fbc: aarch64 target by default maps to '-march=armv8-a' when passing options to gcc/LLVM

- aarch64 is not a permissible value for -march passed to gcc/LLVM
- fbc will accept aarch64, then map to '-march=armv8-a' when calling the backend
- comments in fbc.bas:hCompileStage2Module() listing permissible values

Author: czsgaba

changelog.txt

@@ -3,6 +3,7 @@ Version 1.07.0
 [changed]
 - SADD/STRPTR(wstring) returns WSTRING PTR
 - '-v' be verbose command line option affects '-help' output
+- fbc aarch64 target by default maps to '-march=armv8-a' when passing options to gcc/LLVM (czsgaba)
 
 [added]
 - CVA_LIST type, CVA_START(), CVA_COPY() CVA_END(), CVA_ARG() macros will map to gcc's __builtin_va_list and __builtin_va_* macros in gcc backend
@@ -46,6 +47,7 @@ Version 1.07.0
 - github #116: Fix optimizations in [L/R]TrimAny rtlib functions (SkyFish)
 - github #145: WSTRING concat and assign buffer (&=) overrun
 - sf.net #893: 'Suffixes are only valid in -lang' error message showing incorrect -lang options allowed
+- fbc uses '-march=armv8-a' instead of invalid option '-march=aarch64' when passing options to gcc/LLVM (czsgaba)
 
 
 Version 1.06.0

src/compiler/fb.bas

@@ -232,7 +232,7 @@ dim shared as FBCPUTYPEINFO cputypeinfo(0 to FB_CPUTYPE__COUNT-1) = _
 	( NULL       , @"x86-64"       , FB_CPUFAMILY_X86_64 , 64 ), _ '' FB_CPUTYPE_X86_64
 	( NULL       , @"armv6"        , FB_CPUFAMILY_ARM    , 32 ), _ '' FB_CPUTYPE_ARMV6
 	( NULL       , @"armv7-a"      , FB_CPUFAMILY_ARM    , 32 ), _ '' FB_CPUTYPE_ARMV7A
-	( NULL       , @"aarch64"      , FB_CPUFAMILY_AARCH64, 64 )  _ '' FB_CPUTYPE_AARCH64
+	( @"armv8-a" , @"aarch64"      , FB_CPUFAMILY_AARCH64, 64 )  _ '' FB_CPUTYPE_AARCH64
 }
 
 ''::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

src/compiler/fbc.bas

@@ -2961,7 +2961,39 @@ private function hCompileStage2Module( byval module as FBCIOFILE ptr ) as intege
 		case FB_CPUFAMILY_ARM
 			ln += "-march=arm "
 		case FB_CPUFAMILY_AARCH64
-			ln += "-march=aarch64 "
+			'' From the GCC manual:
+			'' -march=name
+			'' Specify the name of the target architecture and, 
+			'' optionally, one or more feature modifiers. This option 
+			'' has the form ‘-march=arch{+[no]feature}*’.
+			'' 
+			'' The permissible values for arch are 
+			'' 'armv8-a'
+			'' 'armv8.1-a' = 'armv8-a' + ARMv8.1-A 
+			'' 'armv8.2-a' = 'armv8.1-a' + ARMv8.2-A
+			'' 'armv8.3-a' = 'armv8.2-a' + ARMv8.3-A
+			'' 'armv8.4-a' = 'armv8.3-a' + ARMv8.4-A
+			'' 'armv8.5-a' = 'armv8.4-a' + ARMv8.5-A
+			'' 'native' = architecture of the host system
+			'' 
+			'' It enables the '+crc', '+lse', and '+rdma' features.
+			'' 
+			'' The value 'native' is available on native AArch64 
+			'' GNU/Linux and causes the compiler to pick the 
+			'' architecture of the host system. This option has no 
+			'' effect if the compiler is unable to recognize the 
+			'' architecture of the host system, The permissible 
+			'' values for feature are listed in the sub-section on 
+			'' ['-march' and '-mcpu' Feature Modifiers]. Where 
+			'' conflicting feature modifiers are specified, the 
+			'' right-most feature is used. GCC uses name to determine 
+			'' what kind of instructions it can emit when generating 
+			'' assembly code. If '-march' is specified without either 
+			'' of '-mtune' or '-mcpu' also being specified, the code
+			'' is tuned to perform well across a range of target 
+			'' processors implementing the target architecture.
+
+			ln += "-march=armv8-a "
 		end select
 
 		if( fbGetOption( FB_COMPOPT_PIC ) ) then