Commit Graph

192 Commits

Author SHA1 Message Date
Alexander Larsson
e5474684d5 client: Avoid crash if platform_win is NULL
This is a temporary fix as this will be fixed for real when
the cut and paste patches land.
2010-09-29 17:48:28 +02:00
Alexander Larsson
92d286c6b5 client: Re-enable USE_XRANDR_1_2
This was disabled by mistake before.
2010-09-29 15:08:05 +02:00
Alexander Larsson
d8faf813e7 client: Don't hardcode -lrt
-lrt is already included in SPICE_NONPKGCONFIG_LIBS if needed so no
need to add it to the command line manually.
2010-09-29 13:37:48 +02:00
Alexander Larsson
3f3283ee17 client: Fall back to gettimeofday if clock_gettime not found 2010-09-29 13:35:18 +02:00
Alexander Larsson
93dc13c91a client: Include config.h from common.h
config.h should be availible everywhere, so move its inclusion
to the top of common.h.
2010-09-29 13:26:21 +02:00
Alexander Larsson
44631aa023 client: Check for pthread yield function using autoconf 2010-09-29 13:16:45 +02:00
Alexander Larsson
8e14489621 client: Include stdint.h for uint32_t 2010-09-29 12:44:33 +02:00
Alexander Larsson
2626419da6 client: Don't use basename for argv[0] in --help output
This isn't what other apps do, and it had issues in the OSX port,
so just remove this.
2010-09-29 12:40:48 +02:00
Alexander Larsson
a5abceccb2 Fix warning from OSX compiler
The OSX compiler warns about uninitialized variable, so we change
a bit how size is initialized.
2010-09-29 12:36:39 +02:00
Hans de Goede
4efeef3302 spicec-x11: Fix going into a never ending loop upon xrandr event (#628573) (v2)
When handling an xrandr event the event_listener->on_monitors_change()
callback destroys and re-creates the monitor object(s) which results
in the DynamicScreen or MultyMonconstructor being called, which triggers
more xrandr events. This causes a never ending event handling loop making
spicec hang, and eventually making the X-server crash as a backlog
of events builds up and it oom's.

This patches this by explictly processing the xrandr event caused
by the constructor inside the constructor surrounded by the already
present guard code against recursive xrandr events.
2010-09-09 16:25:24 +02:00
Alon Levy
c47def6e48 client: add default agent capabilities 2010-08-31 14:30:00 +03:00
Alon Levy
18466b8b55 client: add announce_capabilities 2010-08-31 11:08:09 +03:00
Alexander Larsson
2df2b3df1e client: Handle async errors from xshm setup
XShmAttach can fail asynchronously, so we need to check the
errors in the x error handler during the XSync.
2010-08-25 12:19:46 +02:00
Alexander Larsson
7c7e3efff3 client: Don't leak xshm segments 2010-08-24 10:23:26 +02:00
Alexander Larsson
7c61dfee89 Don't try xshm any more if it fails for a permanent reason
This is copied from how Gtk+ detects Xshm failures.
2010-08-24 10:23:26 +02:00
Yonit Halperin
7b7feaecdf client: Fix for clipboard sending; It wasn't thread safe. 2010-08-23 08:19:26 +03:00
Yonit Halperin
a5194d37e9 client: fix - exit on bad display-effect argument 2010-08-23 08:19:09 +03:00
Alon Levy
01c06d6eaa client: Application::get_screen: fix double SpicePoint size, second hiding the first 2010-07-29 13:22:54 +03:00
Alon Levy
a8f2925162 add eol after #ifdef in jpeg_encoder.h for n900 scratchbox compiler 2010-07-29 11:10:41 +03:00
Alexander Larsson
ccbb922d5a Swap red and blue when decoding 0.4 mjpeg streams
There was an error in how this was encoded in 0.4, which we need
to handle. There is still some issues with the old streams as
the luminocity handling in 0.4 was not correct.
2010-07-20 11:45:37 +02:00
Alexander Larsson
5f664e0ffb Don't send padding over the network with video data 2010-07-20 11:33:57 +02:00
Alexander Larsson
12b08f2c3e codegen: Various cleanups
Remove all uses of @end in the marshaller, instead just using
the C struct array-at-end-of-struct. To make this work we also remove
all use of @end for switches (making them C unions).

We drop the zero member of the notify message so that we can avoid this
use of @end for a primitive in the marshaller (plus its useless to send
over the wire).

We change the offsets and stuff in the migration messages to real pointers.
2010-07-19 16:28:22 +02:00
Arnon Gilboa
ce03f5449d client: add clipboard support
* windows - untested
 * linux - small strings both ways, large implemented differently:
  * client to guest - support INCR
  * guest to client - we supply a single possibly very large property
 * requires server changes in next patch to work with spice-vmc
2010-07-19 10:30:19 +03:00
Alon Levy
4f8545ed62 fix typo DisplayModeListner -> DisplayModeListener 2010-07-19 10:28:06 +03:00
Yonit Halperin
9877e7ae84 client: command line arguments for setting windows guest monitors'
color depth and disabling some display options (helpful on WAN)
2010-07-19 09:40:11 +03:00
Alexander Larsson
9d8a76fcd0 Disable tunnel by default on windows 2010-07-08 22:25:32 +02:00
Alexander Larsson
633e962ac3 Make tunnel support optional in client too 2010-07-08 22:25:28 +02:00
Alexander Larsson
ceff16cad6 Make distcheck work 2010-07-08 20:20:35 +02:00
Alexander Larsson
1a6e77d6d4 Fix various misspellings
letancy -> latency
compund -> compound
SpicedSubMessage -> SpiceSubMessage
modifaiers -> modifiers
massage -> message
outgoiong -> outgoing
AlphaBlnd -> AlphaBlend
remoth -> remote
modifires -> modifiers
secore -> secure
2010-07-08 18:26:37 +02:00
Alexander Larsson
9cc1582561 Fix inclusion of common files, no need for common/ part 2010-07-08 15:44:15 +02:00
Alexander Larsson
601aecc8ad Remove unused method declarations 2010-07-08 14:19:45 +02:00
Alexander Larsson
0659fb7897 Move in spice/draw.h from spice-protocol to common/ 2010-07-08 13:56:01 +02:00
Alon Levy
0e16cadc70 Make CEGUI optional
This makes the CEGUI dependency optional and off by default. Restoring
previous behaviour of exiting on disconnect if disabled.
2010-07-01 13:48:58 +02:00
Alexander Larsson
d69fd9408c client: Use ASSERT, not assert 2010-07-01 10:50:12 +02:00
Alexander Larsson
5cd86fc45d Update client and protocol to support the new SpiceClipRects 2010-06-30 22:35:16 +02:00
Yonit Halperin
59152e31f0 client: add --include "common.h" for generated marshallers 2010-06-29 13:59:46 +02:00
Yonit Halperin
270a1e6df3 fixed unhandled destroy screen in Application:restore_screens_size 2010-06-29 13:59:46 +02:00
Yonit Halperin
3ca28ddf28 detaching the screen when the primary surface is destoryed and closing the window if the primary surface is not recreated (the monitor has been detached). 2010-06-29 13:59:46 +02:00
Alexander Larsson
7e26ff3c26 Replace log4cpp with custom log function
Also prints a simpler error to stderr for WARN or above so that
we print something on the commandline if something go wrong.
2010-06-28 21:54:53 +02:00
Alexander Larsson
b08b80d13f Fix build error due to member "SpiceMsgEmpty" same name as type 2010-06-28 12:45:07 +02:00
Alexander Larsson
9655dc30e9 Remove support for clip by path
This is not supported currently anyway and was not generated before.
2010-06-24 14:30:00 +02:00
Alexander Larsson
dc8625c7e1 Make pointers 32bit in new protocol format 2010-06-23 16:33:07 +02:00
Alexander Larsson
eb3fe11d94 Fix version mismatch error on connect
Protocol is 0 (auto), 1 (old), or 2 (new). This is (apart from 0) the
same as the major number for the stable protocol. However, the current major
is ~(-1) to signify it being unstable, so don't use the major number as source
for setting or comparing protocol.
2010-06-23 12:18:41 +02:00
Alexander Larsson
f35ac2049a Fix screen resolution changes on old servers
* Correctly check for existing canvas 0
* Don't try to unlock the screen size if there is no active screen
2010-06-23 11:37:01 +02:00
Alexander Larsson
72cf104c28 client: Support connecting to a major==1 server 2010-06-22 17:34:45 +02:00
Alexander Larsson
652c13e71b Implement display_mode message needed for backwards compat 2010-06-22 17:34:45 +02:00
Alexander Larsson
2788b2ec46 Generate marshaller/demarshallers for old protocol 2010-06-22 17:34:39 +02:00
Alexander Larsson
36fd22a9db Add more emacs mode headers 2010-06-22 12:03:21 +02:00
Alexander Larsson
5c05a24ba1 Convert client to use indirect calls for message marshalling
This is required to support multiple versions
2010-06-22 10:54:59 +02:00
Alexander Larsson
9123e24e7b Add destructor for demarshalled messages
This is required because we don't want to free messages that just
refer to the unparsed message (like SpiceMsgData).

Also, in the future we might need it for more complex demarshalling.
2010-06-22 10:53:24 +02:00