November 2005 - Darwin-dev - lists.apple.com

Re: 64-bit question
by Sam Hart 04 Nov '05

04 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 4 Nov 2005, at 14:07, Andrew Gallatin wrote: The problem seems more to be that the system doesn't gracefully handle running out of swap space with 64-bit jobs. My test machine crashed after allocating ~10GB, and I have ~10GB free on my rootfs. interesting This is what I was thinking S a m _______________________________________________ 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… So 64-bits seems to be a factor here. This email sent to site_archiver(a)lists.apple.com

1 0

Re: 64-bit question
by Andrew Gallatin 04 Nov '05

04 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com The problem seems more to be that the system doesn't gracefully handle running out of swap space with 64-bit jobs. My test machine crashed after allocating ~10GB, and I have ~10GB free on my rootfs. It is interesting that if I run the system out of memory by starting lots of 1GB 32-bit jobs, its much more graceful, and does not reboot. So 64-bits seems to be a factor here. Drew _______________________________________________ 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

task_t and dealloc?
by Stephane Sudre 04 Nov '05

04 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com When you do something like this: task_t task; tErrorCode=task_for_pid(mach_task_self(),inPID,&task); mach_port_deallocate(mach_task_self(),task); ? Because I'm seeing 1 warning during compilation when I use this. _______________________________________________ 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… Do you have to deallocate 'task' when you don't need it anymore? If yes, how do you do this? This email sent to site_archiver(a)lists.apple.com

1 0

Re: 64-bit question
by Sam Hart 04 Nov '05

04 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Bob wrote: I am an admin for a box that my department has users doing development work on. We have 8 gigs of ram in the machine and have been waiting to go to Tiger and 64-bit. I just upgraded the machine to tiger and my users are telling me the below conditions Can anyone confirm this or tell me what is going on? why does this crash a 10.4 machine but not Linux. ---------------- Observations from my users ----------------------- 1. The "top" command does not display the correct memory allocation status for a given process. To be precise, it is not able to show the values for memory greater than 4GB so it's a kind of hard to test high-mem applications. Looks like it's just the old 32bit "top", without 64bit capability. 2. We wrote and executed a custom C program to test the 64bit C compiler the system came with. It also tests 64bit environment and a system's general ability to handle high-mem apps. The application was able to allocate about 15GB on your test box and ... then the machine rebooted! Unacceptable! The program is listed belowl, to compile it just execute: gcc -arch ppc64 c-bigmem-test.c -o c-bigmem-test.exe The program is straightforward and it just tries to allocate more and more memory and it doesn't stop by itself (leave it or kill it!). On the screen you'll see the value of current memory taken. We tested it on the 64bit Linux /the memory status obtained from 64bit "top" matched the values printed out by our C program/ and when all system resources (physical memory + swap/hdd/ memory ) ran out, it was killed by the kernel which is correct. Just FYI, in case anyone wanted more data the following output is from a PowerBook with 1Gb wired % gcc -o test test.c % ./test <-- snipped --> MB allocated: 3700 test(399) malloc: *** vm_allocate(size=1003520) failed (error code=3) test(399) malloc: *** error: can't allocate region test(399) malloc: *** set a breakpoint in szone_error to debug Internal malloc failed! zsh: bus error ./test % S a m _______________________________________________ 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… % gcc -v Using built-in specs. Target: powerpc-apple-darwin8 Configured with: /private/var/tmp/gcc/gcc-5026.obj~19/src/configure -- disable-checking --prefix=/usr --mandir=/share/man --enable- languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^+.-]*$/ s/$/-4.0/ --with-gxx-include-dir=/include/gcc/darwin/4.0/c++ -- build=powerpc-apple-darwin8 --host=powerpc-apple-darwin8 -- target=powerpc-apple-darwin8 Thread model: posix gcc version 4.0.0 (Apple Computer, Inc. build 5026) obviously this is 32-bit, the interesting thing here is that the machine doesn't reboot This email sent to site_archiver(a)lists.apple.com

1 0

Re: 64-bit question
by Andrew Gallatin 03 Nov '05

03 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Bob Bader writes: > Can anyone confirm this or tell me what is going on? why does this > crash a 10.4 machine but not Linux. I do drivers, so I had a bit of a kernel crash debugging setup handy. The program crashed my G5 w/10.4.2 as well. The last messages to the console were: (default pager): [KERNEL]: no space in available paging segments (default pager): [KERNEL]: no space in available paging segments disk0s10: I/O error. disk0s10: I/O timeout. disk0s10: I/O timeout. Memory access exception(1,0,0) Entering system dump routine The crash has a stack like this: (gdb) bt #0 0x000acfa0 in memcpy_nop_if_32bit () #1 0x000ab598 in copyinmsg () #2 0x002741ac in sysctl_procargs () #3 0x002724d4 in kern_sysctl () #4 0x00271ac8 in __sysctl () #5 0x002a7a94 in unix_syscall () #6 0x000abcb0 in shandler () #7 0x41414141 in ?? () I don't have more info than this because I don't have a debug kernel handy for 10.4.2. You should file a bug so that Apple can fix it. It should be easy for them to reproduce. The issues with top have been pointed out before, but they have not been addressed as far as I know. Filing a bug against them might help to get them fixed. As to why it does not crash linux -- linux has been a 64-bit OS for nearly a decade, and they have the rough edges polished off. Tiger is only months old. If you need a ppc64 machine for running large memory jobs, you might be better off with linux. At least the IBM XL* compilers are available in 64-bit versions for linux. Good luck, Drew _______________________________________________ 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: Thread critical sections
by Nat! 03 Nov '05

03 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 30.10.2005 um 13:55 schrieb Nat!: Am 29.10.2005 um 02:20 schrieb Matt Watson: On Oct 28, 2005, at 5:15 PM, Steve Checkoway wrote: You really need all threads suspended except the current one? Assuming that I do, how would I go about doing that ? _______________________________________________ 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… Are you writing a garbage collector or something similar? No answer means: It isn't possible ? I tried once to thread_suspend all other threads excluding mine, but that didn't work out too well. Then I gave up. ---------------------------------------------- I'd like to fly But my wings have been so denied -- Cantrell This email sent to site_archiver(a)lists.apple.com

1 0

64-bit question
by Bob Bader 03 Nov '05

03 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thanks Bob The program is listed belowl, to compile it just execute: gcc -arch ppc64 c-bigmem-test.c -o c-bigmem-test.exe "c-bigmem-test.exe" is the output - an executable file you may run. Here is the code -------- #include <stdio.h> #include <string.h> #include <stdlib.h> #define ARRAY_SIZE 1000000 int main(){ char **bigbuf; char buf [ARRAY_SIZE]; long long int i=0; for(i=0;i<ARRAY_SIZE;i++){ buf[i]='A'; } bigbuf=malloc(ARRAY_SIZE*sizeof(char *)); if(bigbuf==NULL){ printf("Main malloc failed!\n"); } i=0; while(++i){ char * str_tmp=malloc(ARRAY_SIZE*sizeof(char)); if(str_tmp==NULL){ printf("Internal malloc failed!\n"); } bigbuf[i]=str_tmp; strcpy(bigbuf[i], buf); if(i%10==0) {printf("MB allocated: %lld\n",i);}; } return 0; } -------- _______________________________________________ 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 am an admin for a box that my department has users doing development work on. We have 8 gigs of ram in the machine and have been waiting to go to Tiger and 64-bit. I just upgraded the machine to tiger and my users are telling me the below conditions Can anyone confirm this or tell me what is going on? why does this crash a 10.4 machine but not Linux. I apologize if this is not the right forum for this question and would appreciate any ideas on where I can get my questions answered. ---------------- Observations from my users ----------------------- 1. The "top" command does not display the correct memory allocation status for a given process. To be precise, it is not able to show the values for memory greater than 4GB so it's a kind of hard to test high-mem applications. Looks like it's just the old 32bit "top", without 64bit capability. 2. We wrote and executed a custom C program to test the 64bit C compiler the system came with. It also tests 64bit environment and a system's general ability to handle high-mem apps. The application was able to allocate about 15GB on your test box and ... then the machine rebooted! Unacceptable! The program is straightforward and it just tries to allocate more and more memory and it doesn't stop by itself (leave it or kill it!). On the screen you'll see the value of current memory taken. We tested it on the 64bit Linux /the memory status obtained from 64bit "top" matched the values printed out by our C program/ and when all system resources (physical memory + swap/hdd/ memory ) ran out, it was killed by the kernel which is correct. This email sent to site_archiver(a)lists.apple.com

1 0

Re: File alteration monitoring
by MDonoughe 02 Nov '05

02 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Resent-date: Wed, 2 Nov 2005 04:44:00 +0000 Resent-from: MDonoughe <mdonoughe(a)bot24.hn.org> Resent-message-id: <48F46667-F88B-418D-A5F4-34BF25B21AEC(a)bot24.hn.org> Resent-to: darwin-dev(a)lists.apple.com On Nov 2, 2005, at 3:34, Craig Rodrigues wrote: On Wed, Nov 02, 2005 at 03:29:40AM +0000, MDonoughe wrote: is accessed, but it doesn't always give a full path. How do programs like Quicksilver and Spotlight get that instant notification? -- Craig Rodrigues rodrigc(a)crodrigues.org MDonoughe http://bot24.hn.org/ mdonoughe(a)bot24.hn.org My GPG key has changed. Verification is on my website. _______________________________________________ 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 I hadn't. That method will work perfectly. I need to add a listener for the directory, every sub-directory, and every file inside? Did you look at the kqueue() example here: http://developer.apple.com/samplecode/filesystem_examples/ filesystem_examples.html PGP.sig

1 0

Re: File alteration monitoring
by Craig Rodrigues 02 Nov '05

02 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Mutt/1.4.2.1i On Wed, Nov 02, 2005 at 03:29:40AM +0000, MDonoughe wrote: > is accessed, but it doesn't always give a full path. How do programs > like Quicksilver and Spotlight get that instant notification? Did you look at the kqueue() example here: http://developer.apple.com/samplecode/filesystem_examples/filesystem_exampl… -- Craig Rodrigues rodrigc(a)crodrigues.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… This email sent to site_archiver(a)lists.apple.com

1 0

File alteration monitoring
by MDonoughe 02 Nov '05

02 Nov '05

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 Hello. I want to synchronize the a folder on my iBook with a folder on my desktop using Unison. It would be much preferred if the synchronization took place automatically when a file changes, rather than when I run the command manually or at set intervals. I do not want to load down the system with polling possibly more than three thousand files. How can I have a program notified when a file inside a folder is created, modified, or deleted? I know fs_usage will show when a file is accessed, but it doesn't always give a full path. How do programs like Quicksilver and Spotlight get that instant notification? PGP.sig

1 0