platform/upstream/efl.git
10 years agoevas/cserve2: Reduce debug logs (s/INF/DBG)
Jean-Philippe Andre [Tue, 3 Sep 2013 02:27:27 +0000 (11:27 +0900)]
evas/cserve2: Reduce debug logs (s/INF/DBG)

10 years agoevas/cserve2: Fallback to normal cache for mmap
Jean-Philippe Andre [Mon, 2 Sep 2013 09:53:18 +0000 (18:53 +0900)]
evas/cserve2: Fallback to normal cache for mmap

cserve2 can't handle virtual files (mmap-only), by design.
Proper support can be added later on, but for now we might want
to just fallback to the normal cache functions.

Fixes photocam test

10 years agoevas/cserve2: Fix font hinting
Jean-Philippe Andre [Mon, 2 Sep 2013 07:17:05 +0000 (16:17 +0900)]
evas/cserve2: Fix font hinting

cserve2 was not handling font hints properly in the server side
Add some consistency checks and separate glyphs per hinting value

10 years agoevas/cserve2: Reset pointers to NULL after close
Jean-Philippe Andre [Mon, 2 Sep 2013 01:10:02 +0000 (10:10 +0900)]
evas/cserve2: Reset pointers to NULL after close

10 years agoevas/cserve2: Add TIMEOUT macro for cs2 client
Jean-Philippe Andre [Thu, 29 Aug 2013 07:46:25 +0000 (16:46 +0900)]
evas/cserve2: Add TIMEOUT macro for cs2 client

This will help debugging as we can disable the timeout on demand.

10 years agoevas/cserve2: Fix segfault in cserve2 slave
Jean-Philippe Andre [Tue, 27 Aug 2013 07:03:28 +0000 (16:03 +0900)]
evas/cserve2: Fix segfault in cserve2 slave

Slaves always crash when loading SVG images.
The svg module was simply not a dynamic module.

10 years agoevas/cserve2: Small performance fixes
Jean-Philippe Andre [Tue, 27 Aug 2013 06:35:22 +0000 (15:35 +0900)]
evas/cserve2: Small performance fixes

10 years agoevas/cserve2: Modify usage of EVAS_CSERVE2 env
Jean-Philippe Andre [Tue, 27 Aug 2013 03:47:51 +0000 (12:47 +0900)]
evas/cserve2: Modify usage of EVAS_CSERVE2 env

Check the value of "EVAS_CSERVE2" instead of just checking
if it is defined in the environment. So we can set to 0
to disable.

10 years agoevas/cserve2: Repack strings index when too fragmented
Jean-Philippe Andre [Tue, 27 Aug 2013 02:14:21 +0000 (11:14 +0900)]
evas/cserve2: Repack strings index when too fragmented

10 years agoevas/cserve2: Fix NULL vs. empty string
Jean-Philippe Andre [Tue, 27 Aug 2013 00:56:50 +0000 (09:56 +0900)]
evas/cserve2: Fix NULL vs. empty string

Server side: Some strings were NULL, some other were empty.
Client side was looking for NULL, but the shared index contained
empty key.

10 years agoevas/cserve2: Rename _evas_cache_ to _evas_cache2_
Jean-Philippe Andre [Mon, 26 Aug 2013 10:59:24 +0000 (19:59 +0900)]
evas/cserve2: Rename _evas_cache_ to _evas_cache2_

10 years agoevas/cserve2: Fix crashes in E17/cserve2
Jean-Philippe Andre [Mon, 26 Aug 2013 08:32:09 +0000 (17:32 +0900)]
evas/cserve2: Fix crashes in E17/cserve2

When remapping the glyph data buffer, we need to reposition
the glyph pointers as well.

10 years agoevas/cserve2: Scan glyphs and add to fash
Jean-Philippe Andre [Mon, 26 Aug 2013 07:48:49 +0000 (16:48 +0900)]
evas/cserve2: Scan glyphs and add to fash

This will save (just) a few socket waits.

10 years agoevas/cserve2: Remap glyph mempool if resized
Jean-Philippe Andre [Mon, 26 Aug 2013 06:40:54 +0000 (15:40 +0900)]
evas/cserve2: Remap glyph mempool if resized

10 years agoevas/cserve2: Use only one Glyph_Map per font
Jean-Philippe Andre [Mon, 26 Aug 2013 06:31:52 +0000 (15:31 +0900)]
evas/cserve2: Use only one Glyph_Map per font

In the new cs2 model, only one shared buffer is allocated
for a font, and it resizes on demand.

10 years agoevas/cserve2: Remap indexes when count is wrong
Jean-Philippe Andre [Fri, 23 Aug 2013 09:31:30 +0000 (18:31 +0900)]
evas/cserve2: Remap indexes when count is wrong

When advertised count is wrong, try to remap the index.
This will avoid errors with string_get().

10 years agoevas/cserve2: Fix lag with invalid font glyphs
Jean-Philippe Andre [Fri, 23 Aug 2013 09:01:22 +0000 (18:01 +0900)]
evas/cserve2: Fix lag with invalid font glyphs

10 years agoevas/cserve2: Add valid flag on Image_Data and File_Data
Jean-Philippe Andre [Fri, 23 Aug 2013 06:10:03 +0000 (15:10 +0900)]
evas/cserve2: Add valid flag on Image_Data and File_Data

Images and Files can be discovered by the client scanner before they
are valid (aka. loaded or opened). We want clients to ignore all
shared objects that are not ready yet, as they are in an undefined
state (values and memory might be invalid).

10 years agoevas/cserve2: Remap string entries when size changed
Jean-Philippe Andre [Fri, 23 Aug 2013 05:52:11 +0000 (14:52 +0900)]
evas/cserve2: Remap string entries when size changed

Lots of files can't be found by the client when the strings table
changed, because it was not remapped properly.

10 years agoevas/cserve2: Optimize shared strings
Jean-Philippe Andre [Thu, 22 Aug 2013 07:53:51 +0000 (16:53 +0900)]
evas/cserve2: Optimize shared strings

Shared string indexes are not repacked, since they live in a
memory pool (where the indexes can be reused).
So, the index in the table is equal to their ID. Add initial
test to check if the item at index n has the ID n.

10 years agoevas/cserve2: Add SHM debug tool (CLI)
Jean-Philippe Andre [Thu, 22 Aug 2013 01:43:53 +0000 (10:43 +0900)]
evas/cserve2: Add SHM debug tool (CLI)

Prints out the content or the shared index.
Does not need any socket connection to the server, read-only.

10 years agoevas/cserve2: Add glyph shm names to Font_Data
Jean-Philippe Andre [Thu, 22 Aug 2013 05:55:07 +0000 (14:55 +0900)]
evas/cserve2: Add glyph shm names to Font_Data

10 years agoevas/cserve2: Remove dead code.
Jean-Philippe Andre [Wed, 21 Aug 2013 07:43:20 +0000 (16:43 +0900)]
evas/cserve2: Remove dead code.

10 years agoevas/cserve2: Add proper tags to shm arrays
Jean-Philippe Andre [Wed, 21 Aug 2013 04:51:15 +0000 (13:51 +0900)]
evas/cserve2: Add proper tags to shm arrays

Use tags such as INDX, FONT, IMAG, FILE.
These can help debugging and for robustness, as they will
be visible in the memory dump.

Update generation_id after repack as well

10 years agoevas/cserve2: Fix cserve2_usage tool
Jean-Philippe Andre [Tue, 20 Aug 2013 09:26:02 +0000 (18:26 +0900)]
evas/cserve2: Fix cserve2_usage tool

Also, print shared arrays usage as it is sent as first message
upon connection to the server.

Add fonts to LRU and flush LRU as needed.
Seems to work okay. Could probably be finetuned a bit.

10 years agoevas/cserve2: Client-side scanning of Font Entries
Jean-Philippe Andre [Tue, 20 Aug 2013 04:09:02 +0000 (13:09 +0900)]
evas/cserve2: Client-side scanning of Font Entries

10 years agoevas/cserve2: Store Glyph_Data in shared mempools
Jean-Philippe Andre [Mon, 19 Aug 2013 03:22:19 +0000 (12:22 +0900)]
evas/cserve2: Store Glyph_Data in shared mempools

- Create one Glyph_Data array per Font_Entry
- Copy Font_Source descriptors in shared array
- Copy font entries in shared index
- Send font index path over socket
- Merge Font_Source and Font_Data in shared memory
- Send font index path on client connect
- Repack font index after free
- Pass font shm name to clients

10 years agoevas/cserve2: Add mempool_buffer_offset_get() api
Jean-Philippe Andre [Wed, 14 Aug 2013 04:57:41 +0000 (13:57 +0900)]
evas/cserve2: Add mempool_buffer_offset_get() api

10 years agoevas/cserve2: Store glyphs in shared arrays
Jean-Philippe Andre [Wed, 7 Aug 2013 09:48:00 +0000 (18:48 +0900)]
evas/cserve2: Store glyphs in shared arrays

Also, change the internal structure of glyph listing:
 - Use lists instead of inlists

Switch to use Glyph_Data instead of Glyph_Entry

10 years agoevas/cserve2: Add compile-time log level flag for cserve2
Jean-Philippe Andre [Wed, 7 Aug 2013 06:45:35 +0000 (15:45 +0900)]
evas/cserve2: Add compile-time log level flag for cserve2

DBG and INF messages pollute the logs and performance reports
from valgrind, let's disable them at build-time for better
benchmarking.

Fix other compilation warnings and clean code a bit

10 years agoevas/cserve2: Optimize cserve2_shared_string_del()
Jean-Philippe Andre [Tue, 6 Aug 2013 05:52:12 +0000 (14:52 +0900)]
evas/cserve2: Optimize cserve2_shared_string_del()

eina_hash_del_by_data is very expensive, so let's delete by key
since we can get the original data.

10 years agoevas/cserve2: Add binary search in server side
Jean-Philippe Andre [Fri, 2 Aug 2013 06:27:20 +0000 (15:27 +0900)]
evas/cserve2: Add binary search in server side

10 years agoevas/cserve2: Fix race condition on font load / socket fail
Jean-Philippe Andre [Fri, 2 Aug 2013 06:16:24 +0000 (15:16 +0900)]
evas/cserve2: Fix race condition on font load / socket fail

If the client exits while the server is loading fonts,
the font entry objects are accessed after being freed.

Note: This is not a proper fix. The race condition should still
be fixed, but delaying client deletion a bit might help for a while...

10 years agoevas/cserve2: Fix Eina_File leak at shutdown
Jean-Philippe Andre [Thu, 1 Aug 2013 09:39:38 +0000 (18:39 +0900)]
evas/cserve2: Fix Eina_File leak at shutdown

cserve2_shutdown was not called

10 years agoevas/cserve2: Scan shared indexes in client side
Jean-Philippe Andre [Wed, 31 Jul 2013 11:29:41 +0000 (20:29 +0900)]
evas/cserve2: Scan shared indexes in client side

Read Image descriptors from the shared arrays

But, accessing the shared index can be a bit expensive, so
try to read from the socket before scanning the index,
without blocking.

10 years agoevas/cserve2: Add _shared_index_remap_check()
Jean-Philippe Andre [Wed, 31 Jul 2013 10:31:50 +0000 (19:31 +0900)]
evas/cserve2: Add _shared_index_remap_check()

Check consistency of index header and remap if necessary.

10 years agoevas/cserve2: Share paths of index tables with clients
Jean-Philippe Andre [Mon, 29 Jul 2013 06:41:23 +0000 (15:41 +0900)]
evas/cserve2: Share paths of index tables with clients

- Send paths to all connected clients
- Send paths upon connection from a new client

- Client side:
 - Fast find opened images
 - Fast find loaded images
 - Optimize load by skipping socket wait

Also, remap indexes when needed (file size changed)
When index file grows, the clients are not notified, so they
must constantly check that their mapping matches the current array
size.

10 years agoevas/cserve2: Prepare shared indexes in clients
Jean-Philippe Andre [Mon, 29 Jul 2013 04:08:36 +0000 (13:08 +0900)]
evas/cserve2: Prepare shared indexes in clients

- Move required struct definitions to common evas_cs2.h

- Open/close shm and map arrays to memory
Pretty stupid for now as it will close & reopen the shm files
even when unchanged.

- Check shm size and array header consistency

- Implement message send

- Send messages from server when calling repack()

10 years agoevas/cserve2: Fix image load premultiplication
Jean-Philippe Andre [Thu, 25 Jul 2013 03:31:45 +0000 (12:31 +0900)]
evas/cserve2: Fix image load premultiplication

There was a color problem in elementary_test since elm
sets premul to 0 while expedite sets it to 1 during image load.

10 years agoevas/cserve2: _cserve2_cache_fast_scaling_check()
Jean-Philippe Andre [Wed, 24 Jul 2013 09:32:07 +0000 (18:32 +0900)]
evas/cserve2: _cserve2_cache_fast_scaling_check()

Adapt function to new Data/Entry model.

10 years agoevas/cserve2: Simplify error logs
Jean-Philippe Andre [Wed, 24 Jul 2013 09:17:52 +0000 (18:17 +0900)]
evas/cserve2: Simplify error logs

10 years agoevas/cserve2: Store Image_Data in shared array
Jean-Philippe Andre [Wed, 24 Jul 2013 05:34:35 +0000 (14:34 +0900)]
evas/cserve2: Store Image_Data in shared array

Split image entries into Image_Data and Image_Entry.
Store, find and use Image_Data descriptors from shared array.
Fix some wrong hash table usages.

10 years agoevas/cserve2: binary search in Shared_Array
Jean-Philippe Andre [Wed, 24 Jul 2013 01:59:51 +0000 (10:59 +0900)]
evas/cserve2: binary search in Shared_Array

10 years agoevas/cserve2: Repack File_Data shared array
Jean-Philippe Andre [Tue, 23 Jul 2013 10:30:54 +0000 (19:30 +0900)]
evas/cserve2: Repack File_Data shared array

10 years agoevas/cserve2: Use Shared_Array storage for File_Data
Jean-Philippe Andre [Mon, 22 Jul 2013 07:28:42 +0000 (16:28 +0900)]
evas/cserve2: Use Shared_Array storage for File_Data

File_Entry and File_Data are now two different types,
one being stored in the shared array, the other one
is only a reference counting entry.

10 years agoevas/cserve2: Rename File_Data pointers to fd
Jean-Philippe Andre [Fri, 19 Jul 2013 04:58:56 +0000 (13:58 +0900)]
evas/cserve2: Rename File_Data pointers to fd

The name fe is also used by Font_Entry, slightly confusing

10 years agoevas/cserve2: Use unique id counter
Jean-Philippe Andre [Mon, 22 Jul 2013 06:27:53 +0000 (15:27 +0900)]
evas/cserve2: Use unique id counter

Merge file id and image id into a single entry id.

10 years agoevas/cserve2: Implement shared index and buffers
Jean-Philippe Andre [Thu, 18 Jul 2013 10:39:33 +0000 (19:39 +0900)]
evas/cserve2: Implement shared index and buffers

Only import the C file for now.
Implement the following features:

- Shared Arrays
Store arrays of elements of fixed size in shm.

- Shared Mempool
Store random sized buffers in shm.
These buffers are indexed in a Shared Array and are
referred to using their index only.

- Shared Strings
Store strings in a shm in a way similar to Eina_Stringshare
(except strings are referred to using an int index).

- Include evas_cserve2_index.c to the compilation.
- Declare shared index functions in header file.
- Call init() and shutdown() on the shared index subsystem.
- Add find and foreach functions

10 years agoevas/cserve2: Add functions to resize shm files
Jean-Philippe Andre [Thu, 18 Jul 2013 10:21:58 +0000 (19:21 +0900)]
evas/cserve2: Add functions to resize shm files

- cserve2_shm_segment_request()
Creates a new shm segment (Shm_Handle) from an existing file,
without changing the current mmap.

- cserve2_shm_resize()
Grows the file and remaps in memory (new virtual address).

- cserve2_shm_id_get()
Returns the shm creation ID.

10 years agoevas/cserve2: Prepare usage of shared strings
Jean-Philippe Andre [Thu, 18 Jul 2013 08:54:06 +0000 (17:54 +0900)]
evas/cserve2: Prepare usage of shared strings

+ Add macros to fake usage of future cserve2 strings functions
+ Rename all File_Entry to fe
+ Other costmetic changes

10 years agoevas/cserve2: Add CRIT() macro
Jean-Philippe Andre [Thu, 18 Jul 2013 10:18:24 +0000 (19:18 +0900)]
evas/cserve2: Add CRIT() macro

For easy debugging

10 years agoevas - cover corner case for native surface shared refs on obj delete
Carsten Haitzler (Rasterman) [Mon, 28 Oct 2013 01:26:32 +0000 (10:26 +0900)]
evas - cover corner case for native surface shared refs on obj delete

10 years agoFix formatting
Chris Michael [Sun, 27 Oct 2013 15:44:31 +0000 (15:44 +0000)]
Fix formatting

Signed-off-by: Chris Michael <devilhorns@comcast.net>
10 years agoFix compiler warning(s) about eina_hash_int64 being passed pointer
Chris Michael [Sun, 27 Oct 2013 15:43:18 +0000 (15:43 +0000)]
Fix compiler warning(s) about eina_hash_int64 being passed pointer
from integer

Signed-off-by: Chris Michael <devilhorns@comcast.net>
10 years agoPKGBUILD cleanup
Doug Newgard [Sat, 26 Oct 2013 14:56:25 +0000 (23:56 +0900)]
PKGBUILD cleanup

Summary:
Some of the files in the "efl" package (emotion engine and others IIRC) were linked to libs in the efl_x11 package. This created a circular dep requiring both to be installed. Should just be one package in that case.
LGPL2 is actually LGPL2.1.
Added 'custom' to license array to cover COPYING.SMALL.
There is no 'dbus-core' package, it's part of 'dbus' now.
Doesn't build without 'check' installed.
provides_efl doesn't work, changed to just provides.
Got rid of tabs in depends and provides, replaced with spaces.
No need to add default 'strip', 'docs', and 'zipman' options, if someone want to override them, there's no reason to stop them.
buildflags doesn't work as a variable, added them to CFLAGS and CXXFLAGS manually. -O2 and -g are already included by default, if someone wants to change them on their system, there's no reason to override them.
Add install script to update mime database.
Change shebang on eina-bench-cmp so it runs with python2, python is 3 on Arch.
make dist doesn't tar Makefile, so we need to check that it exists before running "make clean distclean". Otherwise the build just fails.
Building of docs should be done in the build function.
Quoted all uses of $pkgdir and $srcdir
Each package function gets it's own $pkgdir, so rm-ing them at the beginning of the function isn't needed.
Use cp -a instead of cp -r to make sure permissions are preserved.
Set correct pkgdesc, arch, license, depends, provides, and install for the doc package.

Reviewers: raster

Reviewed By: raster

CC: cedric
Differential Revision: https://phab.enlightenment.org/D240

10 years agoevas gl engine related - fix elm image example 01 (T182).
Carsten Haitzler (Rasterman) [Sat, 26 Oct 2013 12:09:17 +0000 (21:09 +0900)]
evas gl engine related - fix elm image example 01 (T182).

this fixes https://phab.enlightenment.org/T182 as it is an issue with a
surface alloc overwriting an already allocated surface entirely inside
the general software image infra.

10 years agofix ecore evas pointer warping
discomfitor [Fri, 25 Oct 2013 16:14:43 +0000 (17:14 +0100)]
fix ecore evas pointer warping

10 years agoeina: add test for old ABI supported function, but not exposed anymore.
Cedric Bail [Fri, 25 Oct 2013 11:16:10 +0000 (20:16 +0900)]
eina: add test for old ABI supported function, but not exposed anymore.

10 years agoeina: improve test coverage of Eina_File API.
Cedric Bail [Fri, 25 Oct 2013 10:57:57 +0000 (19:57 +0900)]
eina: improve test coverage of Eina_File API.

10 years agoeina: improve test coverage of virtual file.
Cedric Bail [Fri, 25 Oct 2013 10:50:14 +0000 (19:50 +0900)]
eina: improve test coverage of virtual file.

10 years agoeina: use Eina_Barrier to improve coverage and reduce race condition during testing.
Cedric Bail [Fri, 25 Oct 2013 06:12:28 +0000 (15:12 +0900)]
eina: use Eina_Barrier to improve coverage and reduce race condition during testing.

This should hopefully solve the failing test on Jenkins. As I can't reproduce them
here, it is just an educated guess... We will see !

10 years agoEvasGL: Fixed direct rendering not clipping issue
Sung W. Park [Thu, 24 Oct 2013 08:37:22 +0000 (17:37 +0900)]
EvasGL: Fixed direct rendering not clipping issue

Evas GL direct rendering mode didn't properly take into account
the image object's clipping information and clip the region that
it was directly rendering to. Hence there were issues with the
direct rendering region drawing over the objects that are sitting
on top of it.

Also, cleaned up the direct rendering coordinate computation code
and a nasty dependency with image object that should have been
removed a long time ago.  Basically the evas-gl engine was directly
accessing the image object data structure for its data when it
really should have just passed along necessary information.

10 years agoEcore_Legacy.h: fixed typo while reading it.
Daniel Juyung Seo [Wed, 23 Oct 2013 13:23:22 +0000 (22:23 +0900)]
Ecore_Legacy.h: fixed typo while reading it.

10 years agoefreet: Fix file opening
Sebastian Dransfeld [Tue, 22 Oct 2013 10:54:22 +0000 (12:54 +0200)]
efreet: Fix file opening

- Store result from mkstemp so we don't leak file descriptor
- No need to chmod, as we set correct umask before createing file
- Use fdopen to open file from file descriptor, not from path

10 years agoUpdate po files
Sebastian Dransfeld [Tue, 22 Oct 2013 10:49:50 +0000 (12:49 +0200)]
Update po files

10 years agofor ecore system time changed - use a real time in the future, not
Carsten Haitzler (Rasterman) [Tue, 22 Oct 2013 07:11:00 +0000 (08:11 +0100)]
for ecore system time changed - use a real time in the future, not
time 0

for ECORE_EVENT_SYSTEM_TIMEDATE_CHANGED we use a timerfd on linux (and
also support talking to systemd) to detet time/date changes. the
timerfd was set up to go off at the absolute time of 0. since that is
almost always... in the past.. lets set a REAL time in the future.
(almost end of time)

10 years agoEvas: Add API to access private function
Daniel Zaoui [Sun, 13 Oct 2013 03:55:28 +0000 (06:55 +0300)]
Evas: Add API to access private function

We need a function in DnD for the overlapping feature to retrieve all
the objects at a specific position.
A private function exists in Evas (evas_event_objects_event_list) for
that purpose.
We need an access to this function for Elementary but we don't
want the is_frozen flag to be checked.

10 years agoecore examples: move pkg-config at the end of gcc command.
Daniel Juyung Seo [Mon, 21 Oct 2013 15:45:11 +0000 (00:45 +0900)]
ecore examples: move pkg-config at the end of gcc command.

This fixed build break on my machine.

10 years agoevas: change sync all canvases function name.
Ulisses Furquim [Thu, 17 Oct 2013 19:13:24 +0000 (16:13 -0300)]
evas: change sync all canvases function name.

Change evas_render_sync() to evas_all_sync() to better describe what it does
and avoid confusions with evas_render() and friends.

10 years agoecore-evas - handle mouse out then in due to click-to-focus passive grabs
Carsten Haitzler (Rasterman) [Fri, 18 Oct 2013 10:23:38 +0000 (19:23 +0900)]
ecore-evas - handle mouse out then in due to click-to-focus passive grabs

10 years agoevas: Fixed the textblock format to be drawn according to the glyph's horizontal...
Youngbok Shin [Fri, 18 Oct 2013 10:25:59 +0000 (11:25 +0100)]
evas: Fixed the textblock format to be drawn according to the glyph's horizontal advance width.

Summary:
Some characters have different two value on glyph's width and horizontal advance width.
If the glyph's width is smaller than advance width, format can be drawn weird.

Test Plan:
Set underline:on to the entry style and just insert the following characters.





Reviewers: tasn, woohyun

CC: cedric
Differential Revision: https://phab.enlightenment.org/D270

10 years agoonly unswallow deleted swallowed parts if they're in a swallow (wtf.)
Mike Blumenkrantz [Thu, 17 Oct 2013 09:17:07 +0000 (10:17 +0100)]
only unswallow deleted swallowed parts if they're in a swallow (wtf.)

10 years agoedje should probably unswallow objects when they get deleted externally to avoid...
Mike Blumenkrantz [Thu, 17 Oct 2013 09:07:23 +0000 (10:07 +0100)]
edje should probably unswallow objects when they get deleted externally to avoid issues later on?

10 years agoedc.vim: added border_scale_by.
Daniel Juyung Seo [Thu, 17 Oct 2013 08:01:44 +0000 (17:01 +0900)]
edc.vim: added border_scale_by.

10 years agocosmetic/logic move evas_sync call into same if block when applying alpha settings...
discomfitor [Wed, 16 Oct 2013 16:57:07 +0000 (17:57 +0100)]
cosmetic/logic move evas_sync call into same if block when applying alpha settings that might be deferred

10 years agoevas smart callback deletion no longer crashes when called during object deletion
discomfitor [Wed, 16 Oct 2013 16:12:50 +0000 (17:12 +0100)]
evas smart callback deletion no longer crashes when called during object deletion

10 years agoasync render -> alpha set. if not visible dont WAIT. do it now.
Carsten Haitzler (Rasterman) [Wed, 16 Oct 2013 15:00:05 +0000 (00:00 +0900)]
async render -> alpha set. if not visible dont WAIT. do it now.

11 years agoevas - gl - dont allocate image and/or texture until needed on native unset
Carsten Haitzler (Rasterman) [Wed, 16 Oct 2013 08:20:14 +0000 (17:20 +0900)]
evas - gl - dont allocate image and/or texture until needed on native unset

11 years agoRevert "efreet: Use eina_file_mkstemp"
Carsten Haitzler (Rasterman) [Wed, 16 Oct 2013 04:38:40 +0000 (13:38 +0900)]
Revert "efreet: Use eina_file_mkstemp"

This reverts commit bd5a5e7fce3931476bbc68a156be28e58f5d7ef8.

sorry about this... all fine eina mkstemp use and all... but this
*BREAKS EFREET*! ibar icons stop working (vanish). efreet no longer
issues a desktop file update event to e etc. etc. etc. ... so let's
keep code that works for now even if it uses mkstemp and not eina...
:) working is better. :)

11 years agoevas: simplify logic for textblock style parameter parsing.
Cedric Bail [Wed, 16 Oct 2013 09:14:33 +0000 (18:14 +0900)]
evas: simplify logic for textblock style parameter parsing.

11 years agoeina: honor string length in eina_tmpstr_add_length.
Cedric Bail [Wed, 16 Oct 2013 09:14:15 +0000 (18:14 +0900)]
eina: honor string length in eina_tmpstr_add_length.

11 years agoevas: fix doxygen
Yakov Goldberg [Wed, 16 Oct 2013 09:05:27 +0000 (12:05 +0300)]
evas: fix doxygen

11 years agoevas: request ellipsis from the right font.
Cedric Bail [Wed, 16 Oct 2013 08:57:34 +0000 (17:57 +0900)]
evas: request ellipsis from the right font.

11 years agoevas: fix doxygen
Yakov Goldberg [Wed, 16 Oct 2013 08:20:20 +0000 (11:20 +0300)]
evas: fix doxygen

11 years agoeina: just for a little bit of better portability.
Cedric Bail [Wed, 16 Oct 2013 07:54:36 +0000 (16:54 +0900)]
eina: just for a little bit of better portability.

11 years agoeina: let's add some straight forward test.
Cedric Bail [Wed, 16 Oct 2013 07:47:21 +0000 (16:47 +0900)]
eina: let's add some straight forward test.

11 years agoeina: And guard the right function with umask
Sebastian Dransfeld [Wed, 16 Oct 2013 02:56:51 +0000 (04:56 +0200)]
eina: And guard the right function with umask

Thanks Cedric.

11 years agoefreet: Use eina_file_mkstemp
Sebastian Dransfeld [Wed, 16 Oct 2013 02:48:20 +0000 (04:48 +0200)]
efreet: Use eina_file_mkstemp

11 years agoSet secure file permissions for temporary file
Sebastian Dransfeld [Wed, 16 Oct 2013 02:28:53 +0000 (04:28 +0200)]
Set secure file permissions for temporary file

From glibc mkstemp man page:
       In glibc versions 2.06 and earlier, the file is created with
       permissions 0666, that is, read and write for all users.  This old
       behavior may be a security risk, especially since other UNIX flavors
       use 0600, and somebody might overlook this detail when porting
       programs.  POSIX.1-2008 adds a requirement that the file be created
       with mode 0600.

       More generally, the POSIX specification of mkstemp() does not say
       anything about file modes, so the application should make sure its
       file mode creation mask (see umask(2)) is set appropriately before
       calling mkstemp() (and mkostemp()).

And:
http://cwe.mitre.org/data/definitions/377.html

11 years agoevas/map - Fixed the map surface to be updated properly.
ChunEon Park [Tue, 15 Oct 2013 13:24:27 +0000 (22:24 +0900)]
evas/map - Fixed the map surface to be updated properly.

By tranversing the whole of the map target's tree,
it decides to update the map surface to be redrawn if one of the objects is changed except the clipper itself.

11 years agoFix horrid formatting
Chris Michael [Tue, 15 Oct 2013 12:40:50 +0000 (13:40 +0100)]
Fix horrid formatting

Signed-off-by: Chris Michael <cp.michael@samsung.com>
11 years agoFix doxy typo
Chris Michael [Tue, 15 Oct 2013 12:38:15 +0000 (13:38 +0100)]
Fix doxy typo

Signed-off-by: Chris Michael <cp.michael@samsung.com>
11 years agoRemove blank line
Chris Michael [Tue, 15 Oct 2013 12:37:47 +0000 (13:37 +0100)]
Remove blank line

Signed-off-by: Chris Michael <cp.michael@samsung.com>
11 years agoedje: handle reallocation of pointed data properly for limits.
Cedric Bail [Tue, 15 Oct 2013 09:15:07 +0000 (18:15 +0900)]
edje: handle reallocation of pointed data properly for limits.

11 years agoevas - fix textblock handling of effects for shadows and so on in some cases
Carsten Haitzler (Rasterman) [Tue, 15 Oct 2013 07:04:40 +0000 (16:04 +0900)]
evas - fix textblock handling of effects for shadows and so on in some cases

11 years agoEdje - Fixed memory leak in the edje map color transition.
ChunEon Park [Tue, 15 Oct 2013 06:00:13 +0000 (15:00 +0900)]
Edje - Fixed memory leak in the edje map color transition.

Some cases, It doesn't free the interpolated colors.
Now it has guaranteed to free them always.

11 years agoRevert "ecore_evas: window does not go back to normal state if you set iconified...
Carsten Haitzler (Rasterman) [Tue, 15 Oct 2013 01:45:37 +0000 (10:45 +0900)]
Revert "ecore_evas: window does not go back to normal state if you set iconified to EINA_FALSE."

This reverts commit 9bfb730ff57da3d8789f80256abedcbc19e98bfe.

wrong! on deiconify this will now FORCE a ACTIVATE request to the wm
(which asks the wm to focus the window/hilight it and make it active.
might switch desktops or whatever). this is wrong. de-icoifying a
windows does NOT mean ALSO activating it!. the log message also doesnt
say what "go back to normal state"... means.... do they expect/INSIST
windows MUSt be focused after de-iconification? that's wrong as it's a
window management FOCUS policy, not something to do in the lib.

11 years agoRevert "eina: fix CID 1106340: Logically dead code (DEADCODE) reported by coverity."
Carsten Haitzler (Rasterman) [Tue, 15 Oct 2013 01:35:10 +0000 (10:35 +0900)]
Revert "eina: fix CID 1106340: Logically dead code (DEADCODE) reported by coverity."

This reverts commit 1c1b48ca64c4d13cfd3677cde1ee99e4cf6c2c15.

this commit is causing all sorts of hell and crashes, hangs and what
not. like:

Thread 1 (Thread 0x7feb7389e780 (LWP 7149)):
No symbol table info available.
No symbol table info available.
    at lib/eina/eina_inlist.c:422
            l = 0xf38a50
                    __FUNCTION__ = "eina_inlist_demote"
                    #3  0x00007feb6a809867 in
_eina_chained_mempool_alloc_in (
    pool=pool@entry=0x8650a0, p=p@entry=0xf38a50)
        at modules/eina/mp/chained_pool/eina_chained_mempool.c:197
                mem = 0xf3a690
                #4  0x00007feb6a809dff in eina_chained_mempool_malloc
(data=0x8650a0,
    size=<optimized out>)
        at modules/eina/mp/chained_pool/eina_chained_mempool.c:300
                pool = 0x8650a0
                        p = 0xf38a50
                                mem = <optimized out>
                                #5  0x00007feb6a7e451f in
eina_mempool_malloc (size=40, mp=<optimized out>)
    at lib/eina/eina_inline_mempool.x:103
    No locals.
    #6  _eina_list_mempool_list_new (list=0x7feb5402a510)
        at lib/eina/eina_list.c:199
        No locals.
        #7  eina_list_append (list=0x7feb5402a510, data=0xf41720)
            at lib/eina/eina_list.c:534
                    l = <optimized out>
                            __FUNCTION__ = "eina_list_append"
                            #8  0x0000000000445d1a in
e_bindings_signal_add (ctxt=<optimized out>,
    sig=0x947e6c "mouse,clicked,[12]", src=0x92248c "e.event.close",
        mod=E_BINDING_MODIFIER_NONE, any_mod=1, action=0x93479c
"window_close",
    params=0x0) at e_bindings.c:1036
            binding = 0xf41720
            #9  0x00000000004469d4 in e_bindings_init () at
e_bindings.c:68
        ebs = 0x947df0
                ebm = <optimized out>
                        ebw = <optimized out>
                                ebe = <optimized out>
                                        ebk = <optimized out>
                                                eba = <optimized out>
                                                        l = 0x9532e0
                                                        #10
0x0000000000437516 in main (argc=<optimized out>, argv=<optimized out>)
    at e_main.c:941
            safe_mode = 0 '\000'
                    after_restart = 1 '\001'
                            waslocked = 0 '\000'
                                    t = <optimized out>
                                            tstart = <optimized out>
                                                    s = <optimized out>
                                                            buff =
"1381800872.3", '\000' <repeats 19 times>
        action = {__sigaction_handler = {
                    sa_handler = 0x4fca00 <e_sigabrt_act>,
                                sa_sigaction = 0x4fca00
<e_sigabrt_act>}, sa_mask = {__val = {
              0 <repeats 16 times>}}, sa_flags = -1073741820,
                        sa_restorer = 0x0}
                                __FUNCTION__ = "main"
                                Detaching from program:
/usr/local/bin/enlightenment, process 7149

11 years agoecore_file: fix left over borking.
Cedric Bail [Tue, 15 Oct 2013 01:21:03 +0000 (10:21 +0900)]
ecore_file: fix left over borking.

I think I teached raster well, my job here is done !

11 years agoecore-file - remove ifdefs for HAVE_CULR since ecore-con "always" has it
Carsten Haitzler (Rasterman) [Tue, 15 Oct 2013 00:16:26 +0000 (09:16 +0900)]
ecore-file - remove ifdefs for HAVE_CULR since ecore-con "always" has it

now that it's runtime dlopen'ed etc. ecore-con "always" has url
support via curl so ecore-file needs to assume that this can always
work and this just use ecore-con-url anyway.