fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-28 20:48:13 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1
55693 commits 66 branches 937 tags 971 MiB
Commit graph

359 commits

Author SHA1 Message Date
Maarten Lankhorst
7c3d8301af radeon/llvm: Do not link against libgallium when building statically. 
NOTE: This is a candidate for the 9.1 branch.

Tested-by: Vincent Lejeune <vljn@ovi.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
 2013-03-19 20:20:33 +01:00
Christian König
c05483fc00 radeon/llvm: rework input fetch and output store 
Cleanup the code and implement indirect addressing.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-03-19 15:16:18 +01:00
Andreas Boll
36320bfa54 radeon/llvm: Link against libgallium.la to fix an undefined symbol 
Ported from downstream:
http://anonscm.debian.org/gitweb/?p=pkg-xorg/lib/mesa.git;a=blob;f=debian/patches/119-libllvmradeon-link.patch;h=ee47f8a07dbf33c32f8b57faed923680ed6648fb;hb=refs/heads/ubuntu%2B1

Fixes a regression introduced with
f70c385351

NOTE: This is a candidate for the 9.1 branch.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=62434
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
 2013-03-19 12:07:51 +01:00
Tom Stellard
c95177ea88 radeon/llvm: Add missing license headers 
Signed-off-by: Tom Stellard <thomas.stellard@amd.com>
 2013-03-13 16:01:31 +00:00
Tom Stellard
1c4f283151 radeon/llvm: Make radeon_llvm_util.cpp a C file 
All the functions in this file are now implemented in C.
 2013-03-13 16:01:31 +00:00
Tom Stellard
3958c104c6 radeon/llvm: Optimize radeon_llvm_strip_unused_kernels() 
Just delete unused kernels rather than marking them as internal and
running the GlobalDCE pass.

Also implement this function in C and inline it into
radeon_llvm_get_kernel_module()
 2013-03-13 16:01:31 +00:00
Tom Stellard
2ace79dce5 radeon/llvm: Implement radeon_llvm_get_kernel_module() using the C API 2013-03-13 16:01:31 +00:00
Tom Stellard
b34b8576ec radeon/llvm: Implement radeon_llvm_get_num_kernels() using the C API 2013-03-13 16:01:31 +00:00
Tom Stellard
7e9abbea15 radeon/llvm: Implement radeon_llvm_parse_bitcode() using C API 
Also make the function static since it is not used anywhere else.
 2013-03-13 16:01:30 +00:00
Tom Stellard
97bfcddde0 r600g/llvm: Move llvm wrapper functions into the radeon directory 2013-03-13 16:01:30 +00:00
Christian König
eddf33f711 radeon/llvm: document LLVM commit 
We need at least that revision to work correctly now.

Signed-off-by: Christian König <christian.koenig@amd.com>
 2013-03-07 10:06:24 +01:00
Christian König
a7a899584c radeon/llvm: enable LICM and DCE pass v2 
LICM stands for Loop Invariant Code Motion. Instructions that
does not depend of loop index are moved outside of loop body.

DCE is DeadCodeElimination.

v2: updated commit msg, thx to Vincent.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Vincent Lejeune <vljn at ovi.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-03-07 10:03:22 +01:00
Christian König
55fe5ccb39 radeon/llvm: make SGPRs proper function arguments v2 
v2: remove unrelated changes

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-03-07 10:03:22 +01:00
Christian König
b8f4ca3d85 radeon/llvm: replace shader type intrinsic with function attribute 
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-03-07 10:03:22 +01:00
Christian König
886c5085e3 radeon/llvm: fix trivial warnings 
Signed-off-by: Christian König <christian.koenig@amd.com>
 2013-03-06 12:08:54 +01:00
Michel Dänzer
f6b40ddd2d radeon/llvm: Remove stale comment about radeon_llvm_emit_prepare_cube_coords 2013-02-22 13:06:07 +01:00
Tom Stellard
7cd248aa79 radeon/llvm: Fix build with LLVM 3.3 2013-02-19 15:52:55 +00:00
Vincent Lejeune
c116598f86 r600g/llvm: Set Inputs/Outputs count to 32 (api reported value) 
Reviewed-by: Tom Stellard <thomas.stellard at amd.com>
 2013-02-18 15:08:54 +01:00
Vincent Lejeune
90e6f47ac8 r600g/llvm: Fix alpha_to_one piglit tests 
Reviewed-by: Tom Stellard <thomas.stellard at amd.com>
 2013-02-18 15:08:50 +01:00
Vincent Lejeune
ef8fde6acb r600g/llvm: Add support for UBO 
NOTE: This is a candidate for the Mesa stable branch.

Reviewed-by: Tom Stellard <thomas.stellard at amd.com>
 2013-02-18 15:08:45 +01:00
Michel Dänzer
e5fb7347a7 radeonsi: Adapt to sample intrinsics changes. 
Fix up intrinsic names, and bitcast texture address parameters to integers.

NOTE: This is a candidate for the 9.1 branch.
 2013-02-04 17:03:25 +01:00
Vadim Girlin
c9343047cf r600g: improve inputs/interpolation handling with llvm backend 
Get rid of special handling for reserved regs.
Use one intrinsic for all kinds of interpolation.

v2[Vincent Lejeune]: Rebased against current master

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
 2013-01-28 18:30:38 +00:00
Michel Dänzer
a56dfd99e2 radeon/llvm: Handle LP_CHAN_ALL in emit_fetch_immediate(). 
Fixes piglit spec/ARB_sampler_objects/sampler-incomplete and
spec/EXT_texture_swizzle/depth_texture_mode_and_swizzle.

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
 2013-01-22 18:50:02 +01:00
Vincent Lejeune
3b14ce2caf r600g/llvm: tgsi to llvm emits store.swizzle intrinsic for vs/fs output 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-01-18 20:34:26 +00:00
Vincent Lejeune
7b20526466 r600g/llvm: tgsi to llvm emits stream output intrinsics. 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-01-18 20:34:21 +00:00
Vincent Lejeune
ce34ff1ad7 r600g/llvm:translate ARL opcode to a simple cast 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-01-18 20:08:10 +00:00
Vadim Girlin
7d532800d8 r600g/llvm: rework handling of the constants 
Vincent Lejeune:
  - tgsi to llvm now emits pointers for constants

Tom Stellard:
  - Only use texture cache for vtx fetch with compute shaders
  - Change address space used for constant loads to match LLVM
    backend.

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2013-01-18 20:08:10 +00:00
Tom Stellard
4fc11fa3c8 drivers/radeon: Don't link against libgallium.la 
This fixes several duplicate symbol errors.

libllvmradeon is a simple helper library.  If it requires symbols in
other libraries, this should be taken care of by the gallium target that
uses it (e.g. libr600.la)
 2013-01-11 21:40:42 +00:00
Tom Stellard
0dcb9ae0d9 radeon/llvm: Convert to Automake 
v2: Johannes Obermayr <johannesobermayr@gmx.de>
    Fix some undefined symbols.

v3: Johannes Obermayr <johannesobermayr@gmx.de>
    Build it -shared to fix egl_gallium.so on r600/radeonsi builds.
 2013-01-10 22:01:08 +01:00
Tom Stellard
aed37cbee8 radeon/llvm: Remove backend code from Mesa 
This code now lives in an external tree.

For the next Mesa release fetch the code from the master branch
of this LLVM repo:
http://cgit.freedesktop.org/~tstellar/llvm/

For all subsequent Mesa releases, fetch the code from the official LLVM
project:
www.llvm.org
 2013-01-04 21:05:09 +00:00
Johannes Obermayr
05c143cc04 Support LLVM >= 3.2 on radeonsi and opencl. 
Tom Stellard:
 - Backend now has same name for all LLVM versions
 - Add missing LLVM_VERSION_INT definition
 2013-01-04 21:05:09 +00:00
Vadim Girlin
8cf552b182 radeon/llvm: improve cube map handling 
Add support for TEX2, TXB2, TXL2, fix SHADOWCUBE

Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
 2012-12-18 17:40:57 +04:00
Vadim Girlin
3b89fcbe54 radeon/llvm: fix TXQ_LZ handling for cube maps 
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com>
 2012-12-18 17:40:57 +04:00
Michel Dänzer
aac2154729 radeon/llvm: Export prepare_cube_coords helper to driver. 
To be used by radeonsi.

Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
 2012-12-06 20:18:40 +01:00
Vincent Lejeune
00d77e9fe4 r600g: use default action for min/max opcode in tgsi to llvm 
Reveiwed-by: Tom Stellard <thomas.stellard at amd.com>
 2012-12-05 18:31:55 +01:00
Vincent Lejeune
2a03f28e54 r600g: use default action for fdiv/rcp opcode 
Reveiwed-by: Tom Stellard <thomas.stellard at amd.com>
 2012-12-05 18:31:02 +01:00
Vincent Lejeune
0ad1fefd69 r600g: Use default mul/mad function for tgsi-to-llvm 
Reveiwed-by: Tom Stellard <thomas.stellard at amd.com>
 2012-12-05 18:30:16 +01:00
Vincent Lejeune
84b4372132 r600g: make tgsi-to-llvm generates store.pixel* intrinsic for fs 
Reviewed-by: Tom Stellard <thomas.stellard at amd.com>
 2012-11-02 23:19:11 +01:00
Michel Dänzer
f3257d80b0 radeon/llvm: Add intrinsic for reading SI FRONT_FACE VGPR in the pixel shader. 
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-26 15:51:17 +02:00
Tom Stellard
8030cb0ed4 radeon/llvm: Sort tgsi opcode action initialization 
This was done in order to identify and remove duplicate entries.
 2012-10-19 21:25:01 +00:00
Tom Stellard
bd8af8a3dc radeon/llvm: Fix lowering TGSI_OPCODE_SSG 2012-10-19 21:25:00 +00:00
Tom Stellard
4cc530f452 radeon/llvm: Fix build with LLVM 3.2 2012-10-11 21:33:00 +00:00
Vincent Lejeune
5090ce42e4 radeon/llvm: use ceil intrinsic instead of llvm.AMDIL.round.posinf 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:03:33 +02:00
Vincent Lejeune
9a6bb3f645 radeon/llvm: use floor intrinsic instead of llvm.AMDIL.floor 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:03:20 +02:00
Vincent Lejeune
bfdf26892c radeon/llvm: use llvm fabs intrinsic 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:03:03 +02:00
Vincent Lejeune
8db11bc4ed radeon/llvm: use llvm intrinsic for flog2 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:02:45 +02:00
Vincent Lejeune
23e11ac835 radeon/llvm: add support for cos/sin intrinsic 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:02:28 +02:00
Vincent Lejeune
876b42663c radeon/llvm: add a pattern for fsqrt 
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-10 22:02:13 +02:00
Michel Dänzer
dbb4a7f950 radeon/llvm: Disable SI flow control again for now. 
It makes piglit unreliable due to VM protection faults and GPU lockups.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Tom Stellard <thomas.stellard@amd.com>
 2012-10-02 16:50:36 +02:00
Tom Stellard
f2f17fc348 radeon/llvm: Only initialize the AMDGPU target 2012-10-01 21:14:10 +00:00