unno if seccomp can do
that, but if it can, as well as read them out which I assume it can,
you're golden.
Jeff
--
Jeff Dike
AddToIt
978-254-0789 (o)
978-394-8986 (c)
--
Mobile secur
On Tue, Dec 06, 2011 at 07:48:40PM +0100, Riccardo Murri wrote:
> Sorry again for resurrecting an old thread, but I each time I look
> into this issue I realize that I haven't quite understood the details...
You basically have it all right.
> In addition, *every* syscall generates a SIGTRAP to th
The thing about DHCP is that the protocol is confined to an ethernet
broadcast domain, and a TUN/TAP connection to the host is a separate
broadcast domain from the LAN. It's a private ethernet with two hosts
on it.
DHCP will work if one of the following is true:
The DHCP server is the UML
On Thu, Oct 13, 2011 at 01:37:24AM +0200, Riccardo Murri wrote:
> Thus:
>
> - batch system schedulers do righteously consider each UML "thread" as
> a separate process;
>
> - however, UML "threads" do share a large portion of the memory, as
> can be seen from this "ps" output:
>
> PID
On Sun, Feb 28, 2010 at 03:06:21PM +0100, Andrea Gelmini wrote:
> 2010/2/27 Jeff Dike :
> Hi Jeff,
>and thanks for your reply.
>
> > Don't you have anything better to do with your time?
>
> It's a dirty job, but somebody has to do it...
To be a bit mor
On Sat, Feb 27, 2010 at 05:51:47PM +0100, Andrea Gelmini wrote:
> arch/um/sys-x86_64/shared/sysdep/skas_ptrace.h:10: ERROR: code indent should
> use tabs where possible
> arch/um/sys-x86_64/shared/sysdep/skas_ptrace.h:11: ERROR: code indent should
> use tabs where possible
Don't you have anythin
On Fri, Sep 04, 2009 at 04:56:21PM +0100, Richard Smith wrote:
> Within my UML guest kernel, I'm finding that between 5% and
> 10% of reads from /dev/hwrng block, seemingly indefinitely.
> (I've certainly seem reads block for more than 12 hours.)
> All other reads complete in significantly less
On Thu, Jun 25, 2009 at 09:46:44PM -0400, lanas wrote:
> I would like to connect to a UML as if with a plain serial port. No
> protocol, no telnet, no nothng, simply like a serial port, at any
> speed, 8N1. I type 'A' in the host's minicom and the minicom in the UML
> receives 'A'.
>
> How wo
On Thu, Jun 25, 2009 at 09:49:20PM -0400, lanas wrote:
> Is there around a simple diagramming app that would allow naming of
> objects such as rectangles and saving them in XML ? The idea would be
> to use such an app to draw several UMLs and their eth ports, connected
> to each other, and then
On Sun, Jun 07, 2009 at 08:33:53PM -0400, Mark Lidd wrote:
> I have been using UML for while with good results, but this lastest
> issue has got me stumped. I put together a UML jail with a perl
> wrapper. It starts running as root and calls compartment to do the
> chroot, change user, and run
> ~/uml/fs:$ ../linux-2.6.29/linux ubda=hd init=sh &> linux-log
> Segmentation fault
> Warning: unable to open an initial console.
Looks like you need to make a /dev/console. But this isn't fatal.
> Kernel panic - not syncing: Attempted to kill init!
Your init exited. Bad init! Doesn't look
On Fri, Dec 19, 2008 at 03:46:15PM +1300, Cameron Kerr wrote:
> On Thu, Dec 18, 2008 at 09:11:19PM -0500, Diego Ramos V. wrote:
> > Please How can I run UML when My host boot,
> > until now I've run UML using
> > UML eth0=tuntap,,,192.168.0.20
> > HOST uml_mconsole WerTVf config eth0=tuntap,,,192.1
On Wed, Nov 12, 2008 at 11:14:25AM +0100, Nicolas Boullis wrote:
> Does anybody know how I should understand this message? Is there a way I
> can fix the problem?
Try doubling or quadrupuling the host's /proc/sys/vm/max_map_count.
Jeff
--
Work email - jdike at li
On Mon, Nov 03, 2008 at 10:22:04PM +, Antoine Martin wrote:
> FWD: I've had no reply so I've pulled the images.
I had a mail server disaster which sort of crimped my style.
However, I did pull the images from nagafix and will host them on SF
as soon as it stops giving me a connection refused
On Thu, Oct 30, 2008 at 11:31:48AM +, Gordon Russell wrote:
> I would still get you to update the host kernels to something newer.
> That way you are sure it is not a kernel version think.
I agree. There were some 32-bit compat ptrace bugs in that era - I
don't remember version numbers - that
On Sat, Aug 30, 2008 at 08:59:34PM +0700, shah khadafi wrote:
> i want to connect my host Ubuntu 8.04 (ip address 192.168.0.254) with my 2
> UML instances (ip address 192.168.0.253 and 192.168.0.252).
> what that must i using to connects between host with my UML?
Are you following some published H
On Fri, Aug 29, 2008 at 10:14:17AM +0800, guodonn wrote:
> I am new to UML, i download the linux-2.6.24-rc7.bz2 and
> FedoraCora5-x86-root_fs.bz2 from http://user-mode-linux.sourceforge.net/
>
> Then i follow the instructions:
>
> [EMAIL PROTECTED] ~]$ ./UML/linux-2.6.24-rc7 ubda=FedoraCora5-x
On Thu, Aug 28, 2008 at 10:37:38PM +0700, shah khadafi wrote:
> i try to run command to setup networking in UML with host Ubuntu 8.04 hardy
> heron, like this
> [EMAIL PROTECTED]:~$ uml_mconsole hP3SgL config eth0=tuntap192.168.0.254
> OK
>
> then, i run command like this
> [EMAIL PROTECTED]:~
On Wed, Aug 27, 2008 at 11:25:02PM +0530, Prasad Joshi wrote:
> [EMAIL PROTECTED]:~/linuxsrc/linux-2.6$ ./linux mem=256 ubda=./root_fs
mem=256M might work a bit better.
Jeff
--
Work email - jdike at linux dot intel dot com
-
On Mon, Aug 25, 2008 at 01:39:15PM +0100, Jay Shah wrote:
> Can UML use more than 4GB of the hosts memory in total (like MySQL
> can't) due to the 32bit limitation? Or, with High Mem, will it support
> upto 64GB?
On 32-bit, no. There is highmem support, but I don't know that it
still works, and i
On Fri, Aug 22, 2008 at 12:02:35PM -0500, Randall Smith wrote:
> ws-uml:~# aptitude install python
> Reading package lists... Done
> Building dependency tree... Done
> Initializing package states... Done
> Reading task descriptions... Done
> Building tag database... Done
> Segmentation fault
>
> A
On Wed, Aug 20, 2008 at 09:56:20AM +0100, Gordon Russell wrote:
> Firstly:
> If I connect and login, then press CTRL-D before terminating telnet,
> it works fine when I telnet in again.
> If I connect and login, but terminate telnet before CTRL-D, then when
> I telnet in again I get the login promp
changes]
Signed-off-by: Tom Spink <[EMAIL PROTECTED]>
Signed-off-by: Jeff Dike <[EMAIL PROTECTED]>
Signed-off-by: Andrew Morton <[EMAIL PROTECTED]>
Signed-off-by: Linus Torvalds <[EMAIL PROTECTED]>
diff --git a/arch/um/include/os.h b/arch/um/include/os.
Jeff
--
Work email - jdike at linux dot intel dot com
commit cfd28f6695d0fc047478480791a21bdd4967f98e
Author: Jeff Dike <[EMAIL PROTECTED]>
Date: Mon May 12 14:01:53 2008 -0700
uml: fix bad NTP interaction with clock
UML's supposed nanosecond clock interacts badly w
On Tue, Aug 19, 2008 at 10:11:38AM +0700, shah khadafi wrote:
> uml_mconsole.c:39:31: error: readline/readline.h: No such file or directory
> uml_mconsole.c:40:30: error: readline/history.h: No such file or directory
These errors should be pretty self-explanatory. You need to install
readline-dev
On Sat, Aug 16, 2008 at 11:53:01AM +0100, Brendan McNally wrote:
> Can someone please confirm that UML will run on any host Linux distribution?
Do you have any reason to believe that it doesn't?
That's sort of a silly question. I can probably build a distro that
can't run UML for some reason. P
On Mon, Aug 18, 2008 at 11:27:23AM -0500, Randall Smith wrote:
> Initializing software serial port version 1
> ubda: unknown partition table
> ubdb: unknown partition table
> kjournald starting. Commit interval 5 seconds
> EXT3-fs: mounted filesystem with ordered data mode.
> VFS: Mounted root
On Wed, Aug 13, 2008 at 11:01:41PM -0500, Shehla Saleem wrote:
> I am using a setup of two systems. One of them has two instances of uml
> running on it with a uml_switch. I want to send TCP traffic from the first
> uml to the second and then outwards from the host to the second system on
> the net
On Sun, Aug 10, 2008 at 12:43:14PM +, Neo Anderson wrote:
> I install uml on 64bits host and encounter a problem related to
> mismatch between host os (opensuse 64bits) and guest os
> (debian32bits). The error echoes "request_module: runaway loop
> modprobe binfmt-464c." After checking on the
On Tue, Aug 05, 2008 at 04:34:23PM -0500, Shehla Saleem wrote:
> Also, this might sound naive, but I am too confused right now. My host
> is on a network which is not a 192.168.0.0/24 but something else. Do I
> have to give an IP address to uml from the same range as the host or
> can I still use a
On Tue, Jul 22, 2008 at 12:44:56AM +0200, vincent-perrier wrote:
> Is there any news about if and eventually in what
> main stream vanilla kernel will we have the pleasure
> to see SKAS4 mode at uml machine start ?
I'm still undecided about the interface. I think I like what I have
right now, e
On Tue, Jul 22, 2008 at 06:31:00PM +0200, Marek Wodzinski wrote:
> On Mon, 21 Jul 2008, Thomas Winkler wrote:
>
> > I want to connect to my UML using a port connection. For that purpose I've
> > added the following to my UML kernel command line:
> >
> > con7=port:9000
> >
> > When I try to telnet
On Thu, Jul 24, 2008 at 06:11:48PM +0200, Toralf Förster wrote:
> Hhm,
>
> running the same command both at the host :
>
> n22 ~ # time factor 819734028463158891
> 819734028463158891: 3 273244676154386297
>
> real0m6.002s
> user0m5.911s
> sys 0m0.000s
>
> and within UML image :
>
>
On Tue, Jul 29, 2008 at 05:53:48PM -0700, David Francisco Rodriguez Perez wrote:
> So I was able to do the unmapping from the user space process page table
> inside UML using:
>
> pgdp = pgd_offset(mm, addr);
> if (pgdp == NULL) return;
> pudp = pud_offset(pgdp, addr);
> if (pudp
On Fri, Aug 01, 2008 at 02:10:07PM +0100, Jay Shah wrote:
> Does anyone know how UML behaves with multi-core systems? Does the
> process glue to one core for as long as it runs for, or use all cores
> whenever it wants?
Depends on the host scheduler and the load. Linux likes cpu affinity,
within
On Fri, Aug 01, 2008 at 03:14:56PM +0200, Lars Ekman G wrote:
> To be able to run gdb on the UML process I would like to run in
> some SKAS >0. After reading this mail-list and the -devel list and
> checking the UML home-page etc. I come to the conclusion that
> SKAS3 is not supported for 2.6.26, a
On Fri, Aug 01, 2008 at 05:08:21PM -0500, Shehla Saleem wrote:
> debug1: SSH2_MSG_KEX_DH_GEX_INIT sent
> debug1: expecting SSH2_MSG_KEX_DH_GEX_REPLY
> Host key verification failed.
> debug1: Calling cleanup 0x8063550(0x0)
This looks to me like an ssh configuration problem. Have you tried
googling
On Mon, Aug 04, 2008 at 02:25:42PM +0200, Benedict Verheyen wrote:
> Look at the thread titled "ssh can't obtain random bytes"
>
> In Debian, when i checked /var/log/auth.log i found these messages:
> sshd[1712]: fatal: Couldn't obtain random bytes (error 604389476)
>
> See if you can find simila
On Tue, Jul 22, 2008 at 11:27:13AM +0300, Boaz Harrosh wrote:
> I'm not sure I fully understand. Could you please explain what you
> found.
arch/um/include/init.h copies enough definitions from linux/compiler.h
and linux/init.h to implement initcall and exitcall et al. It can't
include the kernel
I believe I figured out the ptrace crash that showed up with 2.6.26,
thanks to Toralf Förster.
The patch is below.
I have one confirmation that this fixes the problem - I'd like a few
more just to be sure.
Jeff
--
Work email - jdike at linux dot intel dot com
Index: li
On Tue, Jul 15, 2008 at 01:07:23AM +0200, david.r wrote:
> Kernel panic - not syncing: do_syscall_stub : PTRACE_SETREGS failed, errno = 5
I'm currently chasing this bug. It can be worked around by reverting
4f81c5350b44bcc501ab6f8a089b16d064b4d2f6, but doing that may introduce
some compilation pr
On Fri, Jul 11, 2008 at 03:04:34PM -0700, David Francisco Rodriguez Perez wrote:
> Program received signal SIGSEGV, Segmentation fault.
> 0x0805c2bb in copy_chunk_from_user (from=148668472, len=4040, arg=0xdd8de48)
Tell gdb to
handle SIGSEGV pass noprint nostop
> Other crash!!!
>
> NI15 # l
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
On Thu, Jul 10, 2008 at 12:38:29PM -0500, Diego Ramos V. wrote:
> Hi everyone, thanks for answering me. I've been working with UML in Debian
> 4.0, but I decided to change to Centos, I knew my last mistakes, so I
> checked, I installe d UML, and executable linux was obtained successfull,
> but I w
On Thu, Jul 10, 2008 at 07:48:01PM +0200, Flavio wrote:
> I tried that patch but it seems to have some problem applying it:
What I said was:
> > Below is the most recent
> > patch for mainline along these lines. You can do something similar
> > for skas.h if you feel up to it.
You'll notice I d
something similar
for skas.h if you feel up to it. Otherwise, I'll fix it for the next
skas4 iteration.
Jeff
--
Work email - jdike at linux dot intel dot com
commit 14c8a77e1bbd693446dad297d2ae2dd22f187e4f
Author: Jeff Dike <[EMAIL PROTECTED]>
Date: Thu Jun
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 lat
On Sat, Jun 21, 2008 at 10:34:23PM +0800, Huang Weiyi wrote:
> Removed duplicated include file in
> arch/um/os-Linux/helper.c.
Yup, I passed along a patch from Ingo without noticing that it had
already gone in.
Jeff
--
Work email - jdike at linux dot intel dot com
---
On Wed, Jun 18, 2008 at 09:51:38AM +0200, Laurent Wandrebeck wrote:
> Random patches are still in place. I'm wondering why this does not work...
> Jeff, any idea ?
Is it dumping messages about not being able to get enough entropy?
If you strace UML on the host, do you see it trying to get entropy
On Tue, Jun 17, 2008 at 04:59:52PM -0500, Diego Ramos V. wrote:
> diegorval:/home/diego/build/tools-20070815# make > anex1
> uml_mount.c:1:32: error: fuse/fuse_lowlevel.h: No existe el fichero o el
> directorio
> uml_mount.c: In function 'init_fuse':
> uml_mount.c:9: error: variable 'args' has init
Below is a patch which fixes compilation errors on distros which have
stopped providing asm/user.h.
This was a known problem, which I fixed. Then I forgot, and as part
of a later fix, added back an include of asm/user.h.
This is headed for mainline, but this is causing trouble now, so I'm
public
On Fri, Jun 06, 2008 at 08:26:08PM +0200, Laurent Wandrebeck wrote:
> > diegorval:/home/diego/build/linux-2.6.24.3/tools-20070815# make >
> > /home/diego/util1.txt
> > ar: creando libuml.a
> > uml_mconsole.c:39:31: error: readline/readline.h: No existe el fichero o el
> > directorio
> > uml_mconso
On Tue, Jun 03, 2008 at 10:16:00AM -0700, David Francisco Rodriguez Perez wrote:
> I wanted to know if I can boot a UML Linux executable that has the initramfs
> (initramfs_data.cpio.gz) image built in it? If so what would I need on the
> command line?
I don't think you need anything special. I t
On Thu, Jun 05, 2008 at 06:08:20PM +0200, Benedict Verheyen wrote:
> I found an option "Hardware random number generator" in the UML config
> that sets CONFIG_UML_RANDOM.
> Together with rng-tools and a daemon program rngd, it allows injection
> of the hosts entropy into the guests /dev/random.
>
On Thu, May 29, 2008 at 12:25:20PM +0200, J wrote:
> UML does not manage to start up and exits with the error:
> Checking advanced syscall emulation patch for ptrace...check_ptrace
> : expected SIGTRAP or (SIGTRAP | 0x80), got status = 256
> System:
> Linux localhost.localdomain 2.6.25.3-18.fc
On Thu, May 29, 2008 at 09:09:24AM +0100, David Chanters wrote:
> With a root_fs image i've made, everything works and it will spawn
> three xterms just fine - one for syslog, and another two with gettys
> attached. This is on my home system with a guest UML version 2.6.25.4
> and a host kernel wi
Is this a 32-bit UML on a 64-bit host?
On Sat, May 24, 2008 at 11:52:08AM +0200, vincent-perrier wrote:
> Locating the top of the address space ... 0xd000
This suggests that it is...
> Detected host type: i386
> (GDT indexes 6 to 9)
This suggests that it isn't...
As for the siginfo stuff
On Tue, May 27, 2008 at 07:58:26PM +0100, David Chanters wrote:
> sudo ~/linux-2.6.25/linux devfs=nomount load_ramdisk=1
> ramdisk_size=3 mem=128m initrd=initrd.gz root=/dev/ram0
> ubd0=root_fs ubd1=/home/david/debian-40r3-i386-netinst.iso ubd2=/cdrom
BTW, most distro installers aren't too UML
On Fri, May 23, 2008 at 06:37:58PM +0200, vincent-perrier wrote:
> Here is strace (attached file)
This time, it actually sorta worked...
> write(1, "\tvcpu ... ", 10vcpu ... ) = 10
> msgsnd(4294967295, {4, ptrace: umoven: Input/output error
> 0x4}, 136966304, 0) = -1 ENOSYS (Funct
On Thu, May 22, 2008 at 09:49:32PM +0200, vincent-perrier wrote:
> si_errno and si_code are both zeros, I am sad to say.
That doesn't necessarily mean there's no information. zero might mean
the signal came from a process.
Can you strace it?
Also, send me the full siginfo? There might be somet
On Fri, May 23, 2008 at 01:47:12PM +0200, Klaus Teller wrote:
> modprobe: FATAL: Could not load /lib/modules/2.6.24/modules.dep: No such file
> or directory
>
> Might that be the problem?
No.
Did you notice this:
> helper_child - execvp of 'uml_net' failed - errno = 2
> helper_child - execvp of
On Fri, May 23, 2008 at 08:58:49AM +0200, Benedict Verheyen wrote:
> >>if you restrict the bisection to arch/um
>
> How would i do that?
git bisect start [ [...]] [--] [...]
So, starting with git bisect start -- arch/um would seem to do the trick.
Jeff
--
Work e
On Thu, May 22, 2008 at 07:02:49PM +0200, Benedict Verheyen wrote:
> I tried and i got this error when starting the UML:
> arch_switch_tls failed, errno = EINVAL
> PTRACE_SET_THREAD_AREA failed, err = -22, index = 0
> PTRACE_SET_THREAD_AREA failed, err = -22, index = 0
> arch_switch_tls failed, err
On Thu, May 22, 2008 at 07:40:09PM +0200, vincent-perrier wrote:
> Detected host type: i386
> (GDT indexes 6 to 9)
> clownix: vcpu returned with signal 19
Odd, you're getting SIGSTOPs for some reason, which is consistent with
this bit later:
> - PTRACE_FAULTINFO...
> [1]+ Stopped
On Wed, May 21, 2008 at 11:52:16PM +0200, vincent-perrier wrote:
> vcpu returned with event = 1
> Failed
> Checking for the skas3 patch in the host:
> - /proc/mm...not found: No such file or directory
> - PTRACE_FAULTINFO...
> [1]+ Stopped ./linux root_fs
> bulk$
That's diffe
On Thu, May 22, 2008 at 11:27:28AM +0200, Benedict Verheyen wrote:
> I'm using Debian 4.0 stable (etch) with a vanilla kernel 2.6.24.3 on the
> host and the guest.
Try upgrading the guest to 2.6.25 (the fix I'm looking at -
7b5cc6ee6cf9001775c348bac09814a45f1276b7
was in 2.6.25-rc1).
> S
On Wed, May 21, 2008 at 02:09:20AM +0200, Klaus Teller wrote:
> Right now what I have is the followinge rror message:
>
> SIOCSIFFLAGS: Operation not permitted
What does the guest's dmesg say?
> I found on the web that this might be related to permissions issue
> related to /dev/net/run. I chang
On Tue, May 20, 2008 at 05:40:00PM +0200, Benedict Verheyen wrote:
> whenever i try to stop an uml to take backups
> and afterwards want to start it again with the "go" command, i always
> get the error "Not stopped".
What version? There were some somewhat recent fixes here, and this
seems to wor
rds
> > Vincent Perrier
> Hello Vincent,
>
> the last skas4 patch has been sent to this mailing list from Jeff
> Dike, as well as for the previous one.
FWIW, the full patch can be had from
http://marc.info/?l=user-mode-linux-devel&m=121088437926424&q=raw
and the broke
On Tue, May 20, 2008 at 03:24:58PM -0700, K D wrote:
> [EMAIL PROTECTED]:~/mods$ make V=1
Where's the ARCH=um?
Jeff
--
Work email - jdike at linux dot intel dot com
-
This SF.net email is sponsored b
On Mon, May 19, 2008 at 05:29:19PM -0700, David Francisco Rodriguez Perez wrote:
> I was able to compile my driver using user space library libc.a.
If libc is all you need, then don't bother trying to pull in libc.a at
all. UML needs libc, so you'll get it anyway.
> So I wanted to compile the d
On Fri, May 16, 2008 at 11:35:20AM -0700, David Francisco Rodriguez Perez wrote:
> I am working on a simulator of a product and so far we have decided to go
> with UML. I have so far 2 questions.
> 1. There is a network namespace patch that is coming into the mainline in
> 2.6.26. Is this patch goi
On Fri, May 16, 2008 at 12:19:43AM +0200, vincent-perrier wrote:
> I use non-patched kernels for the host and for the
> guest, is there hope that in the future, simple
> users that do not patch their hosts (or guest)
> kernels will get the ska4?
Yup, the hope is that the skas4 interfaces are accep
Tidy the existing stub code so that the skas4 code can be dropped in.
diff --git a/arch/um/os-Linux/skas/mem.c b/arch/um/os-Linux/skas/mem.c
index 484e68f..5f09a9b 100644
--- a/arch/um/os-Linux/skas/mem.c
+++ b/arch/um/os-Linux/skas/mem.c
@@ -40,35 +40,69 @@ static unsigned long syscall_regs[MAX_R
/skas_ptrace.h
@@ -1,5 +1,5 @@
/*
- * Copyright (C) 2000, 2001, 2002 Jeff Dike ([EMAIL PROTECTED])
+ * Copyright (C) 2000 - 2008 Jeff Dike ([EMAIL PROTECTED],linux.intel}.com)
* Licensed under the GPL
*/
@@ -7,19 +7,10 @@
#define __SKAS_PTRACE_H
#define PTRACE_FAULTINFO 52
-#define
--- a/arch/um/kernel/syscall.c
+++ b/arch/um/kernel/syscall.c
@@ -1,17 +1,17 @@
/*
- * Copyright (C) 2000 - 2007 Jeff Dike ([EMAIL PROTECTED],linux.intel}.com)
+ * Copyright (C) 2000 - 2008 Jeff Dike ([EMAIL PROTECTED],linux.intel}.com)
* Licensed under the GPL
*/
-#include "linux/file.h"
-
UML already had a new_mm(). This is renamed to make_new_mm().
diff --git a/arch/um/include/skas/skas.h b/arch/um/include/skas/skas.h
index d6cbb4f..061a362 100644
--- a/arch/um/include/skas/skas.h
+++ b/arch/um/include/skas/skas.h
@@ -17,7 +17,7 @@ extern int skas_needs_stub;
extern int user_thr
Macroize stub system call execution loop
Extract the system call execution loop in the stubs into a macro
A later patch will need this loop, so this makes it available.
diff --git a/arch/um/sys-i386/stub.S b/arch/um/sys-i386/stub.S
index 7699e89..a63397f 100644
--- a/arch/um/sys-i386/stub.S
+
;
+ entry->entry_number = i + GDT_ENTRY_TLS_MIN;
+ err = get_thread_area(entry);
+ if (err) {
+ perror("get_thread_area");
+ exit(1);
+ }
+ }
+
+ err = ptrace(PTRACE_GETFPXRE
Tidy arch/um/os-Linux/start_up.c in preparation for host SKAS4 testing.
There are now explicit variables for disabling any particular host
capability.
init_registers is moved out of the skas3 testing because it needs to
happen every time, and the skas3 checks will become optional.
diff -
UML guest SKAS4 support.
diff --git a/arch/um/include/as-layout.h b/arch/um/include/as-layout.h
index cac542d..929d053 100644
--- a/arch/um/include/as-layout.h
+++ b/arch/um/include/as-layout.h
@@ -23,16 +23,15 @@
*/
#ifdef __ASSEMBLY__
-#define _AC(X, Y) (Y)
+#define _C(Y) (Y)
#else
-
copy_siginfo_to_user correctly strips off the high bits of si->si_code
before copying it to userspace.
Unfortunately, ptrace32_siginfo calls copy_siginfo_to_user to put the
siginfo into a temporary userspace buffer and counts on the high bits
still being present so that it may initialize the u
Add CPU trap number and error code to siginfo_t in the SIGSEGV case for x86.
diff --git a/arch/x86/ia32/ia32_signal.c b/arch/x86/ia32/ia32_signal.c
index 5e7771a..a2a4c1c 100644
--- a/arch/x86/ia32/ia32_signal.c
+++ b/arch/x86/ia32/ia32_signal.c
@@ -68,6 +68,8 @@ int copy_siginfo_to_user32(compat_
There were a number of repetitions of the calculation of an stub
address in a process address space. This introduces STUB_ADDR to
eliminate the common code.
diff --git a/arch/um/include/skas/skas.h b/arch/um/include/skas/skas.h
index 061a362..331f343 100644
--- a/arch/um/include/skas/skas.h
+++ b
requires using CPP
to rename the old one out of the way.
diff --git a/arch/um/include/siginfo_segv.h b/arch/um/include/siginfo_segv.h
new file mode 100644
index 000..c000267
--- /dev/null
+++ b/arch/um/include/siginfo_segv.h
@@ -0,0 +1,133 @@
+/*
+ * Copyright (C) 2002- 2007 Jeff Dike ([EMAIL
On Tue, May 13, 2008 at 10:49:36PM -0700, K D wrote:
> > [EMAIL PROTECTED] rdp]# make -C /oldext/downloads/uml-krnl/linux-2.6.24
> > M=`pwd` modules
Can you add "V=1" to the make command?
In /oldext/downloads/uml-krnl/linux-2.6.24, you should have
arch/um/include/sysdep pointing to arch/um/inclu
On Fri, May 09, 2008 at 06:29:39PM -0700, K D wrote:
> When I compile my module with UML kernel source tree I get these errors. I
> must be doing something silly:
>
> [EMAIL PROTECTED] rdp]# make -C /oldext/downloads/uml-krnl/linux-2.6.24
> M=`pwd` modules
> make: Entering directory `/oldext/dow
On Fri, May 09, 2008 at 04:15:52PM +0200, vincent-perrier wrote:
> When I do:
> "date", then I wait a few seconds, then "date --set" with the
> value I had in the first date, then the UML machine
> takes 100% of cpu.
> This occured 5 times out of 5 tries with a 2.6.25.1 kernel uml
> and 2.6.22.15 h
On Wed, May 07, 2008 at 11:57:22PM -0700, K D wrote:
> Coming to my problem. I'm developing a kernel module and I hate to
> reboot my host OS everytime there is any issue in module. I'm hoping
> UML would help me here. When I load my module in host OS, it
> loads/unloads without any issue. However
On Sat, May 03, 2008 at 10:03:36PM +0200, vincent-perrier wrote:
> Hello,
> I want to compile a module with ARCH=um, I get some errors,
> it may come from me as usual, but just in case, here are the errors:
>
> arch/um/include/sysdep/ptrace.h:9:24: error: uml-config.h: No such file
> or directory
On Sat, May 03, 2008 at 08:02:55PM +0200, Martin Kuhn wrote:
> is the PowerPC Architecture supported by user-mode-linux? (I want to run
> uml on a mac mini G4)
Nope, it's i386 and x86_64 only at this point.
Jeff
--
Work email - jdike at linux dot intel dot com
On Sat, May 03, 2008 at 11:42:04AM +0200, Rezarta Jaupi wrote:
> Could anyone help me to examine the network traffic and the network
> throughput of IP layer (chained) puzzle protocol with User mode Linux? I
> want to create 2 scenario in Linux 2.6.12 one with 2 hosts and other with
> 5 hosts, but
On Thu, May 08, 2008 at 09:42:33PM +0800, Huang Weiyi wrote:
> Removed duplicated include file "kern_util.h" in
> arch/um/drivers/ubd_kern.c.
Looks good, thanks.
Jeff
--
Work email - jdike at linux dot intel dot com
-
On Tue, Apr 29, 2008 at 06:35:06PM +0200, Flavio wrote:
> Any news?
The current state of things as I understand them is that we have the
eject ioctl being called in exactly the same way from eject and from
UML and it behaves differently. This, of course, shouldn't happen, so
there must be somethi
On Tue, Apr 29, 2008 at 09:19:55PM -0400, lanas wrote:
> Thanks. It makes me think that I saw once that there's an option to
> do a make modules_install with an actual target directory instead of
> assuming that the modules must be on the same filesystem as the sources
> of the kernel.
INSTALL_M
On Tue, Apr 29, 2008 at 10:27:54AM +0200, Daniel Janzon wrote:
> Thanks for the information. Two follow-up questions though. First, are
> there any certain features of the kernel that can be turned off that
> makes the porting effort substantially smaller?
Turn off hostfs.
You can make the small
On Mon, Apr 28, 2008 at 04:55:44PM -0400, Jonas Meyer wrote:
> I don't seem to have a "slabinfo" in my /proc directory. According to
> some google searching, this is due to a bug in ubuntu gutsy. Should I
> upgrade to hardy to fix this so I can give you the data you want, or
> would that potentia
On Mon, Apr 28, 2008 at 01:09:48PM -0400, Jonas Meyer wrote:
> I'm not sure if this is a UML issue, but my machine is acting very
> strange, so I'm wondering if it is.
>
> I have a guest that has 512Mb of memory assigned to it. After running
> mysql, apache, and mongrel (ruby on rails application
On Thu, Apr 24, 2008 at 03:05:13PM +0200, Joel Palmius wrote:
> I've been trying to get humfs to work, but I can't find it in 2.6.25.
I never sent it to mainline because of bugs I never fixed.
> Is humfs a patch you need to download or something?
Yup - see http://user-mode-linux.sourceforge.net
On Fri, Apr 25, 2008 at 04:02:26PM +0200, Daniel Janzon wrote:
> Are there any reasons UML will not be able to run on for instance ARM
> platforms?
Nope.
> I see in arch/um that there is some processor-dependent code
> for x86 and ppc. Is this code strictly needed in order to run UML? Or is
> on
1 - 100 of 903 matches
Mail list logo
