Update LDC compiler class

[mleise]

The compiler installs as ldc2 nowadays and some of the flags need to be adapted, like
-[enable/disable]-inlining

[s-ludwig]

I'll fix the inline flag. Do you have any other concrete examples that need work? I'll need some time until I can build and test with the latest LDC.

The "ldc2" binary name should be fine, though. Just run with "dub --compiler=ldc2". I'll make it clear in the command line help that arbitrary prefixes/suffixes to the compiler name are supported.

[mleise]

I tried that, but assumed --compiler==ldc2 would use some default. No I have no other examples at the moment. I could dump the command-line help output. Then again this can probably wait. The -inline switch was actually stopping 'simple' release builds with ldc from working.

[mleise]

GitHub will survive it. See if you use any of the flags incorrectly... -d-version and -d-debug ? deprectation warning flags?
Looks like inlining is enabled by default at -O2

OPTIONS:
  -D                                           - Generate documentation
  -Dd=<docdir>                                 - Write documentation file to <docdir> directory
  -Df=<filename>                               - Write documentation file to <filename>
  -H                                           - Generate 'header' file
  -Hd=<hdrdir>                                 - Write 'header' file to <hdrdir> directory
  -Hf=<filename>                               - Write 'header' file to <filename>
  -Hkeep-all-bodies                            - Keep all function bodies in .di files
  -I=<path>                                    - Where to look for imports
  -J=<path>                                    - Where to look for string imports
  -L=<linkerflag>                              - Pass <linkerflag> to the linker
  Setting the optimization level:
    -O                                         - Equivalent to -O3
    -O0                                        - No optimizations (default)
    -O1                                        - Simple optimizations
    -O2                                        - Good optimizations
    -O3                                        - Aggressive optimizations
    -O4                                        - Link-time optimization
    -O5                                        - Link-time optimization
    -Os                                        - Like -O2 with extra optimizations for size
    -Oz                                        - Like -Os but reduces code size further
  -X                                           - generate JSON file
  -Xf=<filename>                               - write JSON file to <filename>
  -asm-verbose                                 - Add comments to directives.
  -bounds-checking-single-trap                 - Use one trap block per function
  -c                                           - Do not link
  -check-printf-calls                          - Validate printf call format strings against arguments
  -code-model                                  - Code model
    =default                                   -   Target default code model
    =small                                     -   Small code model
    =kernel                                    -   Kernel code model
    =medium                                    -   Medium code model
    =large                                     -   Large code model
  -cppfname=<function name>                    - Specify the name of the generated function
  -cppfor=<string>                             - Specify the name of the thing to generate
  -cppgen                                      - Choose what kind of output to generate
    =program                                   -   Generate a complete program
    =module                                    -   Generate a module definition
    =contents                                  -   Generate contents of a module
    =function                                  -   Generate a function definition
    =functions                                 -   Generate all function definitions
    =inline                                    -   Generate an inline function
    =variable                                  -   Generate a variable definition
    =type                                      -   Generate a type definition
  Allow deprecated code/language features:
    -de                                        - Do not allow deprecated features
    -d                                         - Silently allow deprecated features
    -dw                                        - Warn about the use of deprecated features
  -d-debug=<level/idents>                      - Compile in debug code >= <level> or identified by <idents>.
  -d-version=<level/idents>                    - Compile in version code >= <level> or identified by <idents>
  -debuglib=<lib,...>                          - Set default libraries for debug build
  -defaultlib=<lib,...>                        - Set default libraries for non-debug build
  -deps=<filename>                             - Write module dependencies to filename
  -enable-asserts                              - (*) Enable assertions
  -enable-boundscheck                          - (*) Enable array bounds checks
  -disable-d-passes                            - Disable all D-specific passes
  -disable-gc2stack                            - Disable promotion of GC allocations to stack memory
  -enable-preconditions                        - (*) Enable function preconditions
  -disable-red-zone                            - Do not emit code that uses the red zone.
  -disable-simplify-drtcalls                   - Disable simplification of druntime calls
  -disable-simplify-libcalls                   - Disable simplification of well-known C runtime calls
  -disable-spill-fusing                        - Disable fusing of spill code into instructions
  -enable-contracts                            - (*) Enable function pre- and post-conditions
  -enable-correct-eh-support                   - Make the -lowerinvoke pass insert expensive, but correct, EH code
  -enable-inlining                             - Enable function inlining (default in -O2 and higher)
  -enable-invariants                           - (*) Enable invariants
  -enable-load-pre                             - 
  -enable-objc-arc-opts                        - enable/disable all ARC Optimizations
  -enable-postconditions                       - (*) Enable function postconditions
  -enable-tbaa                                 - 
  -fatal-assembler-warnings                    - Consider warnings as error
  -fdata-sections                              - Emit data into separate sections
  -ffunction-sections                          - Emit functions into separate sections
  Generating debug information:
    -g                                         - Generate debug information
    -gc                                        - Same as -g, but pretend to be C
  -help                                        - Display available options (-help-hidden for more)
  -ignore                                      - Ignore unsupported pragmas
  -internalize-public-api-file=<filename>      - A file containing list of symbol names to preserve
  -internalize-public-api-list=<list>          - A list of symbol names to preserve
  -join-liveintervals                          - Coalesce copies (default=true)
  -lib                                         - Create static library
  -limit-float-precision=<uint>                - Generate low-precision inline sequences for some float libcalls
  -linkonce-templates                          - Use linkonce_odr linkage for template symbols instead of weak_odr
  -m32                                         - 32 bit target
  -m64                                         - 64 bit target
  -main                                        - add empty main() (e.g. for unittesting)
  -march=<string>                              - Architecture to generate code for:
  -mattr=<a1,+a2,-a3,...>                      - Target specific attributes (-mattr=help for details)
  -mc-x86-disable-arith-relaxation             - Disable relaxation of arithmetic instruction for X86
  -mcpu=<cpu-name>                             - Target a specific cpu type (-mcpu=help for details)
  -mtriple=<string>                            - Override target triple
  -noasm                                       - Disallow use of inline assembler
  -nodefaultlib                                - Don't add a default library for linking implicitly
  -noruntime                                   - Do not allow code that generates implicit runtime calls
  -o-                                          - Do not write object file
  -od=<objdir>                                 - Write object files to directory <objdir>
  -of=<filename>                               - Use <filename> as output file name
  -op                                          - Do not strip paths from source file
  -oq                                          - Write object files with fully qualified names
  -output-bc                                   - Write LLVM bitcode
  -output-ll                                   - Write LLVM IR
  -output-o                                    - Write native object
  -output-s                                    - Write native assembly
  -pre-RA-sched                                - Instruction schedulers available (before register allocation):
    =default                                   -   Best scheduler for the target
    =list-burr                                 -   Bottom-up register reduction list scheduling
    =source                                    -   Similar to list-burr but schedules in source order when possible
    =list-hybrid                               -   Bottom-up register pressure aware list scheduling which tries to balance latency and register pressure
    =list-ilp                                  -   Bottom-up register pressure aware list scheduling which tries to balance ILP and register pressure
    =vliw-td                                   -   VLIW scheduler
  -print-after-all                             - Print IR after each pass
  -print-before-all                            - Print IR before each pass
  -print-machineinstrs=<pass-name>             - Print machine instrs
  -profile-estimator-loop-weight=<loop-weight> - Number of loop executions used for profile-estimator
  -profile-file=<filename>                     - Profile file loaded by -profile-metadata-loader
  -profile-info-file=<filename>                - Profile file loaded by -profile-loader
  -profile-verifier-noassert                   - Disable assertions
  -property                                    - Enforce property syntax
  -regalloc                                    - Register allocator to use
    =default                                   -   pick register allocator based on -O option
    =fast                                      -   fast register allocator
    =greedy                                    -   greedy register allocator
  -release                                     - Disables asserts, invariants, contracts and boundscheck
  -relocation-model                            - Relocation model
    =default                                   -   Target default relocation model
    =static                                    -   Non-relocatable code
    =pic                                       -   Fully relocatable, position independent code
    =dynamic-no-pic                            -   Relocatable external references, non-relocatable code
  -run=<string>                                - program args...
  -shared                                      - Create shared library
  -shrink-wrap                                 - Shrink wrap callee-saved register spills/restores
  -singleobj                                   - Create only a single output object file
  -spiller                                     - Spiller to use: (default: standard)
    =trivial                                   -   trivial spiller
    =inline                                    -   inline spiller
  -strip-debug                                 - Strip symbolic debug information before optimization
  -struct-path-tbaa                            - 
  -time-passes                                 - Time each pass, printing elapsed time for each on exit
  -unit-at-a-time                              - Enable basic IPO
  -unittest                                    - Compile in unit tests
  -v                                           - Verbose
  -v-cg                                        - Verbose codegen
  -vectorize-loops                             - Run the Loop vectorization passes
  -vectorize-slp                               - Run the SLP vectorization passes
  -vectorize-slp-aggressive                    - Run the BB vectorization passes
  -verify-dom-info                             - Verify dominator info (time consuming)
  -verify-loop-info                            - Verify loop info (time consuming)
  -verify-regalloc                             - Verify during register allocation
  -verify-region-info                          - Verify region info (time consuming)
  -verify-scev                                 - Verify ScalarEvolution's backedge taken counts (slow)
  -version                                     - Display the version of this program
  -vv                                          - Very verbose
  Warnings:
    -w                                         - Enable warnings
    -wi                                        - Enable informational warnings
  -x86-asm-syntax                              - Choose style of code to emit from X86 backend:
    =att                                       -   Emit AT&T-style assembly
    =intel                                     -   Emit Intel-style assembly
  -x86-early-ifcvt                             - Enable early if-conversion on X86
  -x86-use-vzeroupper                          - Minimize AVX to SSE transition penalty

[s-ludwig]

Thanks. I've added cases for the ones I could find. Leaving this open for confirmation.

[mleise]

Not quite there yet. For some reason the build process breaks when invoking the compiler. The relevant part of the output is:

Running ldc2...
ldc2 -w -release -O -enable-inlining -of/tmp/dub/3738160016/fast -oq -od=.dub/obj -d-version=Have_fast -d-version=benchmark -I/home/marco/Entwicklung/D/fast/source source/fast/internal.d source/fast/uniconv.d source/fast/string.d source/std/simd.d
Error: unrecognized file extension d"

Error: Build command failed with exit code 1

Full exception: object.Exception@source/dub/generators/build.d(133): Build command failed with exit code 1

When I run this line manually it works. Do you print the command verbatim?

[s-ludwig]

Sorry, I lost track of this ticket. The command line is passed via a @responsefile, so there may be something hidden - although nothing really looks suspicious...

The response file is generated in the "targetPath" directory and deleted when the compiler process has exited (generators/build.d line 107). You could try to comment out that line and see what is wrong in ".dmd-response-file.txt". Anyway, I'll eventually test with LDC myself but I still have to finish a couple of other things beforehand.