May 2002 - Darwin-kernel - lists.apple.com

Darwin 5.4 xnu kernel CVS Release Tags and SHMMAX
by John Worsley 23 May '02

23 May '02

Greetings, I work for a company producing a commercial distribution of PostgreSQL, an open-source RDBMS. I've found that building it for OS X running the Darwin Kernel produces an error suggesting the Kernel be re-configured to have a higher value for SHMMAX when the requested concurrent connection value is too high. I've read up on the Darwin HOWTOs, and I've checked out xnu release Apple-201-19 from Apple's public CVS server, and I have two questions. 1. The Kernel Version appears to be 5.3, but the Kernel on my iMac running 10.1.4 of OS X reports Darwin Kernel version 5.4...just poking around I don't see any log or diff on the version.minor file in xnu/bsd/conf/ with a value of 4, and browsing the list of Release Tags didn't help much. So: is the version of the Darwin Kernel used with MacOS 10.1.4 (5.4) not available via Darwin CVS? The current snapshot reveals work on Kernel 6.0, which I'm pretty sure is inappropriate for my needs. ;) If it is available, what is the Release Tag? 2. I've found the SHMMAX parameter in xnu/bsd/conf/param.c, lines 98-101, and I found the following: #define SHMMAXPGS 1024 /* XXX until we have more kmap space */ #ifndef SHMMAX #define SHMMAX (SHMMAXPGS * 4096) Increasing the SHMMAXPGS value arbitrarily by a factor of 16 seemed to solve my problem, as far as the RDBMS software goes, but does anyone have an opinion on what effect, if any, this would have on the rest of the system? I'm a little leery of the "until we have more kmap space" warning since I have no idea to what it refers. ;) Any information on either of these topics would be duly appreciated. Best regards, Jw. -- John Worsley - Lead Programmer / Web Developer, Command Prompt, Inc. TheLinuxReview | Resources for Linux Users | http://www.thelinuxreview.com/ (503) 736-4609 | jlx(a)commandprompt.com | webmaster(a)thelinuxreview.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

Re: xnu kernel compile error (invalid parameter)
by Jim Magee 18 May '02

18 May '02

On Saturday, May 18, 2002, at 12:11 PM, hcso(a)csis.hku.hk wrote: The error is: cc1plus: Invalid option `-fapple-kext' Does that mean that I need to use a newer compiler? (the April beta developer tools?) If not, how to fix the problem? Or try an older kernel. I am using the December Developer tools to compile the kernel. The version of xnu is the latest version from CVS on 18-May (tagged as "Apple- 267). It needs the latest gcc3 compiler. But you're not going to have much luck running that kernel on a standard Darwin/ Mac OS X 10.1.x system image. It is too new for the older systems. You'll want to stick to the Apple-201-xxx branches of xnu that match the otherwise available Darwin/Mac OS X systems. --Jim _______________________________________________ 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

xnu kernel compile error (invalid parameter)
by hcso＠csis.hku.hk 18 May '02

18 May '02

Dear all, I tried to compile the xnu kernel according to steps in "Building the XNU kernel on Mac OS X 10.1.1" howto. I followed all the steps and when I typed "make" in xnu directory, error occured. It happened when /usr/bin/c++ tried to compile /xnu/iokit/Kernel/ IOLocks.cpp. The error is: cc1plus: Invalid option `-fapple-kext' Does that mean that I need to use a newer compiler? (the April beta developer tools?) If not, how to fix the problem? I am using the December Developer tools to compile the kernel. The version of xnu is the latest version from CVS on 18-May (tagged as "Apple- 267). thanks for your help. Bill _______________________________________________ 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: compiling Darwin in HFS+ partition
by Jim Magee 17 May '02

17 May '02

On Friday, May 17, 2002, at 02:32 PM, Brian Bergstrand wrote: At 2:05 AM +0800 5/18/02, hcso(a)csis.hku.hk wrote: It said that Darwin kernel should be built in a UFS partition due to case sensitity issue. I'd like to know if there's any problem building Darwin in HFS+ partition. I've built kernels up to xnu-222 on HFS+ partitions without problems. As Brian said, the Darwin kernel builds fine in HFS+. But parts of the rest of Darwin do not. If the HOWTO says the kernel needs UFS, it is wrong. If is simply said "Darwin," then it is correct. if I must build the kernel in UFS partition, is there any ways to convert my extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) Not without re-formatting. If you have Mac OS X, you could use the "Disk Copy" utility to create a UFS disk image as a single file within your HFS+ volume, and then mount that for doing your builds within. You could get more efficient with a reformat, but this is pretty easy. --Jim _______________________________________________ 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: compiling Darwin in HFS+ partition
by David A. Gatwood 17 May '02

17 May '02

On Fri, 17 May 2002, Brian Bergstrand wrote: > At 2:05 AM +0800 5/18/02, hcso(a)csis.hku.hk wrote: > >if I must build the kernel in UFS partition, is there any ways to convert my > >extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) > > Not without re-formatting. Or boot into single-user, use pdisk to change the partition type, then use newfs o create a new UFS filesystem in the partition. The contents of the partition will be lost when you do that. Be certain the partition is the right one; once you run newfs on a partition, there's no going back. Later, David --------------------------------------------------------------------- David A. Gatwood dgatwood(a)gatwood.net Developer Docs Writer dgatwood(a)apple.com Apple Computer dgatwood(a)mklinux.org Check out my weekly web comic: http://www.techmagazine.org _______________________________________________ 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: compiling Darwin in HFS+ partition
by Justin C. Walker 17 May '02

17 May '02

On Friday, May 17, 2002, at 11:05 AM, hcso(a)csis.hku.hk wrote: Dear all, I read the HOWTO document in darwinfo.org. It said that Darwin kernel should be built in a UFS partition due to case sensitity issue. I'd like to know if there's any problem building Darwin in HFS+ partition. That HowTo may be out of date, since the darwin (xnu) kernel has been buildable on HFS+ volumes since forever. It could be referring to problems with other projects that may be needed to build the kernel, but at least, when running Mac OS X, I've not seen any issues. if I must build the kernel in UFS partition, is there any ways to convert my extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) You don't need to worry about it for kernel builds. The only way to convert the partition is to wipe it out and recreate it as a UFS partion. On Mac OS X, use Disk Utility (select the Erase tab); I'm not sure of the tool to use in the case of a pure Darwin system. Regards, Justin -- /~\ The ASCII Justin C. Walker, Curmudgeon-at-Large \ / Ribbon Campaign X Help cure HTML Email / \ _______________________________________________ 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: compiling Darwin in HFS+ partition
by Brian Bergstrand 17 May '02

17 May '02

At 2:05 AM +0800 5/18/02, hcso(a)csis.hku.hk wrote: Dear all, I read the HOWTO document in darwinfo.org. It said that Darwin kernel should be built in a UFS partition due to case sensitity issue. I'd like to know if there's any problem building Darwin in HFS+ partition. I've built kernels up to xnu-222 on HFS+ partitions without problems. if I must build the kernel in UFS partition, is there any ways to convert my extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) Not without re-formatting. Thanks for your help. Bill University of Hong Kong, CE2 -- Brian Bergstrand <http://www.classicalguitar.net/brian/> PGP Key ID: 0xB6C7B6A2 There are no innocents in a world with the F.B.I. and Carnivore. - U.F. 000802 _______________________________________________ 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: compiling Darwin in HFS+ partition
by David Rehring 17 May '02

17 May '02

On 5/17/02 11:05 AM, hcso(a)csis.hku.hk at hcso(a)csis.hku.hk wrote: > Dear all, > > I read the HOWTO document in darwinfo.org. > > It said that Darwin kernel should be built in a UFS partition due to case > sensitity issue. > > I'd like to know if there's any problem building Darwin in HFS+ partition. > > if I must build the kernel in UFS partition, is there any ways to convert my > extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) > > Thanks for your help. > > Bill > University of Hong Kong, CE2 Bill, Some people have suggested that it is easier to create a large UFS disk image instead. You just mount the disk image and build the files on the disk image. Of course, you can't boot off of this image, but you can copy the built kernel to your bootable HFS+ volume. It will be slower than using a partition directly. I can't say how much slower as I haven't had the need to do this. Later, -- David Rehring Psychos do not explode when light hits Senior Software Engineer them, no matter how crazy they are... Atimi Software, Inc. www.atimi.com And totally insane guy! _______________________________________________ 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

compiling Darwin in HFS+ partition
by hcso＠csis.hku.hk 17 May '02

17 May '02

Dear all, I read the HOWTO document in darwinfo.org. It said that Darwin kernel should be built in a UFS partition due to case sensitity issue. I'd like to know if there's any problem building Darwin in HFS+ partition. if I must build the kernel in UFS partition, is there any ways to convert my extra HFS+ partition to UFS? (i've got an extra partition in HFS+ format) Thanks for your help. Bill University of Hong Kong, CE2 _______________________________________________ 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: Code loading and executing
by Adrian MARINESCU 15 May '02

15 May '02

Josh wrote: >Adrian, > You are almost certainly running into a problem with the > G3/G4's split L1 I and D caches. Basically what happens is that the > L1 I cache can contain stale bit after the D cache get loaded with > your code. After you copy the code into your mmap created buffer, > call msync on it with the MS_INVALIDATE option. This should cause > the kernel to flush the dcache for that region then invalidate the > icache for the region. After that, it should behave as expected. Thanks - that was the problem. However, I wrote a small asm routine to flush the caches - msync will not flush them - anyway, now it works just fine [btw - there should be a cacheflush() system call which in my Linux is documented in the man pages but not present in the libraries nor in the headers...] However, I just found another problem. I'm compiling a simple module on Linux/PPC and then I'm trying to load/link it under MacOS. It works just fine, except when I'm calling external functions - it seems to me that calling convension is not the same for Linux/PPC and MacOS. I'm a rookie in PPC assembler, but what I understood is that when entering into a proc the link register contains the return address. Under Linux/PPC, gcc generates code that will save the link register into r0 and then save it into a memory location relative to the r1(sp) register. However, when I'm calling the natively exported MacOS function, the code of the function will first save one of the registers at entry in the same memory location that the Link register was saved - so, when the function will try to return to the calee, the LR will be invalid. Any idea how/if the MacOSX's cc compiler can be forced to generate code for a function to help me here? bye, .mady Adrian MARINESCU, RAV Development Team - http://www.rav.ro PGP Key : http://homepages.gecad.ro/mady/mady.asc Key ID : 1024/1F365951 Adrian MARINESCU <mady(a)gecad.ro> Key fingerprint = 0397 839F F8A6 E720 A49E 7D07 469C BB5C _______________________________________________ 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