Windows manages memory differently

By: rwessel (robertwessel.delete@this.yahoo.com), May 14, 2007 6:09 pm
Room: Moderated Discussions
Tzvetan Mikov (tzvetanmi@yahoo.com) on 5/14/07 wrote:
---------------------------
>You are right, of course, although I don't quite understand why you brought up
>PAE. I was explicitly talking about less than 4 GB RAM. There has been some confusion
>on this because in Linux anything above 1 GB is considered HIGHMEM. This however
>is a design decision specific to Linux and is not necessarily an absolute requirement for a 32-bit x86 OS.
>
>If your applications do not _individually_ require more than 2 GB address space,
>and if your RAM plus your PCI devices is <= 4 GB, why do you really need a 64-bit OS ?
>
>Let's see. Fundamentally, what does the OS kernel need address space for ? Let's assume a 3G:1G user/kernel split.
>
>a) For its own code and data structures, including page tables. 1 GB should be
>more than sufficient for that on a 32-bit system.
>
>b) For memory-mapped hardware. With graphics cards with 512MB RAM and more this
>can be a problem on 32-bit systems. A solution is to map such hardware only in user space.
>
>c) In a system call the OS kernel has to be able access the entire address space
>of the calling process. Obviously this is not a problem, as the kernel can use the user addresses directly.
>
>d) Sometimes the kernel has to access parts of the address space of an arbitrary
>process. Since it needs only parts, not the entire space, it can lock & map the
>necessary address pages to kernel space. Windows does this persistently (the pages
>are mapped in advance). 32-bit Linux ends up doing it dynamically - it maps a page
>if it needs it and it is > 1 GB, and then unmaps it ASAP.
>
>e) For caches. This is the only problematic part. Ideally the kernel should be
>able to use all available RAM as cache, but it can't fit all in the kernel address
>space. However since this is a specialized data structure, it should be possible to map memory in & out as needed.
>
>Am I missing something ?
>
>The point is, the OS doesn't need generic access to all memory at all times.
>So, unless your applications require lots of virtual address space, having 2-3G
>of RAM doesn't necessarily have to be a cause for moving to 64-bit.


It's a long thread, PAE was in there somewhere... ;-)

Anyway, for some OSs, with the limits you listed (<4GB RAM+devices, all small processes), there's no particular pressure to go 64 bit, although that's not universally true.

The one place you run into a problem is with I/O, especially at interrupt time.

Having both the user address space that you're servicing and the real address space accessible at the same time is quite handy for some types of I/O operations. In that case you need more virtual address space than physical. The alternative involves either special casing the driver's environment to always deal with real addresses (extra overhead for many operations), or requiring the driver to know when it needs to use the real or virtual addresses (extra complexity in the driver). In Windows, for example, that’s just part of the basic design of the kernel/driver interface, so in a sense getting to the 4GB limit is “free” because all drivers have had to pay the hidden cost to support it (because of a more complex interface/environment).

There have been OSs where all the address spaces are accessible to the kernel/drivers. Obviously if you do that just with virtual address bits, you'll burn a lot of them, but it does make your life in a driver rather simpler. And having spent a lot of my life in that environment, I think almost anything that simplifies a device driver's existence is a good idea.

Also, there are good reasons to want your virtual address space to be fairly sparse, to make it easier to support things like shared memory. Shared memory in Windows is painful since there's no guarantee that it's mapped to the same place in both processes.
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
Rock/Tukwila rumorsmas2007/05/05 12:59 PM
  Rock/Tukwila rumorsDavid Kanter2007/05/05 02:33 PM
    Rock/Tukwila rumorsDean Kent2007/05/05 03:35 PM
      K8 vs Win64 timelineanonymous2007/05/05 06:19 PM
        Yes, I misremembered...Dean Kent2007/05/05 10:03 PM
    RockDaniel Bizó2007/05/06 02:34 AM
      RockDean Kent2007/05/06 07:11 AM
    Rock/Tukwila rumorsJoe2007/05/06 11:24 AM
      Rock/Tukwila rumorsDean Kent2007/05/06 11:49 AM
      Rock/Tukwila rumorsLinus Torvalds2007/05/06 12:09 PM
      Rock/Tukwila rumorsanon2007/05/07 01:32 AM
        Rock/Tukwila rumorsRakesh Malik2007/05/07 09:36 AM
          Rock/Tukwila rumorsMichael S2007/05/07 10:06 AM
          Rock/Tukwila rumorsanon2007/05/07 09:48 PM
            Rock/Tukwila rumorsRakesh Malik2007/05/08 06:45 AM
              Rock/Tukwila rumorsanon2007/05/08 05:30 PM
                Wow. (nt)Brannon2007/05/08 06:16 PM
                Rock/Tukwila rumorsrwessel2007/05/08 09:48 PM
                  Rock/Tukwila rumorsJS2007/05/08 10:07 PM
                    Rock/Tukwila rumorsJS2007/05/09 06:44 AM
                Rock/Tukwila rumorsRakesh Malik2007/05/09 05:35 AM
                  Much ado about xMichael S2007/05/09 09:39 AM
                    Call it x86-64Linus Torvalds2007/05/09 10:27 AM
                      (i)AMD64Michael S2007/05/09 12:16 PM
                        (i)AMD64Linus Torvalds2007/05/09 12:29 PM
                          (i)AMD64Groo2007/05/09 04:45 PM
                          TIFNAAanonymous2007/05/09 05:49 PM
                            Inspired by FYR Macedonia? (NT)Michael S2007/05/09 11:21 PM
                              More likely...rwessel2007/05/10 12:39 AM
                            TIFNAAGabriele Svelto2007/05/09 11:57 PM
                          (i)AMD64James2007/05/10 02:27 AM
                        i86Dean Kent2007/05/09 12:30 PM
                        (i)AMD64Max2007/05/09 01:28 PM
                          wide86? long86?hobold2007/05/10 05:05 AM
                            x87 perhaps, it is one more. :) (NT)Groo2007/05/10 05:50 AM
                              x86+Dean Kent2007/05/10 08:44 AM
                                Does it really matter?Doug Siebert2007/05/10 09:10 AM
                                  let's stay with x86-64 for now, pleaseMarcin Niewiadomski2007/05/10 11:50 AM
                                    let's stay with x86-64 for now, pleaseDean Kent2007/05/11 06:11 AM
                                      let's stay with x86-64 for now, pleaserwessel2007/05/11 02:46 PM
                                        let's stay with x86-64 for now, pleaseDean Kent2007/05/11 06:03 PM
                                          let's stay with x86-64 for now, pleaseMichael S2007/05/12 10:49 AM
                                            let's stay with x86-64 for now, pleaseDean Kent2007/05/12 01:05 PM
                                              let's stay with x86-64 for now, pleaseMichael S2007/05/12 01:25 PM
                                                let's stay with x86-64 for now, pleaseDean Kent2007/05/12 03:39 PM
                                                  let's stay with x86-64 for now, pleaseJasonB2007/05/13 07:43 AM
                                                    client consolidationMichael S2007/05/13 08:37 AM
                                                  let's stay with x86-64 for now, pleaseTzvetan Mikov2007/05/13 03:44 PM
                                                let's stay with x86-64 for now, pleaserwessel2007/05/14 02:42 PM
                                      What's your point?Doug Siebert2007/05/11 02:56 PM
                                        What's your point?Linus Torvalds2007/05/11 04:15 PM
                                          What's your point?Doug Siebert2007/05/13 03:11 PM
                                            What's your point?Dean Kent2007/05/13 07:04 PM
                                              What's your point?JasonB2007/05/14 02:06 AM
                                                What's your point?Dean Kent2007/05/14 07:20 AM
                                                  What's your point?JasonB2007/05/14 04:35 PM
                                                  What's your point?JasonB2007/05/14 07:35 PM
                                                    What's your point?Dean Kent2007/05/14 08:12 PM
                                        What's your point?Dean Kent2007/05/11 06:06 PM
                                        What's your point?Stephen H2007/05/13 01:55 AM
                                          Why didn't MS take advantage of PAE?David W. Hess2007/05/13 08:37 AM
                                            PAE sucks (Why didn't MS take advantage of PAE?)Linus Torvalds2007/05/13 10:20 AM
                                              PAE sucks (Why didn't MS take advantage of PAE?)Dean Kent2007/05/13 10:49 AM
                                              PAE sucks (Why didn't MS take advantage of PAE?)David W. Hess2007/05/13 12:37 PM
                                              > 1 GB RAM on a 32-bit systemTzvetan Mikov2007/05/13 01:44 PM
                                                > 1 GB RAM on a 32-bit systemS. Rao2007/05/13 03:00 PM
                                                  > 1 GB RAM on a 32-bit systemTzvetan Mikov2007/05/13 05:32 PM
                                                    > 1 GB RAM on a 32-bit systemS. Rao2007/05/14 12:19 AM
                                                > 1 GB RAM on a 32-bit systemLinus Torvalds2007/05/13 03:46 PM
                                                  > 1 GB RAM on a 32-bit systemTzvetan Mikov2007/05/13 05:23 PM
                                                  > 1 GB RAM on a 32-bit systemJasonB2007/05/13 06:37 PM
                                                    Windows manages memory differentlyTzvetan Mikov2007/05/13 08:31 PM
                                                      Windows manages memory differentlyJasonB2007/05/14 01:50 AM
                                                        Windows manages memory differentlyTzvetan Mikov2007/05/14 08:56 AM
                                                          Windows manages memory differentlyrwessel2007/05/14 03:40 PM
                                                            Windows manages memory differentlyDavid W. Hess2007/05/14 04:07 PM
                                                              Windows manages memory differentlyrwessel2007/05/14 04:51 PM
                                                            Windows manages memory differentlyTzvetan Mikov2007/05/14 05:40 PM
                                                              Windows manages memory differentlyrwessel2007/05/14 06:09 PM
                                                      Windows manages memory differentlyHoward Chu2007/05/14 11:17 AM
                                                        Windows manages memory differentlyJukka Larja2007/05/14 11:30 AM
                                                        Windows manages memory differentlyTzvetan Mikov2007/05/14 01:54 PM
                                                          Windows manages memory differentlyHoward Chu2007/05/15 03:35 AM
                                                            Windows manages memory differentlyGroo2007/05/15 07:34 AM
                                                      Anyone know what OS X (10.4, Intel, desktop) does?Matt Sayler2007/05/15 06:23 AM
                                                        Anyone know what OS X (10.4, Intel, desktop) does?Wes Felter2007/05/15 08:37 AM
                                                        Anyone know what OS X (10.4, Intel, desktop) does?Anonymous2007/05/15 10:49 AM
                                                        Anyone know what OS X (10.4, Intel, desktop) does?anon22007/05/15 07:13 PM
                                              PAE sucks (Why didn't MS take advantage of PAE?)Paul2007/05/13 03:40 PM
                                                PAE sucks (Why didn't MS take advantage of PAE?)Peter Arremann2007/05/13 05:38 PM
                                                  PAE sucks (Why didn't MS take advantage of PAE?)Henrik S2007/05/14 03:31 AM
                                              The fragility of your argumentslacker2007/05/13 03:56 PM
                                                The fragility of your argumentnick2007/05/13 05:42 PM
                                                The fragility of your argumentHoward Chu2007/05/14 02:52 AM
                                                  The fragility of your argumentDean Kent2007/05/14 09:19 AM
                                                The fragility of your argumentanon22007/05/14 08:26 AM
                                                  The fragility of your argumentTzvetan Mikov2007/05/14 09:01 AM
                                                  The fragility of your argumentDean Kent2007/05/14 09:16 AM
                                                    The fragility of your argumentLinus Torvalds2007/05/14 11:57 AM
                                                      The fragility of your argumentJasonB2007/05/14 04:48 PM
                                                        The fragility of your argumentDean Kent2007/05/14 07:36 PM
                                                          The fragility of your argumentRicardo B2007/05/16 02:40 AM
                                                            The fragility of your argumentDean Kent2007/05/16 03:32 AM
                                                              The fragility of your argumentRicardo B2007/05/16 06:41 AM
                                                                PSRicardo B2007/05/16 06:50 AM
                                                                The fragility of your argumentDean Kent2007/05/16 09:07 AM
                                                                  Modern web browsingS. Rao2007/05/16 09:16 AM
                                                                    Aha!Dean Kent2007/05/16 09:27 AM
                                                                      Aha!Dean Kent2007/05/16 09:32 AM
                                                                        Aha!S. Rao2007/05/16 10:34 AM
                                                                  The fragility of your argumentRicardo B2007/05/16 10:00 AM
                                                      The fragility of your argumentVincent Diepeveen2007/05/16 10:10 AM
                                                        The fragility of your argumentPaul2007/05/16 03:01 PM
                                                          The fragility of your argumentVincent Diepeveen2007/05/17 03:05 AM
                                                    The fragility of your argumentanon22007/05/15 01:35 AM
                                                    Splits vs page allocations?Matt Sayler2007/05/15 07:33 AM
                                          What's your point?Michael S2007/05/13 08:55 AM
                                            What's your point?anonymous2007/05/13 11:08 AM
                                              What's your point?Michael S2007/05/13 11:31 AM
                                      let's stay with x86-64 for now, pleaseJasonB2007/05/13 07:16 AM
                                x864 =) (NT)some12007/05/15 03:03 AM
    Rock/Tukwila rumorsIntelUser20002007/05/06 02:27 PM
  Rock/Tukwila rumorsm2007/05/13 08:05 AM
  Rock/Tukwila rumorsmas2007/05/15 09:40 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?