March 2006 - Darwin-dev - lists.apple.com

Re: Your source browser sucks
by Graham J Lee 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5 (Macintosh/20051201) I was only joking, but it made for an attention-grabbing headline ;-) * Download of complete source archives for each project Cheers, Graham. -- Graham Lee UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 _______________________________________________ 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 28/3/06 19:55, Kevin Van Vechten wrote: Oh sure, defend Solaris but not my source browser... OK, so what I see is that it's "prettier" (i.e. enscript is run on the source files) and that there's source searching. These features are nice, no doubt. The source searching has been invaluable to me which is why I set this up; I was getting bored of searching google for "symbolName inurl:10.4.5.ppc site:darwinsource.opendarwin.org" However, I'm curious if anyone has any suggestions for integrating a few of our requirements that this doesn't seem to address: * Multiple releases hosted on the site: 10.0, ..., 10.3, ..., 10.4.5 That could be done by grabbing each release and storing it in SCM - opengrok can act like CVSweb in that regard if the SCM repository is local to the server. I only didn't do this because it wasn't immediately of use to me, but it certainly would be possible to come up with a script which: * rsyncs a given release from darwinsource.opendarwin.org * stages it into a temporary area * commits it to a CVS/SVN repository with a message like "changes for release 10.x.y" * updates the opengrok source tree from SCM then rebuilds the tags building this up would take a lot of time, it took the best part of today just to build the tags for 10.4.5.ppc. I'm not sure at the moment how to handle x86-vs-ppc or apple-vs-od branches, but I expect it could be done in the SCM. And as I originally said, the real benefits of using SCM (tracking changes/assigning blame ;-)) can't be used here because we don't have that information. But maybe tracking changes between releases is enough. TBF that's not handled by any of the existing source browsers either, but a separate rsync service. A bit of sqlite magic can accomplish the same thing too of course, using darwinbuild. I'd say that the ability to see which project provides a given file on the installed filesystem would be useful, too - darwinsource.opendarwin.org has this and darwingrok doesn't. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Your source browser sucks
by Kevin Van Vechten 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Oh sure, defend Solaris but not my source browser... * Multiple releases hosted on the site: 10.0, ..., 10.3, ..., 10.4.5 * Download of complete source archives for each project - Kevin On Mar 28, 2006, at 9:22 AM, Graham J Lee wrote: _______________________________________________ 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… OK, so what I see is that it's "prettier" (i.e. enscript is run on the source files) and that there's source searching. These features are nice, no doubt. However, I'm curious if anyone has any suggestions for integrating a few of our requirements that this doesn't seem to address: On 28/3/06 17:51, Andrew Gallatin wrote: Graham J Lee writes: > OpenGrok code browser from OpenSlowlaris. I've just put it up for my Please don't insult Solaris. It may be hard to use, but it is far from "Slowlaris" these days. Heh. Having been running it from 2.7-10, the word's just sortof hard-wired into my brain. You have to say "Stanislaw" before I type Solaris :) In any sort of macro or micro benchmark, it blows the doors off of darwin and/or macosx on the same (intel/ amd) hardware. We'll presumably be able to test Polaris/PPC vs. Darwin/PPC soon, too ;-) This email sent to site_archiver(a)lists.apple.com

1 0

Re: Your source browser sucks
by Andrew Gallatin 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Graham J Lee writes: > On 28/3/06 17:51, Andrew Gallatin wrote: > > Graham J Lee writes: > > > OpenGrok code browser from OpenSlowlaris. I've just put it up for my > > > > Please don't insult Solaris. It may be hard to use, but it is far > > from "Slowlaris" these days. > > Heh. Having been running it from 2.7-10, the word's just sortof > hard-wired into my brain. You have to say "Stanislaw" before I type > Solaris :) Yes, I remember "upgrading" to Solaris 2.1 from SunOS 4 back in 1993 or so. It was really "Slowlaris" then. Now they have a policy that if any other OS is faster at anything, then it is a bug in Solaris and should be fixed. This has helped quite a bit ;) Anwyay, it is cool that you got the OpenGrok code browser working with the darwin sources. I've used it quite a bit with the Solaris sources. 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

Re: Your source browser sucks
by Graham J Lee 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Thunderbird 1.5 (Macintosh/20051201) In any sort of macro or micro benchmark, it blows the doors off of darwin and/or macosx on the same (intel/amd) hardware. We'll presumably be able to test Polaris/PPC vs. Darwin/PPC soon, too ;-) -- Graham Lee UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 _______________________________________________ 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 28/3/06 17:51, Andrew Gallatin wrote: Graham J Lee writes: > OpenGrok code browser from OpenSlowlaris. I've just put it up for my Please don't insult Solaris. It may be hard to use, but it is far from "Slowlaris" these days. Heh. Having been running it from 2.7-10, the word's just sortof hard-wired into my brain. You have to say "Stanislaw" before I type Solaris :) This email sent to site_archiver(a)lists.apple.com

1 0

RE: Your source browser sucks
by Bill Lowrey 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Organization: istation Thread-index: AcZShtpooM80dqxRQ0q3WFaMlsWW1AABBG7w It's very nice. I found an answer to a problem I'd had previously in about 2 minutes using this. My previous search had taken hours and required downloading the source. Thanks! Bill Lowrey -----Original Message----- From: darwin-dev-bounces+billl=istation.com(a)lists.apple.com [mailto:darwin-dev-bounces+billl=istation.com@lists.apple.com] On Behalf Of Graham J Lee Sent: Tuesday, March 28, 2006 10:44 AM To: hackers(a)opendarwin.org; darwin-dev(a)lists.apple.com Subject: Your source browser sucks OK, so don't slashdot me, but I thought some of you might like to play around with this: http://darwingrok.physics.ox.ac.uk:8080/source/ It's the entire source for darwin 10.4.5.ppc expressed via the OpenGrok code browser from OpenSlowlaris. I've just put it up for my own work/interest at the moment, so whether it stays there depends on whether it gets used much and if it's more useful than the two darwinsource browsers [and potentially on other constraints, such as whether the box it's running on needs to be repurposed, or melts]. It currently doesn't have the darwinxref stuff that darwinsource.opendarwin.org does, but I think apart from that it's a better interface. I haven't bothered setting up any history/SCM information mainly because us external types don't have access to anything beyond "this changed somewhere between 10.4.x and 10.4.y" and because for what I do, that's not so interesting. And it would've taken time I don't currently have. Obviously if it looks like it's to become a permanent fixture, I'll do a better job of setting up tomcat ;-) Anyway, have fun, Graham. -- Graham Lee GPG Key ID: 01D5B9D8 UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 01865 273450 _______________________________________________ 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: Your source browser sucks
by Andrew Gallatin 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Graham J Lee writes: > OpenGrok code browser from OpenSlowlaris. I've just put it up for my Please don't insult Solaris. It may be hard to use, but it is far from "Slowlaris" these days. In any sort of macro or micro benchmark, it blows the doors off of darwin and/or macosx on the same (intel/amd) hardware. 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

Your source browser sucks
by Graham J Lee 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Anyway, have fun, Graham. -- Graham Lee GPG Key ID: 01D5B9D8 UNIX Systems Manager, Oxford Physics Practical Course http://users.ox.ac.uk/~wadh1342 01865 273450 _______________________________________________ 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 OK, so don't slashdot me, but I thought some of you might like to play around with this: http://darwingrok.physics.ox.ac.uk:8080/source/ It's the entire source for darwin 10.4.5.ppc expressed via the OpenGrok code browser from OpenSlowlaris. I've just put it up for my own work/interest at the moment, so whether it stays there depends on whether it gets used much and if it's more useful than the two darwinsource browsers [and potentially on other constraints, such as whether the box it's running on needs to be repurposed, or melts]. It currently doesn't have the darwinxref stuff that darwinsource.opendarwin.org does, but I think apart from that it's a better interface. I haven't bothered setting up any history/SCM information mainly because us external types don't have access to anything beyond "this changed somewhere between 10.4.x and 10.4.y" and because for what I do, that's not so interesting. And it would've taken time I don't currently have. Obviously if it looks like it's to become a permanent fixture, I'll do a better job of setting up tomcat ;-) smime.p7s

1 0

Re: db bug with i386: sanity check [BUG FOUND probably]
by Stephane Sudre 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.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… And the culprit is probably the __put_page function from hash_page.c which is not un-swapping a buffer after writing it to disk. So this leads to a corrupted data base in memory and the next call to a sync is saving corrupted data (header is correct but the rest of the file is incorrect). Adding some un-swapping code after the _write call seems to make things run fine on i386 with a BIG_ENDIAN data base. But I won't bet my right hand on this fix for the moment. AFAIK, the bug is also impacting the PPC platform because as soon as the lorder is not the same as BYTE_ORDER, the swap problem is going to happen. Sadly, this means that if you don't use a fix, hash data base saved by db on disks are not cross-endian compatible. I filed a bug report on the FreeBSD site since this is where the original source code is coming from. This email sent to site_archiver(a)lists.apple.com

1 0

Re: zones in NKEs
by Michael Tuexen 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hi Chris, see my comments in-line. Best regards Michael On Mar 28, 2006, at 12:20 AM, Chris Sarcone wrote: Michael -- -- Chris ------------------ 6 Infinite Loop M/S 306-2MS Cupertino CA 95014 phone: (408) 974-4033 fax: (408) 862-7577 email: sarcone(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… we are providing the FreeBSD/KAME implementation of a transport protocol (SCTP), which uses zones, as an NKE. It is just that we want to use the same code as much as possible. I figured out that it works, you can unload the NKE and load it again, but after unloading it, zprint crashes the machine. I would assume that's because you provide a string to zinit and it doesn't copy it locally. When host_zone_info() is called, it strncpy ()'s the zn_name and kerblammo because your NKE is now gone. You could file a Radar for it to do the local copy. However, you probably want to be able to unload and when you load You are correct. I'll file a bug report. again, not create another zone, but re-use the same zone. That might require more work to the zone allocator. Feel free to write some bugs up on whatever else you need too... Yes, reusing it would be fine. I'll report that too. This email sent to site_archiver(a)lists.apple.com

1 0

Re: AuthServer StartupItem & File Paths
by Stéphane Sudre 28 Mar '06

28 Mar '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On mardi, mars 28, 2006, at 01:43 AM, Kevin Van Vechten wrote: _______________________________________________ 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… In a way, you've answered your own question. No it's not "normal" to use relative paths, since most things use absolute paths. But if the real intent of the question is whether this is unsafe, then the answer is no, probably not unsafe. /etc/rc.common sets the path to a relatively restricted set of directories: PATH=/bin:/sbin:/usr/bin:/usr/sbin:/usr/libexec:/System/Library/ CoreServices Ok, thanks. Regarding the safety side, I had no real trouble with that after a quick check of the permissions. This email sent to site_archiver(a)lists.apple.com

1 0