[Spice-space-devel] [PATCH] spice: qxl win driver: change debug level check

2010-01-07 Thread Izik Eidus
Signed-off-by: Izik Eidus --- win/display/driver.c |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/win/display/driver.c b/win/display/driver.c index 2f34240..36aa9db 100644 --- a/win/display/driver.c +++ b/win/display/driver.c @@ -116,7 +116,7 @@ void DebugPrint(PDev

[Spice-space-devel] [PATCH] spice: win qxl driver: split drawing of areas with big clips

2010-01-07 Thread Izik Eidus
split into the clipped areas of the drawing, in case the drawing is highly clipped Signed-off-by: Izik Eidus --- win/display/res.c | 72 +- win/display/res.h |1 + win/display/rop.c | 131 +++- 3 files changed

Re: [Spice-space-devel] integration Question

2010-01-05 Thread Izik Eidus
On 01/05/2010 07:28 PM, Scot Stacey wrote: Hello, I am the Senior Technical Consultant for the software company that I work for. MasterControl, is a leading software for document management in Medical companies and was even recently given the contract for the FDA. All of this aside, we wer

[Spice-space-devel] [PATCH] spice:stable server: add calls to red_pipe_clear_device_resources()

2010-01-05 Thread Izik Eidus
From c90811bad5fbefb7faecba91e737b2a68189d4fd Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Tue, 5 Jan 2010 19:40:08 +0200 Subject: [PATCH] spice:stable server: add calls to red_pipe_clear_device_resources() needed in case we are detaching Signed-off-by: Izik Eidus --- server/red_worker.c

Re: [Spice-space-devel] [PATCH] spice: qxl driver: add modified murmur hash 2a

2010-01-05 Thread Izik Eidus
On Tue, 05 Jan 2010 15:10:20 +0200 Yaniv Kamay wrote: > > +_inline uint32_t MurmurHash2AJump3(const uint32_t * key, uint32_t len, > > uint32_t seed ) > > > > len need to be number of 32bit words > From bcec1feabbd7b478c11245074b1437fe0e9b249a Mon Sep 17 00:

Re: [Spice-space-devel] QXL driver does not start

2009-12-30 Thread Izik Eidus
On Wed, 30 Dec 2009 11:30:35 +0100 (CET) Frédéric Grelot wrote: > > - "Izik Eidus" a écrit : > > > You need to setup windows 2003 ddk enviorment (you can install it on > > windows xp guest) > > the iso for the installtion can be found at: > > h

[Spice-space-devel] [PATCH] spice: qxl driver: add modified murmur hash 2a

2009-12-30 Thread Izik Eidus
>From d9ec7eef5b2c5d63960afb5744224107b8391e36 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Wed, 30 Dec 2009 04:15:31 +0200 Subject: [PATCH] spice: qxl driver: add modified murmur hash 2a Signed-off-by: Izik Eidus --- win/display/lookup3.c | 18 - win/display/re

Re: [Spice-space-devel] QXL driver does not start

2009-12-29 Thread Izik Eidus
On Tue, 29 Dec 2009 18:32:02 -0500 (EST) Yaniv Kamay wrote: > > - "Izik Eidus" wrote: > > > On Tue, 29 Dec 2009 23:43:16 +0100 (CET) > > Frédéric Grelot wrote: > > > > > Thanks for your (fast!) answer. > > > > > > Ok, you

Re: [Spice-space-devel] QXL driver does not start

2009-12-29 Thread Izik Eidus
On Tue, 29 Dec 2009 23:43:16 +0100 (CET) Frédéric Grelot wrote: > Thanks for your (fast!) answer. > > Ok, you confirm that the version-mismatch are problematic... > Since I would be interested in testing the development branch, I would be > very happy to have the precompiled version of the driv

Re: [Spice-space-devel] QXL driver does not start

2009-12-29 Thread Izik Eidus
On Tue, 29 Dec 2009 23:23:15 +0100 (CET) Frédéric Grelot wrote: > Hi all, > > I'm currently testing spice on a fedora 12 server. I installed Windows XP > using the docs available on spice-space.org, with the latest versions of > client and server built from git (2 days ago). I actually have a

[Spice-space-devel] [PATCH] qxl driver: tell windows not to use power managment

2009-12-29 Thread Izik Eidus
>From 8690c7f7db6e967a4feb681d8a71cb5b2be0d1b8 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Wed, 30 Dec 2009 00:06:59 +0200 Subject: [PATCH] qxl driver: tell windows not to use power managment. With latest changes we have problems to sleep, for now we disable this. It appear that this pa

Re: [Spice-space-devel] Compilation on Ubuntu 9.10 64 bits

2009-12-29 Thread Izik Eidus
On Tue, 29 Dec 2009 14:26:54 +0200 Yaniv Kamay wrote: > guer...@free.fr wrote: > > Hello, > > > > I try to compile the spice protocol on an ubuntu 9.10 64 bit server. > > > > I have success compile differents modules : qcairo-1.8.7.1-git74d6b5 > > qpixman-0.13.3-git20090127 slirp spice > > > >

[Spice-space-devel] [PATCH 2/2] qxl: move into new murmur modifed hash

2009-12-28 Thread Izik Eidus
>From 21db4db9b993d3e4a20c309faf6d72c657b1aba0 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Tue, 29 Dec 2009 05:18:12 +0200 Subject: [PATCH] qxl: move into new murmur modifed hash Signed-off-by: Izik Eidus --- win/display/lookup3.c |3 +- win/display/res.c |

[Spice-space-devel] [PATCH 1/2] spice: add new modifed murmur2 hash function

2009-12-28 Thread Izik Eidus
>From 2379eec91c22fd02eca59496ccceb5f86d50e7fe Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Tue, 29 Dec 2009 05:18:51 +0200 Subject: [PATCH] spice: add new modifed murmur2 hash function. optimized for spice usage Signed-off-by: Izik Eidus --- common/murmur_hash2a.h |

[Spice-space-devel] [PATCH 0/2] new modifed murmur2hash - big performence gain

2009-12-28 Thread Izik Eidus
Hi, Following patch`s add a new murmur2a hash function that we have modifed to be optimized for our 32bits with alpha cases. I have saw big performence gains, Windows 7 feels now much faster. Thanks. -- This SF.Net emai

[Spice-space-devel] [PATCH] qemu: add surface0 support to the qxl device v2

2009-12-27 Thread Izik Eidus
>From b09152c01dc7c43f77015b183add3e1db36ccb88 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Sat, 26 Dec 2009 07:39:45 +0200 Subject: [PATCH] qemu: add surface0 support to the qxl device need in order to sync the qxl device with the server latest changes Signed-off-by: Izik Eidus --- q

Re: [Spice-space-devel] [PATCH 1/3] Add surface0 support (spice server side patch)

2009-12-27 Thread Izik Eidus
On Sun, 27 Dec 2009 09:48:33 +1000 Dave Airlie wrote: > Though as long as you don't mind throwing away the static linear > memory requirement > and redoing this later to be useful I don't mind. Above ideas are all interesting, but like you said here, lets first add the 3d support and then we wil

Re: [Spice-space-devel] 512MB mtrr at 0xc0000000

2009-12-23 Thread Izik Eidus
On Thu, 24 Dec 2009 06:51:55 +1000 Dave Airlie wrote: > On Thu, Dec 24, 2009 at 1:33 AM, Izik Eidus wrote: > > On Wed, 23 Dec 2009 10:04:03 +1000 > > Dave Airlie wrote: > > > >> Just wondering why the spice BIOS sets up an MTRR at this point, > >> > &

[Spice-space-devel] [PATCH] spice: fix server crush in case connecting without qxl device

2009-12-23 Thread Izik Eidus
>From 4814aca706372759bfb2ac8ccff78252744a0fe8 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Wed, 23 Dec 2009 23:04:25 +0200 Subject: [PATCH] spice: fix server crush in case connecting without qxl device Signed-off-by: Izik Eidus --- server/red_dispatcher.c |3 +++ 1 files changed

[Spice-space-devel] [PATCH] vdesktop: add shadow rom

2009-12-23 Thread Izik Eidus
>From 5ece87801c97f1fbad118e53a5845adb961cba49 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Wed, 23 Dec 2009 21:55:59 +0200 Subject: [PATCH] vdesktop: add shadow rom. This is needed in order not to let the guesst changes of the rom effect the qxl device. Signed-off-by: Izik Eidus --- q

[Spice-space-devel] [PATCH] spice client: fix wrong gdi-canvas handling of blend_alpha

2009-12-23 Thread Izik Eidus
>From ec07aa2d26ec25fb5a012689f6fdc292427a66a3 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Wed, 23 Dec 2009 17:43:12 +0200 Subject: [PATCH] spice client: fix wrong gdi-canvas handling of blend_alpha Signed-off-by: Izik Eidus --- common/gdi_canvas.c | 20 ++-- 1 fi

Re: [Spice-space-devel] 512MB mtrr at 0xc0000000

2009-12-23 Thread Izik Eidus
On Wed, 23 Dec 2009 10:04:03 +1000 Dave Airlie wrote: > Just wondering why the spice BIOS sets up an MTRR at this point, > > When X starts and tries to setup a write combining mtrr on the video > device it throws an > error because it conflicts with this. The vesa framebuffer start at 0xc00

Re: [Spice-space-devel] alignment of QXL interface structures

2009-12-22 Thread Izik Eidus
On Wed, 23 Dec 2009 07:07:23 +1000 Dave Airlie wrote: > > http://lwn.net/Articles/259732/ > > is probably worth reading, > > http://msdn.microsoft.com/en-us/library/aa290049%28VS.71%29.aspx > > for the Windows folks, > > I've also read that even on x86 unaligned accesses can slow things > do

Re: [Spice-space-devel] alignment of QXL interface structures

2009-12-22 Thread Izik Eidus
On Wed, 23 Dec 2009 06:29:30 +1000 Dave Airlie wrote: > On Tue, Dec 22, 2009 at 10:00 PM, Yaniv Kamay > wrote: > > Hi, > > > > All QXL structures are defined as packed see qxl_dev.h. > > > > This will prove interesting in the future I expect, unaligned accesses > are really > really slow on cer

[Spice-space-devel] [PATCH 2/3] surface0 support qxl device side

2009-12-22 Thread Izik Eidus
Hi, Please review. >From 50ee5a9927859e00ddf7b09e70ec88c9d769a432 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Thu, 3 Dec 2009 15:07:37 +0200 Subject: [PATCH] qemu: add surface0 support to the qxl device need in order to sync the qxl device with the server latest changes Signed-off

[Spice-space-devel] [PATCH] fix spice ssl compilation error

2009-12-21 Thread Izik Eidus
>From 4ed38b3a6376ebc89f2fbcf7390cf24011673e26 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Tue, 22 Dec 2009 06:52:14 +0200 Subject: [PATCH] spice: fix ssl compiling errors (openssl api was changed, so lets have ifdef to compile in all cases) Signed-off-by: Izik Eidus --- cli

Re: [Spice-space-devel] 3D/OpenGL questions

2009-12-21 Thread Izik Eidus
On Mon, 21 Dec 2009 12:57:32 + Daire Byrne wrote: > It builds fine and connecting with either gl_pbuff or gl_fbo both seem > to work fine - thanks. I can't say I see much of a performance > increase though - dragging windows around still seems slower than with > the Windows client. I think ma

Re: [Spice-space-devel] 3D/OpenGL questions

2009-12-21 Thread Izik Eidus
On Mon, 21 Dec 2009 12:57:32 + Daire Byrne wrote: > On Fri, Dec 18, 2009 at 11:29 PM, Yaniv Kamay > wrote: > > > > - "Daire Byrne" wrote: > > You can try cherry-pick from master to 0.4. > > > > in branch 0.4 do > > git cherry-pick 54cf04beed5e7617459cb00fa07be52db6c6e0d7 > > git cherry-

Re: [Spice-space-devel] 3D/OpenGL questions

2009-12-18 Thread Izik Eidus
On Fri, 18 Dec 2009 21:04:21 +0200 Izik Eidus wrote: > On Fri, 18 Dec 2009 17:18:28 + > Daire Byrne wrote: > > > >> > > > We'll do regular releases but anyone can take all master branches > > > and build it from source. The rule > > > w

Re: [Spice-space-devel] 3D/OpenGL questions

2009-12-18 Thread Izik Eidus
On Fri, 18 Dec 2009 17:18:28 + Daire Byrne wrote: > >> > > We'll do regular releases but anyone can take all master branches > > and build it from source. The rule > > we go by is to always keep the repositories in working condition. > > I tried compiling just the client and libspice from th

Re: [Spice-space-devel] [PATCH] make sure spice client won`t use XShem when using X from remote

2009-12-17 Thread Izik Eidus
true; > > } > > > +} > Changed. >From a1665e4a6b7e705821166b9c193fb90e139e6a02 Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Thu, 17 Dec 2009 20:45:09 +0200 Subject: [PATCH] spice: client: add checks to see if it is safe to use XShem. Beacuse that XShem internal chec

[Spice-space-devel] [PATCH] make sure spice client won`t use XShem when using X from remote

2009-12-17 Thread Izik Eidus
Mon Sep 17 00:00:00 2001 From: Izik Eidus Date: Thu, 17 Dec 2009 20:45:09 +0200 Subject: [PATCH] spice: client: add checks to see if it is safe to use XShem. Beacuse that XShem internal checks wont fail when using the spice client from remote, we are adding check on the socket family to see if it

Re: [Spice-space-devel] spice not checking malloc returns?

2009-12-17 Thread Izik Eidus
On Thu, 17 Dec 2009 10:26:41 -0200 Glauber Costa wrote: > Hi, > > I've just downloaded SPICE source so I could take a basic look > and understand more or less what's going on. > > I haven't checked all the source yet, but at least in the files > I've went through, there is simply no error check