10.6.2 - bye bye Atom support
Ric
For anyone hackintoshing. As a side note, how/why would you remove
support for a specific type of X86 CPU? Isn't it just another CPU as
far as the OS is concerned?
James Dore
It's a CPU lacking in any 64-bit usefulness, or other extensions like VT.
I guess Apple will make more extensive use of the 64 bit stuff in Snow
Leopard.
--
James Dore
New College IT Officer
james.dore@new / it-support@new
Rowland McDonnell
> ORic <infob...@gmail.com> wrote:
>
> >
> >http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prow
> >l/#comments
> >
> > For anyone hackintoshing. As a side note, how/why would you remove
> > support for a specific type of X86 CPU? Isn't it just another CPU as
> > far as the OS is concerned?
>
> It's a CPU lacking in any 64-bit usefulness, or other extensions like VT.
> I guess Apple will make more extensive use of the 64 bit stuff in Snow
> Leopard.
`will make more extensive use' of the 64 bit stuff?
Hmm...
AFAICT, with 10.6, Apple's trying to get the entire OS turned 64 bit
ASAP, top to bottom - i.e., it dropped 32 bit code like Jupiter's
gravity was pulling, man.
It's not /will make/ so much as /is making/, if you ask me.
The transition to 64 bit started with 10.3, didn't it? 10.4 provided
quite a lot of 64 bit support. 10.6 is the OS version where 64 bit is
taking over - i.e., Apple's code to run Macs is as fully 64 bit as was
practical to get written and implemented. Hell, we've got a 64 bit
kernel with 10.6, but can't use it because most third party drivers are
still 32 bit...
Rowland.
--
Remove the animal for email address: rowland....@dog.physics.org
Sorry - the spam got to me
http://www.mag-uk.org http://www.bmf.co.uk
UK biker? Join MAG and the BMF and stop the Eurocrats banning biking
Chris
> On Tue, 10 Nov 2009 11:48:27 -0000, Ric <infob...@gmail.com>
> wrote:
>
>>
http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prowl/#comments
>>
>> For anyone hackintoshing. As a side note, how/why would you
>> remove
>> support for a specific type of X86 CPU? Isn't it just another CPU
>> as far as the OS is concerned?
>
> It's a CPU lacking in any 64-bit usefulness, or other extensions
> like VT. I guess Apple will make more extensive use of the 64 bit
> stuff in Snow Leopard.
I was about to disagree with you there, but it seems that intel has
quietly removed 64bit from the Atoms. The original 230/330 have the
64bit extensions, but not the newer ones.
It seems like they are crippling them intentionally, probably to
avoid hurting their cheaper Core-based mobile cpus.
--
The email address is a spam trap. I rarely use it.
Jaimie Vandenbergh
wrote:
>http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prowl/#comments
>
>For anyone hackintoshing.
You can keep the old 10.6.1 /mach_kernel and use that with 10.6.2, to
keep your Atom-based hack going for a bit.
Make a copy of it to /mach_kernel_10.6.1, install the 10.6.2 update,
and before rebooting replace the new /mach_kernel with your old
/mach_kernel_10.6.1. Job's a good'un, apparently.
>As a side note, how/why would you remove
>support for a specific type of X86 CPU? Isn't it just another CPU as
>far as the OS is concerned?
The kernel presumably has a blacklist (or maybe a whitelist) of which
CPUs it supports, plus there's a chance of wedging if the kernel
assumes that the CPU will be able to do something when it can't.
Cheers - Jaimie
--
'Rings! Rings! Wherever they may be
I am the Lord of the Rings,' said he
'And I'll find them all, wherever they may be
And I'll bind them all in the dark,' said he -- Kevin Ahearn
Ric
wrote:
> On Tue, 10 Nov 2009 03:48:27 -0800 (PST), Ric <infobub...@gmail.com>
> wrote:
>
> >http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-...
>
> >For anyone hackintoshing.
>
> You can keep the old 10.6.1 /mach_kernel and use that with 10.6.2, to
> keep your Atom-based hack going for a bit.
>
> Make a copy of it to /mach_kernel_10.6.1, install the 10.6.2 update,
> and before rebooting replace the new /mach_kernel with your old
> /mach_kernel_10.6.1. Job's a good'un, apparently.
>
> >As a side note, how/why would you remove
> >support for a specific type of X86 CPU? Isn't it just another CPU as
> >far as the OS is concerned?
>
> The kernel presumably has a blacklist (or maybe a whitelist) of which
> CPUs it supports, plus there's a chance of wedging if the kernel
> assumes that the CPU will be able to do something when it can't.
>
> Cheers - Jaimie
> --
> 'Rings! Rings! Wherever they may be
> I am the Lord of the Rings,' said he
> 'And I'll find them all, wherever they may be
> And I'll bind them all in the dark,' said he -- Kevin Ahearn
Hmm. there'd better be a good reason for this <mutters darkly> Apple.
David Empson
I doubt that 64-bit support has anything to do with the reason that the
Atom is not working any more.
Snow Leopard works on Macs with Intel Core processors, and they are
32-bit only.
I don't know enough about each of the processor families to come up with
a plausible theory, other than perhaps a compatibility issue with the
chipsets used with the Atom.
--
David Empson
dem...@actrix.gen.nz
Jaimie Vandenbergh
Empson) wrote:
>Chris <ithi...@gmail.com> wrote:
>
>> I was about to disagree with you there, but it seems that intel has
>> quietly removed 64bit from the Atoms. The original 230/330 have the
>> 64bit extensions, but not the newer ones.
>>
>> It seems like they are crippling them intentionally, probably to
>> avoid hurting their cheaper Core-based mobile cpus.
>
>I doubt that 64-bit support has anything to do with the reason that the
>Atom is not working any more.
>
>Snow Leopard works on Macs with Intel Core processors, and they are
>32-bit only.
>
>I don't know enough about each of the processor families to come up with
>a plausible theory, other than perhaps a compatibility issue with the
>chipsets used with the Atom.
Given how many Atom chips and chipsets there are these days, maybe
they're just excluding the whole lot to save wasting effort on things
that aren't in Apple kit. It's the sort of thing I'd do if I were
managing an OS project.
Which, btw, does rather suggest that if the Apple tablet is going to
happen it'll have at least a C2D chip in.
Cheers - Jaimie
--
panic("Foooooooood fight!");
-- /usr/src/linux/drivers/scsi/aha1542.c
Ric
wrote:
> On Wed, 11 Nov 2009 10:27:49 +1300, demp...@actrix.gen.nz (David
>
>
>
>
>
> Empson) wrote:
>
> - Show quoted text -
What I don't understand (and would quite like to!) is how an Atom
differs from, say, an early Macbook 32 bit Intel CPU. I'd imagine
they support the same operations? What's the difference in
functionality between the two? Anyone able to explain this to me, I'm
a bit out of touch when it comes to CPU design...!
ray
>
> Which, btw, does rather suggest that if the Apple tablet is going to
> happen it'll have at least a C2D chip in.
Unless it has the iPhone version of OSX
--
http://www.dream-weaver.com/email.php
Web development promotion and seo
http://www.spaldingcomputers.co.uk
http://www.overseasmovingsolutions.com/
Jaimie Vandenbergh
>Jaimie Vandenbergh <jai...@sometimes.sessile.org> wrote:
>
>>
>> Which, btw, does rather suggest that if the Apple tablet is going to
>> happen it'll have at least a C2D chip in.
>
>Unless it has the iPhone version of OSX
Good point, well made.
Cheers - Jaimie
--
If you can't beat your computer at chess, try kickboxing
chris
> What I don't understand (and would quite like to!) is how an Atom
> differs from, say, an early Macbook 32 bit Intel CPU. I'd imagine
> they support the same operations? What's the difference in
> functionality between the two? Anyone able to explain this to me, I'm
> a bit out of touch when it comes to CPU design...!
My understanding is that the Atom architecture has it's basis in the old
P3, but with the advantage of modern 45nm manufacture.
All the Macbook Intel chips are based on the Core architecture, which
was the 'new' design following the disaster of the Pentium 4. I say
'new', because a lot of it came from the successful Pentium M cpus.
The Atoms have some extensions (e.g. SSE3 and 64bit) to the P3 design to
allow them to compete with current cpus, but they don't have the massive
level 2 and/or 3 caches which are common place now. This, in addition to
the die shrink, makes the Atoms physically tiny and have low energy
requirements.
This is all from memory so some or all of the details may be incorrect ;)
Jim
> http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prowl
> /#comments
>
> For anyone hackintoshing. As a side note, how/why would you remove
> support for a specific type of X86 CPU? Isn't it just another CPU as
> far as the OS is concerned?
It's code they're not using and have no need for. Removing it makes the
kernal smaller, more stable and easier to maintain.
No conspiacies here. It's just good code practice.
Ji
--
"Microsoft admitted its Vista operating system was a 'less good
product' in what IT experts have described as the most ambitious
understatement since the captain of the Titanic reported some
slightly damp tablecloths." http://www.thedailymash.co.uk/
Jim
> Ric <infob...@gmail.com> wrote:
>
> > http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prowl
> > /#comments
> >
> > For anyone hackintoshing. As a side note, how/why would you remove
> > support for a specific type of X86 CPU? Isn't it just another CPU as
> > far as the OS is concerned?
>
> It's code they're not using and have no need for. Removing it makes the
> kernal smaller, more stable and easier to maintain.
>
> No conspiacies here. It's just good code practice.
[sigh] 'conspiracies'
Jim
Ben Shimmin
> Jim <j...@magrathea.plus.com> wrote:
>> Ric <infob...@gmail.com> wrote:
>> > For anyone hackintoshing. As a side note, how/why would you remove
>> > support for a specific type of X86 CPU? Isn't it just another CPU as
>> > far as the OS is concerned?
>>
>> It's code they're not using and have no need for. Removing it makes the
>> kernal smaller, more stable and easier to maintain.
>>
>> No conspiacies here. It's just good code practice.
>
> [sigh] 'conspiracies'
And `kernel' (or `colonel', but not here). :)
b.
--
<b...@bas.me.uk> <URL:http://bas.me.uk/>
`Property, marriage, the law; as the bed to the river, so rule
and convention to the instinct; and woe to him who tampers with
the banks while the flood is flowing.' -- Samuel Butler, _Erewhon_
Jim
> >> > For anyone hackintoshing. As a side note, how/why would you remove
> >> > support for a specific type of X86 CPU? Isn't it just another CPU as
> >> > far as the OS is concerned?
> >>
> >> It's code they're not using and have no need for. Removing it makes the
> >> kernal smaller, more stable and easier to maintain.
> >>
> >> No conspiacies here. It's just good code practice.
> >
> > [sigh] 'conspiracies'
>
> And `kernel' (or `colonel', but not here). :)
Ah, I've never been clear on that one, so generally I'm happy if I'm at
least phonetically close.
Ian McCall
> Ben Shimmin <b...@llamaselector.com> wrote:
>
>> And `kernel' (or `colonel', but not here). :)
>
> Ah, I've never been clear on that one, so generally I'm happy if I'm at
> least phonetically close.
kernel = The Computing World at Large(tm).
kernal = Commodore machines - they spelt it that way.
To address the point of the thread though - I don't believe this
'optimisation' argument, and I do believe it has been deliberately
removed. You add more code checking for an Atom, not less. Unless it's
using some specific SSE-or-whatever extension that the Atom doesn't
have, then I don't believe this is by accident.
Another reason I doubt this is that we're talking about 10.6.2, not
10.7. It's a bug fix minor revision release, not a
bung-in-loads-of-new-toys release. To remove support for a chipset in a
0.0.1 upgrade is just wrong - not merely from a feature planning point
of view, but also from a code stability point of view as well. Making
changes of that magnitude in just a bug fix release would be a poor
approach.
Cheers,
Ian
Woody
> Ric <infob...@gmail.com> wrote:
>
> > http://www.engadget.com/2009/11/09/mac-os-x-10-6-2-update-out-on-the-prowl
> > /#comments
> >
> > For anyone hackintoshing. As a side note, how/why would you remove
> > support for a specific type of X86 CPU? Isn't it just another CPU as
> > far as the OS is concerned?
>
> It's code they're not using and have no need for. Removing it makes the
> kernal smaller, more stable and easier to maintain.
Its not though is it? You don't have to write additional code to work on
an atom. I have an atom netbook running OSX 10.5 and windows XP, both
which were made before the atom was released (and in XPs case, before it
was even thought of) and it runs fine.
No, to remove atom support you have to know of the atom, and go out of
your way to find out if you are running on an atom and deliberately
stop.
So it is more code making the kernel bigger, less stable and require
more testing.
--
Woody
Peter Ceresole
> > And `kernel' (or `colonel', but not here). :)
>
> Ah, I've never been clear on that one, so generally I'm happy if I'm at
> least phonetically close.
At least it's not a major error.
Oh, shit...
--
Peter
Jim
> No, to remove atom support you have to know of the atom, and go out of
> your way to find out if you are running on an atom and deliberately
> stop.
> So it is more code making the kernel bigger, less stable and require
> more testing.
Hmmm. Good point. Also, wouldn't it be the case that for Atom-specific
code to be there, Apple would have had to add it themselves?
On the other hand, if Apple have removed Atom support purely to disable
Hackintoshes then, well, so what? Apple don't make 1p from them (unless
people have actually bought an OS from Apple for the purpose, which is
(a) foolish and (b) na�ve _and_ foolish) so why should Apple continue to
allow them to work?
It could even be argued that they are _legally_ obliged to do this, to
prevent brand dilution.
Jim (currently in 'grumpy old git' mode).
Woody
> Woody <use...@alienrat.co.uk> wrote:
>
> > No, to remove atom support you have to know of the atom, and go out of
> > your way to find out if you are running on an atom and deliberately
> > stop.
> > So it is more code making the kernel bigger, less stable and require
> > more testing.
>
> Hmmm. Good point. Also, wouldn't it be the case that for Atom-specific
> code to be there, Apple would have had to add it themselves?
They would indeed. In fact, there is atom specific code, there is code
to stop it working if it is an atom.
> On the other hand, if Apple have removed Atom support purely to disable
> Hackintoshes then, well, so what?
So what on the grand scheme of things, why are they expending
engineering effort on these things?
No manufacturer such as the pystar clones etc are producing atom based
board, so they are preventing mac enthusiasts from using their system on
hardware better supplied for the purpose (such as my hackintosh)
> Apple don't make 1p from them (unless
> people have actually bought an OS from Apple for the purpose, which is
> (a) foolish and (b) na�ve _and_ foolish) so why should Apple continue to
> allow them to work?
Well, they do. People with hackintoshes often have real macs (I have
several), so I don't see the advantage in trying to upset those people
> It could even be argued that they are _legally_ obliged to do this, to
> prevent brand dilution.
It could be argued, but it would be a very weak argument. It could also
be argued that by the amount of hackintoshes that it is a good
indication that apple aren't producing the hardware people want, and if
they increased their range to do so instead of wasting time programming
code to prevent people using their computers, they could be spending
some time providing that hardware, increasing their market share and
fixing their bugs.
Ultimately it doesn't really matter, someone will just come along and
dig deeper into the code to remove those checks, the people they are
combating here are the people who have already shown they are capable of
hacking the software.
--
Woody
Jonathan White
> Ultimately it doesn't really matter, someone will just come along and
> dig deeper into the code to remove those checks, the people they are
> combating here are the people who have already shown they are capable of
> hacking the software.
in the kernel and that Apple have somehow made the system decide not to
run it. That's only one explanation for the incompatability. It's
equally likely that some of the code they have added is coded for
things the Atom cores don't have and therefore there are no 'checks' to
'hack', it simply won't run on an Atom. Given the CPU grade they're
pushing the Mac OS towards (64 intel) I'd have said even if the
current kernel was atom compatible, that was a happy accident and was
always on borrowed time.
What the hackers would have to do it dig through the whole kernel and
find any bits that aren't atom compatible then code a kernel patch that
replaced that code with a version that is atom compatible. That's
significant'y more work that just changing an 'if' so it's always false.
Jon
Woody
> On 2009-11-15 14:37:25 +0000, use...@alienrat.co.uk (Woody) said:
> > Ultimately it doesn't really matter, someone will just come along and
> > dig deeper into the code to remove those checks, the people they are
> > combating here are the people who have already shown they are capable of
> > hacking the software.
> You're assuming the ability to work on an Atom grade CPU still exists
> in the kernel and that Apple have somehow made the system decide not to
> run it. That's only one explanation for the incompatability.
It is only one explanation for the incompatibility except when the
report of it not running was investigated after someone found some code
introduced in 10.6.2 that specifcially looks for the atom processor and
stops the system from running if found.
> equally likely that some of the code they have added is coded for
> things the Atom cores don't have and therefore there are no 'checks' to
> 'hack', it simply won't run on an Atom.
It runs fine on an atom
> Given the CPU grade they're
> pushing the Mac OS towards (64 intel) I'd have said even if the
> current kernel was atom compatible, that was a happy accident and was
> always on borrowed time.
The same as the original macbooks then and the macbook air?
--
Woody
Alienrat Design Ltd
Flavio Matani
That is the general view (not often held in private..)
--
flavio matani
guitar tuition
http://www.flaviomatani.co.uk
http://fflavio.com
Pd
> Peter Ceresole <pe...@cara.demon.co.uk> wrote:
>
> > Jim <j...@magrathea.plus.com> wrote:
> >
> > > > And `kernel' (or `colonel', but not here). :)
> > >
> > > Ah, I've never been clear on that one, so generally I'm happy if I'm at
> > > least phonetically close.
> >
> > At least it's not a major error.
>
> That is the general view (not often held in private..)
Brigadier and I'll have a look at it. (Sorry, I have a cold.)
--
Pd
Marco Bakker
The Air has had a Core2Duo inside form the beginning, unlike the early
MacBooks which shipped with CoreDuo's.
--
marco