michael@0: #!/usr/bin/perl
michael@0: 
michael@0: my $bigend;  # little/big endian
michael@0: my $nxstack;
michael@0: 
michael@0: $nxstack = 0;
michael@0: 
michael@0: eval 'exec /usr/local/bin/perl -S $0 ${1+"$@"}'
michael@0:     if $running_under_some_shell;
michael@0: 
michael@0: while ($ARGV[0] =~ /^-/) {
michael@0:     $_ = shift;
michael@0:   last if /^--/;
michael@0:     if (/^-n/) {
michael@0:     $nflag++;
michael@0:     next;
michael@0:     }
michael@0:     die "I don't recognize this switch: $_\\n";
michael@0: }
michael@0: $printit++ unless $nflag;
michael@0: 
michael@0: $\ = "\n";      # automatically add newline on print
michael@0: $n=0;
michael@0: 
michael@0: $thumb = 0;     # ARM mode by default, not Thumb.
michael@0: 
michael@0: LINE:
michael@0: while (<>) {
michael@0: 
michael@0:     # For ADRLs we need to add a new line after the substituted one.
michael@0:     $addPadding = 0;
michael@0: 
michael@0:     # First, we do not dare to touch *anything* inside double quotes, do we?
michael@0:     # Second, if you want a dollar character in the string,
michael@0:     # insert two of them -- that's how ARM C and assembler treat strings.
michael@0:     s/^([A-Za-z_]\w*)[ \t]+DCB[ \t]*\"/$1:   .ascii \"/   && do { s/\$\$/\$/g; next };
michael@0:     s/\bDCB\b[ \t]*\"/.ascii \"/                          && do { s/\$\$/\$/g; next };
michael@0:     s/^(\S+)\s+RN\s+(\S+)/$1 .req r$2/                    && do { s/\$\$/\$/g; next };
michael@0:     # If there's nothing on a line but a comment, don't try to apply any further
michael@0:     #  substitutions (this is a cheap hack to avoid mucking up the license header)
michael@0:     s/^([ \t]*);/$1@/                                     && do { s/\$\$/\$/g; next };
michael@0:     # If substituted -- leave immediately !
michael@0: 
michael@0:     s/@/,:/;
michael@0:     s/;/@/;
michael@0:     while ( /@.*'/ ) {
michael@0:       s/(@.*)'/$1/g;
michael@0:     }
michael@0:     s/\{FALSE\}/0/g;
michael@0:     s/\{TRUE\}/1/g;
michael@0:     s/\{(\w\w\w\w+)\}/$1/g;
michael@0:     s/\bINCLUDE[ \t]*([^ \t\n]+)/.include \"$1\"/;
michael@0:     s/\bGET[ \t]*([^ \t\n]+)/.include \"${ my $x=$1; $x =~ s|\.s|-gnu.S|; \$x }\"/;
michael@0:     s/\bIMPORT\b/.extern/;
michael@0:     s/\bEXPORT\b/.global/;
michael@0:     s/^(\s+)\[/$1IF/;
michael@0:     s/^(\s+)\|/$1ELSE/;
michael@0:     s/^(\s+)\]/$1ENDIF/;
michael@0:     s/IF *:DEF:/ .ifdef/;
michael@0:     s/IF *:LNOT: *:DEF:/ .ifndef/;
michael@0:     s/ELSE/ .else/;
michael@0:     s/ENDIF/ .endif/;
michael@0: 
michael@0:     if( /\bIF\b/ ) {
michael@0:       s/\bIF\b/ .if/;
michael@0:       s/=/==/;
michael@0:     }
michael@0:     if ( $n == 2) {
michael@0:         s/\$/\\/g;
michael@0:     }
michael@0:     if ($n == 1) {
michael@0:         s/\$//g;
michael@0:         s/label//g;
michael@0:     $n = 2;
michael@0:       }
michael@0:     if ( /MACRO/ ) {
michael@0:       s/MACRO *\n/.macro/;
michael@0:       $n=1;
michael@0:     }
michael@0:     if ( /\bMEND\b/ ) {
michael@0:       s/\bMEND\b/.endm/;
michael@0:       $n=0;
michael@0:     }
michael@0: 
michael@0:     # ".rdata" doesn't work in 'as' version 2.13.2, as it is ".rodata" there.
michael@0:     #
michael@0:     if ( /\bAREA\b/ ) {
michael@0:         if ( /CODE/ ) {
michael@0:             $nxstack = 1;
michael@0:         }
michael@0:         s/^(.+)CODE(.+)READONLY(.*)/    .text/;
michael@0:         s/^(.+)DATA(.+)READONLY(.*)/    .section .rdata\n    .align 2/;
michael@0:         s/^(.+)\|\|\.data\|\|(.+)/    .data\n    .align 2/;
michael@0:         s/^(.+)\|\|\.bss\|\|(.+)/    .bss/;
michael@0:     }
michael@0: 
michael@0:     s/\|\|\.constdata\$(\d+)\|\|/.L_CONST$1/;       # ||.constdata$3||
michael@0:     s/\|\|\.bss\$(\d+)\|\|/.L_BSS$1/;               # ||.bss$2||
michael@0:     s/\|\|\.data\$(\d+)\|\|/.L_DATA$1/;             # ||.data$2||
michael@0:     s/\|\|([a-zA-Z0-9_]+)\@([a-zA-Z0-9_]+)\|\|/@ $&/;
michael@0:     s/^(\s+)\%(\s)/    .space $1/;
michael@0: 
michael@0:     s/\|(.+)\.(\d+)\|/\.$1_$2/;                     # |L80.123| -> .L80_123
michael@0:     s/\bCODE32\b/.code 32/ && do {$thumb = 0};
michael@0:     s/\bCODE16\b/.code 16/ && do {$thumb = 1};
michael@0:     if (/\bPROC\b/)
michael@0:     {
michael@0:         print "    .thumb_func" if ($thumb);
michael@0:         s/\bPROC\b/@ $&/;
michael@0:     }
michael@0:     s/^(\s*)(S|Q|SH|U|UQ|UH)ASX\b/$1$2ADDSUBX/;
michael@0:     s/^(\s*)(S|Q|SH|U|UQ|UH)SAX\b/$1$2SUBADDX/;
michael@0:     s/\bENDP\b/@ $&/;
michael@0:     s/\bSUBT\b/@ $&/;
michael@0:     s/\bDATA\b/@ $&/;   # DATA directive is deprecated -- Asm guide, p.7-25
michael@0:     s/\bKEEP\b/@ $&/;
michael@0:     s/\bEXPORTAS\b/@ $&/;
michael@0:     s/\|\|(.)+\bEQU\b/@ $&/;
michael@0:     s/\|\|([\w\$]+)\|\|/$1/;
michael@0:     s/\bENTRY\b/@ $&/;
michael@0:     s/\bASSERT\b/@ $&/;
michael@0:     s/\bGBLL\b/@ $&/;
michael@0:     s/\bGBLA\b/@ $&/;
michael@0:     s/^\W+OPT\b/@ $&/;
michael@0:     s/:OR:/|/g;
michael@0:     s/:SHL:/<</g;
michael@0:     s/:SHR:/>>/g;
michael@0:     s/:AND:/&/g;
michael@0:     s/:LAND:/&&/g;
michael@0:     s/CPSR/cpsr/;
michael@0:     s/SPSR/spsr/;
michael@0:     s/ALIGN$/.balign 4/;
michael@0:     s/ALIGN\s+([0-9x]+)$/.balign $1/;
michael@0:     s/psr_cxsf/psr_all/;
michael@0:     s/LTORG/.ltorg/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+EQU/ .set $1,/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+SETL/ .set $1,/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+SETA/ .set $1,/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+\*/ .set $1,/;
michael@0: 
michael@0:     #  {PC} + 0xdeadfeed  -->  . + 0xdeadfeed
michael@0:     s/\{PC\} \+/ \. +/;
michael@0: 
michael@0:     # Single hex constant on the line !
michael@0:     #
michael@0:     # >>> NOTE <<<
michael@0:     #   Double-precision floats in gcc are always mixed-endian, which means
michael@0:     #   bytes in two words are little-endian, but words are big-endian.
michael@0:     #   So, 0x0000deadfeed0000 would be stored as 0x0000dead at low address
michael@0:     #   and 0xfeed0000 at high address.
michael@0:     #
michael@0:     s/\bDCFD\b[ \t]+0x([a-fA-F0-9]{8})([a-fA-F0-9]{8})/.long 0x$1, 0x$2/;
michael@0:     # Only decimal constants on the line, no hex !
michael@0:     s/\bDCFD\b[ \t]+([0-9\.\-]+)/.double $1/;
michael@0: 
michael@0:     # Single hex constant on the line !
michael@0: #    s/\bDCFS\b[ \t]+0x([a-f0-9]{8})([a-f0-9]{8})/.long 0x$1, 0x$2/;
michael@0:     # Only decimal constants on the line, no hex !
michael@0: #    s/\bDCFS\b[ \t]+([0-9\.\-]+)/.double $1/;
michael@0:     s/\bDCFS[ \t]+0x/.word 0x/;
michael@0:     s/\bDCFS\b/.float/;
michael@0: 
michael@0:     s/^([A-Za-z_]\w*)[ \t]+DCD/$1 .word/;
michael@0:     s/\bDCD\b/.word/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+DCW/$1 .short/;
michael@0:     s/\bDCW\b/.short/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+DCB/$1 .byte/;
michael@0:     s/\bDCB\b/.byte/;
michael@0:     s/^([A-Za-z_]\w*)[ \t]+\%/.comm $1,/;
michael@0:     s/^[A-Za-z_\.]\w+/$&:/;
michael@0:     s/^(\d+)/$1:/;
michael@0:     s/\%(\d+)/$1b_or_f/;
michael@0:     s/\%[Bb](\d+)/$1b/;
michael@0:     s/\%[Ff](\d+)/$1f/;
michael@0:     s/\%[Ff][Tt](\d+)/$1f/;
michael@0:     s/&([\dA-Fa-f]+)/0x$1/;
michael@0:     if ( /\b2_[01]+\b/ ) {
michael@0:       s/\b2_([01]+)\b/conv$1&&&&/g;
michael@0:       while ( /[01][01][01][01]&&&&/ ) {
michael@0:         s/0000&&&&/&&&&0/g;
michael@0:         s/0001&&&&/&&&&1/g;
michael@0:         s/0010&&&&/&&&&2/g;
michael@0:         s/0011&&&&/&&&&3/g;
michael@0:         s/0100&&&&/&&&&4/g;
michael@0:         s/0101&&&&/&&&&5/g;
michael@0:         s/0110&&&&/&&&&6/g;
michael@0:         s/0111&&&&/&&&&7/g;
michael@0:         s/1000&&&&/&&&&8/g;
michael@0:         s/1001&&&&/&&&&9/g;
michael@0:         s/1010&&&&/&&&&A/g;
michael@0:         s/1011&&&&/&&&&B/g;
michael@0:         s/1100&&&&/&&&&C/g;
michael@0:         s/1101&&&&/&&&&D/g;
michael@0:         s/1110&&&&/&&&&E/g;
michael@0:         s/1111&&&&/&&&&F/g;
michael@0:       }
michael@0:       s/000&&&&/&&&&0/g;
michael@0:       s/001&&&&/&&&&1/g;
michael@0:       s/010&&&&/&&&&2/g;
michael@0:       s/011&&&&/&&&&3/g;
michael@0:       s/100&&&&/&&&&4/g;
michael@0:       s/101&&&&/&&&&5/g;
michael@0:       s/110&&&&/&&&&6/g;
michael@0:       s/111&&&&/&&&&7/g;
michael@0:       s/00&&&&/&&&&0/g;
michael@0:       s/01&&&&/&&&&1/g;
michael@0:       s/10&&&&/&&&&2/g;
michael@0:       s/11&&&&/&&&&3/g;
michael@0:       s/0&&&&/&&&&0/g;
michael@0:       s/1&&&&/&&&&1/g;
michael@0:       s/conv&&&&/0x/g;
michael@0:     }
michael@0: 
michael@0:     if ( /commandline/)
michael@0:     {
michael@0:         if( /-bigend/)
michael@0:         {
michael@0:             $bigend=1;
michael@0:         }
michael@0:     }
michael@0: 
michael@0:     if ( /\bDCDU\b/ )
michael@0:     {
michael@0:         my $cmd=$_;
michael@0:         my $value;
michael@0:         my $w1;
michael@0:         my $w2;
michael@0:         my $w3;
michael@0:         my $w4;
michael@0: 
michael@0:         s/\s+DCDU\b/@ $&/;
michael@0: 
michael@0:         $cmd =~ /\bDCDU\b\s+0x(\d+)/;
michael@0:         $value = $1;
michael@0:         $value =~ /(\w\w)(\w\w)(\w\w)(\w\w)/;
michael@0:         $w1 = $1;
michael@0:         $w2 = $2;
michael@0:         $w3 = $3;
michael@0:         $w4 = $4;
michael@0: 
michael@0:         if( $bigend ne "")
michael@0:         {
michael@0:             # big endian
michael@0: 
michael@0:             print "        .byte      0x".$w1;
michael@0:             print "        .byte      0x".$w2;
michael@0:             print "        .byte      0x".$w3;
michael@0:             print "        .byte      0x".$w4;
michael@0:         }
michael@0:         else
michael@0:         {
michael@0:             # little endian
michael@0: 
michael@0:             print "        .byte      0x".$w4;
michael@0:             print "        .byte      0x".$w3;
michael@0:             print "        .byte      0x".$w2;
michael@0:             print "        .byte      0x".$w1;
michael@0:         }
michael@0: 
michael@0:     }
michael@0: 
michael@0: 
michael@0:     if ( /\badrl\b/i )
michael@0:     {
michael@0:         s/\badrl\s+(\w+)\s*,\s*(\w+)/ldr $1,=$2/i;
michael@0:         $addPadding = 1;
michael@0:     }
michael@0:     s/\bEND\b/@ END/;
michael@0: } continue {
michael@0:     printf ("%s", $_) if $printit;
michael@0:     if ($addPadding != 0)
michael@0:     {
michael@0:         printf ("   mov r0,r0\n");
michael@0:         $addPadding = 0;
michael@0:     }
michael@0: }
michael@0: #If we had a code section, mark that this object doesn't need an executable
michael@0: # stack.
michael@0: if ($nxstack) {
michael@0:     printf ("    .section\t.note.GNU-stack,\"\",\%\%progbits\n");
michael@0: }