Slashdot Mirror


XP SP2 Can Slow Down Business Apps

An anonymous reader submits "Mobile PC magazine installed XP SP2 on a bunch of notebooks and benchmarked them, finding that SP2 caused a 9-percent performance reduction in business productivity apps. While a couple of notebooks performed better, the majority took a 3- to 22-percent performance hit." For now, the story is just at the top of the Mobile PC website, but they promise more details in an upcoming issue.

17 of 359 comments (clear)

  1. Another benchmark test by Adam9 · · Score: 4, Informative

    Here is another article where they ran different benchmarks on SP2 and SP1. The office productivity test was the one with the biggest difference. The article puts the blame on the new firewall.

    They should compare a PC with SP2 and one with SP1 with a third party firewall.

  2. Coral Cache Link by Anonymous Coward · · Score: 3, Informative

    Cached link in case it gets Slashdotted.

  3. Not a code change.. a compiler flag change.. by bmajik · · Score: 5, Informative

    specifically, the /GS flag to the VC++ compiler.

    The compiler was modified to support automatic stack overflow checking (i.e. canaries). Server 2003 was compiled with this (and as a result, MANY things that are shared-code problems resulting in exploits on other NT based OSes are either ineffective or DoS attacks on Server 2003).

    The idea is that /GS compiled binaries will cause the OS to terminate the app rather then letting code execute. The source code generally doesn't need changes.

    So, its a defense in depth tactic. Ideally, there'd be no BO's in code. But there are. Terminating the program with an explanation as to why is better than letting people run code on your box. :)

    --
    My opinions are my own, and do not necessarily represent those of my employer.
    1. Re:Not a code change.. a compiler flag change.. by darkain · · Score: 2, Informative

      actually, snprintf was marked as unsecure as well in the new CRT. sprintf_s is the replacement for both sprintf and snprintf.

  4. Re:This has to do with... by Anonymous Coward · · Score: 1, Informative

    This is an issue only on CPUs that can slow down their clock speed, as in laptops, desktop Athlon 64s, etc. The cpufreq driver (or whatever it's called on Windows) is broken and apparently does not correctly jack up the clock speed to the maximum when it's needed. So a 2.4 GHz machine may be running at 500 MHz even when demanding tasks need it most.

  5. Re:It was to be expected... by teamhasnoi · · Score: 1, Informative
    Funny, every update I've installed has made my computer faster, and more reliable.

    Then again, I'm running OS X.

    Why doesn't MS copy that?

  6. Hello morons, turn off DEP by Anonymous Coward · · Score: 3, Informative

    http://support.microsoft.com/default.aspx?kbid=875 352&product=windowsxpsp2

    Note the /NoExecute=AlwaysOff option in the article.

    Well known cause for much of the slow down some people find with SP2. Of course, this opens you up to morphic/purposefully overwritten code exploits, but such is life.

    1. Re:Hello morons, turn off DEP by Qrlx · · Score: 2, Informative

      Hmm.... from the KB article

      Currently, the only x86 processors that support No-Execute functionality are the AMD 32/64-bit Opteron and Athlon-64.

      Since this doesn't affect Intel, this can't account for all the slow-downs people are expieriencing.

      (Not that I bothered to read the article in Mobile PC.)

    2. Re:Hello morons, turn off DEP by robhancock · · Score: 3, Informative

      No-execute protection still does something on CPUs not supporting hardware NX, it's just not quite as effective.

    3. Re:Hello morons, turn off DEP by Anonymous Coward · · Score: 1, Informative

      From the "Data Execution Prevention" settings dialog:

      Your computer's processor does not support hardware-based DEP. However, Windows can use DEP software to help prevent some types of attacks.

      This software version of DEP is on by default, and will probably cause some loss of performance.

  7. Re:Putty and SP2 (It's a business app, really! ;) by Anonymous Coward · · Score: 1, Informative
    I had noticed this, but hadn't made the connection. I just tried your suggestion (2k compat mode), which did not resolve the problem.



    I just checked, and I was not running the most recent version of putty (.53 vs .55). Upgrading it to the latest version appears to have fixed the issue.

  8. Re:Buffer checks by cowbutt · · Score: 3, Informative
    Immunix for one. Alternatively, taking a slightly different path towards pro-active security measures, Red Hat has recently included exec-shield (as seen previously in Fedora Core 1 onwards) in RHEL3 update 3. FC2 includes SELinux, so that'll probably turn up in RHEL eventually, too.

    --

  9. Re:Buffer checks by Tim+C · · Score: 2, Informative

    How should he know how a NULL is handled?

    Well, if he wrote the code that needs to handle the null, then he need to know how it should be handled. If he's inexperienced/junior enough to not be able to decide himself, he should speak to someone who can make that decision for him.

  10. Re:Poll question! by vrt3 · · Score: 2, Informative

    Only if you are Roland Piquepaille or whatever his name is.

    --
    This sig under construction. Please check back later.
  11. Re:You might be hitting the socket limit.. by Tim+C · · Score: 2, Informative

    ok don't let the computer make more than 5 simultaneous connections

    That's not what it's doing at all, you can still have as many simultaneous connections as you want. What you can't have is more than 20 connections waiting to completely open; any more than 20 get queued until some of those 20 have completed (or failed).

    Sure, sucks if you run apps that regularly open up a whole bunch of connections to hosts that may or may not be answering right now, but having checked my logs it's only happened to me a handful of times since I installed SP 2, despite running p2p apps while web browsing or playing online games, so it's really not that big a deal for "normal" users.

  12. Re:Buffer checks by dirk · · Score: 2, Informative

    While you are right it is not fair to compare coding from 20 years ago with that of today, it is also unfair to compare OpenBSD with MS. They are aiming at 2 completely seperate goals, so of course they will be different. OpenBSD has the goal of being as secure as possible. They are extremely good at this. They also do not support many of the newest and greatest things and their usability is pretty bad. MS has the opposite goal. They want to have an incredibly usable OS which supports all the latest and greatest hardware and innovations.

    Security and usability are on 2 ends of the same spectrum. If you want usability, you have to give up security and vice versa. MS for years has tried to get as close to usability as they could and they gave up security to do it. They are now trying to strike a balance between the 2. They will never be secure as BSD, because they can't give up all usability like BSD can, because their user base is Joe Average.

    --

    "Information wants to be expensive" - Stewart Brand, the same guy who said "Information wants to be free"
  13. Re:Buffer checks by jazman_777 · · Score: 2, Informative
    given my today's knowledge of Physics and Mathematics that I learn by my twelfth grade, I would have been the most intelligent man alive 400 years ago.

    Learning what other people have figured out is not intelligence, it's education. You'd be more educated, but I hardly think more intelligent.

    --
    Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.