August 2007 - Darwin-kernel

Re: mbuf_pullup practically useless
by Platon Fomichev 30 Aug '07

30 Aug '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Howdy-ho Best regards, Stauff__ On Aug 30, 2007, at 10:33 PM, Stéphane Sudre wrote: On Aug 30, 2007, at 8:25 PM, Platon Fomichev wrote: [...] This can be done at the socket level AFAIK. _______________________________________________ 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.… Thanks for cheering me up but can you point me to the exact func - most of them say in rather 'capital letters' that I should not block while executing them. And only send/receive funcs allow swallowing.. Btw not everything is easy as it seems... For example one of my tasks while developing an NKE (at socket level) was to delay connect() call until user land processes signals that it's ok - rather trivial task as I think - but nope connect() can't be blocked as it is in data path - I can't swallow connect() call and it seems that there is no way I can solve this on Socket NKE level... _______________________________________________ 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/pfomichev% 40alarity.com This email sent to pfomichev(a)alarity.com This email sent to site_archiver(a)lists.apple.com

1 0

Re: mbuf_pullup practically useless
by Stéphane Sudre 30 Aug '07

30 Aug '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Aug 30, 2007, at 8:25 PM, Platon Fomichev wrote: [...] This can be done at the socket level AFAIK. _______________________________________________ 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.… Btw not everything is easy as it seems... For example one of my tasks while developing an NKE (at socket level) was to delay connect () call until user land processes signals that it's ok - rather trivial task as I think - but nope connect() can't be blocked as it is in data path - I can't swallow connect() call and it seems that there is no way I can solve this on Socket NKE level... This email sent to site_archiver(a)lists.apple.com

1 0

Re: mbuf_pullup practically useless
by Andrew Gallatin 30 Aug '07

30 Aug '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Platon Fomichev writes: > possibility to develop it out of kernel as NKE... (f.ex. is there any > possibility to create /dev/pf from NKE as a start?) > It is trivial to create a device node from a normal kernel extension, and from an IOKit driver, so it should be easy to do from an NKE as well. There is an example of creating a device node at http://www.osxbook.com/book/bonus/chapter8/kma/ 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: mbuf_pullup practically useless
by Platon Fomichev 30 Aug '07

30 Aug '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Howdy-ho Chris Best regards, Stauff__ On Aug 30, 2007, at 8:23 PM, Christopher D. Lewis wrote: On Aug 27, 2007, at 2:25 AM, Platon Fomichev wrote: _______________________________________________ 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.… I will try looking at prospectives to port PF as NKE as I've recently developed a bunch of them. If anyone wishes to join me in my efforts on PF port let's unite our forces, although I highly doubt possibility to develop it out of kernel as NKE... (f.ex. is there any possibility to create /dev/pf from NKE as a start?) Btw not everything is easy as it seems... For example one of my tasks while developing an NKE (at socket level) was to delay connect() call until user land processes signals that it's ok - rather trivial task as I think - but nope connect() can't be blocked as it is in data path - I can't swallow connect() call and it seems that there is no way I can solve this on Socket NKE level... So if anyone have some advice about connect() call or is also anxious to devote some time for PF let's communicate. I will surely do this. Btw as a side note questions - are there any plans to include OpenBSD Packet Filter (quite a good piece of software imho) into Mac OS? I am thinking of doing some initial investigations on possibility of this project if no-one is already on this. Not too long after Mr. Hubbard came to Apple, he responded to a question like this (sent by me) by explaining that spending scarce man-hours developing pf was a waste of time because FreeBSD already had a firewall. I understood from his reply that he hadn't actually paid any attention to what I explained were the then- existing advantages of pf (incl. authpf, etc.). Integrated NAT didn't yet exist, so the admin advantages have increased. Questions about features that make administration easier have occasionally been replied-to on the list with retorts about how grandma doesn't want to administer her machine, she wants to use it, which of course flies in the face of Grandma's purpose, which is to have the machine do what she wants so readily that she spends her time using it rather than either administering it or regarding it as broken. Administering is simply the effort needed to prepare the machine to behave as you want as you use it, and it should be a goal to make this time as short and as painless as possible, not pretend every machine will magically take care of needs as configured from the mfg. This email sent to site_archiver(a)lists.apple.com

1 0

Re: mbuf_pullup practically useless
by mm w 30 Aug '07

30 Aug '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=qPJC56EfqRnQoMDe6/Bn027iYSj5LwW56cmOZ6wA2y/yh6vGz9yys89m1TAIrlLQXcdFpmcrzmL1S78lSen2CZPg671lL1pYK7qfQmClDlVt1ils2wba13Tbb1IrqvAB4ZoPTHLLy0UlCWEOlw5pMg8iT6beIzQnYuWDJcy0Etg= 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=KQDAUthvyKw/vYxDowg+qrnH6JlXyt7+MbwqkoBhFm7JktlG77L1s72XLvOOLaQ7GKzSO4hiUC3KZjdmtfYTwVoKEEq+eCS3IYDgsu1QIYebuvIsvtRoXxtdTeokW6M9Evzs2uS9R9GTC3VRex34MI9rnlmzijcE8NuHEUfqqbc= hey kids don't cross your emails you polluate this thread, "respect people" is "respect yourself" Cheers! On 8/30/07, Jeremy Pereira <adc(a)jeremyp.net> wrote: > > On 27 Aug 2007, at 05:40, Vincent Lubet wrote: > > > Platon, > > > > We -- Apple engineers lurking on the darwin mailing list -- really > > do appreciate suggestions like yours. > > > > The best way to make sure your enhancement request does fall > > through the crack is to file a problem report at http:// > > bugreporter.apple.com/ > > Oops! > > > _______________________________________________ > 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/openspecies%40gmail.com > > This email sent to openspecies(a)gmail.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: mbuf_pullup practically useless
by Jeremy Pereira 30 Aug '07

30 Aug '07

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On 27 Aug 2007, at 05:40, Vincent Lubet wrote: Platon, Oops! _______________________________________________ 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.… We -- Apple engineers lurking on the darwin mailing list -- really do appreciate suggestions like yours. The best way to make sure your enhancement request does fall through the crack is to file a problem report at http:// bugreporter.apple.com/ This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cross-compile PPC kernel on Intel?
by Michael Crawford 29 Aug '07

29 Aug '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=ZdcBJSdR08V9FjThG3BiyDN6ED73MqLPHqFTG5ol3TaqeNaP4BGWfxEGqbkpoj1DzkIRjkC4V7HuEIwg4HdxjSgSwuMn+VTmRD0dG9sUE9Qngs9RpgZNbzz3qEngqBXiC51NRfBowXbMv2zQMCWfyQmyossNV9SiedJ62uf8bQw= 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=dtbFg1h3JoXNaChw9n59DjdA/yK1r0O/nxNLbpJSy5cSBVJ5wzIcYJnJdx0Gj1Ig2jGa039+7g9fv14HeZmjXLAriXKZNqH4Lp205+11f0+kzV1c7yUXEw3bJhs2sM70a0H/qirvnE6orlR7ZBXN8q0jkA74AgU8UZqir0gfhWY= On 8/29/07, Michael Crawford <mdcrawford(a)gmail.com> wrote: > I have thought of a workaround: keep the source code on my Intel box, > but do the compile on my PowerPC, using a symbolic link so that the > pathnames to the sources are the same on each machine. > > I'll let you know how it works. It works! Now to find the bug in my driver... -- 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.… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Cross-compile PPC kernel on Intel?
by Michael Crawford 29 Aug '07

29 Aug '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=RC9IjZY6CI4MIugtJrJt3cuCXItVpMYGN+MHIZQk2ueibiwz0bJqH3t5sF5qsxhEoFDGV1xCoe3bsQhxF8OCVbcK+BkrQlGccTjtzHuk4FTLDTMFZfI1m9d3Uik93XZTHasxRRx0CFeFIfHXSOPfjTcdUIdr3nvlFUec/TRIGGM= 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=jwTEbj+9MjVHNtbxDpca9LGIksBXUc/LRpzfz4cYTe079UqPpKE60BFXMiSggkrZHwt0Dnp5WC/WhdlQnNrBQjtZ8KW7mTJ6Wh1ToImA/w3Oo4SouP+j8OgSNNm255SZXH/tVar3YkhkSS9xzbPXJxWsmM14nj4ns4UY0cm1BX0= I have thought of a workaround: keep the source code on my Intel box, but do the compile on my PowerPC, using a symbolic link so that the pathnames to the sources are the same on each machine. I'll let you know how it works. 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: Cross-compile PPC kernel on Intel?
by Michael Crawford 29 Aug '07

29 Aug '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=pW7ogMT/3tCUCodHr7oHb5CKffqnCBvt9eF41KIQAbp8Ucjh8qFvTH5qFQ/8gLAUHvhD5c2X6HlWpz2Njj35OXSC926umyicsSG4o8FWYLDi6mE7I2IYjaJ+ZhwMVkJv6cFdwLnOPedZThLiN5G/u4fSkSaOHWak31phV9KfQgU= 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=Yu2pvgm6VJ8OS99Q9pgruWscVtkBAvon71Prdwrm8YYddBEIfJTNrZOSkt6P0fy8IzPBBa2nnelULnosrRu8w/XOKrvUNqWqPhTlRotQNfVPBcSBb99h0BAIgCzm9sEx8aupzz2ZbmugSCcRCbd9LVDrqmT9mtx/rRwdQ/ou1rc= I'm trying the 10.4.9 sources now, and get the same error but on line 79 of IOWorkLoop.cpp instead of line 81. On 8/29/07, Boris Remizov <catflight(a)gmail.com> wrote: > > /Users/mike/Kernel/Build/xnu-792.12.6/iokit/Kernel/IOWorkLoop.cpp: In > > member function 'virtual bool IOWorkLoop::init()': > > /Users/mike/Kernel/Build/xnu-792.12.6/iokit/Kernel/IOWorkLoop.cpp:81: > > error: converting from `IOReturn (IOWorkLoop::*)(void*, void*, void*, > > void*)' to `IOReturn (*)(OSObject*, void*, void*, void*, void*)' in a > > kext. Use OSMemberFunctionCast() instead. > > Please take a source code for this error. bool IOWorkLoop::init() { // The super init and gateLock allocation MUST be done first if ( !super::init() ) return false; if ( !(gateLock = IORecursiveLockAlloc()) ) return false; if ( !(workToDoLock = IOSimpleLockAlloc()) ) return false; controlG = IOCommandGate:: // *** The line below ---V **** commandGate(this, (IOCommandGate::Action) &IOWorkLoop::_maintRequest); if ( !controlG ) Some source of confusion seems to be that trying to build cctools after having run xnu/SETUP/setup.sh screws things up. When I realized that I started a new terminal session. Thanks -- Mike -- 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.… This email sent to site_archiver(a)lists.apple.com

1 0

Cross-compile PPC kernel on Intel?
by Michael Crawford 29 Aug '07

29 Aug '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:mime-version:content-type:content-transfer-encoding:content-disposition; b=J+gPtKwlk36kwHvYD8Dd459zS5zN3TGrO555zu8TxzZvjCHZRBoeShwKDplyNGFcP3LPM6NES5c++c9/QkzDLznbgdMuwK0KeEX5DlNV08pF2ftR3lEYeAOsqQDD/fjn3X0Mzhy6hL4aobV13RRCr+zWGtkE94odzh5aJ/epTW4= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:to:subject:mime-version:content-type:content-transfer-encoding:content-disposition; b=NdjbcWsy70COa01pj2L0ikSdNVJvAUHGHgmk031LCZQuH5+j05kUcVb6Q0ffMK8tn1g8iU+YAb7XTrNyF3sc8WuuVvhm4pInPIAipyDhINT+UzDttPoGQog9eBeXBL982zNmeOIartsNaBb3DwYU0iH3r6AAW/A11TK156yQhZA= I'm trying to build the 10.4.8 PowerPC kernel and some of its drivers on an Intel Mac. I've used gcc_select to set the gcc version to 3.3. The kernel build command line I use is: make ARCH_CONFIGS=PPC KERNEL_CONFIGS=DEBUG all the make runs for quite a while then gives the following error: /usr/bin/cc -o makedis /Users/mike/Kernel/Build/xnu-792.12.6/osfmk/ddb/makedis.c cc: installation problem, cannot exec `cc1': No such file or directory make[6]: *** [makedis] Error 1 note that cc wasn't given the "-arch ppc" flag; it's trying to compile to Intel code, but gcc 3.3 doesn't support Intel - it was trying to exec a 3.3 cc1 for Intel and there isn't one. So I try using gcc 4.01 (against the advice of the kernel build instructions) and I get: /Users/mike/Kernel/Build/xnu-792.12.6/iokit/Kernel/IOWorkLoop.cpp: In member function 'virtual bool IOWorkLoop::init()': /Users/mike/Kernel/Build/xnu-792.12.6/iokit/Kernel/IOWorkLoop.cpp:81: error: converting from `IOReturn (IOWorkLoop::*)(void*, void*, void*, void*)' to `IOReturn (*)(OSObject*, void*, void*, void*, void*)' in a kext. Use OSMemberFunctionCast() instead. Any tips for me? The reason I'm trying to do this is that I've got a bug in a kernel extension on a PowerPC machine, and I only have one PowerPC machine. I want to build the kernel and the relevant drivers on an Intel Mac that I'll be using as a debug host. Thanks for your help, Mike -- 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.… This email sent to site_archiver(a)lists.apple.com

1 0