December 2006 - Darwin-dev - lists.apple.com

Re: further help on debugging a hang
by Giuliano Gavazzi 15 Dec '06

15 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Giuliano _______________________________________________ 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 have one more data point that puzzle me and that might indicate a way to debug this. The system that hangs, presumably for some raid card/drive problems, while stopped performing an HTTP GET from cron (to a remote server b), managed 15 to 20 minutes later to send an ntp packet to the same server b. (I have pretty good ipfw logs!) So, while cron got stuck, presumably while calling curl from disk, ntp must have been not swapped out and managed to do part of his job. I wonder if a kernel extension could not use remote syslog to send debugging messages.. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Preventing system sleep, QA1340 code does not seem to work
by Ethan Bold 14 Dec '06

14 Dec '06

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: Preventing system sleep, QA1340 code does not seem to work
by Sean McBride 14 Dec '06

14 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: Rogue Research On 2006-12-14 11:04, Ethan Bold said: > >On Dec 13, 2006, at 1:41 PM, Sean McBride wrote: > >> But does 'idle sleep' include the display? That is, if in the >> callback >> I refuse idle sleep will both the display and cpu stay awake? > >No, idle sleep does not include the display. Using the API's per >QA1340 will prevent system sleep, but will allow the display to sleep Thanks for this info. I also need to stop the display sleeping so the QA1340 technique is no good anyway (phew! since I can't get it to work). >normally. It sounds like this might be what you're seeing - the >display sleeping while the system remains awake. I don't think so, because I don't even hit the callback at all. And both my display and system sleep settings are at 1 minute. >So you want to keep the display awake? The short answer is that it's >pretty ugly API to do so; the even shorter answer is that all you have >to do is call the Carbon.framework call UpdateSystemActivity() every >25 seconds. This generates a fake HID UI event, like a mouse move, >frequently enough to prevent idle display sleep AND idle system sleep. Is 25 seconds better? QA1160 uses 30 seconds. I hate magic numbers. :) >Note that the display must idle sleep before the system can idle >sleep. This is a restriction of PM's idle sleep behavior; so if you >prevent disply sleep with NULL HID activity (via >UpdateSystemActivity()), the system will never idle sleep either. Perfect. The UpdateSystemActivity() is working exactly as I need. Thanks! -- ____________________________________________________________ Sean McBride, B. Eng sean(a)rogue-research.com Rogue Research www.rogue-research.com Mac Software Developer Montréal, Québec, Canada _______________________________________________ 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: Preventing system sleep, QA1340 code does not seem to work
by Ethan Bold 14 Dec '06

14 Dec '06

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: ifconfig weirdness due to modification.
by Josh Wisenbaker 14 Dec '06

14 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Importance: Normal User-agent: SquirrelMail/1.4.6-rc1 On Mon, December 11, 2006 6:36 pm, matt jaffa wrote: > Hi, > > I am attempting to include some form of ifconfig.c in my own app. I > compile > it into a lib and then include it in my main code. > When I call main with my args everything is fine the first time. But every > time I call it afterwords when it trys to set the IP address > of an interface it returns ENOMEM. It is working on the first call > everytime > no matter what I do, but the second call to it fails on setting > the new ip address but everything else gets set correctly: e.x. broadcast, > netmask etc. > > Why I am including ifconfig in my app is just for fun. Why it is having > this > weird behavior is probably cause I am doing this weird thing. > Anyone possibly might know why this could be happening? No idea, but Mac OS X is gonna stomp all over your ifconfig changes anyway when configd feels like it. <http://www.afp548.com/article.php?story=20041015131913324> Josh -- Josh Wisenbaker U, U, D, D, L, R, L, R, B, A, Start for your server http://www.afp548.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: gdb isn't looking at DYLD_LIBRARY_PATH - NO frameworks involved.
by Mike Logan 13 Dec '06

13 Dec '06

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: gdb isn't looking at DYLD_LIBRARY_PATH - NO frameworks involved.
by Graham J Lee 13 Dec '06

13 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 13 Dec 2006, at 23:04, Mike Logan wrote: - DYLD_LIBRARY_PATH is set properly. - (gdb) show env gives correct results. - the dylibs do exist and are freshly built. but I don't understand what its telling me. The first line is the wrong path. Why the second line (the correct path) has no address? Cheers, Graham. -- Graham J Lee http://www.thaesofereode.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… Hi, this may be an easy question to answer but much searching of the lists hasn't given me anything... If you have suggestions, please help - I'm out of ideas. I have a bunch of dylib's and the program starts up fine in a terminal shell & Xcode, but in gdb(shell or xcode), it cannot find the libraries (message is below). I tried to "info sharedlibrary all" in gdb, and got this output this particular dylib: 6 libProducer.dylib - - c-init Y Y /Users/ mlogan/dev/KrishnaKumar/NelFly/bin/libProducer.dylib at 0x0 (objfile is) /Users/mlogan/dev/ Producer_1.1/src/Darwin32.Opt/libProducer.dylib Stab in the dark, but does "otool -D /my/problematic/libRary.dylib" have the wrong path? This email sent to site_archiver(a)lists.apple.com

1 0

gdb isn't looking at DYLD_LIBRARY_PATH - NO frameworks involved.
by Mike Logan 13 Dec '06

13 Dec '06

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: Preventing system sleep, QA1340 code does not seem to work
by Sean McBride 13 Dec '06

13 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: Rogue Research On 2006-12-12 14:55, Ethan Bold said: >To be clear, the kIOMessageCanSystemSleep message only allows you to >prevent IDLE sleep. If you close the clamshell or sleep via the Apple >menu, your callback will never receive the kIOMessageCanSystemSleep >message. There is no way to prevent non-idle sleeps. That's ok, I'll take what I can get. :) But does 'idle sleep' include the display? That is, if in the callback I refuse idle sleep will both the display and cpu stay awake? >That said, yes, this code should work in 10.3.9, 10.4.8, and 9A303. I >use it all the time. > >* Are you getting any error returns from any of the IOKit or CF calls >in main()? No. >* Where are your breaks/printfs/asserts - are you breaking on the >invocation of your callback function itself? Or only in the >kIOMessageCanSystemSleep code (which will not get called on a force >sleep)? Right at the start of the callback (MySleepCallBack()). >Yep. I'm stumped. I've put the project (and binary) here if anyone is interested. I've trivially modified it from what's on the page. I'm stumped too! <http://www.rogue-research.com/vtk/SleepTest.zip> -- ____________________________________________________________ Sean McBride, B. Eng sean(a)rogue-research.com Rogue Research www.rogue-research.com Mac Software Developer Montréal, Québec, Canada _______________________________________________ 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: Network Bandwidth
by Stewart.Dickson＠noaa.gov 13 Dec '06

13 Dec '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Priority: normal Please see (for example) web100 based Network Diagnostic Tool (NDT) http://miranda.ctd.anl.gov:7123/ Generally, it tells you where the bottlenecks are and the reasons. It usually comes down to small buffers set for your 1000-Base-T card in your kernel. I don't know, but perhaps someone else on the list can point you to procedures for increasing your Network I/O buffer size in Darwin. Hope this helps. -Stewart http://us.imdb.com/Name?Stewart+Dickson Principal Scientific Subject Matter Expert in Visualization STG Inc., Government Contractor Graphics Department, Office of the Director NOAA National Climatic Data Center 151 Patton Avenue Asheville, NC 28801-5001 <Stewart.Dickson(a)noaa.gov> (828)271-4688 Lecturer, Department of Computer Science, UNC-Asheville http://www.cs.unca.edu/~dickson Research Associate, National Environmental Modeling and Analysis Center http://www.unca.edu/nemac Sculptor, http://emsh.calarts.edu/~mathart/MathArt_siteMap.html _______________________________________________ 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/stewart.dickson%40noaa.gov This email sent to stewart.dickson(a)noaa.gov _______________________________________________ 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