March 2007 - Darwin-kernel - lists.apple.com

Re: large, disposable files?
by Quinn 14 Mar '07

14 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Which, last time I checked, was largely ignored by the kernel. S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Relations, Developer Technical Support, Core OS/Hardware _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… At 21:27 -0700 13/3/07, Michael Smith wrote: It's largely superseded by the (more standard) madvise(2). This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Michael Smith 14 Mar '07

14 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Mar 13, 2007, at 12:05 PM, Michael Crawford wrote: I was going to suggest looking at /usr/include/sys/vadvise.h, but it has this in it: #ifdef __APPLE_API_OBSOLETE The vadvise system call at least used to be a way to inform the virtual memory system of your planned memory access patterns. It's largely superseded by the (more standard) madvise(2). = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Michael Crawford 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=heekzpfhxUUjQGPq1dgXTvvaL+AzijsQRMvVyJG971BPJRUiTPMzsZtvxSHXUaTd+iJWJBYG9iJGFTw37lovNzEKKssPbTufbqyzf0ljV/05Y0gBMOeZyK2CcoysACSPfDQOhIl0EOX9UZfnLBK5M+OvmT8O9i2NwSXy9rzAuTg= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=EOirhyq3/gSy3y8mSukIEH6bznxH9xQDA6a8nmej5xfpZn7RnHveasOeV6ekGrYGsMmNtN20tsUlbHxpEWzWInCXs6o/PObmeHh55Nu1Wz94VWfyn2RwbW9IqrHRPRwOUEFk69sQ/vpTXLEVsxlkHUgqM/79gUcmr0Jn0CMRGn8= I was going to suggest looking at /usr/include/sys/vadvise.h, but it has this in it: #ifdef __APPLE_API_OBSOLETE The vadvise system call at least used to be a way to inform the virtual memory system of your planned memory access patterns. Mike Crawford mdcrawford at gmail dot com http://www.goingware.com/ _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Andrew Gallatin 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Andrew Gallatin writes: > Darwin, since Tru64 was also a "shotgun wedding" of Mach and BSD, > with no servers, etc. Heck, I think they even used Mach-O! Too > bad HP ended up with it rather than you guys. Faulty memory.. It was ecoff.. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Andrew Gallatin 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Michael Smith writes: > > On Mar 13, 2007, at 9:32 AM, Andrew Gallatin wrote: > >>> But before I go trying to re-compile these applications, I'm still > >>> wondering if there is any way to limit the ubc like in Tru64... > >> > >> Limit in what way? > > > > To using no more than a certain percentage of memory, like > > ubc-maxpercent does in Tru64. > FWIW, what I'm doing is nothing "serious", just watching high-def, MPEG-2 files recorded over the air by my MythTV DVR with either MythFrontend or VLC. This is on a Core2 Duo iMac with 1GB RAM. After watching a few minutes of an HD recording, nearly everything on the desktop (like mail clients, web browsers and even portions of the video player not in the critical path for playing) gets paged out, and switching between the video and any other application is miserable. I have nearly the same experience (but it takes longer to happen because the quality is lower, and the file size is smaller) watching shows purchased from iTunes. > File cache pages join the queue in a deprivileged position (at the > head of the inactive list), so they tend not to cannibalise your > working set. They do totally swamp any apps you haven't touched in a while, and create a huge paging storm though, which tends to drive the user experience through the floor. The problem is really exacerbated by the usage pattern watching HD video. You watch 5 or 10 minutes of a video, then you realize you have an email, you pause the video to answer the mail, and curse while the portion of the video player responsible for pausing gets paged in, then curse some more while the mailer gets paged back in, etc. > Because there is no logical division between buffer cache and VM > cache, there's no straightforward way to implement what you describe... Thanks for letting me stop looking, but it sure would be nice if you added one ;). There is a lot of good stuff in Tru64, especially for Darwin, since Tru64 was also a "shotgun wedding" of Mach and BSD, with no servers, etc. Heck, I think they even used Mach-O! Too bad HP ended up with it rather than you guys. At any rate, I've wanted to re-build MythFrontend for a while, maybe this is finally the excuse I need.. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Michael Smith 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com But before I go trying to re-compile these applications, I'm still wondering if there is any way to limit the ubc like in Tru64... To using no more than a certain percentage of memory, like ubc-maxpercent does in Tru64. = Mike _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On Mar 13, 2007, at 9:32 AM, Andrew Gallatin wrote: Limit in what way? File cache pages join the queue in a deprivileged position (at the head of the inactive list), so they tend not to cannibalise your working set. Because there is no logical division between buffer cache and VM cache, there's no straightforward way to implement what you describe... This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Andrew Gallatin 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Shawn Erickson writes: > On 3/13/07, Andrew Gallatin <gallatin(a)cs.duke.edu> wrote: > > > > Andrew Gallatin writes: > > > I doesn't look like Darwin/MacOSX has an O_DIRECT option to disable > > > caching on the file at open time. Does MacOSX/Darwin have a way to > > > disabling caching on a per-mountpoint basis? > > > > I got two, independant suggestions to try fcntl(fd, F_NOCACHE, 1). > > Execellent! > > > > But before I go trying to re-compile these applications, I'm still > > wondering if there is any way to limit the ubc like in Tru64... > > Limit in what way? To using no more than a certain percentage of memory, like ubc-maxpercent does in Tru64. Drew _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: large, disposable files?
by Shawn Erickson 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=hG0KUGUGZF28WurAKhxlzeNafyR0ACPaOVIeamIRgXzwhVaiw0Ic7kkalleYOaHvFJBUs4HkVU+k+G4qKKkcqi6luPFPMkiZ/LvD6aVVeQ4agsf17y9OhsaeM/vYgc6M+wAKp9WwbFhWArTOZsb6S4hbtjawyca2e0VwOc+gVjI= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=GZo5H7DmrKlgDocatsdpjc702IvvE7c4RJ1nHKID1gIv9mj70SzjWKCBwQXo4sSExyb/AeqWNpKK76kXiO7s+dxnTNEpV/Mv+BVHlVavlXA3Y2J7xJ9hhsaNfMvWqPQCSL1Q83TxCAWq9IABwO2/gjaX3uY2RsoyWTy0S1ItA7Q= Andrew Gallatin writes: > I doesn't look like Darwin/MacOSX has an O_DIRECT option to disable > caching on the file at open time. Does MacOSX/Darwin have a way to > disabling caching on a per-mountpoint basis? I got two, independant suggestions to try fcntl(fd, F_NOCACHE, 1). Execellent! But before I go trying to re-compile these applications, I'm still wondering if there is any way to limit the ubc like in Tru64... Limit in what way? -Shawn _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… On 3/13/07, Andrew Gallatin <gallatin(a)cs.duke.edu> wrote: This email sent to site_archiver(a)lists.apple.com

1 0

Re: Beginning with Kernel Extension Concepts
by Francis Derive 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Sorry, Thanks for the support. Francis. On Mar 13, 2007, at 3:07 PM, Michael Crawford wrote: Could you be compiling the code as C and not C++? extern "C" is legal in C++, but not C. You definitely want to compile as C++ for I/O Kit drivers. Mike Crawford mdcrawford at gmail dot com Bonjour à tous, //--------------------- /* add your code here */ #include <IOKit/IOLib.h> #include "HelloIOKit.h" extern "C" { #include <pexpert/pexpert.h> } // For debugging purpose only [...] //------------------- Thanks. Francis. This email sent to mdcrawford(a)gmail.com -- Michael David Crawford mdcrawford at gmail dot com Enjoy my art, photography, music and writing at http://www.geometricvisions.com/ --- Free Music Downloads --- _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… The error didn't come from the cpp file - but from the header where a ending semi-colon was missing. On 3/13/07, Francis Derive <francisderive(a)mac.com> wrote: >From Apple Dev Documentation, I get an error : "expected unqualified-id before string constant" - on line 5 in the HelloIOKit.cpp file of this "Hello I/O Kit : Creating a Device Driver With XCode" chapter. I stare at it and can't think anything brilliant about it. First, I am not sure about where the error raised from : the call to a "C " construct, or the line before ? _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/mdcrawford% 40gmail.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: Beginning with Kernel Extension Concepts
by Francis Derive 13 Mar '07

13 Mar '07

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Mar 13, 2007, at 3:07 PM, Michael Crawford wrote: Could you be compiling the code as C and not C++? extern "C" is legal in C++, but not C. You definitely want to compile as C++ for I/O Kit drivers. Mike Crawford mdcrawford at gmail dot com Bonjour à tous, //--------------------- /* add your code here */ #include <IOKit/IOLib.h> #include "HelloIOKit.h" extern "C" { #include <pexpert/pexpert.h> } // For debugging purpose only [...] //------------------- Thanks. Francis. This email sent to mdcrawford(a)gmail.com -- Michael David Crawford mdcrawford at gmail dot com Enjoy my art, photography, music and writing at http://www.geometricvisions.com/ --- Free Music Downloads --- _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/site_archiver%40lists.… Well, I am using Xcode (2.4.1), the project was initiated with the Assistant ( Kernel Extension/IOKit Driver ), generated this C++ file, and I understand the compilation process is aware of the project being C++ (where the extern "C" form is known to call for a C construct - the included pexpert.h looks like pure C ). So I understand this line is the way to call pure C from C++ - never mind. I have edited the info.plist and the HelloIOKit header file as per the tutorial. And I believe I am compiling C++ code. So ... ? On 3/13/07, Francis Derive <francisderive(a)mac.com> wrote: >From Apple Dev Documentation, I get an error : "expected unqualified-id before string constant" - on line 5 in the HelloIOKit.cpp file of this "Hello I/O Kit : Creating a Device Driver With XCode" chapter. I stare at it and can't think anything brilliant about it. First, I am not sure about where the error raised from : the call to a "C " construct, or the line before ? _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-kernel mailing list (Darwin-kernel(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-kernel/mdcrawford% 40gmail.com This email sent to site_archiver(a)lists.apple.com

1 0