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

Re: Set process' working set size explictly
by Maxim Zhuravlev 23 Mar '09

23 Mar '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:mime-version:received:in-reply-to:references :date:message-id:subject:from:to:cc:content-type :content-transfer-encoding; bh=f3sbSHaobO3GoT1xRHOiYeG62nn5ZGSPNojeQ9AWXmM=; b=D/hgj//zIT11zHbuzowY3ULNXHx7S0+yrbFdeU8eRZntHGQBkUedN6VDUO1/zUeopU y3pZ/wCj6XKnjPY7wxQodddczwgazkL3ti8KIl+JMyFuYs3DavDtB3iMeTW4XcK1YH1t p67zVrJ+y4r5y8G5fdySlzdJeXyg3aFm5tucs= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; b=bV/K/6WisuL1Z90+60309m+rtLStCRwgqB5P7NJYNvpc81aWlkEfkGmR1m9LSZ39wi yxr5Pmt6tkkI3wBvivzQ0ANN8ASfHcmB/5FX8kjjVss+FUk7JK9WTw91FgRomE7Nb4DT McfCYGwp3tboJ8Bn2+8MjuTjVi3vZxJxSg3Vw= Michael, thanks a lot for all the information and your time. -- - Maxim _______________________________________________ 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: /etc/ethers not consulted by ether_hostton() ?
by Terry Lambert 22 Mar '09

22 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -- Terry On Mar 22, 2009, at 2:57 AM, Marc Balmer <marc(a)msys.ch> wrote: Am 22.03.2009 um 08:13 schrieb Terry Lambert: - Marc Balmer _______________________________________________ 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 know for a fact that there is code to support flat file lookup and I even identified it for you by file name and pointed you at the source code. Yeah, there are also module plugins for netinfo and DS and theoretically it's even possible to make other plugins. I'm sure you could also add Open Directory records, and (probably) plist files, to do the same thing. That no one has given you specific steps to activate flat file support probably only means you asked the question on a Friday. On Mar 20, 2009, at 6:49 AM, Marc Balmer wrote: While porting a piece of software from OpenBSD to Mac OS X 10.5.6, I found that /etc/ethers seems not to be read by ether_hostton(). I created an /etc/ethers file with a few entries, but neither my software nor e.g. tcpdump finds those entries. Does anybody here have some information on this? If this is the wrong mailing list to ask, pls tell me which one would be more appropriate. This appears to be handled in SEL_NAME case in the file lookup.subproj/file_module.c in the LibInfo project, which is part of libc/libSystem. Maybe you are formatting the contents incorrectly for parsing, and need to look at the source for the parse routine? Alternately, since it's in the file-module of LibInfo, maybe you have to do some sort of configuration dance to ensure that it's being consulted (I would have no idea what). I expect enumation by ether_line() would work in any case, assuming your file format was correct, since there are no explicit comparison items used in that case, as it's an iterator. Good luck (and before you ask, I have no idea what sort of dance would be necessary, you'll need to trace the code paths to find out). Thanks for your reply. In the meantime, with the help of an opendarwin developer, I found out the right steps. In short, /etc/ethers seems not to be supported on Mac OS X at all and all man pages mentioning seem to be just plain wrong. To get it working, a netinfo Ethernets database has to be built and then ether_hostton() and friends will use that information. I described this on http://mac.vnode.ch/wake.html . This email sent to site_archiver(a)lists.apple.com

1 0

Re: /etc/ethers not consulted by ether_hostton() ?
by Marc Balmer 22 Mar '09

22 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 22.03.2009 um 08:13 schrieb Terry Lambert: - Marc Balmer _______________________________________________ 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 Mar 20, 2009, at 6:49 AM, Marc Balmer wrote: While porting a piece of software from OpenBSD to Mac OS X 10.5.6, I found that /etc/ethers seems not to be read by ether_hostton(). I created an /etc/ethers file with a few entries, but neither my software nor e.g. tcpdump finds those entries. Does anybody here have some information on this? If this is the wrong mailing list to ask, pls tell me which one would be more appropriate. This appears to be handled in SEL_NAME case in the file lookup.subproj/file_module.c in the LibInfo project, which is part of libc/libSystem. Maybe you are formatting the contents incorrectly for parsing, and need to look at the source for the parse routine? Alternately, since it's in the file-module of LibInfo, maybe you have to do some sort of configuration dance to ensure that it's being consulted (I would have no idea what). I expect enumation by ether_line() would work in any case, assuming your file format was correct, since there are no explicit comparison items used in that case, as it's an iterator. Good luck (and before you ask, I have no idea what sort of dance would be necessary, you'll need to trace the code paths to find out). Thanks for your reply. In the meantime, with the help of an opendarwin developer, I found out the right steps. In short, /etc/ethers seems not to be supported on Mac OS X at all and all man pages mentioning seem to be just plain wrong. To get it working, a netinfo Ethernets database has to be built and then ether_hostton() and friends will use that information. I described this on http://mac.vnode.ch/wake.html . This email sent to site_archiver(a)lists.apple.com

1 0

Re: /etc/ethers not consulted by ether_hostton() ?
by Terry Lambert 22 Mar '09

22 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Does anybody here have some information on this? If this is the wrong mailing list to ask, pls tell me which one would be more appropriate. -- 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 Mar 20, 2009, at 6:49 AM, Marc Balmer wrote: While porting a piece of software from OpenBSD to Mac OS X 10.5.6, I found that /etc/ethers seems not to be read by ether_hostton(). I created an /etc/ethers file with a few entries, but neither my software nor e.g. tcpdump finds those entries. This appears to be handled in SEL_NAME case in the file lookup.subproj/ file_module.c in the LibInfo project, which is part of libc/libSystem. Maybe you are formatting the contents incorrectly for parsing, and need to look at the source for the parse routine? Alternately, since it's in the file-module of LibInfo, maybe you have to do some sort of configuration dance to ensure that it's being consulted (I would have no idea what). I expect enumation by ether_line() would work in any case, assuming your file format was correct, since there are no explicit comparison items used in that case, as it's an iterator. Good luck (and before you ask, I have no idea what sort of dance would be necessary, you'll need to trace the code paths to find out). This email sent to site_archiver(a)lists.apple.com

1 0

/etc/ethers not consulted by ether_hostton() ?
by Marc Balmer 20 Mar '09

20 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Does anybody here have some information on this? If this is the wrong mailing list to ask, pls tell me which one would be more appropriate. Marc Balmer, mbalmer(a)openbsd.org _______________________________________________ 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… While porting a piece of software from OpenBSD to Mac OS X 10.5.6, I found that /etc/ethers seems not to be read by ether_hostton(). I created an /etc/ethers file with a few entries, but neither my software nor e.g. tcpdump finds those entries. This email sent to site_archiver(a)lists.apple.com

1 0

Re: JNI dependent library symbols
by Terry Lambert 19 Mar '09

19 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -- Terry Appreciate any helpful comments. Thanks. _______________________________________________ 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/tlambert%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… You may want to ask this question on a mailing list that deals with Java, such as the java-dev mailing list, instead of cross-posting to the darwin-dev and unix-porting lists, which normally have nothing to do with Java. On Mar 18, 2009, at 11:03 PM, Rohitash Panda wrote: Hi, I have some confusion regarding how dependent libraries of a native library are treated on Mac OS X . I am on Mac OS X Leopard version 10.5.4 with JDK 1.4.2 I have a native library A which has a dependent library B . I see that, even if the JNI library A gets loaded and is able to successfuly load C , some sumbols which I should be seeing and which are present in library B don't seem to be exported/un-available. Using the verbose jni option provided I saw that after the native library A is loaded by JAVA, the JNI symbols directly exported from library A are resolved without any issues, but those are expected from library B are not getting resolved . Is this a known problem with this JDK version ? Since the symbols from the dependent libraries of the native library don't get exported in this case , then I guess the right way to fix this is to change the linker directives when building the jnilib . I am already using the flat-namespace option for Unix style loading/ resolving of symbols . Going through the man pages of ld , I found a couple of options which seem to be of relevance to the current context : -reexport-lx This is the same as the -lx but specifies that the all symbols in library x should be avail- able to clients linking to the library being created. This was previously done with a sepa- rate -sub_library option. -reexport_library path_to_library This is the same as listing a file name path to a library on the link line and it specifies that the all symbols in library path should be available to clients linking to the library being created. This was previously done with a separate -sub_library option. I am looking for comments from people who have faced this issue earlier. Also , to add to this, I don't seem to hit upon this issue using JDK 1.5. This email sent to tlambert(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

JNI dependent library symbols
by Rohitash Panda 19 Mar '09

19 Mar '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) Appreciate any helpful comments. Thanks. _______________________________________________ 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 have some confusion regarding how dependent libraries of a native library are treated on Mac OS X . I am on Mac OS X Leopard version 10.5.4 with JDK 1.4.2 I have a native library A which has a dependent library B . I see that, even if the JNI library A gets loaded and is able to successfuly load C , some sumbols which I should be seeing and which are present in library B don't seem to be exported/un-available. Using the verbose jni option provided I saw that after the native library A is loaded by JAVA, the JNI symbols directly exported from library A are resolved without any issues, but those are expected from library B are not getting resolved . Is this a known problem with this JDK version ? Since the symbols from the dependent libraries of the native library don't get exported in this case , then I guess the right way to fix this is to change the linker directives when building the jnilib . I am already using the flat-namespace option for Unix style loading/resolving of symbols . Going through the man pages of ld , I found a couple of options which seem to be of relevance to the current context : -reexport-lx This is the same as the -lx but specifies that the all symbols in library x should be avail- able to clients linking to the library being created. This was previously done with a sepa- rate -sub_library option. -reexport_library path_to_library This is the same as listing a file name path to a library on the link line and it specifies that the all symbols in library path should be available to clients linking to the library being created. This was previously done with a separate -sub_library option. I am looking for comments from people who have faced this issue earlier. Also , to add to this, I don't seem to hit upon this issue using JDK 1.5. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Powerdown notification
by Ethan Bold 18 Mar '09

18 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -Ethan On Mar 18, 2009, at 1:14 PM, Brian Bergstrand wrote: On Mar 18, 2009, at 8:07 PM, Ethan Bold wrote: Hi Matt- What about the following constants: #define kIOMessageCanSystemPowerOff iokit_common_msg(0x240) #define kIOMessageSystemWillPowerOff iokit_common_msg(0x250) #define kIOMessageSystemWillNotPowerOff iokit_common_msg(0x260) They exist in the headers so one assumes they will be sent. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 _______________________________________________ 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… Though your assumption is reasonable, those constants don't work. IOKit doesn't send out messages for those. IOKit does not provide any notifications of system shutdown. The IORegisterForSystemPower() API call that you make below only refers to system sleep & wake power events, not shutdown events - the argument messageType may be kIOPMMessageCanSystemSleep, kIOPMMessageSystemWillSleep, kIOPMMessageSystemWillNotSleep, kIOPMMessageSystemHasPoweredOn, kIOPMMessageSystemWillPowerOn. That's all - no messages sent on shutdown. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Powerdown notification
by Ethan Bold 18 Mar '09

18 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com -Ethan On Mar 18, 2009, at 1:14 PM, Brian Bergstrand wrote: On Mar 18, 2009, at 8:07 PM, Ethan Bold wrote: Hi Matt- What about the following constants: #define kIOMessageCanSystemPowerOff iokit_common_msg(0x240) #define kIOMessageSystemWillPowerOff iokit_common_msg(0x250) #define kIOMessageSystemWillNotPowerOff iokit_common_msg(0x260) They exist in the headers so one assumes they will be sent. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 _______________________________________________ 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… Your assumption is reasonable, but those constants don't work. IOKit doesn't send out messages for those. IOKit does not provide any notifications of system shutdown. The IORegisterForSystemPower() API call that you make below only refers to system sleep & wake power events, not shutdown events - the argument messageType may be kIOPMMessageCanSystemSleep, kIOPMMessageSystemWillSleep, kIOPMMessageSystemWillNotSleep, kIOPMMessageSystemHasPoweredOn, kIOPMMessageSystemWillPowerOn. That's all - no messages sent on shutdown. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Powerdown notification
by Brian Bergstrand 18 Mar '09

18 Mar '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Mar 18, 2009, at 8:07 PM, Ethan Bold wrote: Hi Matt- What about the following constants: #define kIOMessageCanSystemPowerOff iokit_common_msg(0x240) #define kIOMessageSystemWillPowerOff iokit_common_msg(0x250) #define kIOMessageSystemWillNotPowerOff iokit_common_msg(0x260) They exist in the headers so one assumes they will be sent. Brian Bergstrand <http://www.bergstrand.org/brian/> PGP Key ID: 0xB6C7B6A2 _______________________________________________ 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… IOKit does not provide any notifications of system shutdown. The IORegisterForSystemPower() API call that you make below only refers to system sleep & wake power events, not shutdown events - the argument messageType may be kIOPMMessageCanSystemSleep, kIOPMMessageSystemWillSleep, kIOPMMessageSystemWillNotSleep, kIOPMMessageSystemHasPoweredOn, kIOPMMessageSystemWillPowerOn. That's all - no messages sent on shutdown. This email sent to site_archiver(a)lists.apple.com

1 0