Commit Graph

86 Commits

Author SHA1 Message Date
Kenneth Graunke
2b956a093a iris: totally untested icelake support 2019-02-21 10:26:08 -08:00
Kenneth Graunke
eff081cdd9 iris: Support multiple binder BOs, update Surface State Base Address 2019-02-21 10:26:08 -08:00
Kenneth Graunke
5a2257bb2f iris: don't unconditionally emit 3DSTATE_VF / 3DSTATE_VF_TOPOLOGY
this was just laziness on my part
2019-02-21 10:26:08 -08:00
Kenneth Graunke
93c1921ce2 iris: proper cache tracking
this is copied from the i965 aux resolve stuff...minus the aux resolves
2019-02-21 10:26:08 -08:00
Kenneth Graunke
5e30b1083b iris: Move cache tracking to iris_resolve.c 2019-02-21 10:26:08 -08:00
Kenneth Graunke
1d33982e9b iris: track depth/stencil writes enabled 2019-02-21 10:26:08 -08:00
Kenneth Graunke
b75b52530a iris: Move things to iris_shader_state
We didn't originally have this struct, so we had lots of ad-hoc arrays.
Now that we have it, it makes sense to group things there.
2019-02-21 10:26:08 -08:00
Kenneth Graunke
410a555bfb iris: move iris_shader_state from ice->shaders.state to ice->state.shaders
it's more state related...
2019-02-21 10:26:08 -08:00
Kenneth Graunke
2bd7d6fa71 iris: last VUE map NOS, handle > 16 FS inputs
not sure if the UNCOMPILED_FS flagging is still needed, should
reevaluate those hacks at some point
2019-02-21 10:26:08 -08:00
Kenneth Graunke
dfe1ee4f6f iris: comment everything
1. Write the code
2. Add comments
3. PROFIT (or just avoid cost of explaining or relearning things...)
2019-02-21 10:26:08 -08:00
Kenneth Graunke
387a414f2c iris: add minor comments 2019-02-21 10:26:08 -08:00
Dave Airlie
9d39e69219 iris: fix some hangs around null framebuffers
This fixes some cases in fbo-none* and framebuffer_no_attachments.

I'm not sure this is correct otherwise, the tests don't all pass yet

No idea if this is in any way the correct answer
2019-02-21 10:26:08 -08:00
Kenneth Graunke
870f2e8434 iris: implement texture/memory barriers 2019-02-21 10:26:07 -08:00
Kenneth Graunke
b7b061c4e2 iris: fix SSBO indexing
st/nir offsets SSBO indexes by MaxABOs.  This is not what we want,
as it bloats the binding tables.  We'll need to adjust it to use
info->num_abos as the offset and buffer base instead.  For now,
just use the inefficient format to get us rolling.  We can add a
PIPE_CAP later.
2019-02-21 10:26:07 -08:00
Kenneth Graunke
77b9219818 iris: unbind compiled shaders if none are present
avoids the case where you have a stale compiled shader bound, but no
uncompiled shader bound, which is not just boats, but an entire marina
2019-02-21 10:26:07 -08:00
Kenneth Graunke
2d5f545464 iris: implement set_shader_buffers
for SSBOs/ABOs.  We just stream out SURFACE_STATE for now...since it's
a set_* API...and the buffer offset may change...not sure where else
we'd do it.
2019-02-21 10:26:07 -08:00
Kenneth Graunke
541cb60e7e iris: export get_shader_info 2019-02-21 10:26:07 -08:00
Kenneth Graunke
5520a54bc5 iris: vertex ID, instance ID 2019-02-21 10:26:07 -08:00
Kenneth Graunke
d223b316ad iris: NOS mechanics 2019-02-21 10:26:07 -08:00
Kenneth Graunke
b030671298 iris: fix CC_VIEWPORT
I was confusing depth bounds test with depth clamping
2019-02-21 10:26:07 -08:00
Kenneth Graunke
7b8c0f058e iris: actually pin the buffers 2019-02-21 10:26:07 -08:00
Kenneth Graunke
cef0b8b13b iris: streamout 2019-02-21 10:26:07 -08:00
Kenneth Graunke
5c00f5fdca iris: Implement 3DSTATE_SO_DECL_LIST 2019-02-21 10:26:07 -08:00
Kenneth Graunke
a3f77eceb4 iris: slab allocate transfers
apparently we need this for u_threaded_context
2019-02-21 10:26:07 -08:00
Kenneth Graunke
7f39f4843f iris: SBE change stash
not used yet, but want to flag it so I don't forget
2019-02-21 10:26:07 -08:00
Kenneth Graunke
8a080223e6 iris: just malloc one iris_genx_state instead of a bunch of oddball pieces
Things that are gen-specific can go in iris_genx_state.  Things that are
gen-agnostic can go directly in ice->state.
2019-02-21 10:26:07 -08:00
Kenneth Graunke
a7e0edffb6 iris: dead pointer 2019-02-21 10:26:07 -08:00
Kenneth Graunke
ccec5bab5b iris: implement border color, fix other sampler nonsense 2019-02-21 10:26:07 -08:00
Kenneth Graunke
1cea195a95 iris: state ref tuple 2019-02-21 10:26:07 -08:00
Kenneth Graunke
c0e80a8d0a iris: null surface for unbound textures
avoids crashes...may not be really right
2019-02-21 10:26:07 -08:00
Kenneth Graunke
72416a2d0d iris: clears 2019-02-21 10:26:07 -08:00
Kenneth Graunke
81f899c148 iris: scissor count fixes 2019-02-21 10:26:07 -08:00
Kenneth Graunke
5f722bf7c4 iris: remove 4 bytes of padding in iris_compiled_shader 2019-02-21 10:26:07 -08:00
Kenneth Graunke
23987df412 iris: some dirty fixes
two scissor bits, constants not being flagged, ZeroRTA, clip not being
flagged
2019-02-21 10:26:07 -08:00
Kenneth Graunke
ccf37c7da9 iris: bindings dirty tracking 2019-02-21 10:26:07 -08:00
Kenneth Graunke
80dee31846 iris: save pointers to streamed state resources
will be used for cross-batch validation list fixing
2019-02-21 10:26:06 -08:00
Kenneth Graunke
9247546181 iris: state cleaning 2019-02-21 10:26:06 -08:00
Kenneth Graunke
90046b43cc iris: make surface states for cbufs 2019-02-21 10:26:06 -08:00
Kenneth Graunke
4e007dbb30 iris: have more than one const_offset 2019-02-21 10:26:06 -08:00
Kenneth Graunke
9ea05ccf1f iris: completely rewrite binder
now we get a new one per batch, and flush if it fills up
2019-02-21 10:26:06 -08:00
Kenneth Graunke
26cc609927 iris: better ubo handling 2019-02-21 10:26:06 -08:00
Kenneth Graunke
7fb7704b2e iris: more depth stuffs...
still missing stencil
2019-02-21 10:26:06 -08:00
Kenneth Graunke
e2ba98ba39 iris: more blorp 2019-02-21 10:26:06 -08:00
Kenneth Graunke
698d45b725 iris: more blitting code to make readpixels work 2019-02-21 10:26:06 -08:00
Kenneth Graunke
c9d9e44720 iris: bits of blorp code 2019-02-21 10:26:06 -08:00
Kenneth Graunke
60d708bb80 iris: copy over i965's cache tracking
needed to split out vtbl so I can pipe control without ice
2019-02-21 10:26:06 -08:00
Kenneth Graunke
21acc00490 iris: crazy pipe control code
imported from ~kwg/mesa pcx-2, gen < 8 code dropped
2019-02-21 10:26:06 -08:00
Kenneth Graunke
5d2673ba7e iris: set sampler views 2019-02-21 10:26:05 -08:00
Kenneth Graunke
0707ff3f2f iris: assemble SAMPLER_STATE table at bind time
It's useless to allocate SAMPLER_STATEs in GPU memory on creation like
we do for SURFACE_STATES, because they need to be organized into a
contiguous block of memory.  But we can do that at bind time, rather
than draw time.
2019-02-21 10:26:05 -08:00
Kenneth Graunke
6cbd1d1692 iris: binders 2019-02-21 10:26:05 -08:00