Page 3 of 3 FirstFirst 123
Results 21 to 30 of 30

Thread: Knoppix V6.7.1

  1. #21
    Moderator Moderator
    Join Date
    Mar 2004
    Location
    Menlo Park, California
    Posts
    674
    Hi Klaus,
    Your steps are working great, thank you very much !
    (I should stop using synaptic and better start using apt-get...)
    Best Regards,
    Gilles

  2. #22
    Senior Member registered user
    Join Date
    May 2006
    Location
    Columbia, Maryland USA
    Posts
    1,631
    Quote Originally Posted by Werner P. Schulz View Post
    I've found the cause of this strange behaviour; it is the testing version of "lxpanel". After downgrade to stable version "0.5.6-1" I have the menu at my preferred position.
    @ Werner

    PCLinuxOS has another work-around at post #15 of this thread:
    http://www.pclinuxos.com/forum/index...,95512.15.html

    Cheers,

  3. #23
    Senior Member registered user
    Join Date
    May 2006
    Location
    Columbia, Maryland USA
    Posts
    1,631
    @ Werner

    There is also an lxmenu (empty menu) work-around on the LXDE wiki that may interest you at:
    http://wiki.lxde.org/en/LXPanel#gxine_menu_icon_fix

  4. #24
    Moderator Moderator
    Join Date
    Mar 2004
    Location
    Menlo Park, California
    Posts
    674
    Hello,
    Does anyone know why compiling a source tree as "root" is much slower than as the default "Knoppix" user ?
    (by just using "su" before starting the make). Top indicates a much higher percentage of system call (%sy)
    with root than Knoppix and as such there is not much time left for the user apps (%us). Any idea how to fix it ?
    (I know it is usually a bad idea to compile as root anyway but in some cases root is required...)
    Thanks,
    Gilles

  5. #25
    Senior Member registered user
    Join Date
    Dec 2009
    Posts
    423
    Quote Originally Posted by ruymbeke View Post
    Hello,
    Does anyone know why compiling a source tree as "root" is much slower than as the default "Knoppix" user ?
    (by just using "su" before starting the make). Top indicates a much higher percentage of system call (%sy)
    with root than Knoppix and as such there is not much time left for the user apps (%us). Any idea how to fix it ?
    (I know it is usually a bad idea to compile as root anyway but in some cases root is required...)
    Thanks,
    Gilles
    Personally I am not so convinced that compiling things as root will be slower, unless the makefiles do additional things when detecting compiling as root verses normal users. In any case, if one compiles something repeatedly, the speed can be improved by installing/configuring 'ccache'. Therefore if compile as root is indeed slower, perhaps the cache has not been setup for 'root' ? It's easy to verify that, just check if there is a .ccache folder for /home/knoppix. If the folder exists, meaning likely you have been using ccache without realizing it.

  6. #26
    Moderator Moderator
    Join Date
    Mar 2004
    Location
    Menlo Park, California
    Posts
    674
    Hi kl522,
    I do clear ccache: "ccache -C; ccache -z" and do a "make distclean" before the "make"
    to have a fair comparison between the root and knoppix user. It is actually a cross-compile
    but I don't think that it makes any difference...
    Best Regards,
    Gilles

  7. #27
    Senior Member registered user
    Join Date
    Dec 2009
    Posts
    423
    Quote Originally Posted by ruymbeke View Post
    Hi kl522,
    It is actually a cross-compile
    but I don't think that it makes any difference...
    The ccache may not support your cross-compile, you have to check if the "non-ccache" compiler binary name is one of those having the same name as one of those in /usr/lib/ccache. If the binary you used for cross-compile is not any of those, then too bad, debian ccache package does not support your cross-compile.

  8. #28
    Moderator Moderator
    Join Date
    Mar 2004
    Location
    Menlo Park, California
    Posts
    674
    Hi Kl522,
    I do run my test both with and without ccache. As expected, running twice or more with ccache provides some speed improvement.
    But at this point I do not care much about ccache. The problem that I try to understand is the time difference when compiling as root
    or as the knoppix user. Or in other words, in both cases (with or without ccache) running as root is much slower than with the user Knoppix.
    Thanks for your help. Best Regards,
    Gilles

  9. #29
    Senior Member registered user
    Join Date
    Dec 2009
    Posts
    423
    Quote Originally Posted by ruymbeke View Post
    The problem that I try to understand is the time difference when compiling as root
    or as the knoppix user. Or in other words, in both cases (with or without ccache) running as root is much slower than with the user Knoppix.
    Without disclosing any of your proprietary work/sources, are you able to provide with a test-case of compiling as root is slower than compiling as knoppix user ?

  10. #30
    Moderator Moderator
    Join Date
    Mar 2004
    Location
    Menlo Park, California
    Posts
    674
    Hi Kl522,
    After trying to reproduce the problem with some sample code and a toolchain publicly available
    I found out that the problem is linked to the Code Sourcery ARM toolchain (full license).
    Or in other words the problem disappear when the lite version is used. I did post a support
    request with Code Sourcery as this problem is most likely not linked to Knoppix at all.
    Thanks for your help, case closed.
    Best Regards,
    Gilles

Page 3 of 3 FirstFirst 123

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


Dell PowerEdge M620 Blade Server 2x 10C E5-2690v2 64GB Ram 2x 600GB 10k HDD picture

Dell PowerEdge M620 Blade Server 2x 10C E5-2690v2 64GB Ram 2x 600GB 10k HDD

$125.00



Dell PowerEdge FX2s CTO Blade 4 Slot 2U Chassis 2x 2000W picture

Dell PowerEdge FX2s CTO Blade 4 Slot 2U Chassis 2x 2000W

$399.00



HP ProLiant BL460c Gen9 Blade 2x E5-2697v4 2.3GHz =36 Cores 64GB P246 650FLB picture

HP ProLiant BL460c Gen9 Blade 2x E5-2697v4 2.3GHz =36 Cores 64GB P246 650FLB

$279.00



DELL PowerEdge M630 Blade 2x E5-2680v4 2.4GHz =28 Cores 128GB H330 2x10Gb X520 picture

DELL PowerEdge M630 Blade 2x E5-2680v4 2.4GHz =28 Cores 128GB H330 2x10Gb X520

$275.00



Dell PowerEdge M620  Blade Server 2 x Xeon e5 2665 8 x 16 GB(128) RAM - No HD picture

Dell PowerEdge M620 Blade Server 2 x Xeon e5 2665 8 x 16 GB(128) RAM - No HD

$49.99



HP ProLiant BL460c Gen9 Blade Server Intel Xeon E5-2698 v3 - 512GB RAM picture

HP ProLiant BL460c Gen9 Blade Server Intel Xeon E5-2698 v3 - 512GB RAM

$255.42



Google Search Appliance - Dell PowerEdge R710 x2 Xeon X5690 192GB Memory 6TB HDD picture

Google Search Appliance - Dell PowerEdge R710 x2 Xeon X5690 192GB Memory 6TB HDD

$529.99



DELL POWEREDGE FC640 Blade Server FX2 CTO Barebones iDrac Enterprise Warranty picture

DELL POWEREDGE FC640 Blade Server FX2 CTO Barebones iDrac Enterprise Warranty

$199.00



UCSB-B200-M4 UCS Blade Server, 2x E5-2667 V3, 256GB RAM DDR4, 2x 300GB Drives picture

UCSB-B200-M4 UCS Blade Server, 2x E5-2667 V3, 256GB RAM DDR4, 2x 300GB Drives

$299.99



HP BL460c G10 Blade 2x Gold 6140 2.3GHz =36 Cores 128GB P204i-B QMH2672 2x 20GB picture

HP BL460c G10 Blade 2x Gold 6140 2.3GHz =36 Cores 128GB P204i-B QMH2672 2x 20GB

$642.00