summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
5 daysMerge branch 'master' into devs/hermet/lottiedevs/hermet/lottieHermet Park
7 daysecore - efl thread - dont close invalid < 0 fd'sCarsten Haitzler (Rasterman)
fix CID 1396951
7 daysecore - efl exe - check returns of open and dup2 for stdin etc. fdsCarsten Haitzler (Rasterman)
check returns - fix CID 1397002
7 dayseina bench - have default values inside benc dtoaCarsten Haitzler (Rasterman)
addess CID 1400856
7 daysevas test - check return of ftell and malloc and handle properlyCarsten Haitzler (Rasterman)
fix CID 1400871
7 dayseina thread - check return of pthread_attr_initCarsten Haitzler (Rasterman)
fix CID 1401031
7 daysemotion - check return of eet data writeCarsten Haitzler (Rasterman)
fix CID 1401040
7 daysethumb - fix idx range checkCarsten Haitzler (Rasterman)
fix CID 1402612
7 daysethumb - fix index range checkCarsten Haitzler (Rasterman)
fix CID 1402628
7 daysethumb - fix index range checkCarsten Haitzler (Rasterman)
fix CID 1402631
7 daysethumb - check size readCarsten Haitzler (Rasterman)
fix CID 1402633
7 daysethumb - check inoput index from stdinCarsten Haitzler (Rasterman)
fix CID 1402646
7 daysevas - psd loader - check read values to try make coverity happyCarsten Haitzler (Rasterman)
this is trying to fix CID 1402688, 1402687, 1402682, 1402665. not sure it will or will not but it does dot i's and cross t's
7 dayselm - collection view - check return and fix coverity wanringCarsten Haitzler (Rasterman)
fix CID 1405808
7 daysecore con url - curl is actually optional and only needs to init on useCarsten Haitzler (Rasterman)
ecore_con_utl_init is kind of pointless when we can init at the first use of a url and just have it fail. the problem is anyone initting ecore_con's url stuff will then pull in curl at that point and not laer "on first use" which is kind of nw how the dlopen fun was intended, so push it off until then.
8 dayselm_config: remove useless init|shutdown of ecore_win32 and eocore_cocoa ↵Vincent Torri
(done in ecore_evas module), and useless wayland variable ecore_win32 and ecore_cocoa was init. wayland variable was also unused Reviewed-by: Christopher Michael <devilhorns@comcast.net> Differential Revision: https://phab.enlightenment.org/D12149
8 dayscall sd_seat_can_tty to determine if ttys should be used, instead of ↵nerdopolis
hard-coding seat0 Summary: All TTYs are on seat0, but not all seat0s have TTYs Reviewers: #efl, devilhorns Reviewed By: #efl, devilhorns Subscribers: bu5hm4n, stefan_schmidt, raster, cedric, #reviewers, #committers Tags: #efl Reviewed-by: Christopher Michael <devilhorns@comcast.net> Differential Revision: https://phab.enlightenment.org/D12143
9 daysevas_outbuf: Fix memory leak.Hermet Park
Summary: In certain use cases, when image data is created twice, the existing memory pointer is isolated and a memory leak occurs. Reviewers: raster, Hermet Reviewed By: Hermet Subscribers: cedric, #reviewers, #committers Tags: #efl Maniphest Tasks: T8813 Differential Revision: https://phab.enlightenment.org/D12148
9 daysevas_main: Fix memory leak.Hermet Park
Summary: If the inlist has only one item when it removed. the data is not freed in descturctor. Reviewers: Hermet, raster, vtorri Reviewed By: Hermet Subscribers: vtorri, cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12147
9 daysevas_outbuf: Fix memory leak.Woochanlee
Summary: In certain use cases, when image data is created twice, the existing memory pointer is isolated and a memory leak occurs. Reviewers: raster, Hermet Reviewed By: Hermet Subscribers: cedric, #reviewers, #committers Tags: #efl Maniphest Tasks: T8813 Differential Revision: https://phab.enlightenment.org/D12148
9 daysevas_main: Fix memory leak.Woochanlee
Summary: If the inlist has only one item when it removed. the data is not freed in descturctor. Reviewers: Hermet, raster, vtorri Reviewed By: Hermet Subscribers: vtorri, cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12147
10 daysevas vg: check object changed as well 2Shinwoo Kim
Summary: evas cannot render vg object if Efl_Canvas_Vg_Node_Data flag is not EFL_GFX_CHANGE_FLAG_NONE and Evas_Object_Protected_Data.changed is FALSE, when vg object marks its node as changed. Above case could be possible if vg object render_pre is not called, and only nd->vd->obj->changed is set to false by evas_object_change_reset. Reviewers: Hermet, jsuya, herb Reviewed By: Hermet Subscribers: cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12142
10 daysevas_textblock: prevent textnodes with <ps>/<br> without format nodeAli Alzyod
this will prevent textnodes content with <ps> or <br> without format node Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org> Differential Revision: https://phab.enlightenment.org/D12145
11 daysvg_load_svg: Support dasharray attribute for strokeJunsuChoi
Summary: It supports stroke-dasharray, one of the stroke properties of svg. https://developer.mozilla.org/en-US/docs/Web/SVG/Attribute/stroke-dasharray Test Plan: [Test SVG] <svg viewBox="0 0 30 10" xmlns="http://www.w3.org/2000/svg"> <!-- No dashes nor gaps --> <line x1="0" y1="1" x2="30" y2="1" stroke="black" /> <!-- Dashes and gaps of the same size --> <line x1="0" y1="3" x2="30" y2="3" stroke="black" stroke-dasharray="4" /> <!-- Dashes and gaps of different sizes --> <line x1="0" y1="5" x2="30" y2="5" stroke="black" stroke-dasharray="4 1" /> <!-- Dashes and gaps of various sizes with an odd number of values --> <line x1="0" y1="7" x2="30" y2="7" stroke="black" stroke-dasharray="4 1 2" /> <!-- Dashes and gaps of various sizes with an even number of values --> <line x1="0" y1="9" x2="30" y2="9" stroke="black" stroke-dasharray="4 1 2 3" /> </svg> Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: kimcinoo, herb, cedric, #committers, #reviewers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12146
12 daysrelease: Update NEWS and bump version for 1.25.0-beta3 releasev1.25.0-beta3Stefan Schmidt
12 daysMerge branch 'master' into devs/hermet/lottieHermet Park
12 daysMerge branch 'feature/themes/flat' into devs/hermet/lottieHermet Park
2020-09-12TH - icons - module categoriesCarsten Haitzler (Rasterman)
2020-09-11Efl.Canvas.Vg: When VG container has transparency, ector buffer created ↵JunsuChoi
synchronously Summary: If the container has transparency, it internally alpha blends with ector buffer. So ector buffer must be created synchronously. Test Plan: N/A Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: #reviewers, #committers, cedric, herb, kimcinoo Tags: #efl Differential Revision: https://phab.enlightenment.org/D12140
2020-09-11Merge branch 'master' into devs/hermet/lottieHermet Park
2020-09-11vg_load_svg: Add Mask, ClipPath nodees as dummy.JunsuChoi
Summary: Mask and clippath TAG are not supported yet. If this TAG is used, the Child TAG declared with this TAG may cause problems. To prevent that, declare them as Unknown types until they are supported. (display=none) Test Plan: N/A Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: cedric, herb, #reviewers, kimcinoo, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12139
2020-09-11vg_load_svg: Fix missing loader.gradients updateJunsuChoi
Summary: When both <defs> and <svg> gradients are declared, the update of gradient declared in is omitted. Therefore, this patch fixes any missing gradient updates. Test Plan: N/A Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: #reviewers, #committers, cedric, herb, kimcinoo Tags: #efl Differential Revision: https://phab.enlightenment.org/D12138
2020-09-11Efl.Canvas.Vg.Object: Fix typoJunsuChoi
Summary: The 3th parameter of ector_buffer_pixels_get is the width. So chnage px to pw. Test Plan: N/A Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: cedric, herb, #reviewers, kimcinoo, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12141
2020-09-10TH - ctxpopupCarsten Haitzler (Rasterman)
2020-09-10docs: Correct the mismatched group name in Evas.Myoungwoon Roy, Kim
I had found mismatched group name in Evas and fixed them. Reviewed-by: Stefan Schmidt <stefan@datenfreihafen.org> Differential Revision: https://phab.enlightenment.org/D12134
2020-09-10vg_load_svg: Supports case when only rx or ry is declaredJunsuChoi
Summary: In relation to the declaration of rx and ry attribute of rect, the following three cases occur. rx="10" (or ry="10" rx="10" ry = "0" (or rx="0" ry = "10") rx="10" ry = "10" To cover these case, we check the rx and ry declarations. Test Plan: N/A Reviewers: Hermet, smohanty Subscribers: kimcinoo, herb, cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12137
2020-09-10vg_load_svg: Supports case when only rx or ry is declaredJunsuChoi
Summary: In relation to the declaration of rx and ry attribute of rect, the following three cases occur. rx="10" (or ry="10" rx="10" ry = "0" (or rx="0" ry = "10") rx="10" ry = "10" To cover these case, we check the rx and ry declarations. Test Plan: N/A Reviewers: Hermet, smohanty Subscribers: kimcinoo, herb, cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12137
2020-09-09evas vg: check object changed as wellShinwoo Kim
Summary: evas cannot render vg object if Efl_Canvas_Vg_Object_Data.changed is ture and Evas_Object_Protected_Data.changed is false, when vg object marks itself as changed. Above case is possible depending on the draw area which is calculated by eng_output_redraws_next_update_get. If this function returns NULL, the vg object render function is not called, and vd->changed remains true. Only vd->obj->changed is set to false by vas_object_change_reset. Reviewers: Hermet, jsuya, herb, raster Reviewed By: raster Subscribers: raster, cedric, #reviewers, #committers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12136
2020-09-07release: Update NEWS and bump version for 1.25.0-beta2 releasev1.25.0-beta2Stefan Schmidt
2020-09-07elm_gesture_layer: fix crash caused by elm_gesture_layer_cb_del in cbJaehyun Cho
elm_gesture_layer_cb_del can be called in Elm_Gesture_Event_Cb cb set by elm_gesture_layer_cb_set and cb_info of the cb is free in the cb. This causes crash if cb_info is searched with EINA_INLIST_FOREACH.
2020-09-07Merge branch 'feature/themes/flat' into devs/hermet/lottieHermet Park
2020-09-07Merge branch 'master' into devs/hermet/lottieHermet Park
2020-09-05elm cnp - make matchers const and static tooCarsten Haitzler (Rasterman)
2020-09-05elm_cnp: adjust image type to real formatMarcel Hollerbach
we should never say that we have image/png etc. if we do not have the respective format. Otherwise we are just confusing the other client. With this you can copy things from elm to thunderbird / libreoffice writer / firefox / telegram / chromium. Please note if you test this: Start *a new xorg session* it was discovered that xorg clients tent to be totally confused if there was a protocol error before, or some wrong mime type was delivered before, thunderbird even likes to crash if this is not done. This should at some point be extended to more image types.
2020-09-04ecore_evas_x: reply in *any* way with XdndStatusMarcel Hollerbach
otherwise Xorg stops sometimes to continue sending callbacks to us. Which results in a broken state maschine.
2020-09-04ecore x - fix typo in internal backdoor api to get vsync debugCarsten Haitzler (Rasterman)
2020-09-04ecore drm - increase timeout for flips from 0.05 to 2.0 secCarsten Haitzler (Rasterman)
so our timeout is far too agressive. we cant stay hung on buggy devices/drivers though so make the timeout 2s instead of 0.05s so we dont go retrying flips too agressively which causes stuttering @fix
2020-09-04TH - iconsCarsten Haitzler (Rasterman)
2020-09-04TH - border - fix csd issues in wl with contentCarsten Haitzler (Rasterman)
2020-09-04vg_load_svg: Fix pointing to wrong parentJunsuChoi
Summary: If defs type is open empty style, it is not included in stack. Because the current logic is weak to group tags in sibling relationships. If defs of open-empty type is placed before the group tag, the group tag refers to the wrong parent. This patch prevents it. This change is the same concept as the D12127, which was reverted to 5065c9d0c2. Test Plan: N/A Reviewers: Hermet, smohanty Reviewed By: Hermet Subscribers: cedric, herb, kimcinoo, #committers, #reviewers Tags: #efl Differential Revision: https://phab.enlightenment.org/D12129