On Wed, 17 Jun 2026 09:28:27 +0200 Jérémy Zurcher <[email protected]> said:
> Hello,
>
> on some machines, I have to revert 64b81d3256 to complete edj files
> compilation.
>
> rsvg miss -m :
> sed -i "s/\[rsvg\]/[rsvg, cc.find_library('m', required: true)]/"
> src/generic/evas/rsvg/meson.build
>
> with HEAD, this is what I do:
> rm -fr build
> meson setup --prefix=/opt/efl -Defl-one=true -Doptimization=2
> -Dharfbuzz=true -Dbuild-examples=false -Dbuild-tests=false -Ddrm=true
> -Dwl=true -Dbindings= build ninja -C build
>
> it ends up with 262 error codes,
> but the edj files built are valid & cmp says they are the same as those
> produced without 64b81d3256.
>
> [3003/3008] Generating data/ethumb/frames/edje_cc_ethumb_frame with a
> custom command FAILED: [code=262] data/ethumb/frames/default.edj
> /usr/bin/env EFL_RUN_IN_TREE=1 ./build/src/bin/edje/edje_cc -beta
> -fastcomp -sd ./data/ethumb/frames -id ./data/ethumb/frames
> -fd ./data/ethumb/frames ../data/ethumb/frames/default.edc
> data/ethumb/frames/default.edj corrupted size vs. prev_size [3004/3008]
> Generating data/elementary/objects/edje_cc_elm_test_tab_pager.edc with a
> custom command FAILED: [code=262]
> data/elementary/objects/test_tab_pager.edj /usr/bin/env
> EFL_RUN_IN_TREE=1 ./build/src/bin/edje/edje_cc -beta -fastcomp
> -no-warn-unused-images -sd ./data/elementary/themes/snd
> -id ./data/elementary/themes/img -id ./data/elementary/objects
> -fd ./data/elementary/objects ../data/elementary/objects/test_tab_pager.edc
> data/elementary/objects/test_tab_pager.edj corrupted size vs. prev_size ...
>
>
> if I run this, everything compiles nicely, so it leads to a fd close issue in
> multi-threads environment.
>
> for f in cursors font_preview multibuttonentry multip postit_ent test
> test_external test_focus_style test_masking test_pager test_prefs
> test_tab_pager test_tooltip do /usr/bin/env
> EFL_RUN_IN_TREE=1 ./build/src/bin/edje/edje_cc -beta -fastcomp
> -no-warn-unused-images \ -sd ./data/elementary/themes/snd \
> -id ./data/elementary/themes/img \ -id ./data/elementary/objects \
> -fd ./data/elementary/objects \
> "./data/elementary/objects/${f}.edc" "/tmp/${f}.edj"
> done
>
> what are your thoughts on this ?
hmm my async closer should have gotten this right... even if closes are
submitted from multiple threads. grrr..... i guess will have to revert.
--
------------- Codito, ergo sum - "I code, therefore I am" --------------
Carsten Haitzler - [email protected]
_______________________________________________
enlightenment-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/enlightenment-users
