Menu
▾
▴
colinux-devel — Cooperative Linux development mailing list
You can subscribe to this list here.
| 2004 |
Jan
(64) |
Feb
(530) |
Mar
(266) |
Apr
(580) |
May
(360) |
Jun
(161) |
Jul
(185) |
Aug
(164) |
Sep
(123) |
Oct
(160) |
Nov
(59) |
Dec
(84) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2005 |
Jan
(156) |
Feb
(95) |
Mar
(124) |
Apr
(81) |
May
(79) |
Jun
(179) |
Jul
(35) |
Aug
(64) |
Sep
(56) |
Oct
(57) |
Nov
(18) |
Dec
(41) |
| 2006 |
Jan
(65) |
Feb
(37) |
Mar
(59) |
Apr
(73) |
May
(65) |
Jun
(27) |
Jul
(54) |
Aug
(76) |
Sep
(103) |
Oct
(23) |
Nov
(45) |
Dec
(29) |
| 2007 |
Jan
(41) |
Feb
(47) |
Mar
(61) |
Apr
(24) |
May
(14) |
Jun
(6) |
Jul
(23) |
Aug
(30) |
Sep
(16) |
Oct
(9) |
Nov
(53) |
Dec
(36) |
| 2008 |
Jan
(19) |
Feb
(49) |
Mar
(74) |
Apr
(21) |
May
(24) |
Jun
(5) |
Jul
(9) |
Aug
(53) |
Sep
(26) |
Oct
(23) |
Nov
(32) |
Dec
(19) |
| 2009 |
Jan
(47) |
Feb
(49) |
Mar
(39) |
Apr
(61) |
May
(28) |
Jun
(19) |
Jul
(12) |
Aug
(10) |
Sep
(31) |
Oct
(16) |
Nov
(60) |
Dec
(26) |
| 2010 |
Jan
(17) |
Feb
(9) |
Mar
(32) |
Apr
(11) |
May
(24) |
Jun
(33) |
Jul
(5) |
Aug
(2) |
Sep
(7) |
Oct
(8) |
Nov
(17) |
Dec
(7) |
| 2011 |
Jan
(12) |
Feb
(16) |
Mar
(2) |
Apr
(12) |
May
(5) |
Jun
(10) |
Jul
(3) |
Aug
(3) |
Sep
(2) |
Oct
(1) |
Nov
(17) |
Dec
(1) |
| 2012 |
Jan
(9) |
Feb
(9) |
Mar
(8) |
Apr
(4) |
May
(2) |
Jun
(1) |
Jul
(4) |
Aug
(8) |
Sep
(11) |
Oct
(1) |
Nov
(2) |
Dec
(2) |
| 2013 |
Jan
|
Feb
(7) |
Mar
(4) |
Apr
(10) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(5) |
Dec
(3) |
| 2016 |
Jan
|
Feb
(2) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2023 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(1) |
|
From: Henry N. <Hen...@Ar...> - 2008-03-22 18:13:04
|
Óscar Fuentes wrote:
> Henry Nestler writes:
>
>> You wrote:
>>> Host machine: 49 seconds
>>> VMWare: 48 seconds
>>> andLinux: 108 seconds
>> There VMware is faster as native systems? That let me thing it is not
>> real.
>
> It is real. I checked it with a stopwatch.
>
>> Or it is some kind of prefetching cache to have faster disk
>> access? Perhaps VMware loads some of the disk into RAM?
>
> The test suite starts a process more than 1000 times, pipes the output
> into the controlling process, etc. My experience shows that GNU/Linux is
> more effective than Windows for this sort of tasks.
Suggest from you, here is a test script, that only eceute some
processes, without disk io:
#!/bin/bash
loop=10000
while test $loop -gt 0
do
x=`uname -r`
loop=$(($loop - 1))
done
Tested with Intel(R) Pentium(R) 4 Mobile CPU 1.80GHz
22.9s native, kernel 2.6.17
89.7s coLinux 0.8.0, kernel 2.6.22.18
Can you run tis on WVware please?
--
Henry N.
|
|
From: Óscar F. <of...@wa...> - 2008-03-22 17:00:57
|
Henry Nestler <Hen...@Ar...> writes: > You wrote: >> Host machine: 49 seconds >> VMWare: 48 seconds >> andLinux: 108 seconds > > There VMware is faster as native systems? That let me thing it is not > real. It is real. I checked it with a stopwatch. > Or it is some kind of prefetching cache to have faster disk > access? Perhaps VMware loads some of the disk into RAM? The test suite starts a process more than 1000 times, pipes the output into the controlling process, etc. My experience shows that GNU/Linux is more effective than Windows for this sort of tasks. > What is the host CPU type? Intel Pentium M @ 2 GHz (single core). -- Oscar |
|
From: <Use...@zo...> - 2008-03-22 16:59:21
|
Hen...@Ar...(Henry Nestler) 22.03.08 16:33 >You wrote: >> Host machine: 49 seconds >> VMWare: 48 seconds >> andLinux: 108 seconds >There VMware is faster as native systems? That let me thing it is not >real. I would assume a "quantization error". i ran the make twice and got different results with more than 1000ms. I assume thare are other tasks running too. >Or it is some kind of prefetching cache to have faster disk >access? Perhaps VMware loads some of the disk into RAM? Rainer |
|
From: Henry N. <Hen...@Ar...> - 2008-03-22 16:04:36
|
Óscar Fuentes wrote: >> What is the local volume? A image file somethere on disk, or a true >> partition? > > It is a image file (I hope that andLinux didn't partitioned my HD while > installing! :-) Ok. You not using the cofs. That would badly in such tests. >> What version of coLinux is there included? (uname -a) > > Linux andLinux 2.6.12-co-0.7.1 #1 Sat Jul 14 12:13:49 UTC 2007 i686 GNU/Linux Is Ok. That kernel does not use REGPARM. Small improvements comes with coLinux kernel 2.6.22, there REGPARM default. And what kernel you have in VMware? >> coLinux 0.8.0 has some improvemnets on block device with parameter >> "setcobd=async". It's available from http://www.colinux.org/snapshots/ > > I'll try it after learning how to upgrade, but the test suite is CPU > bound (if the OS does a minimally sensible disk caching). "setcobd=async" would no help for the executing time of "./configure". This helps only in systems with reading/writing from multitask or Multi-thread applications. Typicaly configure is a linear task. There, async option does not help. -- Henry N. |
|
From: Henry N. <Hen...@Ar...> - 2008-03-22 15:50:13
|
Rainer Zocholl wrote: > (I wonder why "bogomips" are only 100 on colinux, bogo...) That is the overhead between native and coLinux. In the loop where bogomips are calculated, coLinux does only switching between Windows and Linux. That cost many time. coLinux goes faster, if you mounts filesystem with "noatime". For every file access, coLinux ask the host about current precice time to write the file "last access time". If starts many of processes, it can the speed problem here. -- Henry N. |
|
From: Henry N. <Hen...@Ar...> - 2008-03-22 15:33:25
|
Paolo Minazzi wrote: > On Sat, Mar 22, 2008 at 1:01 AM, Óscar Fuentes <of...@wa...> wrote: >> Óscar Fuentes <of...@wa...> writes: >> >> > Henry Nestler <Hen...@Ar...> writes: >> > >> >> Can I download this test somethere? >> > >> > No, sorry. But I expect the same from the gcc testsuite, for instance, >> > as it is very similar. Compiling is slower too. It is a C++ project and >> > is compiled with g++ 4.1.2 in VMWare (80 seconds) and 4.1.3 in coLinux >> > (124 seconds). >> >> This is a another good test case: >> >> Building GNU Make from >> >> http://ftp.gnu.org/pub/gnu/make/make-3.81.tar.gz >> >> coLinux VMWare >> configure: 42.2 seconds 18.7 seconds >> make : 15.1 seconds 8.5 seconds >> >> As mentioned on my previous message, VMWare is 6.0.0 with Debian etch as >> guest OS, gcc 4.1.2. coLinux is 0.7.1 with Ubuntu Gutsy, gcc 4.1.3. coLinux-0.7.3 native configure: 63s 28s make: 28s 15s There can see, that colinux is ~50% slower as native. I can only compair native with coLinux. Have no VMware under Windows. CoLinux runs with 256MB of native 480MB RAM, gcc 3.3.1, both runs on same installation (dualboot). You wrote: > Host machine: 49 seconds > VMWare: 48 seconds > andLinux: 108 seconds There VMware is faster as native systems? That let me thing it is not real. Or it is some kind of prefetching cache to have faster disk access? Perhaps VMware loads some of the disk into RAM? What is the host CPU type? -- Henry N. |
|
From: <Use...@zo...> - 2008-03-22 15:22:16
|
of...@wa...(?scar Fuentes) 22.03.08 00:43 Once upon a time "?scar Fuentes " shaped the electrons to say... >Use...@zo... (Rainer Zocholl) >writes: >> Take outmost care when benchmarking! >> Times in VMs are not reliable. >My stopwatch is very reliable, and it agrees with both VMWare and >coLinux :-) Ok. You didn't mention independed mesurement. >> AFAIK VMware is "optimzed" to win such bench marks. >> VMware looses on file and netaccess, aka: real world ;-( >This is not a benchmark. It is real-world work and I must execute it >dozens of times per day. Ok, your need CPU time, right, not IO. >> But comparing VMware with colinux is unfair, anyway. >Why? VMware is much much older, developed by a payed team driven by comerical interrest (you know EMC,the owner of VMware.) Colinux is mainly based on vulantary (great) work of some nice guys like Henry et al. But you are right: If there is a factor of 2 to 3 in performance waste, it might be worth a view. But 4 me "stability" is more important: I can simply buy more GHz clocks with the next chip releases. >> Maybe you can post a link to your benchmark programm? >If you insist, I'll locate a project that shows a similar behaviour, >as my test suite uses a fairly common approach. I think that would be a good idea, so i'll "insist" ;-) |
|
From: <Use...@zo...> - 2008-03-22 15:22:14
|
of...@wa...(?scar Fuentes) 22.03.08 01:01 Once upon a time "?scar Fuentes " shaped the electrons to say... >?scar Fuentes <of...@wa...> writes: >> But I expect the same from the gcc testsuite, for >> instance, as it is very similar. Compiling is slower too. It is a >> C++ project and is compiled with g++ 4.1.2 in VMWare (80 seconds) >> and 4.1.3 in coLinux (124 seconds). always exactly 50% "plus". >This is a another good test case: >Building GNU Make from >http://ftp.gnu.org/pub/gnu/make/make-3.81.tar.gz #mkdir bench #cd bench #wget http://ftp.gnu.org/pub/gnu/make/make-3.81.tar.gz #tar xzf make-3.81.tar.gz #cd make-3.81 #time ./configure <... <real 0m46.911s <user 0m10.410s <sys 0m28.630s #time make <real 0m15.610s (validated with manual stopwatch) <user 0m10.150s <sys 0m5.240s #time make <... <real 0m0.340s <user 0m0.220s sys 0m0.090s second try: real 0m43.181s user 0m10.270s sys 0m31.700s real 0m16.280s user 0m10.680s sys 0m5.570s CPU: cat /proc/cpuinfo vendor_id : GenuineIntel cpu family : 6 model : 13 model name : Intel(R) Pentium(R) M processor 1.60GHz stepping : 8 cpu MHz : 1596.000 cache size : 2048 KB bogomips : 104.65 clflush size : 64 colinux:~/work/bench/make-3.81# free total used free shared buffers cached Mem: 125944 88200 37744 0 7172 52172 -/+ buffers/cache: 28856 97088 Swap: 524280 0 524280 #uname -a Linux colinux 2.6.22.18-co-0.8.0 #1 PREEMPT Fri Mar 14 05:23:51 UTC 2008 i686 GNU/Linux # cat /etc/issue Debian GNU/Linux 4.0 \n \l # gcc --version gcc (GCC) 4.1.2 20061115 (prerelease) (Debian 4.1.1-21) ########################################## native CPU debian (dynamically throttled 2,1GHz Dual core AMD CPU): uname -a Linux vdr 2.6.23x2 #2 SMP PREEMPT Sat Oct 20 03:08:47 CEST 2007 i686 GNU/Linux vdr:~/work/bench/make-3.81# gcc --version gcc (GCC) 4.1.2 20061115 (prerelease) (Debian 4.1.1-21) configure: real 0m19.401s user 0m10.430s sys 0m7.220s make: real 0m9.315s user 0m8.510s sys 0m0.680s vdr:~/work/bench/make-3.81# cat /proc/cpuinfo processor : 0 vendor_id : AuthenticAMD cpu family : 15 model : 107 model name : AMD Athlon(tm) X2 Dual Core Processor BE-2350 stepping : 1 cpu MHz : 1000.000 cache size : 512 KB bogomips : 2010.75 ########################### Virtual Box: configure real 0m31.718s user 0m11.001s sys 0m17.721s make: real 0m11.393s user 0m8.489s sys 0m2.424s ubuntu 7.10 (With running)xwindows ~/bench/make-3.81# uname -a Linux zoc-desktop 2.6.22-14-generic #1 SMP Tue Feb 12 07:42:25 UTC 2008 i686 GNU/Linux ~/bench/make-3.81# free total used free shared buffers cached Mem: 527644 447596 80048 0 22424 305236 -/+ buffers/cache: 119936 407708 Swap: 409616 34664 374952 ~/bench/make-3.81# cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : Intel(R) Pentium(R) M processor 1.60GHz stepping : 8 cpu MHz : 1595.741 cache size : 2048 KB bogomips : 3229.85 clflush size : 64 # cat /etc/issue Ubuntu 7.10 \n \l > coLinux VMWare >configure: 42.2 seconds 18.7 seconds >make : 15.1 seconds 8.5 seconds >As mentioned on my previous message, VMWare is 6.0.0 with Debian etch >as guest OS, gcc 4.1.2. coLinux is 0.7.1 with Ubuntu Gutsy, gcc 4.1.3. So i got the same results: -Colinux is the slowest. -VirtualPC third -VMware second (not tested here as too much overhead and too deep manipulations of OS) -Native fastest (surprise, surprise ;-)) (I wonder why "bogomips" are only 100 on colinux, bogo...) |
|
From: Paolo M. <pao...@gm...> - 2008-03-22 09:09:52
|
On Sat, Mar 22, 2008 at 1:01 AM, Óscar Fuentes <of...@wa...> wrote: > Óscar Fuentes <of...@wa...> writes: > > > Henry Nestler <Hen...@Ar...> writes: > > > >> Can I download this test somethere? > > > > No, sorry. But I expect the same from the gcc testsuite, for instance, > > as it is very similar. Compiling is slower too. It is a C++ project and > > is compiled with g++ 4.1.2 in VMWare (80 seconds) and 4.1.3 in coLinux > > (124 seconds). > > This is a another good test case: > > Building GNU Make from > > http://ftp.gnu.org/pub/gnu/make/make-3.81.tar.gz > > coLinux VMWare > configure: 42.2 seconds 18.7 seconds > make : 15.1 seconds 8.5 seconds > > As mentioned on my previous message, VMWare is 6.0.0 with Debian etch as > guest OS, gcc 4.1.2. coLinux is 0.7.1 with Ubuntu Gutsy, gcc 4.1.3. > > > > -- > Oscar > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > coLinux-devel mailing list > coL...@li... > https://lists.sourceforge.net/lists/listinfo/colinux-devel > Hi, I used VMware workstation 5.5 with slackware for several months. Then I changed to colinux. I don't have time measurement. I only have the impression that colinux is a little faster compiling. But I don't have any number to demostrate it. The great advantage of colinux is to have cofs .... the vmware module to access to my host fs didn't work fine. Bye, Paolo |
|
From: coLinux a. <col...@he...> - 2008-03-22 05:07:29
|
The autobuild system has detected a new revision in the source repository. Review last changed from changelog.txt, also attached in mail. Download the compiled version: http://www.henrynestler.com/colinux/autobuild/devel-20080321/ colinux-0.8.0-20080321.src.tgz (597079 Bytes) daemons-0.8.0-20080321.dbg.zip (1314596 Bytes) daemons-0.8.0-20080321.zip (1020592 Bytes) Note, the autobuild compilation does not include an installer. Remember to reload the driver with these commands: colinux-daemon.exe --remove-driver colinux-daemon.exe --install-driver The vmlinux and modules are up to date. Please use last version from http://www.henrynestler.com/colinux/autobuild/devel-20080317/ The autobuild compilations are not official releases of Cooperative Linux software. There is no warranty that any autobuild version is stable. If use this autobuild version, please give us feedback of your experience. Job runs on machine with 64 bit version of gcc 4.1.2. A service from http://gcc.gnu.org/wiki/CompileFarm -- Lots of fun with newest version, Henry Nestler ------------------------------------------------------------------------ r980 | henryn | 2008-03-21 20:09:31 +0000 (Fri, 21 Mar 2008) | 1 line Changed paths: M /branches/devel/src/colinux/arch/i386/passage.c * Comments on passage code. ------------------------------------------------------------------------ r978 | henryn | 2008-03-21 17:00:25 +0000 (Fri, 21 Mar 2008) | 1 line Changed paths: M /branches/devel/NEWS * NEW: Backport from branch stable ------------------------------------------------------------------------ r976 | henryn | 2008-03-21 16:03:05 +0000 (Fri, 21 Mar 2008) | 1 line Changed paths: M /branches/devel/src/colinux/os/winnt/kernel/manager.c * MmAllocatePagesForMdl: Limit non supported PGE ------------------------------------------------------------------------ r973 | henryn | 2008-03-21 15:34:16 +0000 (Fri, 21 Mar 2008) | 1 line Changed paths: M /branches/devel/src/colinux/user/daemon.c M /branches/devel/src/colinux/user/debug/main.c * Warning prints, if COLINUX_DEBUG was not compiled in. ------------------------------------------------------------------------ r963 | henryn | 2008-03-21 13:37:24 +0000 (Fri, 21 Mar 2008) | 1 line Changed paths: M /branches/devel/src/colinux/os/winnt/user/install/colinux.nsi * Installer: No ShowInstDetails/ShowUnInstDetails. (port from stable) ------------------------------------------------------------------------ |
|
From: Óscar F. <of...@wa...> - 2008-03-22 00:01:34
|
Óscar Fuentes <of...@wa...> writes: > Henry Nestler <Hen...@Ar...> writes: > >> Can I download this test somethere? > > No, sorry. But I expect the same from the gcc testsuite, for instance, > as it is very similar. Compiling is slower too. It is a C++ project and > is compiled with g++ 4.1.2 in VMWare (80 seconds) and 4.1.3 in coLinux > (124 seconds). This is a another good test case: Building GNU Make from http://ftp.gnu.org/pub/gnu/make/make-3.81.tar.gz coLinux VMWare configure: 42.2 seconds 18.7 seconds make : 15.1 seconds 8.5 seconds As mentioned on my previous message, VMWare is 6.0.0 with Debian etch as guest OS, gcc 4.1.2. coLinux is 0.7.1 with Ubuntu Gutsy, gcc 4.1.3. -- Oscar |
|
From: Óscar F. <of...@wa...> - 2008-03-21 23:43:27
|
Use...@zo... (Rainer Zocholl) writes: > Take outmost care when benchmarking! > Times in VMs are not relyable. My stopwatch is very reliable, and it agrees with both VMWare and coLinux :-) > How does the host "feel" when the bench is running? > > I assume a 100% load under VMware, host rendered unusable? > while colinux is "coopperative"? The system feels snappy on both cases. As the test suite is CPU bound, there is a 100% load on both cases too, as it should. > AFAIK VMware is "optimzed" to win such bench marks. > VMware looses on file and netaccess, aka: real world ;-( This is not a benchmark. It is real-world work and I must execute it dozens of times per day. > But comparing VMware with colinux is unfair, anyway. Why? >>The result of the current state of >>coLinux? Or an andLinux specific problem? > > I found colinux pretty fast enough. > > Maybe you can post a link to your benchmark programm? If you insist, I'll locate a project that shows a similar behaviour, as my test suite uses a fairly common approach. -- Oscar |
|
From: Óscar F. <of...@wa...> - 2008-03-21 23:36:45
|
Henry Nestler <Hen...@Ar...> writes: > Can I download this test somethere? No, sorry. But I expect the same from the gcc testsuite, for instance, as it is very similar. Compiling is slower too. It is a C++ project and is compiled with g++ 4.1.2 in VMWare (80 seconds) and 4.1.3 in coLinux (124 seconds). > What is the local volume? A image file somethere on disk, or a true > partition? It is a image file (I hope that andLinux didn't partitioned my HD while installing! :-) > What version of coLinux is there included? (uname -a) Linux andLinux 2.6.12-co-0.7.1 #1 Sat Jul 14 12:13:49 UTC 2007 i686 GNU/Linux > coLinux 0.8.0 has some improvemnets on block device with parameter > "setcobd=async". It's available from http://www.colinux.org/snapshots/ I'll try it after learning how to upgrade, but the test suite is CPU bound (if the OS does a minimally sensible disk caching). coLinux is really awesome. Congratulations! -- Oscar |
|
From: Henry N. <Hen...@Ar...> - 2008-03-21 21:39:43
|
Óscar Fuentes wrote: > Hi! > > I've just installed andLinux Beta 1 rc6 (02/12/2008) (minimal / XFCE > version) on WinXP SP2. > > I'm really impressed with andLinux (which implies coLinux!). > > However, I'm dissappointed about its performance: running a test suite > on the host machine, VMWare Workstation 6.0 (Guest: Debian Etch) and > andLinux takes this time: > > Host machine: 49 seconds > VMWare: 48 seconds > andLinux: 108 seconds > > The test suite consists on interpreting more than 1000 little > programs. For every program, the interpreter is executed. > > The terminal is fast (it shows all text on 0.2 seconds, while VMWare > needs 2.5 seconds). I'm using the local volume (i.e. no Samba > involved). There is no shortage of RAM, neither on andLinux (uses 192 MB > and `top' shows several dozens free; the interpreter just needs about 4 > MB for each run) nor on the host Machie (1GB total, 400 MB free). > > Is this performance normal? The result of the current state of coLinux? > Or an andLinux specific problem? > Can I download this test somethere? What is the local volume? A image file somethere on disk, or a true partition? What version of coLinux is there included? (uname -a) coLinux 0.8.0 has some improvemnets on block device with parameter "setcobd=async". It's available from http://www.colinux.org/snapshots/ -- Henry N. |
|
From: <Use...@zo...> - 2008-03-21 21:37:56
|
of...@wa...(?scar Fuentes) 21.03.08 21:27 >Hi! >I've just installed andLinux Beta 1 rc6 (02/12/2008) (minimal / XFCE >version) on WinXP SP2. >I'm really impressed with andLinux (which implies coLinux!). >However, I'm dissappointed about its performance: running a test suite >on the host machine, VMWare Workstation 6.0 (Guest: Debian Etch) and >andLinux takes this time: >Host machine: 49 seconds >VMWare: 48 seconds >andLinux: 108 seconds Take outmost care when benchmarking! Times in VMs are not relyable. How does the host "feel" when the bench is running? I assume a 100% load under VMware, host rendered unusable? while colinux is "coopperative"? >The test suite consists on interpreting more than 1000 little >programs. For every program, the interpreter is executed. >The terminal is fast (it shows all text on 0.2 seconds, while VMWare >needs 2.5 seconds). I'm using the local volume (i.e. no Samba >involved). >There is no shortage of RAM, neither on andLinux (uses 192 >MB and `top' shows several dozens free; the interpreter just needs >about 4 MB for each run) nor on the host Machie (1GB total, 400 MB >free). >Is this performance normal? AFAIK VMware is "optimzed" to win such bench marks. VMware looses on file and netaccess, aka: real world ;-( But comparing VMware with colinux is unfair, anyway. >The result of the current state of >coLinux? Or an andLinux specific problem? I found colinux pretty fast enough. Maybe you can post a link to your benchmark programm? Rainer |
|
From: Óscar F. <of...@wa...> - 2008-03-21 20:30:14
|
Hi! I've just installed andLinux Beta 1 rc6 (02/12/2008) (minimal / XFCE version) on WinXP SP2. I'm really impressed with andLinux (which implies coLinux!). However, I'm dissappointed about its performance: running a test suite on the host machine, VMWare Workstation 6.0 (Guest: Debian Etch) and andLinux takes this time: Host machine: 49 seconds VMWare: 48 seconds andLinux: 108 seconds The test suite consists on interpreting more than 1000 little programs. For every program, the interpreter is executed. The terminal is fast (it shows all text on 0.2 seconds, while VMWare needs 2.5 seconds). I'm using the local volume (i.e. no Samba involved). There is no shortage of RAM, neither on andLinux (uses 192 MB and `top' shows several dozens free; the interpreter just needs about 4 MB for each run) nor on the host Machie (1GB total, 400 MB free). Is this performance normal? The result of the current state of coLinux? Or an andLinux specific problem? -- Oscar |
|
From: coLinux a. <col...@he...> - 2008-03-21 05:07:38
|
The autobuild system has detected a new revision in the source repository. Review last changed from changelog.txt, also attached in mail. Download the compiled version: http://www.henrynestler.com/colinux/autobuild/devel-20080320/ colinux-0.8.0-20080320.src.tgz (596775 Bytes) daemons-0.8.0-20080320.dbg.zip (1314580 Bytes) daemons-0.8.0-20080320.zip (1020574 Bytes) Note, the autobuild compilation does not include an installer. Remember to reload the driver with these commands: colinux-daemon.exe --remove-driver colinux-daemon.exe --install-driver The vmlinux and modules are up to date. Please use last version from http://www.henrynestler.com/colinux/autobuild/devel-20080317/ The autobuild compilations are not official releases of Cooperative Linux software. There is no warranty that any autobuild version is stable. If use this autobuild version, please give us feedback of your experience. Job runs on machine with 64 bit version of gcc 4.1.2. A service from http://gcc.gnu.org/wiki/CompileFarm -- Lots of fun with newest version, Henry Nestler ------------------------------------------------------------------------ r959 | henryn | 2008-03-20 23:21:20 +0000 (Thu, 20 Mar 2008) | 5 lines Changed paths: M /branches/devel/src/colinux/os/winnt/kernel/alloc.c M /branches/devel/src/colinux/os/winnt/kernel/manager.c M /branches/devel/src/colinux/os/winnt/kernel/manager.h * Don't use MmGetPhysicalMemoryRanges, don't search memory ranges. Lower PFN limit at 16MB. Upper PFN limit from hostmem_pages, to not get pages from behind memory holes. That is enought to fix problems with shared video cards and the bug "PFN to high". Reverts mostly of r937. ------------------------------------------------------------------------ |
|
From: coLinux a. <col...@he...> - 2008-03-18 05:12:16
|
The autobuild system has detected a new revision in the source repository. Review last changed from changelog.txt, also attached in mail. Download the compiled version: http://www.henrynestler.com/colinux/autobuild/devel-20080317/ colinux-0.8.0-20080317.src.tgz (597358 Bytes) daemons-0.8.0-20080317.dbg.zip (1315058 Bytes) daemons-0.8.0-20080317.zip (1021021 Bytes) modules-2.6.22.18-co-0.8.0-20080317.tgz (2542530 Bytes) vmlinux-2.6.22.18-co-0.8.0-20080317.zip (1761947 Bytes) Note, the autobuild compilation does not include an installer. Remember to reload the driver with these commands: colinux-daemon.exe --remove-driver colinux-daemon.exe --install-driver Inside coLinux please update modules as follow: rm -rf /lib/modules/*-co-* tar -xzf modules-*-co-*-20080317.tgz -C / The autobuild compilations are not official releases of Cooperative Linux software. There is no warranty that any autobuild version is stable. If use this autobuild version, please give us feedback of your experience. Job runs on machine with 64 bit version of gcc 4.1.2. A service from http://gcc.gnu.org/wiki/CompileFarm -- Lots of fun with newest version, Henry Nestler ------------------------------------------------------------------------ r957 | henryn | 2008-03-17 23:51:37 +0000 (Mon, 17 Mar 2008) | 1 line Changed paths: M /branches/devel/patch/coconsole-core.diff * Fix BSOD from "stty rows N" where n>25. Disable screen resize. (found by Paolo Minazzi) ------------------------------------------------------------------------ |
|
From: Henry N. <Hen...@Ar...> - 2008-03-17 20:39:38
|
Rainer Zocholl wrote: > Cooperative Linux Daemon, 0.8.0 > Daemon compiled on Fri Mar 14 05:26:03 2008 > > [...] > error initializing > daemon: exit code 83278801 > daemon: error - CO_RC_ERROR_ERROR, line 482, file src/colinux/user/daemon.c (25) Please be very carefully with *autobuild* version! There dosn't exist a proper API versions check. The API was changed last steps, but version was not increment. That means: Please remove and install driver any time you used a new build. Also, if the --status-driver give "no version conflicts"! -- Henry N. |
|
From: H. N. <hen...@ar...> - 2008-03-17 19:27:33
|
----- Original Nachricht ---- Von: Use...@zo... (Rainer Zocholl) An: col...@li... Datum: 17.03.2008 00:08 Betreff: Re: [coLinux-devel] autobuild org.colinux.devel 20080313 23:30:14 > col...@he...(coLinux autobuild) 14.03.08 06:28 > > >The autobuild system has detected a new revision in the source > >repository. Review last changed from changelog.txt, also attached in > >mail. > > >Download the compiled version: > >http://www.henrynestler.com/colinux/autobuild/devel-20080313/ > > > colinux-0.8.0-20080313.src.tgz (597301 Bytes) > > daemons-0.8.0-20080313.dbg.zip (1315063 Bytes) > > daemons-0.8.0-20080313.zip (1021023 Bytes) > > modules-2.6.22.18-co-0.8.0-20080313.tgz (2542543 Bytes) > > vmlinux-2.6.22.18-co-0.8.0-20080313.zip (1762056 Bytes) > > modules-2.6.22.18-co-0.8.0-20080313.tgz > seems to be left unzipped? (size 6.5MB?) A view from web site, I can see 2.4MB "modules-2.6.22.18-co-0.8.0-20080313.tgz 14-Mar-2008 14:27 2.4M Linux modules" Perhaps your browser has automaticly unzipped the file? 6871040 is the size of 'tar' without 'gz'. Try right mouse key and say "save link/file as...", or just use the tar as it is. -- Henry Jetzt komfortabel bei Arcor-Digital TV einsteigen: Mehr Happy Ends, mehr Herzschmerz, mehr Fernsehen! Erleben Sie 50 digitale TV Programme und optional 60 Pay TV Sender, einen elektronischen Programmführer mit Movie Star Bewertungen von TV Movie. Außerdem, aktuelle Filmhits und prickelnde Erotik in der Arcor-Videothek. Infos unter www.arcor.de/tv |
|
From: H. N. <hen...@ar...> - 2008-03-17 18:29:16
|
Hello Rainer,
if I'm right, you can not ask for "==" in Windows batch. Errorlevel is not a variable.
The errorelevel can only ask for a level equal or grater, that is implizit.
I native CMD.EXE you can do such
colinux-daemon --status-driver
if errorlevel 14 goto error14
if errorlevel 3 goto error3
if errorlevel 1 goto error1
goto noproblem
:error14
echo error 14 or grater detected
goto exit
:error3
echo error 3 to 13 detected
goto exit
:error1
echo error 1 or 2 detected
goto exit
:noproblem
...
Please read your file NEW (verion 0.7.1):
- Return errorlevel procedures (return from running, --install-driver,
--status-driver, ...) Errorcodes: 0=OK, 3=driver version mismatch,
14=driver no installed, 255=general fail (-1)
The errors are level coded (no binary mask). Only listed codes will be returned.
--
Henry
Rainer Zocholl wrote:
> Hello
>
> .\colinux-daemon.exe -v 3 --status-driver
> do not deliver an errorlevel anymore,
> if the loaded driver is not working.
>
> The batch sequence shown below worked flawlessly for some time:
> If the driver was wrong, the batch unloads the old and
> installs the current.
> But now error level is always 0,
> but conlinux now dies with:
>
> C:\coLinux>.\colinux-daemon.exe -t NT @colinux.conf -v 3
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> using 'vmlinux' as kernel image
> configuring 128 MB of virtual RAM
> mapping cobd0 to \??\C:\colinux..
> mapping cobd1 to \??\C:\colinux...
> mapping cobd4 to \Device\Cdrom0
> mapping hda to cobd0
> mapping hdb to cobd1
> mapping hdc to cobd4
> configured PCAP bridge at 'LAN-ADM851X' device as eth0
> MAC address: 0A:C0:71:65:08:00
> mapping cofs0 to \??\C:\colinux....
> using 'initrd.gz' as initrd image
> mapping cofs31 to \??\C:\coLinux
> mapping ttys0 to COM2
> kernel boot parameters: 'root=/dev/cobd0'
> error initializing
> daemon: exit code 83278801
> daemon: error - CO_RC_ERROR_ERROR, line 482, file src/colinux/user/daemon.c
> (25)
>
> snip-------------------->
>
> set COLINUXEXE=.\colinux-daemon.exe
> :testdriver
> echo on
> %COLINUXEXE% --status-driver
> if %ERRORLEVEL% == 0 goto noproblem
> installnewdriver:
> echo ERROR: Wrong driver!
> net stop colinux
> %COLINUXEXE% --remove-service
> %COLINUXEXE% --remove-service colinux
> %COLINUXEXE% --remove-driver
> %COLINUXEXE% --install-driver
> %COLINUXEXE% --status-driver
> if %ERRORLEVEL% == 0 goto noproblem
> pause
> goto testdriver
>
> :noproblem
>
> <-------------------snap
>
>
> Maybe 3 states, binary combinable?
>
> a) 0 All is well
> 1 ?
> 2 ?
> 4 ?
> 8 ?
> b) 16 Old driver
> c) 32 Wrong linux API
> d) 64 Wrong periphery API
> e) 128 ?
>
>
>
>
>
> C:\coLinux>.\colinux-daemon.exe -v 3 --status-driver
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> checking if the driver is installed
> current state: 4 (fully initialized)
> current number of monitors: 0
> current linux api version: 11
> current periphery api version: 21
> Driver compiled on: Sun Feb 24 18:39:07 2008
>
> C:\coLinux>echo %ERRORLEVEL%
> 0
>
> C:\coLinux>net stop colinux
> Systemfehler 1060 aufgetreten.
>
> Der angegebene Dienst ist kein installierter Dienst.
>
>
> C:\coLinux>.\colinux-daemon.exe --remove-service
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> daemon: removing service 'Cooperative Linux'
> daemon: failed to remove service. OpenService failed: Der angegebene Dienst
> ist
> kein installierter Dienst.
>
> C:\coLinux>.\colinux-daemon.exe --remove-service colinux
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> daemon: removing service 'colinux'
> daemon: failed to remove service. OpenService failed: Der angegebene Dienst
> ist
> kein installierter Dienst.
>
> C:\coLinux>.\colinux-daemon.exe --remove-driver
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
>
> C:\coLinux>.\colinux-daemon.exe --install-driver
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> loading C:\coLinux\linux.sys
> daemon: driver installed
>
> C:\coLinux>.\colinux-daemon.exe --status-driver
> Cooperative Linux Daemon, 0.8.0
> Daemon compiled on Fri Mar 14 05:26:03 2008
>
> checking if the driver is installed
> current state: 4 (fully initialized)
> current number of monitors: 0
> current linux api version: 11
> current periphery api version: 21
> Driver compiled on: Fri Mar 14 05:26:03 2008
>
>
> Now it runs!
>
> Rainer
Jetzt komfortabel bei Arcor-Digital TV einsteigen: Mehr Happy Ends, mehr Herzschmerz, mehr Fernsehen! Erleben Sie 50 digitale TV Programme und optional 60 Pay TV Sender, einen elektronischen Programmführer mit Movie Star Bewertungen von TV Movie. Außerdem, aktuelle Filmhits und prickelnde Erotik in der Arcor-Videothek. Infos unter www.arcor.de/tv
|
|
From: <Use...@zo...> - 2008-03-17 15:09:33
|
Hello
.\colinux-daemon.exe -v 3 --status-driver
do not deliver an errorlevel anymore,
if the loaded driver is not working.
The batch sequence shown below worked flawlessly for some time:
If the driver was wrong, the batch unloads the old and
installs the current.
But now error level is always 0,
but conlinux now dies with:
C:\coLinux>.\colinux-daemon.exe -t NT @colinux.conf -v 3
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
using 'vmlinux' as kernel image
configuring 128 MB of virtual RAM
mapping cobd0 to \??\C:\colinux..
mapping cobd1 to \??\C:\colinux...
mapping cobd4 to \Device\Cdrom0
mapping hda to cobd0
mapping hdb to cobd1
mapping hdc to cobd4
configured PCAP bridge at 'LAN-ADM851X' device as eth0
MAC address: 0A:C0:71:65:08:00
mapping cofs0 to \??\C:\colinux....
using 'initrd.gz' as initrd image
mapping cofs31 to \??\C:\coLinux
mapping ttys0 to COM2
kernel boot parameters: 'root=/dev/cobd0'
error initializing
daemon: exit code 83278801
daemon: error - CO_RC_ERROR_ERROR, line 482, file src/colinux/user/daemon.c (25)
snip-------------------->
set COLINUXEXE=.\colinux-daemon.exe
:testdriver
echo on
%COLINUXEXE% --status-driver
if %ERRORLEVEL% == 0 goto noproblem
installnewdriver:
echo ERROR: Wrong driver!
net stop colinux
%COLINUXEXE% --remove-service
%COLINUXEXE% --remove-service colinux
%COLINUXEXE% --remove-driver
%COLINUXEXE% --install-driver
%COLINUXEXE% --status-driver
if %ERRORLEVEL% == 0 goto noproblem
pause
goto testdriver
:noproblem
<-------------------snap
Maybe 3 states, binary combinable?
a) 0 All is well
1 ?
2 ?
4 ?
8 ?
b) 16 Old driver
c) 32 Wrong linux API
d) 64 Wrong periphery API
e) 128 ?
C:\coLinux>.\colinux-daemon.exe -v 3 --status-driver
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
checking if the driver is installed
current state: 4 (fully initialized)
current number of monitors: 0
current linux api version: 11
current periphery api version: 21
Driver compiled on: Sun Feb 24 18:39:07 2008
C:\coLinux>echo %ERRORLEVEL%
0
C:\coLinux>net stop colinux
Systemfehler 1060 aufgetreten.
Der angegebene Dienst ist kein installierter Dienst.
C:\coLinux>.\colinux-daemon.exe --remove-service
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
daemon: removing service 'Cooperative Linux'
daemon: failed to remove service. OpenService failed: Der angegebene Dienst
ist
kein installierter Dienst.
C:\coLinux>.\colinux-daemon.exe --remove-service colinux
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
daemon: removing service 'colinux'
daemon: failed to remove service. OpenService failed: Der angegebene Dienst
ist
kein installierter Dienst.
C:\coLinux>.\colinux-daemon.exe --remove-driver
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
C:\coLinux>.\colinux-daemon.exe --install-driver
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
loading C:\coLinux\linux.sys
daemon: driver installed
C:\coLinux>.\colinux-daemon.exe --status-driver
Cooperative Linux Daemon, 0.8.0
Daemon compiled on Fri Mar 14 05:26:03 2008
checking if the driver is installed
current state: 4 (fully initialized)
current number of monitors: 0
current linux api version: 11
current periphery api version: 21
Driver compiled on: Fri Mar 14 05:26:03 2008
Now it runs!
Rainer
|
|
From: <Use...@zo...> - 2008-03-16 23:10:01
|
col...@he...(coLinux autobuild) 14.03.08 06:28 >The autobuild system has detected a new revision in the source >repository. Review last changed from changelog.txt, also attached in >mail. >Download the compiled version: >http://www.henrynestler.com/colinux/autobuild/devel-20080313/ > colinux-0.8.0-20080313.src.tgz (597301 Bytes) > daemons-0.8.0-20080313.dbg.zip (1315063 Bytes) > daemons-0.8.0-20080313.zip (1021023 Bytes) > modules-2.6.22.18-co-0.8.0-20080313.tgz (2542543 Bytes) > vmlinux-2.6.22.18-co-0.8.0-20080313.zip (1762056 Bytes) modules-2.6.22.18-co-0.8.0-20080313.tgz seems to be left unzipped? (size 6.5MB?) Rainer |
|
From: coLinux a. <col...@he...> - 2008-03-14 05:47:30
|
The autobuild system has detected a new revision in the source repository. Review last changed from changelog.txt, also attached in mail. Download the compiled version: http://www.henrynestler.com/colinux/autobuild/devel-20080313/ colinux-0.8.0-20080313.src.tgz (597301 Bytes) daemons-0.8.0-20080313.dbg.zip (1315063 Bytes) daemons-0.8.0-20080313.zip (1021023 Bytes) modules-2.6.22.18-co-0.8.0-20080313.tgz (2542543 Bytes) vmlinux-2.6.22.18-co-0.8.0-20080313.zip (1762056 Bytes) Note, the autobuild compilation does not include an installer. Remember to reload the driver with these commands: colinux-daemon.exe --remove-driver colinux-daemon.exe --install-driver Inside coLinux please update modules as follow: rm -rf /lib/modules/*-co-* tar -xzf modules-*-co-*-20080313.tgz -C / The autobuild compilations are not official releases of Cooperative Linux software. There is no warranty that any autobuild version is stable. If use this autobuild version, please give us feedback of your experience. Job runs on machine with 64 bit version of gcc 4.1.2. A service from http://gcc.gnu.org/wiki/CompileFarm -- Lots of fun with newest version, Henry Nestler ------------------------------------------------------------------------ r956 | henryn | 2008-03-13 23:30:14 +0000 (Thu, 13 Mar 2008) | 33 lines Changed paths: M /branches/devel/bin/build-colinux-libs.sh M /branches/devel/bin/build-common.sh M /branches/devel/bin/comake/tools.py A /branches/devel/conf/wxWidgets.conf (from /branches/steve/conf/wxWidgets.conf:954) M /branches/devel/patch/base-2.6.22.diff M /branches/devel/patch/cobd-async-2.6.22.diff M /branches/devel/patch/conet-2.6.22.diff M /branches/devel/patch/conet-pci-core.diff M /branches/devel/patch/pci-core.diff M /branches/devel/patch/scsi-core.diff M /branches/devel/patch/video-core.diff M /branches/devel/src/colinux/common/config.h M /branches/devel/src/colinux/common/ioctl.h M /branches/devel/src/colinux/common/libc.c M /branches/devel/src/colinux/common/libc.h M /branches/devel/src/colinux/kernel/monitor.c M /branches/devel/src/colinux/kernel/pci.c A /branches/devel/src/colinux/kernel/pci.h (from /branches/steve/src/colinux/kernel/pci.h:954) M /branches/devel/src/colinux/kernel/scsi.c M /branches/devel/src/colinux/kernel/scsi.h A /branches/devel/src/colinux/kernel/video.c (from /branches/steve/src/colinux/kernel/video.c:954) M /branches/devel/src/colinux/kernel/video.h M /branches/devel/src/colinux/os/winnt/build/build.comake.py M /branches/devel/src/colinux/os/winnt/kernel/scsi.c M /branches/devel/src/colinux/user/config.c A /branches/devel/src/colinux/user/console-wx (from /branches/steve/src/colinux/user/console-wx:954) R /branches/devel/src/colinux/user/console-wx/build.comake.py (from /branches/steve/src/colinux/user/console-wx/build.comake.py:954) R /branches/devel/src/colinux/user/console-wx/console.h (from /branches/steve/src/colinux/user/console-wx/console.h:954) R /branches/devel/src/colinux/user/console-wx/main.cpp (from /branches/steve/src/colinux/user/console-wx/main.cpp:954) * By Steve Shoecraft scsi: rom updates ... work in progress. changed driver to use CO_DEVICE_SCSI instead of CO_OPERATION_SCSI host driver will create the storage device file if it does not exist. a size can be specified for a scsi device file in the configuration file. size is in megabytes. format is: scsi#=type,<path>,size. BE VERY CAREFULL with this parameter, as it can increase/decrease the size of a file as well! network: added ethtool and mii-tool support video: added video (framebuffer) support. use video#=<size> in config file to enable. size is in megabytes. added wxWidgets library to build system. added a new console, colinux-console-wx. work in progress. pci: enabled dynamic pci configuration, guest pci config comes from the host. user can specify which slot a virtual adapter is in using pci<slot #>=<function #>,<device type>,<device unit #>. example: if user wants video0 at slot 5, function 0: pci5=0,video,0 * By Henry Nestler Build tools: Version cache files removed from md5sum, check version is enough. * Merged from branch steve SVN r954. ------------------------------------------------------------------------ |
|
From: <Use...@zo...> - 2008-03-13 22:52:09
|
Can colinux be used to make span VPNs with IPsec? If, how? Any link? Rainer |
81 messages has been excluded from this view by a project administrator.
