platform/upstream/efl.git
12 years agoadd note about CORK availability
Mike Blumenkrantz [Sat, 10 Dec 2011 08:23:34 +0000 (08:23 +0000)]
add note about CORK availability

SVN revision: 66083

12 years agowe've got NODELAY, may as well add CORK
Mike Blumenkrantz [Sat, 10 Dec 2011 08:14:50 +0000 (08:14 +0000)]
we've got NODELAY, may as well add CORK

SVN revision: 66082

12 years agofix cert flag detection
Mike Blumenkrantz [Sat, 10 Dec 2011 08:09:17 +0000 (08:09 +0000)]
fix cert flag detection

SVN revision: 66081

12 years agosvn 1.7 uses a stupid new error string so now we're gonna have to update all of these
Mike Blumenkrantz [Sat, 10 Dec 2011 07:34:48 +0000 (07:34 +0000)]
svn 1.7 uses a stupid new error string so now we're gonna have to update all of these

SVN revision: 66080

12 years agogitignore
Mike Blumenkrantz [Sat, 10 Dec 2011 07:34:47 +0000 (07:34 +0000)]
gitignore

SVN revision: 66079

12 years agorevert 66063 + related commits: my dreams of zero-copy ecore-con transfers were shatt...
Mike Blumenkrantz [Sat, 10 Dec 2011 07:34:45 +0000 (07:34 +0000)]
revert 66063 + related commits: my dreams of zero-copy ecore-con transfers were shattered by the inconsistency of linux socket operations. way to go kernel developers.

SVN revision: 66078

12 years agofix annoying spankies
Mike Blumenkrantz [Sat, 10 Dec 2011 07:21:53 +0000 (07:21 +0000)]
fix annoying spankies

SVN revision: 66077

12 years agoadd docs for NODELAY flag
Mike Blumenkrantz [Sat, 10 Dec 2011 06:53:54 +0000 (06:53 +0000)]
add docs for NODELAY flag

SVN revision: 66071

12 years agotemp fix while I figure out some weird behavior
Mike Blumenkrantz [Sat, 10 Dec 2011 06:15:00 +0000 (06:15 +0000)]
temp fix while I figure out some weird behavior

SVN revision: 66069

12 years agorewrote server+client deletion code to (hopefully) be failproof.
Mike Blumenkrantz [Sat, 10 Dec 2011 05:42:32 +0000 (05:42 +0000)]
rewrote server+client deletion code to (hopefully) be failproof.
I've also set it to abort() with a big error if it gets to the old fallthrough spot, so make sure to send me a backtrace if this is reached (though it should be impossible)

SVN revision: 66068

12 years agomore error event unification
Mike Blumenkrantz [Sat, 10 Dec 2011 05:25:30 +0000 (05:25 +0000)]
more error event unification

SVN revision: 66067

12 years agosome small cleanups to unify more error code
Mike Blumenkrantz [Sat, 10 Dec 2011 05:17:34 +0000 (05:17 +0000)]
some small cleanups to unify more error code

SVN revision: 66066

12 years agoadd windows error events for new code, condense some connect handlers
Mike Blumenkrantz [Sat, 10 Dec 2011 05:09:47 +0000 (05:09 +0000)]
add windows error events for new code, condense some connect handlers

SVN revision: 66065

12 years agooptimize read mechanics for remote servers using ioctl to determine number of bytes...
Mike Blumenkrantz [Fri, 9 Dec 2011 22:35:12 +0000 (22:35 +0000)]
optimize read mechanics for remote servers using ioctl to determine number of bytes available for reading.
in general, this should provide an improvement which scales with the amount of data being transferred:
* small transfers will incur a small amount of overhead from potentially unneeded memory as I try to account for a bug in FIONREAD which returns a number that is smaller than the actual number of bytes available for read on a socket
* large transfers will no longer require any copies of the data

on systems which do not provide the FIONREAD ioctl(), old functionality will be used
this should work on windows, though I (obviously) can't test it myself

thus ends the longest commit message I have ever written

SVN revision: 66063

12 years agoadd EAPI. Maybe it is missed.
Jaehwan Kim [Fri, 9 Dec 2011 07:55:31 +0000 (07:55 +0000)]
add EAPI. Maybe it is missed.

SVN revision: 66061

12 years agookay I was a little too excited
Mike Blumenkrantz [Thu, 8 Dec 2011 17:27:30 +0000 (17:27 +0000)]
okay I was a little too excited

SVN revision: 66042

12 years agoedje: we never call perror.
Cedric BAIL [Thu, 8 Dec 2011 16:21:49 +0000 (16:21 +0000)]
edje: we never call perror.

SVN revision: 66041

12 years agoecore: we never call perro.
Cedric BAIL [Thu, 8 Dec 2011 16:21:29 +0000 (16:21 +0000)]
ecore: we never call perro.

SVN revision: 66040

12 years agoecore: fix header inclusion.
Cedric BAIL [Thu, 8 Dec 2011 16:19:44 +0000 (16:19 +0000)]
ecore: fix header inclusion.

SVN revision: 66039

12 years agomore binbuf_manage_new!
Mike Blumenkrantz [Thu, 8 Dec 2011 16:08:33 +0000 (16:08 +0000)]
more binbuf_manage_new!

SVN revision: 66038

12 years agouse binbuf_manage_new
Mike Blumenkrantz [Thu, 8 Dec 2011 15:58:01 +0000 (15:58 +0000)]
use binbuf_manage_new

SVN revision: 66037

12 years agoEvas textblock: Added evas_textblock_markup_to_plain.
Tom Hacohen [Thu, 8 Dec 2011 15:12:25 +0000 (15:12 +0000)]
Evas textblock: Added evas_textblock_markup_to_plain.

This function converts a textblock markup to plain text.
It converts for example <br/> to \n and a lot more.

SVN revision: 66034

12 years agoEvas: Updated NEWS file.
Tom Hacohen [Thu, 8 Dec 2011 13:45:09 +0000 (13:45 +0000)]
Evas: Updated NEWS file.

SVN revision: 66032

12 years agoEvas: Updated changelog to include latest textblock changes.
Tom Hacohen [Thu, 8 Dec 2011 13:39:31 +0000 (13:39 +0000)]
Evas: Updated changelog to include latest textblock changes.

SVN revision: 66031

12 years agoevas: use the right message command when usefull.
Cedric BAIL [Thu, 8 Dec 2011 13:29:02 +0000 (13:29 +0000)]
evas: use the right message command when usefull.

SVN revision: 66030

12 years agoeina: use WRN for warning.
Cedric BAIL [Thu, 8 Dec 2011 13:28:32 +0000 (13:28 +0000)]
eina: use WRN for warning.

SVN revision: 66029

12 years agoEina bin/(u)strbuf: Added eina_*buf_manage_new_length.
Tom Hacohen [Thu, 8 Dec 2011 13:10:57 +0000 (13:10 +0000)]
Eina bin/(u)strbuf: Added eina_*buf_manage_new_length.

Same as eina_*buf_manage_new except for the length parameter.

SVN revision: 66028

12 years agoeina: make it possible to disable Eina_Log infrastructure.
Cedric BAIL [Thu, 8 Dec 2011 12:08:53 +0000 (12:08 +0000)]
eina: make it possible to disable Eina_Log infrastructure.

SVN revision: 66025

12 years agoEdje entry: use the newly added self-closing textblock tags.
Tom Hacohen [Thu, 8 Dec 2011 12:07:05 +0000 (12:07 +0000)]
Edje entry: use the newly added self-closing textblock tags.

SVN revision: 66024

12 years agoEvas textblock: Support self-closing formats, i.e <br/>.
Tom Hacohen [Thu, 8 Dec 2011 12:05:56 +0000 (12:05 +0000)]
Evas textblock: Support self-closing formats, i.e <br/>.

Also updated tests to follow this. Using <br>, although will work, is
discouraged, please use <br/> instead.

SVN revision: 66023

12 years agoEvas textblock: Improve internal handling of formats.
Tom Hacohen [Thu, 8 Dec 2011 12:05:47 +0000 (12:05 +0000)]
Evas textblock: Improve internal handling of formats.

SVN revision: 66022

12 years agounify and improve gnutls debug output
Mike Blumenkrantz [Thu, 8 Dec 2011 05:58:47 +0000 (05:58 +0000)]
unify and improve gnutls debug output

SVN revision: 66019

12 years agonews updates
Mike Blumenkrantz [Thu, 8 Dec 2011 05:10:44 +0000 (05:10 +0000)]
news updates

SVN revision: 66018

12 years agoalign debug info
Mike Blumenkrantz [Thu, 8 Dec 2011 05:02:49 +0000 (05:02 +0000)]
align debug info

SVN revision: 66017

12 years agofix some format string errors
Mike Blumenkrantz [Thu, 8 Dec 2011 04:21:25 +0000 (04:21 +0000)]
fix some format string errors

SVN revision: 66016

12 years agofix shadow var warning
Mike Blumenkrantz [Thu, 8 Dec 2011 04:18:12 +0000 (04:18 +0000)]
fix shadow var warning

SVN revision: 66014

12 years agofix eina_log*level_check() to return the value that the docs claim will be returned
Mike Blumenkrantz [Thu, 8 Dec 2011 04:11:54 +0000 (04:11 +0000)]
fix eina_log*level_check() to return the value that the docs claim will be returned
how did nobody catch this sooner?

SVN revision: 66013

12 years agomore informative error messages when cert verification fails
Mike Blumenkrantz [Thu, 8 Dec 2011 03:56:29 +0000 (03:56 +0000)]
more informative error messages when cert verification fails

SVN revision: 66012

12 years agofix type so openssl verification works again
Mike Blumenkrantz [Thu, 8 Dec 2011 03:53:49 +0000 (03:53 +0000)]
fix type so openssl verification works again

SVN revision: 66011

12 years agoca loading function now loads directories
Mike Blumenkrantz [Thu, 8 Dec 2011 03:25:53 +0000 (03:25 +0000)]
ca loading function now loads directories

SVN revision: 66009

12 years agounify openssl debug printing and remove #ifdefs: now this will always print with...
Mike Blumenkrantz [Thu, 8 Dec 2011 02:59:27 +0000 (02:59 +0000)]
unify openssl debug printing and remove #ifdefs: now this will always print with ecore-con:4

SVN revision: 66008

12 years agofix compat on openssl errors
Mike Blumenkrantz [Thu, 8 Dec 2011 02:53:43 +0000 (02:53 +0000)]
fix compat on openssl errors

SVN revision: 66007

12 years agobeef up openssl debug output
Mike Blumenkrantz [Thu, 8 Dec 2011 02:37:00 +0000 (02:37 +0000)]
beef up openssl debug output

SVN revision: 66006

12 years agobetter error messages on certificate verify failure
Mike Blumenkrantz [Thu, 8 Dec 2011 02:11:22 +0000 (02:11 +0000)]
better error messages on certificate verify failure

SVN revision: 66005

12 years agowhoops, fix compile on unrelated change that snuck in
Mike Blumenkrantz [Thu, 8 Dec 2011 01:30:53 +0000 (01:30 +0000)]
whoops, fix compile on unrelated change that snuck in
changelog tbd

SVN revision: 66004

12 years agodamn you ctrl+d!
Mike Blumenkrantz [Thu, 8 Dec 2011 01:23:10 +0000 (01:23 +0000)]
damn you ctrl+d!

SVN revision: 66003

12 years ago+ecore_con_ssl_server_verify_name_set/get
Mike Blumenkrantz [Thu, 8 Dec 2011 01:14:55 +0000 (01:14 +0000)]
+ecore_con_ssl_server_verify_name_set/get

SVN revision: 66002

12 years agoallow certificates to be loaded for STARTTLS
Mike Blumenkrantz [Wed, 7 Dec 2011 23:36:15 +0000 (23:36 +0000)]
allow certificates to be loaded for STARTTLS

SVN revision: 66000

12 years agoupdate docs for eina_file_dir_list
Mike Blumenkrantz [Wed, 7 Dec 2011 22:50:50 +0000 (22:50 +0000)]
update docs for eina_file_dir_list

SVN revision: 65997

12 years agoEina: missing 'static inline'
Vincent Torri [Wed, 7 Dec 2011 20:28:56 +0000 (20:28 +0000)]
Eina: missing 'static inline'

SVN revision: 65992

12 years agoEina: improve a bit the doc
Vincent Torri [Wed, 7 Dec 2011 12:07:18 +0000 (12:07 +0000)]
Eina: improve a bit the doc

SVN revision: 65987

12 years agoNO! you break api. you made my e sit here with a segv in a getenv.
Carsten Haitzler [Wed, 7 Dec 2011 01:10:43 +0000 (01:10 +0000)]
NO! you break api. you made my e sit here with a segv in a getenv.
because now many libraries and api's don't have prototyopes for
malloc/calloc and much more and this goes horribly wrong especially on
64bit! the eina headers have provided these includes historically and
removing them is a BREAK in api. apps that used to compile and run
just fine now don't. it's unacceptable to break api.

i'm stuck here in unity for crying out loud! this deservves a big FAT
REVERT for that! :-P

SVN revision: 65983

12 years agolet's not fail to include standard headers when we need things like..
Carsten Haitzler [Wed, 7 Dec 2011 00:49:00 +0000 (00:49 +0000)]
let's not fail to include standard headers when we need things like..
ooh.. malloc? calloc? printf?....

SVN revision: 65982

12 years agodisable sse3 if immintrin.h not found.
Carsten Haitzler [Wed, 7 Dec 2011 00:10:36 +0000 (00:10 +0000)]
disable sse3 if immintrin.h not found.

SVN revision: 65980

12 years agoEcore: fix warnings
Vincent Torri [Tue, 6 Dec 2011 22:33:01 +0000 (22:33 +0000)]
Ecore: fix warnings

SVN revision: 65976

12 years agoeina: the new endless story of fixing header in all possible configuration.
Cedric BAIL [Tue, 6 Dec 2011 21:33:39 +0000 (21:33 +0000)]
eina: the new endless story of fixing header in all possible configuration.

SVN revision: 65974

12 years agoeina: last header fix.
Cedric BAIL [Tue, 6 Dec 2011 17:05:14 +0000 (17:05 +0000)]
eina: last header fix.

SVN revision: 65970

12 years agoeina: fix system header usage.
Cedric BAIL [Tue, 6 Dec 2011 16:57:28 +0000 (16:57 +0000)]
eina: fix system header usage.

SVN revision: 65969

12 years agoeina: move clist around as needed.
Cedric BAIL [Tue, 6 Dec 2011 16:55:50 +0000 (16:55 +0000)]
eina: move clist around as needed.

SVN revision: 65968

12 years agoeina: only include system header as necessary.
Cedric BAIL [Tue, 6 Dec 2011 16:09:56 +0000 (16:09 +0000)]
eina: only include system header as necessary.

SVN revision: 65966

12 years agoecore: include stdlib.h when required.
Cedric BAIL [Tue, 6 Dec 2011 16:09:05 +0000 (16:09 +0000)]
ecore: include stdlib.h when required.

SVN revision: 65964

12 years agoevas: less unused warning.
Cedric BAIL [Tue, 6 Dec 2011 16:08:30 +0000 (16:08 +0000)]
evas: less unused warning.

SVN revision: 65963

12 years agoevas: include stdlib.h as required.
Cedric BAIL [Tue, 6 Dec 2011 16:07:59 +0000 (16:07 +0000)]
evas: include stdlib.h as required.

SVN revision: 65962

12 years agoemotion: turn fps debugging on also on classic Gstreamer.
Cedric BAIL [Tue, 6 Dec 2011 11:36:29 +0000 (11:36 +0000)]
emotion: turn fps debugging on also on classic Gstreamer.

SVN revision: 65955

12 years agoeet: no reason to duplicate environment variable.
Cedric BAIL [Tue, 6 Dec 2011 11:00:50 +0000 (11:00 +0000)]
eet: no reason to duplicate environment variable.

SVN revision: 65954

12 years agoEthumb: source files should have their names with lower case letters
Vincent Torri [Tue, 6 Dec 2011 06:50:10 +0000 (06:50 +0000)]
Ethumb: source files should have their names with lower case letters

SVN revision: 65948

12 years agoEthumb: formatting, fix future compilation on Windows
Vincent Torri [Tue, 6 Dec 2011 06:31:45 +0000 (06:31 +0000)]
Ethumb: formatting, fix future compilation on Windows

SVN revision: 65946

12 years agoEthumb: fix future compilation on Windows, formatting
Vincent Torri [Tue, 6 Dec 2011 06:12:15 +0000 (06:12 +0000)]
Ethumb: fix future compilation on Windows, formatting

SVN revision: 65945

12 years agoone of these days I'm going to snap and throw my computer off my balcony because...
Mike Blumenkrantz [Tue, 6 Dec 2011 05:52:45 +0000 (05:52 +0000)]
one of these days I'm going to snap and throw my computer off my balcony because of this

SVN revision: 65944

12 years ago@since
Mike Blumenkrantz [Tue, 6 Dec 2011 05:37:29 +0000 (05:37 +0000)]
@since

SVN revision: 65943

12 years agoEthumb: remove and move header file
Vincent Torri [Tue, 6 Dec 2011 05:32:35 +0000 (05:32 +0000)]
Ethumb: remove and move header file

SVN revision: 65942

12 years agoEthumb: remove unused macro
Vincent Torri [Tue, 6 Dec 2011 05:27:13 +0000 (05:27 +0000)]
Ethumb: remove unused macro

SVN revision: 65941

12 years agoEthumb: fix future compilation on Windows
Vincent Torri [Tue, 6 Dec 2011 05:21:30 +0000 (05:21 +0000)]
Ethumb: fix future compilation on Windows

SVN revision: 65940

12 years ago@since
Mike Blumenkrantz [Tue, 6 Dec 2011 05:20:13 +0000 (05:20 +0000)]
@since

SVN revision: 65939

12 years agolast ssl bugger
Mike Blumenkrantz [Tue, 6 Dec 2011 05:14:03 +0000 (05:14 +0000)]
last ssl bugger

SVN revision: 65938

12 years agofix some ssl socks buggies
Mike Blumenkrantz [Tue, 6 Dec 2011 04:48:55 +0000 (04:48 +0000)]
fix some ssl socks buggies

SVN revision: 65937

12 years ago...make that twice
Mike Blumenkrantz [Tue, 6 Dec 2011 03:51:01 +0000 (03:51 +0000)]
...make that twice

SVN revision: 65936

12 years agojust kidding before, had to break svn at least once today
Mike Blumenkrantz [Tue, 6 Dec 2011 03:41:16 +0000 (03:41 +0000)]
just kidding before, had to break svn at least once today

SVN revision: 65935

12 years agointroducinggggggggggggggg <drum roll>:
Mike Blumenkrantz [Tue, 6 Dec 2011 03:32:16 +0000 (03:32 +0000)]
introducinggggggggggggggg <drum roll>:
ECORE-CON-SOCKS! SOCKS ON!!!!

now ecore_con supports socks (v4 and v4a only, so no ipv6) connections natively for making remote connections
for those of you who want their apps to start proxying immediately, just update and export this handy environment variable:
ECORE_CON_SOCKS_V4=[user@]PROXY_IP_ADDRESS:PROXY_PORT[:1] <--use :1 here to enable dns lookups on the proxy

SVN revision: 65934

12 years agoEvas (gl engine): Remove incorrect __UNUSED__ on paramaters that are
Christopher Michael [Tue, 6 Dec 2011 00:44:25 +0000 (00:44 +0000)]
Evas (gl engine): Remove incorrect __UNUSED__ on paramaters that are
actually used.

SVN revision: 65928

12 years agoeina: fix typo.
Cedric BAIL [Mon, 5 Dec 2011 21:19:19 +0000 (21:19 +0000)]
eina: fix typo.

SVN revision: 65921

12 years agoevas: don't randomly move code around !
Cedric BAIL [Mon, 5 Dec 2011 16:50:40 +0000 (16:50 +0000)]
evas: don't randomly move code around !

SVN revision: 65916

12 years agosame for these
Mike Blumenkrantz [Mon, 5 Dec 2011 15:39:04 +0000 (15:39 +0000)]
same for these

SVN revision: 65913

12 years agoalways have the file which includes config.h first!
Mike Blumenkrantz [Mon, 5 Dec 2011 15:38:55 +0000 (15:38 +0000)]
always have the file which includes config.h first!
SPANK SPANK SPANK cedric!

SVN revision: 65912

12 years agoevas: use already existing macro.
Cedric BAIL [Mon, 5 Dec 2011 14:32:10 +0000 (14:32 +0000)]
evas: use already existing macro.

SVN revision: 65906

12 years agoeina: follow Vincent advice.
Cedric BAIL [Mon, 5 Dec 2011 14:25:25 +0000 (14:25 +0000)]
eina: follow Vincent advice.

SVN revision: 65905

12 years agoevas: use eina_lock a little bit more.
Cedric BAIL [Mon, 5 Dec 2011 14:01:42 +0000 (14:01 +0000)]
evas: use eina_lock a little bit more.

SVN revision: 65904

12 years agoevas: more fine grained system detection.
Cedric BAIL [Mon, 5 Dec 2011 14:00:53 +0000 (14:00 +0000)]
evas: more fine grained system detection.

SVN revision: 65903

12 years agoevas: cast as we did in eina.
Cedric BAIL [Mon, 5 Dec 2011 13:59:44 +0000 (13:59 +0000)]
evas: cast as we did in eina.

SVN revision: 65902

12 years agoeina: disable log if fnmatch is not present.
Cedric BAIL [Mon, 5 Dec 2011 13:58:09 +0000 (13:58 +0000)]
eina: disable log if fnmatch is not present.

NOTE: would be better to have an explicit --disable option I think.

SVN revision: 65901

12 years agoeina: improve system header detection.
Cedric BAIL [Mon, 5 Dec 2011 13:57:10 +0000 (13:57 +0000)]
eina: improve system header detection.

SVN revision: 65900

12 years agoFrom: Joerg Sonnenberger <joerg@britannica.bec.de>
Joerg Sonnenberger [Mon, 5 Dec 2011 10:17:57 +0000 (10:17 +0000)]
From: Joerg Sonnenberger <joerg@britannica.bec.de>
Subject: [E-devel] eina 1.1.0beta patches for/from pkgsrc

Several patches to fix warnings and potential on netbsd/solaris. minor
so i think i'll skip changelog for this.

SVN revision: 65898

12 years agochlog++
Carsten Haitzler [Mon, 5 Dec 2011 09:47:49 +0000 (09:47 +0000)]
chlog++

SVN revision: 65897

12 years agoremove warnings!
Carsten Haitzler [Mon, 5 Dec 2011 09:26:04 +0000 (09:26 +0000)]
remove warnings!

SVN revision: 65896

12 years agoremove warning - bug ready to happen with fastpath. didnt pass
Carsten Haitzler [Mon, 5 Dec 2011 09:18:05 +0000 (09:18 +0000)]
remove warning - bug ready to happen with fastpath. didnt pass
sharedlist or direct down.

SVN revision: 65895

12 years agofix gles/egl bugs/issues in sungwoo's new patch
Carsten Haitzler [Mon, 5 Dec 2011 09:05:04 +0000 (09:05 +0000)]
fix gles/egl bugs/issues in sungwoo's new patch

SVN revision: 65894

12 years agoI'm attaching a patch for the initial version of the GL Fastpath
Carsten Haitzler [Mon, 5 Dec 2011 08:54:14 +0000 (08:54 +0000)]
I'm attaching a patch for the initial version of the GL Fastpath
addition to evas gl backend.

Working on different mobile devices, we've noticed that the cost of context
switch (MakeCurrent) in GL can be very expensive depending on the driver
implementation.  To minimize the poorly written driver's context switch
overhead, I've implemented a state tracking layer on top of the driver
implemented GL.

Essentially, this layer wraps all the GL/Glue(GLX/EGL) APIs and manages its own
state changes.  Internally, only one real GL context is created and logical
contexts are created every time a user requests context creation.  The logical
contexts keep track of its states and sets only the necessary states
(the ones that are different than the current ones)
when there is a MakeCurrent request.  The real MakeCurrent gets called when
there is a Surface/Window change request.

The GL library is dlopened and all the APIs are dlsym'ed and wrapped
accordingly.  All the GL functions are in evas_gl_core.{h,c}.

Here's a very simply flow of the code.
   - all the APIs are exported as function pointers (*glsym_glBegin),
     (*glsm_eglCreatContext), and etc.
   - all the native GL/Glue(GLX/EGL) APIs are dlsym'ed as _sym_glBegin,
     _sym_eglCreateContext, and etc.
   - all the fastpath APIs are implmemnted as fpgl_glBegin,
     fpgl_eglCreateContext, and etc.
   - if faspath is seletected, the exported APIs are set accordingly
     ie. glsym_glBegin = fpgl_glBegin;
   - default mode is the regular gl symbols are directly set.
     ie. glsym_glBegin = _sym_glBegin;

I have an Environment variable where you can set it to three different Modes

EVAS_GL_FASTPATH = 0    // Default mode. Regular GL symbols are directly loaded
EVAS_GL_FASTPATH = 1    // Fastpath mode. Takes the path described above.
EVAS_GL_FASTPATH = 2    // Wrapped mode.  All the regular GL functions are
                           wrapped once.  This can be used for various
purposes

Since all the GL symbols are now loaded in this library, I took out all
the gl symbol loading parts in the evas gl backend as you'll see in the patch.
The changes to the engine and the backend itself is pretty minor.

There are still some known issues to hammer out but I thought we're at a good
place for an initial version so that my source doesn't diverge too much.

Known Issues and To Do's
* Current GL Fastpath version doesn't support multiple threads. Instead of
  having one global real context, I would need to do it for each thread. I'll
  get on this soon.

  * Issues running Evas GL on certain conditions.  When running the elementary
  test (with gl engine), if you run ELMGLview test that runs in ON_DEMAND mode,
  everything works fine. BUT, when you run the ELMGLView test in ALWAYS
    mode, the subsequent elm tests shows blank screen. When you destroy the
  GLView window, everything else comes on fine.

  * Resource protection code.  This actually applies to Evas GL code in general
  as well. Since all the resources are shared among all the contexts that get
  created, I would like to eventually have a resource protecting mechanism that
  prevents access to resources outside of its context unless specifically
  specified.

  I'm attaching three files
     - evas_gl_core.h, evas_gl_core.c, fastpath.patch

  To get the code running...
     - copy evas_gl_core.{c,h} to src/modules/engine/gl_common/
     - apply the fastpath.patch
     - compile/install evas
     - to run with fastpath GL (ie. % EVAS_GL_FASTPATH=1 ./evasgl_sample1)

SVN revision: 65891

12 years agogiant patch refactoring ecore-con-url from bluezery <opowel@gmail.com>
Mike Blumenkrantz [Mon, 5 Dec 2011 06:08:46 +0000 (06:08 +0000)]
giant patch refactoring ecore-con-url from bluezery <opowel@gmail.com>
will it blend?!

SVN revision: 65888

12 years agototally meant to write this wrong the first time to see if anyone was watching.
Mike Blumenkrantz [Sun, 4 Dec 2011 23:54:41 +0000 (23:54 +0000)]
totally meant to write this wrong the first time to see if anyone was watching.

SVN revision: 65878

12 years agoanother failure attempt at pleasing Demon Lord Vincent Torri
Mike Blumenkrantz [Sun, 4 Dec 2011 23:16:25 +0000 (23:16 +0000)]
another failure attempt at pleasing Demon Lord Vincent Torri

SVN revision: 65877