Menu
▾
▴
user-mode-linux-user
You can subscribe to this list here.
| 1999 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(15) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2000 |
Jan
(6) |
Feb
(1) |
Mar
(39) |
Apr
(13) |
May
(24) |
Jun
(11) |
Jul
(23) |
Aug
(85) |
Sep
(12) |
Oct
(103) |
Nov
(79) |
Dec
(112) |
| 2001 |
Jan
(52) |
Feb
(82) |
Mar
(84) |
Apr
(65) |
May
(105) |
Jun
(188) |
Jul
(174) |
Aug
(182) |
Sep
(103) |
Oct
(137) |
Nov
(143) |
Dec
(98) |
| 2002 |
Jan
(258) |
Feb
(236) |
Mar
(386) |
Apr
(307) |
May
(238) |
Jun
(170) |
Jul
(252) |
Aug
(230) |
Sep
(278) |
Oct
(394) |
Nov
(336) |
Dec
(194) |
| 2003 |
Jan
(290) |
Feb
(182) |
Mar
(175) |
Apr
(220) |
May
(209) |
Jun
(286) |
Jul
(279) |
Aug
(164) |
Sep
(208) |
Oct
(324) |
Nov
(204) |
Dec
(380) |
| 2004 |
Jan
(344) |
Feb
(332) |
Mar
(395) |
Apr
(357) |
May
(349) |
Jun
(352) |
Jul
(279) |
Aug
(269) |
Sep
(374) |
Oct
(442) |
Nov
(428) |
Dec
(253) |
| 2005 |
Jan
(225) |
Feb
(219) |
Mar
(245) |
Apr
(249) |
May
(203) |
Jun
(157) |
Jul
(171) |
Aug
(194) |
Sep
(200) |
Oct
(232) |
Nov
(190) |
Dec
(195) |
| 2006 |
Jan
(158) |
Feb
(190) |
Mar
(235) |
Apr
(161) |
May
(134) |
Jun
(169) |
Jul
(117) |
Aug
(161) |
Sep
(170) |
Oct
(297) |
Nov
(230) |
Dec
(205) |
| 2007 |
Jan
(197) |
Feb
(132) |
Mar
(151) |
Apr
(97) |
May
(109) |
Jun
(99) |
Jul
(57) |
Aug
(110) |
Sep
(56) |
Oct
(119) |
Nov
(39) |
Dec
(45) |
| 2008 |
Jan
(101) |
Feb
(116) |
Mar
(141) |
Apr
(98) |
May
(133) |
Jun
(61) |
Jul
(43) |
Aug
(76) |
Sep
(20) |
Oct
(32) |
Nov
(22) |
Dec
(41) |
| 2009 |
Jan
(35) |
Feb
(15) |
Mar
(18) |
Apr
(13) |
May
(13) |
Jun
(26) |
Jul
(12) |
Aug
(32) |
Sep
(21) |
Oct
(41) |
Nov
(35) |
Dec
(12) |
| 2010 |
Jan
(3) |
Feb
(35) |
Mar
(28) |
Apr
(20) |
May
(5) |
Jun
(14) |
Jul
(6) |
Aug
(8) |
Sep
(20) |
Oct
(20) |
Nov
(10) |
Dec
(12) |
| 2011 |
Jan
(14) |
Feb
(10) |
Mar
(14) |
Apr
(14) |
May
(13) |
Jun
(43) |
Jul
(13) |
Aug
(50) |
Sep
(30) |
Oct
(23) |
Nov
(15) |
Dec
(49) |
| 2012 |
Jan
(15) |
Feb
(28) |
Mar
(7) |
Apr
|
May
(12) |
Jun
(13) |
Jul
(28) |
Aug
(11) |
Sep
(19) |
Oct
(27) |
Nov
(5) |
Dec
(25) |
| 2013 |
Jan
(18) |
Feb
(19) |
Mar
(56) |
Apr
(26) |
May
(38) |
Jun
(24) |
Jul
(42) |
Aug
(24) |
Sep
(4) |
Oct
(3) |
Nov
(18) |
Dec
(4) |
| 2014 |
Jan
(10) |
Feb
(9) |
Mar
(3) |
Apr
|
May
(12) |
Jun
(34) |
Jul
(8) |
Aug
(18) |
Sep
(3) |
Oct
(27) |
Nov
(2) |
Dec
(1) |
| 2015 |
Jan
|
Feb
(10) |
Mar
(49) |
Apr
(2) |
May
(4) |
Jun
(7) |
Jul
(1) |
Aug
(17) |
Sep
(7) |
Oct
(35) |
Nov
(40) |
Dec
(4) |
| 2016 |
Jan
(9) |
Feb
|
Mar
(6) |
Apr
|
May
(10) |
Jun
(2) |
Jul
|
Aug
|
Sep
(5) |
Oct
|
Nov
|
Dec
(1) |
| 2017 |
Jan
(2) |
Feb
(4) |
Mar
(1) |
Apr
(4) |
May
(31) |
Jun
(9) |
Jul
(1) |
Aug
|
Sep
|
Oct
(1) |
Nov
(1) |
Dec
(2) |
| 2018 |
Jan
|
Feb
|
Mar
(1) |
Apr
(4) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2022 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(2) |
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Yuri P. <yu...@sw...> - 2000-04-13 12:28:06
|
Jeff Dike wrote: > Not much is new in this one. I fixed a crash in the arch code. The brown > paper bag bug in -pre4 was fixed in my 0.19-2.3.99-pre4 release yesterday. arch/um/Makefile needs ptrace.o for um.a, but a corresponding ptrace.c file is missing. I'm trying 2..2.99-pre5 + 0.20-patch-2.3.99-pre5. Thanks, Yuri Pudgorodsky |
|
From: Jeff D. <jd...@ka...> - 2000-04-13 02:42:03
|
Not much is new in this one. I fixed a crash in the arch code. The brown paper bag bug in -pre4 was fixed in my 0.19-2.3.99-pre4 release yesterday. I also put a SuSE installation kit up on the project's download page: http://sourceforge.net/project/filelist.php?group_id=429 Jeff |
|
From: William S. <wst...@po...> - 2000-04-13 02:39:32
|
Good evening, Jeff,
Linux version 2.3.99-pre4-1um (jd...@cc...) (gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)) #143 Tue Apr 11 19:50:22 EST 2000
Just _how_ did you chop the kernel size in half? *smile* Lots of
debugging symbols gone?
All of the following with the debian root filesystem.
As non-root:
pty: 256 Unix98 ptys configured
Initializing stdio console driver
Initializing software serial port version 0
Kernel panic: Out of pty's in getmaster
I think this is an old one...
As root:
The following text stair-steps across the screen:
Initializing software serial port version 0
serial line 0 assigned pty
/dev/ptyp0
ssl receive thread is pid 9314
devfs: v0.93 (20000306) Richard Gooch (rg...@at...)
devfs: devfs_debug: 0x0
devfs: boot_options: 0x0
VFS: Mounted root (ext2 filesystem) readonly.
Mounted devfs on /dev
INIT: version 2.77 booting
usermode:~# dd if=/dev/zero of=/dev/null
dd: /dev/zero: Bad address
0+0 records in
0+0 records out
usermode:~# ls -al /dev/zero
crw-rw-rw- 1 root root 1, 5 Jan 1 1970 /dev/zero
usermode:~# ls -al /dev/null
crw-rw-rw- 1 root root 1, 3 Jan 1 1970 /dev/null
Cheers,
- Bill
---------------------------------------------------------------------------
"Be vewy vewy quiet...I'm hunting Jedi."
-- Darth Fudd
(Courtesy of Peter da Silva <pe...@ba...>)
--------------------------------------------------------------------------
William Stearns (wst...@po...). Mason, Buildkernel, named2hosts,
and ipfwadm2ipchains are at: http://www.pobox.com/~wstearns
LinuxMonth; articles for Linux Enthusiasts! http://www.linuxmonth.com
--------------------------------------------------------------------------
|
|
From: Alan C. <al...@lx...> - 2000-04-13 02:04:53
|
> Are there guidelines anywhere about how best to submit a largish body of code > so it doesn't get dinged for non-functional reasons? One thing that I know > about is changing my 2-char indents to 8-char ones. Are there others? Format it up nicely so it follows the coding style. Read over it and look at all the things that make you think uggh. Then seperate out the changes to the mainstream code and look very hard at them and see if you can sanely avoid them or if they are pointers to generic code that needs fixing or are just plain sensible to add. Then send me a set of diffs |
|
From: Jeff D. <jd...@ka...> - 2000-04-13 00:47:11
|
al...@lx... said: > I'd like to see UML in. Its security value for isolating virtual machines is > fascinating Great! Are there guidelines anywhere about how best to submit a largish body of code so it doesn't get dinged for non-functional reasons? One thing that I know about is changing my 2-char indents to 8-char ones. Are there others? Would some 2.4 releases be better to target than others (i.e. should I try to go for a 2.4pre in preference to 2.4.low)? I'm currently cleaning things up in preparation for 2.4/2.5, so the sooner I know what needs doing, the sooner I can get going on them. Thanks, Jeff |
|
From: Alan C. <al...@lx...> - 2000-04-12 23:49:59
|
> I'm also thinking about seeing about getting it into an early 2.4. If those > IBM people can get the S390 port into 2.2, I think I ought to be allowed to > get um into 2.4... S/390 is a target for early 2.4 or 2.4pre too. I'd like to see UML in. Its security value for isolating virtual machines is fascinating |
|
From: Jeff D. <jd...@ka...> - 2000-04-12 22:46:53
|
> Looks nice... Do you plan to merge it to linus? [In my eyes this is > much more interesting architecure than sh...] (Now you've got the sh fans all mad at you :-) Wherever they are). My plan is to submit it to Linus when he opens up 2.5. I'm also thinking about seeing about getting it into an early 2.4. If those IBM people can get the S390 port into 2.2, I think I ought to be allowed to get um into 2.4... Jeff |
|
From: Pavel M. <pa...@su...> - 2000-04-12 20:54:32
|
Hi! > The 2.3.99-pre4 user-mode port is available. Looks nice... Do you plan to merge it to linus? [In my eyes this is much more interesting architecure than sh...] Pavel -- I'm pa...@uc.... "In my country we have almost anarchy and I don't care." Panos Katsaloulis describing me w.r.t. patents me at di...@li... |
|
From: Jeff D. <jd...@ka...> - 2000-04-12 03:15:20
|
The 2.3.99-pre4 user-mode port is available. The changes include: a.out binaries run the block device shows up in /proc/partitions the console driver sets its window size the network driver frees up skbuffs now all of the remaining system calls are hooked up - the exceptions are things like modify_ldt and vm86, which don't seem to make sense in this kernel various other bug fixes Things that now work include: telnet now works in both directions X clients can display both locally and remotely. In the former case, the local X server is Xnest. Obviously, remote clients can also display on the local Xnest. the SuSE install procedure now runs strace works, gdb doesn't As usual, the goodies can be had from http://sourceforge.net/project/filelist.p hp?group_id=429 A new goody is a Slackware install kit, which contains a README and a slightly modified root disk from which the user-mode kernel will run the Slackware install procedure. You'll need media, obviously. I used Slackware 7.0. Other info can be had from the project home page: http://user-mode-linux.source forge.net Jeff |
|
From: lars b. <la...@no...> - 2000-03-30 09:10:36
|
Jeff Dike <jd...@ka...> writes: > >From Alan's 2.2.15pre16 fix list: > o Allow syscall interception in ptrace (Jeff Dike) > This is very cool. Distros that include 2.2.15 or later will be able to run > the user-mode kernel without patching the kernel. Well done! |
|
From: Jeff D. <jd...@ka...> - 2000-03-29 03:54:45
|
From Alan's 2.2.15pre16 fix list: o Allow syscall interception in ptrace (Jeff Dike) This is very cool. Distros that include 2.2.15 or later will be able to run the user-mode kernel without patching the kernel. I wasn't expecting this until 2.2.16 at the earliest, and I really wasn't expecting anything at all given the null reaction it got on lk. Jeff |
|
From: Jeff D. <jd...@ka...> - 2000-03-29 02:37:51
|
I put a Slackware rootdisk image on the project download page - http://sourceforge.net/project/filelist.php?group_id=429 This is a filesystem image that will boot the kernel into the Slackware installation procedure. You'll need a kernel and Slackware install media to use it. It is somewhat akin to the Red Hat mkrootfs script, except that it's not a horrible kludge. It runs the official install procedure, so you get a genuine installation. Jeff |
|
From: Jeff D. <jd...@ka...> - 2000-03-28 20:56:30
|
The changes include: an update to 2.3.99-pre3 the block driver now falls back from read-write to read-only when it is given a read-only file changed the network device to configure itself correctly more often and give better errors when it doesn't fixed various bugs, including a crash when running it on Mandrake, a hang, and a compilation error It can build kernels (probably including itself, but I haven't checked this), and, even more impressively, run emacs :-) The patch and kernel are available from http://sourceforge.net/project/filelist .php?group_id=429 The project's home page is at http://user-mode-linux.sourceforge.net Jeff |
|
From: lars b. <la...@no...> - 2000-03-21 07:33:16
|
Jeff Dike <jd...@ka...> writes: >Host OS Mhz null null open selct sig sig fork exec sh > call I/O stat clos inst hndl proc proc proc >--------- ------------- --- ---- ---- ---- ---- ----- ---- ---- ---- ---- --- >i686-linu Linux 2.2.5-1 448 0.6 1.0 7 8 0.04K 2.1 3 0.3K 1K 7K >um-linuxo Linux 2.3.99- 448 99.2 168 135 244 0.15K 103 33 9.5K 19K 57K >um-linuxo Linux 2.3.99- 448 46.1 85 80 134 0.09K 49 73 8.7K 17K 48K Getting closer to native speed already! :) |
|
From: Jeff D. <jd...@ka...> - 2000-03-21 03:25:36
|
I finally put together a decent system (Slackware with a, ap, d, and e (base system, non-X apps, development stuff, and emacs)) and gave the kernel a workout. It accomplished two things: a kernel build, and even more impressively it runs emacs :-) Caveat: The successful kernel build was about the fourth attempt. The first failed because it ran out of memory; that problem seemed to disappear after I closed a trivial memory leak. The next two hung for reasons I don't understand. The fourth succeeded. The kernel was a generic i386 kernel, not itself. Trivia of the day: a kernel build executes a little more than a million system calls. One more note: It did a fair amount of fscking, with that fsck panic not showing up, so I'm reasonably hopeful that it's gone. Jeff |
|
From: Jeff D. <jd...@ka...> - 2000-03-20 15:29:39
|
The user-mode port has been updated to 2.3.99-pre2. Other changes include: Redid the system call mechanism to fix a bug that was causing the occasional system call to return a random number, to clean up the code, and to speed it up. I reran lmbench, the results are below. Gave the block driver a little geometry support. Hooked up a couple more system calls. Assorted other bug fixes and cleanups. The patch, kernel, and packages have all been updated and are available from http://sourceforge.net/project/filelist.php?group_id=429 The kernel now has vfat, iso9660, and minix compiled in. The project home page is at http://user-mode-linux.sourceforge.net Jeff here are the lmbench numbers - each group of three kernels is : the native kernel the 2.3.99-pre1 user-mode kernel the 2.3.99-pre2 user-mode kernel L M B E N C H 1 . 9 S U M M A R Y ------------------------------------ (Alpha software, do not distribute) Processor, Processes - times in microseconds - smaller is better ---------------------------------------------------------------- Host OS Mhz null null open selct sig sig fork exec sh call I/O stat clos inst hndl proc proc proc --------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ---- i686-linu Linux 2.2.5-1 448 0.6 1.0 7 8 0.04K 2.1 3 0.3K 1K 7K um-linuxo Linux 2.3.99- 448 99.2 168 135 244 0.15K 103.2 33 9.5K 19K 57K um-linuxo Linux 2.3.99- 448 46.1 85. 80 134 0.09K 49.5 73 8.7K 17K 48K Context switching - times in microseconds - smaller is better ------------------------------------------------------------- Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw --------- ------------- ----- ------ ------ ------ ------ ------- ------- i686-linu Linux 2.2.5-1 1 15 46 15 130 18 166 um-linuxo Linux 2.3.99- 530 557 471 547 um-linuxo Linux 2.3.99- 457 8 14 64 *Local* Communication latencies in microseconds - smaller is better ------------------------------------------------------------------- Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP ctxsw UNIX UDP TCP conn --------- ------------- ----- ----- ---- ----- ----- ----- ----- ---- i686-linu Linux 2.2.5-1 1 7 13 35 61 um-linuxo Linux 2.3.99- 530 1306 1259 um-linuxo Linux 2.3.99- 457 975 1412 File & VM system latencies in microseconds - smaller is better -------------------------------------------------------------- Host OS 0K File 10K File Mmap Prot Page Create Delete Create Delete Latency Fault Fault --------- ------------- ------ ------ ------ ------ ------- ----- ----- i686-linu Linux 2.2.5-1 14 1 26 2 10863 1 0.6K um-linuxo Linux 2.3.99- 52 16 473 43 3578 0.2K um-linuxo Linux 2.3.99- 35 10 304 13 3324 0.2K *Local* Communication bandwidths in MB/s - bigger is better ----------------------------------------------------------- Host OS Pipe AF TCP File Mmap Bcopy Bcopy Mem Mem UNIX reread reread (libc) (hand) read write --------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- ----- i686-linu Linux 2.2.5-1 338 124 -1 170 335 150 145 336 198 um-linuxo Linux 2.3.99- 5 26 -1 129 334 157 151 339 209 um-linuxo Linux 2.3.99- 5 19 -1 147 336 156 151 339 82 Memory latencies in nanoseconds - smaller is better (WARNING - may not be correct, check graphs) --------------------------------------------------- Host OS Mhz L1 $ L2 $ Main mem Guesses --------- ------------- --- ---- ---- -------- ------- i686-linu Linux 2.2.5-1 448 6 49 152 um-linuxo Linux 2.3.99- 448 6 49 152 um-linuxo Linux 2.3.99- 448 6 49 152 |
|
From: Jeff D. <jd...@ka...> - 2000-03-18 05:18:01
|
I've been wanting to get uml to run a distribution install procedure for quite a while. I tried with the RH installation a while ago, and it was pretty cool with each virtual console in an xterm and the main install window working for the first couple of screens. However, it started doing i386 in and out instructions, which caused segfaults which the kernel couldn't deal with. So, I decided to try a Slackware install, on the theory that, since it is the least graphical install I've seen, it might install without trying anything fancy with the hardware. And, after a while, I got it to work. The steps, if anyone is interested, were roughly these: uncompress rootdsks/text.gz (relative to the CD root) make a ext2 file (100 meg isn't enough for the a, ap, and d subsets) in root_slack run fdisk on it and make it a single big partition mount text somewhere mknod dev/fhd0 b 62 0 (and repeat for fhd1 and fhd2) rm dev/hda mknod dev/hda b 62 0 (this is necessary to convince fdisk that there is an installable partition available) build a kernel with minix in it boot it with fhd0=text fhd1=root_slack fhd2r=/dev/cdrom rw devfs=nomount if it hangs on boot, comment the syslogd line in etc/rc, repeat log in mount /dev/fhd1 /mnt mount /dev/fhd2 /cdrom run setup, telling it that you're installing from a mounted directory which is /cdrom/slakware it should go off and happily install whatever you want I got it to fill up my 100 meg file midway through the d install. After making an /etc/fstab, fiddling /etc/inittab and /etc/securetty to tell them about devfs, and copying /bin/true over /sbin/clock, I got a bootable Slackware install. The current kernel won't boot for you because there are some system calls that I had to hook up, but the next one will. The reason I'm so happy about this is that getting distro install procedures to run is a very handy shortcut to creating bootable filesystems. Jeff |
|
From: Jeff D. <jd...@ka...> - 2000-03-17 17:30:02
|
The user-mode port has been updated to 2.3.99-pre1. Other changes include: swapping works again process cleanup happens sooner than before Downloads are available from http://sourceforge.net/project/filelist.php?group_ id=429 The project home page is at http://user-mode-linux.sourceforge.net Jeff |
|
From: Jeff D. <jd...@ka...> - 2000-03-17 02:57:14
|
Well, I'm not so much interested in performance right now as fixing crashes,
so lmbench decides to run right through and give me a complete set of numbers.
One possibility is that the difference is that I'm running .99, Lars was
running .51, and someone fixed a bug somewhere else in the kernel (heh)...
They are pretty similar to Lars' results, with a big difference here and
there. I wasn't trying to give lmbench a completely clean environment, so
some numbers may be perturbed by whatever else I was doing at the time.
Below are two pairs of results for each measurement; the first is mine, for
comparison, the second is what Lars reported. The labels are the same as for
Lars:
> "i686-linu" is Linux 2.2.5 running on hardware, and "um-linuxo" is
> user-mode Linux 2.3.51 on the same machine.
BTW, what are the *K numbers, like "fork proc", "exec proc", and "sh proc" in
the first set? 1000, so that .3K is really 300? Why not just say 300?
Jeff
L M B E N C H 1 . 9 S U M M A R Y
------------------------------------
(Alpha software, do not distribute)
Processor, Processes - times in microseconds - smaller is better
----------------------------------------------------------------
Host OS Mhz null null open selct sig sig fork exec sh
call I/O stat clos inst hndl proc proc proc
--------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ----
i686-linu Linux 2.2.5-1 448 0.6 1.0 7 8 0.04K 2.1 3 0.3K 1K 7K
um-linuxo Linux 2.3.99- 448 99.2 168 135 244 0.15K 103.2 33 9.5K 19K 57K
i686-linu Linux 2.2.5-2 499 0.6 0.9 4 5 0.04K 1.9 3 0.3K 1K 6K
um-linuxo Linux 2.3.51- 499 95.1 159 126 230 0.14K 98.5 31 9.2K 19K 56K
Context switching - times in microseconds - smaller is better
-------------------------------------------------------------
Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw
--------- ------------- ----- ------ ------ ------ ------ ------- -------
i686-linu Linux 2.2.5-1 1 15 46 15 130 18 166
um-linuxo Linux 2.3.99- 530 557 471 547
i686-linu Linux 2.2.5-2 1 14 43 15 128 19 157
um-linuxo Linux 2.3.51- ( no results due to crash )
*Local* Communication latencies in microseconds - smaller is better
-------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
i686-linu Linux 2.2.5-1 1 7 13 35 61
um-linuxo Linux 2.3.99- 530 1306 1259
i686-linu Linux 2.2.5-2 1 6 13 ( no results because I
um-linuxo Linux 2.3.51- 988 1136 didn't measure networking )
File & VM system latencies in microseconds - smaller is better
--------------------------------------------------------------
Host OS 0K File 10K File Mmap Prot Page
Create Delete Create Delete Latency Fault Fault
--------- ------------- ------ ------ ------ ------ ------- ----- -----
i686-linu Linux 2.2.5-1 14 1 26 2 10863 1 0.6K
um-linuxo Linux 2.3.99- 52 16 473 43 3578 0.2K
i686-linu Linux 2.2.5-2 14 1 26 2 22126 1 0.6K
um-linuxo Linux 2.3.51- 48 15 60 18 6635 0.2K
*Local* Communication bandwidths in MB/s - bigger is better
-----------------------------------------------------------
Host OS Pipe AF TCP File Mmap Bcopy Bcopy Mem Mem
UNIX reread reread (libc) (hand) read write
--------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- -----
i686-linu Linux 2.2.5-1 338 124 -1 170 335 150 145 336 198
um-linuxo Linux 2.3.99- 5 26 -1 129 334 157 151 339 209
i686-linu Linux 2.2.5-2 356 121 -1 176 354 150 146 355 194
um-linuxo Linux 2.3.51- -1 -1 -1 -1 -1 -1 -1 -1
Memory latencies in nanoseconds - smaller is better
(WARNING - may not be correct, check graphs)
---------------------------------------------------
Host OS Mhz L1 $ L2 $ Main mem Guesses
--------- ------------- --- ---- ---- -------- -------
i686-linu Linux 2.2.5-1 448 6 49 152
um-linuxo Linux 2.3.99- 448 6 49 152
i686-linu Linux 2.2.5-2 499 6 46 151
um-linuxo Linux 2.3.51- 499 ( no results due to crash )
|
|
From: Jeff D. <jd...@ka...> - 2000-03-15 19:51:51
|
wst...@po... said: > ftp://ftp.bitmover.com/lmbench/ I'm grabbing lmbench-2alpha10.tgz. The only problem I have with it is that it's going on two years old. Lars, is that the version you used? Jeff |
|
From: William S. <wst...@po...> - 2000-03-15 15:53:42
|
On Wed, 15 Mar 2000, Jeff Dike wrote: > > The benchmark crashed the kernel before finishing the context switch > > and memory latency measurements. > > Looks like I've got some debugging to do. Where do you get lmbench? ftp://ftp.bitmover.com/lmbench/ Cheers, - Bill --------------------------------------------------------------------------- Q: Will the tcp/ethernet SMP scaling changes be back-ported to 2.2.x? Mingo: yes, all SMP changes in 2.3 will be backported to 2.2 in the next few months, but to not confuse it with 2.2 it will be named '2.4' ;) -- Ingo Molnar <mi...@ch...> -------------------------------------------------------------------------- William Stearns (wst...@po...). Mason, Buildkernel, named2hosts, and ipfwadm2ipchains are at: http://www.pobox.com/~wstearns LinuxMonth; articles for Linux Enthusiasts! http://www.linuxmonth.com -------------------------------------------------------------------------- |
|
From: Jeff D. <jd...@ka...> - 2000-03-15 15:46:58
|
> User-mode Linux seems to be about 100 times slower than Linux on bare > hardware on fast operations. Slower operations like fork() and exec() > does better. I was expecting it to be pretty bad. Especially things like system calls. On bare hardware a system call is not too many instructions. On uml, it's a whole lot of context switches. > Surprisingly, user-mode Linux seems to be do mmap() and page faults > faster than Linux on hardware. I wonder why? Maybe the hosting kernel usually has the data already in memory, so it's just a matter of mapping it in, whereas on bare hardware, the data has to be read in more often from disk? > The benchmark crashed the kernel before finishing the context switch > and memory latency measurements. Looks like I've got some debugging to do. Where do you get lmbench? > When I try to run the context switch benchmark, uml forks off lots of > subprocesses and I get a lot of > fork: Resource temporarily unavailable messages from the > hosting kernel. What's lmbench doing at that point? Is it possible that uml just isn't killing off dead processes quickly enough? Jeff |
|
From: lars b. <la...@no...> - 2000-03-15 12:01:41
|
lars brinkhoff <la...@no...> writes:
> Here are some results. The benchmark crashed the kernel before finishing
> the context switch and memory latency measurements.
When I try to run the context switch benchmark, uml forks off lots
of subprocesses and I get a lot of
fork: Resource temporarily unavailable
messages from the hosting kernel.
|
|
From: lars b. <la...@no...> - 2000-03-15 10:49:21
|
lars brinkhoff <la...@no...> writes: > Processor, Processes - times in microseconds - smaller is better > ---------------------------------------------------------------- > Host OS Mhz null null open selct sig sig fork exec sh > call I/O stat clos inst hndl proc proc proc > --------- ----------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ---- > i686-linu Linux 2.2.5 499 0.6 0.9 4 5 0.04K 1.9 3 0.3K 1K 6K > um-linuxo Linux 2.3.51 499 95.1 159 126 230 0.14K 98.5 31 9.2K 19K 56K User-mode Linux seems to be about 100 times slower than Linux on bare hardware on fast operations. Slower operations like fork() and exec() does better. > File & VM system latencies in microseconds - smaller is better > -------------------------------------------------------------- > Host OS 0K File 10K File Mmap Prot Page > Create Delete Create Delete Latency Fault Fault > --------- ------------- ------ ------ ------ ------ ------- ----- ----- > i686-linu Linux 2.2.5-2 14 1 26 2 22126 1 0.6K > um-linuxo Linux 2.3.51- 48 15 60 18 6635 0.2K Surprisingly, user-mode Linux seems to be do mmap() and page faults faster than Linux on hardware. I wonder why? |
|
From: lars b. <la...@no...> - 2000-03-15 10:41:46
|
Here are some results. The benchmark crashed the kernel before finishing
the context switch and memory latency measurements.
"i686-linu" is Linux 2.2.5 running on hardware, and "um-linuxo" is
user-mode Linux 2.3.51 on the same machine. I have edited the results
slightly for clarity.
L M B E N C H 1 . 9 S U M M A R Y
------------------------------------
(Alpha software, do not distribute)
Processor, Processes - times in microseconds - smaller is better
----------------------------------------------------------------
Host OS Mhz null null open selct sig sig fork exec sh
call I/O stat clos inst hndl proc proc proc
--------- ------------- ---- ---- ---- ---- ---- ----- ---- ---- ---- ---- ----
i686-linu Linux 2.2.5-2 499 0.6 0.9 4 5 0.04K 1.9 3 0.3K 1K 6K
um-linuxo Linux 2.3.51- 499 95.1 159 126 230 0.14K 98.5 31 9.2K 19K 56K
Context switching - times in microseconds - smaller is better
-------------------------------------------------------------
Host OS 2p/0K 2p/16K 2p/64K 8p/16K 8p/64K 16p/16K 16p/64K
ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw ctxsw
--------- ------------- ----- ------ ------ ------ ------ ------- -------
i686-linu Linux 2.2.5-2 1 14 43 15 128 19 157
um-linuxo Linux 2.3.51- ( no results due to crash )
*Local* Communication latencies in microseconds - smaller is better
-------------------------------------------------------------------
Host OS 2p/0K Pipe AF UDP RPC/ TCP RPC/ TCP
ctxsw UNIX UDP TCP conn
--------- ------------- ----- ----- ---- ----- ----- ----- ----- ----
i686-linu Linux 2.2.5-2 1 6 13 ( no results because I
um-linuxo Linux 2.3.51- 988 1136 didn't measure networking )
File & VM system latencies in microseconds - smaller is better
--------------------------------------------------------------
Host OS 0K File 10K File Mmap Prot Page
Create Delete Create Delete Latency Fault Fault
--------- ------------- ------ ------ ------ ------ ------- ----- -----
i686-linu Linux 2.2.5-2 14 1 26 2 22126 1 0.6K
um-linuxo Linux 2.3.51- 48 15 60 18 6635 0.2K
*Local* Communication bandwidths in MB/s - bigger is better
-----------------------------------------------------------
Host OS Pipe AF TCP File Mmap Bcopy Bcopy Mem Mem
UNIX reread reread (libc) (hand) read write
--------- ------------- ---- ---- ---- ------ ------ ------ ------ ---- -----
i686-linu Linux 2.2.5-2 356 121 -1 176 354 150 146 355 194
um-linuxo Linux 2.3.51- -1 -1 -1 -1 -1 -1 -1 -1
( these are completely bogous )
Memory latencies in nanoseconds - smaller is better
(WARNING - may not be correct, check graphs)
---------------------------------------------------
Host OS Mhz L1 $ L2 $ Main mem Guesses
--------- ------------- --- ---- ---- -------- -------
i686-linu Linux 2.2.5-2 499 6 46 151
um-linuxo Linux 2.3.51- 499 ( no results due to crash )
|
525 messages has been excluded from this view by a project administrator.
