os_readdir might return an empty path if the OS filename doesn't fit
into the buffer after os_wcs_to_utf8. If this empty path is then passed
to os_get_path_extension it causes a crash.
Previously, any short write would be treated elsewhere as a failure. The
easiest solution is to have pipe_write() automatically retry until either
it's written everything, or it couldn't write anything at all.
Complex external systems using the D3D11 device may need to perform
their own device loss handling, the upcoming Windows Graphics Capture
support for example.
A plugin that saves whitespace-only data would crash OBS on startup as
the whitespace stripping would send the string length negative. While
that bug is fixed in this commit, there is also no good reason that OBS
is stripping whitespace to begin with. All data going into the configs
should be well formed and some plugins legitimately wish to save
whitespace as a config option (eg for search / replace characters).
(This commit also modifies libobs)
Hides encoders that are marked as internal. Same general functionality
as DEPRECATED but not actually deprecated, just internal-only.
Returns whether rescaling is enabled for an encoder. This will be used
with texture-based encoders to determine whether to fall back to
RAM-based encoding instead.
`obs_sceneitem_group_add_item` and `obs_Sceneitem_group_remove_item`
were leaving the affected scene in an inconsistent state.
To resolve, simplify and make the code more readable, we replace the
manual linked lists mutation with existing `detach_sceneitem` and
`attach_sceneitem` calls.
Adds API:
obs_scene_add_group2
obs_scene_insert_group2
obs_sceneitem_group_ungroup2
These functions should be used by plugins if they need to use these
functions and need to send a refresh signal. If a major API rework ever
happens the old functions should be removed. The old functions should
eventually be deprecated.
The reason why specifying a 'signal' parameter is useful is because it's
a bit more seamless for the user interface to be able to have custom
handling of these specific cases. It looks better and doesn't require
completely erasing/recreating the entire list, which is visually
unappealing.
Previously the assertion required the idx to be smaller than the number
of elements in the darray. This would mean you could not insert anything
at the end of a darray, and would make it impossible to insert an array
into an empty darray.
SIMDE was introduced for aarch64 support, however, the library itself
supports non-SIMD fallback, which allows us provide support to other
platforms without code changes.
There is another world beyond x86. So we can simply enable SIMDE for
processors without SSE2 support.
Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
This allows the ability to do things such as change defaults or
properties of sources without necessarily breaking older user
configurations that had older defaults or properties.
This uses three methods of obtaining the actual windows version,
RtlGetVersion which is the ntdll version of GetVersionEx that bypasses
the manifest check garbage, looking up the file version of a file that
is most likely to be updated per windows version (ntoskrnl), and the
registry. Of the three values, it chooses the highest windows version
obtained by the three.
Closesobsproject/obs-studio#2294
Allows the ability for manual transitioning to smoothly flow
(interpolate) to the intended transition point over a short period of
time rather than simply setting a hard transition point number. Doing
this allows manual transitioning to occur more smoothly, and in a more
visually pleasant way.
Allows the ability to manually specify the transitioning point so the
user can transition at a custom rate, usually done by a device that can
be used as a T-bar
Co-authored-by: Jim <obs.jim@gmail.com>
When obs_properties_remove_by_name is called on any obs_properties_t*,
it corrupts the pointers for first_property and last which end up
pointing at either unallocated memory or randomly into the heap memory.
Neither of these is a good thing, and it usually leads to rapid
unscheduled program behavior, also known as crashing and security
issues.
This fixes the issue by first checking if the pointer stored in
props->last is identical to &cur->next, then checking if we are the
only element (cur is also prev element), and if we are then the pointer
is fixed to point back at props->first_property. Additionally fixes
props->first_property which was never updated either.
