index
:
mesa.git
gallium_va_encpackedheader01
master
Unnamed repository; edit this file 'description' to name the repository.
summary
refs
log
tree
commit
diff
stats
log msg
author
committer
range
path:
root
/
src
Commit message (
Expand
)
Author
Age
Files
Lines
*
nir/builder: Add a nir_imm_intN_t helper
Jason Ekstrand
2017-11-07
1
-0
/
+12
*
nir/lower_system_values: Lower SUBGROUP_*_MASK based on type
Jason Ekstrand
2017-11-07
1
-2
/
+3
*
nir: Make ballot intrinsics variable-size
Jason Ekstrand
2017-11-07
3
-6
/
+8
*
nir: Add a ssa_dest_init_for_type helper
Jason Ekstrand
2017-11-07
1
-0
/
+9
*
nir: Add a new subgroups lowering pass
Jason Ekstrand
2017-11-07
8
-186
/
+208
*
intel/fs: Don't use automatic exec size inference
Jason Ekstrand
2017-11-07
1
-3
/
+9
*
intel/fs: Explicitly set EXECUTE_1 where needed
Jason Ekstrand
2017-11-07
4
-9
/
+15
*
intel/eu: Explicitly set EXECUTE_1 where needed
Jason Ekstrand
2017-11-07
1
-0
/
+9
*
intel/eu: Make automatic exec sizes a configurable option
Jason Ekstrand
2017-11-07
3
-14
/
+29
*
intel/fs: Rework zero-length URB write handling
Jason Ekstrand
2017-11-07
1
-29
/
+31
*
intel/compiler/fs: Set up subgroup invocation as a system value
Jason Ekstrand
2017-11-07
1
-13
/
+21
*
intel/cs: Push subgroup ID instead of base thread ID
Jason Ekstrand
2017-11-07
9
-36
/
+40
*
intel/cs: Re-run final NIR optimizations for each SIMD size
Jason Ekstrand
2017-11-07
1
-41
/
+69
*
intel/compiler: Move the destructor from vec4_visitor to backend_shader
Jason Ekstrand
2017-11-07
4
-5
/
+5
*
i965/fs: Get rid of the early return in brw_compile_cs
Jason Ekstrand
2017-11-07
1
-13
/
+14
*
intel/cs: Rework the way thread local ID is handled
Jason Ekstrand
2017-11-07
6
-46
/
+32
*
intel/fs: Mark 64-bit values as being contiguous
Jason Ekstrand
2017-11-07
1
-1
/
+4
*
intel/cs: Ignore runtime_check_aads_emit for CS
Jason Ekstrand
2017-11-07
1
-2
/
+1
*
intel/cs: Stop setting dispatch_grf_start_reg
Jason Ekstrand
2017-11-07
2
-3
/
+0
*
intel/cs: Drop max_dispatch_width checks from compile_cs
Jason Ekstrand
2017-11-07
1
-4
/
+8
*
intel/fs: Remove min_dispatch_width from fs_visitor
Jason Ekstrand
2017-11-07
3
-33
/
+25
*
intel/fs: use pull constant locations to check for first compile of a shader
Jason Ekstrand
2017-11-07
2
-2
/
+7
*
intel/fs: Retype dest to match value in read[First]Invocation
Jason Ekstrand
2017-11-07
1
-4
/
+2
*
intel/fs: Uniformize the index in readInvocation
Jason Ekstrand
2017-11-07
1
-1
/
+1
*
intel/fs: Protect opt_algebraic from OOB BROADCAST indices
Jason Ekstrand
2017-11-07
1
-2
/
+11
*
i965/fs/nir: Don't stomp 64-bit values to D in get_nir_src
Jason Ekstrand
2017-11-07
1
-13
/
+24
*
i965/fs/nir: Minor refactor of store_output
Jason Ekstrand
2017-11-07
1
-4
/
+3
*
i965/fs: Return a fs_reg from shuffle_64bit_data_for_32bit_write
Jason Ekstrand
2017-11-07
2
-29
/
+12
*
i965/fs/nir: Simplify 64-bit store_output
Jason Ekstrand
2017-11-07
1
-19
/
+6
*
intel/fs: Use the original destination region for int MUL lowering
Jason Ekstrand
2017-11-07
1
-7
/
+9
*
intel/fs: Fix integer multiplication lowering for src/dst hazards
Jason Ekstrand
2017-11-07
1
-2
/
+8
*
intel/fs: Fix MOV_INDIRECT for 64-bit values on little-core
Jason Ekstrand
2017-11-07
1
-36
/
+39
*
intel/eu: Fix broadcast instruction for 64-bit values on little-core
Jason Ekstrand
2017-11-07
1
-2
/
+24
*
intel/eu/reg: Add a subscript() helper
Jason Ekstrand
2017-11-07
1
-0
/
+16
*
intel/eu: Just modify the offset in brw_broadcast
Jason Ekstrand
2017-11-07
1
-4
/
+5
*
intel/compiler: Add some restrictions to MOV_INDIRECT and BROADCAST
Jason Ekstrand
2017-11-07
3
-0
/
+20
*
intel/fs: Use a pair of 1-wide MOVs instead of SEL for any/all
Jason Ekstrand
2017-11-07
1
-9
/
+33
*
intel/fs: Use an explicit D type for vote any/all/eq intrinsics
Jason Ekstrand
2017-11-07
1
-0
/
+6
*
intel/fs: Don't stomp f0.1 in SIMD16 ballot
Jason Ekstrand
2017-11-07
1
-2
/
+9
*
intel/fs: Use ANY/ALL32 predicates in SIMD32
Jason Ekstrand
2017-11-07
1
-12
/
+30
*
intel/fs: Be more explicit about our placement of [un]zip
Jason Ekstrand
2017-11-07
1
-3
/
+17
*
intel/fs: Pass builders instead of blocks into emit_[un]zip
Jason Ekstrand
2017-11-07
1
-26
/
+35
*
intel/fs: Use a pure vertical stride for large register strides
Jason Ekstrand
2017-11-07
1
-3
/
+13
*
broadcom/vc5: Skip emitting textures that aren't used.
Eric Anholt
2017-11-07
1
-2
/
+4
*
broadcom/vc5: Add missing SRGBA8 ETC2 support.
Eric Anholt
2017-11-07
1
-0
/
+1
*
broadcom/vc5: Disable early Z test when the FS writes Z.
Eric Anholt
2017-11-07
1
-1
/
+2
*
broadcom/vc5: Shift the min/max lod fields by the BASE_LEVEL.
Eric Anholt
2017-11-07
2
-4
/
+15
*
broadcom/vc5: Add support for anisotropic filtering.
Eric Anholt
2017-11-07
1
-0
/
+9
*
broadcom/vc5: Fix mipmap filtering enums.
Eric Anholt
2017-11-07
2
-8
/
+32
*
broadcom/vc5: Fix height padding of small UIF slices.
Eric Anholt
2017-11-07
1
-1
/
+5
[next]