Linus Torvalds (torvalds@osdl.org) on 5/13/07 wrote:
---------------------------
>PAE really really sucks.
>
> ...
>
>Reasons for why you need a bigger virtual space:
>
>- you need to map that physical memory somehow, and no,
>tiny windows into the physical memory simply do not
>cut it! If you cannot have normal pointers to the
>physical space, it immediately means that you need to
>jump through serious hoops to get there.
>
>- you additionally need to be able to remap things
>in alternate ways (ie user space) or make space for
>non-memory issues (virtual page tables, IO, you name it)
>
>Ergo, a factor-of-two is a requirement. PAE was a total
>and utter disaster.
>
>Yes, Linux supported it, and probably did so better than
>anybody else. But "better than anybody else" still wasn't
>very good. Because you couldn't use normal pointers to
>point to arbitrary physical memory, all the memory that
>couldn't be accessed directly (ie anythign that didn't fit
>in the virtual address map, which also had the user
>space memory in it) was basically limited to "special uses
>only".

Wow Linus. I hope you did not spend a lot of time on your post just to answer me because I did not think my post deserved that much attention. :) I am flattered.

I certainly agree that PAE sucks for the reasons you describe but what I wanted to get at was Microsoft's decision not to support PAE on the desktop when they had support already available.

For similar reasons, I questioned AMD's initial Athlon 64 offering for not allowing inexpensive dual processor workstation designs from the start but there I suspect AMD had better reasons then Microsoft for the decisions they made.

>I'm not at all surprised that Windows didn't push PAE
>either. It was a total braindamage. I bet they supported
>it in the server offerings just because they had to, and
>I bet they did a much worse job of it than Linux did, and
>the reason you can do that with servers is that the loads
>are much easier, and you can expect maintainers to set
>magic config entries to tweak it to make it appear
>to work well for any particular load, when in reality it
>is fragile as hell and works only with duct-tape and
>prayers.
>
>That kind of "duct-tape and prayers and lots of specialized
>knowledge about the load" is simply not possible in
>a desktop environment. Yeah, users have prayers, but they
>lack the duct-tape and the knowledge to work around the
>problems.

So it was an applications support issue instead of a marketing issue?

>And dammit, in this age and date when almost everybody
>has a gigabyte of RAM in any new machine, anybody who still
>thinks that "not that many people need 64-bits" is simply
>not aware of what he's speaking of.

I actually have 2 GB on my old Pentium 4 which is still my main machine and the reason I did not get more RAM was inadequate desktop support. Ya, I should be using a real operating system but getting past legacy applications has proven to be difficult for me.

>Go back and play with HIGHMEM.SYS on a 286, and stop
>blathering crap. When you've spent the last ten
>years of your life working with HIGHMEM.SYS, then
>you can come back and tell me that we still don't need
>64 bits. Until that is the case, anybody who still doesn't
>get why 64 bits is a requirement should just shut
>up rather than make a total fool of himself.>

Microsoft's Address Windowing Extensions allow this very use and I agree that it is at best a kludge. I was not advocating the use of 32 bits plus PAE over true 64 bit addressing but instead questioning Microsoft's lack of PAE support on the desktop before x86-64 became available. (Ha. Talk about getting back onto the previous "what should 64 bit x86 be called" topic.)

I did live through all the years of HIMEM.SYS and actually did some application programming at that time but I was more of a QEMM and Desqview guy. The whole situation sucked but at least I avoided Windows until 95 became available except for using Mathematica.

> ...
>
>So no, PAE does not mean that you can use more than
>4GB of RAM. Even before PAE, the practical limit was around
>1GB, and PAE didn't move that post a fraction of an inch!

I never argued that PAE allowed the use of more then 4GB of linear or virtual address space. It does allow the use of more then 4GB of physical address space by multiple seperate applications. I actually recently went through the current Intel documentation on segments, paging, etc. to confirm my thoughts on the subject.