gfapi env module for LevelDB

env_gfapi.cc

// 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.

#include <deque>
#include <set>
#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>
#if defined(LEVELDB_PLATFORM_ANDROID)
#include <sys/stat.h>
#endif
#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"
extern "C" {
#include <glusterfs/api/glfs.h>
}

namespace leveldb {

namespace {

static Status IOError(const std::string& context, int err_number) {
  return Status::IOError(context, strerror(err_number));
}

class GfapiSequentialFile: public SequentialFile {
 private:
  std::string filename_;
  glfs_fd_t* fd_;

 public:
  GfapiSequentialFile(const std::string& fname, glfs_fd_t* fd)
      : filename_(fname), fd_(fd) { }
  virtual ~GfapiSequentialFile() { glfs_close(fd_); }

  virtual Status Read(size_t n, Slice* result, char* scratch) {
// Original:
//    size_t r = fread_unlocked(scratch, 1, n, file_);
    ssize_t r = glfs_read(fd_, scratch, n, 0);
    *result = Slice(scratch, r);

    if (r < 0)
      return IOError(filename_, errno);

// TODO: I'm not sure how I can check EOF with glfs_read().
//       The following is the original code using fread_unlocked().
/*
    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 Status::OK();
  }

  virtual Status Skip(uint64_t n) {
    if (glfs_lseek(fd_, static_cast<off_t>(n), SEEK_CUR) < 0)
      return IOError(filename_, errno);
    return Status::OK();
  }
};

// pread() based random-access
class GfapiRandomAccessFile: public RandomAccessFile {
 private:
  std::string filename_;
  glfs_fd_t* fd_;

 public:
  GfapiRandomAccessFile(const std::string& fname, glfs_fd_t* fd)
      : filename_(fname), fd_(fd) { }
  virtual ~GfapiRandomAccessFile() { glfs_close(fd_); }

  virtual Status Read(uint64_t offset, size_t n, Slice* result,
                      char* scratch) const {
    ssize_t r = glfs_pread(fd_, scratch, n, static_cast<off_t>(offset), 0);
    *result = Slice(scratch, (r < 0) ? 0 : r);
    if (r < 0) {
      // An error: return a non-ok status
      return IOError(filename_, errno);
    }
    return Status::OK();
  }
};

class GfapiWritableFile: public WritableFile {
 private:
  std::string filename_;
  glfs_fd_t* fd_;

 public:
  GfapiWritableFile(const std::string& fname, glfs_fd_t* fd)
      : filename_(fname), fd_(fd) { }
  virtual ~GfapiWritableFile() {
    GfapiWritableFile::Close();
  }

  virtual Status Append(const Slice& data) {
    const char* src = data.data();
    size_t left = data.size();
    size_t ret = glfs_write(fd_, src, left, 0);
    if (ret < 0)
      return IOError(filename_, errno);
    return Status::OK();
  }

  virtual Status Close() {
    if (fd_ != NULL) {
      int ret = glfs_close(fd_);
      fd_ = NULL;
      if (ret < 0)
        return IOError(filename_, errno);
    }
    return Status::OK();
  }

  virtual Status Flush() {
    return Status::OK();
  }

  virtual Status Sync() {
    if (glfs_fsync(fd_) < 0)
      return IOError(filename_, errno);
    return Status::OK();
  }
};

//static int LockOrUnlock(int fd, bool lock) {
static int LockOrUnlock(glfs_fd_t* fd, bool lock) {
  errno = 0;
  struct flock f;
  memset(&f, 0, sizeof(f));
  f.l_type = (lock ? F_WRLCK : F_UNLCK);
  f.l_whence = SEEK_SET;
  f.l_start = 0;
  f.l_len = 0;        // Lock/unlock entire file
  return glfs_posix_lock(fd, F_SETLK, &f);
}

class GfapiFileLock : public FileLock {
 public:
  glfs_fd_t* fd_;
  std::string name_;
};

// Set of locked files.  We keep a separate set instead of just
// relying on fcntrl(F_SETLK) since fcntl(F_SETLK) does not provide
// any protection against multiple uses from the same process.
class GfapiLockTable {
 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 GfapiEnv : public Env {
 private:
  glfs_t* fs_;

 public:
  GfapiEnv();
  virtual ~GfapiEnv() {
    fprintf(stderr, "Destroying Env::Default()\n");
    exit(1);
  }

  virtual Status NewSequentialFile(const std::string& fname,
                                   SequentialFile** result) {
    glfs_fd_t* fd;
    fd = glfs_open(fs_, fname.c_str(), O_RDONLY);

    if (fd == NULL) {
      *result = NULL;
      return IOError(fname, errno);
    } else {
      *result = new GfapiSequentialFile(fname, fd);
      return Status::OK();
    }
  }

  virtual Status NewRandomAccessFile(const std::string& fname,
                                     RandomAccessFile** result) {
    *result = NULL;

    glfs_fd_t* fd;
    fd = glfs_open(fs_, fname.c_str(), O_RDONLY);    
    if (fd == NULL) {
      *result = NULL;
      return IOError(fname, errno);
    } else {
      *result = new GfapiRandomAccessFile(fname, fd);
      return Status::OK();
    }
  }

  virtual Status NewWritableFile(const std::string& fname,
                                 WritableFile** result) {
    glfs_fd_t* fd;

    if (GfapiEnv::FileExists(fname))
      GfapiEnv::DeleteFile(fname);
    fd = glfs_creat(fs_, fname.c_str(), O_RDWR | O_APPEND, 0644);
    if (fd == NULL) {
      *result = NULL;
    } else {
      *result = new GfapiWritableFile(fname, fd);
      return Status::OK();
    }
    return IOError(fname, errno);
  }

  virtual bool FileExists(const std::string& fname) {
    return glfs_access(fs_, fname.c_str(), F_OK) == 0;
  }

  virtual Status GetChildren(const std::string& dir,
                             std::vector<std::string>* result) {
    result->clear();
    glfs_fd_t* d = glfs_opendir(fs_, dir.c_str());
    if (d == NULL)
      return IOError(dir, errno);
    struct dirent entry;
    struct dirent* res;
    while (true) {
      int ret = glfs_readdir_r(d, &entry, &res); 
      if (ret)
        return Status::IOError(dir.c_str(), "failed to read dentry");
      if (res == NULL)
        break;
      result->push_back(entry.d_name);
    }
    glfs_closedir(d);
    return Status::OK();
  }

  virtual Status DeleteFile(const std::string& fname) {
    if (glfs_unlink(fs_, fname.c_str()) != 0)
      return IOError(fname, errno);
    return Status::OK();
  }

  virtual Status CreateDir(const std::string& name) {
    if (glfs_mkdir(fs_, name.c_str(), 0755) != 0)
      return IOError(name, errno);
    return Status::OK();
  }

  virtual Status DeleteDir(const std::string& name) {
    if (glfs_rmdir(fs_, name.c_str()) != 0)
      return IOError(name, errno);
    return Status::OK();
  }

  virtual Status GetFileSize(const std::string& fname, uint64_t* size) {
    struct stat sbuf;
    if (glfs_stat(fs_, fname.c_str(), &sbuf) != 0) {
      *size = 0;
      return IOError(fname, errno);
    } else {
      *size = sbuf.st_size;
    }
    return Status::OK();
  }

  virtual Status RenameFile(const std::string& src, const std::string& target) {
    if (glfs_rename(fs_, src.c_str(), target.c_str()) != 0)
      return IOError(src, errno);
    return Status::OK();
  }

  virtual Status LockFile(const std::string& fname, FileLock** lock) {
    *lock = NULL;

    glfs_fd_t* fd;
    if (FileExists(fname)) {
      fd = glfs_open(fs_, fname.c_str(), O_RDWR);
    } else {
      fd = glfs_creat(fs_, fname.c_str(), O_RDWR, 0644);
    }

    if (fd == NULL) {
      return IOError(fname, errno);
    } else if (!locks_.Insert(fname)) {
      glfs_close(fd);
      return Status::IOError("lock " + fname, "already held by process");
    } else if (LockOrUnlock(fd, true) == -1) {
      glfs_close(fd);
      locks_.Remove(fname);
      return IOError("lock " + fname, errno);
    } else {
      GfapiFileLock* my_lock = new GfapiFileLock;
      my_lock->fd_ = fd;
      my_lock->name_ = fname;
      *lock = my_lock;
    }
    return Status::OK();
  }

  virtual Status UnlockFile(FileLock* lock) {
    GfapiFileLock* my_lock = reinterpret_cast<GfapiFileLock*>(lock);
    if (LockOrUnlock(my_lock->fd_, false) == -1) {
      return IOError("unlock", errno);
    }
    locks_.Remove(my_lock->name_);
    glfs_close(my_lock->fd_);
    delete my_lock;
    return Status::OK();
  }

  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, &GfapiEnv::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));
      exit(1);
    }
  }

  // BGThread() is the body of the background thread
  void BGThread();
  static void* BGThreadWrapper(void* arg) {
    reinterpret_cast<GfapiEnv*>(arg)->BGThread();
    return NULL;
  }

  size_t page_size_;
  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_;

  GfapiLockTable locks_;
//  MmapLimiter mmap_limit_;
};

GfapiEnv::GfapiEnv() : page_size_(getpagesize()),
                       started_bgthread_(false) {

  const char* gfvol = getenv("GF_VOLUME");
  const char* gfserver = getenv("GF_SERVER");
  int ret;

  if (gfvol == NULL || gfserver == NULL) {
    fprintf(stderr, "Need to set environmental variables GF_SERVER and GF_VOLUME\n");
    exit(-1);
  }
  fs_ = glfs_new(gfvol);
  glfs_set_volfile_server(fs_, "tcp", gfserver, 24007);
  ret = glfs_init(fs_);
  if (ret) {
    fprintf(stderr, "Failed to connect GlusterFS server/volume: %s/%s\n", gfserver, gfvol);
    exit(ret);
  }
  PthreadCall("mutex_init", pthread_mutex_init(&mu_, NULL));
  PthreadCall("cvar_init", pthread_cond_init(&bgsignal_, NULL));
}

void GfapiEnv::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,  &GfapiEnv::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 GfapiEnv::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 GfapiEnv::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 GfapiEnv; }

Env* Env::Default() {
  pthread_once(&once, InitDefaultEnv);
  return default_env;
}

}  // namespace leveldb