summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2016-09-19eina: populate memory in the right limit.efl-1.8Cedric BAIL
2015-10-08ecore_evas: fix forgotten call to post render callback.Cedric BAIL
This fix using expedite with buffer backend.
2015-03-24evas: fix use after free.Cedric BAIL
Thanks to Thanatermesis for tracking this in T2135.
2015-01-07evas textgrid - fix drawing of codepoints that are 0Carsten Haitzler (Rasterman)
this fixes artifacts if you have cells with 0 codepoints and also have underlines or strikethroughs in them. @fix
2014-10-20Revert "evas: GL_X11 context need to always be with alpha or it will fail to ↵Eduardo Lima (Etrunko)
change." This reverts commit 8862fc38b451fb8a598876111871336fe0749387.
2014-10-09evas: GL_X11 context need to always be with alpha or it will fail to change.Cedric BAIL
@fix
2014-09-30evas: GL_X11 - make sure to re-create the gl context when window visual changeJulien Isorce
Reviewers: cedric Subscribers: seoz, capOM, cedric Differential Revision: https://phab.enlightenment.org/D1229 Signed-off-by: Cedric BAIL <cedric@osg.samsung.com>
2014-09-02evas: actually just define the right number of events.Cedric BAIL
2014-09-02evas: forgotten evas canvas event EVAS_CANVAS_EVENT_DEVICE_CHANGED.Cedric BAIL
2014-09-02evas: fix evas events order and count.Cedric BAIL
Before this patch, if you were hooking to any event above EVAS_OBJECT_EVENT_IMAGE_PRELOADED you will end up receiving another event instead and be fooled into receiving the right one. In the mean time I did discover that our table is missing one item, the equivalent to EVAS_CALLBACK_DEVICE_CHANGED. I don't know how to properly fix that one for now, so put NULL instead to at least avoid a random memory access.
2014-08-21xlib: fix undefined behaviour when freeing colormap too earlyJulien Isorce
Summary: On Ubuntu 14.04 it makes a 32 bit depth window un-responsive to any XEvent. Reviewers: cedric, raster Reviewed By: raster Subscribers: raster, capOM, cedric Differential Revision: https://phab.enlightenment.org/D1236
2014-08-21ecore_evas: Ecore_Evas_X - make sure that ecore evas post render is always ↵Cedric BAIL
called. This fix a case where when doing asynchronous rendering and no update was done, the post render evas event would never be triggered and so no call to ecore evas post render callback were made. Now the async X backend follow all other backend behavior. @fix
2014-07-14lz4: fix another security issue on ARM 32bits.Cedric BAIL
2014-07-04Evas: Update for giflib 5.1 api changeDoug Newgard
Summary: Slight change in the file close function. Reviewers: jpeg CC: cedric Differential Revision: https://phab.enlightenment.org/D916
2014-06-30lz4: fix possible security issue.Cedric BAIL
See : - http://fastcompression.blogspot.fr/2014/06/lets-move-on.html - http://www.openwall.com/lists/oss-security/2014/06/26/25 @fix
2014-06-12efreet: remove code duplicationSebastian Dransfeld
If the menu starts with either kde- or gnome- the resulting action is the same. (cherry picked from commit a10d4167f11074f338f9d8cf6f8f8023fc438b05) Signed-off-by: Thanatermesis <thanatermesis@gmail.com>
2014-06-12efreet: improve handling of merged menusSebastian Dransfeld
If the menu is of the format ${XDG_MENU_PREFIX}applications.menu, we want to merge applications-merged/, not ${XDG_MENU_PREFIX}applications-merged (cherry picked from commit 9415f8a7ef24e051a5ebf812324fe0f659cf89ea) Signed-off-by: Thanatermesis <thanatermesis@gmail.com>
2014-06-12efreet: Correctly handle default merge dirSebastian Dransfeld
When main menu file is gnome-applications.menu or kde-applications.menu, we should be using applications as base for merged dir. Part of T1284 (cherry picked from commit ffda8635fe8d7140c28044d0affc64697f1eb550) Signed-off-by: Thanatermesis <thanatermesis@gmail.com>
2014-06-06Updating catalan translationmaxerba
2014-05-26Updating galician, lithuanian and adding turkish translationsmaxerba
2014-05-19Updating translationsmaxerba
2014-03-17edje: check font change in edje text part cache infra.Cedric BAIL
@fix
2014-03-07edje/entry: fix to not emit "changed" signal in unnecessary cases of ↵Youngbok Shin
password mode. Summary: When text is hidden by password character, "changed" signal should be emitted. But, even if there is no visible text, the signal was emitted. @fix Reviewers: woohyun, tasn, cedric Reviewed By: cedric CC: cedric Differential Revision: https://phab.enlightenment.org/D604 Signed-off-by: Cedric BAIL <cedric.bail@samsung.com>
2014-03-03release: Update NEWS and bump version for 1.8.6 releasev1.8.6Stefan Schmidt
2014-03-03Revert "evas: replace EINA_LIST_FREE to EINA_LIST_FOREACH_SAFE."Stefan Schmidt
This reverts commit eb022cba1d1df351d5f5043879eade5cee9d918c. This does not even compile. Blindly backported without any testing.
2014-03-03evas: replace EINA_LIST_FREE to EINA_LIST_FOREACH_SAFE.WooHyun Jung
EINA_LIST_FREE does eina_list_remove_list, and clip_unset does the same thing to the same list pointer. So, EINA_LIST_FOREACH_SAFE is proper for this case.
2014-02-28examples/evas: Set a proper format string for fprintfStefan Schmidt
Instead of abusing the format string we should set it. Was made aware of this problme by chinmaya061 in https://phab.enlightenment.org/D585 @fix
2014-02-27ecore: @fix race condition when using ecore_main_loop_thread_safe_call_sync.Cedric Bail
2014-02-19Evas text utils: Fixed walking compound clusters.Tom Hacohen
In some scripts, like Devanagari, clusters can be split across more than just one glyph. This is now fixed. Thanks to YoungBok Shin for reporting.
2014-02-19Evas textblock tests: Add a wrapping test for complex cluster wrapping.Tom Hacohen
This test uses some Devanagari text that should have more complex clusters than what latin text can provide. This is a more complex wrapping case that should be tested and haven't been tested until now.
2014-02-12eina: allow eina_time_get to fall back to other clocks if the first one failsAlbin Tonnerre
Summary: eina_time_get tries to use only one clock which is defined at compile-time and returns the result of that one. This causes problems on platforms where eg. CLOCK_PROCESS_CPUTIME_ID is defined but the clock is actually not implemented (ie. clock_gettime returns EINVAL), as we simply don't get any time at all. Instead, make sure we include the code for all defined clocks and simply fall back to other clocks if the previous ones aren't implemented. Reviewers: cedric, raster Reviewed By: cedric CC: cedric Differential Revision: https://phab.enlightenment.org/D547 Signed-off-by: Cedric BAIL <cedric.bail@samsung.com>
2014-02-10Evas textblock: Fix rendering of spaces in some cases.Tom Hacohen
The issue was with a textblock that's being resized and a space between formats. The problem is, that the text would get trimmed when wrapping, and then not restored, because it had nothing to merge to. This fixes T924.
2014-02-10Evas textblock: Added a test for rendering of spaces in some cases.Tom Hacohen
This checks that spaces are one again used, in a specific case with wrapping.
2014-02-07Evas textblock: Fixed *_cursor_char_coord_set with some y values.Tom Hacohen
Specifically y values that are between par->y and the first ln->y. This should fix some selection issues with elm.
2014-02-07Evas textblock: Added tests for *_cursor_char_coord_set.Tom Hacohen
There are issues with values between the beginning of the first par and the beginning of the first line.
2014-02-05eina: restore magic debug option for ABI compatibility purpose.Cedric BAIL
We can't remove that option as it break ABI for people who did turn it off in 1.7. - cherry-pick me -
2014-02-03edje freeze count can no longer become negativeMike Blumenkrantz
2014-01-30evas: fix tiler bug (makes evas get stuck in an infinite loop there in ↵Daniel Kolesa
certain scenarios) backport of master commit 871597f8c20f3fc418df64669f36f9fae76d89a8.
2014-01-29evas: textgrid - fix crash in terminology with rare characters.Jean-Philippe Andre
So I have a weird crash in terminology. Reproduction path: eet -x /path/to/elm/theme/default.edj edje/images/537 Scroll back in the terminal buffer, to show the entire file: CRASH. Reviewers: cedric, tasn CC: cedric, raster Differential Revision: https://phab.enlightenment.org/D468 Signed-off-by: Cedric BAIL <cedric.bail@samsung.com>
2014-01-29Evas textblock: Fix selection with bidi text.Tom Hacohen
This is a regression introduced in 548e548632333f5ef533079a7aaa54aadd20ef8d. This is really bad, and essentially broke selection geometry for bidi text. Very serious. The problematic code assumed that the range comparison for the items assumed the item marked with 1 is always logically before the item marked with 2, which is just not true.
2014-01-29Evas textblock: Added a test for range geometry with bidi text.Tom Hacohen
2014-01-28Edje entry: Fix memory corruption (of by one in allocation).Tom Hacohen
This fixes T867.
2014-01-27Edje entry: Fix anchor href parsing.Tom Hacohen
It should now correctly work with arguments surrounded with either quotes or spaces.
2014-01-27release: Update NEWS and bump version for 1.8.5 releasev1.8.5Stefan Schmidt
2014-01-24embryo: force the insertion of a '/' in the path to be sure that file path ↵Cedric BAIL
make sense.
2014-01-20eet_data: guarantee double-word alignment for temporary stack buffer.Gustavo Sverzut Barbieri
The code was giving enough memory to store doubles and longs, but they could be unaligned as "unsigned char" allows 1-byte alignment, while double may require 8 bytes. By specifying the array as "long long" we force certain alignment in a platform independent way. As this array is small enough and short-lived, the number of items were not changed, this results in more bytes on the stack but it shouldn't matter.
2014-01-20eina/mp/one_big: fix alignment issues.Gustavo Sverzut Barbieri
When over-allocating (past "pool->max" items) a memory slice will be allocated to the new item as a linked list using Eina_Inlist. The original code was placing the Eina_Inlist structure (3 pointers) at the beginning of the allocated memory. However the item must have proper alignment based on "pool->item_size", otherwise a structure may end with unaligned members. Take for example MIPS 32 bits, it uses 4 bytes pointers with 8 bytes double. A structure containing a double could have it unaligned as 12 % 8 = 4 (12 is the size of Eina_Inlist, that contains 3 pointers), and MIPS doesn't allow unaligned access. Albin Tonnerre (Lutin) spotted this in his Debian MIPS test machine, it was breaking at eet_data_get_double() that was storing an unaligned double. This was being called from within edje test suite. The current code will place the list node after the requested "pool->item_size", of course guaranteeing the pointer inside the node is aligned (otherwise a "char" or "short" would break its alignment).
2014-01-20configure: include required NEON build flags.Gustavo Sverzut Barbieri
It is required to use "-mfpu=neon" to build NEON. Patch-by: Albin Tonnerre <lutin@debian.org>
2014-01-20eina-inlist - handle prepend/append relative if list is null (empty)Carsten Haitzler (Rasterman)
stable release - cherry-pick me!
2014-01-11eina: fix memory leak of inserting the wrong value in the hash.Cedric Bail
We where inserting the pointer data instead of the pointer, leading to unaligned access on Sparc (Thanks Lutin to report it and Debian tools/infra to help us catch it) and also a memory leak.