Re: [Spice-devel] Various problems with "-vga qxl"

2011-05-22 Thread David Jaša
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 IIRC, qxl device is provided by qemu-kvm, so you may need to backport newer qemu-kvm to your hosts. David Jaša Dne 22.5.2011 20:06, Peter Daum napsal(a): > Hi, > > Inspired by a talk at last week's "Linuxtag" that sounded pretty promising, I > st

Re: [Spice-devel] [RFCv5 00/47] server: multiple client support

2011-05-22 Thread Yaniv Kaul
On 05/22/2011 08:38 PM, Yonit Halperin wrote: Hi, Though implementing this feature using multiple instances of the core data structures, one instance per client, is straightforward and comprises a clean solution, I think that we should implement a more parsimonious solution with respect to mem

[Spice-devel] Various problems with "-vga qxl"

2011-05-22 Thread Peter Daum
Hi, Inspired by a talk at last week's "Linuxtag" that sounded pretty promising, I started playing with qemu/spice to evaluate whether this would be a good solution for a problem I'm working on: serving remote (mostly windows)-: desktops with decent performance over slow WAN links. I got the co

Re: [Spice-devel] [RFCv5 00/47] server: multiple client support

2011-05-22 Thread Yonit Halperin
Hi, Though implementing this feature using multiple instances of the core data structures, one instance per client, is straightforward and comprises a clean solution, I think that we should implement a more parsimonious solution with respect to memory and cpu, so we can support a large number

Re: [Spice-devel] [PATCH] add io ports for sleep (S3) support

2011-05-22 Thread Alon Levy
On Sun, May 22, 2011 at 04:23:55PM +0300, Alon Levy wrote: Please ignore for now - I'm not sure this is worthwhile if we have to support S4. > related RHBZ: 63 > > Sleep is supported by the following flow: > (G stands for Guest, D is the device, not the driver) > G: QXL_IO_ABOUT_TO_SLE

[Spice-devel] [PATCH] add io ports for sleep (S3) support

2011-05-22 Thread Alon Levy
related RHBZ: 63 Sleep is supported by the following flow: (G stands for Guest, D is the device, not the driver) G: QXL_IO_ABOUT_TO_SLEEP G: acpi reset D: qxl_reset_handler (ignored) G: QXL_IO_RETURNED_FROM_SLEEP Thereby the spice-server and the driver never reset during sleep. After

Re: [Spice-devel] [PATCH 0/2] client/windows: fix broken build

2011-05-22 Thread Uri Lublin
On 05/22/2011 03:04 PM, Arnon Gilboa wrote: Arnon Gilboa (2): common: add WIN64 ifdef for spice_bit_find_msb (fix broken windows x64 build) client/windows: remove precompiled header for common.h (fix broken windows debug build) client/windows/redc.vcproj |8 commo

[Spice-devel] [PATCH 2/2] client/windows: remove precompiled header for common.h (fix broken windows debug build)

2011-05-22 Thread Arnon Gilboa
-Release currently doesn't use precompiled headers at all -Debug is broken since common/*.c files don't include common.h -PCH can be enabled for all but specifically-chosen c-files --- client/windows/redc.vcproj |8 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/client

[Spice-devel] [PATCH 1/2] common: add WIN64 ifdef for spice_bit_find_msb (fix broken windows x64 build)

2011-05-22 Thread Arnon Gilboa
inline __asm is not supported in x64, so use the naive implementation until x64 asm implemented. --- common/bitops.h |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/common/bitops.h b/common/bitops.h index 4823662..449409b 100644 --- a/common/bitops.h +++ b/common/bitops

[Spice-devel] [PATCH 0/2] client/windows: fix broken build

2011-05-22 Thread Arnon Gilboa
Arnon Gilboa (2): common: add WIN64 ifdef for spice_bit_find_msb (fix broken windows x64 build) client/windows: remove precompiled header for common.h (fix broken windows debug build) client/windows/redc.vcproj |8 common/bitops.h|4 ++-- 2 files changed,