On 2/23/11 5:55 AM, Martin Koegler wrote:
> Why does the trace/error infrastructure of libjpeg puts all parameters
> for formating the message into a structure and then calls via a
> pointer a functions (emit_message), which determines, if the message
> should be printed?
Not sure. That is inheri
On Sun, Feb 20, 2011 at 06:27:02PM +0100, Martin Koegler wrote:
> I did some profiling:
>
> With TLS, the vncviewer hot spots are:
> * ~23% in libgcrypt
> * ~14% in the jpeg code
> * ~14% again in the jpeg code
>
> With TLS, the Xvnc hot spots are:
> * 17.4% in zlib
> * 14.3% libcrypt
> * 13% rfb
On Sun, Feb 20, 2011 at 11:45:37AM -0600, DRC wrote:
> On 2/20/11 11:27 AM, Martin Koegler wrote:
> > If somebody wants to improve the overall performance of Xvnc,
> > ComparingUpdateTracker::compareRect could be of interest.
>
> It's on my radar for sure. One of the things I had to do for TurboV
On 2/20/11 11:27 AM, Martin Koegler wrote:
> Adding SSH Tunnling to the comparison would be interesting (the CPU
> time of ssh must be added to the server/client time, as the SSH
> process does the crypto operations on behalf of TigerVNC).
To be clear, the CPU time in my measurements was the syste
On Sat, Feb 19, 2011 at 04:17:14PM -0600, DRC wrote:
> On 2/18/11 4:55 AM, Adam Tkac wrote:
> Definite improvement. Here's what I now see with VirtualGL:
>
> Machine 1: Pentium 4 Xeon 2.8 GHz single core
> Machine 2: AMD X2 5050e 2.6 GHz dual core
>
> 1-->2
> -
>
On 2/18/11 4:55 AM, Adam Tkac wrote:
>> The attached patch improves frame rates to 9-11 fps.
>
> Good catch, I improved your patch a little and commited it as r4295.
>
> With the patch GNUTLS debug routines are registered only when we
> need their output.
Definite improvement. Here's what I now
On Fri, Feb 18, 2011 at 08:59:50AM +0100, Martin Koegler wrote:
> On Thu, Feb 17, 2011 at 09:06:33AM +0100, Martin Koegler wrote:
> > On Wed, Feb 16, 2011 at 03:21:38AM -0600, DRC wrote:
> > > However, I did discover that playing the 1920x1080 version of Big Buck
> > > Bunny from the beginning usin
On Thu, Feb 17, 2011 at 09:06:33AM +0100, Martin Koegler wrote:
> On Wed, Feb 16, 2011 at 03:21:38AM -0600, DRC wrote:
> > However, I did discover that playing the 1920x1080 version of Big Buck
> > Bunny from the beginning using VLC in full-screen mode makes a good 2D
> > benchmark. The beginning
On Wed, Feb 16, 2011 at 03:21:38AM -0600, DRC wrote:
> However, I did discover that playing the 1920x1080 version of Big Buck
> Bunny from the beginning using VLC in full-screen mode makes a good 2D
> benchmark. The beginning of this movie has full-scene motion on almost
> every frame, so if you s
On Mon, Feb 14, 2011 at 04:37:16PM -0600, DRC wrote:
> IIRC, there is some way to override this path, either at build time or
> run time or both. I will look into it and see if I can build the DRI
> library into the distribution. I didn't really care about it because I
> always just use VirtualGL
On 2/14/11 3:59 PM, Martin Koegler wrote:
>> OK, let me see if I can get you a 2D benchmark to use.
>
> Thanks.
I've been doing more digging. I have verified that indeed the software
OpenGL renderer is creating way too much overhead to be useful for the
purposes of this testing, even when just
IIRC, there is some way to override this path, either at build time or
run time or both. I will look into it and see if I can build the DRI
library into the distribution. I didn't really care about it because I
always just use VirtualGL, but the software renderer really should work
for the purpos
This is an annoying problem I ran into when I was building and
distributing to a couple of machines. Using the build-xorg scripts,
it leaves the path hard set to the location of the xorg
libraries/modules instead of setting a path and installing them to
it with the main
On Mon, Feb 14, 2011 at 02:35:42PM -0600, DRC wrote:
> On 2/14/11 2:25 AM, Martin Koegler wrote:
> I disagree. I think there is something system-specific that is
> occurring and that you would be observing this problem very plainly if
> you were using my system. Others on different systems may ru
On 2/14/11 2:25 AM, Martin Koegler wrote:
> The performance problems are triggered by running running high
> performance applications like virtualgl, which is not part of any
> major distribution]. So any user of default installation of a major
> linux distribution has either an Xvnc without 3D sup
On Sun, Feb 13, 2011 at 04:32:46PM -0600, DRC wrote:
> So what exactly is a "default" Linux installation? I don't think anyone
> can reasonably conclude that what is true for their Linux installation
> is true for all Linux installations.
The performance problems are triggered by running running
So what exactly is a "default" Linux installation? I don't think anyone
can reasonably conclude that what is true for their Linux installation
is true for all Linux installations.
I still don't fully buy that OpenGL isn't causing some amount of CPU
overhead that may still be limiting the frame ra
I'm not following you. We already established that the auth extensions
can be used without building the code with GnuTLS support, so why is
this still an issue?
On 2/13/11 3:47 PM, Martin Koegler wrote:
> On Fri, Feb 11, 2011 at 10:53:48AM -0600, DRC wrote:
>> I would still be interested in know
On Fri, Feb 11, 2011 at 10:53:48AM -0600, DRC wrote:
> I would still be interested in knowing whether it would be feasible to
> use the auth extensions without GnuTLS.
The main question is:
How much more fragementation should be added to the VNC world?
qemu has eg. implemented VeNCrypt:
http://
On Sat, Feb 12, 2011 at 03:19:48PM -0600, DRC wrote:
> You have to run glxspheres with vglrun. Otherwise, it will use the
> software OpenGL renderer in TigerVNC, and that will be the bottleneck,
> not TigerVNC's image pipeline.
>
> Also realize that the methodology is not just running GLXspheres
You have to run glxspheres with vglrun. Otherwise, it will use the
software OpenGL renderer in TigerVNC, and that will be the bottleneck,
not TigerVNC's image pipeline.
Also realize that the methodology is not just running GLXspheres and
reading the output of that benchmark. You have to run
v
On Fri, Feb 11, 2011 at 10:53:48AM -0600, DRC wrote:
> On 2/11/11 2:17 AM, Martin Koegler wrote:
> >> What do you mean "still get 20 Mbit"? You mean Megapixels/second? Or
> >> do you mean Megabits/second of throughput? Because without VeNCrypt, my
> >
> > I mean the value showen in the viewer i
On Fri, 11 Feb 2011 10:53:48 -0600
DRC wrote:
>
> All of the Internet chatter I'm seeing about it points to a fundamental
> performance limitation in GnuTLS. I don't think that optimizing the way
> we use it is going to help much. I guess what I'm interested in hearing
> is why GnuTLS is funda
On 2/11/11 2:17 AM, Martin Koegler wrote:
>> What do you mean "still get 20 Mbit"? You mean Megapixels/second? Or
>> do you mean Megabits/second of throughput? Because without VeNCrypt, my
>
> I mean the value showen in the viewer info dialog. What tools are you using?
A thorough description o
On Thu, Feb 10, 2011 at 04:40:13AM -0600, DRC wrote:
> On 2/10/11 2:24 AM, Martin Koegler wrote:
> >> I really wonder how useful of a feature built-in session encryption is
> >> if it's this slow. Without encryption, I can get approximately 20
> >> Megapixels/second. Tunneling through SSh, about
On 2/10/11 2:24 AM, Martin Koegler wrote:
>> I really wonder how useful of a feature built-in session encryption is
>> if it's this slow. Without encryption, I can get approximately 20
>> Megapixels/second. Tunneling through SSh, about the same. With TLSVnc,
>> I get about 7.5 Megapixels/second.
On Wed, 09 Feb 2011 18:25:22 -0600
DRC wrote:
>
> An unrelated issue is that whenever I try to use the vncviewer GUI to
> set the Zlib level to something other than the default, the server
> crashes (the log says "ZlibOutStream: deflate failed".)
>
Somewhat known. I think we have a bug entry.
On Wed, Feb 09, 2011 at 10:09:25PM -0600, DRC wrote:
> > An unrelated issue is that whenever I try to use the vncviewer GUI to
> > set the Zlib level to something other than the default, the server
> > crashes (the log says "ZlibOutStream: deflate failed".)
In my self compiled windows viewer, I ca
Ugh. OK, so this was due to VeNCrypt. It is apparently enabled by
default if both server and viewer support it. I do not like that at
all. It's particularly problematic because, once enabled, there is no
way to subsequently disable it without disconnecting and reconnecting
(the VeNCrypt boxes i
When attempting to test the latest build (r4280) running from Linux
server to Linux client, I am seeing very poor performance, about half of
what I can achieve with 1.0.1 using identical settings. It appears to
be an issue with Xvnc, because replacing Xvnc with Xvnc 1.0.1 brings it
back to normal.
30 matches
Mail list logo