some man page leads me to believe that it's a syscall, yet I can't find it in any lists of syscalls in libc. This email sent to site_archiver(a)lists.apple.com

1 0

libcc_kext?
by Matt Darland 23 Jan '07

23 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com ---- 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/site_archiver%40lists.app… I'm trying to build the 10.4.8 xnu project. I'm all the way to the final link step which barfs trying to find a libcc_kext. I've found a libcpp_kext project, but can't find any references to libcc_kext. Are they the same, or where can I find libcc_kext? This email sent to site_archiver(a)lists.apple.com

1 0

Network Transfer Speed
by Anees Alappat 23 Jan '07

23 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com 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=ZkhIzb+wflKTu0u0fvEN1X1xA5axVqhjb0DzBfN/Wq+N163yw35hT4SO5S5epf6LhxHuyxk59Ogok9c1AP+FedhwJEtcvwKuo6v41krKCs4RWpVpV0VDymKvOBIhfEvlkvdtpWAXy8H42hmBAbrhvYcJFUel7HklJHq0uF3G2PU= _______________________________________________ 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: stricmp?
by seebs＠plethora.net 20 Jan '07

20 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com In message <45B177A5.60406(a)michael-amorose.com>, Mike writes: >Where is stricmp defined in the Mac OS X 10.4 SDK and what .h file do I >need to include to use it? It's not. strcasecmp, however, is in <strings.h>. Yes, that's different from <string.h>; this is because these functions are not standard C. -s _______________________________________________ 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

stricmp?
by Mike 20 Jan '07

20 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5.0.9 (Macintosh/20061207) Thanks, 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… Where is stricmp defined in the Mac OS X 10.4 SDK and what .h file do I need to include to use it? This email sent to site_archiver(a)lists.apple.com

1 0

Re: e: TASK_BASIC_INFO question (Steve Thompson)
by Steve Thompson 19 Jan '07

19 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Fri, 19 Jan 2007, Michael Smith wrote: This is a critical functionality for my application, so hopefully this will be fixed before /dev/kmem support goes away for good, otherwise I will not be able to continue to support OS X. Steve ---------------------------------------------------------------------------- Steve Thompson E-mail: smt(a)nospam.vgersoft.com Voyager Software LLC Web: http://www.vgersoft.com 39 Smugglers Path VSW Support: support(a)nospam.vgersoft.com Ithaca, NY 14850 "186,300 miles per second: it's not just a good idea, it's the law" ---------------------------------------------------------------------------- _______________________________________________ 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'm not quite sure whether this is hyperbole or blackmail, but it seems a bit flat either way. Sorry, not intended as either. It meant simply that if I don't have the ability to get this information, the functionality of my application will be severely impacted. What are you trying to do? If you have a priori knowledge of these processes, you can wrap them and gather statistics on termination. If you know who they are but cannot be complicit in their invocation, you might consider the acct(2) facility and/or the use of the sa(8) utility. It's actually a component of a batch processing engine. Imagine the situation where a daemon launches a shell script on behalf of another user. We have no control over what that shell does; all we know is the name of the shell and the input file. However, we need to know periodically, while the top-level shell is still running, what its resource usage is so that we can report back to the user on the progress of the job. All I know about the running shell is the PID of the top-level process; by doing a sysctl(CTL_KERN,KERN_PROC,KERN_PROC_ALL) I can trace through the process relationships and work out the total resource consumption, but only if I can get at the proc->p_stats->p_cru fields (and right now I am doing a kvm_read to get this). This application has been run on other platforms for several years, but only recently did I get a request to port it to OS X (Intel and ppc), so I am very new (4 days) to this sort of bsd programming. This email sent to site_archiver(a)lists.apple.com

1 0

e: TASK_BASIC_INFO question (Steve Thompson)
by Michael Smith 19 Jan '07

19 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com = 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… However, it appears that unless one enables the /dev/kmem interface and uses kvm_read(), there is no supported API for an application (running as root) to obtain the resources used by terminated children of an arbitrary pid. That's correct, and not typically unreasonable. Without knowing what that child was, knowing its resource consumption is typically not very useful. This is a critical functionality for my application, so hopefully this will be fixed before /dev/kmem support goes away for good, otherwise I will not be able to continue to support OS X. I'm not quite sure whether this is hyperbole or blackmail, but it seems a bit flat either way. What are you trying to do? If you have a priori knowledge of these processes, you can wrap them and gather statistics on termination. If you know who they are but cannot be complicit in their invocation, you might consider the acct(2) facility and/or the use of the sa(8) utility. This email sent to site_archiver(a)lists.apple.com

1 0

Re: TASK_BASIC_INFO question
by Steve Thompson 19 Jan '07

19 Jan '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Wed, 17 Jan 2007, Steve Thompson wrote: [...] count = TASK_BASIC_INFO_COUNT; task_info (task, TASK_BASIC_INFO, (task_info_t) &tinf, &count); [...] -s ---------------------------------------------------------------------------- Steve Thompson E-mail: smt(a)nospam.vgersoft.com Voyager Software LLC Web: http://www.vgersoft.com 39 Smugglers Path VSW Support: support(a)nospam.vgersoft.com Ithaca, NY 14850 "186,300 miles per second: it's not just a good idea, it's the law" ---------------------------------------------------------------------------- _______________________________________________ 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… Well, I must have been asleep or something when I posted that utter rubbish. That problem is solved. However, it appears that unless one enables the /dev/kmem interface and uses kvm_read(), there is no supported API for an application (running as root) to obtain the resources used by terminated children of an arbitrary pid. This is a critical functionality for my application, so hopefully this will be fixed before /dev/kmem support goes away for good, otherwise I will not be able to continue to support OS X. This email sent to site_archiver(a)lists.apple.com

1 0