diff options
author | Tom Stellard <[email protected]> | 2012-01-06 17:38:37 -0500 |
---|---|---|
committer | Tom Stellard <[email protected]> | 2012-04-13 10:32:06 -0400 |
commit | a75c6163e605f35b14f26930dd9227e4f337ec9e (patch) | |
tree | 0263219cbab9282896f874060bb03d445c4de891 /src/gallium/drivers/radeon/AMDILIntrinsics.td | |
parent | e55cf4854d594eae9ac3f6abd24f4e616eea894f (diff) |
radeonsi: initial WIP SI code
This commit adds initial support for acceleration
on SI chips. egltri is starting to work.
The SI/R600 llvm backend is currently included in mesa
but that may change in the future.
The plan is to write a single gallium driver and
use gallium to support X acceleration.
This commit contains patches from:
Tom Stellard <[email protected]>
Michel Dänzer <[email protected]>
Alex Deucher <[email protected]>
Vadim Girlin <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
The following commits were squashed in:
======================================================================
radeonsi: Remove unused winsys pointer
This was removed from r600g in commit:
commit 96d882939d612fcc8332f107befec470ed4359de
Author: Marek Olšák <[email protected]>
Date: Fri Feb 17 01:49:49 2012 +0100
gallium: remove unused winsys pointers in pipe_screen and pipe_context
A winsys is already a private object of a driver.
======================================================================
radeonsi: Copy color clamping CAPs from r600
Not sure if the values of these CAPS are correct for radeonsi, but the
same changed were made to r600g in commit:
commit bc1c8369384b5e16547c5bf9728aa78f8dfd66cc
Author: Marek Olšák <[email protected]>
Date: Mon Jan 23 03:11:17 2012 +0100
st/mesa: do vertex and fragment color clamping in shaders
For ARB_color_buffer_float. Most hardware can't do it and st/mesa is
the perfect place for a fallback.
The exceptions are:
- r500 (vertex clamp only)
- nv50 (both)
- nvc0 (both)
- softpipe (both)
We also have to take into account that r300 can do CLAMPED vertex colors only,
while r600 can do UNCLAMPED vertex colors only. The difference can be expressed
with the two new CAPs.
======================================================================
radeonsi: Remove PIPE_CAP_OUTPUT_READ
This CAP was dropped in commit:
commit 04e324008759282728a95a1394bac2c4c2a1a3f9
Author: Marek Olšák <[email protected]>
Date: Thu Feb 23 23:44:36 2012 +0100
gallium: remove PIPE_SHADER_CAP_OUTPUT_READ
r600g is the only driver which has made use of it. The reason the CAP was
added was to fix some piglit tests when the GLSL pass lower_output_reads
didn't exist.
However, not removing output reads breaks the fallback for glClampColorARB,
which assumes outputs are not readable. The fix would be non-trivial
and my personal preference is to remove the CAP, considering that reading
outputs is uncommon and that we can now use lower_output_reads to fix
the issue that the CAP was supposed to workaround in the first place.
======================================================================
radeonsi: Add missing parameters to rws->buffer_get_tiling() call
This was changed in commit:
commit c0c979eebc076b95cc8d18a013ce2968fe6311ad
Author: Jerome Glisse <[email protected]>
Date: Mon Jan 30 17:22:13 2012 -0500
r600g: add support for common surface allocator for tiling v13
Tiled surface have all kind of alignment constraint that needs to
be met. Instead of having all this code duplicated btw ddx and
mesa use common code in libdrm_radeon this also ensure that both
ddx and mesa compute those alignment in the same way.
v2 fix evergreen
v3 fix compressed texture and workaround cube texture issue by
disabling 2D array mode for cubemap (need to check if r7xx and
newer are also affected by the issue)
v4 fix texture array
v5 fix evergreen and newer, split surface values computation from
mipmap tree generation so that we can get them directly from the
ddx
v6 final fix to evergreen tile split value
v7 fix mipmap offset to avoid to use random value, use color view
depth view to address different layer as hardware is doing some
magic rotation depending on the layer
v8 fix COLOR_VIEW on r6xx for linear array mode, use COLOR_VIEW on
evergreen, align bytes per pixel to a multiple of a dword
v9 fix handling of stencil on evergreen, half fix for compressed
texture
v10 fix evergreen compressed texture proper support for stencil
tile split. Fix stencil issue when array mode was clear by
the kernel, always program stencil bo. On evergreen depth
buffer bo need to be big enough to hold depth buffer + stencil
buffer as even with stencil disabled things get written there.
v11 rebase on top of mesa, fix pitch issue with 1d surface on evergreen,
old ddx overestimate those. Fix linear case when pitch*height < 64.
Fix r300g.
v12 Fix linear case when pitch*height < 64 for old path, adapt to
libdrm API change
v13 add libdrm check
Signed-off-by: Jerome Glisse <[email protected]>
======================================================================
radeonsi: Remove PIPE_TRANSFER_MAP_PERMANENTLY
This was removed in commit:
commit 62f44f670bb0162e89fd4786af877f8da9ff607c
Author: Marek Olšák <[email protected]>
Date: Mon Mar 5 13:45:00 2012 +0100
Revert "gallium: add flag PIPE_TRANSFER_MAP_PERMANENTLY"
This reverts commit 0950086376b1c8b7fb89eda81ed7f2f06dee58bc.
It was decided to refactor the transfer API instead of adding workarounds
to address the performance issues.
======================================================================
radeonsi: Handle PIPE_VIDEO_CAP_PREFERED_FORMAT.
Reintroduced in commit 9d9afcb5bac2931d4b8e6d1aa571e941c5110c90.
======================================================================
radeonsi: nuke the fallback for vertex and fragment color clamping
Ported from r600g commit c2b800cf38b299c1ab1c53dc0e4ea00c7acef853.
======================================================================
radeonsi: don't expose transform_feedback2 without kernel support
Ported from r600g commit 15146fd1bcbb08e44a1cbb984440ee1a5de63d48.
======================================================================
radeonsi: Handle PIPE_CAP_GLSL_FEATURE_LEVEL.
Ported from r600g part of commit 171be755223d99f8cc5cc1bdaf8bd7b4caa04b4f.
======================================================================
radeonsi: set minimum point size to 1.0 for non-sprite non-aa points.
Ported from r600g commit f183cc9ce3ad1d043bdf8b38fd519e8f437714fc.
======================================================================
radeonsi: rework and consolidate stencilref state setting.
Ported from r600g commit a2361946e782b57f0c63587841ca41c0ea707070.
======================================================================
radeonsi: cleanup setting DB_SHADER_CONTROL.
Ported from r600g commit 3d061caaed13b646ff40754f8ebe73f3d4983c5b.
======================================================================
radeonsi: Get rid of register masks.
Ported from r600g commits
3d061caaed13b646ff40754f8ebe73f3d4983c5b..9344ab382a1765c1a7c2560e771485edf4954fe2.
======================================================================
radeonsi: get rid of r600_context_reg.
Ported from r600g commits
9344ab382a1765c1a7c2560e771485edf4954fe2..bed20f02a771f43e1c5092254705701c228cfa7f.
======================================================================
radeonsi: Fix regression from 'Get rid of register masks'.
======================================================================
radeonsi: optimize r600_resource_va.
Ported from r600g commit 669d8766ff3403938794eb80d7769347b6e52174.
======================================================================
radeonsi: remove u8,u16,u32,u64 types.
Ported from r600g commit 78293b99b23268e6698f1267aaf40647c17d95a5.
======================================================================
radeonsi: merge r600_context with r600_pipe_context.
Ported from r600g commit e4340c1908a6a3b09e1a15d5195f6da7d00494d0.
======================================================================
radeonsi: Miscellaneous context cleanups.
Ported from r600g commits
e4340c1908a6a3b09e1a15d5195f6da7d00494d0..621e0db71c5ddcb379171064a4f720c9cf01e888.
======================================================================
radeonsi: add a new simple API for state emission.
Ported from r600g commits
621e0db71c5ddcb379171064a4f720c9cf01e888..f661405637bba32c2cfbeecf6e2e56e414e9521e.
======================================================================
radeonsi: Also remove sbu_flags member of struct r600_reg.
Requires using sid.h instead of r600d.h for the new CP_COHER_CNTL definitions,
so some code needs to be disabled for now.
======================================================================
radeonsi: Miscellaneous simplifications.
Ported from r600g commits 38bf2763482b4f1b6d95cd51aecec75601d8b90f and
b0337b679ad4c2feae59215104cfa60b58a619d5.
======================================================================
radeonsi: Handle PIPE_CAP_QUADS_FOLLOW_PROVOKING_VERTEX_CONVENTION.
Ported from commit 8b4f7b0672d663273310fffa9490ad996f5b914a.
======================================================================
radeonsi: Use a fake reloc to sleep for fences.
Ported from r600g commit 8cd03b933cf868ff867e2db4a0937005a02fd0e4.
======================================================================
radeonsi: adapt to get_query_result interface change.
Ported from r600g commit 4445e170bee23a3607ece0e010adef7058ac6a11.
Diffstat (limited to 'src/gallium/drivers/radeon/AMDILIntrinsics.td')
-rw-r--r-- | src/gallium/drivers/radeon/AMDILIntrinsics.td | 705 |
1 files changed, 705 insertions, 0 deletions
diff --git a/src/gallium/drivers/radeon/AMDILIntrinsics.td b/src/gallium/drivers/radeon/AMDILIntrinsics.td new file mode 100644 index 00000000000..ef361f4e6af --- /dev/null +++ b/src/gallium/drivers/radeon/AMDILIntrinsics.td @@ -0,0 +1,705 @@ +//===- AMDILIntrinsics.td - Defines AMDIL Intrinscs -*- tablegen -*-===// +// +// The LLVM Compiler Infrastructure +// +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. +// +//==-----------------------------------------------------------------------===// +// +// This file defines all of the amdil-specific intrinsics +// +//===---------------------------------------------------------------===// + +let TargetPrefix = "AMDIL", isTarget = 1 in { +//------------- Synchronization Functions - OpenCL 6.11.9 --------------------// + def int_AMDIL_fence : GCCBuiltin<"mem_fence">, + UnaryIntNoRetInt; + def int_AMDIL_fence_global : GCCBuiltin<"mem_fence_global">, + UnaryIntNoRetInt; + def int_AMDIL_fence_local : GCCBuiltin<"mem_fence_local">, + UnaryIntNoRetInt; + def int_AMDIL_fence_region : GCCBuiltin<"mem_fence_region">, + UnaryIntNoRetInt; + def int_AMDIL_fence_read_only : GCCBuiltin<"read_mem_fence">, + UnaryIntNoRetInt; + def int_AMDIL_fence_read_only_global : GCCBuiltin<"read_mem_fence_global">, + UnaryIntNoRetInt; + def int_AMDIL_fence_read_only_local : GCCBuiltin<"read_mem_fence_local">, + UnaryIntNoRetInt; + def int_AMDIL_fence_read_only_region : GCCBuiltin<"read_mem_fence_region">, + UnaryIntNoRetInt; + def int_AMDIL_fence_write_only : GCCBuiltin<"write_mem_fence">, + UnaryIntNoRetInt; + def int_AMDIL_fence_write_only_global : GCCBuiltin<"write_mem_fence_global">, + UnaryIntNoRetInt; + def int_AMDIL_fence_write_only_local : GCCBuiltin<"write_mem_fence_local">, + UnaryIntNoRetInt; + def int_AMDIL_fence_write_only_region : GCCBuiltin<"write_mem_fence_region">, + UnaryIntNoRetInt; + + def int_AMDIL_early_exit : GCCBuiltin<"__amdil_early_exit">, + UnaryIntNoRetInt; + + def int_AMDIL_cmov_logical : GCCBuiltin<"__amdil_cmov_logical">, + TernaryIntInt; + def int_AMDIL_fabs : GCCBuiltin<"__amdil_fabs">, UnaryIntFloat; + def int_AMDIL_abs : GCCBuiltin<"__amdil_abs">, UnaryIntInt; + + def int_AMDIL_bit_extract_i32 : GCCBuiltin<"__amdil_ibit_extract">, + TernaryIntInt; + def int_AMDIL_bit_extract_u32 : GCCBuiltin<"__amdil_ubit_extract">, + TernaryIntInt; + def int_AMDIL_bit_reverse_u32 : GCCBuiltin<"__amdil_ubit_reverse">, + UnaryIntInt; + def int_AMDIL_bit_count_i32 : GCCBuiltin<"__amdil_count_bits">, + UnaryIntInt; + def int_AMDIL_bit_find_first_lo : GCCBuiltin<"__amdil_ffb_lo">, + UnaryIntInt; + def int_AMDIL_bit_find_first_hi : GCCBuiltin<"__amdil_ffb_hi">, + UnaryIntInt; + def int_AMDIL_bit_find_first_sgn : GCCBuiltin<"__amdil_ffb_signed">, + UnaryIntInt; + def int_AMDIL_media_bitalign : GCCBuiltin<"__amdil_bitalign">, + TernaryIntInt; + def int_AMDIL_media_bytealign : GCCBuiltin<"__amdil_bytealign">, + TernaryIntInt; + def int_AMDIL_bit_insert_u32 : GCCBuiltin<"__amdil_ubit_insert">, + QuaternaryIntInt; + def int_AMDIL_bfi : GCCBuiltin<"__amdil_bfi">, + TernaryIntInt; + def int_AMDIL_bfm : GCCBuiltin<"__amdil_bfm">, + BinaryIntInt; + def int_AMDIL_mad_i32 : GCCBuiltin<"__amdil_imad">, + TernaryIntInt; + def int_AMDIL_mad_u32 : GCCBuiltin<"__amdil_umad">, + TernaryIntInt; + def int_AMDIL_mad : GCCBuiltin<"__amdil_mad">, + TernaryIntFloat; + def int_AMDIL_mulhi_i32 : GCCBuiltin<"__amdil_imul_high">, + BinaryIntInt; + def int_AMDIL_mulhi_u32 : GCCBuiltin<"__amdil_umul_high">, + BinaryIntInt; + def int_AMDIL_mul24_i32 : GCCBuiltin<"__amdil_imul24">, + BinaryIntInt; + def int_AMDIL_mul24_u32 : GCCBuiltin<"__amdil_umul24">, + BinaryIntInt; + def int_AMDIL_mulhi24_i32 : GCCBuiltin<"__amdil_imul24_high">, + BinaryIntInt; + def int_AMDIL_mulhi24_u32 : GCCBuiltin<"__amdil_umul24_high">, + BinaryIntInt; + def int_AMDIL_mad24_i32 : GCCBuiltin<"__amdil_imad24">, + TernaryIntInt; + def int_AMDIL_mad24_u32 : GCCBuiltin<"__amdil_umad24">, + TernaryIntInt; + def int_AMDIL_carry_i32 : GCCBuiltin<"__amdil_carry">, + BinaryIntInt; + def int_AMDIL_borrow_i32 : GCCBuiltin<"__amdil_borrow">, + BinaryIntInt; + def int_AMDIL_min_i32 : GCCBuiltin<"__amdil_imin">, + BinaryIntInt; + def int_AMDIL_min_u32 : GCCBuiltin<"__amdil_umin">, + BinaryIntInt; + def int_AMDIL_min : GCCBuiltin<"__amdil_min">, + BinaryIntFloat; + def int_AMDIL_max_i32 : GCCBuiltin<"__amdil_imax">, + BinaryIntInt; + def int_AMDIL_max_u32 : GCCBuiltin<"__amdil_umax">, + BinaryIntInt; + def int_AMDIL_max : GCCBuiltin<"__amdil_max">, + BinaryIntFloat; + def int_AMDIL_media_lerp_u4 : GCCBuiltin<"__amdil_u4lerp">, + TernaryIntInt; + def int_AMDIL_media_sad : GCCBuiltin<"__amdil_sad">, + TernaryIntInt; + def int_AMDIL_media_sad_hi : GCCBuiltin<"__amdil_sadhi">, + TernaryIntInt; + def int_AMDIL_fraction : GCCBuiltin<"__amdil_fraction">, + UnaryIntFloat; + def int_AMDIL_clamp : GCCBuiltin<"__amdil_clamp">, + TernaryIntFloat; + def int_AMDIL_pireduce : GCCBuiltin<"__amdil_pireduce">, + UnaryIntFloat; + def int_AMDIL_round_nearest : GCCBuiltin<"__amdil_round_nearest">, + UnaryIntFloat; + def int_AMDIL_round_neginf : GCCBuiltin<"__amdil_round_neginf">, + UnaryIntFloat; + def int_AMDIL_round_posinf : GCCBuiltin<"__amdil_round_posinf">, + UnaryIntFloat; + def int_AMDIL_round_zero : GCCBuiltin<"__amdil_round_zero">, + UnaryIntFloat; + def int_AMDIL_acos : GCCBuiltin<"__amdil_acos">, + UnaryIntFloat; + def int_AMDIL_atan : GCCBuiltin<"__amdil_atan">, + UnaryIntFloat; + def int_AMDIL_asin : GCCBuiltin<"__amdil_asin">, + UnaryIntFloat; + def int_AMDIL_cos : GCCBuiltin<"__amdil_cos">, + UnaryIntFloat; + def int_AMDIL_cos_vec : GCCBuiltin<"__amdil_cos_vec">, + UnaryIntFloat; + def int_AMDIL_tan : GCCBuiltin<"__amdil_tan">, + UnaryIntFloat; + def int_AMDIL_sin : GCCBuiltin<"__amdil_sin">, + UnaryIntFloat; + def int_AMDIL_sin_vec : GCCBuiltin<"__amdil_sin_vec">, + UnaryIntFloat; + def int_AMDIL_pow : GCCBuiltin<"__amdil_pow">, BinaryIntFloat; + def int_AMDIL_div : GCCBuiltin<"__amdil_div">, BinaryIntFloat; + def int_AMDIL_udiv : GCCBuiltin<"__amdil_udiv">, BinaryIntInt; + def int_AMDIL_sqrt: GCCBuiltin<"__amdil_sqrt">, + UnaryIntFloat; + def int_AMDIL_sqrt_vec: GCCBuiltin<"__amdil_sqrt_vec">, + UnaryIntFloat; + def int_AMDIL_exp : GCCBuiltin<"__amdil_exp">, + UnaryIntFloat; + def int_AMDIL_exp_vec : GCCBuiltin<"__amdil_exp_vec">, + UnaryIntFloat; + def int_AMDIL_exn : GCCBuiltin<"__amdil_exn">, + UnaryIntFloat; + def int_AMDIL_log : GCCBuiltin<"__amdil_log">, + UnaryIntFloat; + def int_AMDIL_log_vec : GCCBuiltin<"__amdil_log_vec">, + UnaryIntFloat; + def int_AMDIL_ln : GCCBuiltin<"__amdil_ln">, + UnaryIntFloat; + def int_AMDIL_sign: GCCBuiltin<"__amdil_sign">, + UnaryIntFloat; + def int_AMDIL_fma: GCCBuiltin<"__amdil_fma">, + TernaryIntFloat; + def int_AMDIL_rsq : GCCBuiltin<"__amdil_rsq">, + UnaryIntFloat; + def int_AMDIL_rsq_vec : GCCBuiltin<"__amdil_rsq_vec">, + UnaryIntFloat; + def int_AMDIL_length : GCCBuiltin<"__amdil_length">, + UnaryIntFloat; + def int_AMDIL_lerp : GCCBuiltin<"__amdil_lerp">, + TernaryIntFloat; + def int_AMDIL_media_sad4 : GCCBuiltin<"__amdil_sad4">, + Intrinsic<[llvm_i32_ty], [llvm_v4i32_ty, + llvm_v4i32_ty, llvm_i32_ty], []>; + + def int_AMDIL_frexp_f64 : GCCBuiltin<"__amdil_frexp">, + Intrinsic<[llvm_v2i64_ty], [llvm_double_ty], []>; + def int_AMDIL_ldexp : GCCBuiltin<"__amdil_ldexp">, + Intrinsic<[llvm_anyfloat_ty], [llvm_anyfloat_ty, llvm_anyint_ty], []>; + def int_AMDIL_drcp : GCCBuiltin<"__amdil_rcp">, + Intrinsic<[llvm_double_ty], [llvm_double_ty], []>; + def int_AMDIL_convert_f16_f32 : GCCBuiltin<"__amdil_half_to_float">, + ConvertIntITOF; + def int_AMDIL_convert_f32_f16 : GCCBuiltin<"__amdil_float_to_half">, + ConvertIntFTOI; + def int_AMDIL_convert_f32_i32_rpi : GCCBuiltin<"__amdil_float_to_int_rpi">, + ConvertIntFTOI; + def int_AMDIL_convert_f32_i32_flr : GCCBuiltin<"__amdil_float_to_int_flr">, + ConvertIntFTOI; + def int_AMDIL_convert_f32_f16_near : GCCBuiltin<"__amdil_float_to_half_near">, + ConvertIntFTOI; + def int_AMDIL_convert_f32_f16_neg_inf : GCCBuiltin<"__amdil_float_to_half_neg_inf">, + ConvertIntFTOI; + def int_AMDIL_convert_f32_f16_plus_inf : GCCBuiltin<"__amdil_float_to_half_plus_inf">, + ConvertIntFTOI; + def int_AMDIL_media_convert_f2v4u8 : GCCBuiltin<"__amdil_f_2_u4">, + Intrinsic<[llvm_i32_ty], [llvm_v4f32_ty], []>; + def int_AMDIL_media_unpack_byte_0 : GCCBuiltin<"__amdil_unpack_0">, + ConvertIntITOF; + def int_AMDIL_media_unpack_byte_1 : GCCBuiltin<"__amdil_unpack_1">, + ConvertIntITOF; + def int_AMDIL_media_unpack_byte_2 : GCCBuiltin<"__amdil_unpack_2">, + ConvertIntITOF; + def int_AMDIL_media_unpack_byte_3 : GCCBuiltin<"__amdil_unpack_3">, + ConvertIntITOF; + def int_AMDIL_dp2_add : GCCBuiltin<"__amdil_dp2_add">, + Intrinsic<[llvm_float_ty], [llvm_v2f32_ty, + llvm_v2f32_ty, llvm_float_ty], []>; + def int_AMDIL_dp2 : GCCBuiltin<"__amdil_dp2">, + Intrinsic<[llvm_float_ty], [llvm_v2f32_ty, + llvm_v2f32_ty], []>; + def int_AMDIL_dp3 : GCCBuiltin<"__amdil_dp3">, + Intrinsic<[llvm_float_ty], [llvm_v4f32_ty, + llvm_v4f32_ty], []>; + def int_AMDIL_dp4 : GCCBuiltin<"__amdil_dp4">, + Intrinsic<[llvm_float_ty], [llvm_v4f32_ty, + llvm_v4f32_ty], []>; +//===---------------------- Image functions begin ------------------------===// + def int_AMDIL_image1d_write : GCCBuiltin<"__amdil_image1d_write">, + Intrinsic<[], [llvm_ptr_ty, llvm_v2i32_ty, llvm_v4i32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_read_norm : GCCBuiltin<"__amdil_image1d_read_norm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_read_unnorm : GCCBuiltin<"__amdil_image1d_read_unnorm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_info0 : GCCBuiltin<"__amdil_image1d_info0">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image1d_info1 : GCCBuiltin<"__amdil_image1d_info1">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image1d_array_write : GCCBuiltin<"__amdil_image1d_array_write">, + Intrinsic<[], [llvm_ptr_ty, llvm_v2i32_ty, llvm_v4i32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_array_read_norm : GCCBuiltin<"__amdil_image1d_array_read_norm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_array_read_unnorm : GCCBuiltin<"__amdil_image1d_array_read_unnorm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image1d_array_info0 : GCCBuiltin<"__amdil_image1d_array_info0">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image1d_array_info1 : GCCBuiltin<"__amdil_image1d_array_info1">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image2d_write : GCCBuiltin<"__amdil_image2d_write">, + Intrinsic<[], [llvm_ptr_ty, llvm_v2i32_ty, llvm_v4i32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_read_norm : GCCBuiltin<"__amdil_image2d_read_norm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_read_unnorm : GCCBuiltin<"__amdil_image2d_read_unnorm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_info0 : GCCBuiltin<"__amdil_image2d_info0">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image2d_info1 : GCCBuiltin<"__amdil_image2d_info1">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image2d_array_write : GCCBuiltin<"__amdil_image2d_array_write">, + Intrinsic<[], [llvm_ptr_ty, llvm_v2i32_ty, llvm_v4i32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_array_read_norm : GCCBuiltin<"__amdil_image2d_array_read_norm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_array_read_unnorm : GCCBuiltin<"__amdil_image2d_array_read_unnorm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image2d_array_info0 : GCCBuiltin<"__amdil_image2d_array_info0">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image2d_array_info1 : GCCBuiltin<"__amdil_image2d_array_info1">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image3d_write : GCCBuiltin<"__amdil_image3d_write">, + Intrinsic<[], [llvm_ptr_ty, llvm_v4i32_ty, llvm_v4i32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image3d_read_norm : GCCBuiltin<"__amdil_image3d_read_norm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image3d_read_unnorm : GCCBuiltin<"__amdil_image3d_read_unnorm">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty, llvm_i32_ty, llvm_v4f32_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_image3d_info0 : GCCBuiltin<"__amdil_image3d_info0">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + + def int_AMDIL_image3d_info1 : GCCBuiltin<"__amdil_image3d_info1">, + Intrinsic<[llvm_v4i32_ty], [llvm_ptr_ty], []>; + +//===---------------------- Image functions end --------------------------===// + + def int_AMDIL_append_alloc_i32 : GCCBuiltin<"__amdil_append_alloc">, + Intrinsic<[llvm_i32_ty], [llvm_ptr_ty], [IntrReadWriteArgMem]>; + def int_AMDIL_append_consume_i32 : GCCBuiltin<"__amdil_append_consume">, + Intrinsic<[llvm_i32_ty], [llvm_ptr_ty], [IntrReadWriteArgMem]>; + def int_AMDIL_append_alloc_i32_noret : GCCBuiltin<"__amdil_append_alloc_noret">, + Intrinsic<[llvm_i32_ty], [llvm_ptr_ty], [IntrReadWriteArgMem]>; + def int_AMDIL_append_consume_i32_noret : GCCBuiltin<"__amdil_append_consume_noret">, + Intrinsic<[llvm_i32_ty], [llvm_ptr_ty], [IntrReadWriteArgMem]>; + + def int_AMDIL_get_global_id : GCCBuiltin<"__amdil_get_global_id_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_local_id : GCCBuiltin<"__amdil_get_local_id_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_group_id : GCCBuiltin<"__amdil_get_group_id_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_num_groups : GCCBuiltin<"__amdil_get_num_groups_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_local_size : GCCBuiltin<"__amdil_get_local_size_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_global_size : GCCBuiltin<"__amdil_get_global_size_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_global_offset : GCCBuiltin<"__amdil_get_global_offset_int">, + Intrinsic<[llvm_v4i32_ty], [], []>; + def int_AMDIL_get_work_dim : GCCBuiltin<"get_work_dim">, + Intrinsic<[llvm_i32_ty], [], []>; + def int_AMDIL_get_printf_offset : GCCBuiltin<"__amdil_get_printf_offset">, + Intrinsic<[llvm_i32_ty], []>; + def int_AMDIL_get_printf_size : GCCBuiltin<"__amdil_get_printf_size">, + Intrinsic<[llvm_i32_ty], []>; + +/// Intrinsics for atomic instructions with no return value +/// Signed 32 bit integer atomics for global address space +def int_AMDIL_atomic_add_gi32_noret : GCCBuiltin<"__atomic_add_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_gi32_noret : GCCBuiltin<"__atomic_sub_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_gi32_noret : GCCBuiltin<"__atomic_rsub_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_gi32_noret : GCCBuiltin<"__atomic_xchg_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_gi32_noret : GCCBuiltin<"__atomic_inc_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_gi32_noret : GCCBuiltin<"__atomic_dec_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_gi32_noret : GCCBuiltin<"__atomic_cmpxchg_gi32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_gi32_noret : GCCBuiltin<"__atomic_min_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_gi32_noret : GCCBuiltin<"__atomic_max_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_gi32_noret : GCCBuiltin<"__atomic_and_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_gi32_noret : GCCBuiltin<"__atomic_or_gi32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_gi32_noret : GCCBuiltin<"__atomic_xor_gi32_noret">, + BinaryAtomicIntNoRet; + + + +/// Unsigned 32 bit integer atomics for global address space +def int_AMDIL_atomic_add_gu32_noret : GCCBuiltin<"__atomic_add_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_gu32_noret : GCCBuiltin<"__atomic_sub_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_gu32_noret : GCCBuiltin<"__atomic_rsub_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_gu32_noret : GCCBuiltin<"__atomic_xchg_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_gu32_noret : GCCBuiltin<"__atomic_inc_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_gu32_noret : GCCBuiltin<"__atomic_dec_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_gu32_noret : GCCBuiltin<"__atomic_cmpxchg_gu32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_gu32_noret : GCCBuiltin<"__atomic_min_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_gu32_noret : GCCBuiltin<"__atomic_max_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_gu32_noret : GCCBuiltin<"__atomic_and_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_gu32_noret : GCCBuiltin<"__atomic_or_gu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_gu32_noret : GCCBuiltin<"__atomic_xor_gu32_noret">, + BinaryAtomicIntNoRet; + + +/// Intrinsics for atomic instructions with a return value +/// Signed 32 bit integer atomics for global address space +def int_AMDIL_atomic_add_gi32 : GCCBuiltin<"__atomic_add_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_gi32 : GCCBuiltin<"__atomic_sub_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_gi32 : GCCBuiltin<"__atomic_rsub_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_gi32 : GCCBuiltin<"__atomic_xchg_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_gi32 : GCCBuiltin<"__atomic_inc_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_gi32 : GCCBuiltin<"__atomic_dec_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_gi32 : GCCBuiltin<"__atomic_cmpxchg_gi32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_gi32 : GCCBuiltin<"__atomic_min_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_gi32 : GCCBuiltin<"__atomic_max_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_gi32 : GCCBuiltin<"__atomic_and_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_gi32 : GCCBuiltin<"__atomic_or_gi32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xor_gi32 : GCCBuiltin<"__atomic_xor_gi32">, + BinaryAtomicInt; + +/// 32 bit float atomics required by OpenCL +def int_AMDIL_atomic_xchg_gf32 : GCCBuiltin<"__atomic_xchg_gf32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_gf32_noret : GCCBuiltin<"__atomic_xchg_gf32_noret">, + BinaryAtomicIntNoRet; + +/// Unsigned 32 bit integer atomics for global address space +def int_AMDIL_atomic_add_gu32 : GCCBuiltin<"__atomic_add_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_gu32 : GCCBuiltin<"__atomic_sub_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_gu32 : GCCBuiltin<"__atomic_rsub_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_gu32 : GCCBuiltin<"__atomic_xchg_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_gu32 : GCCBuiltin<"__atomic_inc_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_gu32 : GCCBuiltin<"__atomic_dec_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_gu32 : GCCBuiltin<"__atomic_cmpxchg_gu32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_gu32 : GCCBuiltin<"__atomic_min_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_gu32 : GCCBuiltin<"__atomic_max_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_gu32 : GCCBuiltin<"__atomic_and_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_gu32 : GCCBuiltin<"__atomic_or_gu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xor_gu32 : GCCBuiltin<"__atomic_xor_gu32">, + BinaryAtomicInt; + + +/// Intrinsics for atomic instructions with no return value +/// Signed 32 bit integer atomics for local address space +def int_AMDIL_atomic_add_li32_noret : GCCBuiltin<"__atomic_add_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_li32_noret : GCCBuiltin<"__atomic_sub_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_li32_noret : GCCBuiltin<"__atomic_rsub_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_li32_noret : GCCBuiltin<"__atomic_xchg_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_li32_noret : GCCBuiltin<"__atomic_inc_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_li32_noret : GCCBuiltin<"__atomic_dec_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_li32_noret : GCCBuiltin<"__atomic_cmpxchg_li32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_li32_noret : GCCBuiltin<"__atomic_min_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_li32_noret : GCCBuiltin<"__atomic_max_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_li32_noret : GCCBuiltin<"__atomic_and_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_li32_noret : GCCBuiltin<"__atomic_or_li32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_mskor_li32_noret : GCCBuiltin<"__atomic_mskor_li32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_li32_noret : GCCBuiltin<"__atomic_xor_li32_noret">, + BinaryAtomicIntNoRet; + +/// Signed 32 bit integer atomics for region address space +def int_AMDIL_atomic_add_ri32_noret : GCCBuiltin<"__atomic_add_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_ri32_noret : GCCBuiltin<"__atomic_sub_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_ri32_noret : GCCBuiltin<"__atomic_rsub_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_ri32_noret : GCCBuiltin<"__atomic_xchg_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_ri32_noret : GCCBuiltin<"__atomic_inc_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_ri32_noret : GCCBuiltin<"__atomic_dec_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_ri32_noret : GCCBuiltin<"__atomic_cmpxchg_ri32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_ri32_noret : GCCBuiltin<"__atomic_min_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_ri32_noret : GCCBuiltin<"__atomic_max_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_ri32_noret : GCCBuiltin<"__atomic_and_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_ri32_noret : GCCBuiltin<"__atomic_or_ri32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_mskor_ri32_noret : GCCBuiltin<"__atomic_mskor_ri32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_ri32_noret : GCCBuiltin<"__atomic_xor_ri32_noret">, + BinaryAtomicIntNoRet; + + + +/// Unsigned 32 bit integer atomics for local address space +def int_AMDIL_atomic_add_lu32_noret : GCCBuiltin<"__atomic_add_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_lu32_noret : GCCBuiltin<"__atomic_sub_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_lu32_noret : GCCBuiltin<"__atomic_rsub_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_lu32_noret : GCCBuiltin<"__atomic_xchg_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_lu32_noret : GCCBuiltin<"__atomic_inc_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_lu32_noret : GCCBuiltin<"__atomic_dec_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_lu32_noret : GCCBuiltin<"__atomic_cmpxchg_lu32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_lu32_noret : GCCBuiltin<"__atomic_min_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_lu32_noret : GCCBuiltin<"__atomic_max_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_lu32_noret : GCCBuiltin<"__atomic_and_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_lu32_noret : GCCBuiltin<"__atomic_or_lu32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_mskor_lu32_noret : GCCBuiltin<"__atomic_mskor_lu32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_lu32_noret : GCCBuiltin<"__atomic_xor_lu32_noret">, + BinaryAtomicIntNoRet; + +/// Unsigned 32 bit integer atomics for region address space +def int_AMDIL_atomic_add_ru32_noret : GCCBuiltin<"__atomic_add_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_sub_ru32_noret : GCCBuiltin<"__atomic_sub_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_rsub_ru32_noret : GCCBuiltin<"__atomic_rsub_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_ru32_noret : GCCBuiltin<"__atomic_xchg_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_inc_ru32_noret : GCCBuiltin<"__atomic_inc_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_dec_ru32_noret : GCCBuiltin<"__atomic_dec_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_cmpxchg_ru32_noret : GCCBuiltin<"__atomic_cmpxchg_ru32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_min_ru32_noret : GCCBuiltin<"__atomic_min_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_max_ru32_noret : GCCBuiltin<"__atomic_max_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_and_ru32_noret : GCCBuiltin<"__atomic_and_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_or_ru32_noret : GCCBuiltin<"__atomic_or_ru32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_mskor_ru32_noret : GCCBuiltin<"__atomic_mskor_ru32_noret">, + TernaryAtomicIntNoRet; +def int_AMDIL_atomic_xor_ru32_noret : GCCBuiltin<"__atomic_xor_ru32_noret">, + BinaryAtomicIntNoRet; + +def int_AMDIL_get_cycle_count : GCCBuiltin<"__amdil_get_cycle_count">, + VoidIntLong; + +def int_AMDIL_compute_unit_id : GCCBuiltin<"__amdil_compute_unit_id">, + VoidIntInt; + +def int_AMDIL_wavefront_id : GCCBuiltin<"__amdil_wavefront_id">, + VoidIntInt; + + +/// Intrinsics for atomic instructions with a return value +/// Signed 32 bit integer atomics for local address space +def int_AMDIL_atomic_add_li32 : GCCBuiltin<"__atomic_add_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_li32 : GCCBuiltin<"__atomic_sub_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_li32 : GCCBuiltin<"__atomic_rsub_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_li32 : GCCBuiltin<"__atomic_xchg_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_li32 : GCCBuiltin<"__atomic_inc_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_li32 : GCCBuiltin<"__atomic_dec_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_li32 : GCCBuiltin<"__atomic_cmpxchg_li32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_li32 : GCCBuiltin<"__atomic_min_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_li32 : GCCBuiltin<"__atomic_max_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_li32 : GCCBuiltin<"__atomic_and_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_li32 : GCCBuiltin<"__atomic_or_li32">, + BinaryAtomicInt; +def int_AMDIL_atomic_mskor_li32 : GCCBuiltin<"__atomic_mskor_li32">, + TernaryAtomicInt; +def int_AMDIL_atomic_xor_li32 : GCCBuiltin<"__atomic_xor_li32">, + BinaryAtomicInt; + +/// Signed 32 bit integer atomics for region address space +def int_AMDIL_atomic_add_ri32 : GCCBuiltin<"__atomic_add_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_ri32 : GCCBuiltin<"__atomic_sub_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_ri32 : GCCBuiltin<"__atomic_rsub_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_ri32 : GCCBuiltin<"__atomic_xchg_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_ri32 : GCCBuiltin<"__atomic_inc_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_ri32 : GCCBuiltin<"__atomic_dec_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_ri32 : GCCBuiltin<"__atomic_cmpxchg_ri32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_ri32 : GCCBuiltin<"__atomic_min_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_ri32 : GCCBuiltin<"__atomic_max_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_ri32 : GCCBuiltin<"__atomic_and_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_ri32 : GCCBuiltin<"__atomic_or_ri32">, + BinaryAtomicInt; +def int_AMDIL_atomic_mskor_ri32 : GCCBuiltin<"__atomic_mskor_ri32">, + TernaryAtomicInt; +def int_AMDIL_atomic_xor_ri32 : GCCBuiltin<"__atomic_xor_ri32">, + BinaryAtomicInt; + +/// 32 bit float atomics required by OpenCL +def int_AMDIL_atomic_xchg_lf32 : GCCBuiltin<"__atomic_xchg_lf32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_lf32_noret : GCCBuiltin<"__atomic_xchg_lf32_noret">, + BinaryAtomicIntNoRet; +def int_AMDIL_atomic_xchg_rf32 : GCCBuiltin<"__atomic_xchg_rf32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_rf32_noret : GCCBuiltin<"__atomic_xchg_rf32_noret">, + BinaryAtomicIntNoRet; + +/// Unsigned 32 bit integer atomics for local address space +def int_AMDIL_atomic_add_lu32 : GCCBuiltin<"__atomic_add_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_lu32 : GCCBuiltin<"__atomic_sub_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_lu32 : GCCBuiltin<"__atomic_rsub_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_lu32 : GCCBuiltin<"__atomic_xchg_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_lu32 : GCCBuiltin<"__atomic_inc_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_lu32 : GCCBuiltin<"__atomic_dec_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_lu32 : GCCBuiltin<"__atomic_cmpxchg_lu32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_lu32 : GCCBuiltin<"__atomic_min_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_lu32 : GCCBuiltin<"__atomic_max_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_lu32 : GCCBuiltin<"__atomic_and_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_lu32 : GCCBuiltin<"__atomic_or_lu32">, + BinaryAtomicInt; +def int_AMDIL_atomic_mskor_lu32 : GCCBuiltin<"__atomic_mskor_lu32">, + TernaryAtomicInt; +def int_AMDIL_atomic_xor_lu32 : GCCBuiltin<"__atomic_xor_lu32">, + BinaryAtomicInt; + +/// Unsigned 32 bit integer atomics for region address space +def int_AMDIL_atomic_add_ru32 : GCCBuiltin<"__atomic_add_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_sub_ru32 : GCCBuiltin<"__atomic_sub_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_rsub_ru32 : GCCBuiltin<"__atomic_rsub_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_xchg_ru32 : GCCBuiltin<"__atomic_xchg_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_inc_ru32 : GCCBuiltin<"__atomic_inc_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_dec_ru32 : GCCBuiltin<"__atomic_dec_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_cmpxchg_ru32 : GCCBuiltin<"__atomic_cmpxchg_ru32">, + TernaryAtomicInt; +def int_AMDIL_atomic_min_ru32 : GCCBuiltin<"__atomic_min_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_max_ru32 : GCCBuiltin<"__atomic_max_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_and_ru32 : GCCBuiltin<"__atomic_and_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_or_ru32 : GCCBuiltin<"__atomic_or_ru32">, + BinaryAtomicInt; +def int_AMDIL_atomic_mskor_ru32 : GCCBuiltin<"__atomic_mskor_ru32">, + TernaryAtomicInt; +def int_AMDIL_atomic_xor_ru32 : GCCBuiltin<"__atomic_xor_ru32">, + BinaryAtomicInt; + +/// Semaphore signal/wait/init +def int_AMDIL_semaphore_init : GCCBuiltin<"__amdil_semaphore_init">, + Intrinsic<[], [llvm_ptr_ty, llvm_i32_ty]>; +def int_AMDIL_semaphore_wait : GCCBuiltin<"__amdil_semaphore_wait">, + Intrinsic<[], [llvm_ptr_ty]>; +def int_AMDIL_semaphore_signal : GCCBuiltin<"__amdil_semaphore_signal">, + Intrinsic<[], [llvm_ptr_ty]>; +def int_AMDIL_semaphore_size : GCCBuiltin<"__amdil_max_semaphore_size">, + Intrinsic<[llvm_i32_ty], []>; +} |