// Copyright (c) 2011 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
#include "chrome/browser/crash_handler_host_linux.h"
#include <stdint.h>
#include <stdlib.h>
#include <sys/socket.h>
#include <sys/syscall.h>
#include <unistd.h>
#include "base/eintr_wrapper.h"
#include "base/file_path.h"
#include "base/format_macros.h"
#include "base/linux_util.h"
#include "base/logging.h"
#include "base/memory/singleton.h"
#include "base/message_loop.h"
#include "base/path_service.h"
#include "base/rand_util.h"
#include "base/string_util.h"
#include "base/task.h"
#include "base/threading/thread.h"
#include "breakpad/src/client/linux/handler/exception_handler.h"
#include "breakpad/src/client/linux/minidump_writer/linux_dumper.h"
#include "breakpad/src/client/linux/minidump_writer/minidump_writer.h"
#include "chrome/app/breakpad_linux.h"
#include "chrome/common/chrome_paths.h"
#include "chrome/common/env_vars.h"
#include "content/browser/browser_thread.h"
using google_breakpad::ExceptionHandler;
namespace {
// The length of the control message:
const unsigned kControlMsgSize =
CMSG_SPACE(2*sizeof(int)) + CMSG_SPACE(sizeof(struct ucred));
// The length of the regular payload:
const unsigned kCrashContextSize = sizeof(ExceptionHandler::CrashContext);
// Handles the crash dump and frees the allocated BreakpadInfo struct.
void CrashDumpTask(CrashHandlerHostLinux* handler, BreakpadInfo* info) {
if (handler->IsShuttingDown())
return;
HandleCrashDump(*info);
delete[] info->filename;
delete[] info->process_type;
delete[] info->crash_url;
delete[] info->guid;
delete[] info->distro;
delete info;
}
} // namespace
// Since classes derived from CrashHandlerHostLinux are singletons, it's only
// destroyed at the end of the processes lifetime, which is greater in span than
// the lifetime of the IO message loop.
DISABLE_RUNNABLE_METHOD_REFCOUNT(CrashHandlerHostLinux);
CrashHandlerHostLinux::CrashHandlerHostLinux()
: shutting_down_(false) {
int fds[2];
// We use SOCK_SEQPACKET rather than SOCK_DGRAM to prevent the process from
// sending datagrams to other sockets on the system. The sandbox may prevent
// the process from calling socket() to create new sockets, but it'll still
// inherit some sockets. With PF_UNIX+SOCK_DGRAM, it can call sendmsg to send
// a datagram to any (abstract) socket on the same system. With
// SOCK_SEQPACKET, this is prevented.
CHECK_EQ(socketpair(AF_UNIX, SOCK_SEQPACKET, 0, fds), 0);
static const int on = 1;
// Enable passcred on the server end of the socket
CHECK_EQ(setsockopt(fds[1], SOL_SOCKET, SO_PASSCRED, &on, sizeof(on)), 0);
process_socket_ = fds[0];
browser_socket_ = fds[1];
BrowserThread::PostTask(
BrowserThread::IO, FROM_HERE,
NewRunnableMethod(this, &CrashHandlerHostLinux::Init));
}
CrashHandlerHostLinux::~CrashHandlerHostLinux() {
HANDLE_EINTR(close(process_socket_));
HANDLE_EINTR(close(browser_socket_));
}
void CrashHandlerHostLinux::Init() {
MessageLoopForIO* ml = MessageLoopForIO::current();
CHECK(ml->WatchFileDescriptor(
browser_socket_, true /* persistent */,
MessageLoopForIO::WATCH_READ,
&file_descriptor_watcher_, this));
ml->AddDestructionObserver(this);
}
void CrashHandlerHostLinux::InitCrashUploaderThread() {
SetProcessType();
uploader_thread_.reset(
new base::Thread(std::string(process_type_ + "_crash_uploader").c_str()));
uploader_thread_->Start();
}
void CrashHandlerHostLinux::OnFileCanWriteWithoutBlocking(int fd) {
DCHECK(false);
}
void CrashHandlerHostLinux::OnFileCanReadWithoutBlocking(int fd) {
DCHECK_EQ(fd, browser_socket_);
// A process has crashed and has signaled us by writing a datagram
// to the death signal socket. The datagram contains the crash context needed
// for writing the minidump as well as a file descriptor and a credentials
// block so that they can't lie about their pid.
const size_t kIovSize = 7;
struct msghdr msg = {0};
struct iovec iov[kIovSize];
// Freed in WriteDumpFile();
char* crash_context = new char[kCrashContextSize];
// Freed in CrashDumpTask();
char* guid = new char[kGuidSize + 1];
char* crash_url = new char[kMaxActiveURLSize + 1];
char* distro = new char[kDistroSize + 1];
char* tid_buf_addr = NULL;
int tid_fd = -1;
uint64_t uptime;
char control[kControlMsgSize];
const ssize_t expected_msg_size =
kCrashContextSize +
kGuidSize + 1 +
kMaxActiveURLSize + 1 +
kDistroSize + 1 +
sizeof(tid_buf_addr) + sizeof(tid_fd) +
sizeof(uptime);
iov[0].iov_base = crash_context;
iov[0].iov_len = kCrashContextSize;
iov[1].iov_base = guid;
iov[1].iov_len = kGuidSize + 1;
iov[2].iov_base = crash_url;
iov[2].iov_len = kMaxActiveURLSize + 1;
iov[3].iov_base = distro;
iov[3].iov_len = kDistroSize + 1;
iov[4].iov_base = &tid_buf_addr;
iov[4].iov_len = sizeof(tid_buf_addr);
iov[5].iov_base = &tid_fd;
iov[5].iov_len = sizeof(tid_fd);
iov[6].iov_base = &uptime;
iov[6].iov_len = sizeof(uptime);
msg.msg_iov = iov;
msg.msg_iovlen = kIovSize;
msg.msg_control = control;
msg.msg_controllen = kControlMsgSize;
const ssize_t msg_size = HANDLE_EINTR(recvmsg(browser_socket_, &msg, 0));
if (msg_size != expected_msg_size) {
LOG(ERROR) << "Error reading from death signal socket. Crash dumping"
<< " is disabled."
<< " msg_size:" << msg_size
<< " errno:" << errno;
file_descriptor_watcher_.StopWatchingFileDescriptor();
return;
}
if (msg.msg_controllen != kControlMsgSize ||
msg.msg_flags & ~MSG_TRUNC) {
LOG(ERROR) << "Received death signal message with the wrong size;"
<< " msg.msg_controllen:" << msg.msg_controllen
<< " msg.msg_flags:" << msg.msg_flags
<< " kCrashContextSize:" << kCrashContextSize
<< " kControlMsgSize:" << kControlMsgSize;
return;
}
// Walk the control payload an extract the file descriptor and validated pid.
pid_t crashing_pid = -1;
int partner_fd = -1;
int signal_fd = -1;
for (struct cmsghdr *hdr = CMSG_FIRSTHDR(&msg); hdr;
hdr = CMSG_NXTHDR(&msg, hdr)) {
if (hdr->cmsg_level != SOL_SOCKET)
continue;
if (hdr->cmsg_type == SCM_RIGHTS) {
const unsigned len = hdr->cmsg_len -
(((uint8_t*)CMSG_DATA(hdr)) - (uint8_t*)hdr);
DCHECK_EQ(len % sizeof(int), 0u);
const unsigned num_fds = len / sizeof(int);
if (num_fds != 2) {
// A nasty process could try and send us too many descriptors and
// force a leak.
LOG(ERROR) << "Death signal contained wrong number of descriptors;"
<< " num_fds:" << num_fds;
for (unsigned i = 0; i < num_fds; ++i)
HANDLE_EINTR(close(reinterpret_cast<int*>(CMSG_DATA(hdr))[i]));
return;
} else {
partner_fd = reinterpret_cast<int*>(CMSG_DATA(hdr))[0];
signal_fd = reinterpret_cast<int*>(CMSG_DATA(hdr))[1];
}
} else if (hdr->cmsg_type == SCM_CREDENTIALS) {
const struct ucred *cred =
reinterpret_cast<struct ucred*>(CMSG_DATA(hdr));
crashing_pid = cred->pid;
}
}
if (crashing_pid == -1 || partner_fd == -1 || signal_fd == -1) {
LOG(ERROR) << "Death signal message didn't contain all expected control"
<< " messages";
if (partner_fd >= 0)
HANDLE_EINTR(close(partner_fd));
if (signal_fd >= 0)
HANDLE_EINTR(close(signal_fd));
return;
}
// Kernel bug workaround (broken in 2.6.30 at least):
// The kernel doesn't translate PIDs in SCM_CREDENTIALS across PID
// namespaces. Thus |crashing_pid| might be garbage from our point of view.
// In the future we can remove this workaround, but we have to wait a couple
// of years to be sure that it's worked its way out into the world.
// The crashing process closes its copy of the signal_fd immediately after
// calling sendmsg(). We can thus not reliably look for with with
// FindProcessHoldingSocket(). But by necessity, it has to keep the
// partner_fd open until the crashdump is complete.
uint64_t inode_number;
if (!base::FileDescriptorGetInode(&inode_number, partner_fd)) {
LOG(WARNING) << "Failed to get inode number for passed socket";
HANDLE_EINTR(close(partner_fd));
HANDLE_EINTR(close(signal_fd));
return;
}
HANDLE_EINTR(close(partner_fd));
pid_t actual_crashing_pid = -1;
if (!base::FindProcessHoldingSocket(&actual_crashing_pid, inode_number)) {
LOG(WARNING) << "Failed to find process holding other end of crash reply "
"socket";
HANDLE_EINTR(close(signal_fd));
return;
}
if (actual_crashing_pid != crashing_pid) {
crashing_pid = actual_crashing_pid;
// The crashing TID set inside the compromised context via sys_gettid()
// in ExceptionHandler::HandleSignal is also wrong and needs to be
// translated.
//
// We expect the crashing thread to be in sys_read(), waiting for use to
// write to |signal_fd|. Most newer kernels where we have the different pid
// namespaces also have /proc/[pid]/syscall, so we can look through
// |actual_crashing_pid|'s thread group and find the thread that's in the
// read syscall with the right arguments.
std::string expected_syscall_data;
// /proc/[pid]/syscall is formatted as follows:
// syscall_number arg1 ... arg6 sp pc
// but we just check syscall_number through arg3.
base::StringAppendF(&expected_syscall_data, "%d 0x%x %p 0x1 ",
SYS_read, tid_fd, tid_buf_addr);
pid_t crashing_tid =
base::FindThreadIDWithSyscall(crashing_pid, expected_syscall_data);
if (crashing_tid == -1) {
// We didn't find the thread we want. Maybe it didn't reach sys_read()
// yet, or the kernel doesn't support /proc/[pid]/syscall or the thread
// went away. We'll just take a guess here and assume the crashing
// thread is the thread group leader.
crashing_tid = crashing_pid;
}
ExceptionHandler::CrashContext* bad_context =
reinterpret_cast<ExceptionHandler::CrashContext*>(crash_context);
bad_context->tid = crashing_tid;
}
// Sanitize the string data a bit more
guid[kGuidSize] = crash_url[kMaxActiveURLSize] = distro[kDistroSize] = 0;
// Freed in CrashDumpTask();
BreakpadInfo* info = new BreakpadInfo;
info->process_type_length = process_type_.length();
char* process_type_str = new char[info->process_type_length + 1];
process_type_.copy(process_type_str, info->process_type_length);
process_type_str[info->process_type_length] = '\0';
info->process_type = process_type_str;
info->crash_url_length = strlen(crash_url);
info->crash_url = crash_url;
info->guid_length = strlen(guid);
info->guid = guid;
info->distro_length = strlen(distro);
info->distro = distro;
info->upload = (getenv(env_vars::kHeadless) == NULL);
info->process_start_time = uptime;
BrowserThread::PostTask(
BrowserThread::FILE, FROM_HERE,
NewRunnableMethod(this,
&CrashHandlerHostLinux::WriteDumpFile,
info,
crashing_pid,
crash_context,
signal_fd));
}
void CrashHandlerHostLinux::WriteDumpFile(BreakpadInfo* info,
pid_t crashing_pid,
char* crash_context,
int signal_fd) {
DCHECK(BrowserThread::CurrentlyOn(BrowserThread::FILE));
FilePath dumps_path("/tmp");
PathService::Get(base::DIR_TEMP, &dumps_path);
if (!info->upload)
PathService::Get(chrome::DIR_CRASH_DUMPS, &dumps_path);
const uint64 rand = base::RandUint64();
const std::string minidump_filename =
StringPrintf("%s/chromium-%s-minidump-%016" PRIx64 ".dmp",
dumps_path.value().c_str(), process_type_.c_str(), rand);
if (!google_breakpad::WriteMinidump(minidump_filename.c_str(),
crashing_pid, crash_context,
kCrashContextSize)) {
LOG(ERROR) << "Failed to write crash dump for pid " << crashing_pid;
}
delete[] crash_context;
// Freed in CrashDumpTask();
char* minidump_filename_str = new char[minidump_filename.length() + 1];
minidump_filename.copy(minidump_filename_str, minidump_filename.length());
minidump_filename_str[minidump_filename.length()] = '\0';
info->filename = minidump_filename_str;
BrowserThread::PostTask(
BrowserThread::IO, FROM_HERE,
NewRunnableMethod(this,
&CrashHandlerHostLinux::QueueCrashDumpTask,
info,
signal_fd));
}
void CrashHandlerHostLinux::QueueCrashDumpTask(BreakpadInfo* info,
int signal_fd) {
DCHECK(BrowserThread::CurrentlyOn(BrowserThread::IO));
// Send the done signal to the process: it can exit now.
struct msghdr msg = {0};
struct iovec done_iov;
done_iov.iov_base = const_cast<char*>("\x42");
done_iov.iov_len = 1;
msg.msg_iov = &done_iov;
msg.msg_iovlen = 1;
HANDLE_EINTR(sendmsg(signal_fd, &msg, MSG_DONTWAIT | MSG_NOSIGNAL));
HANDLE_EINTR(close(signal_fd));
uploader_thread_->message_loop()->PostTask(
FROM_HERE,
NewRunnableFunction(&CrashDumpTask, this, info));
}
void CrashHandlerHostLinux::WillDestroyCurrentMessageLoop() {
file_descriptor_watcher_.StopWatchingFileDescriptor();
// If we are quitting and there are crash dumps in the queue, turn them into
// no-ops.
shutting_down_ = true;
uploader_thread_->Stop();
}
bool CrashHandlerHostLinux::IsShuttingDown() const {
return shutting_down_;
}
GpuCrashHandlerHostLinux::GpuCrashHandlerHostLinux() {
InitCrashUploaderThread();
}
GpuCrashHandlerHostLinux::~GpuCrashHandlerHostLinux() {
}
void GpuCrashHandlerHostLinux::SetProcessType() {
process_type_ = "gpu-process";
}
// static
GpuCrashHandlerHostLinux* GpuCrashHandlerHostLinux::GetInstance() {
return Singleton<GpuCrashHandlerHostLinux>::get();
}
PluginCrashHandlerHostLinux::PluginCrashHandlerHostLinux() {
InitCrashUploaderThread();
}
PluginCrashHandlerHostLinux::~PluginCrashHandlerHostLinux() {
}
void PluginCrashHandlerHostLinux::SetProcessType() {
process_type_ = "plugin";
}
// static
PluginCrashHandlerHostLinux* PluginCrashHandlerHostLinux::GetInstance() {
return Singleton<PluginCrashHandlerHostLinux>::get();
}
RendererCrashHandlerHostLinux::RendererCrashHandlerHostLinux() {
InitCrashUploaderThread();
}
RendererCrashHandlerHostLinux::~RendererCrashHandlerHostLinux() {
}
void RendererCrashHandlerHostLinux::SetProcessType() {
process_type_ = "renderer";
}
// static
RendererCrashHandlerHostLinux* RendererCrashHandlerHostLinux::GetInstance() {
return Singleton<RendererCrashHandlerHostLinux>::get();
}
PpapiCrashHandlerHostLinux::PpapiCrashHandlerHostLinux() {
InitCrashUploaderThread();
}
PpapiCrashHandlerHostLinux::~PpapiCrashHandlerHostLinux() {
}
void PpapiCrashHandlerHostLinux::SetProcessType() {
process_type_ = "ppapi";
}
// static
PpapiCrashHandlerHostLinux* PpapiCrashHandlerHostLinux::GetInstance() {
return Singleton<PpapiCrashHandlerHostLinux>::get();
}