the-tree/scripts/markup_oops.pl

0001 #!/usr/bin/env perl
0002 # SPDX-License-Identifier: GPL-2.0-only
0003
0004 use File::Basename;
0005 use Math::BigInt;
0006 use Getopt::Long;
0007
0008 # Copyright 2008, Intel Corporation
0009 #
0010 # This file is part of the Linux kernel
0011 #
0012 # Authors:
0013 #   Arjan van de Ven <arjan@linux.intel.com>
0014
0015
0016 my $cross_compile = "";
0017 my $vmlinux_name = "";
0018 my $modulefile = "";
0019
0020 # Get options
0021 Getopt::Long::GetOptions(
0022     'cross-compile|c=s' => \$cross_compile,
0023     'module|m=s'        => \$modulefile,
0024     'help|h'        => \&usage,
0025 ) || usage ();
0026 my $vmlinux_name = $ARGV[0];
0027 if (!defined($vmlinux_name)) {
0028     my $kerver = `uname -r`;
0029     chomp($kerver);
0030     $vmlinux_name = "/lib/modules/$kerver/build/vmlinux";
0031     print "No vmlinux specified, assuming $vmlinux_name\n";
0032 }
0033 my $filename = $vmlinux_name;
0034
0035 # Parse the oops to find the EIP value
0036
0037 my $target = "0";
0038 my $function;
0039 my $module = "";
0040 my $func_offset = 0;
0041 my $vmaoffset = 0;
0042
0043 my %regs;
0044
0045
0046 sub parse_x86_regs
0047 {
0048     my ($line) = @_;
0049     if ($line =~ /EAX: ([0-9a-f]+) EBX: ([0-9a-f]+) ECX: ([0-9a-f]+) EDX: ([0-9a-f]+)/) {
0050         $regs{"%eax"} = $1;
0051         $regs{"%ebx"} = $2;
0052         $regs{"%ecx"} = $3;
0053         $regs{"%edx"} = $4;
0054     }
0055     if ($line =~ /ESI: ([0-9a-f]+) EDI: ([0-9a-f]+) EBP: ([0-9a-f]+) ESP: ([0-9a-f]+)/) {
0056         $regs{"%esi"} = $1;
0057         $regs{"%edi"} = $2;
0058         $regs{"%esp"} = $4;
0059     }
0060     if ($line =~ /RAX: ([0-9a-f]+) RBX: ([0-9a-f]+) RCX: ([0-9a-f]+)/) {
0061         $regs{"%eax"} = $1;
0062         $regs{"%ebx"} = $2;
0063         $regs{"%ecx"} = $3;
0064     }
0065     if ($line =~ /RDX: ([0-9a-f]+) RSI: ([0-9a-f]+) RDI: ([0-9a-f]+)/) {
0066         $regs{"%edx"} = $1;
0067         $regs{"%esi"} = $2;
0068         $regs{"%edi"} = $3;
0069     }
0070     if ($line =~ /RBP: ([0-9a-f]+) R08: ([0-9a-f]+) R09: ([0-9a-f]+)/) {
0071         $regs{"%r08"} = $2;
0072         $regs{"%r09"} = $3;
0073     }
0074     if ($line =~ /R10: ([0-9a-f]+) R11: ([0-9a-f]+) R12: ([0-9a-f]+)/) {
0075         $regs{"%r10"} = $1;
0076         $regs{"%r11"} = $2;
0077         $regs{"%r12"} = $3;
0078     }
0079     if ($line =~ /R13: ([0-9a-f]+) R14: ([0-9a-f]+) R15: ([0-9a-f]+)/) {
0080         $regs{"%r13"} = $1;
0081         $regs{"%r14"} = $2;
0082         $regs{"%r15"} = $3;
0083     }
0084 }
0085
0086 sub reg_name
0087 {
0088     my ($reg) = @_;
0089     $reg =~ s/r(.)x/e\1x/;
0090     $reg =~ s/r(.)i/e\1i/;
0091     $reg =~ s/r(.)p/e\1p/;
0092     return $reg;
0093 }
0094
0095 sub process_x86_regs
0096 {
0097     my ($line, $cntr) = @_;
0098     my $str = "";
0099     if (length($line) < 40) {
0100         return ""; # not an asm istruction
0101     }
0102
0103     # find the arguments to the instruction
0104     if ($line =~ /([0-9a-zA-Z\,\%\(\)\-\+]+)$/) {
0105         $lastword = $1;
0106     } else {
0107         return "";
0108     }
0109
0110     # we need to find the registers that get clobbered,
0111     # since their value is no longer relevant for previous
0112     # instructions in the stream.
0113
0114     $clobber = $lastword;
0115     # first, remove all memory operands, they're read only
0116     $clobber =~ s/\([a-z0-9\%\,]+\)//g;
0117     # then, remove everything before the comma, thats the read part
0118     $clobber =~ s/.*\,//g;
0119
0120     # if this is the instruction that faulted, we haven't actually done
0121     # the write yet... nothing is clobbered.
0122     if ($cntr == 0) {
0123         $clobber = "";
0124     }
0125
0126     foreach $reg (keys(%regs)) {
0127         my $clobberprime = reg_name($clobber);
0128         my $lastwordprime = reg_name($lastword);
0129         my $val = $regs{$reg};
0130         if ($val =~ /^[0]+$/) {
0131             $val = "0";
0132         } else {
0133             $val =~ s/^0*//;
0134         }
0135
0136         # first check if we're clobbering this register; if we do
0137         # we print it with a =>, and then delete its value
0138         if ($clobber =~ /$reg/ || $clobberprime =~ /$reg/) {
0139             if (length($val) > 0) {
0140                 $str = $str . " $reg => $val ";
0141             }
0142             $regs{$reg} = "";
0143             $val = "";
0144         }
0145         # now check if we're reading this register
0146         if ($lastword =~ /$reg/ || $lastwordprime =~ /$reg/) {
0147             if (length($val) > 0) {
0148                 $str = $str . " $reg = $val ";
0149             }
0150         }
0151     }
0152     return $str;
0153 }
0154
0155 # parse the oops
0156 while (<STDIN>) {
0157     my $line = $_;
0158     if ($line =~ /EIP: 0060:\[\<([a-z0-9]+)\>\]/) {
0159         $target = $1;
0160     }
0161     if ($line =~ /RIP: 0010:\[\<([a-z0-9]+)\>\]/) {
0162         $target = $1;
0163     }
0164     if ($line =~ /EIP is at ([a-zA-Z0-9\_]+)\+0x([0-9a-f]+)\/0x[a-f0-9]/) {
0165         $function = $1;
0166         $func_offset = $2;
0167     }
0168     if ($line =~ /RIP: 0010:\[\<[0-9a-f]+\>\]  \[\<[0-9a-f]+\>\] ([a-zA-Z0-9\_]+)\+0x([0-9a-f]+)\/0x[a-f0-9]/) {
0169         $function = $1;
0170         $func_offset = $2;
0171     }
0172
0173     # check if it's a module
0174     if ($line =~ /EIP is at ([a-zA-Z0-9\_]+)\+(0x[0-9a-f]+)\/0x[a-f0-9]+\W\[([a-zA-Z0-9\_\-]+)\]/) {
0175         $module = $3;
0176     }
0177     if ($line =~ /RIP: 0010:\[\<[0-9a-f]+\>\]  \[\<[0-9a-f]+\>\] ([a-zA-Z0-9\_]+)\+(0x[0-9a-f]+)\/0x[a-f0-9]+\W\[([a-zA-Z0-9\_\-]+)\]/) {
0178         $module = $3;
0179     }
0180     parse_x86_regs($line);
0181 }
0182
0183 my $decodestart = Math::BigInt->from_hex("0x$target") - Math::BigInt->from_hex("0x$func_offset");
0184 my $decodestop = Math::BigInt->from_hex("0x$target") + 8192;
0185 if ($target eq "0") {
0186     print "No oops found!\n";
0187     usage();
0188 }
0189
0190 # if it's a module, we need to find the .ko file and calculate a load offset
0191 if ($module ne "") {
0192     if ($modulefile eq "") {
0193         $modulefile = `modinfo -F filename $module`;
0194         chomp($modulefile);
0195     }
0196     $filename = $modulefile;
0197     if ($filename eq "") {
0198         print "Module .ko file for $module not found. Aborting\n";
0199         exit;
0200     }
0201     # ok so we found the module, now we need to calculate the vma offset
0202     open(FILE, $cross_compile."objdump -dS $filename |") || die "Cannot start objdump";
0203     while (<FILE>) {
0204         if ($_ =~ /^([0-9a-f]+) \<$function\>\:/) {
0205             my $fu = $1;
0206             $vmaoffset = Math::BigInt->from_hex("0x$target") - Math::BigInt->from_hex("0x$fu") - Math::BigInt->from_hex("0x$func_offset");
0207         }
0208     }
0209     close(FILE);
0210 }
0211
0212 my $counter = 0;
0213 my $state   = 0;
0214 my $center  = -1;
0215 my @lines;
0216 my @reglines;
0217
0218 sub InRange {
0219     my ($address, $target) = @_;
0220     my $ad = "0x".$address;
0221     my $ta = "0x".$target;
0222     my $delta = Math::BigInt->from_hex($ad) - Math::BigInt->from_hex($ta);
0223
0224     if (($delta > -4096) && ($delta < 4096)) {
0225         return 1;
0226     }
0227     return 0;
0228 }
0229
0230
0231
0232 # first, parse the input into the lines array, but to keep size down,
0233 # we only do this for 4Kb around the sweet spot
0234
0235 open(FILE, $cross_compile."objdump -dS --adjust-vma=$vmaoffset --start-address=$decodestart --stop-address=$decodestop $filename |") || die "Cannot start objdump";
0236
0237 while (<FILE>) {
0238     my $line = $_;
0239     chomp($line);
0240     if ($state == 0) {
0241         if ($line =~ /^([a-f0-9]+)\:/) {
0242             if (InRange($1, $target)) {
0243                 $state = 1;
0244             }
0245         }
0246     }
0247     if ($state == 1) {
0248         if ($line =~ /^([a-f0-9][a-f0-9][a-f0-9][a-f0-9][a-f0-9][a-f0-9]+)\:/) {
0249             my $val = $1;
0250             if (!InRange($val, $target)) {
0251                 last;
0252             }
0253             if ($val eq $target) {
0254                 $center = $counter;
0255             }
0256         }
0257         $lines[$counter] = $line;
0258
0259         $counter = $counter + 1;
0260     }
0261 }
0262
0263 close(FILE);
0264
0265 if ($counter == 0) {
0266     print "No matching code found \n";
0267     exit;
0268 }
0269
0270 if ($center == -1) {
0271     print "No matching code found \n";
0272     exit;
0273 }
0274
0275 my $start;
0276 my $finish;
0277 my $codelines = 0;
0278 my $binarylines = 0;
0279 # now we go up and down in the array to find how much we want to print
0280
0281 $start = $center;
0282
0283 while ($start > 1) {
0284     $start = $start - 1;
0285     my $line = $lines[$start];
0286     if ($line =~ /^([a-f0-9]+)\:/) {
0287         $binarylines = $binarylines + 1;
0288     } else {
0289         $codelines = $codelines + 1;
0290     }
0291     if ($codelines > 10) {
0292         last;
0293     }
0294     if ($binarylines > 20) {
0295         last;
0296     }
0297 }
0298
0299
0300 $finish = $center;
0301 $codelines = 0;
0302 $binarylines = 0;
0303 while ($finish < $counter) {
0304     $finish = $finish + 1;
0305     my $line = $lines[$finish];
0306     if ($line =~ /^([a-f0-9]+)\:/) {
0307         $binarylines = $binarylines + 1;
0308     } else {
0309         $codelines = $codelines + 1;
0310     }
0311     if ($codelines > 10) {
0312         last;
0313     }
0314     if ($binarylines > 20) {
0315         last;
0316     }
0317 }
0318
0319
0320 my $i;
0321
0322
0323 # start annotating the registers in the asm.
0324 # this goes from the oopsing point back, so that the annotator
0325 # can track (opportunistically) which registers got written and
0326 # whos value no longer is relevant.
0327
0328 $i = $center;
0329 while ($i >= $start) {
0330     $reglines[$i] = process_x86_regs($lines[$i], $center - $i);
0331     $i = $i - 1;
0332 }
0333
0334 $i = $start;
0335 while ($i < $finish) {
0336     my $line;
0337     if ($i == $center) {
0338         $line =  "*$lines[$i] ";
0339     } else {
0340         $line =  " $lines[$i] ";
0341     }
0342     print $line;
0343     if (defined($reglines[$i]) && length($reglines[$i]) > 0) {
0344         my $c = 60 - length($line);
0345         while ($c > 0) { print " "; $c = $c - 1; };
0346         print "| $reglines[$i]";
0347     }
0348     if ($i == $center) {
0349         print "<--- faulting instruction";
0350     }
0351     print "\n";
0352     $i = $i +1;
0353 }
0354
0355 sub usage {
0356     print <<EOT;
0357 Usage:
0358   dmesg | perl $0 [OPTION] [VMLINUX]
0359
0360 OPTION:
0361   -c, --cross-compile CROSS_COMPILE Specify the prefix used for toolchain.
0362   -m, --module MODULE_DIRNAME       Specify the module filename.
0363   -h, --help                Help.
0364 EOT
0365     exit;
0366 }