March 2010 - Darwin-kernel - lists.apple.com

Re: Possible bug with nanosleep()?
by Chris Wilson 02 Mar '10

02 Mar '10

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com User-agent: Alpine 1.00 (DEB 882 2007-12-20) Hi Michael, On Tue, 2 Mar 2010, Michael Smith wrote: I think you might be a bit confused about how time and timing works. Cheers, Chris. -- _ ___ __ _ / __/ / ,__(_)_ | Chris Wilson <0000 at qwirx.com> - Cambs UK | / (_/ ,\/ _/ /_ \ | Security/C/C++/Java/Perl/SQL/HTML Developer | \ _/_/_/_//_/___/ | We are GNU-free your mind-and your software | _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This looks suspiciously like code one would use to implement a polling loop. That's generally a mistake. I don't think it's a polling loop, I just want to damn well be sure that I sleep for the time requested, regardless of interruption. This function is used in timing-sensitive unit tests. If all you care about is whether at least the requested time has elapsed, you should be calling gettimeofday() once nanosleep returns. Yes, you're right, I should do that and ignore the remaining time returned by nanosleep(). But I still think that the remaining time returned is wrong, and that there's a bug in the nanosleep implementation on OSX. I'm not interested at all in signals in this code, I wish they wouldn't happen, I just want to sleep for the appointed time and nothing else. You can't do that; other threads may run, resulting in you not executing for considerably longer. You can ensure that you sleep for no less than the requested time by adjusting your signal mask appropriately, or looping with a call to gettimeofday to ensure that you have not been woken early. You can't guarantee that your thread will suspend and resume execution at a precise time. I agree, this is not a real-time application. What I meant is that I'd really prefer it if my nanosleep was not interrupted by signals, and if it is, I just want to sleep again until the appointed time. I don't want to mess about with signal masks as nanosleep might (now or in future) be internally implemented using signals on some platforms (I've seen this with sleep()). This email sent to site_archiver(a)lists.apple.com

1 0

Re: Possible bug with nanosleep()?
by Michael Smith 02 Mar '10

02 Mar '10

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Possible bug with nanosleep()?
by Chris Wilson 02 Mar '10

02 Mar '10

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com User-agent: Alpine 1.00 (DEB 882 2007-12-20) Hi Terry and all, For reference, this is the code I have in the application now: void safe_sleep(int seconds) { struct timespec ts; memset(&ts, 0, sizeof(ts)); ts.tv_sec = seconds; ts.tv_nsec = 0; while (nanosleep(&ts, &ts) == -1 && errno == EINTR) { // FIXME evil hack for OSX, where ts.tv_sec contains // a negative number interpreted as unsigned 32-bit // when nanosleep() returns later than expected. int32_t secs = (int32_t) ts.tv_sec; int64_t remain_ns = (secs * 1000000000) + ts.tv_nsec; if (remain_ns < 0) { BOX_WARNING("nanosleep interrupted " << ((float)(0 - remain_ns) / 1000000000) << " secs late"); return; } BOX_TRACE("nanosleep interrupted with " << (remain_ns / 1000000000) << " secs remaining, " "sleeping again"); } } int32_t secs = (int32_t) ts.tv_sec; then it hangs forever on OSX. Cheers, Chris. -- _ ___ __ _ / __/ / ,__(_)_ | Chris Wilson <0000 at qwirx.com> - Cambs UK | / (_/ ,\/ _/ /_ \ | Security/C/C++/Java/Perl/SQL/HTML Developer | \ _/_/_/_//_/___/ | We are GNU-free your mind-and your software | _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Mon, 1 Mar 2010, Terry Lambert wrote: You need to be looking at tr, not ts, or you need to do the structure assign immediately after the EINTR, before your trace statements. The tr contains the remainder time, the ts structure contents are irrelevant after the nanosleep() call. You're BOX_ macros appear to indicate you wanted the remainder time. Sorry, my mistake, my code does that but I edited it somewhat while writing the email and forgot to update that part of the sample that I pasted. And if I change both int32_t to long (or __darwin_time_t, the type used in the structure) on this line: This looks suspiciously like code one would use to implement a polling loop. That's generally a mistake. I don't think it's a polling loop, I just want to damn well be sure that I sleep for the time requested, regardless of interruption. This function is used in timing-sensitive unit tests. I think if you are getting signals often enough for this to be an issue, you'd be better off passing the same address structure in for both parameters and specifying SA_RESTART in the flags field. Actually I originally wrote the code to pass the same structure for both, and was wondering whether that was allowed (it's not documented as supported or unsupported) so I wanted to separate the two to make sure that people wouldn't claim that I was using the function inappropriately and thus ignore the rest of my message. Even then, this is probably a bad use of signals, since multiple signals being sent won't necessarily result in multiple notifications. I'm not interested at all in signals in this code, I wish they wouldn't happen, I just want to sleep for the appointed time and nothing else. Any ideas why I'm seeing (1<<32)-1 in tr.tv_sec after the call, when the call finishes late? This is the actual problem that I'm having. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Possible bug with nanosleep()?
by Terry Lambert 02 Mar '10

02 Mar '10

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Hi all, The code in question does this: void safe_sleep(int seconds) { struct timespec ts, tr; memset(&ts, 0, sizeof(ts)); ts.tv_sec = seconds; ts.tv_nsec = 0; while (nanosleep(&ts, &tr) == -1 && errno == EINTR) { BOX_TRACE("nanosleep interrupted with " << ts.tv_sec << "." << ts.tv_nsec << " secs remaining, sleeping again"); -- Terry _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Feb 28, 2010, at 11:48 AM, Chris Wilson wrote: I'm developing open source software that runs on MacOS among other platforms. I recently discovered a problem where code that works fine on other platforms is hanging indefinitely on OSX 10.6.2. I'm sure it worked in 10.3 on PPC hardware. if (ts.tv_sec >= seconds) { BOX_WARNING("nanosleep returned with junk in " << "struct: " << ts.tv_sec << "." << ts.tv_nsec); return; } ts = tr; /* sleep again */ } } You need to be looking at tr, not ts, or you need to do the structure assign immediately after the EINTR, before your trace statements. The tr contains the remainder time, the ts structure contents are irrelevant after the nanosleep() call. You're BOX_ macros appear to indicate you wanted the remainder time. This looks suspiciously like code one would use to implement a polling loop. That's generally a mistake. I think if you are getting signals often enough for this to be an issue, you'd be better off passing the same address structure in for both parameters and specifying SA_RESTART in the flags field. Even then, this is probably a bad use of signals, since multiple signals being sent won't necessarily result in multiple notifications. Signals are defined as persistent conditions, not events, which means, for example, if you had multiple child processes die in a narrow time window, the last one to happen will be overwriting the siginfo information (assuming you set the SA_SIGINFO flag in the sa_flags field, and used sigaction() rather than signal() so that you are getting siginfo in the first place). This is typically why things that intend to reap child process exit status loop calling waitpid(-1, &statusvar, WNOHANG) until it returns -1 with an errno of ECHILD. If this isn't the specific case, then you'd be a lot better off avoiding signals, and using a reliable IPC mechanism instead. This email sent to site_archiver(a)lists.apple.com

1 0

wifi
by overneinthousand＠yahoo.com 01 Mar '10

01 Mar '10

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Received:X-Yahoo-SMTP:X-YMail-OSG:X-Yahoo-Newman-Property:X-rim-org-msg-ref-id:Message-ID:Content-Transfer-Encoding:Reply-To:X-Priority:Sensitivity:Importance:To:Subject:From:Date:Content-Type:MIME-Version; b=eqzQmB/vmTVCUzq9toAztJ/tYrFMPndQ5OejlJxID7Gz3LL2IT+rfSiy7JDxVSQldnfPlQOSf5Qkh4wjW5iPSDhep7YqEDrhYCKi5DI00bXDwm2KwMpO62/1LfdIlCov2tGXr+WHMbVkXJ+B5WY/6vxf3Zzfr8f2YM6ejUmkNoY= ; Importance: Normal Sensitivity: Normal Sorry if that last message got throught (human error). Ok, I've been trying for a while within the darwin kernel to connect to my wireless network, but with no luck. Are there any particular daemons that I should use? And if so, how do I access them? all help is appreciated. Sent on the Sprint® Now Network from my BlackBerry® _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

wifi
by overneinthousand＠yahoo.com 01 Mar '10

01 Mar '10

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Received:X-Yahoo-SMTP:X-YMail-OSG:X-Yahoo-Newman-Property:X-rim-org-msg-ref-id:Message-ID:Content-Transfer-Encoding:Reply-To:X-Priority:Sensitivity:Importance:To:Subject:From:Date:Content-Type:MIME-Version; b=xUoF9I/Cs0sa52jeQviClSie9rKzF3KJKMcwB3O7urx6DaKY+ttqKK1Lc69I3FNmrEqfX6ffSomn9oG0uErphd+qxjZztX06/dkZDqLUYjZ4f8Z9G15upicWoxcaxatiFiSWxXuchGXXLEXRoniD+xWZnl1C0Jzwzi5+jLygooE= ; Importance: Normal Sensitivity: Normal Sent on the Sprint® Now Network from my BlackBerry® _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0