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

Re: Darwin-dev Digest, Vol 2, Issue 125
by Mark Day 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 25, 2005, at 2:40 PM, darwin-dev-request(a)lists.apple.com wrote: Okay, I think I understand, but that still doesn't fully explain the following behaviour: 1) $ mdfind hamishtest3 2) $ cat > foo.txt 3) hamishtest3 4) ^D 5) $ mdfind hamishtest3 6) /Users/hamish/foo.txt 7) $ mkdir bar 8) $ ln foo.txt bar/bar.txt 9) $ mdfind hamishtest3 10) /Users/hamish/foo.txt 11) $ cat >> bar/bar.txt 12) extra 13) ^D 14) $ mdfind hamishtest3 15) /Users/hamish/bar/foo.txt 16) $ rm foo.txt 17) $ mdfind hamishtest3 18) $ mdimport bar/bar.txt 19) $ mdfind hamishtest3 20) /Users/hamish/bar/foo.txt In line 15 above, the last time the vnode for foo.txt was looked up was via /Users/hamish/bar/, so that's where foo.txt is reported. But there is no file /Users/hamish/bar/foo.txt, and nor has there ever been, so something is wrong with the caching. Whatever part of the system is caching foo.txt shouldn't be doing so once the write to bar/bar.txt has happened. And why, even after I've removed foo.txt and imported bar/bar.txt, hasn't it been updated yet in line 20? It would be good if you could return all of them: anything which would want a vnode-to-path-lookup would surely benefit from having them all? Instead of caching a single parent node, you would maintain a linked list (or whatever) of all parent nodes. -Mark _______________________________________________ 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 about line 11, the lookup of bar/bar.txt caused the parent directory associated with the (bar.txt and foo.txt) vnode to become the bar directory. It probably should have also changed the name associated with the vnode to bar.txt. It would be worth filing a bug on that. Because the vnode continued to exist for the entire duration, leaving the vnode's associated name and parent unchanged. Had the vnode been recycled and looked up again, that lookup would have created a new vnode with new parent and name. For a fun experiment, try the following between lines 13 and 14: $ find -x /System -ls &> /dev/null I'll bet that the output in line 15 would be "/Users/hamish/bar/ bar.txt". First of all, it's not just a list of parent nodes. It would be a list of (parent,name) tuples. And if you allowed hard links to directories, it would have to be a list of complete paths from the root. Note that you'd have to maintain this list both on disk and in memory. (You'd want to get a list of all paths, regardless of whether they were used recently to look up the file, right?) As far as I know, NTFS is the only popular volume format that maintains this kind of mapping from one "inode" to multiple directory entries. It would certainly be more complex. I'd be concerned about locking issues, though I haven't thought that out all the way. This email sent to site_archiver(a)lists.apple.com

1 0

Re: xattr on SMB and NTFS
by Paul Nelson 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Microsoft-Entourage/11.1.0.040913 Two things: In Tiger, the Finder deletes the meta file (._xxx) if the file has no resource fork. If you sniff a copy across the network, you may see the meta file get created, but later deleted. This is different from Panther. Second, No release of OS X by Apple supports the NTFS streams used by Services for Mac. If you are trying to convert NTFS files on the fly (as needed) stored by SFM, DAVE and ExtremeZ, this would be very difficult with Apple's SMB client. You would need to write code to convert them all over, then stop using these products. This is a case of Apple not following a de facto standard used by all other vendors I know of. Paul Nelson CTO Thursby Software Systems, Inc. > From: "Philip D. Wasson" <pwasson(a)maned.com> > Date: Wed, 25 May 2005 19:27:11 -0400 > To: <darwin-dev(a)lists.apple.com> > Subject: xattr on SMB and NTFS > > I'm not sure if this is the correct list for this question, but it > seems like it. > > I'm responsible for code that might have to manipulate Mac-created > files, stored on Windows fileservers, from Windows, so I'm looking into > how Mac OS X stores extra information in AppleDouble files. Now that > Tiger has extended attribute support, I thought I should look into how > that stuff is stored too. So I wrote a little program that can list, > get, or set xattrs on a specified file (if there's already a program > that does this, I don't know what it's called, but no big deal). I > created an empty file, stored two different attributes, copied it using > Finder to a fileserver mounted via SMB (stored on an NTFS volume), and > then looked at its xattrs. To my surprise, the attributes I had set > were missing, but I didn't get the error I expected. And I was able to > set new attributes. When I listed them, I was again surprised to find > an AFP_AfpInfo attribute, which is the name of the NTFS stream in which > various servers store Mac file metadata. Presumably that attribute was > created by a Services for Macintosh server running on that server > machine. Plus there was no extra AppleDouble file. I used the SetFile > utility to set the filetype of the file, and an AppleDouble file > appeared; the AFP_AfpInfo xattr still had a default filetype in it, not > the one I'd just set. Oh, and the attribute I set on the file after > copying, when it resided on the server, showed up in a Windows program > that lists additional data streams. > > This all seems rather inconsistent. xattrs don't get copied by the > Finder when copying over SMB to NTFS even though it's supposed to be > using the vaunted new system-supplied file-copying engine library. > Setting xattrs on SMB-mounted files stores them in NTFS streams, but > Mac metadata ,and presumably resource forks, set on Mac go into > AppleDouble files even though they could store it in a stream the way > that MS' Services for Macintosh, ExtremeZ-IP, and DAVE do. Is there > some order to this that I'm missing? Are there docs that say what's > stored where and when? It looks like if I want to copy a Mac file from > Windows, I have to copy both the streams and the AppleDouble file; I > figured it would be one or the other. > > ---------------------------------------------------------------------- > Philip D. Wasson pwasson at maned dot com > Software Engineer Managing Editor Inc. > > _______________________________________________ > 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/nelson%40thursby.com > > This email sent to nelson(a)thursby.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

Re: Undefined symbols: _rtrequest
by Josh Graessley 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com This just sort of scratches the surface. -josh On May 26, 2005, at 8:36 AM, Philip George wrote: On May 26, 2005, at 10:32 AM, Justin Walker wrote: What are you trying to do? adding and deleting custom routes. running 10.3.9 thanks. - philip This email sent to jgraessley(a)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 If you're in user space, you need to use a routing socket. If you're in the kernel, you are better off using a routing socket from the kernel than using the rtrequest function directly. rtrequest is off limits in 10.4. For 10.3.x there should be two symbol kexts you can link against as well as the kernel itself when declaring your OSBundleLibraries in your kext's Info.plist. I can't recall the names but you should be able to figure them out using kextstat. kextstat will list the kernel library kexts at the top. They typically have an address, size, and length of 0. From Tiger (10.4.1) the library list is: 1 1 0x0 0x0 0x0 com.apple.kernel (8.1.0) 2 13 0x0 0x0 0x0 com.apple.kpi.bsd (8.1.0) 3 14 0x0 0x0 0x0 com.apple.kpi.iokit (8.1.0) 4 16 0x0 0x0 0x0 com.apple.kpi.libkern (8.1.0) 5 14 0x0 0x0 0x0 com.apple.kpi.mach (8.1.0) 6 15 0x0 0x0 0x0 com.apple.kpi.unsupported (8.1.0) 7 1 0x0 0x0 0x0 com.apple.iokit.IONVRAMFamily (8.1.0) 8 1 0x0 0x0 0x0 com.apple.driver.AppleNMI (8.1.0) 9 1 0x0 0x0 0x0 com.apple.iokit.IOSystemManagementFamily (8.1.0) 10 1 0x0 0x0 0x0 com.apple.iokit.ApplePlatformFamily (8.1.0) 11 66 0x0 0x0 0x0 com.apple.kernel.6.0 (7.9.9) 12 1 0x0 0x0 0x0 com.apple.kernel.bsd (7.9.9) 13 1 0x0 0x0 0x0 com.apple.kernel.iokit (7.9.9) 14 1 0x0 0x0 0x0 com.apple.kernel.libkern (7.9.9) 15 1 0x0 0x0 0x0 com.apple.kernel.mach (7.9.9) The com.apple.kpi.* are the new Tiger symbol sets. com.apple.kernel.* are the backwards compatible ones that have been stripped of everything that isn't supported on Tiger. The goal was to allow existing drivers using IOKit to continue to work by exporting the few symbols that provided functionality they didn't get from IOKit. com.apple.kernel is the kernel itself. This will give you access to every symbol in the kernel. If you link against this, you will break with every software update since the compatible version of the kernel always matches the current version. rtrequest() is a procedure in the kernel; it is not in any library. In 10.4 and later, it may be off-limits (not available via the KPI); I haven't verified this. can you point me in the direction of instructions on how i can get that function linked into my binary. _______________________________________________ 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/jgraessley% 40apple.com smime.p7s

1 0

Re: Undefined symbols: _rtrequest
by Philip George 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 26, 2005, at 10:32 AM, Justin Walker wrote: adding and deleting custom routes. running 10.3.9 thanks. - philip _______________________________________________ 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… What are you trying to do? rtrequest() is a procedure in the kernel; it is not in any library. In 10.4 and later, it may be off-limits (not available via the KPI); I haven't verified this. can you point me in the direction of instructions on how i can get that function linked into my binary. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Undefined symbols: _rtrequest
by Justin Walker 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On May 26, 2005, at 2:46, Philip George wrote: What are you trying to do? Regards, Justin -- Justin C. Walker, Curmudgeon-At-Large Institute for General Semantics -------- "Weaseling out of things is what separates us from the animals. Well, except the weasel." - Homer J Simpson -------- _______________________________________________ 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… which library do i need to link in to get rtrequest()? rtrequest() is a procedure in the kernel; it is not in any library. In 10.4 and later, it may be off-limits (not available via the KPI); I haven't verified this. This email sent to site_archiver(a)lists.apple.com

1 0

Undefined symbols: _rtrequest
by Philip George 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com which library do i need to link in to get rtrequest()? - philip _______________________________________________ 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: Bizarre out of memory bug
by Alexey Proskuryakov 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Microsoft-Entourage/11.1.0.040913 On 25.05.2005 14:40, "Tim Seufert" <tas(a)mindspring.com> wrote: > Bizarre thing #2: So, "yobgorgle" is really "yobgorgle.local" since I > have "local" in my search domains list. In other words, it's a > Ren^h^h^h Bonjour name. On a whim, I decided to SSH to the actual IP > address: > > [mightygorilla:~] tas% ssh 192.168.1.40 When you connect to a Bonjour address, ssh probably uses IPv6 - at least, Little Snitch'es reaction to such connections is very different on my machine. Try connecting as ssh -4 yobgorgle.local I have also seen very strange things with scp being unable to copy files unless Remote Login were enabled on the source(!) machine (hadn't time to investigate and file in Radar). - WBR, Alexey Proskuryakov _______________________________________________ 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

__builtin_frame_address()
by Steve Checkoway 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com struct Frame { struct Frame *sp; long cr; void *lr; }; register void *r1 __asm__("r1"); Frame *frame = (Frame *)r1; while (frame) { if (frame->lr) /* Do something with the return address */ frame = frame->sp; } If it makes any difference, I am doing this in c++. - Steve _______________________________________________ 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 Is __builtin_frame_address() broken? I have some backtracing code that walks the stack essentially using: It seems to me that I could remove the register void *r1 __asm__ ("r1"); and change r1 in the next line to __builtin_frame_address(0); but this doesn't seem to work. The value returned is completely bizarre. On a slightly related note, the first few values that frame->lr holds are in unrelated functions that are not on the stack at the time. This is new in gcc 4. Prior versions seemed to provide sane addresses. After the first few functions, the stack seems correct. smime.p7s

1 0

Re: GCC 4.0 version
by Bill Northcott 26 May '05

26 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Inspecting the source code of the latest Darwin release (8.1), I see that the GCC 4.0 version shipped is 4.0.0 20041026 (Apple Computer, Inc. build 4061) Are there plans to upgrade this? The official GNU version 4.0 includes bugfixes commited after 2004/10/26. I'm asking because Darwin 8.1 is 6 months newer than the GCC code it ships with. Bill Northcott _______________________________________________ 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 26/05/2005, at 4:05 AM, darwin-dev-request(a)lists.apple.com wrote: You can build more recent versions of the Apple compiler from the cvs code. Of course cautions about pre-release software apply. For details of accessing the cvs see http://gcc.gnu.org. To get Apple code, you need to use a tag to extract cvs code. Currently tags like apple-ppc-50xx get code which has been merged with the gcc-4.0.0 release code. It is Apple's practice to freeze compiler code several months before system releases. If building the cvs code is daunting, you can find a binary for apple- gcc-5017 at http://www.swarm.org/wiki/Swarm:_MacOS_X_binaries This email sent to site_archiver(a)lists.apple.com

1 0

xattr on SMB and NTFS
by Philip D. Wasson 25 May '05

25 May '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com ---------------------------------------------------------------------- Philip D. Wasson pwasson at maned dot com Software Engineer Managing Editor Inc. _______________________________________________ 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 not sure if this is the correct list for this question, but it seems like it. I'm responsible for code that might have to manipulate Mac-created files, stored on Windows fileservers, from Windows, so I'm looking into how Mac OS X stores extra information in AppleDouble files. Now that Tiger has extended attribute support, I thought I should look into how that stuff is stored too. So I wrote a little program that can list, get, or set xattrs on a specified file (if there's already a program that does this, I don't know what it's called, but no big deal). I created an empty file, stored two different attributes, copied it using Finder to a fileserver mounted via SMB (stored on an NTFS volume), and then looked at its xattrs. To my surprise, the attributes I had set were missing, but I didn't get the error I expected. And I was able to set new attributes. When I listed them, I was again surprised to find an AFP_AfpInfo attribute, which is the name of the NTFS stream in which various servers store Mac file metadata. Presumably that attribute was created by a Services for Macintosh server running on that server machine. Plus there was no extra AppleDouble file. I used the SetFile utility to set the filetype of the file, and an AppleDouble file appeared; the AFP_AfpInfo xattr still had a default filetype in it, not the one I'd just set. Oh, and the attribute I set on the file after copying, when it resided on the server, showed up in a Windows program that lists additional data streams. This all seems rather inconsistent. xattrs don't get copied by the Finder when copying over SMB to NTFS even though it's supposed to be using the vaunted new system-supplied file-copying engine library. Setting xattrs on SMB-mounted files stores them in NTFS streams, but Mac metadata ,and presumably resource forks, set on Mac go into AppleDouble files even though they could store it in a stream the way that MS' Services for Macintosh, ExtremeZ-IP, and DAVE do. Is there some order to this that I'm missing? Are there docs that say what's stored where and when? It looks like if I want to copy a Mac file from Windows, I have to copy both the streams and the AppleDouble file; I figured it would be one or the other. This email sent to site_archiver(a)lists.apple.com

1 0