November 2011 - Darwin-kernel

Re: how do I enable MACH_KDB on 10.8.0 ?
by Shantonu Sen 16 Nov '11

16 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com The kdb inline serial debugger is not supported, and was never ported to the 64 bit kernel. You should use KDP over Ethernet or FireWire. Sent from my iPhone On Nov 16, 2011, at 2:30 PM, Andrew Gallatin <gallatin(a)cs.duke.edu> wrote: > I'm trying to build 10.8.0 with kdb. > I edited osfmk/conf/MASTER and added a line: > options MACH_KDB > > Then I rebuilt, but now the link is failing: > > CC version.o > LD mach_kernel.sys > Undefined symbols: > "_getsymtab", referenced from: > _aout_db_sym_init in db_aout.o > ld: symbol(s) not found > > > > This is because libkern/kernel_mach_header.c has that > compiled out when MACH_KDB is not defined. For now, > I just hacked this into ./libkern/libkern/kernel_mach_header.h > #ifndef MACH_KDB > #define MACH_KDB 1 > #endif > > But is there a more proper way? > > > BTW, I haven't crashed yet, so I don't know if kdb actually still > works ;) > > Thanks, > > 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/ssen%40apple.com > > This email sent to ssen(a)apple.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

how do I enable MACH_KDB on 10.8.0 ?
by Andrew Gallatin 16 Nov '11

16 Nov '11

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=1321482654; bh=YBotPSJDftuLNIoPkRYuhhlxzJmUomwT7Rbx2/j6H7I=; h=Message-ID:Date:From:MIME-Version:To:Subject:Content-Type: Content-Transfer-Encoding; b=chf45mMw96FODNC4RD92vCmx3EgZvkJqn9ZXjtt2VzZPeE8zEpv9fBFCyFEpC5oZl oOrh94Ny+w7OcC/JcYhSoaBm41cP4TpUVuOs9P4Jq8tcUOPyx35tXsP4aQe4HZ12fn plfeJ0RYdgx8nqUD2lznmJCJ/OMUWycSNSvf2p65GKhpg+NKj1mLMd9wuKpyEEluzu j2sEV4XQ3PSOmxtA1yvXfAXFt0s6nKirOQbWnffApYCZ6Y+1m3+UJxF1H+L0Nx+gc8 6bAujjbIHdTtCuiVTYX2gP5E2pISzQLJAs5Ruh55hqHiCYBl5D1yKZdP+PdbY8dSu/ Z4qhu1q3FZhlA== User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:7.0.1) Gecko/20110929 Thunderbird/7.0.1 I'm trying to build 10.8.0 with kdb. I edited osfmk/conf/MASTER and added a line: options MACH_KDB Then I rebuilt, but now the link is failing: CC version.o LD mach_kernel.sys Undefined symbols: "_getsymtab", referenced from: _aout_db_sym_init in db_aout.o ld: symbol(s) not found This is because libkern/kernel_mach_header.c has that compiled out when MACH_KDB is not defined. For now, I just hacked this into ./libkern/libkern/kernel_mach_header.h #ifndef MACH_KDB #define MACH_KDB 1 #endif But is there a more proper way? BTW, I haven't crashed yet, so I don't know if kdb actually still works ;) Thanks, 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: IOMalloc / IOFree while holding a mutex
by Shantonu Sen 16 Nov '11

16 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com There's some layers of sublety here. You can do a lot of things at a technical level. Certainly there are no assertions or consistency checks that prevent you from holding an IOLock/lck_mtx_t. However if you are writing a storage driver that is on the paging path, consider what happens if you take the lock (or run under a command gate), attempt the IOMalloc, which in turn needs to free up pages by paging out user data, which issues an I/O, which tries to acquire the mutex previously acquired? You can try to massage the code to accept an asyncReadWrite() without taking any locks, so you don't end up deadlocking, but it requires some thought. This is one of the reasons why memory allocations on the I/O path is frowned upon. Shantonu Sen ssen(a)apple.com On Nov 16, 2011, at 6:17 AM, chas williams - CONTRACTOR wrote: > On Wed, 16 Nov 2011 11:09:07 +0530 (IST) > sheetal phirke <sheetal_phirke06(a)yahoo.co.in> wrote: > >> Can I call IOMalloc / IOFree while holding mutex lock? > > yes. > _______________________________________________ > 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/ssen%40apple.com > > This email sent to ssen(a)apple.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: IOMalloc / IOFree while holding a mutex
by chas williams - CONTRACTOR 16 Nov '11

16 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Wed, 16 Nov 2011 07:10:13 -0800 Michael Crawford <mdcrawford(a)gmail.com> wrote: > It will also slow everything else down in the system, even if not > CPU-bound, because all the tasks that depend on spinlock acquisition > are designed with the assumption that it never takes very long to > acquire one. i believe this is why apple strongly suggests that you do not use spin locks. the manual says "should be used only in places where a mutex cannot be used". the only reason to use a spinlock would be taking interrupts directly which isnt very common (or suggested). even "interrupt handlers" in darwin run in a sleep-safe context. _______________________________________________ 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: IOMalloc / IOFree while holding a mutex
by Ken Hornstein 16 Nov '11

16 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com >It will also slow everything else down in the system, even if not >CPU-bound, because all the tasks that depend on spinlock acquisition >are designed with the assumption that it never takes very long to >acquire one. That's not actually a problem in practice, because if you block (in IOMalloc, or try to acquire a mutex that someone else is holding, or some other reason) while holding a spinlock you'll panic; problem solved! :-) --Ken _______________________________________________ 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: IOMalloc / IOFree while holding a mutex
by Michael Crawford 16 Nov '11

16 Nov '11

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=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type:content-transfer-encoding; bh=a+gPu/fttUButxJAxRhKmvo3INVLZgNRioJKHTF6JI4=; b=BrD1qyNIwyAE0fbDJoCZAhry/Ws1H3sok/PrOy3eNk4D0R2tvizqSq5dDu9nVa6E6m JVfURvetLXfO1ZaMmlOHlIlmkJzcQJ2JDT5cbC0Skaow8HmWUrsMc+Cyvz8LGHQFj+6j 7k4Oe1ykwkWee4ExczBEz3If5FEpG+yqZhFAQ= > On Wed, 16 Nov 2011 11:09:07 +0530 (IST) > sheetal phirke <sheetal_phirke06(a)yahoo.co.in> wrote: > >> Can I call IOMalloc / IOFree while holding mutex lock? > > yes. I expect that the reason you can do that is that attempting to acquire a mutex lock can block. Spinlocks are much faster than mutexes because they never block. You never have to worry about context switch time. But if you are blocked when some other thread tries to acquire the same spinlock that you are holding, it might hold that spinlock for a very long time, until your call to IOMalloc or IOFree returns. Because spinlocks just waste CPU cycles until the spinlock object is acquired, doing this will waste CPU resources in a dramatic way. It will also slow everything else down in the system, even if not CPU-bound, because all the tasks that depend on spinlock acquisition are designed with the assumption that it never takes very long to acquire one. -- Michael David Crawford mdcrawford at gmail dot com Custom Software Development for the iPhone and Mac OS X http://www.dulcineatech.com/custom-software-development/ _______________________________________________ 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: IOMalloc / IOFree while holding a mutex
by chas williams - CONTRACTOR 16 Nov '11

16 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com On Wed, 16 Nov 2011 11:09:07 +0530 (IST) sheetal phirke <sheetal_phirke06(a)yahoo.co.in> wrote: > Can I call IOMalloc / IOFree while holding mutex lock? yes. _______________________________________________ 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

IOMalloc / IOFree while holding a mutex
by sheetal phirke 16 Nov '11

16 Nov '11

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=yahoo.co.in; s=s1024; t=1321421948; bh=qD5HA5WyvLEY++G9N6gWGTFIUCkI0PSGfKoZS3IIP2g=; h=X-YMail-OSG:Received:X-Mailer:Message-ID:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type; b=lA09tVZxzyxm6gS9JccFO8RCMgwiWys6VT44ikyaw0g/t+9/k9mpV4dcYURiTklOi5DaC59UIrdLupKADbU7Bgy8hfUlFZL5ERPMWkOeh5ExNI3nosuyo1i4koAfw3HNcr/wucK+Gi6zV5cAszf4M33lcwNKh1Ln48xo+7Jqq/0= Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.co.in; h=X-YMail-OSG:Received:X-Mailer:Message-ID:Date:From:Reply-To:Subject:To:MIME-Version:Content-Type; b=oZWv1pOE4QKZ3dOz/HCAVE8rJK8UOmrO3NMxwod+gqTGBFZg6cVp8FaBtD3SV4/cWXPPOk19JBEPyVIT+SBEXD9giBhsoDtCefUSAPOdnzQxZmEdqUstsQXmppMDIMzuRyTdYLPVqENazz5fmuHAKR91DNFAN2cQdy0kf0fNGkU=; _______________________________________________ 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

Basic Kernel Extension Programming questions
by mailist＠ericgorr.net 09 Nov '11

09 Nov '11

site_archiver(a)lists.apple.com Delivered-To: darwin-kernel(a)lists.apple.com User-agent: Roundcube Webmail/0.5.4 IOMemoryDescriptor* descriptor; if ( descriptor ) { IOLog("MyBufferMethod prepare kIODirectionOutIn \n" ); IOReturn ioErr = descriptor->prepare( kIODirectionOut ); if ( ioErr == kIOReturnSuccess ) { IOLog("MyBufferMethod reading\n" ); char buffer[ 200 ]; IOByteCount count = descriptor->readBytes( 0, buffer, 10 ); if ( count > 0 ) IOLog("MyBufferMethod done...count > 0\n" ); else IOLog("MyBufferMethod done...count <= 0" ); } } Thank you. _______________________________________________ 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.… (1) In Apple's Kernel Extension Programming Topics book there is a tutorial titled "Creating a Device Driver with Xcode." I went through this tutorial and was able to use kextload to load the device driver and can see it in the list after executing kextstat. However, I was unable to get any code to actually execute. I was then able to locate the SimpleUserClient sample code and have been able to play around with it successfully. The primary difference between the tutorial and the sample project is that in the sample project there is an extra layer in the kext - the SimpleUserClient.cpp/h which appear to provide an interface between the actual driver code and the SimpleUserClientTool. So, what I am wondering is if the tutorial is missing this extra layer which would have allowed some of it code to be executed or if I should have been able to see output from any of the IOLog calls...? (2) I was interested in learning how to use an IOMemoryDescriptor and wrote the follow code: descriptor = IOMemoryDescriptor::withPhysicalAddress( 5000, 5010, kIODirectionOut ); It, of course, fails to read any bytes...I see the message "MyBufferMethod done...count <= 0". My questions is that if I simply wanted to read the bytes contained at an arbitrary location within physical memory, how can I do that? It seems like this approach should work, so I believe I am missing something obvious. Or, do I need to look into using something like /dev/mem to access the bytes stored within physical memory? (3) One can create an IOMemoryDescriptor with withAddressRange or withPhysicalAddress. If one passes kernel_task for the task_t argument, does this effectively do the same thing as creating the IOMemoryDescriptor using withPhysicalAddress? This email sent to site_archiver(a)lists.apple.com

1 0

Re: Safe Assumptions(?)
by José Cornado 09 Nov '11

09 Nov '11

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=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=iRgZzIUOBlkumCMsG1niLru/NIKXveXL3DC//Swhn5s=; b=q1S/uVLZHgz+heqzeFA8TwIX2dcw623I8im2Z3o8oqlHbxPgLuVtza4VcVnwQb4s7u 1GDW8yckUW3/IjnKS+8FpCphqXL5KhmMrdpxxVLyXMNCrqrjFl+yiQgBHtiULCYH7mfj PH3ZVJIceMbdIFHLZsqyJGtEiHT6OULSbXRvo= The information contained in this electronic mail transmission may be privileged and confidential, and therefore, protected from disclosure. If you have received this communication in error, please notify us immediately by replying to this message and deleting it from your computer without copying or disclosing it. _______________________________________________ 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