The hashtable stores a hash generated from the key, so there is no need
to duplicate the key, as otherwise it is a leak.
Found through address sanitizer.
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30204>
Also optimize the denorm scaling path by only emitting the expensive trans op once
and allowing fma for the final muliplication.
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/29245>
We were generating odd instructions like:
math inv(8) g93<1>HF g85<8,8,1>HF null<8,8,1>F { align1 1Q @7 $4 };
It's unclear whether the type of the null operand matters, but sometimes
these things don't get ignored properly. Out of caution, retype the
null source to match the actual operand's type. It'll at least look
less surprising in assembly dumps.
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30193>
We had some helpers for this at one point but the old ones were super
clunky and didn't really do what we wanted so they were removed.
However, we have a lot of manual banging in opt_copy_prop and we're
about to add more. These new helpers will make it all a lot safer.
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30230>
When a symbol is looked up for the first time, the associated function
is built, and the building process seems to be not thread-safe.
Use a mutex to protect the symbol looking up process, which should be
serialized when the function is not built, and fast when the function is
built.
Signed-off-by: Icenowy Zheng <uwu@icenowy.me>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30217>
xserver's loader will look for drivers this way, but there's no reason
we need to keep hurting ourselves like that.
Co-authored-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28378>
This provides just enough of a "DRI driver" for non-glamor-using
xservers to initialize GLX and enable direct clients. We build it if you
build DRI support for GLX, or if you build any X11 client support for
EGL. We only build it at this point, we'll swap it into the install
next.
Co-authored-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Co-authored-by: Pavel Ondračka <pavel.ondracka@gmail.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28378>
Move it into src/gallium/include/ to make it absolutely clear this is a Mesa detail.
While we're at it, clean up its include sites, including some places
where we can just include kopper_interface.h instead since it includes
mesa_interface.h as its first act.
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28378>
We're about to split the latter off as a compatibility detail for older
versions of Xorg, and the former includes the latter at this point, so
this should be just to prove no functional change.
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/28378>
Event::set_status only called the cbs for the passed in status, but we
need it to call all cbs up to the passed in status. This simplifies error
handling.
Cc: mesa-stable
Reviewed-by: @LingMan
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30215>
If the driver needs to create two different DCC retile pipelines which
is based on the image swizzle, it will just overwrite the existing
layouts.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30233>
