August 2009 - Darwin-dev - lists.apple.com

Re: ucontext.h changes between 10.4 and 10.5?
by Terry Lambert 13 Aug '09

13 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com ... HAL_SavedRegisters regs; regs.edi = SIGCONTEXT_MEMBER( context, edi ); regs.esi = SIGCONTEXT_MEMBER( context, esi ); ...etc... } #include <stdio.h> #include <ucontext.h> /* A small program that proves we actually accessed them */ int main(int ac, char *av[]) { #if __LP64 unsigned long long save_rdi = fee->__ss.__rdi; #else unsigned long save_edi = fee->__ss.__edi; #endif printf("I compiled with the definition in scope\n"); 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… On Aug 12, 2009, at 3:25 PM, Rick Mann wrote: I'm guessing I'm going to need to change this code. It uses ucontext to get the sigcontext, and then uses that to save off some registers and optionally deal with some error state. Is there a way to do that without relying on a fragile API? Can you get chip vendors to stop making new chips with new features that might need information saved on a context switch? POSIX pretty much nails us to a non-opaque cross for all time for the ucontext_t and mcontext_t because it mandates that we expose the ucontext_t elements directly off a cast pointer, and it mandates that getcontext() take a ucontext_t that might have been statically declared, instead of treating it opaquely, and there's no corresponding destroycontext() (a pointer to a pointer would have been ideal for the ucontext_t type, with accessor and mutator functions which could be vendor defined, the way it's possible to implement posix_spawn()). I can (probably) post some of our code if it helps. But we basically do this: #if defined( HAVE_BSD_STYLE_SIGNALS ) #define SIGCONTEXT_MEMBER(x, y) ( static_cast< unsigned long >( x- >sc_##y ) ) #endif void Global_sigaction_handler( int signum, siginfo_t *sinfo, void *arg ) { struct ucontext * uc = reinterpret_cast<struct ucontext *>( arg ); struct sigcontext * context = (struct sigcontext *)&uc- >uc_mcontext; Is there an alternative way to implement this? Realize the code will not necessarily be portable between versions of the OS (POSIX is what added the __'s to endure against namespace pollution), and the binaries will not necessarily be portable between software updates, since we reserve the right to plop down pretty much any modified version of a context structure on the user stack we need to in order to support a hardware platform, and then pass a pointer to that as the new ucontext_t/mcontext_t type. Also realize that mcontext_t is implementation defined, and we've defined it as a pointer that points someplace on the stack after the ucontext_t, without definition anything about packing or anything else it'd be convenient to nail down, if you didn't want to have to rev your product on software updates, just like we potentially have to rev gdb, Java, and our other tools that know about the internals of these things on software updates. This cruft is NOT API. If you break on an SU because we change it, you fix it. -- Terry /////////////////////////////////// #define _XOPEN_SOURCE 600L /* This is pretty clearly documented in <ucontexh.h> */ /* * some fake data so I don't really have to write a signal handler and make it fire * to demonstrate successfully accessing structure members of the mcontext... */ ucontext_t foo; char *buf[1024]; mcontext_t fee = (void *)buf;; This email sent to site_archiver(a)lists.apple.com

1 0

Re: ucontext.h changes between 10.4 and 10.5?
by Rick Mann 13 Aug '09

13 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Aug 12, 2009, at 15:55:08, Terry Lambert wrote: Any ideas? Terry, thanks for the info. How can I get the registers? TIA, Rick _______________________________________________ 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… On Aug 12, 2009, at 11:16 AM, Rick Mann wrote: I have some code that was compiling when the project was set to 10.4 or "Use Base SDK" (I'm not really sure what that means. In my project, Base SDK is set to 10.5). But if I compile it with 10.5 (the Overview menu), it fails in a lot of ways, including issues around ucontext and sigcontext. I'm not able to significantly change this code. It's deep in our custom OS code. Is it just a matter of changing the name of the struct we're referencing or include a different header? There were two types of changes made to the ucontext/mcontext header file definitions: (1) Changes for UNIX conformance; typically these have to do with renaming variables that are not allowed to be defined in scope in a UNIX conformant compilation environment (which 10.5 / 64 bit is by default). (2) Changes to make certain definitions opaque to prevent their use by third parties, without enough difficulty that it should be clear to the third party that they are messing areound with SPI rather than API, and that therefore they should expect to potentially need to recompile their code on each minor release or software update, if they want it to continue working. You can always specif a deployment target of 10.4 to get the old (namespace polluting) header files, or you can update your code to only use the non-opaque fields, living with some of the renaming (usually, the addition of __ to things like register names), or you can include the header files from the relevent system frameworks and/ or published xnu source code, and take your chances on the next software update. Personally, I'd break the code in question out into a single compilation unit, and have everyone call functions from that compilation unit to mess with the contents of these structures. Worst case, you could target that one compilation unit to 10.4 to avoid changes now, and then cleanup breakage if it happens later. I think I'm going to go ahead and fix our code, but I need to check with someone else first. In the meantime, I can easily fix the ucontext name change, but I'm not sure how to get at the machine registers...is there a way? We log some information during the signal which includes $eip, and we save off registers in our HAL (I might be able to change that to use the get/setcontext() calls). This email sent to site_archiver(a)lists.apple.com

1 0

Re: ucontext.h changes between 10.4 and 10.5?
by Terry Lambert 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Any ideas? -- Terry _______________________________________________ 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… On Aug 12, 2009, at 11:16 AM, Rick Mann wrote: I have some code that was compiling when the project was set to 10.4 or "Use Base SDK" (I'm not really sure what that means. In my project, Base SDK is set to 10.5). But if I compile it with 10.5 (the Overview menu), it fails in a lot of ways, including issues around ucontext and sigcontext. I'm not able to significantly change this code. It's deep in our custom OS code. Is it just a matter of changing the name of the struct we're referencing or include a different header? There were two types of changes made to the ucontext/mcontext header file definitions: (1) Changes for UNIX conformance; typically these have to do with renaming variables that are not allowed to be defined in scope in a UNIX conformant compilation environment (which 10.5 / 64 bit is by default). (2) Changes to make certain definitions opaque to prevent their use by third parties, without enough difficulty that it should be clear to the third party that they are messing areound with SPI rather than API, and that therefore they should expect to potentially need to recompile their code on each minor release or software update, if they want it to continue working. You can always specif a deployment target of 10.4 to get the old (namespace polluting) header files, or you can update your code to only use the non-opaque fields, living with some of the renaming (usually, the addition of __ to things like register names), or you can include the header files from the relevent system frameworks and/ or published xnu source code, and take your chances on the next software update. Personally, I'd break the code in question out into a single compilation unit, and have everyone call functions from that compilation unit to mess with the contents of these structures. Worst case, you could target that one compilation unit to 10.4 to avoid changes now, and then cleanup breakage if it happens later. This email sent to site_archiver(a)lists.apple.com

1 0

Re: ucontext.h changes between 10.4 and 10.5?
by Rick Mann 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Is there a way to do that without relying on a fragile API? ... HAL_SavedRegisters regs; regs.edi = SIGCONTEXT_MEMBER( context, edi ); regs.esi = SIGCONTEXT_MEMBER( context, esi ); ...etc... } Is there an alternative way to implement this? On Aug 12, 2009, at 11:16:46, Rick Mann wrote: Any ideas? TIA, Rick _______________________________________________ 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/rmann%40latencyzero.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… I'm guessing I'm going to need to change this code. It uses ucontext to get the sigcontext, and then uses that to save off some registers and optionally deal with some error state. I can (probably) post some of our code if it helps. But we basically do this: #if defined( HAVE_BSD_STYLE_SIGNALS ) #define SIGCONTEXT_MEMBER(x, y) ( static_cast< unsigned long >( x- >sc_##y ) ) #endif void Global_sigaction_handler( int signum, siginfo_t *sinfo, void *arg ) { struct ucontext * uc = reinterpret_cast<struct ucontext *>( arg ); struct sigcontext * context = (struct sigcontext *)&uc- >uc_mcontext; I have some code that was compiling when the project was set to 10.4 or "Use Base SDK" (I'm not really sure what that means. In my project, Base SDK is set to 10.5). But if I compile it with 10.5 (the Overview menu), it fails in a lot of ways, including issues around ucontext and sigcontext. I'm not able to significantly change this code. It's deep in our custom OS code. Is it just a matter of changing the name of the struct we're referencing or include a different header? This email sent to rmann(a)latencyzero.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: Open Source Release for Mac OS X 10.5.8
by Stuart Crook 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 12 Aug 2009, at 22:30, Éric Lévénez wrote: - Stuart Crook _______________________________________________ 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… For me, Darwin OS (the standalone OS from Apple) is a little disappointment because of it's silence cancellation. There are some projects to continue the story, but that is not the same thing. Maybe PureDarwin will be THE Darwin distribution for the future. Maybe a new release next Xmas :-) I can't speak for the other couple of guys working on PureDarwin, but personally I'm waiting to make a fresh start with the Darwin 10 sources rather than put in any more work right now, so an Xmas release sounds about right. Maybe you'd like to lend a hand. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Open Source Release for Mac OS X 10.5.8
by Éric Lévénez 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Le 12 août 09 à 22:32, Dave Schroeder a écrit : On Aug 12, 2009, at 1:59 PM, Éric Lévénez wrote: Not at all. I like Darwin. Since my first NeXTstation. :-) Of course Mac OS X is alive, and well. Éric -- Éric Lévénez -- <http://www.levenez.com/> Unix is not only an OS, it's a way of life _______________________________________________ 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 Darwin as a bootable OS *from Apple* is dead, and has been for quite some time. OpenDarwin took up the torch for some time, but is also no more. Darwin as the UNIX core of Mac OS X is very much alive. Yes, sure :-) Do you disagree? The UNIX core of Mac OS X is called Darwin. If you dispute the statement that, "Darwin as the UNIX core of Mac OS X is very much alive," then that would imply you don't think Mac OS X as an OS is alive. ;-) Darwin as a collection of open source projects and software components included in Mac OS X is very much alive; some with considerable community participation and adoption; see e.g. http://www.macosforge.org Darwin as a bootable OS from the community is as alive as the community can make it; see e.g. http://www.puredarwin.org/ Darwin, OpenDarwin, PureDarwin, FreeDarwin, GNU-Darwin... some are more dead than other :-) Not really, actually. Darwin as a bootable OS from Apple and OpenDarwin are dead, PureDarwin is alive, and saying it's "dead" is like saying any Linux distro is "dead" because it doesn't have some subjective level of activity. In the bigger picture, yes, I'll concede it's probably fair to say that any "Darwin" OS (aside from Mac OS X) is dead. But Darwin is very much alive inside Mac OS X; the fact that it's not available from Apple as a standalone OS called Darwin was a decision that certainly impacted other Darwin-as-an-OS efforts, but I'm sure you'll agree Mac OS X is something of a common UNIX OS -- some might say by far the most common. ;-) For me, Darwin OS (the standalone OS from Apple) is a little disappointment because of it's silence cancellation. There are some projects to continue the story, but that is not the same thing. Maybe PureDarwin will be THE Darwin distribution for the future. Maybe a new release next Xmas :-) smime.p7s

1 0

Re: Open Source Release for Mac OS X 10.5.8
by Dave Schroeder 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Aug 12, 2009, at 1:59 PM, Éric Lévénez wrote: - Dave _______________________________________________ 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 Darwin as a bootable OS *from Apple* is dead, and has been for quite some time. OpenDarwin took up the torch for some time, but is also no more. Darwin as the UNIX core of Mac OS X is very much alive. Yes, sure :-) Do you disagree? The UNIX core of Mac OS X is called Darwin. If you dispute the statement that, "Darwin as the UNIX core of Mac OS X is very much alive," then that would imply you don't think Mac OS X as an OS is alive. ;-) Darwin as a collection of open source projects and software components included in Mac OS X is very much alive; some with considerable community participation and adoption; see e.g. http://www.macosforge.org Darwin as a bootable OS from the community is as alive as the community can make it; see e.g. http://www.puredarwin.org/ Darwin, OpenDarwin, PureDarwin, FreeDarwin, GNU-Darwin... some are more dead than other :-) Not really, actually. Darwin as a bootable OS from Apple and OpenDarwin are dead, PureDarwin is alive, and saying it's "dead" is like saying any Linux distro is "dead" because it doesn't have some subjective level of activity. In the bigger picture, yes, I'll concede it's probably fair to say that any "Darwin" OS (aside from Mac OS X) is dead. But Darwin is very much alive inside Mac OS X; the fact that it's not available from Apple as a standalone OS called Darwin was a decision that certainly impacted other Darwin-as-an-OS efforts, but I'm sure you'll agree Mac OS X is something of a common UNIX OS -- some might say by far the most common. ;-) smime.p7s

1 0

Re: Open Source Release for Mac OS X 10.5.8
by Éric Lévénez 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Le 12 août 09 à 20:41, William Siegrist a écrit : On Aug 12, 2009, at 10:47 AM, Éric Lévénez wrote: Is Darwin dead ? So "Darwin" will be back for the next announcement :-) Ok. Éric -- Éric Lévénez -- <http://www.levenez.com/> Unix is not only an OS, it's a way of life _______________________________________________ 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 At the beginning of Darwin project, the announcement was about Darwin OS, with Darwin release number. The last one was, I think, Darwin 9.7. It seems to me that the Darwin operating system is fading out. For Mac OS X 10.5.8, no word about Darwin 9.8. The lack of Darwin version number was just an oversight. Yes, this is the collection of projects for "Darwin 9.8". smime.p7s

1 0

Re: Open Source Release for Mac OS X 10.5.8
by Éric Lévénez 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Le 12 août 09 à 19:55, Dave Schroeder a écrit : Yes, sure :-) Ok, thank you. Éric -- Éric Lévénez -- <http://www.levenez.com/> Unix is not only an OS, it's a way of life _______________________________________________ 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 Darwin as a bootable OS *from Apple* is dead, and has been for quite some time. OpenDarwin took up the torch for some time, but is also no more. Darwin as the UNIX core of Mac OS X is very much alive. Darwin as a collection of open source projects and software components included in Mac OS X is very much alive; some with considerable community participation and adoption; see e.g. http://www.macosforge.org Darwin as a bootable OS from the community is as alive as the community can make it; see e.g. http://www.puredarwin.org/ Darwin, OpenDarwin, PureDarwin, FreeDarwin, GNU-Darwin... some are more dead than other :-) You won't get an Apple statement saying "Darwin is dead", because Darwin means so many things. You also (probably) won't get an Apple statement saying "Darwin OS is dead". And since "Darwin" is the core of Mac OS X, and Mac OS X (as of 10.5.x) is one of few OSes actually officially able to be labeled "UNIX", you may still want to reference Darwin with respect to your descriptions for Mac OS X. smime.p7s

1 0

Re: Open Source Release for Mac OS X 10.5.8
by William Siegrist 12 Aug '09

12 Aug '09

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Aug 12, 2009, at 10:47 AM, Éric Lévénez wrote: Is Darwin dead ? -Bill _______________________________________________ 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… At the beginning of Darwin project, the announcement was about Darwin OS, with Darwin release number. The last one was, I think, Darwin 9.7. It seems to me that the Darwin operating system is fading out. For Mac OS X 10.5.8, no word about Darwin 9.8. The lack of Darwin version number was just an oversight. Yes, this is the collection of projects for "Darwin 9.8". This email sent to site_archiver(a)lists.apple.com

1 0