Age | Commit message (Collapse) | Author | Files | Lines |
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Copied from libxcursor@f807ac9c786714ef4e86ad7edfa60f92baf0b4a6
Signed-off-by: Thomas E. Dickey <dickey@invisible-island.net>
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Copied from libxcursor@897213f36baf6926daf6d192c709cf627aa5fd05
Reported-by: @ithinkapps in libxcb-cursor issue #11
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Mihail Konev <k.mvc@ya.ru>
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Place quotes around the $srcdir, $ORIGDIR and $0 variables to prevent
fall-outs, when they contain space.
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Syncs the invocation of configure with the one from the server.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
See http://people.gnome.org/~walters/docs/build-api.txt
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
configure.ac:37: warning: AC_OUTPUT should be used without arguments.
configure.ac:37: You should run autoupdate.
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
When building something out-of-source, make's VPATH mechanism is used so that
files in the source directory are also found. This makes make automatically
check that other directory for needed files and use those when needed.
However, make doesn't magically copy those files over. We have to explicitly use
the file that make found. Do so by using $< for refering to shape_to_id.gperf
instead of hardcoding the name of the source file.
This fixes out-of-source git builds.
Out-of-source release builds already worked before this, because releases come
with a pre-generated shape_to_id.c.
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=80153
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
The documentation for xcb_cursor_load_cursor() say that the resulting
cursor must be freed, but the example in the documentation does not do
so. It does not matter much for this example, but I think it is a good
idea to use it to hint to this requirement. Also, the example already
uses xcb_cursor_context_free(), so it tries to clean up after itself.
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
|
|
This works (hopefully) the same way as the same option does for libXcursor.
Signed-off-by: Uli Schlachter <uli.schlachter@informatik.uni-oldenburg.de>
|
|
|
|
The passed in cursor name to xcb_cursor_load_cursor might not match
one of the predefined font cursor values. Without the check the
call to create glyph cursor will fail with a BadValue error, but
the library returns the id allocated for the xcb_cursor_t. A user
of the library gets a value which looks like a valid cursor, but
when using it for e.g. a cursor value in xcb_create_window it raises
a BadCursor error.
|
|
Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
|
|
Exit early with an informative message if the submodules are missing,
since they are needed. Without this autoconf throws a bunch of
uninformative errors which does not point to the actual problem.
This was taken from util-keysyms.
Signed-off-by: Niclas Zeising <zeising@daemonic.se>
Reviewed-By: Arnaud Fontaine <arnau@debian.org>
|
|
le32toh.
|
|
|
|
This fixes https://bugs.freedesktop.org/show_bug.cgi?id=71060
|
|
|
|
The rule that cursor/Makefile.am uses for calling gperf can be found on the
internet[0]. That page explains three cases that should be handled:
1. gperf succeeded
2. gperf failed
3. gperf is missing
The reasoning for the third case is:
If $(GPERF) does not answer to --version, it is certainly missing, and missing
already suggested to install Gperf. Then remove the temporary output file, and
let the compilation proceed by updating the timestamp of the output file.
That's a best effort, essentially helping users who get the project with
broken timestamps.
However, this assumes that the user is building from a tarball which already
contains a working version of the output C file. When building from git, this
file does not exist and instead of updating the timestamp of the output file,
this Makefile rule would create the output file.
Thus, the following four cases need to be handled:
1. gperf succeeded
2. gperf failed
3. gperf is missing and the output file already exists
4. gperf is missing and the output file is missing, too
For the third case, the above reasoning applies. However, in the fourth case,
the build would continue and produce a broken library. For xcb-util-cursor this
means that linking to this library would fail with "undefined reference to
`cursor_shape_to_id'". So in this case the build should fail.
[0]: http://www.lrde.epita.fr/~akim/ccmp/doc/gnuprog2/Using-Gperf-with-the-GNU-Build-System.html
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
This adds support for automake's silent mode to the gperf rule. In silent mode,
make will not print the whole command lines, but only a short version of what it
does.
For this rule, "GEN shape_to_id.c" will be printed.
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com>
|
|
RENDER's CreateCursor request was added in version 0.5 and CreateAnimCursor was
added in version 0.8. Hence, just having the RENDER extension is not enough and
we need to check the exact version that the server provides.
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
If someone e.g. defines Xcursor.theme twice in the resource information, this
would cause c->rm[RM_XCURSOR_THEME] to be set via strdup() twice which means
that the first pointer is leaked.
Fix this by freeing the old value in this case.
Signed-off-by: Uli Schlachter <psychon@znc.in>
|
|
|
|
In early development, I hardcoded these to make debugging easier, but
then forgot to use the actual variables :).
|
|
|
|
|
|
FreeBSD have the le32toh macro in <sys/endian.h>, and this file is not
indirectly included by any other header in util-cursor. Add a configure
check that checks for <endian.h> (Linux) and <sys/endian.h> (FreeBSD),
and include the right file based on this check.
Signed-off-by: Niclas Zeising <zeising@daemonic.se>
|
|
|
|
This fixes a compilation problem on NetBSD
|
|
|
|
|
|
|
|
|
|
|
|
…and therefore does not need to be freed in xcb_cursor_context_free()
|
|
|
|
|
|
This makes “make distcheck” work
|
|
|