October 2007 - Darwin-dev - lists.apple.com

RE: How can I use a .dylib created in Xcode in my gcc-compiled program?
by Milford Brimdash 16 Oct '07

16 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Importance: Normal Boris Remizov said: > Hmmm...> Try to play with extern "C" directive before your func (). That is only part of the solution. For example, if I have a library I made named libmine.dylib, I can compile my program in two ways: g++ main.cpp libmine.dylib -o run (assuming libmine.dylib is in the same directory) or g++ main.cpp -L. -lmine -o run In both cases it compiles (what is the difference??), but I get the following error when I try to execute the program: $ ./run dyld: Library not loaded: /usr/local/lib/libmine.dylib Referenced from: /Users/milford/program/./run Reason: image not found Trace/BPT trap Another thing I've noticed...if I compile a .dylib in Xcode with the C compiler, I can call a function from the library in a separate program, but if I recompile the .dylib using the C++ compiler (no changes to the code though!) I can't even compile and link the program that uses the library. I'm aware these are very trivial problems, and ordinarily I can figure them out myself with ease, but not in this case because of the total lack of documentation! I can't find documentation on building .dylibs with Xcode or on using them. I am pretty green with OS X development, if anyone could point me in the right direction that'd be great. If documentation exists, I can figure it out myself, but without it I'm just grappling in the darkness :( Thanks, Milford _________________________________________________________________ Boo! Scare away worms, viruses and so much more! Try Windows Live OneCare! http://onecare.live.com/standard/en-us/purchase/trial.aspx?s_cid=wl_hotmail… _______________________________________________ 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: Question about bigendian/littleendian dependencies with universal binaries.
by Dave Zarzycki 15 Oct '07

15 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com http://developer.apple.com/documentation/MacOSX/Conceptual/universal_binary… davez On Oct 15, 2007, at 11:24 AM, Terry Simons wrote: Hi, Any help is appreciated. - 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/zarzycki%40apple.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… Yes, we needed to adjust our autoconf programs to use compiler supplied macros in order to build Universal Binaries correctly on Mac OS X. The following URL discusses what to do: I'm working on an application that needs to deal with big and little endian differences beyond the standard networking "host to network" and "network to host" API calls, and I'm curious how this will affect the creation of a Universal Binary. We're using the autoconf tools to configure the project, and I know that it's possible to use AC_C_BIGENDIAN to elicit -DWORDS_BIGENDIAN (or not) on the appropriate platform, but it seems as if this would always be defined if the build host is big endian, and never defined if the build host is little endian. So what's the proper way to deal with a situation such as this? Is there some sort of define that magically happens for me when the compiler is targeting x86 on a ppc and vice versa? Should we be using some sort of magic token to do the determination at runtime? If so, is there an example somewhere of how to do this? This email sent to zarzycki(a)apple.com This email sent to site_archiver(a)lists.apple.com

1 0

Question about bigendian/littleendian dependencies with universal binaries.
by Terry Simons 15 Oct '07

15 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi, Any help is appreciated. - 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… I'm working on an application that needs to deal with big and little endian differences beyond the standard networking "host to network" and "network to host" API calls, and I'm curious how this will affect the creation of a Universal Binary. We're using the autoconf tools to configure the project, and I know that it's possible to use AC_C_BIGENDIAN to elicit -DWORDS_BIGENDIAN (or not) on the appropriate platform, but it seems as if this would always be defined if the build host is big endian, and never defined if the build host is little endian. So what's the proper way to deal with a situation such as this? Is there some sort of define that magically happens for me when the compiler is targeting x86 on a ppc and vice versa? Should we be using some sort of magic token to do the determination at runtime? If so, is there an example somewhere of how to do this? This email sent to site_archiver(a)lists.apple.com

1 0

How to Identify a Bootable System?
by Duane Murphy 15 Oct '07

15 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com How can I determine bootable partitions like the Startup Disk Preference Pane? Our application needs to be able to identify disk partitions that maybe bootable in order to properly guide the user of the application. How does the Startup DIsk Preference Pane identify a bootable partition? Thanks, ...Duane _______________________________________________ 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: How can I use a .dylib created in Xcode in my gcc-compiled program?
by Boris Remizov 11 Oct '07

11 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:in-reply-to:references:mime-version:content-type:message-id:cc:from:subject:date:to:x-mailer; bh=0m2NFlqDmr33iK8dPj958zlJvp6UBDbq2pKJQGEFJ1o=; b=QX0S1TZhaP3ldncxwHNKAAorYW7UGrIjOeAG4trFb84eE3TYDc2BTE8HvS8ZYRKzCFz/qJgJTRjd0RVIfk4Zwi4xNNeAFkculVTWz4It74gp4pUGdqi9VfnmBUeVm5SFrhq8aIQOHQRCpwjBqGagftJ1aKjcp4PFh3oqY5P5HVo= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:in-reply-to:references:mime-version:content-type:message-id:cc:from:subject:date:to:x-mailer; b=BflYY8OWNhiXV4/JCztmnAUZCQgE3NSOwDMv4YF3sQ5IwAvOVSiH+ORkOanfmVUmvkyJWgf9/EN/U3npPCR6joEyEyywizc1whtNJf+u1keNwQXIWscSGzk8Sc58cnM/PLPqIhsUvoXWHwrFZSzNOajnzvmeiMsxLmbajEarWFk= _______________________________________________ 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

How can I use a .dylib created in Xcode in my gcc-compiled program?
by Milford Brimdash 10 Oct '07

10 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Importance: Normal Howdy, I've spent a few hours trying to do somethin' very simple: I have a .dylib I created in Xcode (BSD Dynamic Library) that I want to use in a command-line program I am compilin' with gcc. How do I do this? The following page is the only help I've found on includin' a .dylib when using gcc: http://developer.apple.com/documentation/DeveloperTools/Conceptual/DynamicL… Based on it, I am tryin': $ g++ main.cpp Lib.dylib -o test But I get the followin' errors: /usr/bin/ld: Undefined symbols: func() ... etc. The queer thing is, if I take .c and .h files from the Xcode project and compile 'em into a .dylib by hand, like so: $ g++ -dynamiclib Lib.c -o Lib.dylib Then I can compile my program just fine, as above! This solution ain't acceptable, however, because then the .dylib won't work with Maple11 worksheet for some reason. When I look at the build details in Xcode, I see all sortsa flags like like -fpascal-strings and -no_farm_animals but I have no idea why the same exact code compiled by Xcode works in Maple11 but not with my gcc-compiled program, but the gcc-compiled .dylib works with my gcc-compiled program but not Maple11! (Tarnation!) Any help will be 'ppreciated. Milford Brimdash (ps. usin' Xcode to write my command-line program is out, since when I right-click on "Link Binary with Libraries" and add the Lib.dylib, it then complains durin' build that "can't locate file for: -ILib") _________________________________________________________________ Help yourself to FREE treats served up daily at the Messenger Café. Stop by today. http://www.cafemessenger.com/info/info_sweetstuff2.html?ocid=TXT_TAGLM_OctW… _______________________________________________ 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: Identifying directory changes with kqueue/kevent
by Uli Kusterer 07 Oct '07

07 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Cheers, -- M. Uli Kusterer http://www.zathras.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… Am 19.09.2007 um 17:43 schrieb Fred Leboucher: I'm trying to watch a directory for changes, so the couple kqueue/ kevent could do the job (aside from the Carbon FNSubscribe). Only problem is that, like FNSubscribe, kqueue/kevent can't tell what file in the directory changed (or maybe I missed something ?). So, is there any simple way to identify the changed file, or do I have to keep track of all files in the directory, and compare this to the directory every time a change occurs ? This could be quite expensive and handling events like file renaming would be a real pain... I don't think you necessarily have to keep track of much. One of the dates (last accessed, last changed, created etc.) will likely have changed by the time you get your kevent (that's a guess, back when I did stuff like that, I didn't really think of that approach, so you will have to verify that part). If my guess is right, you could probably just note down the newest date at the time of your last check, and any file with a newer date must have changed in the meantime. Also, FSGetCatalogInfoBulk() or maybe FSCatSearch() or whatever it was called should be fast enough to get that information for a folder when it changes. I've found that, since you've already narrowed down the files to be examined using kqueue, scanning a single folder should work pretty well. In addition, you could use a timer to coalesce several messages on the same folder that arrive in a short succession, to avoid having to re-scan the same folder several times. This email sent to site_archiver(a)lists.apple.com

1 0

[OT] Do dylibs have to be marked as executable?
by Karan, Cem (Civ, ARL/CISD) 04 Oct '07

04 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AcgGfbgv+EUWiTVfQ+GdfwCLDuIIRw== Thread-topic: [OT] Do dylibs have to be marked as executable? This is an off-topic question, but I think this group has the best chance of answering it correctly. I've been working on hardening a OS X machine I have (not plain Darwin) and one of the things I've been doing is changing the permissions of files so that they are the minimal necessary for them to work correctly. Part of that is changing files that are marked as executable to no longer being executable, and I ran into a peculiar problem; I don't know if *.dylib and *.so can be safely set to not being executable. Part of me says that it is OK because they aren't applications, but are merely linked into applications, which are marked as executable. Another part of me says that they are code, and so need to be executable. ` find / \! -perm +100 -type f -name '*.dylib'` does turn up some files that are marked as not being executable in this manner for both, but you can also find a bunch that are executable. So, which is it? Should they be executable or not? Thanks, Cem Karan _______________________________________________ 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: Strange Behavior When Looking for Hex Values
by Joe van Tunen 02 Oct '07

02 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AcgE+WGMoBGBjHDsEdyz+wADk61ysgAI5OVY Thread-topic: Strange Behavior When Looking for Hex Values User-agent: Microsoft-Entourage/11.3.6.070618 Power PC instructions are all 4 bytes long so it could take two instructions to load a 4 byte value into a register. The instructions are not always next to each other or in the same order or the same instruction. The 4 bytes may not appear in the code at all. For example, loading 0xDEADBEEF into register r3 could look like this: 3C60DEAE lis r3,-8530 <-- loads 0xDEAE instead of 0xDEAD ... <-- any number of instructions here 3863BEEF subi r3,r3,16657 <-- finish loading r3 You should download a Motorola PowerPC instruction manual (like "PowerPC 601 RISC Microprocessor User's Manual" or a "Programming Environments" manual). You need something like the objdump (or gobjdump) or otool command line tools to examine a binary file properly. They can dump the various sections of the binary file and disassemble the code sections. > From: James Kelly <JKelly(a)nemetschek.net> > Date: Tue, 02 Oct 2007 09:37:31 -0400 > To: <darwin-dev(a)lists.apple.com> > > Well, this behavior is strange to me at least, and my co-workers seem a bit > surprised as well. I'm willing to entertain the idea that we are missing > something fundamental, however (in fact, I'd say that's the only way to > explain this). > > I am searching for a series of hex values in a binary file, they are defined > as two Uint32s. On the intel side of the universal binary, they can be > searched for as expected, using a hex editor for example. When looking at > the ppc code, expecting to see the values reversed, we don't find them as > their whole Uint32 selves, but broken up. Interestingly enough, they are > broken into smaller pieces and interwoven with each other. An example would > probably better explain. Say my hex values were as follows (for > simplicity's sake, we'll assume these are already in the ppc byteorder): > > 0x12345678 > 0x9ABCDEF0 > > In the ppc side of the code, the values will be found like so (using 00 to > represent random fill-in values): > > 12 34 00 00 9A BC 00 00 56 78 00 00 DE F0 > > So as you can see, the first four hex digits from the first value, then four > unrelated hex digits, then the first four hex digits from the second value, > then four unrelated hex digits, then the last four hex digits from the first > value, and so on. > > Is this the way it is supposed to be? Is there anyone who could point me to > a resource to read up on this behavior and similar issues? Thanks in > advance. And I apologize for the verbosity, I just wanted to be clear on > the problem. > > Jim > **************************************************************************** This email may contain material confidential to Pearson. If you were not an intended recipient, please notify the sender and delete all copies. We may monitor email to and from our network. **************************************************************************** _______________________________________________ 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

Strange Behavior When Looking for Hex Values
by James Kelly 02 Oct '07

02 Oct '07

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Thread-index: AcgE+WGMoBGBjHDsEdyz+wADk61ysg== Thread-topic: Strange Behavior When Looking for Hex Values User-agent: Microsoft-Entourage/11.3.6.070618 Well, this behavior is strange to me at least, and my co-workers seem a bit surprised as well. I'm willing to entertain the idea that we are missing something fundamental, however (in fact, I'd say that's the only way to explain this). I am searching for a series of hex values in a binary file, they are defined as two Uint32s. On the intel side of the universal binary, they can be searched for as expected, using a hex editor for example. When looking at the ppc code, expecting to see the values reversed, we don't find them as their whole Uint32 selves, but broken up. Interestingly enough, they are broken into smaller pieces and interwoven with each other. An example would probably better explain. Say my hex values were as follows (for simplicity's sake, we'll assume these are already in the ppc byteorder): 0x12345678 0x9ABCDEF0 In the ppc side of the code, the values will be found like so (using 00 to represent random fill-in values): 12 34 00 00 9A BC 00 00 56 78 00 00 DE F0 So as you can see, the first four hex digits from the first value, then four unrelated hex digits, then the first four hex digits from the second value, then four unrelated hex digits, then the last four hex digits from the first value, and so on. Is this the way it is supposed to be? Is there anyone who could point me to a resource to read up on this behavior and similar issues? Thanks in advance. And I apologize for the verbosity, I just wanted to be clear on the problem. Jim _______________________________________________ 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