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

Re: flushing HFS cache to disk
by Timothy J. Wood 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com But all the filesystem operations _do_ have transactional semantics. fs_begin_transaction(xxx); { ... do stuff ... } fs_{commit,rollback}_transaction(); -tim _______________________________________________ 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 Nov 30, 2005, at 1:53 PM, Peter Bierman wrote: The issue here is that someone is trying to read from the disk _behind the back_ of the filesystem. There is NO provision for that while the filesystem is in use. They have transactional semantics in that each individual operation is atomic, but they don't in that you can't group fs operations into a transaction. Imagine some new syscalls: In this particular case, you'd like 'xxx' to specify "don't show me writes made by other processes", but a mode where you can prevent your changes from becoming visible until you commit would be very useful too (say, if you want to create two files and have them show up atomically). Obviously there are a whole ton of design issues with doing this, problems this couldn't solve, and filesystems where this couldn't be implemented easily (if at all). This email sent to site_archiver(a)lists.apple.com

1 0

Re: Determining the canonical case for a file name on HFS+
by Adam Nohejl 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com 2005/11/30 v 22:05, Camillo Lugaresi: Hello to all subscribers. -- Adam _______________________________________________ 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 am looking for a way to determine the canonical form for a file name. Suppose that there is a file named "FILE" in the directory / folder/, which is on an HFS+ volume: then the canonical form of the path "/folder/file" would be "/folder/FILE". Both paths refer to the same file, but the representation on disk is FILE. I need to know this for the purpose of comparison, and I cannot simply do a case-insensitive comparison because I do not know if / folder/ is on a case-insensitive or case-sensitive volume. I have thought of opening the file and then using fstat on the descriptor, but is there a better way to do it? >From a broader point of view, what I need is a reliable solution for determining if two paths refer to the same file, without assuming anything about the filesystem on which it is located. I have searched the archives of this list, but I could not find an answer to this question. I wouldn't open the file, but using stat() and comparing inodes seems like a good solution to me. This email sent to site_archiver(a)lists.apple.com

1 0

Re: flushing HFS cache to disk
by Peter Bierman 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com What happens after that synchronous call returns...? Are you expecting it to lock out the file system from any other access? It other words a synchronous call serves no purpose since at any moment another thread can come along and modify the state of the file system as long as it is mounted and writable. But all the filesystem operations _do_ have transactional semantics. -pmb _______________________________________________ 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 1:35 PM -0800 11/30/05, Dave Yost wrote: At 12:22 PM -0800 2005-11-30, Shawn Erickson wrote: On 11/30/05, Dave Yost <ListMailForDave(a)yost.com> wrote: At 01:32 PM -0800 2005-11-29, Mark Day wrote: >On Nov 29, 2005, at 12:03 PM, darwin-dev-request(a)lists.apple.com wrote: > >>>that's what sync(2) does. >> >>It doesn't seem to. > >Beware that sync(2) is an asynchronous call. It just begins the flushing of data to all volumes. If you're reading the disk immediately, the changes may not have been written yet. Surely there should be a synchronous call one can use. Well, it could be specified to guarantee that on return, everything that was dirty at the time of the call has been flushed to disk, whether or not anything has been modified during the interval. That is a purpose that would be served. Someday, I wish for a file system that is really a true database, with transactional semantics. The issue here is that someone is trying to read from the disk _behind the back_ of the filesystem. There is NO provision for that while the filesystem is in use. This email sent to site_archiver(a)lists.apple.com

1 0

Re: flushing HFS cache to disk
by Dave Yost 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com At 12:22 PM -0800 2005-11-30, Shawn Erickson wrote: >On 11/30/05, Dave Yost <ListMailForDave(a)yost.com> wrote: >> At 01:32 PM -0800 2005-11-29, Mark Day wrote: >> >On Nov 29, 2005, at 12:03 PM, darwin-dev-request(a)lists.apple.com wrote: >> > >> >>>that's what sync(2) does. >> >> >> >>It doesn't seem to. >> > >> >Beware that sync(2) is an asynchronous call. It just begins the flushing of data to all volumes. If you're reading the disk immediately, the changes may not have been written yet. >> >> Surely there should be a synchronous call one can use. > >What happens after that synchronous call returns...? Are you expecting >it to lock out the file system from any other access? > >It other words a synchronous call serves no purpose since at any >moment another thread can come along and modify the state of the file >system as long as it is mounted and writable. Well, it could be specified to guarantee that on return, everything that was dirty at the time of the call has been flushed to disk, whether or not anything has been modified during the interval. That is a purpose that would be served. Someday, I wish for a file system that is really a true database, with transactional semantics. 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

1 0

Determining the canonical case for a file name on HFS+
by Camillo Lugaresi 30 Nov '05

30 Nov '05

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:mime-version:content-transfer-encoding:message-id:content-type:to:from:subject:date:x-mailer; b=oH2t3hrKBHFGE9fDbcWOj2Tn8TwRsy3wE5wialH7GQbsqF2AX9n3IT6/VStBXgCPKPS7lN5fqeFPRBUnrWucxWBGH3XvZxg4u/GwceMI9LGvDKaRfgyl/pNISKKK9jsQUOU9HL7EWMEdr2zbAm2B9RIcXhbFVJZdTQdVgEVzwQs= Hello to all subscribers. Thanks, 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… I am looking for a way to determine the canonical form for a file name. Suppose that there is a file named "FILE" in the directory / folder/, which is on an HFS+ volume: then the canonical form of the path "/folder/file" would be "/folder/FILE". Both paths refer to the same file, but the representation on disk is FILE. I need to know this for the purpose of comparison, and I cannot simply do a case-insensitive comparison because I do not know if / folder/ is on a case-insensitive or case-sensitive volume. I have thought of opening the file and then using fstat on the descriptor, but is there a better way to do it? >From a broader point of view, what I need is a reliable solution for determining if two paths refer to the same file, without assuming anything about the filesystem on which it is located. I have searched the archives of this list, but I could not find an answer to this question. This email sent to site_archiver(a)lists.apple.com

1 0

Re: missing project archive needed.
by Kevin Van Vechten 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Thanks, - Kevin On Nov 29, 2005, at 6:32 PM, Dale Walsh wrote: Any ideas??? _______________________________________________ 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 don't think the PerlObjCBridge is open source. But if you're referring to a dead-link then please reply with the URLs so I can investigate further. Going through the modules I need to build, some appear to be appear to be missing although I believe I can get away with the last one. PerlObjCBridge 22.1 PerlObjCBridge 22.3 PerlObjCBridge 22.5 Also this module (version unknown) for darwin 8 would also be appreciated. This email sent to site_archiver(a)lists.apple.com

1 0

Re: perl 5.8.1 issue on 10.3.9 server [ semi solved ]
by Kevin Van Vechten 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On Nov 30, 2005, at 12:32 AM, Sam Hart wrote: Dale wrote: The only other question I have is this, it seems that when building under 10.4 the binaries are being built for ppc and i386, is there a way to reduce the build to ppc only ??? - Kevin _______________________________________________ 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… RC_ARCHS=ppc Yes, the RC_ARCHS environment variable can be used while building, or the lipo(1) tool can be used to reduce a universal binary to one architecture after the fact. This email sent to site_archiver(a)lists.apple.com

1 0

Re: flushing HFS cache to disk
by Shawn Erickson 30 Nov '05

30 Nov '05

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:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=g6jyQ4xUQhIXB8TpOJN+wAFIbcZmgWQK0RsXLNkCLkowD4HShdiw3qSYYZX9WQMrtzeRk3OwIr4zC2M1PV9x4XKfRSZLqtQ5YMNPxENILNx8VrlbcTPnhSAkZm74qcOT1vpzpigHFu9zjdU/dKjXQLnMF+GyFk9AvJZPt+7Ywao= On 11/30/05, Dave Yost <ListMailForDave(a)yost.com> wrote: > At 01:32 PM -0800 2005-11-29, Mark Day wrote: > >On Nov 29, 2005, at 12:03 PM, darwin-dev-request(a)lists.apple.com wrote: > > > >>>that's what sync(2) does. > >> > >>It doesn't seem to. > > > >Beware that sync(2) is an asynchronous call. It just begins the flushing of data to all volumes. If you're reading the disk immediately, the changes may not have been written yet. > > Surely there should be a synchronous call one can use. What happens after that synchronous call returns...? Are you expecting it to lock out the file system from any other access? It other words a synchronous call serves no purpose since at any moment another thread can come along and modify the state of the file system as long as it is mounted and writable. _______________________________________________ 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: flushing HFS cache to disk
by Dave Yost 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com At 01:32 PM -0800 2005-11-29, Mark Day wrote: >On Nov 29, 2005, at 12:03 PM, darwin-dev-request(a)lists.apple.com wrote: > >>>that's what sync(2) does. >> >>It doesn't seem to. > >Beware that sync(2) is an asynchronous call. It just begins the flushing of data to all volumes. If you're reading the disk immediately, the changes may not have been written yet. Surely there should be a synchronous call one can use. 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

1 0

Re: Leaf vs. Non-leaf functions & crash reporting
by Jonas Maebe 30 Nov '05

30 Nov '05

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 30 nov 2005, at 00:23, Nathan Herring wrote: Jonas _______________________________________________ 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… However, it appears that Apple's crash reporter manages to figure it out nicely. I've been staring at it for some time, but I cannot figure out how to do it that doesn't involve some code inspection to see if lr points to some code between ssr0 and the prolog of the function ssr0 is in (and that's assuming it has a prolog). This one includes some stack crawling code by Apple (in MIB-Libraries/ MoreDebugging/MoreBacktrace.c): http://developer.apple.com/samplecode/MoreIsBetter/MoreIsBetter.html This email sent to site_archiver(a)lists.apple.com

1 0