aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorH. Peter Anvin <hpa@zytor.com>2018-12-29 20:14:50 -0800
committerH. Peter Anvin <hpa@zytor.com>2018-12-29 20:14:50 -0800
commitef4f23d76a6ce89c93b40150f48ba6803d42341e (patch)
treef239dbe58f0186c3bb10f3ea79707b054abef6ec
parent6a4f0b36c803f2028810022f6ad2d9c22d1483e7 (diff)
downloadnasm-ef4f23d76a6ce89c93b40150f48ba6803d42341e.tar.gz
nasm-ef4f23d76a6ce89c93b40150f48ba6803d42341e.tar.xz
nasm-ef4f23d76a6ce89c93b40150f48ba6803d42341e.zip
tokhash.pl: zero all the fields for a not-found token
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>
-rwxr-xr-xasm/tokhash.pl63
1 files changed, 37 insertions, 26 deletions
diff --git a/asm/tokhash.pl b/asm/tokhash.pl
index 186924cd..72f3fee4 100755
--- a/asm/tokhash.pl
+++ b/asm/tokhash.pl
@@ -65,14 +65,17 @@ while (defined($line = <ID>)) {
# Single instruction token
if (!defined($tokens{$token})) {
$tokens{$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 {
# Conditional instruction
foreach $cc (@conditions) {
- if (!defined($tokens{$token.$cc})) {
- $tokens{$token.$cc} = scalar @tokendata;
- push(@tokendata, "\"${token}${cc}\", TOKEN_INSN, C_\U$cc\E, 0, I_${insn}");
+ my $etok = $token.$cc;
+ if (!defined($tokens{$etok})) {
+ $tokens{$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";
}
$tokens{$reg} = scalar @tokendata;
- if ($reg_flag eq '') {
- push(@tokendata, "\"${reg}\", TOKEN_REG, 0, 0, R_\U${reg}\E");
- } else {
- push(@tokendata, "\"${reg}\", TOKEN_REG, 0, ${reg_flag}, R_\U${reg}\E");
- }
+ $reg_flag = '0' if ($reg_flag eq '');
+ push(@tokendata, "\"${reg}\", ".length($reg).", TOKEN_REG, 0, ${reg_flag}, R_\U${reg}\E");
if (defined($reg_prefix)) {
$reg_nr++;
@@ -153,23 +153,23 @@ while (defined($line = <TD>)) {
$data =~ s/\*/\U$token/g;
- push(@tokendata, "\"$token\", $data");
+ push(@tokendata, "\"$token\", ".length($token).", $data");
}
}
close(TD);
+$max_len = 0;
+foreach $token (keys(%tokens)) {
+ if (length($token) > $max_len) {
+ $max_len = length($token);
+ }
+}
+
if ($output eq 'h') {
#
# tokens.h
#
- $max_len = 0;
- foreach $token (keys(%tokens)) {
- if (length($token) > $max_len) {
- $max_len = length($token);
- }
- }
-
print "/*\n";
print " * This file is generated from insns.dat, regs.dat and token.dat\n";
print " * by tokhash.pl; do not edit.\n";
@@ -218,9 +218,10 @@ if ($output eq 'h') {
# 64-bit machines and 12 bytes on 32-bit machines.
print "struct tokendata {\n";
print " const char *string;\n";
+ print " uint16_t len;\n";
print " int16_t tokentype;\n";
- print " int8_t aux;\n";
- print " int8_t tokflag;\n";
+ print " int16_t aux;\n";
+ print " uint16_t tokflag;\n";
print " int32_t num;\n";
print "};\n";
print "\n";
@@ -247,7 +248,8 @@ if ($output eq 'h') {
}
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) {
print " { ", $d, " },\n";
}
@@ -255,29 +257,38 @@ if ($output eq 'h') {
print " uint32_t k1, k2;\n";
print " uint64_t crc;\n";
+ print " size_t len;\n";
# For correct overflow behavior, "ix" should be unsigned of the same
# width as the hash arrays.
print " uint16_t ix;\n";
print " const struct tokendata *data;\n";
print "\n";
- printf " tv->t_flag = 0;\n";
- printf " crc = crc64(UINT64_C(0x%08x%08x), token);\n",
+ print " len = strlen(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];
print " k1 = (uint32_t)crc;\n";
print " k2 = (uint32_t)(crc >> 32);\n";
print "\n";
printf " ix = hash1[k1 & 0x%x] + hash2[k2 & 0x%x];\n", $n-1, $n-1;
printf " if (ix >= %d)\n", scalar(@tokendata);
- print " return tv->t_type = TOKEN_ID;\n";
+ print " goto notfound;\n";
print "\n";
print " data = &tokendata[ix];\n";
-
- print " if (strcmp(data->string, token))\n";
- print " return tv->t_type = TOKEN_ID;\n";
+ print " if (data->len != len || memcmp(data->string, token, len))\n";
+ print " goto notfound;\n";
print "\n";
print " tv->t_integer = data->num;\n";
print " tv->t_inttwo = data->aux;\n";
print " tv->t_flag = data->tokflag;\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";
}