March 2003 - Darwin-kernel - lists.apple.com

Re: realloc
by Jonas Maebe 17 Mar '03

17 Mar '03

On maandag, maa 17, 2003, at 04:05 Europe/Brussels, Eric Grant wrote: Based on my experience with some simple test programs (under OS X 10.2.3), and confirmed by my skimming of the sources (Revision 1.1.1.4 of malloc.c), it appears that realloc does not "free" memory (in any meaningful sense) realloc is a libc function, not a kernel function. As such, when you free memory using realloc, this memory is put back in libc's memory allocator pool for reuse for other malloc/realloc/calloc's. Jonas _______________________________________________ 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: realloc
by Justin Walker 17 Mar '03

17 Mar '03

On Sunday, Mar 16, 2003, at 19:05 US/Pacific, Eric Grant wrote: Based on my experience with some simple test programs (under OS X 10.2.3), and confirmed by my skimming of the sources (Revision 1.1.1.4 of malloc.c), it appears that realloc does not "free" memory (in any meaningful sense) when passed a size smaller than the size originally malloc'd. That is, despite the supposed downsizing of the object, none of the logical address space encompassed by the original object is available to satisfy subsequent allocation requests. For example: p1 = malloc (2 * 1024 * 1024 * 1024); // 2GB p2 = realloc (p1, 1); The result is that half the logical address space has basically disappeared (though it appears to be available to satisfy a request to realloc p2 to a size up to 2GB). To my mind, this result violates the standard (from man realloc): The realloc() function changes the size of the previously allocated mem- ory referenced by ptr to size bytes. Moreover, it is difficult to see how this result is otherwise a "feature." Good? Bad? Ugly? None of the above? I'd say it was normal from my experience. Nothing violates the description you quote above, at any rate. AFAIK, this is the way most unices act for realloc, and in addition, address space acquired in the process of malloc()ing space is never returned. This is really a topic for darwin-developers, not darwin-kernel. The kernel just provides resources for user-mode operation, and it's user-mode that dictates policy (at least for this issue). Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | Some people have a got a mental | horizon of radius zero, and call | it their point of view | -- David Hilbert *--------------------------------------*-------------------------------* _______________________________________________ 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

realloc
by Eric Grant 17 Mar '03

17 Mar '03

Based on my experience with some simple test programs (under OS X 10.2.3), and confirmed by my skimming of the sources (Revision 1.1.1.4 of malloc.c), it appears that realloc does not "free" memory (in any meaningful sense) when passed a size smaller than the size originally malloc'd. That is, despite the supposed downsizing of the object, none of the logical address space encompassed by the original object is available to satisfy subsequent allocation requests. For example: p1 = malloc (2 * 1024 * 1024 * 1024); // 2GB p2 = realloc (p1, 1); The result is that half the logical address space has basically disappeared (though it appears to be available to satisfy a request to realloc p2 to a size up to 2GB). To my mind, this result violates the standard (from man realloc): > The realloc() function changes the size of the previously allocated mem- > ory referenced by ptr to size bytes. Moreover, it is difficult to see how this result is otherwise a "feature." Good? Bad? Ugly? Cheers, Eric Grant _______________________________________________ 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

help? trouble setting Shared Memory parameters in OSX kernel
by R Blake 16 Mar '03

16 Mar '03

hi all, i've successfully built postgreSQL 7.3.2-STABLE on Mac OSX 10.2.4. much trouble launching it led me to the discussions (<http://www.postgresql.org/docs/view.php?version=7.3&idoc=1&file=kernel-re sources.html>) on setting shared memory parameters -- especially SHMMAX -- in OSX's "/System/Library/StartupItems/SystemTuning/SystemTuning" so, as instructed, i changed the defaults (near the bottom of SystemTuning): sysctl -w kern.sysv.shmmax=4194304 sysctl -w kern.sysv.shmmin=1 sysctl -w kern.sysv.shmmni=32 sysctl -w kern.sysv.shmseg=8 sysctl -w kern.sysv.shmall=1024 to: sysctl -w kern.sysv.shmmax=33554432 # bytes sysctl -w kern.sysv.shmmin=1 sysctl -w kern.sysv.shmmni=32 sysctl -w kern.sysv.shmseg=8 sysctl -w kern.sysv.shmall=8192 # 4K pages and defined (in postgresql.conf): shared_buffers = 2048 max_coinnection = 1024 and rebooted. unfortunately, immediately after system launch, i still couldn't launch postgres eventually, a random check with "sysctl -a" gave: kern.sysv.shmmax: -1 kern.sysv.shmmin: -1 kern.sysv.shmmni: -1 kern.sysv.shmseg: -1 kern.sysv.shmall: -1 OBVIOUSLY, a problem! i CAN enter the sysctl settings at the command line, verify success with 'sysctl -a', and then postgres launches/operates just fine ..... QUESTIONS: any idea why these setting are being written at boot to values of "-1"? how do i get these setting to 'stick'? thanks, richard _______________________________________________ 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

Max number partitions per disk??
by Timothy Standing 15 Mar '03

15 Mar '03

I remember reading on this list a year ago (or it could have been two) that there was a limit on the number of partitions the kernel supported on each disk. The post said that the limit was 16 partitions per disk. (I am assuming that the limit still exists.) I would prefer to keep users from running into this limit so I have a few questions about it: - Does anyone know whether this limit includes the partition which contains the partition map itself? - Does it include Apple_Free partitions? - Where is the limit implemented? Thanks in advance, Tim Standing SoftRAID, LLC _______________________________________________ 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: Different headers definitions between Kernel.framework and /usr/include/sys/socketvar.h
by Justin Walker 14 Mar '03

14 Mar '03

On Friday, Mar 14, 2003, at 07:53 US/Pacific, Stiphane Sudre wrote: SS_ISDISCONNECTED is not declared in Kernel.framework. Is it correct? Incorrect. It is defined in <sys/socketvar.h>, which, on my system, is in both the /usr/include and the Kernel.framework trees. Side question: When you're first calling sendto/sendmsg, am I right to believe the socket state is at more SS_PRIV? I don't understand your question. The SS_PRIV bit is set at the time the socket is created, and at no other time. It is *only* set if the creating process is owned by root. Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | Men are from Earth. | Women are from Earth. | Deal with it. *--------------------------------------*-------------------------------* _______________________________________________ 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

Different headers definitions between Kernel.framework and /usr/include/sys/socketvar.h
by Stéphane Sudre 14 Mar '03

14 Mar '03

SS_ISDISCONNECTED is not declared in Kernel.framework. Is it correct? Side question: When you're first calling sendto/sendmsg, am I right to believe the socket state is at more SS_PRIV? _______________________________________________ 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: deciphering powerpc registers
by Garth Cummings 13 Mar '03

13 Mar '03

Hi Brian, On Thursday, March 13, 2003, at 10:03 AM, Brian Daugherty wrote: hi, I'm trying to debug a kernel panic. You're in luck! Take a look at Technical Note 2063: "Understanding and Debugging Kernel Panics" <http://developer.apple.com/technotes/tn2002/tn2063.html> The TN also contains references to IBM and Motorola PowerPC docs. --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: deciphering powerpc registers
by Justin Walker 13 Mar '03

13 Mar '03

On Thursday, Mar 13, 2003, at 10:42 US/Pacific, Brian Bergstrand wrote: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday, March 13, 2003, at 12:03 PM, Brian Daugherty wrote: PC = Program Counter. This is actually r2 (I think) since the PPC does not actually have a PC register. Whereever the PC may be, it's not r2, which is a general-purpose register, available for any use (ignoring the ABI, of course). The PC is not accessible, but it's there in some form, internally. DAR = Data register, this is usually r3, or the first arg to a function call. No, again, r3 is a general-purpose register. This one (DAR) is the Data Address Register (kernel-accessible) that contains the address that caused a Data Access fault. LR = Link Register which should be the previous function in the call stack. It contains the address to return to after a procedure call (bl, == 'branch and link'); or contains the target of one of the "branch-to-link-register" instructions. I forget what MSR and DSISR are. Machine State Register and Data Storage/Instruction Storage Register. Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large * Institute for General Semantics | If you're not confused, | You're not paying attention *--------------------------------------*-------------------------------* _______________________________________________ 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: deciphering powerpc registers
by Justin Walker 13 Mar '03

13 Mar '03

Look at the 'mot.com' website. They have PowerPC doc on-line. You can also look in the kernel (/usr/include/...) header files for some of this, in a more cryptic form. Regards, Justin On Thursday, Mar 13, 2003, at 10:03 US/Pacific, Brian Daugherty wrote: hi, I'm trying to debug a kernel panic. I'd like to lookup the meaning of the codes contained in the DSISR, MSR and DAR. Exception state (sv=0x0DC62C80) PC=0x90000E2C; MSR=0x0000D030; DAR=0x0DCDB000; DSISR=0x42000000; LR=0x0001 -- 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