def _db_backtrace_to_satyr(self, db_backtrace):
stacktrace = satyr.Kerneloops()
if len(db_backtrace.threads) < 1:
self.log_warn("Backtrace #{0} has no usable threads"
.format(db_backtrace.id))
return None
db_thread = db_backtrace.threads[0]
if len(db_thread.frames) < 1:
self.log_warn("Thread #{0} has no usable frames"
.format(db_thread.id))
return None
for db_frame in db_thread.frames:
frame = satyr.KerneloopsFrame()
if db_frame.symbolsource.symbol is not None:
frame.function_name = db_frame.symbolsource.symbol.name
else:
frame.function_name = "??"
frame.address = db_frame.symbolsource.offset
frame.function_offset = db_frame.symbolsource.func_offset
frame.reliable = db_frame.reliable
if frame.address < 0:
frame.address += (1 << 64)
stacktrace.frames.append(frame)
if self.normalize:
stacktrace.normalize()
return stacktrace
def ureport1to2(ureport1):
"""
Convert uReport1 to uReport2. This has hardcodes plugin names for python,
coredump and kerneloops. It also adds a soft dependency on satyr library.
The process is best-effort and strange results may be obtained. Be sure to
validate the resulting uReport2 after conversion.
"""
# Too many branches
# pylint: disable-msg=R0912
ureport2 = {"ureport_version": 2, "problem": {}, "os": {}, "packages": []}
# make sure we are converting uReport1
if "ureport_version" not in ureport1 or ureport1["ureport_version"] != 1:
raise FafError("uReport1 is required")
# problem type
if "type" in ureport1:
if ureport1["type"].lower() == "userspace":
ureport2["problem"]["type"] = "core"
else:
ureport2["problem"]["type"] = ureport1["type"].lower()
# operating system - quite straightforward
if "os" in ureport1:
if "name" in ureport1["os"]:
ureport2["os"]["name"] = ureport1["os"]["name"].lower()
if "version" in ureport1["os"]:
ureport2["os"]["version"] = ureport1["os"]["version"]
if "architecture" in ureport1:
ureport2["os"]["architecture"] = ureport1["architecture"]
# selinux
if "selinux" in ureport1:
ureport2["os"]["selinux"] = {}
if "mode" in ureport1["selinux"]:
ureport2["os"]["selinux"]["mode"] = ureport1["selinux"]["mode"]
if "context" in ureport1["selinux"]:
ureport2["os"]["selinux"]["context"] = \
ureport1["selinux"]["context"]
if "policy_package" in ureport1["selinux"]:
newpkg = dict(ureport1["selinux"]["policy_package"])
newpkg["package_role"] = "selinux_policy"
ureport2["packages"].append(newpkg)
# component
if "component" in ureport1:
ureport2["problem"]["component"] = ureport1["component"]
# executable
if "executable" in ureport1:
ureport2["problem"]["executable"] = ureport1["executable"]
# reason + exception name / signal number
if "reason" in ureport1:
ureport2["reason"] = ureport1["reason"]
if ureport1["type"].lower() == "python":
ureport2["problem"]["exception_name"] = "reason"
if ureport1["type"].lower() == "userspace":
match = RE_SIGNAL.match(ureport1["reason"])
if match:
ureport2["problem"]["signal"] = int(match.group(1))
# just copy reporter - the format did not change
if "reporter" in ureport1:
ureport2["reporter"] = ureport1["reporter"]
# packages - the NEVRA dict remains the same
if "installed_package" in ureport1:
newpkg = dict(ureport1["installed_package"])
newpkg["package_role"] = "affected"
ureport2["packages"].append(newpkg)
if "running_package" in ureport1:
ureport2["packages"].append(dict(ureport1["running_package"]))
if "related_packages" in ureport1:
for relpkg in ureport1["related_packages"]:
if "running_package" in relpkg:
ureport2["packages"].append(dict(relpkg["running_package"]))
if "installed_package" in relpkg:
ureport2["packages"].append(dict(relpkg["installed_package"]))
# map core backtrace into uReport2 structures
if "core_backtrace" in ureport1:
# python
if ureport1["type"].lower() == "python":
ureport2["problem"]["stacktrace"] = []
cb = sorted(ureport1["core_backtrace"], key=lambda f: f["frame"])
for frame in cb:
newframe = {
"line_contents": "",
}
if "path" in frame:
newframe["file_name"] = frame["path"]
if "offset" in frame:
newframe["file_line"] = frame["offset"]
if "funcname" in frame:
if (frame["funcname"].startswith("<")
and frame["funcname"].endswith(">")):
funcname = frame["funcname"].strip("<>")
newframe["special_function"] = funcname
else:
newframe["function_name"] = frame["funcname"]
if "buildid" in frame:
newframe["build_id"] = frame["buildid"]
ureport2["problem"]["stacktrace"].append(newframe)
# coredump
if ureport1["type"].lower() == "userspace":
ureport2["problem"]["stacktrace"] = []
# group frame list into threads
threads = {}
cb = sorted(ureport1["core_backtrace"],
key=lambda f: (f["thread"], f["frame"]))
for frame in cb:
if frame["thread"] not in threads:
threads[frame["thread"]] = []
threads[frame["thread"]].append(frame)
for thread, frames in sorted(threads.items()):
newthread = {
"crash_thread": thread == ureport1["crash_thread"],
"frames": []
}
for frame in frames:
newframe = {
"address": 0,
}
if "buildid" in frame:
newframe["build_id"] = frame["buildid"]
if "offset" in frame:
newframe["build_id_offset"] = frame["offset"]
if "path" in frame:
newframe["file_name"] = frame["path"]
if "funcname" in frame:
newframe["function_name"] = frame["funcname"]
if "funchash" in frame:
newframe["fingerprint"] = frame["funchash"]
newthread["frames"].append(newframe)
ureport2["problem"]["stacktrace"].append(newthread)
# kerneloops
if ureport1["type"].lower() == "kerneloops":
# in uReport1 there is not enough information to fill uReport2
# let's try to use satyr library to parse the raw kerneloops
# or skip this entirely even though the result will be an invalid
# report
if "oops" in ureport1 and satyr is not None:
ureport2["problem"]["raw_oops"] = ureport1["oops"]
if "core_backtrace" in ureport1:
for frame in ureport1["core_backtrace"]:
if "buildid" in frame:
ureport2["problem"]["version"] = frame["buildid"]
break
koops = satyr.Kerneloops(ureport1["oops"])
ureport2["problem"]["modules"] = koops.modules
ureport2["problem"]["frames"] = []
for frame in koops.frames:
newframe = {
"address": frame.address,
"reliable": frame.reliable,
"function_name": frame.function_name,
"function_offset": frame.function_offset,
"function_length": frame.function_length,
}
if frame.from_function_offset is not None:
newframe["from_function_offset"] = \
frame.from_function_offset
if frame.from_function_length is not None:
newframe["from_function_length"] = \
frame.from_function_length
if (frame.module_name is not None
and frame.module_name != "vmlinux"):
newframe["module_name"] = frame.module_name
ureport2["problem"]["frames"].append(newframe)
# older versions of satyr do not export taint_flags
if hasattr(koops, "taint_flags"):
ureport2["problem"]["taint_flags"] = []
for flag, value in koops.taint_flags.items():
if value:
ureport2["problem"]["taint_flags"].append(flag)
# coredump requires user specs, use some defaults
if ureport1["type"].lower() == "userspace":
ureport2["problem"]["user"] = {"local": True, "root": False}
else:
ureport2["problem"]["user"] = {}
# sometimes even uReport1 provides the user data but in a different format
if "user_type" in ureport1:
if ureport1["user_type"].lower() == "root":
ureport2["problem"]["user"]["root"] = True
if ureport1["user_type"].lower() == "nologin":
ureport2["problem"]["user"]["nologin"] = True
if ureport1["user_type"].lower() == "local":
ureport2["problem"]["user"]["local"] = True
if ureport1["user_type"].lower() == "remote":
ureport2["problem"]["user"]["local"] = False
# raw kernel oops text
if ureport1["type"].lower() == "kerneloops" and "oops" in ureport1:
ureport2["problem"]["raw_oops"] = ureport1["oops"]
return ureport2