| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A driver which doesn't have async flip support will queue up flips without any
way to replace them afterwards. This means we've got a scanout buffer pinned
as soon as we schedule a flip and so we need another buffer to keep from
stalling.
When vblank_mode=0, if there are only three buffers we do:
current scanout buffer = 0 at MSC 0
Render frame 1 to buffer 1
PresentPixmap for buffer 1 at MSC 1
This is sitting down in the kernel waiting for vblank to
become the next scanout buffer
Render frame 2 to buffer 2
PresentPixmap for buffer 2 at MSC 1
This cannot be displayed at MSC 1 because the
kernel doesn't have any way to replace buffer 1 as the pending
scanout buffer. So, best case this will get displayed at MSC 2.
Now we block after this, waiting for one of the three buffers to become idle.
We can't use buffer 0 because it is the scanout buffer. We can't use buffer 1
because it's sitting in the kernel waiting to become the next scanout buffer
and we can't use buffer 2 because that's the most recent frame which will
become the next scanout buffer if the application doesn't manage to generate
another complete frame by MSC 2.
With four buffers, we get:
current scanout buffer = 0 at MSC 0
Render frame 1 to buffer 1
PresentPixmap for buffer 1 at MSC 1
This is sitting down in the kernel waiting for vblank to
become the next scanout buffer
Render frame 2 to buffer 2
PresentPixmap for buffer 2 at MSC 1
This cannot be displayed at MSC 1 because the
kernel doesn't have any way to replace buffer 1 as the pending
scanout buffer. So, best case this will get displayed at MSC
2. The X server will queue this swap until buffer 1 becomes
the scanout buffer.
Render frame 3 to buffer 3
PresentPixmap for buffer 3 at MSC 1
As soon as the X server sees this, it will replace the pending
buffer 2 swap with this swap and release buffer 2 back to the
application
Render frame 4 to buffer 2
PresentPixmap for buffer 2 at MSC 1
Now we're in a steady state, flipping between buffer 2 and 3
waiting for one of them to be queued to the kernel.
...
current scanout buffer = 1 at MSC 1
Now buffer 0 is free and (e.g.) buffer 2 is queued in
the kernel to be the scanout buffer at MSC 2
Render frames, flipping between buffer 0 and 3
When the system can replace a queued buffer, and we update Present to take
advantage of that, we can use three buffers and get:
current scanout buffer = 0 at MSC 0
Render frame 1 to buffer 1
PresentPixmap for buffer 1 at MSC 1
This is sitting waiting for vblank to become the next scanout
buffer
Render frame 2 to buffer 2
PresentPixmap for buffer 2 at MSC 1
Queue this for display at MSC 1
1. There are three possible results:
1) We're still before MSC 1. Buffer 1 is released,
buffer 2 is queued waiting for MSC 1.
2) We're now after MSC 1. Buffer 0 was released at MSC 1.
Buffer 1 is the current scanout buffer.
a) If the user asked for a tearing update, we swap
scanout from buffer 1 to buffer 2 and release buffer
1.
b) If the user asked for non-tearing update, we
queue buffer 2 for the MSC 2.
In all three cases, we have a buffer released (call it 'n'),
ready to receive the next frame.
Render frame 3 to buffer n
PresentPixmap for buffer n
If we're still before MSC 1, then we'll ask to present at MSC
1. Otherwise, we'll ask to present at MSC 2.
Present already does this if the driver offers async flips, however it does
this by waiting for the right vblank event and sending an async flip right at
that point.
I've hacked the intel driver to offer this, but I get tearing at the top of
the screen. I think this is because flips are always done from within the
ring, and so the latency between the vblank event and the async flip happening
can cause tearing at the top of the screen.
That's why I'm keying the need for the extra buffer on the lack of 2D
driver support for async flips.
Signed-off-by: Keith Packard <[email protected]>
Acked-by: Jason Ekstrand <[email protected]>
Tested-by: Dylan Baker <[email protected]>
|
| |
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
| |
Need to unwrap the indirect resource otherwise bad things will happen.
Fixes random crashes and timeouts with piglit's arb_indirect_draw tests.
Reviewed-by: Jose Fonseca <[email protected]>
|
|
|
|
| |
Reviewed-by: Jose Fonseca <[email protected]>
|
|
|
|
|
|
|
| |
IVB had a restriction that prevented us from emitting compressed
three-source instructions, and although that was lifted on Haswell,
Haswell had a new restriction that said BFI instructions specifically
couldn't be compressed.
|
|
|
|
|
|
|
| |
These checks were intended for Gen 7 only. None of these restrictions
apply to Gen 8.
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
| |
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Transform
sqrt a, b
rcp c, a
into
sqrt a, b
rsq c, b
The improvement here is that we've broken a dependency between these
instructions. Leads to 330 fewer INV instructions and 330 more RSQ.
Reviewed-by: Anuj Phogat <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Transform
sqrt a, b
rcp c, a
into
sqrt a, b
rsq c, b
In most cases the sqrt's result is still used, so the improvement here
is that we've broken a dependency between these instructions. Leads to
80 fewer INV instructions and 80 more RSQ.
Occasionally the sqrt's result is no longer used, leading to:
instructions in affected programs: 5005 -> 4949 (-1.12%)
Reviewed-by: Anuj Phogat <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The next patch adds an algebraic optimization for the pattern
sqrt a, b
rcp c, a
and turns it into
sqrt a, b
rsq c, b
but many vertex shaders do
a = sqrt(b);
var1 /= a;
var2 /= a;
which generates
sqrt a, b
rcp c, a
rcp d, a
If we apply the algebraic optimization before CSE, we'll end up with
sqrt a, b
rsq c, b
rcp d, a
Applying CSE combines the RCP instructions, preventing this from
happening.
No shader-db changes.
Reviewed-by: Anuj Phogat <[email protected]>
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
| |
Helps a handful of programs in Serious Sam 3 that use do-while loops.
instructions in affected programs: 16114 -> 16075 (-0.24%)
Reviewed-by: Ian Romanick <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Do not rely on LLVMMCJITMemoryManagerRef being available.
The c binding to the memory manager objects only appeared
on llvm-3.4.
The change is based on an initial patch of Brian Paul.
Reviewed-by: Brian Paul <[email protected]>
Tested-by: Brian Paul <[email protected]>
Reviewed-by: Roland Scheidegger <[email protected]>
Signed-off-by: Mathias Froehlich <[email protected]>
|
|
|
|
|
|
|
|
| |
Blitter can still have transfers hanging around which it frees in
util_blitter_destroy(). So let it clean up before we yank the
transfer_pool from under it.
Signed-off-by: Rob Clark <[email protected]>
|
|
|
|
|
|
|
| |
Indirect registers consume an additional token. Try to clean up the
token calculation math a bit, and fix it at the same time.
Signed-off-by: Rob Clark <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
If the name is just going to get dropped, don't bother making it. If
the name is made, release it sooner (rather than later).
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
If the name is just going to get dropped, don't bother making it. If
the name is made, release it sooner (rather than later).
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 74 40,578,719,715 67,762,208 62,263,404 5,498,804 0
After (32-bit): 52 40,565,579,466 66,359,800 61,187,818 5,171,982 0
Before (64-bit): 74 37,129,541,061 95,195,160 87,369,671 7,825,489 0
After (64-bit): 76 37,134,691,404 93,271,352 85,900,223 7,371,129 0
A real savings of 1.0MiB on 32-bit and 1.4MiB on 64-bit.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
These few places were using ir_var_auto for seemingly no reason. The
names were not added to the symbol table.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
| |
No change Valgrind massif results for a trimmed apitrace of dota2.
v2: Minor rebase on _mesa_init_constants changes.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
Later patches will give every ir_var_temporary the same name in release
builds. Adding a bunch of variables named "compiler_temp" to the symbol
table can only cause problems.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
Specifically, ir_var_temporary variables constructed with a NULL name
will all have the name "compiler_temp" in static storage.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 44 40,577,049,140 68,118,608 62,441,063 5,677,545 0
After (32-bit): 71 40,583,408,411 67,761,528 62,263,519 5,498,009 0
Before (64-bit): 63 37,122,829,194 95,153,008 87,333,600 7,819,408 0
After (64-bit): 67 37,123,303,706 95,150,544 87,333,600 7,816,944 0
A real savings of 173KiB on 32-bit and no change on 64-bit.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
At least one of these pointers must be NULL, and we can determine which
will be NULL by looking at other fields. Use this information to store
both pointers in the same location.
If anyone can think of a better name for the union than "u", I'm all
ears.
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 63 40,574,239,515 68,117,280 62,618,607 5,498,673 0
After (32-bit): 44 40,577,049,140 68,118,608 62,441,063 5,677,545 0
Before (64-bit): 53 37,126,451,468 95,150,256 87,711,304 7,438,952 0
After (64-bit): 63 37,122,829,194 95,153,008 87,333,600 7,819,408 0
A real savings of 173KiB on 32-bit and 368KiB on 64-bit.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also move num_state_slots inside ir_variable_data for better packing.
The payoff for this will come in a few more patches.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The payoff for this will come in a few more patches.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
warn_extension_index was moved to improve packing.
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 73 40,580,476,304 68,488,400 62,796,151 5,692,249 0
After (32-bit): 73 40,575,751,558 68,116,528 62,618,607 5,497,921 0
Before (64-bit): 71 37,124,890,613 95,889,584 88,089,008 7,800,576 0
After (64-bit): 62 37,123,578,526 95,150,784 87,711,304 7,439,480 0
A real savings of 173KiB on 32-bit and 368KiB on 64-bit.
v2: Use the enum name with the bit-field and remove the extra casts.
Suggested by Ken.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Kenneth Graunke <[email protected]>
Reviewed-by: Matt Turner <[email protected]> [v1]
Reviewed-by: Tapani Pälli <[email protected]> [v1]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Also move the new warn_extension_index into ir_variable::data. This
enables slightly better packing.
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 82 40,580,040,531 68,488,992 62,973,695 5,515,297 0
After (32-bit): 73 40,580,476,304 68,488,400 62,796,151 5,692,249 0
Before (64-bit): 65 37,124,013,542 95,892,768 88,466,712 7,426,056 0
After (64-bit): 71 37,124,890,613 95,889,584 88,089,008 7,800,576 0
A real savings of 173KiB on 32-bit and 368KiB on 64-bit.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
| |
The payoff for this will come in the next patch.
No change Valgrind massif results for a trimmed apitrace of dota2.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Reviewed-by: Tapani Pälli <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
After compilation (and before linking) we can eliminate quite a few
built-in variables. Basically, any uniform or constant (e.g.,
gl_MaxVertexTextureImageUnits) that isn't used (with one exception) can
be eliminated. System values, vertex shader inputs (with one
exception), and fragment shader outputs that are not used and not
re-declared in the shader text can also be removed.
gl_ModelViewProjectMatrix and gl_Vertex are used by the built-in
function ftransform. There are some complications with eliminating
these variables (see the comment in the patch), so they are not
eliminated.
Valgrind massif results for a trimmed apitrace of dota2:
n time(i) total(B) useful-heap(B) extra-heap(B) stacks(B)
Before (32-bit): 46 40,661,487,174 75,116,800 68,854,065 6,262,735 0
After (32-bit): 50 40,564,927,443 69,185,408 63,683,871 5,501,537 0
Before (64-bit): 64 37,200,329,700 104,872,672 96,514,546 8,358,126 0
After (64-bit): 59 36,822,048,449 96,526,888 89,113,000 7,413,888 0
A real savings of 4.9MiB on 32-bit and 7.0MiB on 64-bit.
v2: Don't remove any built-in with Transpose in the name.
v3: Fix comment typo noticed by Anuj.
Signed-off-by: Ian Romanick <[email protected]>
Suggested-by: Eric Anholt <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Acked-by: Anuj Phogat <[email protected]>
Cc: Eric Anholt <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
| |
All built-in uniforms are supposed to be backed by some GL state. The
state_slots field describes this backing state.
This helped me track down a bug in a later patch.
Signed-off-by: Ian Romanick <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
Acked-by: Anuj Phogat <[email protected]>
|
|
|
|
|
| |
Otherwise, we'd replace the stencil in our packed depth/stencil with 0s.
Fixes about 50 piglit tests.
|
|
|
|
|
|
|
|
|
|
|
| |
Reuse the LLVMContext already allocated in llvmpipe_context
for draw_llvm if ppossible. This should decrease the memory
footprint of an llvmpipe context.
v2: Fix compile with llvm disabled.
Reviewed-by: Jose Fonseca <[email protected]>
Signed-off-by: Mathias Froehlich <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This fixes the remaining problem with the recently introduced
global jit memory manager. This change again uses a memory manager
that is local to gallivm_state. This implementation still frees
the majority of the memory immediately after compilation.
Only the generated code is deferred until this code is no longer used.
This change and the previous one using private LLVMContext instances
I can now safely run several independent OpenGL contexts driven
by llvmpipe from different threads.
v3: Rebase on llvm-3.6 compile fixes.
Reviewed-by: Jose Fonseca <[email protected]>
Signed-off-by: Mathias Froehlich <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is one step to make llvmpipe thread safe as mandated by the OpenGL
standard. Using the global LLVMContext is obviously a problem for
that kind of use pattern. The patch introduces two LLVMContext
instances that are private to an OpenGL context and used for all
compiles. One is put into struct draw_llvm and the other
one into struct llvmpipe_context.
Reviewed-by: Jose Fonseca <[email protected]>
Signed-off-by: Mathias Froehlich <[email protected]>
|
|
|
|
|
|
|
| |
The big pile of patches I just pushed regresses about 25 piglit tests on
SNB. This fixes the regressions.
Signed-off-by: Jason Ekstrand <[email protected]>
|
|
|
|
|
|
|
|
|
| |
The screen argument isn't actually used by lp_jit_screen_init() at this
time, but let's move the call so that we pass a valid pointer.
v2: don't leak screen if lp_jit_screen_init() fails.
Reviewed-by: Roland Scheidegger <[email protected]>
|
|
|
|
|
|
|
| |
Assign the sem_name parameter, not TGSI_SEMANTIC_GENERIC.
Fixes polygon stipple regression.
Reviewed-by: Charmaine Lee <[email protected]>
|
|
|
|
|
|
|
|
|
| |
For cube resources, the array_size value should be 6. So handle
that case as we do for array texture resources. But assert that
array_size==6 just to be safe.
Reviewed-by: Ilia Mirkin <[email protected]>
Reviewed-by: Marek Olšák <[email protected]>
|
|
|
|
|
|
| |
As with the previous patch for llvmpipe.
Reviewed-by: Ilia Mirkin <[email protected]>
|
|
|
|
|
|
|
| |
layers (aka array_size) should be 6 for cube textures so we don't need
to special-case it. But add an assertion just to be safe.
Reviewed-by: Ilia Mirkin <[email protected]>
|
|
|
|
|
|
| |
Just to be clear, and echo the description for resource_create().
Reviewed-by: Ilia Mirkin <[email protected]>
|
|
|
|
|
|
|
| |
Earlier in the function we assert layers==6 for PIPE_TEXTURE_CUBE so
there's no reason to special-case the pt.array_size = layers assignment.
Reviewed-by: Ilia Mirkin <[email protected]>
|
|
|
|
|
|
|
| |
The core sw primitive restart code is still around, because i965 uses it
in some cases, but there are no drivers that want it on all the time.
Reviewed-by: Rob Clark <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The drivers not flagging primitive restart support are r300 swtcl, svga,
nv30, and vc4.
The point of primitive restart is to slightly reduce draw call overhead
for apps by batching multiple draws. If we do an extra pass to read the
index buffer and split back into multiple draws, we've entirely missed the
point. This is particularly bad for drivers that otherwise have hardware
IB reads, where the readback is probably uncached.
Reviewed-by: Rob Clark <[email protected]>
|
|
|
|
|
|
|
| |
calculate_register_pressure
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On gen 7, the MRF was removed and we gained the ability to do send
instructions directly from the GRF. This commit enables that
functinoality for FB writes.
v2: Make handling of components more sane.
i965/fs: Force a high register for the final FB write
v2: Renamed the array for the range mappings and added a comment
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
| |
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|
|
|
|
|
|
|
|
| |
If we are going to use LOAD_PAYLOAD operations to fill MRF registers, then
we will need this.
Signed-off-by: Jason Ekstrand <[email protected]>
Reviewed-by: Matt Turner <[email protected]>
|