blob: c75b06cf9ce4f5d7d67a47b6c746e79de9c53e6b [file] [log] [blame]
// Copyright (c) 2011 The LevelDB Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file. See the AUTHORS file for names of contributors.
// Copyright 2016 The Fuchsia Authors. All rights reserved.
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
// * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
// * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#include <dirent.h>
#include <errno.h>
#include <fcntl.h>
#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/mman.h>
#include <sys/stat.h>
#include <sys/time.h>
#include <sys/types.h>
#include <time.h>
#include <unistd.h>
#include <deque>
#include <set>
#include "leveldb/env.h"
#include "leveldb/slice.h"
#include "port/port.h"
#include "util/logging.h"
#include "util/mutexlock.h"
#include "util/posix_logger.h"
namespace leveldb {
namespace {
static Status IOError(const std::string& context, int err_number) {
return Status::IOError(context, strerror(err_number));
}
class FuchsiaSequentialFile: public SequentialFile {
private:
std::string filename_;
FILE* file_;
public:
FuchsiaSequentialFile(const std::string& fname, FILE* f)
: filename_(fname), file_(f) { }
virtual ~FuchsiaSequentialFile() { fclose(file_); }
virtual Status Read(size_t n, Slice* result, char* scratch) {
Status s;
size_t r = fread_unlocked(scratch, 1, n, file_);
*result = Slice(scratch, r);
if (r < n) {
if (feof(file_)) {
// We leave status as ok if we hit the end of the file
} else {
// A partial read with an error: return a non-ok status
s = IOError(filename_, errno);
}
}
return s;
}
virtual Status Skip(uint64_t n) {
if (fseek(file_, n, SEEK_CUR)) {
return IOError(filename_, errno);
}
return Status::OK();
}
};
// pread() based random-access
class FuchsiaRandomAccessFile: public RandomAccessFile {
private:
std::string filename_;
int fd_;
public:
FuchsiaRandomAccessFile(const std::string& fname, int fd)
: filename_(fname), fd_(fd) { }
virtual ~FuchsiaRandomAccessFile() { close(fd_); }
virtual Status Read(uint64_t offset, size_t n, Slice* result,
char* scratch) const {
Status s;
ssize_t r = pread(fd_, scratch, n, static_cast<off_t>(offset));
*result = Slice(scratch, (r < 0) ? 0 : r);
if (r < 0) {
// An error: return a non-ok status
s = IOError(filename_, errno);
}
return s;
}
};
// Helper class to limit mmap file usage so that we do not end up
// running out virtual memory or running into kernel performance
// problems for very large databases.
class MmapLimiter {
public:
// Up to 1000 mmaps for 64-bit binaries; none for smaller pointer sizes.
MmapLimiter() {
SetAllowed(sizeof(void*) >= 8 ? 1000 : 0);
}
// If another mmap slot is available, acquire it and return true.
// Else return false.
bool Acquire() {
if (GetAllowed() <= 0) {
return false;
}
MutexLock l(&mu_);
intptr_t x = GetAllowed();
if (x <= 0) {
return false;
} else {
SetAllowed(x - 1);
return true;
}
}
// Release a slot acquired by a previous call to Acquire() that returned true.
void Release() {
MutexLock l(&mu_);
SetAllowed(GetAllowed() + 1);
}
private:
port::Mutex mu_;
port::AtomicPointer allowed_;
intptr_t GetAllowed() const {
return reinterpret_cast<intptr_t>(allowed_.Acquire_Load());
}
// REQUIRES: mu_ must be held
void SetAllowed(intptr_t v) {
allowed_.Release_Store(reinterpret_cast<void*>(v));
}
MmapLimiter(const MmapLimiter&);
void operator=(const MmapLimiter&);
};
// mmap() based random-access
class FuchsiaMmapReadableFile: public RandomAccessFile {
private:
std::string filename_;
void* mmapped_region_;
size_t length_;
MmapLimiter* limiter_;
public:
// base[0,length-1] contains the mmapped contents of the file.
FuchsiaMmapReadableFile(const std::string& fname, void* base, size_t length,
MmapLimiter* limiter)
: filename_(fname), mmapped_region_(base), length_(length),
limiter_(limiter) {
}
virtual ~FuchsiaMmapReadableFile() {
munmap(mmapped_region_, length_);
limiter_->Release();
}
virtual Status Read(uint64_t offset, size_t n, Slice* result,
char* scratch) const {
Status s;
if (offset + n > length_) {
*result = Slice();
s = IOError(filename_, EINVAL);
} else {
*result = Slice(reinterpret_cast<char*>(mmapped_region_) + offset, n);
}
return s;
}
};
class FuchsiaWritableFile : public WritableFile {
private:
std::string filename_;
FILE* file_;
public:
FuchsiaWritableFile(const std::string& fname, FILE* f)
: filename_(fname), file_(f) { }
~FuchsiaWritableFile() {
if (file_ != NULL) {
// Ignoring any potential errors
fclose(file_);
}
}
virtual Status Append(const Slice& data) {
size_t r = fwrite_unlocked(data.data(), 1, data.size(), file_);
if (r != data.size()) {
return IOError(filename_, errno);
}
return Status::OK();
}
virtual Status Close() {
Status result;
if (fclose(file_) != 0) {
result = IOError(filename_, errno);
}
file_ = NULL;
return result;
}
virtual Status Flush() {
if (fflush_unlocked(file_) != 0) {
return IOError(filename_, errno);
}
return Status::OK();
}
Status SyncDirIfManifest() {
const char* f = filename_.c_str();
const char* sep = strrchr(f, '/');
Slice basename;
std::string dir;
if (sep == NULL) {
dir = ".";
basename = f;
} else {
dir = std::string(f, sep - f);
basename = sep + 1;
}
Status s;
if (basename.starts_with("MANIFEST")) {
int fd = open(dir.c_str(), O_RDONLY);
if (fd < 0) {
s = IOError(dir, errno);
} else {
if (fsync(fd) < 0) {
s = IOError(dir, errno);
}
close(fd);
}
}
return s;
}
virtual Status Sync() {
// Ensure new files referred to by the manifest are in the filesystem.
Status s = SyncDirIfManifest();
if (!s.ok()) {
return s;
}
if (fflush_unlocked(file_) != 0 ||
fdatasync(fileno(file_)) != 0) {
s = Status::IOError(filename_, strerror(errno));
}
return s;
}
};
class FuchsiaFileLock : public FileLock {
public:
std::string name_;
};
// Set of locked files. This is used to guard against multiple uses from one
// process.
class LockTable {
private:
port::Mutex mu_;
std::set<std::string> locked_files_;
public:
bool Insert(const std::string& fname) {
MutexLock l(&mu_);
return locked_files_.insert(fname).second;
}
void Remove(const std::string& fname) {
MutexLock l(&mu_);
locked_files_.erase(fname);
}
};
class FuchsiaEnv : public Env {
public:
FuchsiaEnv();
virtual ~FuchsiaEnv() {
char msg[] = "Destroying Env::Default()\n";
fwrite(msg, 1, sizeof(msg), stderr);
abort();
}
virtual Status NewSequentialFile(const std::string& fname,
SequentialFile** result) {
FILE* f = fopen(fname.c_str(), "r");
if (f == NULL) {
*result = NULL;
return IOError(fname, errno);
} else {
*result = new FuchsiaSequentialFile(fname, f);
return Status::OK();
}
}
virtual Status NewRandomAccessFile(const std::string& fname,
RandomAccessFile** result) {
*result = NULL;
Status s;
int fd = open(fname.c_str(), O_RDONLY);
if (fd < 0) {
s = IOError(fname, errno);
} else if (mmap_limit_.Acquire()) {
uint64_t size;
s = GetFileSize(fname, &size);
if (s.ok()) {
void* base = mmap(NULL, size, PROT_READ, MAP_SHARED, fd, 0);
if (base != MAP_FAILED) {
*result = new FuchsiaMmapReadableFile(fname, base, size, &mmap_limit_);
} else {
s = IOError(fname, errno);
}
}
close(fd);
if (!s.ok()) {
mmap_limit_.Release();
}
} else {
*result = new FuchsiaRandomAccessFile(fname, fd);
}
return s;
}
virtual Status NewWritableFile(const std::string& fname,
WritableFile** result) {
Status s;
FILE* f = fopen(fname.c_str(), "w");
if (f == NULL) {
*result = NULL;
s = IOError(fname, errno);
} else {
*result = new FuchsiaWritableFile(fname, f);
}
return s;
}
virtual Status NewAppendableFile(const std::string& fname,
WritableFile** result) {
Status s;
FILE* f = fopen(fname.c_str(), "a");
if (f == NULL) {
*result = NULL;
s = IOError(fname, errno);
} else {
*result = new FuchsiaWritableFile(fname, f);
}
return s;
}
virtual bool FileExists(const std::string& fname) {
return access(fname.c_str(), F_OK) == 0;
}
virtual Status GetChildren(const std::string& dir,
std::vector<std::string>* result) {
result->clear();
DIR* d = opendir(dir.c_str());
if (d == NULL) {
return IOError(dir, errno);
}
struct dirent* entry;
while ((entry = readdir(d)) != NULL) {
result->push_back(entry->d_name);
}
closedir(d);
return Status::OK();
}
virtual Status DeleteFile(const std::string& fname) {
Status result;
if (unlink(fname.c_str()) != 0) {
result = IOError(fname, errno);
}
return result;
}
virtual Status CreateDir(const std::string& name) {
Status result;
if (mkdir(name.c_str(), 0755) != 0) {
result = IOError(name, errno);
}
return result;
}
virtual Status DeleteDir(const std::string& name) {
Status result;
if (rmdir(name.c_str()) != 0) {
result = IOError(name, errno);
}
return result;
}
virtual Status GetFileSize(const std::string& fname, uint64_t* size) {
Status s;
struct stat sbuf;
if (stat(fname.c_str(), &sbuf) != 0) {
*size = 0;
s = IOError(fname, errno);
} else {
*size = sbuf.st_size;
}
return s;
}
virtual Status RenameFile(const std::string& src, const std::string& target) {
Status result;
if (rename(src.c_str(), target.c_str()) != 0) {
result = IOError(src, errno);
}
return result;
}
// Our implementation uses the |locks_| table to guard against creating
// multiple databases backed by the same file within one process. This does
// not guard against multiple processes using the same file concurrently.
virtual Status LockFile(const std::string& fname, FileLock** lock) {
*lock = NULL;
Status result;
if (!locks_.Insert(fname)) {
result = Status::IOError("lock " + fname, "already held by process");
} else {
FuchsiaFileLock* my_lock = new FuchsiaFileLock;
my_lock->name_ = fname;
*lock = my_lock;
}
return result;
}
virtual Status UnlockFile(FileLock* lock) {
FuchsiaFileLock* my_lock = reinterpret_cast<FuchsiaFileLock*>(lock);
Status result;
locks_.Remove(my_lock->name_);
delete my_lock;
return result;
}
virtual void Schedule(void (*function)(void*), void* arg);
virtual void StartThread(void (*function)(void* arg), void* arg);
virtual Status GetTestDirectory(std::string* result) {
const char* env = getenv("TEST_TMPDIR");
if (env && env[0] != '\0') {
*result = env;
} else {
char buf[100];
snprintf(buf, sizeof(buf), "/tmp/leveldbtest-%d", int(geteuid()));
*result = buf;
}
// Directory may already exist
CreateDir(*result);
return Status::OK();
}
static uint64_t gettid() {
pthread_t tid = pthread_self();
uint64_t thread_id = 0;
memcpy(&thread_id, &tid, std::min(sizeof(thread_id), sizeof(tid)));
return thread_id;
}
virtual Status NewLogger(const std::string& fname, Logger** result) {
FILE* f = fopen(fname.c_str(), "w");
if (f == NULL) {
*result = NULL;
return IOError(fname, errno);
} else {
*result = new PosixLogger(f, &FuchsiaEnv::gettid);
return Status::OK();
}
}
virtual uint64_t NowMicros() {
struct timeval tv;
gettimeofday(&tv, NULL);
return static_cast<uint64_t>(tv.tv_sec) * 1000000 + tv.tv_usec;
}
virtual void SleepForMicroseconds(int micros) {
usleep(micros);
}
private:
void PthreadCall(const char* label, int result) {
if (result != 0) {
fprintf(stderr, "pthread %s: %s\n", label, strerror(result));
abort();
}
}
// BGThread() is the body of the background thread
void BGThread();
static void* BGThreadWrapper(void* arg) {
reinterpret_cast<FuchsiaEnv*>(arg)->BGThread();
return NULL;
}
pthread_mutex_t mu_;
pthread_cond_t bgsignal_;
pthread_t bgthread_;
bool started_bgthread_;
// Entry per Schedule() call
struct BGItem { void* arg; void (*function)(void*); };
typedef std::deque<BGItem> BGQueue;
BGQueue queue_;
LockTable locks_;
MmapLimiter mmap_limit_;
};
FuchsiaEnv::FuchsiaEnv() : started_bgthread_(false) {
PthreadCall("mutex_init", pthread_mutex_init(&mu_, NULL));
PthreadCall("cvar_init", pthread_cond_init(&bgsignal_, NULL));
}
void FuchsiaEnv::Schedule(void (*function)(void*), void* arg) {
PthreadCall("lock", pthread_mutex_lock(&mu_));
// Start background thread if necessary
if (!started_bgthread_) {
started_bgthread_ = true;
PthreadCall(
"create thread",
pthread_create(&bgthread_, NULL, &FuchsiaEnv::BGThreadWrapper, this));
}
// If the queue is currently empty, the background thread may currently be
// waiting.
if (queue_.empty()) {
PthreadCall("signal", pthread_cond_signal(&bgsignal_));
}
// Add to priority queue
queue_.push_back(BGItem());
queue_.back().function = function;
queue_.back().arg = arg;
PthreadCall("unlock", pthread_mutex_unlock(&mu_));
}
void FuchsiaEnv::BGThread() {
while (true) {
// Wait until there is an item that is ready to run
PthreadCall("lock", pthread_mutex_lock(&mu_));
while (queue_.empty()) {
PthreadCall("wait", pthread_cond_wait(&bgsignal_, &mu_));
}
void (*function)(void*) = queue_.front().function;
void* arg = queue_.front().arg;
queue_.pop_front();
PthreadCall("unlock", pthread_mutex_unlock(&mu_));
(*function)(arg);
}
}
namespace {
struct StartThreadState {
void (*user_function)(void*);
void* arg;
};
}
static void* StartThreadWrapper(void* arg) {
StartThreadState* state = reinterpret_cast<StartThreadState*>(arg);
state->user_function(state->arg);
delete state;
return NULL;
}
void FuchsiaEnv::StartThread(void (*function)(void* arg), void* arg) {
pthread_t t;
StartThreadState* state = new StartThreadState;
state->user_function = function;
state->arg = arg;
PthreadCall("start thread",
pthread_create(&t, NULL, &StartThreadWrapper, state));
}
} // namespace
static pthread_once_t once = PTHREAD_ONCE_INIT;
static Env* default_env;
static void InitDefaultEnv() { default_env = new FuchsiaEnv; }
Env* Env::Default() {
pthread_once(&once, InitDefaultEnv);
return default_env;
}
} // namespace leveldb