Jean-Philippe Andre
22e79af097
evas/cserve2: return in case of select timeout.
...
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
79b102f6fd
evas/cserve2: add some more error debug (client side).
...
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
65fcf390b4
evas/loaders: init/shutdown EET from loader module.
...
cserve2 slave requires Eet to be initialized before use :)
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
a09efda8d2
evas/loaders: add log domain for ICO loader.
...
This is to remove the dependency of the ico loader on Evas itself.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
8db3d4d2bf
evas/cserve2: remove dead code (old loaders API).
...
Remove: cserve-specific loaders and deprecated code.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
90d8647c02
evas/cserve2: fix crash with YUV tests in expedite.
...
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
48d97f7a90
evas/cserve2: use Evas loaders from slave
...
Use common loaders with evas, and ignore specific modules.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
c39b714868
evas/modules: Add preprocessor ifs
...
We'll want to distinguish full builds for evas and partial
builds for cserve2 slaves.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
26e01c0ff6
evas: build evas_module within cserve slave
...
Objective: use common loaders from cserve2
Ideally evas_module should be a static library but it would
then require static/dynamic linking to too many modules unused
by cserve2 (eg. engines & savers)
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
db6cba1429
evas/cserve2: avoid useless SETOPT messages.
...
Most of the time, the SETOPT message is sent only
to carry 0 data. Let's add a flag to the OPEN message to
avoid useless socket calls.
Server-side: implement with message faking.
Also, fix image ID logic on client side:
- alloc data entry asap,
- assign image_id during call to OPEN
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
f9a7820e02
evas/cserve2: Simplify debug logs (server side).
...
Remove newlines.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
096a8ece9f
evas/cserve2: Set default LRU size to 4Mb or env-based
...
The LRU should not be limited to 0 byte otherwise all scaled
images will be dropped and reloaded constantly, killing the
performance.
The size is inspired from the scalecache's size of 4Mb.
Variable: EVAS_CSERVE2_SIZE (number in Mb)
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
29bf164c75
evas/cserve2: pass loader data during open.
...
Note: this path seems never to be used.
Loader options are set by the file loader or after open.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
6a2a0187ff
evas/cserve2: write() can return other errors.
...
In case of error, return from the function and don't decrement
the send variable.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
f7c6545170
evas/cserve2: minor fixes (error paths).
...
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:10 +09:00
Jean-Philippe Andre
d712a13a7a
evas/cserve2: use null-terminated strings everywhere.
...
Some hashtable lookups failed because the keys were added
either with eina_stringshare_add() or e_s_add_lenth() but
looked for with e_s_add_length() only.
Signed-off-by: Cedric Bail <cedric.bail@samsung.com>
2013-07-02 16:58:09 +09:00
ChunEon Park
acf4728ca7
evas - updated missed CNN
2013-07-02 15:47:13 +09:00
ChunEon Park
627beaae71
evas/proxy rendering - optimize proxy rendering.
...
re-render proxies if sources are changed. this patch increases the proxy rendering up to 2x.
2013-07-02 15:38:42 +09:00
ChunEon Park
5f4ae9abeb
evas/render - refactoring about checking smart member changed.
...
move those code to the evas_object_smart.
2013-07-02 15:28:56 +09:00
Cedric Bail
e0b51e3a37
efl: cleanup NEWS.
2013-07-02 14:50:02 +09:00
Cedric Bail
2fdc05863d
eet: force thread to always run during eet_cache_concurrency test.
2013-07-02 13:45:30 +09:00
Daniel Willmann
b6aaef651b
edje_text: Clarify that the block and else are independent
...
Sometimes whitespace helps
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-07-01 16:04:27 +01:00
Tom Hacohen
e37b83110c
Fixed ChaneLog formatting.
2013-07-01 14:19:05 +01:00
Tom Hacohen
827885d794
Evas textblock: Fixed format parsing to not be confused by single-quotes.
...
Before this commit, having a single quote anywhere in the format would mess
up all of the format parsing.
Thanks to MinSu Seo for reporting.
2013-07-01 14:18:52 +01:00
Cedric Bail
cc012b754c
eo: little micro optimization.
2013-07-01 18:35:55 +09:00
Cedric Bail
c435968f69
eo: a little more inlining, give me a 10% speed improvement.
2013-07-01 18:18:40 +09:00
Daniel Willmann
6aa69a0c7d
Revert "configure: Link eina to librt. This is needed for shm_* calls"
...
This reverts commit 6445fea318
.
Jenkins seems lost without it, but this is wrong on some platforms. Need
to solve it another way.
2013-07-01 09:18:25 +01:00
Jiyoun Park
958a430f5e
* Evas: Fix jpeg loader cannot deal with exif information correctly
2013-06-29 15:00:43 +09:00
Daniel Willmann
7927c0f9fa
edje_test: Test geometry of a more complex layout
...
It seems someone thought it was fun to have rel2.offset behave
differently than rel1.offset...
A rel2.offset of 0 actually means the size is increased by one pixel.
Acocunt for that in the tests.
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-06-28 14:25:30 +01:00
Daniel Willmann
db29fbb286
edje_test: Make the variable returned in test_layout_get static
...
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-06-28 14:25:30 +01:00
Chris Michael
f9348a5015
Fix klockwork possible memleak report:
...
If we get an error trying to "make current" an allocated surface, then
we should free that allocated surface.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 14:23:57 +01:00
Chris Michael
2a67e78bde
Fix klockwork memleak: We should free 'texts' that we malloc'd earlier
...
if we are going to error out.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 14:14:44 +01:00
Chris Michael
694d1600e8
We should free glyph_out also (as that gets malloc'd too) on error.
...
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 14:13:09 +01:00
Chris Michael
dd136b82e2
Fix klockwork memleak: If we allocate space for a Font Glyph (through
...
evas_common_font_int_cache_glyph_get), then we should free that
allocation when we have an error.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 14:02:00 +01:00
Chris Michael
7e222b949c
Fix klockwork memleak: If we allocate a colormap above, then we need
...
to free it on error.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 13:57:56 +01:00
Chris Michael
204212131b
Fix klockwork error: Null pointer may be dereferenced
...
Check for valid out_file before trying to use it.
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 13:48:21 +01:00
Chris Michael
d92edb51a4
Fix memleak reported by klockwork:
...
If a jpeg image is rotated and we have allocated space for ptr_rotate,
then we should free it Always ... Not Only if the degrees are
180...because we allocate it always
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-28 13:39:45 +01:00
Carsten Haitzler
6828cf2e29
evas: efl 1.8 dev bug only... update add bug in image objects throws away some updates.
2013-06-28 17:26:14 +09:00
Carsten Haitzler
2cdc41fd36
update po's
2013-06-28 17:26:14 +09:00
Cedric Bail
dcf65a8eec
eo: add a new macro to make the code easier to read and use it.
2013-06-28 15:28:30 +09:00
Cedric Bail
7849c5de32
eo: 30% speed improvement in message propagation.
2013-06-28 10:18:39 +09:00
Cedric Bail
383592b444
eo: eo is absolutely necessary by now, can't be disabled any more.
2013-06-28 10:18:39 +09:00
Cedric Bail
0ed0d79611
evas: prevent people from including header directly.
2013-06-28 10:18:39 +09:00
Eduardo Lima (Etrunko)
2e7479f8db
eldbus: Missing dbus_connection_unref() in error path
...
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
2013-06-27 17:31:22 -03:00
Eduardo Lima (Etrunko)
f6295c5828
eldbus: Call dbus_bus_register() for TYPE_ADDRESS connections
...
From D-Bus documentation:
http://dbus.freedesktop.org/doc/api/html/group__DBusBus.html
dbus_bus_register():
If you open a bus connection with dbus_connection_open() or
dbus_connection_open_private() you will have to dbus_bus_register()
yourself, or make the appropriate registration method calls yourself.
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
2013-06-27 17:24:46 -03:00
Eduardo Lima (Etrunko)
6f255a7ccc
eldbus: Add proxy call to org.freedesktop.DBus.Hello
...
Signed-off-by: Eduardo Lima (Etrunko) <eduardo.lima@intel.com>
2013-06-27 16:43:24 -03:00
Daniel Willmann
5763195418
edje_test: Test geometry and color of edje part
...
Just a simple test that uses the existing test_layout
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-06-27 17:51:30 +01:00
Daniel Willmann
a23ff7b5ab
edje_test: Make test_layout_get() configurable
...
This will help when we want to test different edj files
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-06-27 17:50:23 +01:00
Daniel Willmann
3e5ff22ab9
eina_test_fp: Improve speed of eina_fp test
...
Test less values in between, but still have a decent combination of
large/small values to test.
Fixes -WTasn
Signed-off-by: Daniel Willmann <d.willmann@samsung.com>
2013-06-27 14:34:02 +01:00
Chris Michael
d263e31c7f
When getting crtc info, fill in missing "rotations" field so that the
...
value is initialized.
NB: Spank Spank whoever did this ... oh wait, that was me :( Ouch !!
Signed-off-by: Chris Michael <cp.michael@samsung.com>
2013-06-27 10:46:34 +01:00