September 2006 - Darwin-dev

Re: Passing too many parameters to a function?
by Mailing list subscriptions 18 Sep '06

18 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com El 18/09/2006, a las 22:06, Camillo Lugaresi escribió: Which may not be all that easy to handle... _______________________________________________ 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… Of course, the later versions of the callee must be prepared to receive garbage in the extra arguments from older callers. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Passing too many parameters to a function?
by Camillo Lugaresi 18 Sep '06

18 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:in-reply-to:references:mime-version:content-type:message-id:cc:content-transfer-encoding:from:subject:date:to:x-mailer; b=aL83bo/75k4oIqBTNtom5rAm0Edb0ZrAq9Sqizd1GCPtH0x82PRh3fcQNpexTN+0N3pC4DEO/YIkjfCb48nWU1VU4u7L+jFjI8Fhh5h/hEu6A+891jnFq1uaxecWkDJ5pq8b+TqZESHub4qxVy/xyJHQzsJWfwkYO1VrJzQhDMU= On 20/ago/06, at 12:17, Uli Kusterer wrote: Hi, int main() or int main( int argc, const char* argv* ) Camillo _______________________________________________ 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… you know how main() can be declared with varying parameter counts, like and even more. Does this work for any function? If I pass a function more parameters than it actually uses, is that safe with OS X's current Mach-O ABIs? I would like to have plugins in my app be prepared for a future expansion of the plugin protocol, which would require them to pass an additional parameter (that my app isn't using right now) to the callbacks they send. Since I'm gonna be releasing a new plugin SDK with the first universal version, I thought this would be a good opportunity to add this change, so that all recompiled Xes would run unchanged on the future new system and I can drop the old code that doesn't have this parameter. I trid googling for "Mach O ABI too many parameters" and stuff like that, but I just couldn't find any clues there, so if anyone knows where I can find this documented or even has an answer, that would be cool! I just read this thread, and although it's old, it seems that it did not reach a conclusion, so forgive me for resurrecting it. :) If you look at the PPC and IA32 ABIs, it seems that what you want to do is safe. Here is the PPC32 ABI: <http://developer.apple.com/documentation/DeveloperTools/Conceptual/ LowLevelABI/Articles/32bitPowerPC.html> Note that space for the parameters is allocated in the caller's stack frame (most parameters are passed in registers, but space is still allocated for them on the stack). On entry, the called function sees a 24-byte linkage area at SP, and the parameter area at SP+24. The parameter area is large enough to hold parameters for all functions the caller wants to call; the callee does not need to know its exact size, since arguments are allocated in left-to-right order at increasing addresses starting at SP+24: the callee just looks at the lowest chunk of the parameter area, corresponding to the leftmost arguments (however many it expects), and ignores the rest. The same goes for the arguments passed in registers: the callee will use the ones it needs, and ignore/clobber the rest. Here is the IA32 ABI: <http://developer.apple.com/documentation/ DeveloperTools/Conceptual/LowLevelABI/Articles/IA32.html> The parameter area is allocated and deallocated by the caller, and arguments are allocated in order starting at SP and growing upwards. The situation is the same as on PPC. Therefore, I'd say you're safe to add parameters as you wish. Of course, the later versions of the callee must be prepared to receive garbage in the extra arguments from older callers. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Is /bin/pax buggy in Mac OS X 10.4.7?
by Stéphane Sudre 18 Sep '06

18 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On lundi, septembre 18, 2006, at 10:18 AM, Markus Hitter wrote: Am 17.09.2006 um 16:17 schrieb Stéphane Sudre: _______________________________________________ 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… As a matter of fact, file "./IcebergAlpha/Library/Application Support/Iceberg/Projects Templates/Plug-ins/Spotlight Importer/._template.packproj" really does _NOT_ exist so there does not seem to be any valuable reason why pax would need to read it. Three _is_ a reason for ._template.packproj: it's the Finder Info and resource fork stuff. Every file on a HFSx partition gets split on the fly when copied to a non-resource fork aware file system or archive. Same for cp, tar, rsync, ... Hmm, probably not all the files that should be split are. Based on the output of the /Developers/Tools/SplitFork tool, some files that should are not. So if the code is shared with the OS... In the case I described, I was on a HFS+ volume for both source and destination. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Is /bin/pax buggy in Mac OS X 10.4.7?
by Markus Hitter 18 Sep '06

18 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Am 17.09.2006 um 16:17 schrieb Stéphane Sudre: Markus - - - - - - - - - - - - - - - - - - - Dipl. Ing. Markus Hitter http://www.jump-ing.de/ _______________________________________________ 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… As a matter of fact, file "./IcebergAlpha/Library/Application Support/Iceberg/Projects Templates/Plug-ins/Spotlight Importer/._template.packproj" really does _NOT_ exist so there does not seem to be any valuable reason why pax would need to read it. Three _is_ a reason for ._template.packproj: it's the Finder Info and resource fork stuff. Every file on a HFSx partition gets split on the fly when copied to a non-resource fork aware file system or archive. Same for cp, tar, rsync, ... Obviously, the implementation of this on-the-fly split isn't savy against the assumptions made when pax was coded. This email sent to site_archiver(a)lists.apple.com

1 0

[SOLVED] Re: Is /bin/pax buggy in Mac OS X 10.4.7?
by Stéphane Sudre 18 Sep '06

18 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On dimanche, septembre 17, 2006, at 04:17 PM, Stéphane Sudre wrote: I'm investigating an issue related to the pax tool (/bin/pax). _______________________________________________ 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… Short answer: Yes, pax is buggy on Mac OS X 10.4.7. I will file a bug report. Long answer: The problem is related to ACL being enabled on /. Thanks to Matt for the tip. [...] This email sent to site_archiver(a)lists.apple.com

1 0

Is /bin/pax buggy in Mac OS X 10.4.7?
by Stéphane Sudre 17 Sep '06

17 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com I'm investigating an issue related to the pax tool (/bin/pax). I'm seeing really strange errors: When I run the following command: % sudo /bin/pax -v -w -z -x cpio -f /tmp/toto.pax.gz . on this hierarchy, I get this in system.log: and this in the terminal output: This problem does not happen on Mac OS X 10.2.8. _______________________________________________ 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… Let's say I have a file hierarchy for a package (an Installer.app package): Sep 17 16:02:35 Selene /bin/pax: open on /tmp/pax.md.625l: Cannot allocate memory Sep 17 16:02:35 Selene /bin/pax: close on : Bad file descriptor Sep 17 16:02:35 Selene /bin/pax: error closing files: Bad file descriptor Sep 17 16:02:35 Selene /bin/pax: open on /tmp/pax.md.aCjr: Cannot allocate memory pax: Unable to open ./IcebergAlpha/Library/Application Support/Iceberg/Projects Templates/Plug-ins/Spotlight Importer/._template.packproj to read <No such file or directory> ./IcebergAlpha/Library/Application Support/Iceberg/Projects Templates/Plug-ins/Spotlight Importer/template.packproj pax: Read fault on ./IcebergAlpha/Library/Frameworks <Is a directory> As a matter of fact, file "./IcebergAlpha/Library/Application Support/Iceberg/Projects Templates/Plug-ins/Spotlight Importer/._template.packproj" really does _NOT_ exist so there does not seem to be any valuable reason why pax would need to read it. It can not be a memory issue as this computer has 5 GB of RAM and 12 GB of disk space available for the system partition. Is there any known issues with pax? Has pax recently developed an allergy to AppleDouble format files? This email sent to site_archiver(a)lists.apple.com

1 0

CFNetwork compiled for Win32
by Steve Graesser 15 Sep '06

15 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: White Lab Consulting Hi All, I am trying to port my apps networking code to run on Windows. The networking code is written using the CFHTTPMessage and CFReadStream functions. I was hoping to create a Win32 version of the CFNetwork library. I started with the CoreFoundation on Win32 projects provided by Jose Cruanyes and posed to this list. See <http://lists.apple.com/archives/ Darwin-development/2004/Jan/msg00072.html>. I was able to successfully compile the CF-368.27 code and run some sample apps using CW Pro 9. I then tried to compile the CFNetwork-129.16 source code. I noticed in the CFNetwork Makefile has a set of compile options for the Cygwin platform. However, the CFNETWORK_BUILDING_DLL is defined but never referenced in the source code. I also found the CFNetwork_EXPORT define in the CFNetworkDefs.h header file. But, the CFNetwork_EXPORT define is only used twice in the _CFURLAccess.c source file and the CFNetwork_EXPORT is not set correctly for Win32 builds. Even after modifying the source to use a corrrectly define CFNetwork_EXPORT and ignoring calls to the SSL framework, I can't compile and link the CFNetwork library due to dependency on the Security and SystemConfiguration frameworks. Has anyone been able to successfully compile and link the CFNetwork library for Win32? Steve Graesser White Lab Consulting LLC PGP Fingerprint: 8AA4 6514 A87C FD26 A135 DBA0 F8AF 3E1B 41C5 4B5C _______________________________________________ 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: Stripping IOKit kernel extensions
by Geoffrey Schmit 13 Sep '06

13 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:date:from:subject:to:x-priority:message-id:mime-version:content-type:content-transfer-encoding:x-mailer; b=POCODZx5p42pmCfNaeMCpAcJXJej1jv7y20z25Dybr1b5eIkEHCJ+rAyXKr6l9exYG2dF5DbAUmKMDtQn3xrgX6yZ38jMFviaLqZthFrAg4B7EdSZEyjOBm/FiED78KCgr4a0mt7eW5c5LiZnltimT4OmZjPkFLM+VQ34yatYpw= On 12/09/2006 at 2:28 PM, Ryan McGann <rmcgann(a)mac.com> wrote: > I am trying to strip my IOKit driver and am running into problems. > Spefically, kextload crashes (bus error) when it attempts to load the > kext when I strip it. It loads fine without stripping. > > I found a discussion about this very issue in the archives, but I > didn't see a resolution for the OP. > > I have a symbols file which contains _kmod_info, and all symbols for > my driver (both static data and methods). When I strip the executable, > I use strip -u -s <symbols.txt> <path to kext binary>. I get a nice > trim binary that contains only _kmod_info and the following symbols > (I've included their demanged form, but the symbols file contains the > mangled names of course): > > [...] > > I'm not sure what other symbols I need to include in the binary to prevent kextload from > crashing. Is there a way to do this from Xcode? If not, what's the proper strip/nmedit > command needed to do what is necessary? This is all on Tiger with Xcode 2.3. You may have found the thread that I started on darwin-drivers (kextload Bus Error When Patching C++ Code). The resolution is that it isn't possible to strip a kext in the manner you describe :) There are a number of symbols that must not be stripped. Dean Reece provided the following list: > Make sure you've used the OSDefine* and OSDeclare* macros correctly in > your project. Also, make sure you are not stripping all the symbols. > The kext patcher needs at least the following symbols to be able to > patch up your vtables if a superclass you use changes: > > For every class that inherits from OSObject Keep: > __ZTV<##><classname> > __ZN<##><classname>10gMetaClassE > __ZN<##><classname>10superClassE > > Also keep the following symbols, as they are used to load & unload > your kext. > _kmod_info > __antimain > __realmain > __start > __stop In my experience additional symbols are also required. However, this is somewhat moot due to the following issue: On 26/02/2005 at 12:03 PM, Godfrey van der Linden <gvdl(a)apple.com> wrote: > There is a latent bug in the patcher that may also be tripping you. A > reliable driver must have at least one local/non-exported i.e. static > symbol. This has to do with the way that I walk the symbol table > before patching. The code expects a local symbol, all of my test > cases had them (sigh), otherwise it will skip the first exported or > undefined symbol it finds. By invoking strip in the above manner, there will not be any local symbols left. So, I tried to find a way to strip a subset of the local symbols (i.e., those not required by Apple's kext patcher). However, the strip tool doesn't support stripping only a subset of local symbols. Apple's response to my bug report is: > Engineering has determined that this issue behaves as intended > based on the following information: > > "strip -S" is the appropriate incantation for current releases. We > may be able to tighten that in the future, but the kext system does > need access to some local symbols for the patching process. I wasn't thrilled with this option, so I devised the following workaround: I examined the symbols in each object file after it is compiled. If the object file doesn't contain any local symbols that need to be preserved, I strip it (i.e., strip -x). If it does, I leave it alone. Not perfect, but much better than my previous solution of preserving all local symbols. Later, when linking the kernel extension, I invoked strip with the -R option and pass a file that enumerates those global symbols that should be stripped. After these steps the kernel extension only exports the global symbols that I want and the local symbols that are required for patching. I didn't use Xcode at the time I did this; so, I'm not sure how easy it would be to integrate this functionality into it. geoff _______________________________________________ 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

Stripping IOKit kernel extensions
by Ryan McGann 12 Sep '06

12 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello, I am trying to strip my IOKit driver and am running into problems. Spefically, kextload crashes (bus error) when it attempts to load the kext when I strip it. It loads fine without stripping. I found a discussion about this very issue in the archives, but I didn't see a resolution for the OP. I have a symbols file which contains _kmod_info, and all symbols for my driver (both static data and methods). When I strip the executable, I use strip -u -s <symbols.txt> <path to kext binary>. I get a nice trim binary that contains only _kmod_info and the following symbols (I've included their demanged form, but the symbols file contains the mangled names of course): com_foo_bar_service_kext::gMetaClass com_foo_bar_service_kext::superClass com_foo_bar_service_kext::start(IOService*) com_foo_bar_service_kext::MetaClass::MetaClass() com_foo_bar_service_kext::MetaClass::MetaClass() #com_foo_bar_service_kext::MetaClass::~MetaClass() com_foo_bar_service_kext::metaClass com_foo_bar_service_kext::terminate(unsigned long) com_foo_bar_service_kext::com_foo_bar_service_kext(OSMetaClass const*) com_foo_bar_service_kext::com_foo_bar_service_kext() com_foo_bar_service_kext::com_foo_bar_service_kext(OSMetaClass const*) com_foo_bar_service_kext::com_foo_bar_service_kext() com_foo_bar_service_kext::~com_foo_bar_service_kext() com_foo_bar_service_kext::~com_foo_bar_service_kext() com_foo_bar_service_kext::getMetaClass() const com_foo_bar_service_kext::MetaClass::alloc() const vtable for com_foo_bar_service_kext vtable for com_foo_bar_service_kext::MetaClass I'm not sure what other symbols I need to include in the binary to prevent kextload from crashing. Is there a way to do this from Xcode? If not, what's the proper strip/nmedit command needed to do what is necessary? This is all on Tiger with Xcode 2.3. Thanks, Ryan -- Ryan rmcgann(a)mac.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

1 0

userspace access to PCI config space?
by Andrew Gallatin 12 Sep '06

12 Sep '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Is there any builtin (or even 3rd party) way to access PCI configuration space from a user process on Darwin/ppc? I need to port a tool of last resort that we have which accesses our NIC via PCI config space. I'm looking for something similar to what pciutils does on Linux/FreeBSD/Solaris. I found _chudReadPCIConfig and _chudWritePCIConfig in /System/Library/PrivateFrameworks/CHUD.framework/Frameworks/CHUDCore.framework/CHUDCore, and I figure this must be what "Reggie se" uses. Is there any hope of finding the header file for this? If an engineer is reading, I'd be happy to sign an NDA. Thanks, 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