Ensures early algebraic passes aren't called again following late
algebraic passes, so that the latter's opts aren't undone (e.g.
unfusing ffmas).
Signed-off-by: Simon Perretta <simon.perretta@imgtec.com>
Acked-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37724>
The rounding behaviour on [iu]2f32 ops needs to be explicitly set in
order to match the implicit behaviour described in the
KHR_shader_float_controls properties.
Fixes: e306abc6e6 ("pvr: implement KHR_shader_float_controls")
Signed-off-by: Simon Perretta <simon.perretta@imgtec.com>
Acked-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37724>
The Imagination PowerVR Vulkan driver is now conformant on BXS-4-64, so drop the
PVR_I_WANT_A_BROKEN_VULKAN_DRIVER runtime check for this GPU. This eliminates
the need for the user to explicitly opt in via an environment variable.
Signed-off-by: Ashish Chauhan <Ashish.Chauhan@imgtec.com>
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37761>
The kernel driver zero initialises device memory allocations for us, so all that
needs to be done is to advertise support for the extension.
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Reviewed-by: Karmjit Mahil <karmjit.mahil@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37732>
This is so that passes and backends can tell if a coherent load/store is
atomic or not, instead of having to assume it could be either.
Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/36602>
This header is only used in a single place, which will generate macros
based on these definitions. Let's remove the header, and just include
the separate files in pvr_csb.h instead.
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37675>
Set the has_f2i32_rtne shader compiler option to indicate hardware
support for it. This enables NIR's late algebraic optimization pass to
generate more efficient code for float-to-int conversions.
Signed-off-by: Christian Gmeiner <cgmeiner@igalia.com>
Reviewed-by: Simon Perretta <simon.perretta@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37680>
This header contains a lot of includes, and gets everywhere. Let's make
sure we don't include it from headers, which makes this much easier to
manage.
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37432>
To avoid some circular dependencies due to pvr_private.h, split out
pvr_framebuffer etc into their own header.
We often only need to peek into the framebuffer, so this seems like a
good idea anyway. We can reconsider this once dynamic rendering has
landed, and we know how much remains here.
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37432>
This is used in a lot of places in the driver, and doesn't naturally
belong in any of the smaller modules that we're about to introduce.
Reviewed-by: Frank Binns <frank.binns@imgtec.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37432>
Enabling DS resolve attachment entails a blit for a ZLS subtile region, in such
case the TQ needs to know the number of samples in the source surface in order to
pack the texture state.
Signed-off-by: Luigi Santivetti <luigi.santivetti@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
This commit enables the VK_KHR_depth_stencil_resolve extension in the PVR
Vulkan driver. Only worth to note that this implementation is based on the
renderpass driver path and it will need a rework as part of the dynamic
rendering enablement.
Signed-off-by: leonperianu <leon.perianu@imgtec.com>
Co-authored-by: Luigi Santivetti <luigi.santivetti@imgtec.com>
Co-authored-by: Ashish Chauhan <ashish.chauhan@imgtec.com>
Co-authored-by: Simon Perretta <simon.perretta@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
Do not conflate api and internal driver use of copy commands and make the
correct assumptions for the api path, according to the Vulkan spec.
Signed-off-by: Luigi Santivetti <luigi.santivetti@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
The driver already implements the regular render pass functions in terms of the
VK_KHR_create_renderpass2 functions. However, the extension couldn't be
advertised due to missing support for VK_KHR_multiview. Now multiview is
supported, renderpass2 can be advertised as well.
Signed-off-by: Frank Binns <frank.binns@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
- Pass view index to fragment shader via linked flat varying.
- Use view index instead of layer id for input attachments when required.
Signed-off-by: Simon Perretta <simon.perretta@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
- Fix ds address offset for multiview.
- Fix multiview depth/stencil address by restoring it after every kick.
- Enable empty tile processing after first subpass for multiview.
Signed-off-by: Peter Quayle <peter.quayle@imgtec.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37512>
