The mesa state tracker can emulate this, but we should pass
it through properly to the host in case it has a more efficent path.
Reviewed-By: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
When I added indirect draw support, I forgot to add support
for the binding flag for command args, we need this later,
but in order to introduce support without breaking things,
we should fix the bug first and add a separate cap for it.
Reviewed-By: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
When we add new feature checks on the host side that is used to enable
a cap conditionally in the guest that was enabled unconditionally before
we might end up with a feature regression when a new mesa version is
used with an old virglrenderer version that doesn't check for that cap.
To work around this problem add a version id to the caps that corresponds
to the features that are actually checked on the host so that it can be
checked in the guest whether this cap was actually checked for or whether
it should be enabled unconditionally.
The id should be incremented whenever a new feature check is added that
might result in a feature regression in the guest when run on an old host.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Pohsien Wang <pwang@chromium.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The shaders are issued backwards starting from the fragment
shader and, therefore, in the first pass when we fill out
the shader keys for the TES no TCS will be present.
So don't bail out in this case and assume there is a VS.
Later, when the VS shader is compiled it will check whether
a TES is present without a TCS and report the error.
Fixes: 956a6ceb8d
shader: Pass information about the layout of generics and patches to the next stage
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: <Gurchetan Singh gurchetansingh@chromium.org>
Softpipe doesn't support ARB_GLES3_1_compatibility but can support
NV_shader_atomic_float and this is needed for some dEQP-GLES31 tests to be
run on a softpipe GL host.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
(u|i)mulExtended are provided by ARB_gpu_shaders5 but also by
MESA_shader_integer_functions and only the latter is supported by softpipe
so fall back to this extension if needed.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
This is needed for using the right extension in the shaders.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
On GLES GL_EXT_disjoint_timer_query provides this functionality, so we
should make use of it.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Makes things a tiny bit more consistent and easier to read.
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
[airlied: rebased and fixed these up]
Signed-off-by: Dave Airlie <airlied@redhat.com>
Use explicit named initializers for the enum to string mappings.
Makes the code tiny bit easier to follow and grep through.
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
[airlied: rebased these and fix up the fallout]
Signed-off-by: Dave Airlie <airlied@redhat.com>
This fixes running Metro Redux 2033 on the GLES host.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Fixes:
KHR-GL41.viewport_array.dynamic_viewport_index
KHR-GL41.viewport_array.draw_mulitple_viewports_with_single_invocation
when run in a batch as KHR-GL41.viewport_array.*
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
With OES_viewport_array available the depth ranges can also be set
on GLES.
Fixes on GLES when run individually:
KHR-GL41.viewport_array.*
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Apart from fp64 about which we lie aynway the features required by
GLSL level 4.10 on top of 4.00 can be supported by GLES, so report
this higher feature level and it will give us GL 4.1 in the guest.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This enables indexed access to viewports on GLES hosts.
v2: Also add the extension to shaders when needed
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This helps a lot locating errors in the shaders.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Check args of the following function:
- vrend_decode_set_vertex_buffers
- vrend_decode_set_shader_buffers
- vrend_decode_set_atomic_buffers
- vrend_decode_set_shader_images
And change variable type to uint as the protocol should never send negative number.
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Left shift int 1 by 31 bit is undefined behavior, which causes fuzzer
failure. Change 1 to 1u to avoid the runtime error.
Reviewed-by: Elie Tournier <elie.tournier@collabora.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
There's little reason to confuse lod-bias and cubemap seamlessness. So
let's give lod-bias its own warning string.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
This flag is *always* passed as zero, so it carries no information.
Let's just get rid of it.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
This flag is *always* passed as zero, so it carries no information.
Let's just get rid of it.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Alexandros Frantzis <alexandros.frantzis@collabora.com>
These already have to be the same as VIRGL_TRANSFER_{TO,FROM}_HOST, as
defined by the protocol. There's no point in having a separate set of
internal names for them.
While we're at it, rewrite tan if as a switch for clarity. This allows
us to detect internal usage of invalid values.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Virglrenderer sometimes tries to remove resources from the hash table
twice. Which will mess up the ressource hash table and reference counts
and therefore and leads to qemu/virglrenderer crashes.
Reproducer:
(a) guest creates resource foo, id 42.
(b) guest creates an object bar referencing resource foo.
(c) guest unreferences resource foo.
-> resource id 42 is removed from hash.
(d) guest creates a new resource baz, re-using id 42.
(e) guest destroys object bar.
-> resource foo refcount goes down to zero.
-> resource id 42 gets removed from hash the second time,
but id 42 entry points to resource baz not foo now.
Oops.
Note that most linux kernel drivers will never ever re-use resource ids
due to a bug in the virtio-gpu kms driver, in which case this bug
doesn't cause any harm.
Root cause is that vrend_renderer_resource_destroy() may call
vrend_resource_remove(), depending on the call chain. This is wrong.
Only vrend_renderer_resource_unref() which is called when the guest
unreferences a resource should remove the id from the hash table by
calling vrend_resource_remove().
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: <Gurchetan Singh gurchetansingh@chromium.org>
mesa optimizes away the information about input layouts for generics and
patches for TCS, TES, and GEOM shaders (all that pass these inputs as
arrays), but when input arrays are allowed, then these varyings may
actually have overlapping layouts (at least some piglits do, even though
they don't specifically require ARB_enhanced_layouts).
To be able to link shaders like these with enables arrays pass this info
to the next stage.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-By: Gurchetan Singh <gurchetansingh@chromium.org>
Arrays of arrays are way easier to handle and they are available on GLES
3.1, D-GL 4.3, and all hardware that is supported by mesa. The old code
path is still able to use blocks for passing parameters to and from
tesselation and geometry stages, but for the new code path they are
required.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-By: Gurchetan Singh <gurchetansingh@chromium.org>
Set the pre stage correctly when the TCS shader is missing on OpenGL and
warn about it when on OpenGL ES.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
We already have a helper for this, let's use it.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
When calling vrend_renderer_transfer_write_iov through
vrend_renderer_transfer_iov, we sometimes come down these
code-paths with the context 0 active, but the non-zero context being
passed as the context-pointer. Trying to do any caching based on this
will only lead to incorrect behavior.
There's little point in trying to fixup the caching in the ctx0 for
this call-site, as there's no other code-path using ctx0 that tries
to play ball with the cache.
So let's get rid ot the context-argument when possible, and pass a
NULL-pointer when it can come either from ctx0 or a rendering-context.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Without this, the next patch gets trickier, because we try to
*avoid* passing down the wrong data rather than trying to react to it.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
This argument is unused, and the function is not a part of the external
API, so there's little point in passing it just to discard it again.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
Context name length was being ignored and strncpy does not enforce
null termination but the destination was being used as a null terminated
string.
Change-Id: I715394b364130578a1a6a319dc16927261523d71
Signed-off-by: David Riley <davidriley@chromium.org>
Reviewed-By: Gert Wollny <gert.wollny@collabora.com>
EXT_texture_query_lod on GLES is proposed as extension providing for
GLES what ARB_texture_query_lod provides for D-GL.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
GL_TEXTURE_LOD_BIAS is supported in GLES at least since 3.0,
This also fixes a number of piglits when run on a GLES host.
GL_TEXTURE_SEAMLESS_CUBE_MAP is not supported on the other hand.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
This GLES extension has been merged into the OpenGL registry, so it may
now become available.
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
To avoid that the guest sends mixed color attachments when this is not
supported test this and add an according caps flag.
Related: #88 (a full fix needs a mesa side patch to make use of the
flag sent)
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This tracks more accurately what exactly needs to be updated.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
We're going to have to do this anyway, so let's just do it cleanly
right away.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
These settings can't change in the lifetime of the texture-view,
so let's set them up front instead of re-setting them on every re-emit.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
There's no point in changing the texture-binding unless we're going to
do something with the texture-stage.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
This means less needless work per frame, as this binding will never
change.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
The result is the same, we just skip a bit quicker over unused entries.
Signed-off-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
A transfer won't occur in context zero if the transfer command
comes from submit_cmd. We need to restore prior state.
With the recent framebuffer binding cleanups, we really only need
to worry about the framebuffer.
Reviewed-by: Gert Wollny <gert.wollny@collabora.com>
d9a4be4d0a broke a number of tests
when running on a GLES host because tgsitargettogltarget may
return PIPE_BUFFER which is 0. So, instead of simply reverting the commit
make it more explicit that we check for PIPE_BUFFER to make it clearer
that this condition may actually trigger.
Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Signed-off-by: Jakob Bornecrantz <jakob@collabora.com>
Dave Airlie
Gert Wollny
Emil Velikov
Emil Velikov
Po-Hsien Wang
Erik Faye-Lund
Gerd Hoffmann
Stéphane Marchesin
Elie Tournier
David Riley
Gurchetan Singh