October 2004 - Darwin-kernel

Re: CarbonPriv header file?
by Felix Kronlage 31 Oct '04

31 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Oct 30, 2004, at 10:38, Jim Witte wrote: - -fkr iD8DBQFBhLwiDTGb09msdNARAv/iAKCBBZcye51vX8PGd4D4EAK5JkpaaQCfelmP jkRSrSngwthBU6c0E78et4E= =dC/7 -----END PGP SIGNATURE----- _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… I want to compile the ppp package, the sample program for which (MiniTerm) requires CarbonPriv.h (I assume MiniTerm is the example program, and Google turns this as the ONLY thing in Darwin that requires this header..) Where can I get it? this header file is not publically available. This is btw not the only pitfall in building ppp. I advise to grab the patches found here[1]. [1] <url: http://cvs.opendarwin.org/index.cgi/src/Ports/ppp/files/apple/> - -- 076E 1E87 3E05 1C7F B1A0 8A48 0D31 9BD3 D9AC 74D0 http://fkr.hazardous.org/ | http://opendarwin.org/~fkr/ | FKR-RIPE -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (Darwin) This email sent to site_archiver(a)lists.apple.com

1 0

Re: truncate()
by Steve Dekorte 30 Oct '04

30 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Please ignore that question. I found the problem and it was my fault. -- Steve _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Oct 30, 2004, at 1:04 AM, Steve Dekorte wrote: Has anyone else had problems with truncate() locking up OSX 10.3 filesystem? If so, is there a solution / workaround? This email sent to site_archiver(a)lists.apple.com

1 0

CarbonPriv header file?
by Jim Witte 30 Oct '04

30 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com smKeyEnableRomanOnly // enable only roman keyboard KeyScript(smKeyEnableRomanOnly); // contact pppd to get serial descriptor and exit code communication channel sockfd = socket(AF_LOCAL, SOCK_STREAM, 0);

1 0

truncate()
by Steve Dekorte 30 Oct '04

30 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com -- Steve _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Has anyone else had problems with truncate() locking up OSX 10.3 filesystem? If so, is there a solution / workaround? This email sent to site_archiver(a)lists.apple.com

1 0

Re: Problem with Kernel thread and soreceive
by Matt Jaffa 30 Oct '04

30 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, nvm I figured it out, the: aiov2[0].iov_base = test2; aiov2[0].iov_len = 1500; auio2.uio_iov = aiov2; auio2.uio_iovcnt = 1; auio2.uio_rw = UIO_WRITE | UIO_READ; auio2.uio_segflg = UIO_SYSSPACE; auio2.uio_offset = 0; auio2.uio_resid = 1500; Matt _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… needed to be inside the while loop, with printf statements I was realizing that it was only allowing my while loop to receive at max 1500 chars, and when i saw the above code not in the for loop I realized that the 1500 was being decremented by the size returned from soreceive so putting the above code in the while loop always initializes it to 1500 each time. Shoulda realized this sooner. This email sent to site_archiver(a)lists.apple.com

1 0

Problem with Kernel thread and soreceive
by Matt Jaffa 30 Oct '04

30 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi, Thanks, Matt _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… I have a thread that I create in my nke_start function called when my nke is loaded. And as you can see in my thread below with the code I include I create a socket(created in the start function of the nke, this is a global struct socket). and then bind it to port 9856 and this is a UDP socket, the soreceive is in a while always receiving data. This is working fine but after awhile and when the data is somewhat big, it causes the soreceive to do nothing but return so it doesn't want on it anymore and causes my NKE to eat up the cpu like crazy and stop receiving data. Notice anything bad in the code below? void mythread(void * one) { int funnel_state; funnel_state = thread_funnel_set(network_flock, TRUE); struct ifnet *ifp; struct ifnet *sendg = NULL; TAILQ_FOREACH(ifp, &ifnet, if_link) { printf("This is the name: %s, This is the number: %d \n",ifp->if_name, ifp->if_unit); if(( bcmp(ifp->if_name,"en",strlen(ifp->if_name))== 0) && (ifp->if_unit == 0) ) { sendg = ifp; printf("IS SENDG NULL?: %d\n", sendg == NULL); //printf("Install filter on bsdname = %s\n", bsdName); //here is where we } } theconnect->so_snd.sb_timeo = 15; theconnect->so_rcv.sb_flags |= SB_NOINTR; theconnect->so_snd.sb_flags |= SB_NOINTR; //int test = 0; printf("The thread was called without crashing\n"); struct mbuf *cm2 = 0; struct iovec aiov2[1]; struct uio auio2 = { 0 }; char test2[1600]; aiov2[0].iov_base = test2; aiov2[0].iov_len = 1500; auio2.uio_iov = aiov2; auio2.uio_iovcnt = 1; auio2.uio_rw = UIO_WRITE | UIO_READ; auio2.uio_segflg = UIO_SYSSPACE; auio2.uio_offset = 0; auio2.uio_resid = 1500; printf("Check\n"); struct sockaddr_in name; name.sin_addr.s_addr = htonl(INADDR_ANY); name.sin_len = sizeof(struct sockaddr_in); name.sin_family = AF_INET; // Set the server's family name.sin_port = htons((unsigned short) 9856); // Set the port to connect to... int tes = sobind(theconnect,(struct sockaddr*)&name); if(tes == 0) { while(1) { //printf("rerror: %d\n", 33); /*test++; if(test >= 9) return;*/ int flags = MSG_DONTWAIT; printf("Socket state: %x\n", theconnect->so_state); int rerror = soreceive(theconnect, (struct sockaddr **)0, &auio2, &cm2, 0, 0); printf("rerror: %d\n", rerror); if (rerror) { } else { if(cm2 != NULL) { printf("cm2 is not null\n"); //char * final = mtod(cm2, char*); printf("sendg == NULL2nd: %d\n", sendg == NULL); if(sendg != NULL) { int retval = (*sendg->if_output)(sendg, cm2); printf("What is retval: %d\n", retval); } } } } } thread_funnel_set(network_flock, funnel_state); } This email sent to site_archiver(a)lists.apple.com

1 0

Re: [APPL:DarwinKernel] Re: type sizes
by Quinn 28 Oct '04

28 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com * has a 32-bit kernel address space regardless of host CPU; * has some system libraries optimized for the host CPU, which includes taking advantage of 64-bit operations when on G5; * supports 32-bit user address spaces on any CPU, and additionally 64-bit user address spaces (depending on whether the binary is 32/64-bit) on G5. Accurate, but not complete. The item you're missing is: * only a small subset of system libraries are available to 64-bit programs (basically you get System framework and anything that it depends on) ps This point is actually described quite clearly on our public web site. The enhanced kernel, plus a 64-bit version of libSystem, let command-line programs, background daemons and network services directly manipulate up to 16 exabytes of virtual memory, more than four billion times the memory addressable by today's 32-bit applications. -- <http://www.apple.com/macosx/tiger/64bit.html> S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… At 12:34 +1300 28/10/04, Koryn Grant wrote: Is this an accurate summary? This is enough to get the majority of traditional UNIX programs running 64-bit, but not for 64-bit GUI programs. This email sent to site_archiver(a)lists.apple.com

1 0

Re: [APPL:DarwinKernel] Re: type sizes
by Koryn Grant 27 Oct '04

27 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Thu, 28 Oct 2004, Quinn wrote: > At 10:10 -0500 26/10/04, Jay A. Kreibich wrote: > >The big question is how the kernel is going to run. > > The Tiger kernel will run 32 bit, even on G5. That is, sizeof(void > *) == 4 for kernel code. > > Portions of the kernel have been modified to support 64 bit user > address spaces but those portions represent user addresses using a 64 > bit integer type (that is, they don't use a C pointer). So the Tiger release: * has a 32-bit kernel address space regardless of host CPU; * has some system libraries optimized for the host CPU, which includes taking advantage of 64-bit operations when on G5; * supports 32-bit user address spaces on any CPU, and additionally 64-bit user address spaces (depending on whether the binary is 32/64-bit) on G5. Is this an accurate summary? Cheers, Koryn -- Dr Koryn Grant phone +64 7 8346729 Software Engineering Manager mobile +64 21 488386 Endace Technology Ltd email koryn(a)endace.com Hamilton, New Zealand web http://www.endace.com/ _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: [APPL:DarwinKernel] Re: type sizes
by Quinn 27 Oct '04

27 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… At 10:10 -0500 26/10/04, Jay A. Kreibich wrote: The big question is how the kernel is going to run. The Tiger kernel will run 32 bit, even on G5. That is, sizeof(void *) == 4 for kernel code. Portions of the kernel have been modified to support 64 bit user address spaces but those portions represent user addresses using a 64 bit integer type (that is, they don't use a C pointer). This email sent to site_archiver(a)lists.apple.com

1 0

Re: mach_msg and mach_msg_type_t missing
by Quinn 27 Oct '04

27 Oct '04

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com This sounds like a real bad idea. S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Technical Support * Networking, Communications, Hardware _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… At 23:45 +1000 26/10/04, Matt Connolly wrote: Any help, or pointers to some more detailed documentation for using mach_msg to communicate from the kernel to a user client would be much appreciated. 1. In general, trying to 'call' from kernel space to user space is a bad idea. You have to be really careful that you don't run the kernel out of resources, or put the kernel in a position where it's waiting, potentially forever, for user space. A much better design is to have your user space agent call into the kernel and ask it "Is there anything to do?" If there's nothing to do, it blocks. You can implement this model via a UNIX domain socket or with an I/O Kit user client, depending on what part of the kernel you're running in. Of course, you still have to design the kernel code so that it can't consume unbounded resources. 2. Sending Mach messages from kernel space to user space is particularly tricky. Jim Magee has talked about this on the list in the past. If you Google for "site:lists.apple.com Jim Magee naked" (*snicker*), you'll find his posts. Can you give me a high-level description of what you're trying to do? Perhaps I can suggest the right direction to proceed. This email sent to site_archiver(a)lists.apple.com

1 0