On Tue, Oct 03, 2006 at 08:18:46AM +0200, Michael Sternberg wrote:
Two questions:
1. What the situation in Linux ? Is free returns memory to OS ?
No. free() returns memory to the standard C library, which may or may
not then return it to the OS.
2. If not - what the proper call to flush
On Tue, Sep 26, 2006 at 04:12:55PM +0300, Dan Shimshoni wrote:
perhaps is this mkufs problem (or because
of running it in conjunction with loop device, which is less
likely).
It's not the loop device, I get the same thing with a raw partition.
Cheers,
Muli
On Tue, Sep 26, 2006 at 02:56:10PM +0300, Dan Shimshoni wrote:
mount -o loop -t ufs /test/disk-image /loopDir
I get:
mount: wrong fs type, bad option, bad superblock on /dev/loop1,
or too many mounted file systems
(could this be the IDE device where you in fact use
On Tue, Sep 26, 2006 at 04:32:08PM +0300, Rami Rosen wrote:
Hi,
Try ./mkufs -O 1 /test/disk-image instead of
/mkufs /test/disk-image
there are 2
file system formats for mkufs : 1 = UFS1, 2 = UFS2
see also : mkufs -help.
This should work.
I doubt it (did you try it?). It's obvious
On Tue, Sep 26, 2006 at 03:47:28PM +0300, Dan Shimshoni wrote:
Did anyone had success with this ?
Ok, I took a quick look at it. It appears that mkufs and fs/ufs/ do
not agree about the superblock location and/or the magic number:
UFSD (/home/muli/w/iommu/calgary/linux/fs/ufs/super.c, 612):
On Tue, Sep 26, 2006 at 09:02:44PM +0300, Rami Rosen wrote:
Hi Muli,
I doubt it (did you try it?).
Now I tried it, on Fedora Core 4 (on x86).
It works like a charm! feel free to try it ...
I did, on Suse 9 SP2 (x86-64) with a 2.6.18 kernel with both -O1 and
-O2 and neither worked. Maybe
On Thu, Sep 21, 2006 at 01:38:17PM +, Oded Arbel wrote:
Can Xen run both x86-64 and x86 (ia32) domU OSs at the same time?
(on x86-64 hardware, of course. no hardware VT)
I've looked in Google, but apparently I don't know how to ask the
question properly :-(
You'd usually get better
On Thu, Sep 21, 2006 at 08:30:42PM +0300, Oded Arbel wrote:
If I want to build some servers using Xen, which needs to run several
flavors of Linux - lets say various RedHat based distros - and I have
not used Xen before, nor do I want to invest too much into learning it
(I have other stuff I
On Thu, Sep 07, 2006 at 04:48:59PM +1000, Amos Shapira wrote:
You follow Digg too? :)
reddit is better.
/me runs
Cheers,
Muli
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g.,
On Mon, Aug 28, 2006 at 11:21:36PM +0300, Gil Freund wrote:
I was asked about putting a firewall in Virtual Server environment.
As far as I can tell, XEN will allow me to assign a NIC as a PCI
desvice to a DomU.
Yes, but that makes the domU trusted (an attacker with root access ot
the domU
On Tue, Aug 29, 2006 at 12:17:53AM +0300, Shachar Shemesh wrote:
If you do set it up like that (and I did), please be sure to turn off
hardware checksum generation for TCP/IP, or you'll have trouble
connecting from the Xen machines that are behind the firewall to the
internet
There were
On Tue, Aug 29, 2006 at 11:36:27AM +0300, Shachar Shemesh wrote:
Solution: RTFM ethtool for the command line option to disable hardware
offloading of checksum calculation.
ethtool -K ethN tx off
Muli,
This seems to me like a conceptual bug in the way Xen determines which
is the true
On Tue, Aug 29, 2006 at 08:37:13PM +0300, Gil Freund wrote:
Yes, but that makes the domU trusted (an attacker with root access ot
the domU can easily take down the entire machine - unless you have an
isolation capable IOMMU)
Can you elaborate a little more? Does this mean that if the
On Mon, Jul 31, 2006 at 09:12:09AM +0300, Geoffrey S. Mendelson wrote:
On Mon, Jul 31, 2006 at 08:16:11AM +0300, Muli Ben-Yehuda wrote:
Everyone loses from binary only modules. I point you again to
http://www.kroah.com/log/linux/ols_2006_keynote.html as well as
http://lwn.net/Articles
On Sun, Jul 30, 2006 at 07:13:24PM +0300, Geoffrey S. Mendelson wrote:
To keep this relevant to Linux, there are also sorts of accomodations made
in the Linux Kernel and applications for varying sofware licenses. Just
because
the Kernel itself is GPL, that does not prevent someone from
On Sun, Jul 30, 2006 at 09:04:12PM +0300, Geoffrey S. Mendelson wrote:
On Sun, Jul 30, 2006 at 07:32:52PM +0300, Muli Ben-Yehuda wrote:
Dual licensing is subtle, but doable. I'm not sure what totaly
freeware means from a license point of view, but OCO (more commonly
referred to as binary
On Sun, Jul 30, 2006 at 11:01:28PM +0300, Diego Iastrubni wrote:
?, 30 ? 2006, 22:51, ?? ??? Muli Ben-Yehuda:
Just because something is illegal does not mean people do not indulge
in it. It does mean they shouldn't, though. In my humble opinion it is
only a matter of time
On Thu, Jul 27, 2006 at 12:36:22PM +0300, Michael Vasiliev wrote:
Hello all,
I believe these might interest some of you. OLS 2006 recordings, courtesy of
Shawn Starr.
http://christian-leber.de/~ijuz/ols2006.torrent
Since you mentioned it, proceedings are at
On Mon, Jul 24, 2006 at 02:18:18PM +0300, Leonid Podolny wrote:
Hi,
Either I'm missing something very obvious, or udev doesn't support
generic (i.e. non-tty, non-dev-null, etc) character devices. Can it
possibly be? As a matter of fact, I tried to compare the state of the
/sys tree before and
On Mon, Jul 10, 2006 at 12:04:37PM +0300, Shachar Shemesh wrote:
There are two parts to this equation (para vs. fully virtualized host).
One is the use of some impossible to easily virtualize CPU commands on
normal pentium CPUs. The two workarounds are:
1. Tell the virtualized machine not to
On Mon, Jul 10, 2006 at 12:55:10PM +0300, Gilboa Davara wrote:
Hello all,
I'm thinking about screwing around with the Linux VMM layer (mostly
I/O/pagecache/etc).
Can anyone point me to a good book about it?
Mel Gorman's Understanding the Linux Virtual Memory Manager is a great
code walk
On Tue, Jul 04, 2006 at 04:17:04PM +0300, Michael Ben-Nes wrote:
If you have Hardware virtualization like Intel vt-x you can use Xen.
And the performance overhead is from 0.3% to 5% ( it will kick vmware butt )
These numbers are for para-virtualized guests, not fully virtualized
guests (and
On Mon, Jul 03, 2006 at 09:59:13AM +0100, Andre Bar'yudin wrote:
In big companies it is much more complicated.
Not necessarily - it all depends on the company. I work for one of the
biggest, and all you need to do to get a Linux desktop is ask for
it. It's supported and in some departments,
On Mon, Jul 03, 2006 at 11:03:27AM +0300, guy keren wrote:
note: i only worked in small companies, which tend to be flexible about
these issues. they say for a lousy 3000 shmeckels i don't intend to
reject a work candidate that i want to employ. note that this is simply
part of the standard
On Mon, Jul 03, 2006 at 12:36:09PM +0100, Andre Bar'yudin wrote:
Are we talking 10,000+ employees?
330,000 or so.
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the
On Sun, Jul 02, 2006 at 01:53:01PM +0300, Orna Agmon wrote:
This Monday, at 18:30, Haifa Linux Club will once again gather to hear
Muli Ben-Yehuda talk about
Utilizing IOMMUs for Virtualization in Linux and Xen
The slides are now available at
http://www.mulix.org/lectures/using-iommus
On Sun, Jul 02, 2006 at 01:53:01PM +0300, Orna Agmon wrote
The talk will be given in English.
Actually - I'll do it in Hebrew, unless there's someone who prefers to
hear it in English. Pig latin optional.
Cheers,
Muli
=
To
On Sun, Jul 02, 2006 at 02:12:39PM +0300, Gilad Ben-Yossef wrote:
Howdie list,
I'm trying to build the be-old, end-all exception signal signal handler for
SIGSEGV, SIGILL, SIGFPE and their ilk. One thing that will be useful to do
in such a handler is dump the state of the CPU registers
On Fri, Jun 30, 2006 at 11:15:49PM +0100, Andre Bar'yudin wrote:
What's the official repository for OS and Linux freelancers, then?
google, of course.
Cheers,
Muli
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word
On Wed, Jun 28, 2006 at 04:45:09PM +0300, Shachar Shemesh wrote:
Hi list,
Is there a way to turn off hardware calculation of the TCP checksum?
ethtool -K interface rx off tx off
=
To unsubscribe, send mail to [EMAIL
On Wed, Jun 28, 2006 at 05:00:46PM +0300, Shachar Shemesh wrote:
As usual, you have all been a great help
Careful, we might start charging you the going consultancy rate :-)
Cheers,
Muli
=
To unsubscribe, send mail to [EMAIL
On Sun, Jun 25, 2006 at 10:59:33AM +0300, Ami Chayun wrote:
This is an extremely bad move. 'Headless' servers with no keyboard and mouse
have very few ways to create random entropy. Since the server in question is
a web server, it both generates little amount of entropy via disk events, and
On Wed, Jun 14, 2006 at 08:20:20AM +0300, Marc A. Volovic wrote:
Quoth Marc A. Volovic:
I use Xen with Debian (woody, sarge and etch) with not problem whatsoever.
The dom0 is etch, domUs are a mix, all very very small (sub 60MB).
That is - xen 2.4. All my attempts to build xen 3 so far
On Wed, Jun 14, 2006 at 09:12:44AM +0300, Marc A. Volovic wrote:
Quoth Muli Ben-Yehuda:
More details please... with xen3 we switched to a single kernel for
both dom0 and domU, the -xen kernel. Are you passing
'KERNELS=...-domU' perchance? I wouldn't expect that to work.
Duh! I am. Haha
On Wed, Jun 14, 2006 at 01:01:23PM +0300, Tzahi Fadida wrote:
I am curious,
When you say windows xp emulation, you mean by XEN itself
because of the VT feature
Yes.
or in QEMU or VMWARE in the XEN
guest system. If it is directly with XEN, can the same thing be
done with AMD pacifica which
On Tue, Jun 13, 2006 at 06:00:39PM +0300, Gilad Ben-Yossef wrote:
The Fedora Core line is a non production distribution *by design*.
Since when has this ever stopped a customer from actually using it in
exactly that capacity. I heard a talk at USENIX last week[1] where
Pixar's VP of Technology
On Tue, Jun 13, 2006 at 06:41:37PM +0300, Marc A. Volovic wrote:
Speaking of which - whoever among you people is responsible for Xen's build
system should be compelled to use FC.
That would be the Cambridge folks. They've proven remarkably resistant
to attempts to fix it.
Cheers,
Muli
On Tue, Jun 13, 2006 at 06:25:55PM +0300, Shachar Shemesh wrote:
Not to rub it in for anyone (hi Ira), even as I'm writing this email
we're running a Windows XP installation inside a XEN VM running on
Debian.
Which CPU and where did you buy it?
And for the sake of technical accuracy (not to
On Wed, Jun 14, 2006 at 12:16:44AM +0300, Oleg Goldshmidt wrote:
I will not defend the Xen build system, but FC is one of the platforms
(RHEL is another) that Xen can actually live with. So far I cannot say
the same about, say, SuSE (never heard of anyone who managed to have a
working network
On Wed, May 17, 2006 at 03:27:43PM +0300, Oleg Goldshmidt wrote:
Gilad Ben-Yossef [EMAIL PROTECTED] writes:
Muli Ben-Yehuda wrote:
.. but that requires a proprietary kernel module, which is we know,
is The Root of All Evil.
So does VMware.
They do provide sources... I patched
for the time being. That something else is usually
VMWare.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe
VNC support built in. The speed with KQEMU is
not bad (not as good as VMWare though).
.. but that requires a proprietary kernel module, which is we know,
is The Root of All Evil.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
someone who can't
tell AJAX from WINDEX edited your resume to fit the job, in the
extremely improbable event that you'll get it.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe
is different.
Maybe the most practical thing to do is to ASK which format is
preferred.
Ah, but that's sensible. We don't do sensible here.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
://www.appleintelfaq.com/#10.1 and on the Xen mailing list. You
might want to re-confirm.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word
On Wed, May 17, 2006 at 09:28:07AM +1000, Amos Shapira wrote:
On 5/16/06, Muli Ben-Yehuda [EMAIL PROTECTED] wrote:
http://wiki.xensource.com/xenwiki/IntelVT
Thanks for the pointer.
Right now I still don't see any mention of availability of the AMD
equivalent, only to be revealed
and not read the message yet.
There's no way to select or poll message queue, unfortunately, so you
have to manually poll it every so often. To poll, use either a
non-blocking msgrcv() or msgctl(qid, IPC_STAT, buf) and look at
buf.q_qnum. If it's 0, you have a message waiting.
Cheers,
Muli
--
Muli Ben
each time. The first
message is used to wake up the receiving process and then discarded. It
could be zero-length message.
Isn't this racy - if the sender context switches between the two
sends, the receiver may block on the second receive.
Cheers,
Muli
--
Muli Ben-Yehuda
http
:-)
Nit-pickingly-yours,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo
On Tue, May 02, 2006 at 10:34:14PM +0300, Yedidyah Bar-David wrote:
How is it tunable? A quick google search finds (for me) only an external
patch providing /proc/sys/fs/pipe-sz. Is there such a thing in
vanilla?
It's not tunable - the kernel will just do the right thing. It is an
area in
On Tue, May 02, 2006 at 09:16:14PM +0300, Yedidyah Bar-David wrote:
When you run 'prog1 | prog2', both run in parallel, having a small
buffer (IIRC in Linux one page = 4KB, not sure if tunable)
Maximum capacity is 16 pages at the moment.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]
the compiler worse for AMD than it would
be otherwise. In other words, making it better for Intel's CPUs is
fine, making it worse for AMD's isn't.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
. If you want a quick and
dirty solution - system(ldd /proc/$PID/exe).
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe
!)
Attendance is free of charge, and everyone are welcome.
s/are/is/
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe
, books. Roughly
in order of usefullness.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]
On Mon, Mar 06, 2006 at 08:47:32AM +0200, Gilad Ben-Yossef wrote:
I wonder if enough people are interested in a talk about early user
space.
+1 from me. I'm always interested in kernel talks.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
. This happened to me with 2.6.13.1 and 2.6.15 and 2.6.15.4.
What should I check? I'm running Debian SID, BTW.
make version? assuming it's 3.81rc1, it's a known make bug - see
http://marc.theaimsgroup.com/?l=linux-kernelm=114150857516839w=2
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org
having an old kernel source tree. Looks like Debian SID users have
already stepped on that rake.
.. and the universe blow up? not quite, you just suffer a few full
recompilations. I'm sure someone will come up with the appropriate fix
for every kernel version.
Cheers,
Muli
--
Muli Ben-Yehuda
http
then shift-shift refused to work,
but alt-shift worked fine.
Option XkbOptions grp:switch,grp:alt_shift_toggle,grp_led:scroll
/me shrugs.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
into 2.4-compatibility mode,
or another creative idea?
No way to do the former; for the latter, just wrap it up with
ptrace or LD_PRELOAD and intercept the calls to open, read and close
/proc/stat.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
announcements
are fine but only every so often.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run
have fun :-)
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe
Last time I checked, with no hardware VT extensions, Xen cannot run
Windows.
Am I wrong?
No, Xen does require VT (or AMD's SVM) to run Windows. But VT enabled
processors are now available, and SVM will be available soon.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http
).
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL
of the Intel Pentium's ability to support a secure virtual
machine monitor, Robin J, Irvine C,
http://citeseer.lcs.mit.edu/robin00analysis.html
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
, but reports from the lab say it works. xen-unstable
tree includes both VT and SVM support, in a unified framework called
'hvm' which was merged a couple of days ago (just after 3.0.1 was
released).
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
to patch your kernel. I don't think
there is any distribution that ships kernels pre-patched with this
patch, but I may be wrong on this one.
Got a URI for the patch?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
anyway.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe
on a uni-processor
system.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
does the GPL help me here...? I still have no way to really
trust that application
I suggest rereading Reflections on Trusting Trust.
Also, http://free.oszoo.org/
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
,
provided you use a file that has an in-memory backing store
(e.g. resides on tmpfs), your buffers are page aligned,
etc. http://lwn.net/Articles/24468/. I am entirely unconvinced however
that the cost of the page table fiddling is worth it for simplifying
the processing.
Cheers,
Muli
--
Muli Ben
to create x86 selectors to
test some 16 bit addressing and 32 bit with segmentation cases.
I assume that wine uses is for similar reasons, but haven't checked.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
if this is purely an implementation issue or whether
there's a reason why TASK_UNMAPPED_BASE isn't below 1GB in the virtual
address space?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe
On Tue, Dec 13, 2005 at 04:14:04AM +0200, guy keren wrote:
now, why didn't anyone tell me of this one before?
We were too busy posting about it to our blogs ;-)
http://www.technorati.com/search/everybody+loves+esr
http://www.livejournal.com/users/mulix/163495.html
Cheers,
Muli
--
Muli Ben
inefficient
(O(n^2) worst case, match every substring against every substring):
http://aspn.activestate.com/ASPN/Cookbook/Rx/Recipe/326097
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
, it will not bother with the remaining characters.
shouldn't that be possible with an NFA as well? if I remember automata
theory correctly, NFAs and DFAs are completely equivalent in their
descriptive power and can be mechanically traslated from one form to
the other.
Cheers,
Muli
--
Muli Ben-Yehuda
http
for the explanation. I look forward to seeing the code.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message
I'm forwarding this for a colleague. Please CC Marcel on your
replies.
Hi,
I'm looking for a version of pthreads for Linux that has the API
pthread_atfork implemented.
I'm using an Intel Xeon Hyperthread machine with two 3.20GHz CPUs.
Kernel version is 2.6.4-52.smp. Current pthread/glibc
sibling CPUs (one
physical CPU) differently than two physical CPUs, aka HT aware
scheduler. See http://lwn.net/Articles/8720/.
AFAICS this was added very early in the 2.6 development cycle, so I
don't know what RH means by the implementation being non-comparable.
Cheers,
Muli
--
Muli Ben-Yehuda
the Xeons and use the AMD Opteron instead), though the
cache trashing does sound reasonable.
What was the test? HT compared to what? two physical CPUs? single
physical CPU?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
On Mon, Dec 05, 2005 at 07:43:28PM +0200, Gilboa Davara wrote:
On Mon, 2005-12-05 at 17:49 +0200, Muli Ben-Yehuda wrote:
On Mon, Dec 05, 2005 at 05:05:04PM +0200, Gilboa Davara wrote:
The application we tested on was a kernel based network system that did
a lot of I/O (with network
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail [EMAIL PROTECTED]
framework have hooks at
the appropriate VFS level.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g
translation.
As long as you can find professional programmers at $smallnum dollars,
you won't have a business proposition, because a free community effort
will depend precisely upon the people who make a living from
programming.
Need I say more?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org
Arcangeli's cpushare[0], and specifically seccomp[1].
[0] http://www.cpushare.com/about
[1]
http://kernel.org/hg/linux-2.6/?cmd=file;file=kernel/seccomp.c;filenode=e872174f0339a7410f6d429bb5318426f59ed569
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
On Fri, Nov 11, 2005 at 03:16:49AM +0200, Omer Zak wrote:
But how much effort will be needed to make it work with current
kernels?
A month or two of after-hours hacking to rewrite the two kernel
modules. Any volunteers? :-)
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http
accesses is
inotify. See e.g http://www.linuxjournal.com/article/8478
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe
a flame war, but mostly open is like mostly
dead or mostly pregnant, just not good enough. It's either open or it
isn't, and if it isn't, it's not interesting.
RMS-ly yours,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
.
That is your prerogative. Just please don't post any bug reports to
the kernel mailing list when your nvidia driver breaks and/or breaks
some other unrelated part of the kernel.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
for cooperation. Almost as much as I'm against closed source
drivers.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe
that correspond to those values. You might not find them either,
in which case check your kernel sources.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL
and you're
not expected to be doing it; ask yourself why do you need to do it,
and are there existing mechanisms that do what you need? the answer is
almost certainly yes.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
they do if you must. (the size is the size of the module in
bytes, but I don't recall at which stage - older modutils actually
implement most of a dynamic linker in userspace, newer modutils do it
in the kernel).
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
If anyone who attended go-linux today feels like writing a short
summary of what transpired, it will be appreciated.
Terminally curiously yours,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
On Tue, Nov 01, 2005 at 04:13:59PM +0200, Shlomo Solomon wrote:
Just got back. Unfortunately, I don't have the time to do a full summary, but
I can make a few points:
[snipped]
Thanks! how many people would you estimate were there?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org
?
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com/
=
To unsubscribe, send mail to [EMAIL PROTECTED] with
the word unsubscribe in the message body, e.g., run the command
echo unsubscribe | mail
is a special way to open files
that has some limitations, but is much faster (and in some situations,
can be easier to program with).
It's not faster to open the file, but it can be faster than seek +
read/write, although for most workloads, it's not.
Cheers,
Muli
--
Muli Ben-Yehuda
http
words, if the application is kind enough to
tell the system how many pages it will need, the system, if ulimit -m
is set, will tell it yes or no. Not very useful, I agree.
Cheers,
Muli
--
Muli Ben-Yehuda
http://www.mulix.org | http://mulix.livejournal.com
101 - 200 of 783 matches
Mail list logo
