January 2004 - Darwin-kernel

Re: sosend, soreceive, soconnect
by Justin Walker 20 Jan '04

20 Jan '04

On Monday, January 19, 2004, at 09:05 PM, Matt Jaffa wrote: Amanda and others: I would love to implement what SimpleUserClient is doing but with my application, instead of the KERNEL socket AF_UNIX implementation, I don't know exactly how to do it. You can't use the UserClient interface for your project, since it is designed for IOKit usage, and yours is not an IOKit kext. So I have a KEXT that is a socket/Protocol filter. And I insert my code to do all the checking within the send function that is within my filter. How am i supposed to do this, You need to review kernel code. I don't think there is anyone else with enough time to design and program your project for you. my daemon will keep checking with the KEXT to see if it has work to do? And within my send function i block it with a semaphore??? Until the daemon picks it up, signals then does work and then the daemon goes back to checking for work to do???? As I mentioned earlier, there is code in the 'bsd/netat' directory that does something similar. NFS clients and servers do similar things, but the code is much more complex than the AURP 'daemon' that runs (or, more accurately, ran) in kernel mode. As long as you want to block awaiting incoming datagrams, you need a thread on which to sleep, and therefore you need to create a kernel process. And what if the daemon is not running, it won't be polling for work to do, so the stuff that is sitting in semaphore_wait mode will always wait and nothing will work. If your user-mode component is not running, the kernel-mode component should exit. The AURP code should give a hint regarding the way to do this. See also the NFS client and server code. Please advise me in this area, I am looking to finish my project quickly. You will need to adjust your goals. It doesn't appear that you have much experience with this kind of development, and this is tricky code. What do you have running so far? Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | It's not whether you win or lose... | It's whether *I* win or lose. *--------------------------------------*-------------------------------* _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: sosend, soreceive, soconnect
by Matt Jaffa 20 Jan '04

20 Jan '04

Amanda and others: I would love to implement what SimpleUserClient is doing but with my application, instead of the KERNEL socket AF_UNIX implementation, I don't know exactly how to do it. So I have a KEXT that is a socket/Protocol filter. And I insert my code to do all the checking within the send function that is within my filter. How am i supposed to do this, my daemon will keep checking with the KEXT to see if it has work to do? And within my send function i block it with a semaphore??? Until the daemon picks it up, signals then does work and then the daemon goes back to checking for work to do???? And what if the daemon is not running, it won't be polling for work to do, so the stuff that is sitting in semaphore_wait mode will always wait and nothing will work. Please advise me in this area, I am looking to finish my project quickly. Thanks, Matt On Jan 19, 2004, at 8:06 PM, Amanda Walker wrote: On Jan 19, 2004, at 8:47 AM, Matt Jaffa wrote: I did search the archives, and there was no answer unless i missed the right one. I need to use sosend, and the others in the kernel so that i can communicate with my daemon in user space, so i am doing it in user space. Communicating between the kernel and user space with a UNIX domain socket is not a very good approach. Darwin has a mechanism called "User Clients", which allow an IOKit module to communicate with one or more user space processes. For sample code, see: http://developer.apple.com/samplecode/Sample_Code/ Devices_and_Hardware/IOKit/SimpleUserClient.htm Amanda Walker _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: sosend, soreceive, soconnect
by Amanda Walker 20 Jan '04

20 Jan '04

On Jan 19, 2004, at 8:47 AM, Matt Jaffa wrote: I did search the archives, and there was no answer unless i missed the right one. I need to use sosend, and the others in the kernel so that i can communicate with my daemon in user space, so i am doing it in user space. Communicating between the kernel and user space with a UNIX domain socket is not a very good approach. Darwin has a mechanism called "User Clients", which allow an IOKit module to communicate with one or more user space processes. For sample code, see: http://developer.apple.com/samplecode/Sample_Code/Devices_and_Hardware/ IOKit/SimpleUserClient.htm Amanda Walker _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Problems building xnu-517 or xnu-517.3.7
by Kevin Van Vechten 20 Jan '04

20 Jan '04

I've re-posted the xnu sources for 10.3.2, correcting the issues reported here. The fixes are the same as the patch Felix volunteered, so anyone who has already incorporated that patch need not grab new sources. <http://www.opensource.apple.com/darwinsource/tarballs/apsl/xnu- 517.3.7.tar.gz> -- Kevin Van Vechten <kvv(a)apple.com> Open Source Software Engineer Apple Computer On Saturday, January 3, 2004, at 11:29 AM, Felix Kronlage wrote: On Fri, Jan 02, 2004 at 07:29:57AM -0800, Timothy Standing wrote: When I try and build xnu-517, I end up with the build ending with the following error: [...] /bin/sh: line 1: /usr/local/bin/kextsymboltool: No such file or directory you lack parts of the toolchain. Shantonu made a tarball with the binaries needed. You can grab all necessary parts here[1]. When I try and build xnu-517.3.7, I get the following errors [..] of course you still need [1], but also I suggest applying the diff found at [2], since otherwise the code won't compile either. -fkr [1] http://opendarwin.org/~ssen/xnu-tools-darwin-7.0.tar.gz [1] http://opendarwin.org/~fkr/xnu/make-xnu-517.3.7-compile.diff -- gpg-fingerprint: 076E 1E87 3E05 1C7F B1A0 8A48 0D31 9BD3 D9AC 74D0 Felix Kronlage | FKR-RIPE | http://www.hazardous.org/fkr/ fkr(a){grummel.net|opendarwin.org} | http://opendarwin.org/~fkr/ [demime 0.98b removed an attachment of type application/pgp-signature] _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: TAKE2: Problems testing with IOStorageFamily.kext
by Garth Cummings 19 Jan '04

19 Jan '04

Hi Chris, On Jan 18, 2004, at 4:58 PM, Chris Walters wrote: I am not sure why the debug driver does not load. You can verify it with the following command after installing it. Be certain that the source it built on the appropriate system version as well. kextload -nt /System/Library/Extensions/IOStorageFamily.kext I ran this and it said the kext appears to be valid. If I reboot with the partition as the default startup system it displays an apple with a black line above it (appearing like it is displaying text, maybe the reason why it fails, but the box isn't big enough). If I boot with the partition not as default then a "no entry" sign appears. This usually means that you've tried to run with an unstripped version of a kernel extension. All of the symbols in your KEXT have most likely chewed up all of the kernel's memory map. On Panther, the kernel's map was expanded quite a bit, so this limit is harder to reach. But you should always strip KEXTs you plan to load, especially shipping ones. Note that if you're doing two-machine debugging, the symbols go on the debugging system, not the target system. And if you've just embedded extra IOLogs or the like, stripping won't affect that functionality. HTH, --gc _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: scheduling question
by Jeff Nathan 19 Jan '04

19 Jan '04

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Jan 15, 2004, at 4:31 AM, Stefan Werner wrote: Hi, On Wednesday, Jan 14, 2004, at 22:05 Europe/Berlin, Jim Magee wrote: There are "helper" threads that process the data within the in-kernel network code somewhat isolating your thread from directly accessing the network hardware in that way. Your realtime thread may even interfere with the responsiveness of these helper threads in ways that might affect your own data. Even if you don't interfere with the processing of the in-kernel data, your response-time characteristics are going to be dominated by the scheduling behavior of these in-kernel threads. There are also situation where the network stack just borrows another user-level thread (that just happened to be in the network stack at the time) to drive completion of your packet handling (inbound and outbound). Obviously, you can't make these all realtime threads, so that too will affect your latency. Are there other tweaks to enhance the latency of networking? I don't mind a decrease in throughput, I'm in one of the rare cases where bandwidth is not the limit. There are these popular "bandwidth tuner" tweaks where people change the system's TCP buffer size with sysctls, would shrinking the buffers give me any latency improvements or will that result in dropped packets? If I understand the question, you're talking about receive buffer space. I'm not positive this answer is right, but I think a larger buffer can be used for a number of things. First, to negotiate the TCP window scaling option in which case the TCP window size can grow an additional 14 bits. Without SACK (selective acknowledgments) this is probably somewhat useless. If your stack can negotiate a large window size (meaning the other side of the TCP has to have a somewhat similar window scale.. I think the lowest is agreed upon), then selective acknowledgments are important to ensure either side of the connection has to re-send a large amount of data if some is dropped. The default receive window is 1/2 the size of a full TCP window... you could always try to grow it to full size. In reality, it's a combination of things that can improve performance. - -Jeff - -- http://cerberus.sourcefire.com/~jeff (gpg/pgp key id 6923D3FD) "I want to know God's thoughts... the rest are details." - Albert Einstein -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (Darwin) iD8DBQFADEW1Eqr8+Gkj0/0RAvC9AJ90DRHgx8PZ1pCIzHT8/W1vWwBHUwCdH5OM +wqjZ3eIlVQUfyygnQ97+i8= =LWtj -----END PGP SIGNATURE----- _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: sosend, soreceive, soconnect
by chuck remes 19 Jan '04

19 Jan '04

Search for "sosend howto" on darwin-kernel and look for the date 9/16/2003. There was a 3 day thread that covered a lot of this. Also, check out the darwin-development archives and search for similar strings. There are threads going back to 2001 covering many aspects of sosend() in the kernel. Matt was right... most of the time the correct answer is "do it in userland" but YMMV. cr On Jan 19, 2004, at 7:47 AM, Matt Jaffa wrote: I did search the archives, and there was no answer unless i missed the right one. I need to use sosend, and the others in the kernel so that i can communicate with my daemon in user space, so i am doing it in user space. Matt On Jan 19, 2004, at 1:03 AM, Michael Smith wrote: Matt. 1) Please search the archives before asking questions here. This has been answered at length, some might say ad nauseam. 2) Do it in user space. = Mike On Jan 18, 2004, at 11:45 PM, Matt Jaffa wrote: I am looking for a rundown version of the sosend, soconnect, etc. >From what I understand the regular socket, connect, accept, recv, send, don't work within the KERNEL so you have to use the Kernel specific ones, sosend, socreate, etc. I am looking for a good example of one of these that uses AF_UNIX, haven't been able to get the socreate, sobind, so etc. working. Thanks, Matt _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: sosend, soreceive, soconnect
by Matt Jaffa 19 Jan '04

19 Jan '04

I did search the archives, and there was no answer unless i missed the right one. I need to use sosend, and the others in the kernel so that i can communicate with my daemon in user space, so i am doing it in user space. Matt On Jan 19, 2004, at 1:03 AM, Michael Smith wrote: Matt. 1) Please search the archives before asking questions here. This has been answered at length, some might say ad nauseam. 2) Do it in user space. = Mike On Jan 18, 2004, at 11:45 PM, Matt Jaffa wrote: I am looking for a rundown version of the sosend, soconnect, etc. >From what I understand the regular socket, connect, accept, recv, send, don't work within the KERNEL so you have to use the Kernel specific ones, sosend, socreate, etc. I am looking for a good example of one of these that uses AF_UNIX, haven't been able to get the socreate, sobind, so etc. working. Thanks, Matt _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored. _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

Re: Processor Temperature
by William Kucharski 19 Jan '04

19 Jan '04

Robert J. Accettura said: >Well, I found some info here: >http://e-www.motorola.com/files/32bit/doc/app_note/AN1800.pdf > >But I'm missing something :-( I guess. Most likely licensing is in the >way. Perhaps someone knows better. That app note brings up several good points. 1) The TAU in the Motorola CPUs is only considered reasonably accurate in the MPC750 (Motorola G3.) 2) Each individual CPU MUST be calibrated for the numbers returned by the TAU to resemble reality at all; the calibration range provided for means the TAU can be off +/- 34 degrees C. 3) Chip errata for the MPC74XX (Motorola G4) family note that Motorola essentially dropped support for the TAU in the 74XX family of CPUs, as it may report erroneous values in the MPC7450 and MPC7455 CPUs and is not even implemented in the MPC7457. See Errata No. 3 in the MPC74XX family errata: http://e-www.motorola.com/files/32bit/doc/errata/MPC7457CE.pdf 4) IBM has STILL not published a CPU-specific manual for the PPC970, so we have no idea what may or may not be supported on the G5. In short, the TAU is useless on any CPU other than a G3, and even then all you can really tell is RELATIVE rather than ABSOLUTE temperature without first calibrating each individual CPU using the techniques given in the Motorola app note you referenced. William Kucharski kucharsk(a)mac.com _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0

sosend, soreceive, soconnect
by Matt Jaffa 19 Jan '04

19 Jan '04

Hi, I am looking for a rundown version of the sosend, soconnect, etc. >From what I understand the regular socket, connect, accept, recv, send, don't work within the KERNEL so you have to use the Kernel specific ones, sosend, socreate, etc. I am looking for a good example of one of these that uses AF_UNIX, haven't been able to get the socreate, sobind, so etc. working. Thanks, Matt _______________________________________________ darwin-kernel mailing list | darwin-kernel(a)lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.

1 0