Re: [uml-user] what is modified?

[Nix]

On 31 Mar 2008, fang zheng outgrape:

> hi:
>To port linux kernel to linux host,the linux kernel must be modified.
> but what is modified?

It's a new kind of architecture. Look in arch/um and include/asm-um.

-- 
`The rest is a tale of post and counter-post.' --- Ian Rawlings
   describes USENET

-
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Register now and save $200. Hurry, offer ends at 11:59 p.m., 
Monday, April 7! Use priority code J8TLD2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] skas4 on x86-32 -> boom at startup, and fallback failing too

[Nix]

I felt unhappy recently because UML was working perfectly well and I had
no complaints (except for the absence of skas3 on 2.6.25
hosts). Obviously this was unacceptable.

So I tried out skas4, host+guest (the guest also has your rng patches,
time fixes, and SIGWINCH fixes applied).

The result:

,
| Script started on Sun 15 Jun 2008 21:54:41 BST
| loki:/mirror/uml# /usr/bin/nice -n -10 su firewall -c "uml-esperi 
con1=port:163800 con0=fd:0,fd:1 fakeide mem=96M 
ubd0=/mirror/uml/esperi-root-cow.image,/mirror/uuml/esperi-root.image 
umid=esperi eth0=tuntap,tap0,02:60:97:79:e2:c1 eth1=tuntap,,tap1"
| 
| Locating the top of the address space ... 0xc000
| Core dump limits :
|   soft - 102400
|   hard - 102400
| Checking that ptrace can change system call numbers...OK
| Checking syscall emulation patch for ptrace...OK
| Checking advanced syscall emulation patch for ptrace...OK
| Checking for tmpfs mount on /dev/shm...OK
| Checking PROT_EXEC mmap in /dev/shm/...OK
| Checking for SKAS4 support in the host:
|   /proc/self/mm ... OK
|   new_mm ... OK
|   switching over ... switched back ... OK
|   PTRACE_SWITCH_MM ... OK
|   Full CPU fault information in siginfo_t ... OK
|   Full CPU fault information in PTRACE_GETSIGINFO ... OK
|   vcpu ... Host TLS support detected
| Detected host type: i386
|  (GDT indexes 6 to 9)
| vcpu returned with event = 1
| Failed
`

Oops. Now it starts searching for skas3 (which isn't there), and things
go more pear-shaped:

,
| Checking for the skas3 patch in the host:
|   - /proc/mm...not found: No such file or directory
|   - PTRACE_FAULTINFO...
| [1]+  Stopped /usr/bin/nice -n -10 su firewall -c "uml-esperi 
con1=port:16380 con0=fd:0,fd:1 fakeide mem=96M 
ubd0=/mirror/uml/esperi-root-cow.image,/mirror/uml/esperi-root.image 
umid=esperi eth0=tuntap,tap0,02:60:97:79:e2:c1 eth1=tuntap,tap1"
`

fg it, and we see:

,
| loki:/mirror/uml# fg
| /usr/bin/nice -n -10 su firewall -c "uml-esperi con1=port:16380 
con0=fd:0,fd:1 fakeide mem=96M 
ubd0=/mirror/uml/esperi-root-cow.image,/mirror/uml/esperi-root.image 
umid=esperi eth0=tuntap,tap0,02:60:97:79:e2:c1 eth1=tuntap,tap1"
`

[dead halt until we hit it with a SIGKILL, whereupon it dies]

Thus, not only does SKAS4 not work for me, the fallback to SKAS0 doesn't seem 
to be working.

Is the SKAS4 patch supposed to work on x86-32 yet?

(Without the skas4 patch on the guest, everything works fine, SKAS0ing
happily.)

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] skas4 on x86-32 -> boom at startup, and fallback failing too

[Nix]

On 25 Jun 2008, Benedict Verheyen uttered the following:
> I have skas4 working on 2.6.25.4. You need to patch both the guest & 
> host kernel but from you said i conclude that you already knew that.

Yes, they were both patched.

-- 
`If you are having a "ua luea luea le ua le" kind of day, I can only
 assume that you are doing no work due [to] incapacitating nausea caused 
 by numerous lazy demons.' --- Frossie

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Running UML over Vmware

[Nix]

On 7 Jul 2008, Jeff Dike told this:

> On Sat, Jul 05, 2008 at 07:27:44PM +0200, David Fernández wrote:
>> Any hint about the cause of this big performance difference? Should
>> the installation of the SKAS3/4 patch on the host help to solve it?  
>
> skas4 will certainly help.  I get ~86% of native performance on a
> kernel build with the latest skas4 patch.

I'll have to try it again one of these days. I miss skas3's speed,
and skas4 failed horribly last time I tried it...

-
Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
Studies have shown that voting for your favorite open source project,
along with a healthy diet, reduces your potential for chronic lameness
and boredom. Vote Now at http://www.sourceforge.net/community/cca08
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Random Lockups

[Nix]

On 11 Jul 2008, Jeff Dike verbalised:

> On Fri, Jul 11, 2008 at 06:01:32PM +0100, Jay Shah wrote:
>> I'm running a 2.6.20.20 host and 2.6.24.3 guest. The guest will
>> randomly 'freeze' and also stop/go doesn't work at all on any
>> machines. Would upgrading to a new(er) kernel fix this [see bottom for
>> final output]? 
>
> Is the UML running tickless (CONFIG_NOHZ=y)?  Is NTP on the host
> fiddling the time around when the hang happens?  If so, current git
> has these problems fixed.

Since it was an intermittent fault I may as well confirm that I've had
zero problems with my ntp-running guest since your most recent patch.
This is an ex-bug.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] re: peculiar intermittent network seizures with uml-2.4.27-1 on 2.4.28

[Nix]

On Thu, 16 Dec 2004, Antoine Martin spake:
> Looks to me like this is a routing problem, check your arp tables and
> your routes. It may be that something is occasionally causing a packet
> to come out of the wrong interface/wrong mac, which confuses the lower
> network layers.

Alas, I don't think it's that simple :(

My arp tables are simple; virtually everything has the MAC of the ADSL
router.

Every packet going out of the wrong interface gets syslogged, and I see
no such syslog messages, just heaps of NFS timeout messages.

(See <http://www.esperi.demon.co.uk/nix/iptables> for my firewall rules;
a few sensitive external IP addresses have been elided. Look at the log
rules therein.)

ip addr:

1: lo:  mtu 16436 qdisc noqueue 
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 brd 127.255.255.255 scope host lo
2: teql0:  mtu 1500 qdisc noop qlen 100
link/void 
3: dummy0:  mtu 1500 qdisc noop 
link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff
4: gordianet:  mtu 1458 qdisc pfifo_fast qlen 
100
link/ether 00:60:97:79:e2:c1 brd ff:ff:ff:ff:ff:ff
inet 192.168.14.1/24 brd 192.168.14.255 scope global gordianet
5: adsl:  mtu 1458 qdisc cbq qlen 10
link/ether 00:60:97:79:e2:c1 brd ff:ff:ff:ff:ff:ff
inet 194.247.41.52/24 brd 255.255.255.255 scope global adsl
inet 192.168.14.159/32 scope global adsl


ip route:

192.168.14.160 dev adsl  proto static  scope link  src 192.168.14.159 
194.247.41.0/24 dev adsl  proto kernel  scope link  src 194.247.41.52 
192.168.14.0/24 dev gordianet  proto kernel  scope link  src 192.168.14.1 
224.0.0.0/4 dev gordianet  scope link 
default via 194.247.41.52 dev adsl 


(note an arcaneness: the ADSL router is on 192.168.14.160, the same
subnet as all the stuff on the other interface: it's accessed via a
different source address (to avoid being masqueraded).


... hm. Some more info. This is definitely an interface-specific thing.
The blockages appear to affect different interfaces at different times:
i.e., one interface freezes, in both directions --- and it's not just
one destination, it's everyone except for myself. e.g., pinging the
broadcast address on the local net:

64 bytes from 192.168.14.1: icmp_seq=343 ttl=64 time=9.7 ms
64 bytes from 192.168.14.16: icmp_seq=342 ttl=64 time=5010.4 ms (DUP!)
64 bytes from 192.168.14.14: icmp_seq=342 ttl=64 time=5011.0 ms (DUP!)
64 bytes from 192.168.14.18: icmp_seq=342 ttl=64 time=5011.5 ms (DUP!)

I'm starting to suspect the host's TUN/TAP here, y'know: the fact that
it's a simultaneous bidirectional freeze, and that the packet to myself
(automatically sent directly to myself by the kernel) is *not* delayed
is possibly significant.

(But there were no TUN/TAP changes between 2.4.26 and 2.4.28. Damn.
Maybe it's TCP Vegas or something? *clutches at straws in
ChangeLog-2.4.2[78]*)

-- 
`The sword we forged has turned upon us
 Only now, at the end of all things do we see
 The lamp-bearer dies; only the lamp burns on.'


---
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/
___
User-mode-linux-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] peculiar intermittent network seizures with uml-2.4.27-1 on 2.4.28

[Nix]

Ever since upgrading to the 2.4.27-1 UML-applied-to-2.4.28 (from
2.4.24-2-applied-to-2.4.26), I've been seeing decidedly peculiar network 
interaction.

Notably, the UML ceases responding to network packets for up to a minute
and a half at a time, then restarts again just as quietly: the
console-on-network-port freezes at the same time. The UML continues
running and packets get queued on either side of the... blockage, and
are allowed through when it lifts.

Example, pinging every five seconds (this one happened to be a fairly
short blockage):

64 bytes from 192.168.14.1: icmp_seq=57 ttl=64 time=1.1 ms
64 bytes from 192.168.14.1: icmp_seq=58 ttl=64 time=1.3 ms
64 bytes from 192.168.14.1: icmp_seq=59 ttl=64 time=1.3 ms
64 bytes from 192.168.14.1: icmp_seq=60 ttl=64 time=1.3 ms
64 bytes from 192.168.14.1: icmp_seq=61 ttl=64 time=7993.1 ms
64 bytes from 192.168.14.1: icmp_seq=62 ttl=64 time=14206.3 ms
64 bytes from 192.168.14.1: icmp_seq=63 ttl=64 time=9219.6 ms
64 bytes from 192.168.14.1: icmp_seq=64 ttl=64 time=4219.6 ms
64 bytes from 192.168.14.1: icmp_seq=65 ttl=64 time=1.3 ms

The blockages happen quite intermittently, but I'd estimate the mean
time between blockages to be about ten minutes.

The UML has two network interfaces, both bridged tun/tap interfaces: the
one I'm pinging down here, and another, bridged to an ADSL router,
traffic-shaped, and firewalled with iptables.

(This may be an iptables bug: I'll build a UML without iptables and with
only one interface tonight, and see if the problem still occurs.)


Anyone seen anything like this before? Any suggestions as to how I might
debug it?

-- 
`The sword we forged has turned upon us
 Only now, at the end of all things do we see
 The lamp-bearer dies; only the lamp burns on.'


---
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/
___
User-mode-linux-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] re: peculiar intermittent network seizures with uml-2.4.27-1 on 2.4.28

[Nix]

On Fri, 17 Dec 2004, roland whispered secretively:
> maybe this isn`t a networking problem but a problem with the UML process 
> itself ?
> Keep in mind that a UML is a program which runs in userspace - and it`s no 
> realtime
> application. So, under certain circumstances the response of that application 
> probably
> could be slower than normal - e.g. when paging/swapping occurs.

This happens when the box is not paging; e.g. it happened just now:

19400 99768 uml-esperi con=port:16380 con0=fd:0,fd:1 fakeide mem=96M ubd0=
 416   476 [uml-esperi]
19404 99768 uml-esperi con=port:16380 con0=fd:0,fd:1 fakeide mem=96M ubd0=
19404 99768 uml-esperi con=port:16380 con0=fd:0,fd:1 fakeide mem=96M ubd0=
19404 99768 uml-esperi con=port:16380 con0=fd:0,fd:1 fakeide mem=96M ubd0=
1420 99768 uml-esperi con=port:16380 con0=fd:0,fd:1 fakeide mem=96M ubd0=/

and paging activity is effectively nil (a lot is swapped out, sure, but
the firewall's not doing much more than routing packets at this
particular instant).

I tried a wchan: it's not useful: the same stuff is shown when frozen
as when running. (Further evidence that it's not UML alone at fault,
perhaps.)

I think I'll need to gdb it.

gdbing my firewall with a box on the other end expecting DHCP lease
renewals every thirty seconds. Oh joy. ;)

-- 
`The sword we forged has turned upon us
 Only now, at the end of all things do we see
 The lamp-bearer dies; only the lamp burns on.'


---
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/
___
User-mode-linux-user mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML scheduling

[Nix]

On Thu, 24 Feb 2005, Gordon Russell said:
> monitor that by creating the swap with sparse blocks and using du.

Last I tried (in 2.4), sparse swapfiles were a no-no.

-- 
> ...Hires Root Beer...
What we need these days is a stable, fast, anti-aliased root beer
with dynamic shading. Not that you can let just anybody have root.
 --- John M. Ford


---
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] uml and iptables.

[Nix]

On Fri, 1 Apr 2005, Jim Carter spake:
> A maximally paranoid sysop will disable module loading, but this gives only 
> a small benefit in security, because having done a root exploit the hacker 
> can write nefarious code into /dev/kmem, as easily as he can load an 
> inimical module or install a hacked version of a userspace binary.

Not if you remove CAP_RAWIO from the bounding set. :)

-- 
This is like system("/usr/funky/bin/perl -e 'exec sleep 1'");
   --- Peter da Silva


---
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] tls patch

[Nix]

On Tue, 10 May 2005, [EMAIL PROTECTED] mused:
> Exporting in the environment LD_ASSUME_KERNEL=2.4.1 should also work, and is 
> a 
> bit more reliable in theory (because a distro *could* compile /lib/tls to 
> even work on i486, I've heard; since no distro does it,  this remains 
> theory).

Indeed: configuring glibc with --without-__thread, putting the result in
/lib/tls, and linking its dynamic linker into /lib yields an NPTL
installation that works on i586 without trouble.

The only downside is a loss of binary compatibility for programs that
use thread cancellation, but these seem to be very rare.

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] tls patch

[Nix]

On Tue, 10 May 2005, [EMAIL PROTECTED] wrote:
> On Tuesday 10 May 2005 17:23, Nix wrote:
>> Indeed: configuring glibc with --without-__thread, putting the result in
>> /lib/tls, and linking its dynamic linker into /lib yields an NPTL
>> installation that works on i586 without trouble.
> 
>> The only downside is a loss of binary compatibility for programs that
>> use thread cancellation, but these seem to be very rare.
> Thanks for the info, anyhow that's probably enough to refrain many distros 
> from doing it.

Indeed it is; but LinuxThreads isn't long for this world and i586
machines are by no means dead. So I just mentioned it in case anyone
wanted to do it themselves (it doesn't seem to be documented anywhere,
and if you get the options slightly wrong you get peculiar glibc build
failures...)

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

So I'm trying to migrate from 2.4 to 2.6 across the board. I've been at
2.6 everywhere but UML for some time, but this problem is defeating me.

I'm running UML on a GCC-3.4.3-compiled 2.6.11.7-skas3-v8-rc5 host (to
be .9 when next I reboot) running glibc-2.3.5, like this (simplified
from longer line which also fails):

/usr/bin/nice -n -10 su firewall -c "LD_ASSUME_KERNEL=2.4.29 uml-esperi"

Now on a 2.4 UML that works fine (well, obviously *that* line panics
because it doesn't specify a root fs. But the longer line works, and in
2.6 both fail the same way). In 2.6 I see this:

Checking for /proc/mm...found
Checking for the skas3 patch in the host...found
[dead]

A gdb backtrace shows this:

Attaching to program: /usr/packages/linux/i586-esperi/linux, process 6587
Reading symbols from /lib/libutil.so.1...done.
Loaded symbols for /lib/libutil.so.1
Reading symbols from /lib/libc.so.6...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
0x0805df94 in __const_udelay ()
(gdb) bt
#0  0x0805df94 in __const_udelay ()
#1  0x0807284b in panic ()
#2  0x081b9220 in __func__.4 ()
#3  0x0001 in ?? ()
#4  0x08072831 in panic ()
#5  0x000a in ?? ()
#6  0x19bc in ?? ()
#7  0x0806c95f in init_registers ()
#8  0x19bc in ?? ()
#9  0x19bc in ?? ()
#10 0x19bc in ?? ()
#11 0x19bc in ?? ()
#12 0x0805ab01 in can_do_skas ()
#13 0x in ?? ()
#14 0xb2e0 in ?? ()
#15 0x in ?? ()
#16 0x in ?? ()
#17 0x000a in ?? ()
#18 0xb7ee9000 in ?? ()
#19 0x08200cc0 in system_state ()
#20 0x in ?? ()
#21 0xb444 in ?? ()
#22 0x0805edf0 in linux_main ()
#23 0x0001 in ?? ()
#24 0x081cb478 in __initcall_stderr_console_init ()
#25 0x08242050 in ?? ()
#26 0x000a in ?? ()
#27 0x08059159 in main ()

which looks to me like a panic. (Why can't I see it? Where's the panic
message going? All the *other* messages go straight to the console, as
they ought.)

The crashing routine is (probably) init_registers(), which checks for
the presence of PTRACE_GETREGS, PTRACE_GETFPXREGS, and PTRACE_GETFPREGS
on the host --- which certainly look like they're there to me.

How to test if they *work* I have not a clue. Any suggestions?

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Sat, 14 May 2005, [EMAIL PROTECTED] spake:
> On Friday 13 May 2005 23:18, Nix wrote:
>> Checking for /proc/mm...found
>> Checking for the skas3 patch in the host...found
>> [dead]
> add stderr=1 to make it more verbose.

Hm, nothing new appears.

>> The crashing routine is (probably) init_registers(), which checks for
>> the presence of PTRACE_GETREGS, PTRACE_GETFPXREGS, and PTRACE_GETFPREGS
>> on the host --- which certainly look like they're there to me.
> What's your host processor? Can you try upgrading to UML 2.6.11-bs5? I've 
> recently fixed one problem with this checking, which affected a VIA C3 
> Samuel2 (because it didn't support PTRACE_GETFPXREGS). cat /proc/cpuinfo will 
> reveal the presence of the "fxsr" feature; if it's missing upgrade as I said 
> and you'll get the fix.

That's probably it: there's absolutely no chance that this 1997-vintage
Pentium has fxsr :)

flags   : fpu vme de pse tsc msr mce cx8 mmx

It's late here now (mind you, it must be later where you are), but I'll
do a test build tonight and check it in the morning.


(And thank you!)

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Sat, 14 May 2005, [EMAIL PROTECTED] whispered secretively:
> On Saturday 14 May 2005 02:00, Nix wrote:
>> On Sat, 14 May 2005, [EMAIL PROTECTED] spake:
>> > What's your host processor? Can you try upgrading to UML 2.6.11-bs5? I've
>> > recently fixed one problem with this checking, which affected a VIA C3
>> > Samuel2 (because it didn't support PTRACE_GETFPXREGS). cat /proc/cpuinfo
>> > will reveal the presence of the "fxsr" feature; if it's missing upgrade
>> > as I said and you'll get the fix.
> 
>> That's probably it: there's absolutely no chance that this 1997-vintage
>> Pentium has fxsr :)
> Well, so thanks for running GDB, because otherwise I'd never have linked to 
> that problem.

I tend to assume that nobody can get anywhere debugging a problem
without at least a traceback, especially if it comes with as little
attached information as that panic did.

>   How much does it takes to compile a UML kernel? (Have you 

time(1) says

1838.05s user 241.21s system 85% cpu 40:29.98 total

for this kernel (with iptables and traffic shaping adding about ten
minutes to that).

> *only* this Pentium or is it an auxiliary machine?)

It's my pandemonium host (i.e. it runs most of the internal-network-visible
daemons) and my firewall host (running the UMLs that are the firewall ---
obviously those UMLs run very little themselves!)

It's not usually my desktop, but it can run X and has been known to do
desktop duty when my i686 is dead from yet another fan failure. (The
actual desktop programs run on an UltraSPARC in any case.)

>> , but I'll do a test build tonight and check it in the morning.

Hm, well, it failed, because you link against -lutil but don't have any
library search paths set up in vmlinux.lds. Trivial fix, which doesn't
use quite the same (arch-dependent) search path as the usual scripts,
but which certainly should suffice to catch libutil:

--- linux-2.6.11.9-bs5/arch/um/kernel/vmlinux.lds.orig  2005-05-14 
11:54:40.0 +0100
+++ linux-2.6.11.9-bs5/arch/um/kernel/vmlinux.lds   2005-05-14 
11:53:30.0 +0100
@@ -3,6 +3,8 @@
 ENTRY(_start)
 jiffies = jiffies_64;
 
+SEARCH_DIR("/lib"); SEARCH_DIR("/usr/lib"); SEARCH_DIR("/usr/local/lib");
+
 SECTIONS
 {
   PROVIDE (__executable_start = 0x8048000);


With that fix, it works :) --- well, it gets far enough to panic because
my quick test didn't bother with anything as radical as an actual root
filesystem.

Again, thank you!

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Sat, 14 May 2005, [EMAIL PROTECTED] prattled cheerily:
> On Saturday 14 May 2005 13:00, Nix wrote:
>> On Sat, 14 May 2005, [EMAIL PROTECTED] whispered secretively:
>> > On Saturday 14 May 2005 02:00, Nix wrote:
>> >> On Sat, 14 May 2005, [EMAIL PROTECTED] spake:
> 
>> Hm, well, it failed, because you link against -lutil but don't have any
>> library search paths set up in vmlinux.lds.
> Why the hell it works anywhere else? binutils dependant?

Yes, I think so. The implementation of linker search dir stuff has
changed quite a lot recently, and there are no longer any implicitly
searched directories: you've got to specify them all in the
scripts. (IIRC.)

At any rate, specifying the paths explicitly won't do any harm
anywhere. :)

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Sat, 14 May 2005, Jeff Dike announced authoritatively:
> On Fri, May 13, 2005 at 10:18:22PM +0100, Nix wrote:
>> Now on a 2.4 UML that works fine (well, obviously *that* line panics
>> because it doesn't specify a root fs. But the longer line works, and in
>> 2.6 both fail the same way). In 2.6 I see this:
>> 
>> Checking for /proc/mm...found
>> Checking for the skas3 patch in the host...found
>> [dead]
> 
> Since you have it in gdb, 'printf "%s", log_buf' will dump the printk buffer,
> which may be interesting.

I think I know the cause now (lack of fxsr), but if it happens again
I'll certainly do that :)

-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Tue, 17 May 2005, [EMAIL PROTECTED] yowled:
> On Saturday 14 May 2005 14:00, Nix wrote:
>> At any rate, specifying the paths explicitly won't do any harm
>> anywhere. :)
> On 2.6 we just removed -L/usr/lib from the linker command line  because of 
> some warnings.

EXPN? (Are they serious? What are they?)

>So, we could conditionalize that (maybe -D_SEARCH_PATH=). I'm 
> not going to merge it in this form, so send me the cleaned up version.

If I could work out what to conditionalize it on, sure.

We could search only /lib, and skip /usr/lib: I think libutil is there
on every Linux platform, at least. (Do we care about portability to
non-Linux hosts? I don't *think* so...)

> Also, I'd like to ask some questions to you, my dear compiler guy.

I can try to answer, though I'm a tyro compared to the *real* compiler
gods. (Most of these questions seem binutils/glibc-related anyhow.)

> 0) When is RH going to drop LinuxThreads support (as opposed to NPTL)?

Um, ask RH? All I know is that Ulrich Drepper has said that the next
major release of glibc will not have LinuxThreads in it. Daniel
Jacobowitz has indicated a willingness to keep it alive separately.

I don't expect the distributors to drop it for a while: they care about
backward compatibility. (My systems at home have all dropped it, except
for those that run UML: nothing else I run seems to have cared.)

> 1) UML does not compile with TT mode enabled when the host has a NPTL-only 
> glibc installed.

Um, oo-er? (Good thing I've compiled in skas mode.)

>  This showed up on Gentoo, and will show up on newer RedHat 
> since I heard they're dropping LinuxThreads away.

Sooner or later, I expect them to. I just can't say *when*. Only they
can do that. :)

But fundamentally, LinuxThreads isn't being actively maintained by
anyone, and NPTL is, and NPTL is better in just about every way (except
that it places more extreme demands on the kernel and binutils/ABI, so
some platforms don't support it yet). So eventually I expect it to die.

> Now, the problem we're currently having is due to this:
> 
> arch/um/kernel/uml.lds.S:
> 
>   .thread_private : {
> __start_thread_private = .;
> errno = .;
> . += 4;
> arch/um/kernel/tt/unmap_fin.o (.data)
> __end_thread_private = .;
>   }
> 
> That gives a "overflow of program headers, allocated #n needed #m > n" (more 
> or less). Would that be fixable at the linker script level? What it does is 
> to provide a errno definition for that segment of code since it remaps the 
> code while UML is executing.

eee-ick. That'll really collide nastily with glibc's assumption that
errno is located in thread-local storage, I fear.

In fact, glibc's assumptions are even worse. I spent entirely too long
working this stuff out early this year, and because misery loves company
I'll spread the word as best I understand it. (I think only Roland,
Ulrich and possibly a few other deities of that order understand it
fully.)

- inside the dynamic linker is a hidden symbol rtld_errno
  (see glibc/include/errno.h). This is what ld-linux.so.2 uses for errno.
  Nobody else can see it: I mention it only for completeness.

- outside libc, errno expands to a call to the function __errno_location()
  (see e.g. glibc/sysdeps/unix/sysv/linux/bits/errno.h)

- in libc proper, __errno_location returns, um, the address of the errno
  variable. This seems like a waste of time, but it means that the errno
  variable can be *hidden* from outside glibc, and its address can
  change. And it does. (see glibc/sysdeps/generic/errno-loc.c).

- under LinuxThreads, in single-threaded programs, `errno' is a real
  variable, and in multi-threaded programs on non-TLS platforms or those
  not supporting __thread, __errno_location() is overridden by libpthread
  to return a value appropriate to the thread using LinuxThreads's nasty
  and inefficient approximation to TLS, which requires
  (see glibc/linuxthreads/errno.c.)

- Under NPTL, and under LinuxThreads on a __thread, TLS-supporting
  platform (i.e., on glibc compiled for i686 and above, with GCC 3.4+),
  glibc exports a GLIBC_PRIVATE thread_local symbol named errno,
  which __errno_location() picks up. This symbol is exported in the
  GLIBC_2.0 symbol version set if TLS is not in use: if it is, then
  that symbol has no sensible value, and it's not exported at all.

Note that the actual storage of errno under NPTL is a matter of
cooperation between GCC, glibc and GNU ld: glibc declares errno with the
__thread storage class specifier, and GCC proceeds to emit a stereotyped
instruction sequence with suitable relocations (see
gcc-3.4.x/gcc/config/i386/i386.md, search for `Thread-local') which
CPU-specific code in the libbfd library used by GNU ld carefully relaxes

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Tue, 17 May 2005, [EMAIL PROTECTED] suggested tentatively:
> On Tue, 17 May 2005, [EMAIL PROTECTED] yowled:
>>So, we could conditionalize that (maybe -D_SEARCH_PATH=). I'm 
>> not going to merge it in this form, so send me the cleaned up version.
> 
> If I could work out what to conditionalize it on, sure.

If `cleaned up' means `not referring to a generated file', here you are
(whoops: didn't spot that, sorry; I shouldn't work in non-clean
trees...)

(Also refers only to /lib now.)

diff -durN linux-2.6.11.10-bs5-orig/arch/um/kernel/uml.lds.S 
linux-2.6.11.10-bs5/arch/um/kernel/uml.lds.S
--- linux-2.6.11.10-bs5-orig/arch/um/kernel/uml.lds.S   2005-03-28 
13:03:54.0 +0100
+++ linux-2.6.11.10-bs5/arch/um/kernel/uml.lds.S2005-05-17 
07:47:06.0 +0100
@@ -5,6 +5,8 @@
 ENTRY(_start)
 jiffies = jiffies_64;
 
+SEARCH_DIR("/lib");
+
 SECTIONS
 {
   /*This must contain the right address - not quite the default ELF one.*/


-- 
`End users are just test loads for verifying that the system works, kind of
 like resistors in an electrical circuit.' - Kaz Kylheku in c.o.l.d.s


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

[Sorry for the delay: holiday and illness collide.]

On Tue, 17 May 2005, [EMAIL PROTECTED] announced authoritatively:
> On Tuesday 17 May 2005 01:17, Nix wrote:
>> On Tue, 17 May 2005, [EMAIL PROTECTED] yowled:
>> > On Saturday 14 May 2005 14:00, Nix wrote:
>> >> At any rate, specifying the paths explicitly won't do any harm
>> >> anywhere. :)
>> >
>> > On 2.6 we just removed -L/usr/lib from the linker command line  because
>> > of some warnings.
>>
>> EXPN?
> What's EXPN?

`Expand', from SMTP. i.e., `what warnings?' :)

> Also, -lutil is passed to GCC, so what the hell is happening here? And which 
> binutils/GCC versions are you using?

binutils 2.15.94.0.2.2, 2.16.90.0.3; GCC-3.4.3.

> However, here's probably the solution (taken from .../tt/Makefile):
> 
> gcc -print-file-name=libutil.so
> /usr/lib/gcc/i686-pc-linux-gnu/3.4.3-20050110/../../../libutil.so
> (We need to choose libutil.a / libutil.so by hand, but that's not bad).
> 
> I guess that a cross-GCC will make that work well.

Yes, -print-file-name is the right approach, I think.

>> - in libc proper, __errno_location returns, um, the address of the errno
>>   variable. This seems like a waste of time, but it means that the errno
>>   variable can be *hidden* from outside glibc, and its address can
>>   change. And it does. (see glibc/sysdeps/generic/errno-loc.c).
> Yes, it's to make it TLS.

Well, it's to make it TLS without bothering the userspace apps with the
fact that it's TLS, and allowing the implementation of TLS to change under
the covers without changing the ABI.

>> - under LinuxThreads, in single-threaded programs, `errno' is a real
>>   variable, and in multi-threaded programs on non-TLS platforms or those
>>   not supporting __thread, __errno_location() is overridden by libpthread
>>   to return a value appropriate to the thread using LinuxThreads's nasty
>>   and inefficient approximation to TLS, which requires
>>   (see glibc/linuxthreads/errno.c.)
> 
> I wondered about this. The other day I saw it implemented in the thread_desc 
> (something like that) struct.

Yep. Basically it has to hack TLS itself, and the way it does it incurs
extra overhead on every context switch, unlike TLS.

>> - Under NPTL, and under LinuxThreads on a __thread, TLS-supporting
>>   platform (i.e., on glibc compiled for i686 and above, with GCC 3.4+),
>>   glibc exports a GLIBC_PRIVATE thread_local symbol named errno,
>>   which __errno_location() picks up. This symbol is exported in the
>>   GLIBC_2.0 symbol version set if TLS is not in use: if it is, then
>>   that symbol has no sensible value, and it's not exported at all.
> 
>> Note that the actual storage of errno under NPTL is a matter of
>> cooperation between GCC, glibc and GNU ld: glibc declares errno with the
>> __thread storage class specifier, and GCC proceeds to emit a stereotyped
>> instruction sequence with suitable relocations
> Is all this a resume of what "tls.pdf" from Ulrich Drepper site goes 
> describing in detail? If so, I'll study it later.

Well, the tls.pdf site is describing *how TLS is implemented*; this is more
`what errno uses in addition to TLS'. At bottom NPTL errno is a TLS
variable, but glibc goes through considerable dancing so that apps that
link against it don't need to know that.

Alas, that means that you can't portably rely upon the TLS stuff not
changing :(

> Also, about TLS and such: do you know where I can download the userspace 
> example code?

Is there any? Given a suitable glibc, you should just be able to declare
a variable __thread and have it Just Work.

>   I'd like to use some kind of semaphores in UML for cooperation 
> between host threads. Currently it suspends by read/writing a byte on a pipe 

TLS should work --- but last I tried it UML was still allergic to
TLS. (I guess the whole point of all this is to fix that, though.)

>> (Um, why is UML linking against libpcap? Or don't I want to know?)
> The net pcap transport. Only when it's enabled UML links with pcap (verify 
> only with 2.4, not with 2.6).

Oh gods, I'd quite forgotten about the myriads of non-TUN/TAP network
drivers. (How many people use the non-TUN/TAP stuff, anyway?)

>> > The questions are:
>> >  - shouldn't common symbols in libraries be forbidden
>>
>> Since GCC generates unintialized non-static globals that way (because it
>> saves space in the executable), er, no. Normally you *want* globals to
>> be, er, global. :)
> No, I don't mean BSS, I mean common symbols. It's the linker practice which 
> makes:
>
> int i

Re: [uml-user] user-mode-linux 2.4.27 -> 2.6.11 woes

[Nix]

On Tue, 17 May 2005, [EMAIL PROTECTED] moaned:
> On Tuesday 17 May 2005 01:46, Blaisorblade wrote:
>> Also, about TLS and such: do you know where I can download the userspace
>> example code?
> Here I refer to futexes, sorry.

/pub/linux/kernel/people/rusty/futex-2.2.tar.gz, I think.

-- 
`Once again, I must remark on the far-reaching extent of my
 ladylike nature.' --- Rosie Taylor


---
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Running SKAS: 3 processes ok, one terminated

[Nix]

On Sat, 4 Jun 2005, [EMAIL PROTECTED] moaned:
> Hi all,
> 
> I am running UML with the SKAS patch applied. All works fine although my
> host system shows one UML-process as terminated:
> 
>  2605 pts/1S  0:00 ./linux (uml1) [/sbin/getty]
>  2607 pts/1T  0:00 [linux]
>  2612 pts/1S  0:00 ./linux (uml1) [/sbin/getty]
>  2613 pts/1S  0:00 ./linux (uml1) [/sbin/getty]

T is not terminated, it's stopped (actually being ptrace()d).

And yes, it's normal.

-- 
`Once again, I must remark on the far-reaching extent of my
 ladylike nature.' --- Rosie Taylor


---
This SF.Net email is sponsored by: NEC IT Guy Games.  How far can you shotput
a projector? How fast can you ride your desk chair down the office luge track?
If you want to score the big prize, get to know the little guy.  
Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] UML using 100% CPU after mounting root filesystem

[Nix]

On Fri, 22 Jul 2005, [EMAIL PROTECTED] yowled:
> On Friday 22 July 2005 18:53, Ruaidhri Power wrote:
>> On Thu, Jul 21, 2005 at 01:23:46AM +0200, Blaisorblade wrote:
>> > Which was the old UML version? Which is your guest distro?
>> > Also, have you tested the old UML releases on this new host kernel
>> > version? Since UML address space is getting shrinked down to 32M, I fear
>> > address space randomization (introduced in 2.6.12) may be playing a role.
> 
>> It looks like VA space randomization is the culprit.  The problem only
>> shows up when the host is 2.6.12, and can be solved by setting the
>> kernel.randomize_va_space sysctl to zero.  Thanks for the pointer there.
[...]
> Doesn't this disappear when shutting off address space randomization?

Odd. I'm using a 2.6.13.2 host and 2.6.11.9-bs5 guest, and
randomize_va_space works fine. It worked fine when I had a 2.6.12.x host,
as well.

Is there something subarch-specific here (unlikely), or is it Yet
Another bad NPTL interoperation or something like that?

-- 
`But of course, GR is the very best relativity for the masses.'
 --- Wayne Throop


---
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Hardware Requirements -Some Questions about UML by a Newbie

[Nix]

On Thu, 11 Aug 2005, eric leitner wondered:
> - SKAS / TT
> Is SKAS really worth the trouble? I mean is the performance better? Or
> are there other issues why one should use SKAS Mode?

Drastically better performance, easier debuggability, lower load on the
host, better security. Use it unless you have no choice.

> - Memory Support
> The server has 12 GB of RAM - will that be addressed by UML? Or is
> there some kind of limit? Are there performance penalties with > 4 GB
> larger than usual?

Well, UML is a 32-bit process, so it can only address the usual 2--3Gb
limit for such processes. If you want more than that, I'd say it makes
sense to run separate UML instances.

-- 
`I work in computers so, of course, I'm an expert on everything.'
 --- Simon Rumble


---
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] 2.6.12.[35]-bs{7,11} boot failure

[Nix]

(Those are all the 2.6.12 UMLs I've tried.)

Basically, userspace never starts: the last two lines I see are

VFS: Mounted root (ext2 filesystem) readonly.
line_ioctl: tty0: ioctl KDSIGACCEPT called
[halt]

(If I reboot into 2.6.11.9-bs5 with the same root fs, the next line
I see is init's banner.)

A trace of the main thread shows it sleeping, perhaps because, well,
userspace hasn't started yet:

USER   PID  PPID  C STIME TTY  TIME CMD
firewall 14958 14942  0 20:40 ?00:00:00 [uml-esperi]
firewall 14962 14942  0 20:40 ?00:00:00 uml-esperi con=port:16380 
con0=fd:0,fd:1 fakeide mem=96M ubd0=/mirror/uml/esperi-roo
firewall 14963 14942  0 20:40 ?00:00:00 uml-esperi con=port:16380 
con0=fd:0,fd:1 fakeide mem=96M ubd0=/mirror/uml/esperi-roo
firewall 14942 14941  1 20:40 ?00:00:01 uml-esperi con=port:16380 
con0=fd:0,fd:1 fakeide mem=96M ubd0=/mirror/uml/esperi-roo

Attaching to program: /usr/packages/linux/variants/i586-esperi/linux, process 
14942
Reading symbols from /lib/libutil.so.1...done.
Loaded symbols for /lib/libutil.so.1
Reading symbols from /lib/libc.so.6...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
0xb7ee91e5 in nanosleep () from /lib/libc.so.6
(gdb) bt
#0  0xb7ee91e5 in nanosleep () from /lib/libc.so.6
#1  0x0805e28f in idle_sleep ()
#2  0x081f3bb4 in init_thread_union ()
#3  0x in ?? ()
#4  0xbfd765d8 in ?? ()
#5  0x0805b601 in kernel_thread ()
#6  0x000a in ?? ()
#7  0x in ?? ()
#8  0x in ?? ()
#9  0x0805b81c in default_idle ()
#10 0x000a in ?? ()
#11 0xbfd765d8 in ?? ()
#12 0x0805b82f in default_idle ()
#13 0x081f in ?? ()
#14 0x081f in ?? ()
#15 0x081f in ?? ()
#16 0x081f in ?? ()
#17 0x081f in ?? ()
#18 0x in ?? ()
#19 0x08061804 in init_idle_skas ()
#20 0x08049598 in start_kernel ()
#21 0x in ?? ()
#22 0x in ?? ()
#23 0x0822aba0 in uml_start ()
#24 0x0806183b in start_kernel_proc ()
#25 0x08061810 in init_idle_skas ()
#26 0x in ?? ()
#27 0x0001 in ?? ()
#28 0x081f5220 in init_task ()
#29 0x081f5220 in init_task ()
#30 0x in ?? ()
#31 0x08061810 in init_idle_skas ()
#32 0x0805b35b in run_kernel_thread ()
#33 0x0001 in ?? ()
#34 0x1000 in ?? ()
#35 0x in ?? ()
#36 0x in ?? ()
#37 0x in ?? ()
#38 0x in ?? ()
#39 0x in ?? ()
#40 0x in ?? ()
#41 0x in ?? ()
#42 0x in ?? ()
#43 0x in ?? ()
#44 0x in ?? ()
#45 0x in ?? ()
#46 0x in ?? ()
#47 0x in ?? ()
#48 0x in ?? ()
#49 0x in ?? ()
#50 0x in ?? ()
#51 0x in ?? ()
#52 0x in ?? ()
#53 0x in ?? ()
#54 0x in ?? ()
#55 0x in ?? ()
#56 0x in ?? ()
#57 0x in ?? ()
#58 0x in ?? ()
#59 0x in ?? ()
#60 0x in ?? ()
#61 0x in ?? ()
#62 0x in ?? ()
#63 0x in ?? ()
#64 0x in ?? ()
#65 0x in ?? ()
#66 0x in ?? ()
#67 0x in ?? ()
#68 0x08061810 in init_idle_skas ()
#69 0x in ?? ()
#70 0x081f5220 in init_task ()
#71 0x08060fd8 in thread_wait ()
#72 0x in ?? ()
#73 0x081f in ?? ()
#74 0x081f in ?? ()
#75 0x in ?? ()
#76 0x081f in ?? ()
#77 0x081f in ?? ()
#78 0x081f in ?? ()
#79 0x081f in ?? ()
#80 0x081f in ?? ()
#81 0x3a5e in ?? ()
#82 0x0600 in ?? ()
#83 
#84 0xb7e8d321 in kill () from /lib/libc.so.6
#85 0xb7e8cf94 in siglongjmp () from /lib/libc.so.6
#86 0xb7e7a4ae in __libc_start_main () from /lib/libc.so.6
#87 0x080580d1 in _start ()


Any ideas?

(And sorry for not getting around to getting a backtrace for so long
after the release of 2.6.12.*.)

-- 
`I work in computers so, of course, I'm an expert on everything.'
 --- Simon Rumble


---
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

[uml-user] user-mode-linux 2.6.13.2-bs3 oddity

[Nix]

I'm using a COWed block device as my root filesystem. After boot, I see
this:

Size mismatch (140789027962880 vs 209715200) of COW header vs backing file
Failed to open '/mirror/uml/esperi-root-cow.image', errno = 22
VFS: Cannot open root device "98:0" or unknown-block(98,0)
Please append a correct "root=" boot option
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(98,0)

With 2.6.12.5-bs11, it mounts fine.

There don't seem to be any new userspace utilities since April 2004 (at least
not on sourceforge) --- so it can't be that the COW file format has changed.
This looks more like some sort of byte-ordering problem.

Any ideas? What dumb thing have I done now? (Is it just that I need to let
UML recreate the COW file itself? Are COW files not portable between UMLs
or something?)

-- 
`Next: FEMA neglects to take into account the possibility of
fire in Old Balsawood Town (currently in its fifth year of drought
and home of the General Grant Home for Compulsive Arsonists).'
--- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.6.13.2-bs3 oddity

[Nix]

On Mon, 3 Oct 2005, [EMAIL PROTECTED] yowled:
> On Sunday 02 October 2005 22:52, Nix wrote:
>> Size mismatch (140789027962880 vs 209715200) of COW header vs backing file
>> Failed to open '/mirror/uml/esperi-root-cow.image', errno = 22
>> VFS: Cannot open root device "98:0" or unknown-block(98,0)
>> Please append a correct "root=" boot option
>> Kernel panic - not syncing: VFS: Unable to mount root fs on
>> unknown-block(98,0)
> 
>> With 2.6.12.5-bs11, it mounts fine.
> Have you checked now? I was wondering if you increased the backing file 
> size... but the mismatch is excessive, so ignore this.

Not by that much. I *wish* I had that much disk space ;)

(1.4 exabytes, is that?)

> What the bloody hell... I'll look into this ASAP, but I don't remember any 
> changes there.

Looks like there are some :(

>> This looks more like some sort of byte-ordering problem.
> I hope...

Well, 140789027962880 divides evenly into 2^32. Byte-ordering or a
one-long misalignment of some kind or I'll eat my hat. (I haven't got a
hat so I'll have to get one first and then eat it.)

-- 
`Next: FEMA neglects to take into account the possibility of
fire in Old Balsawood Town (currently in its fifth year of drought
and home of the General Grant Home for Compulsive Arsonists).'
--- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.6.13.2-bs3 oddity

[Nix]

On Sun, 9 Oct 2005, [EMAIL PROTECTED] yowled:
> On Monday 03 October 2005 21:50, Nix wrote:
>> Well, 140789027962880 divides evenly into 2^32. Byte-ordering or a
>> one-long misalignment of some kind or I'll eat my hat. (I haven't got a
>> hat so I'll have to get one first and then eat it.)
> 
> Ok, no need to buy any hat. I would have liked very much to see you doing 
> that, but it will be for next time ;-).

Oh, good, I was wondering if I'd have to find somewhere that sold
shortbread headgear. :)

> Compliments.

Hey, you're the one who deserves the complients. You fixed the bug. I just
pontificated. I could at least have bisected for it...

> In fact, between 2.6.13-rc3 and -rc4, a "global style fixup" (which was 
> deemed 
> trivial, and which didn't even reach neither me nor Jeff) affected 

Ah, isn't it wonderful when people break your code without telling you?

> diff --git a/arch/um/drivers/cow.h b/arch/um/drivers/cow.h
> --- a/arch/um/drivers/cow.h
> +++ b/arch/um/drivers/cow.h
> @@ -3,10 +3,10 @@
> 
>  #include 
> 
> -#if __BYTE_ORDER == __BIG_ENDIAN
> +#if defined(__BIG_ENDIAN)
>  # define ntohll(x) (x)
>  # define htonll(x) (x)
> -#elif __BYTE_ORDER == __LITTLE_ENDIAN
> +#elif defined(__LITTLE_ENDIAN)
>  # define ntohll(x)  bswap_64(x)
>  # define htonll(x)  bswap_64(x)
>  #else

Someone didn't notice the significance of that UML-and-userspace-
specific __BYTE_ORDER, I fear.

> grep [nh]to[hn]ll arch/um/drivers/{cow*,ubd*}
> 
> shows all the culprit calls.

(well, with quoting to stop the shell interpreting that regexp :) )

> And they affect exactly size.

You've got it, I'd say.

> Patches are attached - to apply in this order. Actually the first should fix 
> everything, but I'm gonna merge all three ones in -bs4, so you can test all 
> them together (I actually compared the preprocessed source).

... yep, confirmed; with those patches, it works.


Thank you!

-- 
`Next: FEMA neglects to take into account the possibility of
fire in Old Balsawood Town (currently in its fifth year of drought
and home of the General Grant Home for Compulsive Arsonists).'
--- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] user-mode-linux 2.6.13.2-bs3 oddity

[Nix]

On Tue, 11 Oct 2005, [EMAIL PROTECTED] uttered the following:
> On Monday 10 October 2005 00:01, Nix wrote:
>> > Patches are attached - to apply in this order. Actually the first should
>> > fix everything, but I'm gonna merge all three ones in -bs4, so you can
>> > test all them together (I actually compared the preprocessed source).
>>
>> ... yep, confirmed; with those patches, it works.
> 
> Can you try -rc4 with Jeff's AIO patch revertal, and confirm that it works 
> there too?
> 
> Currently there should be (no) known issues with -rc4, with that patch.

It works fine.

(Apologies for the delay: medical bureaucratic crap intervened.)

-- 
`Next: FEMA neglects to take into account the possibility of
fire in Old Balsawood Town (currently in its fifth year of drought
and home of the General Grant Home for Compulsive Arsonists).'
--- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Gentoo stage3 for uml on amd64 host

[Nix]

On Sat, 15 Oct 2005, Joel Palmius wrote:
> On Sat, 15 Oct 2005, Blaisorblade wrote:
> 
>> One question - -march=athlon64 is also supported by GCC (even if I don't know
>> how much it helps).
> 
> Yup. Or -march=k8.

These are synonyms for the same thing, as is -march=opteron and
-march=athlon-fx.

-march=x86-64 is similar but disables use of 3DNow intrinsics (not that
those are used much yet unless you use vector stuff a lot: I guess when
OpenMP and autovectorization start to turn up in GCC trunks, we'll see
more use from them).

> The recommendation for building a 32-bit chroot in
> the gentoo docs says -march=athlon-xp though,

Indeed, this is almost identical to specifying athlon64 except that
it's 32-bit.

>From gcc-4.0.2/gcc/config/i386/i386.c's processor_alias_table:

,
|  {"k6", PROCESSOR_K6, PTA_MMX},
|  {"k6-2", PROCESSOR_K6, PTA_MMX | PTA_3DNOW},
|  {"k6-3", PROCESSOR_K6, PTA_MMX | PTA_3DNOW},
|  {"athlon", PROCESSOR_ATHLON, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW
|   | PTA_3DNOW_A},
|  {"athlon-tbird", PROCESSOR_ATHLON, PTA_MMX | PTA_PREFETCH_SSE
| | PTA_3DNOW | PTA_3DNOW_A},
|  {"athlon-4", PROCESSOR_ATHLON, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW
|| PTA_3DNOW_A | PTA_SSE},
|  {"athlon-xp", PROCESSOR_ATHLON, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW
|  | PTA_3DNOW_A | PTA_SSE},
|  {"athlon-mp", PROCESSOR_ATHLON, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW
|  | PTA_3DNOW_A | PTA_SSE},
|  {"x86-64", PROCESSOR_K8, PTA_MMX | PTA_PREFETCH_SSE | PTA_64BIT
|   | PTA_SSE | PTA_SSE2 },
|  {"k8", PROCESSOR_K8, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW | PTA_64BIT
|  | PTA_3DNOW_A | PTA_SSE | PTA_SSE2},
|  {"opteron", PROCESSOR_K8, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW | PTA_64BIT
|  | PTA_3DNOW_A | PTA_SSE | PTA_SSE2},
|  {"athlon64", PROCESSOR_K8, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW | PTA_64BIT
|  | PTA_3DNOW_A | PTA_SSE | PTA_SSE2},
|  {"athlon-fx", PROCESSOR_K8, PTA_MMX | PTA_PREFETCH_SSE | PTA_3DNOW | 
PTA_64BIT
|  | PTA_3DNOW_A | PTA_SSE | PTA_SSE2},
`

(As far as I can tell PROCESSOR_K8 and PROCESSOR_ATHLON are identical in
GCC except for bitness, a few predefined macros, and scheduling.

Of course the first of those changes ia a biggie, but you're not losing
much by having GCC consider 32-bit code on your K8 to be running on an
Athlon.)

>   possibly because there
> was some confusion about whether you'd really get it 32 bit if you set
> k8 or athlon64. I don't know, I didn't test, so I simply went with the
> recommendation.

I hope the stuff above is useful, then. (I didn't test either: I don't
even have a K8. But the source never lies. :) )

-- 
`Next: FEMA neglects to take into account the possibility of
fire in Old Balsawood Town (currently in its fifth year of drought
and home of the General Grant Home for Compulsive Arsonists).'
--- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Gentoo stage3 for uml on amd64 host

[Nix]

On Mon, 17 Oct 2005, [EMAIL PROTECTED] stated:
> On Sunday 16 October 2005 19:31, Nix wrote:
>> These are synonyms for the same thing, as is -march=opteron and
>> -march=athlon-fx.
> 
>> -march=x86-64 is similar but disables use of 3DNow intrinsics (not that
>> those are used much yet unless you use vector stuff a lot: I guess when
>> OpenMP and autovectorization start to turn up in GCC trunks, we'll see
>> more use from them).
> 
>> (As far as I can tell PROCESSOR_K8 and PROCESSOR_ATHLON are identical in
>> GCC except for bitness, a few predefined macros, and scheduling.

I was wrong: the K8 has faster vector ops, and GCC can use those.

> I don't know, but I exactly expected scheduling to be changed.

I was wrong; except for the vector-instruction stuff, Athlon scheduling
is used. I shouldn't try to read GCC source while thick with cold.

Instruction costs differ between K8 and Athlon, though (unlike Intel AMD
still provide instruction timings, so they're accurate, too). That might
well affect scheduling.

> You may 
> correct 
> me (you know more about GCC), but I expect scheduling to make some 
> difference. Btw, I still haven't done the tricks to make UML use host cpu 
> features - it's still compiled with default GCC settings for this.

Not a killer: those of us who really care will have made GCC default anyway
by specifying --with-cpu or a target-specific triplet to GCC configure.

> Actually, for the host kernel (which uses arch=k8, see arch/i386/Makefile)) 
> SSEx makes no difference (the kernel can't allow GCC to use FP on its own, as 
> you probably know - the kernel must explicitly enable FP state saving, and 
> even then there are strict rules on FP usage, for instance you must be 
> atomic).

Well, actually, if you don't specify -ffreestanding, GCC can implement
large block moves by going via SSE or MMX regs. (How common this is I'm
not sure. Probably not very, especially for MMX.)

-- 
`"Gun-wielding recluse gunned down by local police" isn't the epitaph
 I want. I am hoping for "Witnesses reported the sound up to two hundred
 kilometers away" or "Last body part finally located".' --- James Nicoll


---
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] line_write_room: tty1: no room left in buffer

[Nix]

On Tue, 29 Nov 2005, [EMAIL PROTECTED] announced authoritatively:
> Uff, it's the no. 1 UML FAQ - move /lib/tls away in the guest image, UML 
> doesn't support NPTL (yet).

There's actually two halves to this:

- UML doesn't support NPTL in its guests, move /lib/tls away

- UML doesn't work if an NPTL glibc is used at all, run it with
  LD_ASSUME_KERNEL=2.4.30 on a suitably configured distro (i.e.
  not Fedora Core 4). (Disclaimer, last tested last year and may
  no longer be true: please tell me it isn't true 'cos this is
  the last thing I need my non-NPTL glibc for on my hosts!)

-- 
`Y'know, London's nice at this time of year. If you like your cities
 freezing cold and full of surly gits.' --- David Damerell



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] line_write_room: tty1: no room left in buffer

[Nix]

On Wed, 30 Nov 2005, Jeff Dike spake:
> On Wed, Nov 30, 2005 at 11:03:56AM +0000, Nix wrote:
>> There's actually two halves to this:
>> 
>> - UML doesn't support NPTL in its guests, move /lib/tls away
> 
> My current patchset is running NPTL just fine.

Yeah, sorry, I'm just using 2.6.14.3-bs1, which doesn't support
NPTL. I'm aware that your excellent patchset will finally let
me go NPTL-everywhere, which is great. :)

>> - UML doesn't work if an NPTL glibc is used at all, run it with
>>   LD_ASSUME_KERNEL=2.4.30 on a suitably configured distro (i.e.
>>   not Fedora Core 4). (Disclaimer, last tested last year and may
>>   no longer be true: please tell me it isn't true 'cos this is
>>   the last thing I need my non-NPTL glibc for on my hosts!)
> 
> I build UML on FC4 all the time with no LD_* tricks.

So it runs without that patchset on a host with no non-NPTL glibc?

woo. I'll test this this evening (but as my only live UML root right now
runs my firewall and I'm sshing in through it, testing it now is a bit
difficult ;) )

If so, three machines' groaning root filesystems thank you :)

-- 
`Y'know, London's nice at this time of year. If you like your cities
 freezing cold and full of surly gits.' --- David Damerell



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] line_write_room: tty1: no room left in buffer

[Nix]

On Wed, 30 Nov 2005, Jeff Dike uttered the following:
> If the patchset were so excellent, it would be in mainline by now.  It still
> needs work, even though it is correct AFAICT.

Merely the existence of the patchset is excellent :)

-- 
`Y'know, London's nice at this time of year. If you like your cities
 freezing cold and full of surly gits.' --- David Damerell



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Three newbie questions

[Nix]

On Fri, 2 Dec 2005, Klaus-J. Wolf suggested tentatively:
> Hi,
> 
> I am aware of the fact that my small troubles tend to be rather 
> unspecified, but maybe someone can help anyway.
> 
> I run 2.6.14-ck6-skas3-v8.2 as host and 2.4.28-bs2 as guest OS.
> 
> When I start "linux":
> 
>  1. There are many files created in the current directory on the host, 
> named like "1133497700-955364", containing fragments of the output, e.g. 
> "Switching to runlevel: 0". This is really annoying.

It's actually a feature for the paranoid; UML-2.4 (though not 2.6) can
log all that passes through all its TTYs :)

To stop it doing that, turn off TTY logging in the kernel configuration.
To make it put them somewhere else, pass tty_log_dir=/some/directory on
the UML command line.

-- 
`Y'know, London's nice at this time of year. If you like your cities
 freezing cold and full of surly gits.' --- David Damerell



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Three newbie questions

[Nix]

On Fri, 2 Dec 2005, Rob Landley gibbered uncontrollably:
> On Friday 02 December 2005 07:09, Nix wrote:
> 
>> >  1. There are many files created in the current directory on the host,
>> > named like "1133497700-955364", containing fragments of the output, e.g.
>> > "Switching to runlevel: 0". This is really annoying.
>>
>> It's actually a feature for the paranoid; UML-2.4 (though not 2.6) can
>> log all that passes through all its TTYs :)
> 
> Ah, is that it?

At some point I might figure out how it worked and reimplement it, 'cos
I was using it and scanning the resulting logs for suspicious content,
correlating their file times with times I knew the system was in use by
legitimate users, and so on...

> I never could get 2.4 UM to work, and if I was going to have to debug 
> something anyway, 2.6 seemed the logical choice...

I had great trouble getting 2.6 to work (it had trouble working on boxes
older than i686 until 2.6.11 or thereabouts; Blaisorblade fixed it with
alacrity once I mentioned the problem).

-- 
`Y'know, London's nice at this time of year. If you like your cities
 freezing cold and full of surly gits.' --- David Damerell



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Help compiling kernel 2.6.7

[Nix]

On Wed, 7 Dec 2005, noel anderson murmured woefully:
>  Am trying to compile kernel 2..6.7 with the equivlent patch 2.6.7-1. Any 
> help is appreciated. 
>  I'm using FC4.

There's *very* little chance that a UML patch of such great age will
work reliably. IIRC, UML started working reliably around 2.6.10.
(It seemed to work for others before then, but I had little luck.
Perhaps it was just allergic to me.)

Why not upgrade? That kernel is over a year old, and has many *many*
known security holes (some remote).

-- 
`Don't confuse the shark with the remoras.' --- Rob Landley



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Sockets question

[Nix]

On Thu, 8 Dec 2005, Christopher Barry moaned:
> when two virtual machines use sockets between one another on a single
> local system, does that end up routed through the loopback interface?

Well, it shouldn't end up routed through any physical network cards,
but it shouldn't go through `lo'.

At least it doesn't for me, with a pair of TUN/TAPs bridged together.

-- 
`Don't confuse the shark with the remoras.' --- Rob Landley



---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] newbe questions

[Nix]

On Wed, 22 Mar 2006, [EMAIL PROTECTED] whispered secretively:
> On Wednesday 22 March 2006 17:46, David Lang wrote:
>> Configuring network interfaces...make: /etc/mail/Makefile: No such file or
>> directory
>> make: *** No rule to make target `/etc/mail/Makefile'.  Stop.
>> make: /etc/mail/Makefile: No such file or directory
>> make: *** No rule to make target `/etc/mail/Makefile'.  Stop.
> 
> (DHCP is likely bogus inside UML, unless you have a DHCP server on the host 
> listening there or you setup bridging with the main interface).

Running make as root is even odder. Is this some new distro trick for
tracking network daemon dependencies or something that I haven't heard of?

(If so, the typical placement of make in /usr/bin torpedoes mounting /usr
over the network...)

-- 
`Come now, you should know that whenever you plan the duration of your
 unplanned downtime, you should add in padding for random management
 freakouts.'


---
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] newbe questions

[Nix]

On Wed, 22 Mar 2006, David Lang uttered the following:
> when you do ifup debian immediatly attempts to startup sendmail
> (apparently by doing make as you see here). if you can't do name
> resolution until other interfaces come up you end up with a 30 second
> delay in your machine booting.

I actually do exactly the same thing in my firewall UML instance --- but
then its entire purpose is to route packets and act as a barrier between
my machine and the net so this is hardly a catastrophe, and by that
point the daemon which kicks my crappy router until it brings the line
up has been started, so it's all asynchronous and causes no net delay...

... but if the interface startup doesn't proceed asynchronously with
respect to daemon startup, then it's bad, indeed.

-- 
`Come now, you should know that whenever you plan the duration of your
 unplanned downtime, you should add in padding for random management
 freakouts.'


---
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] about mem=nM

[Nix]

On Wed, 28 Jun 2006, paul john uttered the following:
> if start um without mem=256M , it's ok , if start um with mem=256M it hangs
> there .report "mount devfs on /dev"

Do you have at least 256Mb of space free in /dev/shm?

(In older UMLs, pre-2.6.17, you'll need to make sure /tmp has enough
room. If TMPFS is set, that's used instead.)

-- 
`NB: Anyone suggesting that we should say "Tibibytes" instead of
 Terabytes there will be hunted down and brutally slain.
 That is all.' --- Matthew Wilcox

Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] about mem=nM

[Nix]

On Tue, 27 Jun 2006, Jeff Dike mused:
> On Wed, Jun 28, 2006 at 12:36:19AM +0100, Nix wrote:
>> On Wed, 28 Jun 2006, paul john uttered the following:
>> > if start um without mem=256M , it's ok , if start um with mem=256M it hangs
>> > there .report "mount devfs on /dev"
>> 
>> Do you have at least 256Mb of space free in /dev/shm?
>> 
>> (In older UMLs, pre-2.6.17, you'll need to make sure /tmp has enough
>> room. If TMPFS is set, that's used instead.)
> 
> This one will show up as a SIGBUS (signal 7) sometime after boot.  At this
> point in the boot, init is the only thing running and there's no way
> that it should have eaten 256M.

Hm, true, it's only going to allocate space that's touched.

> However, if there really is no space in /tmp (or /dev/shm), then maybe.

That doesn't seem to be the case, though: something else is wrong.

-- 
`NB: Anyone suggesting that we should say "Tibibytes" instead of
 Terabytes there will be hunted down and brutally slain.
 That is all.' --- Matthew Wilcox

Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Starting udev:

[Nix]

On Thu, 17 Aug 2006, Jeff Dike wrote:
> On Thu, Aug 17, 2006 at 12:54:18PM -0400, Rick Spillane wrote:
>> I was wondering why this phase of the boot process is so slow. Is
>> there some way I can speed it up? If there is some configuration I can
>> tweak to skip this step, that would be great, if it would require
>> changing a lot of kernel source or what-not, then just tell me that
>> also, and I will live with it. Any explanation as to why this is so
>> slow though would be greatly appreciated.
> 
> It creates a hell of a lot of short-lived processes.  Process creation
> is slow for UML, so this makes this part of the boot particularly
> slow.

If you use udevtrigger from udev 088 or higher instead of the shell
script stanza, this should no longer be a problem.

-- 
`In typical emacs fashion, it is both absurdly ornate and
 still not really what one wanted.' --- jdev

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Boot problem

[Nix]

On 12 Oct 2006, Christopher Marshall told this:
> The other thing you can try is to make a simple initrd file with some
> busybox commands in it, enough to give you the ash shell and some
> commands like mount, pivot_root, chroot, so you can try to mount the
> root filesystem yourself then pivot_root and chroot into it.

As an aside, initrd is obsolete and rather annoying to construct. If
I were you I'd use an initramfs instead, which is utterly trivial to
build (modulo the usual annoyances getting the tools together) and
can't *ever* get out of synch with the kernel because they're built
at the same time.

(It seems likely that a large part of the boot process will move
obligatorily into an initramfs quite soon. Nobody is likely to notice
unless they're watching the kernel build closely. That's how transparent
it is to users.)


Rough procedure for an uClibc/busybox-based initrd:

 - make a cross-compiler and cross-assembler for uClibc. (The uClibc
   site has patches to GCC and binutils to add a *-pc-linux-uclibc
   target.)
 - make uClibc from CVS
 - make busybox, 1.2 or later
 - write a file usr/initramfs in the kernel tree, containing a list
   of files to include in the initramfs, viz

,
| file /init usr/init 0755 0 0
| dir /bin 0755 0 0
| file /bin/busybox /usr/i686-pc-linux-uclibc/bin/busybox 0755 0 0
| slink /bin/sh /bin/busybox 0755 0 0
| slink /bin/ash /bin/busybox 0755 0 0
[...]
| # supporting directories
| dir /proc 0755 0 0
| dir /sys 0755 0 0
| dir /new-root 0755 0 0
| dir /etc 0755 0 0
[...]
| # initial device files required (mdev creates the rest)
| dir /dev 0755 0 0
| nod /dev/console 0600 0 0 c 5 1
| nod /dev/null 0666 0 0 c 1 3
`

Make a shell script usr/init (what whever you like: you can name it anything
and rename it to /init with the usr/initramfs file, which you can *also*
name whatever you want). This script gets passed *all* parameters that the
kernel was passed (except for a few which it parses itself) and can do
whatever it likes with them. e.g. the start of mine:

,
| export PATH=/sbin:/bin
| 
| /bin/mount -t proc proc /proc
| /bin/mount -t sysfs sysfs /sys
| CMDLINE=`cat /proc/cmdline`
| 
| # Populate /dev from /sys
| 
| /bin/mount -t tmpfs tmpfs /dev
| /sbin/mdev -s
| 
| # Locate the root filesystem's fstab entry; collapse spaces and tabs in it:
| # extract its significant components. (There are three raw tabs in the next
| # line.)
| 
| FSENT=`sed -n '/[ ]\/[]/ { s,[][  ]*, ,g; p; }' < 
/etc/fstab`
| ROOT="`echo $FSENT | tr ' ' '\n' | sed -n '1p'`"
| TYPE="`echo $FSENT | tr ' ' '\n' | sed -n '3p'`"
| OPTS="`echo $FSENT | tr ' ' '\n' | sed -n '4p'`"
| 
| # Parse arguments, engaging trace mode or dropping to rescue or emergency 
shells
| # as needed. If there is a forced init program, root filesystem, root fs type 
or
| # root fs options, accept the forcing.
| 
| INIT_ARGS=
| 
| for param in $CMDLINE; do
| case "$param" in
| init=*) eval "$param";;
|   -b|single|s|S|[1-5]) INIT_ARGS="$INIT_ARGS $param";;
| trace) echo "Tracing init script.";
|set -x;;
| rescue) echo "Rescue boot mode: invoking ash.";
| init=/bin/ash;
| INIT_ARGS="-";;
| emergency) echo "Emergency boot mode. Dropping to a minimal shell.";
|echo "Reboot with Ctrl-Alt-Delete.";
|exec /bin/sh;;
| root=LABEL=*) ROOT=$(echo $1 | cut -d= -f3-);;
| root-type=*) TYPE=$(echo $1 | cut -d= -f2-);;
| root-options=*) OPTS=$(echo $1 | cut -d= -f2-);;
| esac
| done
`

The last thing it does (after finding the root filesystem, of course)
should be to unmount everything but / and the newly mounted root, and
switch_root into it:

,
| if [ -z "$init" ]; then
| init=/sbin/init
| fi
| 
| # Unmount everything and switch root filesystems for good:
| # exec the real init and begin the real boot process.
| /bin/umount -l /proc
| /bin/umount -l /sys
| /bin/umount -l /dev
| 
| echo "Switching to /new-root and running '$init'"
| exec switch_root /new-root $init $INIT_ARGS
`

(switch_root is a tool in busybox that recursively deletes everything on
the same filesystem as /, as initramfses are nonswappable and anything
left there will eat memory forever. Then it chroots to the new
filesystem and exec's init. --- so look out, you'd better make sure you
still have PID 1 at that point...

The filesystem will be auto-built and gzipped into the kernel image at
kernel build time. As of 2.6.18, any C code mentioned in the initramfs
source file will also be compiled for you!


The last part in particular is *vastly* less fiddly than switching to
the real root filesystem with an initrd, which has about three possible
conflicting methods none of which work terribly well in my experience :(

-- 
`When we are born we have plenty of Hydrogen but as we age our
 Hydrogen pool becomes depleted.'

Re: [uml-user] Boot problem

[Nix]

On 25 Oct 2006, Christopher Marshall verbalised:
> Thanks for all the information about initramfs, that was a very informative 
> post.

I can but try. :)

I've stuck my boot scripts online at
. I don't know if
they'll be any use, but at least it gives you a known-working
example. (Well, it works for me and for a few other people, and nobody
has told me that it *doesn't* work, although of course that might be
because they can no longer boot ;))) )

> It also looks like you may have solved a long standing mystery to me.
> Back when I was experimenting with the initrd process linuxrc,
> pivot_root, chroot, and init, I noticed that unless I invoked init
> with the "-i" argument, which tells init to assume it was called from
> the kernel to boot the system, init would always exit with an error
> message to the effect that it didn't know what to do.
>
> I'll bet init was doing that because it's PID was not 1, as you allude
> to above.  So it assumed it was not being called from the kernel.

Indeed. See the top of main() in sysvinit's src/init.c.

> Since it had no other arguments, it didn't know what to do and printed
> out a syntax guide.  I'll have to see if I can reproduce that bug
> sometime and see if the PID was something other than 1.

Almost certainly. If your /init script is PID 1 (which it normally is)
you have to remember to run init via `exec'. busybox's switch_root
command had a rather bad diagnostic for this case when I tried to set up
an initramfs for the first time, and I spent some time scattering
printf()s around until I worked out which of the dozen-or-so failure
paths in switch_root was failing. (It has better diagnostics now.)

-- 
`When we are born we have plenty of Hydrogen but as we age our
 Hydrogen pool becomes depleted.'

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Boot problem

[Nix]

On 26 Oct 2006, [EMAIL PROTECTED] uttered the following:
> On Wednesday 25 October 2006 23:37, Nix wrote:
>> On 25 Oct 2006, Christopher Marshall verbalised:
>> > Thanks for all the information about initramfs, that was a very
>> > informative post.
>>
>> I can but try. :)
> Have you got a UML+initramfs kernel? Ah ok, you embedded the initramfs... my 

Well, yes, of course. I really dislike the non-embedded versions: they
get out of synch with the kernel and it's hard to have different
initramfses for different kernels without constantly fiddling with the
bootloader configuration and they don't get autobuilt with the kernel
and and and.

However, my UML is the one Linux installation I administer which I'm
*not* yet booting from initramfs: I plan to fix this in very short order
(i.e., next weekend).

> personal bugzilla (i.e. my brain) has a NEW (CONFIRMED) bug that 
> initrd=$initramfs_file won't work with UML, so Gentoo install doesn't work 
> either. Can you tell me whether you falsified my report or it is true for you 
> too?

No idea yet. I'll know soon :)

-- 
`When we are born we have plenty of Hydrogen but as we age our
 Hydrogen pool becomes depleted.'

-
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Apache segfaults at every request

[Nix]

On 13 Nov 2006, [EMAIL PROTECTED] said:

> On Sunday 12 November 2006 21:38, Stephane Bortzmeyer wrote:
>> On Fri, Nov 10, 2006 at 11:53:36PM +0100,
>>  Blaisorblade <[EMAIL PROTECTED]> wrote
>>
>>  a message of 34 lines which said:
>> > 2.6 kernels can run on 2.4 systems
>>
>> Each time I try, I get a "FATAL kernel is too old".
> Strange - that is only valid for a software using TLS.

Nah, you can get that message from any glibc if the glibc was compiled
with --enable-kernel= requesting a later kernel than the one being used.

-- 
`The main high-level difference between Emacs and (say) UNIX, Windows,
 or BeOS... is that Emacs boots quicker.' --- PdS

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] NPTL on Host Issue?

[Nix]

On 2 Dec 2006, eTecc Support spake thusly:

> Wow; I feel silly. Enabling UBD support in the UML Kernel. But now I get 
> this:
>
>   ubda: unknown partition table
> kjournald starting.  Commit interval 5 seconds
> EXT3-fs: mounted filesystem with ordered data mode.
> VFS: Mounted root (ext3 filesystem) readonly.
> Kernel panic - not syncing: No init found.  Try passing init= option to 
> kernel.

That means init won't start (maybe a glibc problem, as you suggested:
maybe it's not even there!)

Personally I test ubds by loopback-mounting them (mount -o loop) in the
host first and chrooting into them, making sure that things like bash,
telinit and so forth work. It's a lot easier to fix problems on the
filesystem image if you have a system that can write to it without
relying on its working first :)

-- 
`The main high-level difference between Emacs and (say) UNIX, Windows,
 or BeOS... is that Emacs boots quicker.' --- PdS

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Strange clock problem

[Nix]

On 3 Jan 2007, [EMAIL PROTECTED] outgrape:
> On Tuesday 26 December 2006 11:46, Stephane Bortzmeyer wrote:
>> FATAL kernel is too old
>>
>> (And I cannot change the host at will, only the guest.)
>
> According to Nix, you're using binaries compiled on a more recent 
> distribution:

Not so much `binaries', more `glibc'. (See the definition of
DL_SYSDEP_OSCHECK in glibc-*/sysdeps/unix/sysv/linux/dl-osinfo.h.)

You can fix it by recompiling/reinstalling glibc with the same options
as the distro used, and specifying an earlier kernel version in the
--enable-kernel= configure option.

Of course, you can't run glibc 2.5 on kernel 2.4.x at all if you want
multithreading, because NPTL is the only threading system supported in
that release, and NPTL requires kernel 2.6 (or a patched 2.4). (Often
you can get away with building an older glibc; you can run binaries
compiled against glibc 2.4 and 2.5 on glibc 2.3.[56], as comparatively
few existing symbols received new versions in those releases. There were
a good few new symbols, though... but in practice I'd say
`upgrade'. There are precious few reasons to run a 2.4 kernel these
days.)

-- 
`He accused the FSF of being "something of a hypocrit", which
 shows that he neither understands hypocrisy nor can spell.'
   --- jimmybgood

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] Strange clock problem

[Nix]

On 9 Jan 2007, Paolo Giarrusso told this:
> Nix <[EMAIL PROTECTED]> ha scritto:
>> Not so much `binaries', more `glibc'. (See the definition of
>> DL_SYSDEP_OSCHECK in glibc-*/sysdeps/unix/sysv/linux/dl-osinfo.h.)
>
> My point (which maybe I didn't state clearly) was about static UML
> binaries. I think he's getting:
> $ ./linux
> FATAL kernel too old
> I.e. a problem from a static binary linked with a recent glibc, A

Oh. Right. You're stuck, in that case. (Well, unless you upgrade to a
suitably recent kernel, and it's hard to figure out how recent that
needs to be...)

-- 
`He accused the FSF of being "something of a hypocrit", which
 shows that he neither understands hypocrisy nor can spell.'
   --- jimmybgood

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] /proc/mm in 2.6.23

[Nix]

On 14 Oct 2007, Antoine Martin verbalised:
> It's very rough around the edges but works for me(tm):
> http://uml.nagafix.co.uk/skas-2.6.23.patch.bz2
> (can't enabled visible process cmdlines - some struct has changed)

I'm afraid that on x86-32 I get an extremely efficient coredump from all
UML guests with this patch installed :/ however, I haven't tried
building a skasless 2.6.23 host kernel yet: maybe I get an extremely
efficient coredump there, too.

(2.6.22.x hosts work fine; 2.6.23.x guests work fine.)

-- 
`Some people don't think performance issues are "real bugs", and I think 
such people shouldn't be allowed to program.' --- Linus Torvalds

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] /proc/mm in 2.6.23

[Nix]

On 20 Oct 2007, [EMAIL PROTECTED] said:
> I'm afraid that on x86-32 I get an extremely efficient coredump from all
> UML guests with this patch installed :/ however, I haven't tried
> building a skasless 2.6.23 host kernel yet: maybe I get an extremely
> efficient coredump there, too.

It works with `noprocmm'; thus, a trivial forward-porting of the skas
patch is broken in 2.6.23.

SKAS0 is tolerable (and far better than tt mode!) but it spams my
process accounting logs and is significantly slower than SKAS3 :/

-- 
`Some people don't think performance issues are "real bugs", and I think 
such people shouldn't be allowed to program.' --- Linus Torvalds

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] /proc/mm in 2.6.23

[Nix]

On 21 Oct 2007, Antoine Martin verbalised:

> Nix wrote:
>> It works with `noprocmm'; thus, a trivial forward-porting of the skas
>> patch is broken in 2.6.23.
>> 
>> SKAS0 is tolerable (and far better than tt mode!) but it spams my
>> process accounting logs and is significantly slower than SKAS3 :/
>> 
> AFAIK guests on 64-bit hosts have never worked in full skas3 mode, only
> with noprocmm.

I'm on a 32-bit host, though. No 64-bit anything visible anywhere.

-- 
`Some people don't think performance issues are "real bugs", and I think 
such people shouldn't be allowed to program.' --- Linus Torvalds

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user

Re: [uml-user] skas3 on 2.6.23? switch_mm_skas - PTRACE_SWITCH_MM failed, errno = 22

[Nix]

On 28 Nov 2007, Jeff Dike said:

> On Mon, Nov 26, 2007 at 07:05:21PM -0800, Jason Lunz wrote:
>> Is anyone still using skas3 on modern kernels? I've been booting on
>> 2.6.23 with noprocmm to avoid crashing with "Kernel panic - not syncing:
>> switch_mm_skas - PTRACE_SWITCH_MM failed, errno = 22".
>
> I'm going to take a look at this and get the patch up to speed.

Yay! :)))

-- 
`The rest is a tale of post and counter-post.' --- Ian Rawlings
   describes USENET

-
SF.Net email is sponsored by: The Future of Linux Business White Paper
from Novell.  From the desktop to the data center, Linux is going
mainstream.  Let it simplify your IT future.
http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4
___
User-mode-linux-user mailing list
User-mode-linux-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user