[crashlog] Implement parser for JSON encoded crashlogs

Add a parser for JSON crashlogs. The CrashLogParser now defers to either
the JSONCrashLogParser or the TextCrashLogParser. It first tries to
interpret the input as JSON, and if that fails falling back to the
textual parser.

Differential revision: https://reviews.llvm.org/D91130

9 files changed, 280 insertions, 1 deletions

diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/Assertion.check b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/Assertion.check
new file mode 100644
index 0000000..d92fb50
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/Assertion.check
@@ -0,0 +1 @@
+# CHECK-NOT: AssertionError
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.crash b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.crash
new file mode 100644
index 0000000..27ffd9e
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.crash
@@ -0,0 +1,49 @@
+Process:               a.out [21606]
+Path:                  /private/tmp/a.out
+Identifier:            a.out
+Version:               0
+Code Type:             X86-64 (Native)
+Parent Process:        fish [88883]
+User ID:               501
+
+Date/Time:             2020-11-11 14:47:34.600 -0800
+OS Version:            macOS 11.0.1
+Report Version:        12
+Bridge OS Version:     redacted
+Anonymous UUID:        DCEF35CB-68D5-F524-FF13-060901F52EA8
+
+
+Time Awake Since Boot: 400000 seconds
+
+System Integrity Protection: enabled
+
+Crashed Thread:        0  Dispatch queue: com.apple.main-thread
+
+Exception Type:        EXC_BAD_ACCESS (SIGSEGV)
+Exception Codes:       KERN_INVALID_ADDRESS at 0x0000000000000000
+Exception Note:        EXC_CORPSE_NOTIFY
+
+Termination Signal:    Segmentation fault: 11
+Termination Reason:    Namespace SIGNAL, Code 0xb
+Terminating Process:   exc handler [21606]
+
+Thread 0 Crashed:: Dispatch queue: com.apple.main-thread
+0   a.out                         	@foo@ foo + 16 (test.c:3)
+1   a.out                         	@bar@ bar + 9 (test.c:6)
+2   a.out                         	@main@ main + 20 (test.c:8)
+3   libdyld.dylib                 	0x0000000100000000 start + 1
+
+Thread 0 crashed with X86 Thread State (64-bit):
+  rax: 0x0000000000000000  rbx: 0x0000000000000000  rcx: 0x00007ffee42d81d0  rdx: 0x00007ffee42d8080
+  rdi: 0x0000000000000001  rsi: 0x00007ffee42d8070  rbp: 0x00007ffee42d8020  rsp: 0x00007ffee42d8020
+   r8: 0x0000000000000000   r9: 0x0000000000000000  r10: 0x0000000000000000  r11: 0x0000000000000000
+  r12: 0x0000000000000000  r13: 0x0000000000000000  r14: 0x0000000000000000  r15: 0x0000000000000000
+  rip: 0x000000010b92af70  rfl: 0x0000000000010202  cr2: 0x0000000000000000
+
+Logical CPU:     2
+Error Code:      0x00000006 (no mapping for user data write)
+Trap Number:     14
+
+
+Binary Images:
+       0x100000000 -        0x200000000 +a.out (0) <@UUID@> @EXEC@
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.ips b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.ips
new file mode 100644
index 0000000..703acce
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/a.out.ips
@@ -0,0 +1,96 @@
+{"app_name":"a.out","timestamp":"2020-11-11 16:12:18.00 -0800","app_version":"","slice_uuid":"9b76648c-9b4e-33a9-a97e-10856e911631","build_version":"","platform":1,"share_with_app_devs":1,"is_first_party":1,"bug_type":"309","os_version":"macOS 11.0.1","incident_id":"598C4706-28B0-4D96-A2F9-AE6973BEC635","name":"a.out"}
+{
+  "uptime" : 180,
+  "procLaunch" : "2020-11-11 16:12:12.4375 -0800",
+  "procRole" : "Unspecified",
+  "exception" : {
+    "type" : "EXC_BAD_ACCESS",
+    "signal" : "SIGSEGV",
+    "subtype" : "KERN_INVALID_ADDRESS at 0x00000000"
+  },
+  "userID" : 501,
+  "modelCode" : "iMacPro1,1",
+  "coalitionID" : 471,
+  "osVersion" : {
+    "train" : "macOS 11.0.1",
+    "build" : "",
+    "releaseType" : ""
+  },
+  "captureTime" : "2020-11-11 16:12:12.6267 -0800",
+  "incident" : "598C4706-28B0-4D96-A2F9-AE6973BEC635",
+  "pid" : 2187,
+  "cpuType" : "X86-64",
+  "procName" : "a.out",
+  "procPath" : "\/private\/tmp\/a.out",
+  "parentProc" : "fish",
+  "parentPid" : 1651,
+  "coalitionName" : "io.alacritty",
+  "crashReporterKey" : "DCEF35CB-68D5-F524-FF13-060901F52EA8",
+  "responsiblePid" : 428,
+  "responsibleProc" : "alacritty",
+  "bridgeVersion" : {"build":"","train":""},
+  "sip" : "enabled",
+  "is_corpse" : 1,
+  "termination" : {"reason":"Namespace SIGNAL, Code 0xb","signal":"Segmentation fault: 11","byProc":"exc handler","code":11,"namespace":"SIGNAL","byPid":2187,"flags":0},
+  "asi" : ["dyld2 mode"],
+  "extMods" : {"caller":{"thread_create":0,"thread_set_state":0,"task_for_pid":0},"system":{"thread_create":0,"thread_set_state":0,"task_for_pid":2067},"targeted":{"thread_create":0,"thread_set_state":0,"task_for_pid":0},"warnings":0},
+  "threads" : [{"triggered":true,"id":22172,"queue":"com.apple.main-thread","frames":[[0,16240],[0,16265],[0,16292],[1,87601]]}],
+  "threadState" : {
+  "r13" : 0,
+  "rax" : 0,
+  "rflags" : 66054,
+  "cpu" : 6,
+  "rsi" : 140732908048520,
+  "r14" : 0,
+  "trap_description" : "(no mapping for user data write)",
+  "r8" : 0,
+  "cr2" : 0,
+  "rdx" : 140732908048536,
+  "r10" : 0,
+  "r9" : 0,
+  "r15" : 0,
+  "rbx" : 0,
+  "trap" : 14,
+  "err" : 6,
+  "r11" : 0,
+  "rip" : 4307689328,
+  "rbp" : 140732908048432,
+  "rsp" : 140732908048432,
+  "r12" : 0,
+  "rcx" : 140732908048880,
+  "flavor" : "x86_THREAD_STATE",
+  "rdi" : 1
+},
+  "usedImages" : [
+  [
+    "@UUID@",
+    0,
+    "P"
+  ],
+  [
+    "6a1f593e-3705-314d-bb40-e7f9d502bf81",
+    140733737017344,
+    "P"
+  ]
+],
+  "legacyInfo" : {
+  "imageExtraInfo" : [
+    {
+      "size" : 16384,
+      "arch" : "x86_64",
+      "path" : "@EXEC@",
+      "name" : "@NAME@"
+    },
+    {
+      "size" : 241664,
+      "arch" : "x86_64",
+      "path" : "\/usr\/lib\/system\/libdyld.dylib",
+      "name" : "libdyld.dylib"
+    }
+  ],
+  "threadTriggered" : {
+    "index" : 0,
+    "queue" : "com.apple.main-thread"
+  }
+}
+}
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/test.c b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/test.c
new file mode 100644
index 0000000..3f0819b
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/Inputs/test.c
@@ -0,0 +1,8 @@
+void foo() {
+  int *i = 0;
+  *i = 1;
+}
+
+void bar() { foo(); }
+
+int main(int argc, char **argv) { bar(); }
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/json.test b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/json.test
new file mode 100644
index 0000000..c92be0b
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/json.test
@@ -0,0 +1,10 @@
+# RUN: %clang_host -g %S/Inputs/test.c -o %t.out
+# RUN: cp %S/Inputs/a.out.ips %t.crash
+# RUN: python %S/patch-crashlog.py %t.out %t.crash
+# RUN: %lldb %t.out -o 'command script import lldb.macosx.crashlog' -o 'crashlog %t.crash' 2>&1 | FileCheck %s
+
+# CHECK: Thread[0] EXC_BAD_ACCESS (SIGSEGV) (KERN_INVALID_ADDRESS at 0x00000000)
+# CHECK: [  0] {{.*}}out`foo + 16 at test.c
+# CHECK: [  1] {{.*}}out`bar + 8 at test.c
+# CHECK: [  2] {{.*}}out`main + 19 at test.c
+# CHECK: [  3] {{.*}}start
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/parser_json.test b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/parser_json.test
new file mode 100644
index 0000000..50da725
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/parser_json.test
@@ -0,0 +1,45 @@
+#                                                                 -*- python -*-
+# RUN: cd %S/../../../../../examples/python && cat %s | %lldb 2>&1 > %t.out
+# RUN: cat %t.out | FileCheck %S/Inputs/Assertion.check
+script
+import crashlog
+import json
+
+parser = crashlog.JSONCrashLogParser("", "", False)
+
+process_info_json = json.loads('{"pid" : 287, "procName" : "mediaserverd", "procPath" : "\/usr\/sbin\/mediaserverd"}')
+parser.parse_process_info(process_info_json)
+
+assert parser.crashlog.process_id == 287
+assert parser.crashlog.process_identifier == "mediaserverd"
+assert parser.crashlog.process_path == "/usr/sbin/mediaserverd"
+
+crash_reason_json = json.loads('{"type" : "EXC_BAD_ACCESS", "signal" : "SIGSEGV", "subtype" : "KERN_INVALID_ADDRESS"}')
+assert parser.parse_crash_reason(crash_reason_json) == "EXC_BAD_ACCESS (SIGSEGV) (KERN_INVALID_ADDRESS)"
+
+crash_reason_json = json.loads('{"type" : "EXC_BAD_ACCESS", "signal" : "SIGSEGV"}')
+assert parser.parse_crash_reason(crash_reason_json) == "EXC_BAD_ACCESS (SIGSEGV)"
+
+crash_reason_json = json.loads('{"type" : "EXC_BAD_ACCESS", "signal" : "SIGSEGV", "codes" : "0x0000000000000000, 0x0000000000000000"}')
+assert parser.parse_crash_reason(crash_reason_json) == "EXC_BAD_ACCESS (SIGSEGV) (0x0000000000000000, 0x0000000000000000)"
+
+thread_state_json = json.loads('{"x":[268451845,117442566],"lr":7309751904,"cpsr":1073741824,"fp":6093236784,"sp":6093236704,"esr":1442840704,"pc":7309755088}')
+registers = parser.parse_thread_registers(thread_state_json)
+assert registers['x0'] == 268451845
+assert registers['x1'] == 117442566
+assert registers['lr'] == 7309751904
+assert registers['cpsr'] ==1073741824
+assert registers['fp'] == 6093236784
+assert registers['sp'] == 6093236704
+assert registers['esr'] == 1442840704
+assert registers['pc'] == 7309755088
+
+parser.data = json.loads('{"usedImages":[["f4d85377-f215-3da3-921e-3fe870e622e9",7309737984,"P"]],"legacyInfo":{"imageExtraInfo":[{"size":204800,"arch":"arm64e","path":"/usr/lib/system/libsystem_kernel.dylib","name":"libsystem_kernel.dylib"}]}}')
+thread_json = json.loads('[{"triggered":true,"id":3835,"queue":"com.apple.bwgraph.devicevendor","frames":[[0,101472],[0,408892]]}]')
+parser.parse_threads(thread_json)
+assert len(parser.crashlog.threads) == 1
+assert parser.crashlog.threads[0].queue == "com.apple.bwgraph.devicevendor"
+assert len(parser.crashlog.threads[0].frames) == 2
+assert parser.crashlog.threads[0].frames[0].pc == 7309839456
+assert parser.crashlog.threads[0].frames[0].description == 101472
+exit()
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/crashlog.test b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/parser_text.test
index 6ac9392..7251d85 100644
--- a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/crashlog.test
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/parser_text.test
@@ -4,7 +4,7 @@
 # CHECK-LABEL: {{S}}KIP BEYOND CHECKS
 script
 import crashlog
-crash_log_parser = crashlog.CrashLogParser
+crash_log_parser = crashlog.TextCrashLogParser
 crash_log = crashlog.CrashLog
 images = [
 "0x10b60b000 - 0x10f707fff com.apple.LLDB.framework (1.1000.11.38.2 - 1000.11.38.2) <96E36F5C-1A83-39A1-8713-5FDD9701C3F1> /Applications/Xcode.app/Contents/SharedFrameworks/LLDB.framework/LLDB",
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/patch-crashlog.py b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/patch-crashlog.py
new file mode 100644
index 0000000..ba69547
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/patch-crashlog.py
@@ -0,0 +1,60 @@
+#!/usr/bin/env python
+
+import json
+import os
+import re
+import subprocess
+import sys
+
+
+class CrashLogPatcher:
+
+    SYMBOL_REGEX = re.compile(r'^([0-9a-fA-F]+) T _(.*)$')
+    UUID_REGEX = re.compile(r'UUID: ([-0-9a-fA-F]+) \(([^\(]+)\) .*')
+
+    def __init__(self, data, binary, offsets):
+        self.data = data
+        self.binary = binary
+        self.offsets = offsets
+
+    def patch_executable(self):
+        self.data = self.data.replace("@EXEC@", self.binary)
+        self.data = self.data.replace("@NAME@", os.path.basename(self.binary))
+
+    def patch_uuid(self):
+        output = subprocess.check_output(['dwarfdump', '--uuid', self.binary])
+        m = self.UUID_REGEX.match(output)
+        if m:
+            self.data = self.data.replace("@UUID@", m.group(1))
+
+    def patch_addresses(self):
+        if not self.offsets:
+            return
+        output = subprocess.check_output(['nm', self.binary])
+        for line in output.splitlines():
+            m = self.SYMBOL_REGEX.match(line)
+            if m:
+                address = m.group(1)
+                symbol = m.group(2)
+                if symbol in self.offsets:
+                    patch_addr = int(m.group(1), 16) + int(
+                        self.offsets[symbol])
+                    self.data = self.data.replace("@{}@".format(symbol),
+                                                  str(hex(patch_addr)))
+
+
+if __name__ == '__main__':
+    binary = sys.argv[1]
+    crashlog = sys.argv[2]
+    offsets = json.loads(sys.argv[3]) if len(sys.argv) > 3 else None
+
+    with open(crashlog, 'r') as file:
+        data = file.read()
+
+    p = CrashLogPatcher(data, binary, offsets)
+    p.patch_executable()
+    p.patch_uuid()
+    p.patch_addresses()
+
+    with open(crashlog, 'w') as file:
+        file.write(p.data)
diff --git a/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/text.test b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/text.test
new file mode 100644
index 0000000..1375784
--- /dev/null
+++ b/lldb/test/Shell/ScriptInterpreter/Python/Crashlog/text.test
@@ -0,0 +1,10 @@
+# RUN: %clang_host -g %S/Inputs/test.c -o %t.out
+# RUN: cp %S/Inputs/a.out.crash %t.crash
+# RUN: python %S/patch-crashlog.py %t.out %t.crash '{"main":20, "bar":9, "foo":16}'
+# RUN: %lldb %t.out -o 'command script import lldb.macosx.crashlog' -o 'crashlog %t.crash' 2>&1 | FileCheck %s
+
+# CHECK: Thread[0] EXC_BAD_ACCESS (SIGSEGV) (KERN_INVALID_ADDRESS at 0x0000000000000000)
+# CHECK: [  0] {{.*}}out`foo + 16 at test.c
+# CHECK: [  1] {{.*}}out`bar + 8 at test.c
+# CHECK: [  2] {{.*}}out`main + 19 at test.c
+# CHECK: [  3] {{.*}}start + 1