Changes between Version 11 and Version 12 of ClrxChangeLog


Ignore:
Timestamp:
09/21/18 12:01:51 (6 years ago)
Author:
matszpk
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • ClrxChangeLog

    v11 v12  
    11= Change Log =
     2
     3CLRadeonExtender 0.1.8:
     4
     5* add chapter about binary formats to CLRX documentation
     6* add some informations about compilation under FreeBSD
     7* add '.nosectdiffs' to disable new section difference behaviour if new ROCm format choosen
     8* small optimization in the AsmScope destructor.
     9* add extra info about setting up number of the SGPRs register in documentation
     10* fixed OpenCL detection for AMDGPU-PRO
     11* add '.enum' pseudo-op to simplify defining enumerations
     12* add CLRX_VERSION_NUMBER and CLRX_POLICY_UNIFIED_SGPR_COUNT
     13* add policy to unify SGPR counting for all binary formats (by default disabled)
     14* in documentation fix some some mistakes about building
     15* add preliminary support for CPU architectures (untested): SPARC, IA64 and MIPS
     16* add new '.dims' syntax for distinguish vector group ids and scalar local ids
     17* improve CLZ32/64 for MSVC
     18* introduce CTZ32/64
     19* while disassemblying determine minimal AMD driver version for GPU device type
     20  (better code detection while disassemblying)
     21* fixed some types in documentation
     22* update list of GPU devices in documentation
     23* fix stupid and old bug in ImageMix sample
     24* change a GPU device name for VEGA11 to GFX902
     25* fixed segfault when attempt to disassemble old Gallium binaries using new Gallium binary format
     26* sort the kernels by an offset order by disassemblying
     27* better input data checking while disassemblying code
     28* add HSALayout mode for AMDCL2 format (similar code layout like in ROCm and Gallium formats)
     29* introduce kernel code parts ('.kcode' and '.kcodeend') to AMDCL2
     30* check sanity of use LDS in AMD VEGA architecture (can be used only in SCRATCH and GLOBAL)
     31* in source code add new types: GPUArchMask, AsmKernelId and AsmSectionId type.
     32* allow constant literals in sym regranges
     33* fixed symreg ranges checking
     34* fixed handling some the symbol names similar to some register names (like exec_masc)
     35* add new GPU devices to list (gfx904, gfx905, gfx906 and gfx907)
     36* add AMD VEGA 20 instruction set
     37* add much stuff to handle register allocation (still it doesn't work and it wasnot finished)
     38* add a DTree structure to save memory in storing register allocation structures
     39* fixed possible segfault while preparing to write when ASMKERN_INNER is present
    240
    341CLRadeonExtender 0.1.7: