December 2008 - Darwin-kernel

Re: Allocate contiguous wired memory and obtain physical address
by Terry Lambert 21 Dec '08

21 Dec '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Dec 21, 2008, at 9:43 AM, "John D." <johndenkar(a)gmail.com> wrote: Hi, I would like to know what KPIs are exported to kernel extensions for allocating contiguous wired memory (non pageable) and then obtain its physical address. Do the IOKit IOMalloc functions provide this functionality? IOMallocContiguous -- Terry _______________________________________________ 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.… You can read more about it on <http://developer.apple.com>; it's the first result you get from typing /contiguous wired memory/ into the search box there. The third result tells you why you should not be doing it. This email sent to site_archiver(a)lists.apple.com

1 0

Allocate contiguous wired memory and obtain physical address
by John D. 21 Dec '08

21 Dec '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:message-id:date:from:to :subject:mime-version:content-type:content-transfer-encoding :content-disposition; bh=X+LUp2HhUwYbCn0rI3KB1jSGvhn0ExHhaD4BY09rYLM=; b=cmPupcPk5m4ZVlxtMJqK4Gdx7Wqrx764Q1d6vwnlRh2wK+NJ3lyEXBwTLwL9KcOCmE dZdBStDWYRI6SQAIEoBaszgQtLqtwXDKgdu6o4C6F2vLCfULQPxKmL0zSfAIQpzSvOCc 608Opwu5WO9p+ty98vSJvP818wTqLtW15f60k= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:mime-version:content-type :content-transfer-encoding:content-disposition; b=mfSp3T+lL1OtCeLUc8ov4+qILxZOYjMDbt0/r8hmDCZNIp97BE9he0L4WpNip21Gy1 6QgGRHLb421EUPxNklNlFccFDi/nGCAoiuKHCRjVc1aivVSSZF16+HOKLLzx+9bOg3qx d2L4g4dKNdvLCoKqdhy/YDJ01+J9POvlNmlUo= Hi, I would like to know what KPIs are exported to kernel extensions for allocating contiguous wired memory (non pageable) and then obtain its physical address. Do the IOKit IOMalloc functions provide this functionality? Merry Christmas everyone. -- John Denkar _______________________________________________ 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: NKE programming: ipf_input filter arguments
by John D. 21 Dec '08

21 Dec '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; 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; bh=aSl1Rm1ptU8olRQouqjrZ8AE82v27oIUlwTMmt6mGXM=; b=TNb4Yy1i7QbVDdmWHqImsOAUV++xbwuE9Yqx3W7SqN8IL4mhUwz29jprytm3p3bPbH 2u4QHx7RJ5vjakjVANDWblu0GlTNSgT1tDdBNIwBeIE6hHflcnhCZqYWBNEHWLIOSeB/ OYbYvrSaEX97/3VEf9nLIX+fyxutwSt9hD5AM= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=tvJSG+Rc6ikx/jMMuKm7lSQZM5fDcEGAVHQHK89Zq0iiwaWtX1lVanbKoEkbpI3/2g VZhacnap3lrLT+1k6H3YM3zsZ3lbNKzJLTHvWp/L0EnK5nM/KS6mAoc4E7JKQLPVZn1X lncofCGm7sAB2n3wbvEVb+vdL/5aizaYVxdCs= Thanks, I fixed this by using both offset and calculating the one to packet body from th_off etc. The values must be validated in the filter since they don't seem to be checked elsewhere. John. On Thu, Dec 18, 2008 at 6:10 PM, Stéphane Sudre <ssudre(a)intego.com> wrote: > > On Dec 14, 2008, at 5:10 PM, John D. wrote: > >> Hi, >> >> I'm having trouble with my IP (v4) ipf_input_func. It registers >> properly and works fine through my initialization code, but the data >> argument is confusing, as well as the offset one. The documentation >> says data (which is a mbuf_t pointer, aka struct __mbuf **) points at >> start of data with the IP header. My code only inspects TCP packets, >> checking the protocol argument and ignoring the packet (returning zero >> immediately) if it's not IPPROTO_TCP. >> >> The problem arises when I try to read the tcphdr structure from the >> data argument. The documentation again says that I should read >> data+offset, and doing so yields unconsistent results. Source and >> destination ports are incorrect after converting them back to host >> native endianness (little endian here, thus using ntohs()) and the >> whole structure seems to be incorrect, which makes me think I'm >> reading a wrong location. I tested with UDP packets and it's the same >> situation. >> >> What information is contained in reality within the data argument? >> Does the offset argument count the IP header itself too? >> >> For example: >> >> struct udphdr *udphdr = (struct udphdr *) (data + offset); >> struct tcphdr *tcphdr = (struct tcphdr *) (data + offset); >> >> The IP filter function arguments: >> (void *cookie, mbuf_t *data, int offset, u_int8_t protocol) > > data is the pointer to the mbuf_t not the pointer to the packet data. > > You need to use the accessor: > > struct ip * tIPv4Header; > > tIPv4Header=(struct ip *) mbuf_data(*data); > > > My $0.02 _______________________________________________ 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: Kernel extensions
by Quinn 19 Dec '08

19 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com It depends on the type of kernel extension: o VFS plug-ins can be discussed on filesystem-dev <http://lists.apple.com/mailman/listinfo/filesystem-dev> o NKEs can be discussed on MacNetworkProg <http://www.lists.apple.com/macnetworkprog> o I/O Kit KEXTs have a number of different family specific lists <http://lists.apple.com/mailman/listinfo> Anything else can be discussed here. 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 8:14 -0800 19/12/08, J. Todd Slack-Moehrle wrote: I dont see a specific mailing list for kernel extension development. I did see this list though. Where can I ask some questions about developing kernel extensions? This email sent to site_archiver(a)lists.apple.com

1 0

Kernel extensions
by J. Todd Slack-Moehrle 19 Dec '08

19 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Hi All, Where can I ask some questions about developing kernel extensions? Thanks, -Jason _______________________________________________ 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 dont see a specific mailing list for kernel extension development. I did see this list though. This email sent to site_archiver(a)lists.apple.com

1 0

Re: NKE programming: ipf_input filter arguments
by Stéphane Sudre 18 Dec '08

18 Dec '08

site_archiver(a)lists.apple.com Delivered-To: Darwin-kernel(a)lists.apple.com On Dec 14, 2008, at 5:10 PM, John D. wrote: Hi, I'm having trouble with my IP (v4) ipf_input_func. It registers properly and works fine through my initialization code, but the data argument is confusing, as well as the offset one. The documentation says data (which is a mbuf_t pointer, aka struct __mbuf **) points at start of data with the IP header. My code only inspects TCP packets, checking the protocol argument and ignoring the packet (returning zero immediately) if it's not IPPROTO_TCP. The problem arises when I try to read the tcphdr structure from the data argument. The documentation again says that I should read data+offset, and doing so yields unconsistent results. Source and destination ports are incorrect after converting them back to host native endianness (little endian here, thus using ntohs()) and the whole structure seems to be incorrect, which makes me think I'm reading a wrong location. I tested with UDP packets and it's the same situation. What information is contained in reality within the data argument? Does the offset argument count the IP header itself too? For example: struct udphdr *udphdr = (struct udphdr *) (data + offset); struct tcphdr *tcphdr = (struct tcphdr *) (data + offset); The IP filter function arguments: (void *cookie, mbuf_t *data, int offset, u_int8_t protocol) data is the pointer to the mbuf_t not the pointer to the packet data. You need to use the accessor: struct ip * tIPv4Header; tIPv4Header=(struct ip *) mbuf_data(*data); My $0.02 _______________________________________________ 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: Obtaining non-exported symbol from kernel on runtime (without the debug symbols)
by Quinn 16 Dec '08

16 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com You might thing that by looking at the headers, but you'd be wrong )-: <http://developer.apple.com/qa/qa2007/qa1574.html> <http://developer.apple.com/qa/qa2007/qa1575.html> 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 21:57 +0100 16/12/08, John D. wrote: That said, I wonder if Apple has provided means to let KEXTs interact with the MAC framework, set its own hooks, etc. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Obtaining non-exported symbol from kernel on runtime (without the debug symbols)
by John D. 16 Dec '08

16 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; 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; bh=OYjwVvxn29v7wRnw8DNw/BHZOZ3js7cGKw0NOPK38vU=; b=xi6Kb+QS8ghW9FLXTxfLL+sbdJBlcvKrecYjycgSJPVR+qH6oVaiNzluNvV8cS2Gv5 1Z4f2dqQFA/tGggbF9eXit+qH22lU9LA38hf+Yvkc/2jXA+wVOujs7msAeztblVtwmlZ eZgFhAZUXTUluK6soelhBvd0brtR5zWDZPeJQ= Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:date:from:to:subject:cc:in-reply-to:mime-version :content-type:content-transfer-encoding:content-disposition :references; b=L23fQRmbBLi8Fshww6F0EPE/QcSC5eWP3zAonm2AF4uqDXKKuArJhXTmUyqzmxnpf8 shjIjIUE720BQpPnxrpgmFC1t5H/rmLssEgLWQ6vkOuqiSVkro/70XE5vQQQzrxk3tiW MP9UMPI16nWju6lljVzd/oap/LbEEvAQFvV9I= SEBSD/SEDarwin are both external projects (developed by Sparta Inc. and contributors), with inspiration from the original Flask project (now SElinux since a few years...). I wonder if the AVC and such are exact implementations of the original design. That said, I wonder if Apple has provided means to let KEXTs interact with the MAC framework, set its own hooks, etc. In addition, Seatbelt is closed source and Apple itself discourages its usage by third-parties since the interfaces are subject to changes, besides being largely undocumented. On Tue, Dec 16, 2008 at 5:49 AM, Michael Smith <drivers(a)mu.org> wrote: > > On Dec 14, 2008, at 10:16 PM, John D. wrote: > >> On other hand. I still think kauth is severely limited (crippled) as it is >> implemented right now. > > Kauth was a means to an end. It has been largely superseded in a broad > sense by the MAC framework, which in turn facilitates Seatbelt and several > other useful features. You will note that this supercession was not > achieved by removing the Kauth symbols from the KPI. > > = Mike > > -- - John Denkar _______________________________________________ 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: Obtaining non-exported symbol from kernel on runtime (without the debug symbols)
by Michael Smith 16 Dec '08

16 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Dec 14, 2008, at 10:16 PM, John D. wrote: = 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 other hand. I still think kauth is severely limited (crippled) as it is implemented right now. Kauth was a means to an end. It has been largely superseded in a broad sense by the MAC framework, which in turn facilitates Seatbelt and several other useful features. You will note that this supercession was not achieved by removing the Kauth symbols from the KPI. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Obtaining non-exported symbol from kernel on runtime (without the debug symbols)
by Andrew Gallatin 15 Dec '08

15 Dec '08

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com Dkim-signature: v=1; a=rsa-sha256; c=simple/simple; d=cs.duke.edu; s=mail; t=1229375000; bh=Gq0Fm4WsVwRXavGDGI6uL7bUAx70bOtL3W5vhk518ZE=; h=Message-ID:Date:From:MIME-Version:To:CC:Subject:References: In-Reply-To:Content-Type:Content-Transfer-Encoding; b=Es8bSeJxwZ99 S8vPEBKylCcgoYt85DIQzUOPnL2HVv2L18oEsFDKtzgHYhMESRDmZvdQfHQJulnfeau IKjh7DKCqu61dn6cMr/0Xd4YkdthL6BP7SHQ0T9m26yRi9u+arn6HzO03VoQNPEvcR5 cwmIAFl4so8cFPmdatU+VjSvo= User-agent: Thunderbird 2.0.0.16 (X11/20080724) Before these obfuscations, I wrote a device driver for an HPC interconnect which does ethernet emulation (really encapsulation) which shared the ethernet driver transmit and receive code between FreeBSD, MacOSX, and Tru64. The shims for the mbuf accessor functions were no big deal, I still use them in current products that support 10.3. My memory is a bit hazy, but I seem to remember having problems creating an ifnet and hooking it into the system in 10.4 and higher. I don't remember exactly what was the problem, but at that point we were working on a new product which wouldn't support Tru64, so the opportunities for code sharing were diminished. So I just decided to write a "normal" IOEthernet derived driver for the next generation HPC software. 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.… Amanda Walker wrote: On Dec 15, 2008, at 10:03 AM, Andrew Gallatin wrote: Interesting. I'd be curious about how extensive the changes were to use through the KPIs introduced in Darwin 8. I haven't had to do that conversion, but just looking at kpi_interface.h, it appears as though the changes would be pretty minimal (ifnet_t and its accessors and mutators strike me as more ecapsulation than obfuscation). This email sent to site_archiver(a)lists.apple.com

1 0