// Copyright 2013 Google Inc. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

//
// These are weird things we need to do to get this compiling on
// random systems [subset].

#ifndef BASE_PORT_H_
#define BASE_PORT_H_

#include <string.h>         // for memcpy()
#include "integral_types.h"

namespace CLD2 {

// Portable handling of unaligned loads, stores, and copies.
// On some platforms, like ARM, the copy functions can be more efficient
// then a load and a store.

#if defined(ARCH_PIII) || defined(ARCH_ATHLON) || defined(ARCH_K8) || defined(_ARCH_PPC)

// x86 and x86-64 can perform unaligned loads/stores directly;
// modern PowerPC hardware can also do unaligned integer loads and stores;
// but note: the FPU still sends unaligned loads and stores to a trap handler!

#define UNALIGNED_LOAD16(_p) (*reinterpret_cast<const uint16 *>(_p))
#define UNALIGNED_LOAD32(_p) (*reinterpret_cast<const uint32 *>(_p))
#define UNALIGNED_LOAD64(_p) (*reinterpret_cast<const uint64 *>(_p))

#define UNALIGNED_STORE16(_p, _val) (*reinterpret_cast<uint16 *>(_p) = (_val))
#define UNALIGNED_STORE32(_p, _val) (*reinterpret_cast<uint32 *>(_p) = (_val))
#define UNALIGNED_STORE64(_p, _val) (*reinterpret_cast<uint64 *>(_p) = (_val))

#elif defined(__arm__) && \
      !defined(__ARM_ARCH_5__) && \
      !defined(__ARM_ARCH_5T__) && \
      !defined(__ARM_ARCH_5TE__) && \
      !defined(__ARM_ARCH_5TEJ__) && \
      !defined(__ARM_ARCH_6__) && \
      !defined(__ARM_ARCH_6J__) && \
      !defined(__ARM_ARCH_6K__) && \
      !defined(__ARM_ARCH_6Z__) && \
      !defined(__ARM_ARCH_6ZK__) && \
      !defined(__ARM_ARCH_6T2__) && \
      !defined(__ARM_ARCH_7__) && \
      !defined(__ARM_ARCH_7A__) && \
      !defined(__ARM_ARCH_7M__) && \
      !defined(__ARM_ARCH_7R__)

// ARMv7 and newer support native unaligned accesses, but only of 16-bit
// and 32-bit values (not 64-bit); older versions either raise a fatal signal,
// do an unaligned read and rotate the words around a bit, or do the reads very
// slowly (trip through kernel mode). There's no simple #define that says just
// “ARMv7 or higher”, so we have to filter away all ARMv5 and ARMv6
// sub-architectures. Newer gcc (>= 4.6) set an __ARM_FEATURE_ALIGNED #define,
// so in time, maybe we can move on to that.
//
// Note that even if a chipset supports unaligned access, it might not be
// enabled in any given system, e.g.:
// http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0491c/CIHCGCFD.html
// Therefore, it's generally just not safe to allow unaligned access on any ARM
// variant.
//
// This is a mess, but there's not much we can do about it.

#define UNALIGNED_LOAD16(_p) (*reinterpret_cast<const uint16 *>(_p))
#define UNALIGNED_LOAD32(_p) (*reinterpret_cast<const uint32 *>(_p))

#define UNALIGNED_STORE16(_p, _val) (*reinterpret_cast<uint16 *>(_p) = (_val))
#define UNALIGNED_STORE32(_p, _val) (*reinterpret_cast<uint32 *>(_p) = (_val))

// TODO(sesse): NEON supports unaligned 64-bit loads and stores.
// See if that would be more efficient on platforms supporting it,
// at least for copies.

inline uint64 UNALIGNED_LOAD64(const void *p) {
  uint64 t;
  memcpy(&t, p, sizeof t);
  return t;
}

inline void UNALIGNED_STORE64(void *p, uint64 v) {
  memcpy(p, &v, sizeof v);
}

#else

#define NEED_ALIGNED_LOADS

// These functions are provided for architectures that don't support
// unaligned loads and stores.

inline uint16 UNALIGNED_LOAD16(const void *p) {
  uint16 t;
  memcpy(&t, p, sizeof t);
  return t;
}

inline uint32 UNALIGNED_LOAD32(const void *p) {
  uint32 t;
  memcpy(&t, p, sizeof t);
  return t;
}

inline uint64 UNALIGNED_LOAD64(const void *p) {
  uint64 t;
  memcpy(&t, p, sizeof t);
  return t;
}

inline void UNALIGNED_STORE16(void *p, uint16 v) {
  memcpy(p, &v, sizeof v);
}

inline void UNALIGNED_STORE32(void *p, uint32 v) {
  memcpy(p, &v, sizeof v);
}

inline void UNALIGNED_STORE64(void *p, uint64 v) {
  memcpy(p, &v, sizeof v);
}

#endif

}       // End namespace CLD2

#endif  // BASE_PORT_H_