summaryrefslogtreecommitdiff
path: root/src/gallium/drivers/vc4/vc4_qir.c
AgeCommit message (Expand)AuthorFilesLines
2022-10-25vc4: consolidate shader-db outputJose Maria Casanova Crespo1-3/+3
2022-02-22vc4: remove redundant initializationJuan A. Suarez Romero1-1/+0
2019-10-28util: rename list_empty() to list_is_empty()Timothy Arceri1-2/+2
2018-08-07vc4: Extend dumping of uniforms in QIR and in the command stream.Eric Anholt1-8/+54
2017-02-24vc4: Track the last block we emitted at the top level.Eric Anholt1-0/+1
2017-02-10vc4: Avoid emitting small immediates for UBO indirect load address guards.Eric Anholt1-0/+1
2016-11-29vc4: Add support for coalescing ALU ops into tex_[srtb] MOVs.Eric Anholt1-0/+11
2016-11-29vc4: Split optimizing VPM writes from VPM reads.Eric Anholt1-0/+1
2016-11-29vc4: Restructure texture insts as ALU ops with tex_[strb] as the dst.Eric Anholt1-7/+65
2016-11-29vc4: Refactor qir_get_op_nsrc(enum qop) to qir_get_nsrc(struct qinst *).Eric Anholt1-7/+5
2016-11-29vc4: Replace the qinst src[] with a fixed-size array.Eric Anholt1-2/+0
2016-11-29vc4: Remove qir_inst4().Eric Anholt1-20/+0
2016-11-12vc4: Add a thread switch QIR instruction.Eric Anholt1-0/+2
2016-11-09vc4: Print a reg pressure estimate in our reg allocation failure dump.Eric Anholt1-0/+5
2016-08-25vc4: Add support for MUL output rotation.Eric Anholt1-0/+3
2016-08-25vc4: Add support for the 2-bit LOAD_IMM variants.Eric Anholt1-0/+2
2016-08-25vc4: Add a QIR value for the QPU element register.Eric Anholt1-0/+1
2016-07-13vc4: Emit resets of the uniform stream at the starts of blocks.Eric Anholt1-0/+1
2016-07-13vc4: Move the QPU instructions to schedule into each block.Eric Anholt1-0/+1
2016-07-12vc4: Define a QIR branch instructionEric Anholt1-9/+29
2016-07-12vc4: Print live variable start/ends during QIR dumping.Eric Anholt1-0/+45
2016-07-12vc4: Implement live intervals using a CFG.Eric Anholt1-0/+47
2016-07-12vc4: Create a basic block structure and move the instructions into it.Eric Anholt1-12/+67
2016-07-04vc4: Regularize instruction emit macrosEric Anholt1-2/+26
2016-07-04vc4: Optimize out redundant SF updates.Eric Anholt1-1/+1
2016-07-04vc4: Move SF removal to a separate peephole pass.Eric Anholt1-0/+1
2016-07-04vc4: Mark texturing setup instructions as having side effects.Eric Anholt1-5/+5
2016-05-06vc4: Add support for loading immediate values in QIR.Eric Anholt1-0/+6
2016-05-06vc4: Add a small QIR validate pass.Eric Anholt1-0/+1
2016-05-06vc4: Fix the src count on exp2/log2.Eric Anholt1-2/+2
2016-05-06vc4: Reuse QPU disasm's cond flags in QIR.Eric Anholt1-14/+5
2016-05-02vc4: Remove the CSE pass.Eric Anholt1-1/+0
2016-04-08vc4: Move FRAG_X/Y/REV_FLAG to a QFILE like VPM or TLB color writes.Eric Anholt1-3/+3
2016-04-08vc4: Allow TLB Z/color/stencil writes from any ALU operation in QIR.Eric Anholt1-8/+34
2016-04-08vc4: Add a helper function for the construction of qregs.Eric Anholt1-5/+3
2016-04-08vc4: Drop the multi_instruction distinction for QIR instructions.Eric Anholt1-13/+5
2016-03-16vc4: Move discard handling to the condition flag.Eric Anholt1-1/+0
2016-03-16vc4: Don't make a temporary for setting flags.Eric Anholt1-1/+2
2016-03-16vc4: Add a safety check for setting flags.Eric Anholt1-0/+3
2016-03-15vc4: Coalesce instructions using VPM reads into the VPM read.Varad Gautam1-1/+1
2016-02-15vc4: Don't treat conditional MOVs as raw MOV.Eric Anholt1-0/+1
2016-01-06vc4: Replace the SSA-style SEL operators with conditional MOVs.Eric Anholt1-31/+15
2016-01-06vc4: Don't try the SF coalescing unless it's on a def.Eric Anholt1-3/+3
2015-12-11vc4: Fix handling of src packs on in qir_follow_movs().Eric Anholt1-2/+8
2015-12-08vc4: Add support for multisample framebuffer operations.Eric Anholt1-0/+1
2015-12-04vc4: Add support for storing sample mask.Eric Anholt1-0/+2
2015-11-17vc4: Add support for nir_op_uge, using the carry bit on QPU_A_SUB.Eric Anholt1-0/+8
2015-10-26vc4: Rewrite the pack instructions as a MOV with a dst pack flagEric Anholt1-6/+4
2015-10-26vc4: Switch the unpack ops to being unpack flags on a mov.Eric Anholt1-41/+4
2015-10-26vc4: Fix up the test for whether the unpack can be from r4.Eric Anholt1-0/+24