Not 100% improvement, there are still a variety of failures with
GLXWindows, but getting there. At least it fixes more things than its
breaks...
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Complete the task started in cd4b2d843b by
remembering to compute the insertion hash on the original font face as
well as the lookup hashes. Also take advantage by deferring resolution
of the implementation for the font face until after we fail to find an
already constructed scaled font.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Ordinary patterns have their device offsets pre-baked into the pattern
matrix. The difference with raster patterns is that the surface is
generated upon use by the pdf,ps and so its device offset is not known
until then.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
I need to remember that the translation for the pattern matrix is the
inverse of the translation for drawing.
Fixes: subsurface-modify-parent
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Rather than store a command to copy from a reference of another recording
surface onto this, just replay the other recording surface.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Borrow the technique for installing a proxy-snapshot developed for
acquiring the image from a recording surface.
This prevents the cairo from exploding and using up all memory due to
infinite recursion, but it does break a few tests that perform
self-copies with differing transformations.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
I forgot that creating any image by defaults results in pixman
calloc'ing the pixel data and so cairo helpfully sets the is-clear flag
on the surface.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Rather than compress the copies into a clip + copy, iterate over and
perform each copy separately so as to avoid the confusion for
window-to-window copies and the solitary GC->pCompositeClip.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
As _cairo_surface_get_source() returns the limits of the source, a chunk
of code to query the source extents became redundant.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
In the midst of trying to simply export the
_cairo_image_surface_get_font_options() for subclassing, I accidentally
decoupled it from the image surface backend. Mea culpa.
Fixes regression from 8bea52bb (Add preliminary damage tracking), which
should not have even touched the backend structure except for an ugly
rebasing.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
and make it more obvious that we're not doing OOB accesses.
They were not possible because of the parity constraint, but by
guarding the highest index with which we access the data in the loop,
the correctness information is completely contained in the loop code.
Fixes:
cairo-stroke-style.c:199:2: warning: cannot optimize loop, the loop
counter may overflow [-Wunsafe-loop-optimizations]
_cairo_path_fixed_last_op() contains an assertion, which gcc doesn't
like to inline. Since it is a static function, which basically
accesses a value, gcc will inline it anyway when assertions are
disabled, so remove the "inline" hint to reduce gcc warning noise when
doing debug builds.
Fixes:
cairo-path-fixed.c: In function '_cairo_path_fixed_drop_line_to':
cairo-path-fixed.c:373:1: warning: inlining failed in call to
'_cairo_path_fixed_last_op.isra.5.part.6': call is unlikely and code
size would grow [-Winline]
cairo-path-fixed.c:400:1: warning: called from here [-Winline]
...
As we return the child image to the user and so perform the reference
tracking on it and not the parent win32 display surface, we need to add
a call to destroy the parent from the image surface. This of course
complicates the normal scenario of destroying the parent first, and so
in that case we need to unhook the image->parent before freeing the
surface->image.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
I suspect I may split the win32 code into a few more files, so move it
to its own directory to reduce the clutter.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
This is initially based around the requirements for handling internal
fallbacks to the image compositor and reducing the number of pixels
required to be transferred.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
The pattern creation function call in the example code has one _mesh too
much. This hopefully fixes the mesh mismatch mess :)
Signed-off-by: Nis Martensen <nis.martensen@web.de>
Signed-off-by: Uli Schlachter <psychon@znc.in>
Update to latest surface backend structure, removed obsolete functions
and fixed functions API in order to be compatible with backend definition.
Fixed compilation with Qt5
GCC complains that:
cairo-surface-observer.c:289:26: warning: ignoring return value of
'cairo_device_acquire', declared with attribute warn_unused_result
[-Wunused-result]
Explain why it's safe to ignore it in a comment.
Making cairo_surface_observer_print() and
cairo_device_observer_print() return the status of the observer or of
the stream makes it possible to correctly track what kind of error
happens if the print is not successful.
This makes the functions more consistent with existing API with a
similar signature like cairo_surface_write_to_png_stream().
If a bounded recording surface doesn't have its extents .x and .y at the
surface's origin, this code was translating all clips to make up for that.
However, the clip already is in device space.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41583
Signed-off-by: Uli Schlachter <psychon@znc.in>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
