Re: [Spice-devel] Does spice in xen-4.2.0 support QXL?

2012-11-22 Thread Qian Hu
Hi, Fabio Thank you for your advice! I have successfully added the patch to xen-4.2.0 and set the parameters as you said. It works and the client seems to be better than before. Still, there is a little problem that the screen is upside down when the win7 started, exactly the moment the four

Re: [Spice-devel] [PATCH vdagent-linux] vdagentd: send monitors config to currently active agent only

2012-11-22 Thread Hans de Goede
Hi, On 11/22/2012 12:27 AM, Marc-André Lureau wrote: Sending monitors config to all agents will results in unnecessary work from some of them, or racy behaviour. Currently, several session agent sharing the same X session may want to create or modify the same mode resulting in X errors:

Re: [Spice-devel] [PATCH spice-server 1/7] char_device.c: add ref count for write-to-device buffers

2012-11-22 Thread Hans de Goede
Hi, Looks good, one minor comment though, see my remarks inline. On 11/21/2012 08:42 PM, Yonit Halperin wrote: The ref count is used in order to keep buffers that were in the write queue and now are part of migration data, in case the char_device state is destroyed before we complete sending

Re: [Spice-devel] [PATCH spice-server 2/7] char_device.c: when the state is destroyed, also free the buffer that is being written to the device

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: --- server/char_device.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/server/char_device.c b/server/char_device.c index 1c48719..5231bf0 100644 --- a/server/char_device.c +++ b/server/char_device.c

Re: [Spice-devel] [PATCH spice-server 4/7] red_worker.c: fix marshalling of migration data

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: fix calling spice_marhsaller_add_ref with memory on stack --- server/red_worker.c | 11 +-- 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/server/red_worker.c b/server/red_worker.c

Re: [Spice-devel] [PATCH spice-server 5/7] smartcard.c: avoid marshalling migration data with reference to a memory that might be released before send has completed

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: The current solution just copy the buffer. Currently data that is read from the guest is always copied before sending it to the client. When we will have ref count for these buffers, we can also use it for

Re: [Spice-devel] [PATCH vdagent-linux] vdagentd: send monitors config to currently active agent only

2012-11-22 Thread flooding Controlled
Hi Lureau: It is currently available under http://spice-space.org/download/gtk/windows/ 3a009d5b008067d4e5b5f0f898ee9818 virt-viewer-0.5.3_x86.exe b988a7c88fa577e2e95f36422f4f9b80 virt-viewer-0.5.3_x64.exe I am now quite interested in howto setup the enviroment to make the exes. Do you have

Re: [Spice-devel] [PATCH spice-server 6/7] red_worker.c: fix not sending all pending messages when the device is stopped

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: red_wait_outgoing_item only waits till the currently outgoing msg is completely sent. red_wait_outgoing_items does the same for multi-clients. handle_dev_stop erroneously called red_wait_outgoing_items,

Re: [Spice-devel] Very poor performance on KVM with Ubuntu 12.10 guest

2012-11-22 Thread David Jaša
Jodi Curtis píše v St 21. 11. 2012 v 17:22 +: Hi The file doesn't exist on the host You should _create_ the file in the _guest_. and I had already written off Ubuntu 12.x, the host is command line only, I moved to Xubuntu, which offers better performance, That's actually no

Re: [Spice-devel] [PATCH spice-server 7/7] red_worker.c: fix memory corruption when data from client is bigger than 1024 bytes

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: Previously, there was no check for the size of the message received from the client, and all messages were read into a buffer of size 1024. However, migration data can be bigger than 1024. In such cases, memory

Re: [Spice-devel] [PATCH] char_device.c: fix call to spice_marshaller_add_ref with memory on stack

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:53 PM, Yonit Halperin wrote: rhbz#862352 --- server/char_device.c | 14 -- 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/server/char_device.c b/server/char_device.c index ac2632d..141ec88 100644 ---

Re: [Spice-devel] [PATCH spice-server 3/7] reds.c: fix calls to spice_marshaller_add_ref with ptr to memory that might be released before sending

2012-11-22 Thread Hans de Goede
Hi, Looks good, ACK. Regards, Hans On 11/21/2012 08:42 PM, Yonit Halperin wrote: --- server/reds.c | 24 ++-- 1 file changed, 14 insertions(+), 10 deletions(-) diff --git a/server/reds.c b/server/reds.c index 98c8706..b99d01f 100644 --- a/server/reds.c +++

Re: [Spice-devel] SSL connection problem

2012-11-22 Thread Christophe Fergeau
Hey, On Wed, Nov 21, 2012 at 07:42:46PM +, Jodi Curtis wrote: I am still having problems connecting via SSL after resolving the apparmor.d problem with reading the key directory contents On attempts to connect via virsh I am given this warning spice channels 1 should be encrypted, I'm

Re: [Spice-devel] [Users] Spice Issue

2012-11-22 Thread Christophe Fergeau
On Thu, Nov 22, 2012 at 05:32:52PM +0800, flooding Controlled wrote: No one tried this ? and how the exe to be made ? I told you how it was made in the mail below... This was cross-compiled on a linux machine using the mingw-w64 toolchain. Christophe 2012/11/21 Christophe Fergeau

Re: [Spice-devel] Very poor performance on KVM with Ubuntu 12.10 guest

2012-11-22 Thread Christophe Fergeau
On Thu, Nov 22, 2012 at 11:06:12AM +0100, David Jaša wrote: Jodi Curtis píše v St 21. 11. 2012 v 17:22 +: I also have a problem with en-gb map not mapping some keys despite being configured in spice and on the guest os (#~ not working at all, @' mapping to ¬` and ¬` not working at all)

Re: [Spice-devel] Does spice in xen-4.2.0 support QXL?

2012-11-22 Thread David Jaša
Doesn't that happen with some other videos as well? It sounds familiar to these two past bugs: https://bugzilla.redhat.com/show_bug.cgi?id=667689 https://bugs.freedesktop.org/show_bug.cgi?id=34262 David Qian Hu píše v Čt 22. 11. 2012 v 15:46 +0800: Hi, Fabio Thank you for your advice! I

Re: [Spice-devel] [PATCH vdagent-linux] vdagentd: send monitors config to currently active agent only

2012-11-22 Thread Marc-André Lureau
Hi - Mensaje original - +/* Send monitor config to currently active agent */ +udscs_write(active_session_conn, VDAGENTD_MONITORS_CONFIG, 0, 0, +(uint8_t *)mon_config, size); /* Acknowledge reception of monitors config to spice server /

Re: [Spice-devel] [Users] Spice Issue

2012-11-22 Thread flooding Controlled
Sorry, I see it. :) thanks. I will have a try. 2012/11/22 Christophe Fergeau cferg...@redhat.com On Thu, Nov 22, 2012 at 05:32:52PM +0800, flooding Controlled wrote: No one tried this ? and how the exe to be made ? I told you how it was made in the mail below... This was cross-compiled on

[Spice-devel] [PATCH] vdservice: support Windows 8 Server 2012

2012-11-22 Thread Arnon Gilboa
vdagent was ok, but vdservice required adding these to the version check. Both are handled as the rest of the Win7 class. Tested on Win8x64. --- vdservice/vdservice.cpp |4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/vdservice/vdservice.cpp b/vdservice/vdservice.cpp

Re: [Spice-devel] [PATCH vdagent-linux] vdagentd: send monitors config to currently active agent only

2012-11-22 Thread Hans de Goede
Hi, On 11/22/2012 11:58 AM, Marc-André Lureau wrote: Hi - Mensaje original - +/* Send monitor config to currently active agent */ +udscs_write(active_session_conn, VDAGENTD_MONITORS_CONFIG, 0, 0, +(uint8_t *)mon_config, size); /* Acknowledge reception

Re: [Spice-devel] Guest agent for Windows 8

2012-11-22 Thread Arnon Gilboa
Erik Lotspeich wrote: On 11/21/2012 07:13 AM, Christophe Fergeau wrote: On Wed, Nov 21, 2012 at 02:59:44PM +0200, Arnon Gilboa wrote: Christophe Fergeau wrote: On Tue, Nov 20, 2012 at 11:34:56AM -0600, Erik Lotspeich wrote: I was wondering if there will be a Windows 8

Re: [Spice-devel] Guest agent for Windows 8

2012-11-22 Thread Arnon Gilboa
Arnon Gilboa wrote: Erik Lotspeich wrote: On 11/21/2012 07:13 AM, Christophe Fergeau wrote: On Wed, Nov 21, 2012 at 02:59:44PM +0200, Arnon Gilboa wrote: Christophe Fergeau wrote: On Tue, Nov 20, 2012 at 11:34:56AM -0600, Erik Lotspeich wrote: I was wondering if there

Re: [Spice-devel] [PATCH vdagent-linux] vdagentd: send monitors config to currently active agent only

2012-11-22 Thread David Jaša
Hans de Goede píše v Čt 22. 11. 2012 v 12:16 +0100: Hi, On 11/22/2012 11:58 AM, Marc-André Lureau wrote: Hi - Mensaje original - +/* Send monitor config to currently active agent */ +udscs_write(active_session_conn, VDAGENTD_MONITORS_CONFIG, 0, 0, +

[Spice-devel] virt-viewer for windows build instructions

2012-11-22 Thread Marc-André Lureau
Hi - Mensaje original - Hi Lureau: It is currently available under http://spice-space.org/download/gtk/windows/ 3a009d5b008067d4e5b5f0f898ee9818 virt-viewer-0.5.3_x86.exe b988a7c88fa577e2e95f36422f4f9b80 virt-viewer-0.5.3_x64.exe I am now quite interested in howto setup the

Re: [Spice-devel] Does spice in xen-4.2.0 support QXL?

2012-11-22 Thread Fabio Fantoni
Il 22/11/2012 08:46, Qian Hu ha scritto: Hi, Fabio Thank you for your advice! I have successfully added the patch to xen-4.2.0 and set the parameters as you said. It works and the client seems to be better than before. Still, there is a little problem that the screen is upside down when

[Spice-devel] [PATCH vdagent-linux] randr: do not set crtc mode to NULL when changing resolution

2012-11-22 Thread Marc-André Lureau
The agent creates a unique mode vdagent-N per output, that is deleted when the custom resolution is changed. In order to be deleted, it can't be the current active crtc config. Now, the vdagent just sets the mode to NULL, but that causes a change of resolution probably due to

Re: [Spice-devel] [PATCH vdagent-linux] randr: do not set crtc mode to NULL when changing resolution

2012-11-22 Thread Marc-André Lureau
Note with this patch, the previous patch is unnecessary: the configuration can be sent to all agents (gdm user) the creation/deletion races will be ignored. On Thu, Nov 22, 2012 at 4:08 PM, Marc-André Lureau marcandre.lur...@gmail.com wrote: The agent creates a unique mode vdagent-N per

Re: [Spice-devel] [PATCH vdagent-linux] randr: do not set crtc mode to NULL when changing resolution

2012-11-22 Thread Hans de Goede
Hi, ACK, but please fix a small whitespace problem before pushing (see comment below) On 11/22/2012 04:08 PM, Marc-André Lureau wrote: The agent creates a unique mode vdagent-N per output, that is deleted when the custom resolution is changed. In order to be deleted, it can't be the current

Re: [Spice-devel] [PATCH vdagent-linux] randr: do not set crtc mode to NULL when changing resolution

2012-11-22 Thread Hans de Goede
Hi, On 11/22/2012 05:07 PM, Marc-André Lureau wrote: Note with this patch, the previous patch is unnecessary: the configuration can be sent to all agents (gdm user) the creation/deletion races will be ignored. Cool, good way to test the race robustness of the code. Still I think it would

Re: [Spice-devel] [PATCH v2] vdservice: support Windows 8 Server 2012

2012-11-22 Thread Christophe Fergeau
On Thu, Nov 22, 2012 at 01:38:07PM +0200, Arnon Gilboa wrote: vdagent was ok, but vdservice required adding these to the version check. Both are handled as the rest of the Win7 class. Tested on Win8x64. --- vdservice/vdservice.cpp |4 ++-- 1 files changed, 2 insertions(+), 2