[DynInst_API:] [dyninst/dyninst] bbfdea: Added support for gfx908 based on the XML-ISA-DROP...

Mailing List Archives Authenticated access	UW Madison Computer Sciences Department Computer Systems Lab

Date:	Tue, 28 Mar 2023 09:27:50 -0700
From:	bbiiggppiigg <noreply@xxxxxxxxxx>
Subject:	[DynInst_API:] [dyninst/dyninst] bbfdea: Added support for gfx908 based on the XML-ISA-DROP...

  Branch: refs/heads/amdgpu-codegen
  Home:   https://github.com/dyninst/dyninst
  Commit: bbfdeabe7fd42272a0016c2b4993c74381dfc633
      https://github.com/dyninst/dyninst/commit/bbfdeabe7fd42272a0016c2b4993c74381dfc633
  Author: Hsuan-Heng Wu <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M common/h/dyn_regs.h
    M instructionAPI/CMakeLists.txt
    M instructionAPI/src/AMDGPU/cdna2/InstructionDecoder-amdgpu-cdna2.C
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.C
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.h
    M instructionAPI/src/AMDGPU/gfx908/amdgpu_gfx908_decoder_impl.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.h
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C
    M instructionAPI/src/AMDGPU/vega/InstructionDecoder-amdgpu-vega.C
    A instructionAPI/src/debug_decode.C
    A instructionAPI/src/debug_decode.h

  Log Message:
  -----------
  Added support for gfx908 based on the XML-ISA-DROP for MI100

Fixed a bug in the instruction decoder that returns incorrect
value on the last instruction.


  Commit: 62f11fa283707b6840acb08fdc52b0c34f696253
      https://github.com/dyninst/dyninst/commit/62f11fa283707b6840acb08fdc52b0c34f696253
  Author: Hsuan-Heng Wu <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/amdgpu_gfx908_decoder_impl.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.h

  Log Message:
  -----------
  fix compiler warnings for unused parameter and unused variable for gfx908


  Commit: 0ea4c9daed8802161ab6c766a20a0c7a5c535478
      https://github.com/dyninst/dyninst/commit/0ea4c9daed8802161ab6c766a20a0c7a5c535478
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  updated based on new xml-isa drop


  Commit: 1c09be1b4448fd84a853b5eee69c5b58150d563b
      https://github.com/dyninst/dyninst/commit/1c09be1b4448fd84a853b5eee69c5b58150d563b
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  fix decoding of operand for v_div_scale based on 220804 isa drop


  Commit: 1dc3c11036472ec5e9b91d4f7f8d39c40d5a6f2c
      https://github.com/dyninst/dyninst/commit/1dc3c11036472ec5e9b91d4f7f8d39c40d5a6f2c
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/CMakeLists.txt
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.C
    R instructionAPI/src/debug_decode.C
    R instructionAPI/src/debug_decode.h

  Log Message:
  -----------
  revert changes related to adding debug interface


  Commit: ca6ae0a1d0659618245e900b35c5f8e9a9c6fd4b
      https://github.com/dyninst/dyninst/commit/ca6ae0a1d0659618245e900b35c5f8e9a9c6fd4b
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  compress finalizeOperands.C by combining cases with same decoding logic


  Commit: 5a01da4f049f2860faf976b07f7349dbc4ede602
      https://github.com/dyninst/dyninst/commit/5a01da4f049f2860faf976b07f7349dbc4ede602
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.h
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  Add support for decoding smem_offset operand type

Add support for decoding smem_offset type,
which is still lacking from the ISA spec


  Commit: 139eec81f0f88b2eae88630d8f14c0b8d9a22960
      https://github.com/dyninst/dyninst/commit/139eec81f0f88b2eae88630d8f14c0b8d9a22960
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M common/h/dyn_regs.h

  Log Message:
  -----------
  Add definition for waitcnt counters


  Commit: 8c1cd08a45ffaedf795e0d522ab089680ffbbc94
      https://github.com/dyninst/dyninst/commit/8c1cd08a45ffaedf795e0d522ab089680ffbbc94
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.C
    M instructionAPI/src/AMDGPU/gfx908/InstructionDecoder-amdgpu-gfx908.h
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.C
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.h
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C
    M instructionAPI/src/ArchSpecificFormatters.C

  Log Message:
  -----------
  Add formatting for consecutive registers


  Commit: dea6a58fdebc4d17666efa39c595c10fff8f7dfd
      https://github.com/dyninst/dyninst/commit/dea6a58fdebc4d17666efa39c595c10fff8f7dfd
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C
    M instructionAPI/src/ArchSpecificFormatters.C

  Log Message:
  -----------
  fix formatting regarding implicit operands


  Commit: f386645434488c487b2fac9ea6ed4f920d7f5bb1
      https://github.com/dyninst/dyninst/commit/f386645434488c487b2fac9ea6ed4f920d7f5bb1
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  remove duplicate branch target operand


  Commit: f62c3e99a5001779261da5f93f70996e6e3c0096
      https://github.com/dyninst/dyninst/commit/f62c3e99a5001779261da5f93f70996e6e3c0096
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/finalizeOperands.C

  Log Message:
  -----------
  fix the logic for operands addr and saddr for flat instruction encoding family based on the manual


  Commit: 665772342c4a6b25245d3c419d3eab7201c3b561
      https://github.com/dyninst/dyninst/commit/665772342c4a6b25245d3c419d3eab7201c3b561
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/gfx908/decodeOperands.C

  Log Message:
  -----------
  fix logic for decoding src_literal, so it correctly consume the next 32 bit as data instead of decoding them as instruction


  Commit: 01e6c1f3dc093c531cf2aab58f537654c2e5e703
      https://github.com/dyninst/dyninst/commit/01e6c1f3dc093c531cf2aab58f537654c2e5e703
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M instructionAPI/src/AMDGPU/cdna2/amdgpu_cdna2_decoder_impl.C
    M instructionAPI/src/AMDGPU/cdna2/amdgpu_cdna2_decoder_impl.h
    M instructionAPI/src/AMDGPU/cdna2/amdgpu_cdna2_insn_entry.h
    M instructionAPI/src/AMDGPU/cdna2/amdgpu_cdna2_opcode_tables.C
    M instructionAPI/src/AMDGPU/cdna2/finalizeOperands.C

  Log Message:
  -----------
  files updated based on new xml isa-drop


  Commit: ebddd8582faaf9e24cc11a452b4f89dbd46bf601
      https://github.com/dyninst/dyninst/commit/ebddd8582faaf9e24cc11a452b4f89dbd46bf601
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M dyninstAPI/CMakeLists.txt
    A dyninstAPI/h/arch-amdgpu.h
    A dyninstAPI/h/arch-forward-decl.h
    A dyninstAPI/h/codegen-amdgpu.h
    A dyninstAPI/src/arch-amdgpu.C
    A dyninstAPI/src/codegen-amdgpu.C

  Log Message:
  -----------
  initial support for assembly-level codegen for amdgpu


  Commit: cf0bcd7f3e96bfb7bd7605b4fb6c97cafec89d82
      https://github.com/dyninst/dyninst/commit/cf0bcd7f3e96bfb7bd7605b4fb6c97cafec89d82
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M dyninstAPI/h/codegen-amdgpu.h
    M dyninstAPI/src/codegen-amdgpu.C

  Log Message:
  -----------
  add support for v_add instruction codegen


  Commit: 8f2531a172e5c0c4362000aab959dd1c92ba8e6a
      https://github.com/dyninst/dyninst/commit/8f2531a172e5c0c4362000aab959dd1c92ba8e6a
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M dyninstAPI/h/codegen-amdgpu.h
    M dyninstAPI/src/codegen-amdgpu.C

  Log Message:
  -----------
  add support for sopp smem instructions


  Commit: fd0dc49bf1c00608c96dd9e68796f639d546d8d1
      https://github.com/dyninst/dyninst/commit/fd0dc49bf1c00608c96dd9e68796f639d546d8d1
  Author: wuxx1279 <bbiiggppiigg@xxxxxxxxx>
  Date:   2023-03-15 (Wed, 15 Mar 2023)

  Changed paths:
    M dyninstAPI/h/codegen-amdgpu.h
    M dyninstAPI/src/codegen-amdgpu.C

  Log Message:
  -----------
  Fix issue with amdgpu s_nop codegen

Previous SOPP use a wrong encoding.
Remove unecessary parameter to gen_s_nop.


Compare: https://github.com/dyninst/dyninst/compare/2cc19da383fb...fd0dc49bf1c0

[← Prev in Thread]	Current Thread	[Next in Thread→]
[DynInst_API:] [dyninst/dyninst] bbfdea: Added support for gfx908 based on the XML-ISA-DROP..., bbiiggppiigg <=

Previous by Date:	[DynInst_API:] [dyninst/dyninst] 77b7b1: Update variable names in dependency-version CI check, Tim Haines
Next by Date:	[DynInst_API:] [dyninst/dyninst] 35532e: Rename AMDGPU cdna2 to gfx90a, decoder updated, bbiiggppiigg
Previous by Thread:	[DynInst_API:] [dyninst/dyninst] bb7771: Mismatched printf format arguments, Tim Haines
Next by Thread:	[DynInst_API:] [dyninst/dyninst] bbfe4c: Fix UB in dwarf/dwarf_subrange.cpp::get_type, Tim Haines
Indexes:	[Date] [Thread]

Mailing List Archives

Authenticated access

[DynInst_API:] [dyninst/dyninst] bbfdea: Added support for gfx908 based on the XML-ISA-DROP...