summaryrefslogtreecommitdiffstats
path: root/src/gallium
Commit message (Collapse)AuthorAgeFilesLines
* freedreno/ir3: insert extra move into phiRob Clark2016-04-051-0/+10
| | | | | | | | | | | | | | | | | We had an implicit assumption that the phi src was assigned in it's source (pred) block leading into the phi. But this is not true with NIR, so we can't just ignore the source block specified in the nir_phi_src. Insert an extra mov in the source block. If it is not required the CP pass will take it back out again. Fixes: ./tests/spec/glsl-1.10/execution/vs-call-in-nested-loop.shader_test ./tests/spec/glsl-1.10/execution/vs-inner-loop-modifies-outer-loop-var.shader_test and probably others. Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: eliminate unnecessary absneg'sRob Clark2016-04-052-3/+26
| | | | | | | | | | | | | | | | The frontend inserts (abs) and (neg)'s to convert between NIR boolean (~0/0) and native boolean (1/0). So we'd end up with things like: cmps.s.ge r1.x, ... absneg.s r1.x, (neg)r1.x absneg.s r1.x, (abs)r1.x sel.b32 r2.x, r0.x, r1.x, r0.y The (neg) already gets collapsed due to the following (abs). Now by realizing that r1.x comes from a cmps.s instruction, we can drop the (abs) as well. Signed-off-by: Rob Clark <[email protected]>
* clover: Fix build against clang SVN >= r265359Michel Dänzer2016-04-051-0/+3
| | | | | Signed-off-by: Michel Dänzer <[email protected]> Reviewed-by: Tom Stellard <[email protected]>
* radeonsi: use bounded indexing for samplersBas Nieuwenhuizen2016-04-051-1/+4
| | | | | Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: use bounded indexing for constant buffersBas Nieuwenhuizen2016-04-051-2/+3
| | | | | Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* gallium/radeon: allow multiple exports of the same texture with different usageMarek Olšák2016-04-051-21/+33
| | | | | | | | | Instead of failing an assertion, disable DCC and CMASK on the first export that needs it, and merge the external usage flags. v2: clear the EXPLICIT_FLUSH flag if it's not set; whitespace fixes Reviewed-by: Michel Dänzer <[email protected]>
* freedreno/ir3: deal with duplicate phi sourcesRob Clark2016-04-041-5/+20
| | | | | | | | | | | | | Otherwise we end up with funny things like: mov.f32f32 r0.x, r1.y mov.f32f32 r0.x, r1.y (It doesn't happen as much after fixing the problem w/ CP into phi src, but it can still happen since we aren't too clever about generating phi sources in the first place.) Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: fix silly brain-fart in RARob Clark2016-04-041-2/+1
| | | | | | | We want to consider all the vars, not 1/32nd of them, when extending live-ranges. Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: don't cp into phi'sRob Clark2016-04-041-0/+6
| | | | | | | | The block defining a phi source might not have been executed. If we allow copy propagation, we could end up pointing to a src instruction in the wrong block. Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: we can't store immediate valuesRob Clark2016-04-041-0/+13
| | | | | | | | Fixes some transform-feedback piglits, like: bin/ext_transform_feedback-nonflat-integral Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: add dumping for use/def/live-in/live-outRob Clark2016-04-043-10/+42
| | | | | | Turned out to be useful to debug an issue in RA. Let's keep it. Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: drop unused instr category argRob Clark2016-04-045-114/+108
| | | | | | No longer used, so drop the extra arg to ir3_instr_create() Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: remove ir3_instruction::categoryRob Clark2016-04-0410-93/+84
| | | | Signed-off-by: Rob Clark <[email protected]>
* freedreno/ir3: encode instruction category in opc_tRob Clark2016-04-045-192/+201
| | | | | | | | | | | | | Been on my TODO list for a while. If nothing else this will make gdb properly grok the opc_t enum. This first step preserves ir3_instruction::category (with an added assert that category matches what is encoded in opc_t). Next step is to drop the category field (and arg to ir3_instr_create()), but that is split into next commit for bisectability and so that we can run piglit in the intermediate state to flush out any problems. Signed-off-by: Rob Clark <[email protected]>
* nvc0: add hardware ETC2 and ASTC support on GK20A and GM107+Ilia Mirkin2016-04-043-2/+64
| | | | Signed-off-by: Ilia Mirkin <[email protected]>
* gallivm: Introduce lp_format_intrinsic.Jose Fonseca2016-04-043-14/+54
| | | | | | | | | | For adding .v4f32 like suffixes to intrinsics, taking special care for scalar case, which was being often neglected. This fixes invalid IR when doing mipmap filtering on SSE2 (the only case where we'd use intrinsics with scalars.) Reviewed-by: Roland Scheidegger <[email protected]>
* gallivm: Use llvm.fabs.Jose Fonseca2016-04-031-8/+3
| | | | | | Exactly the same code. Reviewed-by: Roland Scheidegger <[email protected]>
* gallivm: Prefer backend agnostic intrinsic for rounding.Jose Fonseca2016-04-031-7/+39
| | | | | | | | | We could unconditionally use these instrinsics, but performance with SSE2 would suck, as LLVM falls back to calling libm. lp_test_arit. Reviewed-by: Roland Scheidegger <[email protected]>
* gallivm: Add debug option to force SSE2.Jose Fonseca2016-04-031-11/+14
| | | | | | For simulating less capable machines. Reviewed-by: Roland Scheidegger <[email protected]>
* llvmpipe: Test abs.Jose Fonseca2016-04-031-0/+1
| | | | Trivial.
* llvmpipe: Build lp_test_arit on MSVC too.Jose Fonseca2016-04-031-3/+1
| | | | | | It builds fine now. Probably due to C99 support. Trivial.
* gallivm: Fix performance regressions due to vector selects.Jose Fonseca2016-04-031-22/+18
| | | | | | | | | LLVM often can't determine the mask elements are all ones/zeros, and there doesn't seem to be a good way to hint that. Thanks to Roland Scheidegger for spotting and analyzing the issue. Reviewed-by: Roland Scheidegger <[email protected]>
* gallivm: Remove lp_build_load_volatile.Jose Fonseca2016-04-032-12/+0
| | | | | | | No longer needed. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]>
* gallivm: Use standard LLVMSetAlignment from LLVM 3.4 onwards.Jose Fonseca2016-04-039-27/+39
| | | | | | | | | Only provide a fallback for LLVM 3.3. One less dependency on LLVM C++ interface. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]>
* gm107/ir: add OP_SELP emission, used in DSQRT loweringIlia Mirkin2016-04-021-0/+30
| | | | | | | | The current DSQRT lowering code emits an OP_SELP, so we have to handle its emission. This will eventually go away, but no harm supporting this op. Signed-off-by: Ilia Mirkin <[email protected]>
* nv50/ir: we can't load local memory directly into an outputIlia Mirkin2016-04-021-1/+2
| | | | | | | | | | | This fixes piglit tests like tests/spec/glsl-1.10/execution/variable-indexing/vs-output-array-float-index-wr.shader_test and related ones. Signed-off-by: Ilia Mirkin <[email protected]> Cc: "11.1 11.2" <[email protected]>
* nv50/ir: fix envyas variants when building the code libSamuel Pitoiset2016-04-021-2/+2
| | | | | | | nvc0 and nve4 have been respectively replaced by gf100 and gk104. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* svga: remove unused svga_compile_key::texture_msaa fieldBrian Paul2016-04-022-2/+0
| | | | Reviewed-by: Jose Fonseca <[email protected]>
* svga: check TXF instruction's target to determine MSAABrian Paul2016-04-021-1/+1
| | | | | | | | | | Rather than the currently bound texture. This goes along with the earlier patch to get away from examining bound textures and sampler views during shader translation. Fixes VMware bug 1632739. Reviewed-by: Jose Fonseca <[email protected]>
* tgsi: add simple tgsi_is_msaa_target() helperBrian Paul2016-04-021-0/+8
| | | | Reviewed-by: Jose Fonseca <[email protected]>
* gallium: distinguish between shader IR in get_compute_paramBas Nieuwenhuizen2016-04-0211-33/+47
| | | | | | | | | | | | | For radeonsi, native and TGSI use different compilers and this results in different limits for different IR's. The set we strictly need for radeonsi is only the MAX_BLOCK_SIZE and MAX_THREADS_PER_BLOCK params, but I added a few others as shader related that seemed like they would also typically depend on the compiler. Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Dave Airlie <[email protected]>
* gallium: add global buffer memory barrier bitBas Nieuwenhuizen2016-04-022-0/+3
| | | | | | | | | Currently radeonsi synchronizes after every dispatch and Clover does nothing to synchronize. This is overzealous, especially with GL compute, so add a barrier for global buffers. Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Dave Airlie <[email protected]>
* gallium: add threads per block TGSI propertyBas Nieuwenhuizen2016-04-023-1/+13
| | | | | | | | | | The value 0 for unknown has been chosen to so that drivers using tgsi_scan_shader do not need to detect missing properties if they zero-initialize the struct. Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]> Reviewed-by: Dave Airlie <[email protected]>
* gallium: add compute shader IR typeBas Nieuwenhuizen2016-04-024-1/+6
| | | | | | Signed-off-by: Bas Nieuwenhuizen <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]> Reviewed-by: Dave Airlie <[email protected]>
* nvc0: enable compute shaders on GK104 and GM107+Samuel Pitoiset2016-04-011-1/+2
| | | | | | | | | Compute support on GK110 is still unstable for weird reasons, but this can be fixed later as the NVF0_COMPUTE envvar prevent using compute. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: bump the maximum number of UBOs for compute on KeplerSamuel Pitoiset2016-04-012-3/+0
| | | | | | | | The maximum number of uniform blocks (MAX_COMPUTE_UNIFORM_BLOCKS) per compute program must be at least 12. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0/ir: do not lower shared+atomics on GM107+Samuel Pitoiset2016-04-011-7/+10
| | | | | | | | For Maxwell, the ATOMS instruction can be used to perform atomic operations on shared memory instead of this load/store lowering pass. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0/ir: add atomics support on shared memory for KeplerSamuel Pitoiset2016-04-012-1/+108
| | | | | Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0/ir: fix wrong pred emission for ld lock on GK104Samuel Pitoiset2016-04-011-1/+4
| | | | | | | | This fixes 84b9b8f (nvc0/ir: add missing emission of locked load predicate). Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0/ir: add support for compute UBOs on KeplerSamuel Pitoiset2016-04-012-1/+57
| | | | | | | | Make sure to avoid out of bounds access in presence of indirect array indexing by loading the size from the driver constant buffer. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: add indirect compute support on KeplerSamuel Pitoiset2016-04-011-34/+77
| | | | | | | | | | The grid size is stored as three 32-bits integers in the indirect buffer but the launch descriptor uses a 32-bits integer for both griddim_y and griddim_z like this (z << 16) | y. To make it work, the 16 high bits of griddim_y are overwritten by griddim_z. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: reduce likelihood of collision for real buffers on KeplerSamuel Pitoiset2016-04-011-2/+2
| | | | | | | | | | | Reduce likelihood of collision with real buffers by placing the hole at the top of the 4G area. This fixes some indirect draw+compute tests with large buffers. Suggested by Ilia Mirkin. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: store ubo info to the driver constbuf on KeplerSamuel Pitoiset2016-04-014-1/+30
| | | | | Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: bind user uniforms for compute on KeplerSamuel Pitoiset2016-04-012-27/+55
| | | | | | | | | | | Uniform buffer objects will be sticked to the driver constant buffer like buffers because the launch descriptor only allows 8 CBs. Input kernel parameters for OpenCL are still uploaded to screen->parm which is bound on c0, but this will be changed later with a new series. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: bind shader buffers for compute on KeplerSamuel Pitoiset2016-04-012-3/+39
| | | | | Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* nvc0: bind driver cb for compute on c7[] for KeplerSamuel Pitoiset2016-04-014-45/+37
| | | | | | | | | | | | Instead of using the screen->parm buffer object which will be removed, upload auxiliary constants to uniform_bo to be consistent regarding what we already do for Fermi. This breaks surfaces support (for compute only) but this will be properly re-introduced later for ARB_shader_image_load_store. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* gallivm: Prevent disassembly debug output from being truncated.Jose Fonseca2016-04-011-9/+9
| | | | | | | | | | | By using os_log_message directly, as _debug_vprintf truncates messages to 4K. Also cleanup the disassemble interface. Spotted by Roland. Trivial.
* radeonsi: use util_strchrnul() to fix android build errorMauro Rossi2016-04-011-1/+2
| | | | | | | | | | | | | | Android Bionic does not support strchrnul() string function, gallium auxiliary util/u_string.h provides util_strchrnul() This change avoids the following building error: external/mesa/src/gallium/drivers/radeonsi/si_shader.c:3863: error: undefined reference to 'strchrnul' collect2: error: ld returned 1 exit status Cc: [email protected] Signed-off-by: Emil Velikov <[email protected]>
* gallivm: Use vector selects on LLVM 3.3+.Jose Fonseca2016-04-011-3/+5
| | | | | | | | | | | | This is an old patch I had around. Vector selects seem to work well from LLVM 3.3. Using them should improve code quality, as it might make constant propagation pass more effective. Tested lp_test_* Reviewed-by: Roland Scheidegger <[email protected]>
* nv50,nvc0: add PIPE_BIND_LINEAR support to is_format_supportedIlia Mirkin2016-03-312-0/+18
| | | | | | | vdpau has recently come to rely on this, so make sure to check it properly. Signed-off-by: Ilia Mirkin <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-29 05:30:19 +0000