aniani/nasm
0
0
Fork 1
mirror of https://github.com/netwide-assembler/nasm.git synced 2025-07-24 10:25:42 -04:00
Code

tokhash.pl: zero all the fields for a not-found token

Browse Source 
Make sure we zero all the token fields if we don't find something in
the hash.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
This commit is contained in:
H. Peter Anvin 2018-12-29 20:14:50 -08:00
parent 6a4f0b36c8
commit ef4f23d76a
1 changed files with 37 additions and 26 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

63
asm/tokhash.pl
View File

@ -65,14 +65,17 @@ while (defined($line = <ID>)) {
# Single instruction token # Single instruction token
if (!defined($tokens{$token})) { if (!defined($tokens{$token})) {
$tokens{$token} = scalar @tokendata; $tokens{$token} = scalar @tokendata;
push(@tokendata, "\"${token}\", TOKEN_INSN, C_none, 0, I_${insn}"); push(@tokendata, "\"${token}\", ".length($token).
", TOKEN_INSN, C_none, 0, I_${insn}");
} }
} else { } else {
# Conditional instruction # Conditional instruction
foreach $cc (@conditions) { foreach $cc (@conditions) {
if (!defined($tokens{$token.$cc})) { my $etok = $token.$cc;
$tokens{$token.$cc} = scalar @tokendata; if (!defined($tokens{$etok})) {
push(@tokendata, "\"${token}${cc}\", TOKEN_INSN, C_\U$cc\E, 0, I_${insn}"); $tokens{$etok} = scalar @tokendata;
push(@tokendata, "\"${etok}\", ".length($etok).
", TOKEN_INSN, C_\U$cc\E, 0, I_${insn}");
} }
} }
} }
@ -105,11 +108,8 @@ while (defined($line = <RD>)) {
die "Duplicate definition: $reg\n"; die "Duplicate definition: $reg\n";
} }
$tokens{$reg} = scalar @tokendata; $tokens{$reg} = scalar @tokendata;
if ($reg_flag eq '') { $reg_flag = '0' if ($reg_flag eq '');
push(@tokendata, "\"${reg}\", TOKEN_REG, 0, 0, R_\U${reg}\E"); push(@tokendata, "\"${reg}\", ".length($reg).", TOKEN_REG, 0, ${reg_flag}, R_\U${reg}\E");
} else {
push(@tokendata, "\"${reg}\", TOKEN_REG, 0, ${reg_flag}, R_\U${reg}\E");
}
if (defined($reg_prefix)) { if (defined($reg_prefix)) {
$reg_nr++; $reg_nr++;
@ -153,23 +153,23 @@ while (defined($line = <TD>)) {
$data =~ s/\*/\U$token/g; $data =~ s/\*/\U$token/g;
push(@tokendata, "\"$token\", $data"); push(@tokendata, "\"$token\", ".length($token).", $data");
} }
} }
close(TD); close(TD);
$max_len = 0;
foreach $token (keys(%tokens)) {
if (length($token) > $max_len) {
$max_len = length($token);
}
}
if ($output eq 'h') { if ($output eq 'h') {
# #
# tokens.h # tokens.h
# #
$max_len = 0;
foreach $token (keys(%tokens)) {
if (length($token) > $max_len) {
$max_len = length($token);
}
}
print "/*\n"; print "/*\n";
print " * This file is generated from insns.dat, regs.dat and token.dat\n"; print " * This file is generated from insns.dat, regs.dat and token.dat\n";
print " * by tokhash.pl; do not edit.\n"; print " * by tokhash.pl; do not edit.\n";
@ -218,9 +218,10 @@ if ($output eq 'h') {
# 64-bit machines and 12 bytes on 32-bit machines. # 64-bit machines and 12 bytes on 32-bit machines.
print "struct tokendata {\n"; print "struct tokendata {\n";
print " const char *string;\n"; print " const char *string;\n";
print " uint16_t len;\n";
print " int16_t tokentype;\n"; print " int16_t tokentype;\n";
print " int8_t aux;\n"; print " int16_t aux;\n";
print " int8_t tokflag;\n"; print " uint16_t tokflag;\n";
print " int32_t num;\n"; print " int32_t num;\n";
print "};\n"; print "};\n";
print "\n"; print "\n";
@ -247,7 +248,8 @@ if ($output eq 'h') {
} }
print " };\n"; print " };\n";
printf " static const struct tokendata tokendata[%d] = {\n", scalar(@tokendata); printf " static const struct tokendata tokendata[%d] = {\n",
scalar(@tokendata);
foreach $d (@tokendata) { foreach $d (@tokendata) {
print " { ", $d, " },\n"; print " { ", $d, " },\n";
} }
@ -255,29 +257,38 @@ if ($output eq 'h') {
print " uint32_t k1, k2;\n"; print " uint32_t k1, k2;\n";
print " uint64_t crc;\n"; print " uint64_t crc;\n";
print " size_t len;\n";
# For correct overflow behavior, "ix" should be unsigned of the same # For correct overflow behavior, "ix" should be unsigned of the same
# width as the hash arrays. # width as the hash arrays.
print " uint16_t ix;\n"; print " uint16_t ix;\n";
print " const struct tokendata *data;\n"; print " const struct tokendata *data;\n";
print "\n"; print "\n";
printf " tv->t_flag = 0;\n"; print " len = strlen(token);\n";
printf " crc = crc64(UINT64_C(0x%08x%08x), token);\n", print " if (unlikely(len > $max_len))\n";
print " goto notfound;\n";
print "\n";
printf " crc = crc64b(UINT64_C(0x%08x%08x), token, len);\n",
$$sv[0], $$sv[1]; $$sv[0], $$sv[1];
print " k1 = (uint32_t)crc;\n"; print " k1 = (uint32_t)crc;\n";
print " k2 = (uint32_t)(crc >> 32);\n"; print " k2 = (uint32_t)(crc >> 32);\n";
print "\n"; print "\n";
printf " ix = hash1[k1 & 0x%x] + hash2[k2 & 0x%x];\n", $n-1, $n-1; printf " ix = hash1[k1 & 0x%x] + hash2[k2 & 0x%x];\n", $n-1, $n-1;
printf " if (ix >= %d)\n", scalar(@tokendata); printf " if (ix >= %d)\n", scalar(@tokendata);
print " return tv->t_type = TOKEN_ID;\n"; print " goto notfound;\n";
print "\n"; print "\n";
print " data = &tokendata[ix];\n"; print " data = &tokendata[ix];\n";
print " if (data->len != len || memcmp(data->string, token, len))\n";
print " if (strcmp(data->string, token))\n"; print " goto notfound;\n";
print " return tv->t_type = TOKEN_ID;\n";
print "\n"; print "\n";
print " tv->t_integer = data->num;\n"; print " tv->t_integer = data->num;\n";
print " tv->t_inttwo = data->aux;\n"; print " tv->t_inttwo = data->aux;\n";
print " tv->t_flag = data->tokflag;\n"; print " tv->t_flag = data->tokflag;\n";
print " return tv->t_type = data->tokentype;\n"; print " return tv->t_type = data->tokentype;\n";
print "\n";
print "notfound:\n";
print " tv->t_integer = 0;\n";
print " tv->t_inttwo = 0;\n";
print " tv->t_flag = 0;\n";
print " return tv->t_type = TOKEN_ID;\n";
print "}\n"; print "}\n";
} }