mirror of
https://github.com/dolphin-emu/dolphin.git
synced 2024-11-15 22:09:19 -07:00
335cf4f2db
A deep-copy method CopyReader has been added to BlobReader (virtual) and all of its subclasses (override). This should create a second BlobReader to open the same set of data but with an independent read pointer so that it doesn't interfere with any reads done on the original Reader. As part of this, IOFile has added code to create a deep copy IOFile pointer onto the same file, with code based on the platform in question to find the file ID from the file pointer and open a new one. There has also been a small piece added to FileInfo to enable a deep copy, but its only subclass at this time already had a copy constructor so this was relatively minor.
402 lines
12 KiB
C++
402 lines
12 KiB
C++
// Copyright 2008 Dolphin Emulator Project
|
|
// SPDX-License-Identifier: GPL-2.0-or-later
|
|
|
|
#include "DiscIO/CompressedBlob.h"
|
|
|
|
#include <algorithm>
|
|
#include <cstdio>
|
|
#include <cstring>
|
|
#include <memory>
|
|
#include <string>
|
|
#include <utility>
|
|
#include <vector>
|
|
|
|
#include <zlib.h>
|
|
|
|
#ifdef _WIN32
|
|
#include <windows.h>
|
|
#include <io.h>
|
|
#endif
|
|
|
|
#include "Common/Assert.h"
|
|
#include "Common/CommonTypes.h"
|
|
#include "Common/FileUtil.h"
|
|
#include "Common/Hash.h"
|
|
#include "Common/IOFile.h"
|
|
#include "Common/Logging/Log.h"
|
|
#include "Common/MsgHandler.h"
|
|
#include "DiscIO/Blob.h"
|
|
#include "DiscIO/DiscScrubber.h"
|
|
#include "DiscIO/MultithreadedCompressor.h"
|
|
#include "DiscIO/Volume.h"
|
|
|
|
namespace DiscIO
|
|
{
|
|
bool IsGCZBlob(File::IOFile& file);
|
|
|
|
CompressedBlobReader::CompressedBlobReader(File::IOFile file, const std::string& filename)
|
|
: m_file(std::move(file)), m_file_name(filename)
|
|
{
|
|
m_file_size = m_file.GetSize();
|
|
m_file.Seek(0, File::SeekOrigin::Begin);
|
|
m_file.ReadArray(&m_header, 1);
|
|
|
|
SetSectorSize(m_header.block_size);
|
|
|
|
// cache block pointers and hashes
|
|
m_block_pointers.resize(m_header.num_blocks);
|
|
m_file.ReadArray(m_block_pointers.data(), m_header.num_blocks);
|
|
m_hashes.resize(m_header.num_blocks);
|
|
m_file.ReadArray(m_hashes.data(), m_header.num_blocks);
|
|
|
|
m_data_offset = (sizeof(CompressedBlobHeader)) +
|
|
(sizeof(u64)) * m_header.num_blocks // skip block pointers
|
|
+ (sizeof(u32)) * m_header.num_blocks; // skip hashes
|
|
|
|
// A compressed block is never ever longer than a decompressed block, so just header.block_size
|
|
// should be fine.
|
|
// I still add some safety margin.
|
|
const u32 zlib_buffer_size = m_header.block_size + 64;
|
|
m_zlib_buffer.resize(zlib_buffer_size);
|
|
}
|
|
|
|
std::unique_ptr<CompressedBlobReader> CompressedBlobReader::Create(File::IOFile file,
|
|
const std::string& filename)
|
|
{
|
|
if (IsGCZBlob(file))
|
|
return std::unique_ptr<CompressedBlobReader>(
|
|
new CompressedBlobReader(std::move(file), filename));
|
|
|
|
return nullptr;
|
|
}
|
|
|
|
CompressedBlobReader::~CompressedBlobReader()
|
|
{
|
|
}
|
|
|
|
std::unique_ptr<BlobReader> CompressedBlobReader::CopyReader() const
|
|
{
|
|
return Create(m_file.Duplicate("rb"), m_file_name);
|
|
}
|
|
|
|
// IMPORTANT: Calling this function invalidates all earlier pointers gotten from this function.
|
|
u64 CompressedBlobReader::GetBlockCompressedSize(u64 block_num) const
|
|
{
|
|
u64 start = m_block_pointers[block_num];
|
|
if (block_num < m_header.num_blocks - 1)
|
|
return m_block_pointers[block_num + 1] - start;
|
|
else if (block_num == m_header.num_blocks - 1)
|
|
return m_header.compressed_data_size - start;
|
|
else
|
|
ERROR_LOG_FMT(DISCIO, "{} - illegal block number {}", __func__, block_num);
|
|
return 0;
|
|
}
|
|
|
|
bool CompressedBlobReader::GetBlock(u64 block_num, u8* out_ptr)
|
|
{
|
|
bool uncompressed = false;
|
|
u32 comp_block_size = (u32)GetBlockCompressedSize(block_num);
|
|
u64 offset = m_block_pointers[block_num] + m_data_offset;
|
|
|
|
if (offset & (1ULL << 63))
|
|
{
|
|
if (comp_block_size != m_header.block_size)
|
|
ERROR_LOG_FMT(DISCIO, "Uncompressed block with wrong size");
|
|
uncompressed = true;
|
|
offset &= ~(1ULL << 63);
|
|
}
|
|
|
|
// clear unused part of zlib buffer. maybe this can be deleted when it works fully.
|
|
memset(&m_zlib_buffer[comp_block_size], 0, m_zlib_buffer.size() - comp_block_size);
|
|
|
|
m_file.Seek(offset, File::SeekOrigin::Begin);
|
|
if (!m_file.ReadBytes(m_zlib_buffer.data(), comp_block_size))
|
|
{
|
|
ERROR_LOG_FMT(DISCIO, "The disc image \"{}\" is truncated, some of the data is missing.",
|
|
m_file_name);
|
|
m_file.ClearError();
|
|
return false;
|
|
}
|
|
|
|
// First, check hash.
|
|
const u32 block_hash = Common::HashAdler32(m_zlib_buffer.data(), comp_block_size);
|
|
if (block_hash != m_hashes[block_num])
|
|
{
|
|
ERROR_LOG_FMT(DISCIO,
|
|
"The disc image \"{}\" is corrupt.\n"
|
|
"Hash of block {} is {:08x} instead of {:08x}.",
|
|
m_file_name, block_num, block_hash, m_hashes[block_num]);
|
|
}
|
|
|
|
if (uncompressed)
|
|
{
|
|
std::copy(m_zlib_buffer.begin(), m_zlib_buffer.begin() + comp_block_size, out_ptr);
|
|
}
|
|
else
|
|
{
|
|
z_stream z = {};
|
|
z.next_in = m_zlib_buffer.data();
|
|
z.avail_in = comp_block_size;
|
|
if (z.avail_in > m_header.block_size)
|
|
{
|
|
ERROR_LOG_FMT(DISCIO, "Compressed block size is larger than uncompressed block size");
|
|
}
|
|
z.next_out = out_ptr;
|
|
z.avail_out = m_header.block_size;
|
|
inflateInit(&z);
|
|
int status = inflate(&z, Z_FULL_FLUSH);
|
|
u32 uncomp_size = m_header.block_size - z.avail_out;
|
|
if (status != Z_STREAM_END)
|
|
{
|
|
// this seem to fire wrongly from time to time
|
|
// to be sure, don't use compressed isos :P
|
|
ERROR_LOG_FMT(DISCIO, "Failure reading block {} - out of data and not at end.", block_num);
|
|
}
|
|
inflateEnd(&z);
|
|
if (uncomp_size != m_header.block_size)
|
|
{
|
|
ERROR_LOG_FMT(DISCIO, "Wrong block size");
|
|
return false;
|
|
}
|
|
}
|
|
return true;
|
|
}
|
|
|
|
struct CompressThreadState
|
|
{
|
|
CompressThreadState() : z{} {}
|
|
~CompressThreadState() { deflateEnd(&z); }
|
|
|
|
// z_stream will stop working if it changes address, so this object must not be moved
|
|
CompressThreadState(const CompressThreadState&) = delete;
|
|
CompressThreadState(CompressThreadState&&) = delete;
|
|
CompressThreadState& operator=(const CompressThreadState&) = delete;
|
|
CompressThreadState& operator=(CompressThreadState&&) = delete;
|
|
|
|
std::vector<u8> compressed_buffer;
|
|
z_stream z;
|
|
};
|
|
|
|
struct CompressParameters
|
|
{
|
|
std::vector<u8> data{};
|
|
u32 block_number = 0;
|
|
u64 inpos = 0;
|
|
};
|
|
|
|
struct OutputParameters
|
|
{
|
|
std::vector<u8> data{};
|
|
u32 block_number = 0;
|
|
bool compressed = false;
|
|
u64 inpos = 0;
|
|
};
|
|
|
|
static ConversionResultCode SetUpCompressThreadState(CompressThreadState* state)
|
|
{
|
|
return deflateInit(&state->z, 9) == Z_OK ? ConversionResultCode::Success :
|
|
ConversionResultCode::InternalError;
|
|
}
|
|
|
|
static ConversionResult<OutputParameters> Compress(CompressThreadState* state,
|
|
CompressParameters parameters, int block_size,
|
|
std::vector<u32>* hashes, int* num_stored,
|
|
int* num_compressed)
|
|
{
|
|
state->compressed_buffer.resize(block_size);
|
|
|
|
int retval = deflateReset(&state->z);
|
|
state->z.next_in = parameters.data.data();
|
|
state->z.avail_in = block_size;
|
|
state->z.next_out = state->compressed_buffer.data();
|
|
state->z.avail_out = block_size;
|
|
|
|
if (retval != Z_OK)
|
|
{
|
|
ERROR_LOG_FMT(DISCIO, "Deflate failed");
|
|
return ConversionResultCode::InternalError;
|
|
}
|
|
|
|
const int status = deflate(&state->z, Z_FINISH);
|
|
|
|
state->compressed_buffer.resize(block_size - state->z.avail_out);
|
|
|
|
OutputParameters output_parameters;
|
|
if ((status != Z_STREAM_END) || (state->z.avail_out < 10))
|
|
{
|
|
// let's store uncompressed
|
|
++*num_stored;
|
|
output_parameters = OutputParameters{std::move(parameters.data), parameters.block_number, false,
|
|
parameters.inpos};
|
|
}
|
|
else
|
|
{
|
|
// let's store compressed
|
|
++*num_compressed;
|
|
output_parameters = OutputParameters{std::move(state->compressed_buffer),
|
|
parameters.block_number, true, parameters.inpos};
|
|
}
|
|
|
|
(*hashes)[parameters.block_number] =
|
|
Common::HashAdler32(output_parameters.data.data(), output_parameters.data.size());
|
|
|
|
return std::move(output_parameters);
|
|
}
|
|
|
|
static ConversionResultCode Output(OutputParameters parameters, File::IOFile* outfile,
|
|
u64* position, std::vector<u64>* offsets, int progress_monitor,
|
|
u32 num_blocks, CompressCB callback)
|
|
{
|
|
u64 offset = *position;
|
|
if (!parameters.compressed)
|
|
offset |= 0x8000000000000000ULL;
|
|
(*offsets)[parameters.block_number] = offset;
|
|
|
|
*position += parameters.data.size();
|
|
|
|
if (!outfile->WriteBytes(parameters.data.data(), parameters.data.size()))
|
|
return ConversionResultCode::WriteFailed;
|
|
|
|
if (parameters.block_number % progress_monitor == 0)
|
|
{
|
|
const int ratio =
|
|
parameters.inpos == 0 ? 0 : static_cast<int>(100 * *position / parameters.inpos);
|
|
|
|
const std::string text = Common::FmtFormatT("{0} of {1} blocks. Compression ratio {2}%",
|
|
parameters.block_number, num_blocks, ratio);
|
|
|
|
const float completion = static_cast<float>(parameters.block_number) / num_blocks;
|
|
|
|
if (!callback(text, completion))
|
|
return ConversionResultCode::Canceled;
|
|
}
|
|
|
|
return ConversionResultCode::Success;
|
|
};
|
|
|
|
bool ConvertToGCZ(BlobReader* infile, const std::string& infile_path,
|
|
const std::string& outfile_path, u32 sub_type, int block_size,
|
|
CompressCB callback)
|
|
{
|
|
ASSERT(infile->GetDataSizeType() == DataSizeType::Accurate);
|
|
|
|
File::IOFile outfile(outfile_path, "wb");
|
|
if (!outfile)
|
|
{
|
|
PanicAlertFmtT(
|
|
"Failed to open the output file \"{0}\".\n"
|
|
"Check that you have permissions to write the target folder and that the media can "
|
|
"be written.",
|
|
outfile_path);
|
|
return false;
|
|
}
|
|
|
|
callback(Common::GetStringT("Files opened, ready to compress."), 0);
|
|
|
|
CompressedBlobHeader header;
|
|
header.magic_cookie = GCZ_MAGIC;
|
|
header.sub_type = sub_type;
|
|
header.block_size = block_size;
|
|
header.data_size = infile->GetDataSize();
|
|
|
|
// round upwards!
|
|
header.num_blocks = (u32)((header.data_size + (block_size - 1)) / block_size);
|
|
|
|
std::vector<u64> offsets(header.num_blocks);
|
|
std::vector<u32> hashes(header.num_blocks);
|
|
|
|
// seek past the header (we will write it at the end)
|
|
outfile.Seek(sizeof(CompressedBlobHeader), File::SeekOrigin::Current);
|
|
// seek past the offset and hash tables (we will write them at the end)
|
|
outfile.Seek((sizeof(u64) + sizeof(u32)) * header.num_blocks, File::SeekOrigin::Current);
|
|
|
|
// Now we are ready to write compressed data!
|
|
u64 inpos = 0;
|
|
u64 position = 0;
|
|
int num_compressed = 0;
|
|
int num_stored = 0;
|
|
int progress_monitor = std::max<int>(1, header.num_blocks / 1000);
|
|
|
|
const auto compress = [&](CompressThreadState* state, CompressParameters parameters) {
|
|
return Compress(state, std::move(parameters), block_size, &hashes, &num_stored,
|
|
&num_compressed);
|
|
};
|
|
|
|
const auto output = [&](OutputParameters parameters) {
|
|
return Output(std::move(parameters), &outfile, &position, &offsets, progress_monitor,
|
|
header.num_blocks, callback);
|
|
};
|
|
|
|
MultithreadedCompressor<CompressThreadState, CompressParameters, OutputParameters> compressor(
|
|
SetUpCompressThreadState, compress, output);
|
|
|
|
std::vector<u8> in_buf(block_size);
|
|
for (u32 i = 0; i < header.num_blocks; i++)
|
|
{
|
|
if (compressor.GetStatus() != ConversionResultCode::Success)
|
|
break;
|
|
|
|
const u64 bytes_to_read = std::min<u64>(block_size, header.data_size - inpos);
|
|
|
|
if (!infile->Read(inpos, bytes_to_read, in_buf.data()))
|
|
{
|
|
compressor.SetError(ConversionResultCode::ReadFailed);
|
|
break;
|
|
}
|
|
|
|
std::fill(in_buf.begin() + bytes_to_read, in_buf.begin() + header.block_size, 0);
|
|
|
|
inpos += block_size;
|
|
|
|
compressor.CompressAndWrite(CompressParameters{in_buf, i, inpos});
|
|
}
|
|
|
|
compressor.Shutdown();
|
|
|
|
header.compressed_data_size = position;
|
|
|
|
const ConversionResultCode result = compressor.GetStatus();
|
|
|
|
if (result != ConversionResultCode::Success)
|
|
{
|
|
// Remove the incomplete output file.
|
|
outfile.Close();
|
|
File::Delete(outfile_path);
|
|
}
|
|
else
|
|
{
|
|
// Okay, go back and fill in headers
|
|
outfile.Seek(0, File::SeekOrigin::Begin);
|
|
outfile.WriteArray(&header, 1);
|
|
outfile.WriteArray(offsets.data(), header.num_blocks);
|
|
outfile.WriteArray(hashes.data(), header.num_blocks);
|
|
|
|
callback(Common::GetStringT("Done compressing disc image."), 1.0f);
|
|
}
|
|
|
|
if (result == ConversionResultCode::ReadFailed)
|
|
PanicAlertFmtT("Failed to read from the input file \"{0}\".", infile_path);
|
|
|
|
if (result == ConversionResultCode::WriteFailed)
|
|
{
|
|
PanicAlertFmtT("Failed to write the output file \"{0}\".\n"
|
|
"Check that you have enough space available on the target drive.",
|
|
outfile_path);
|
|
}
|
|
|
|
return result == ConversionResultCode::Success;
|
|
}
|
|
|
|
bool IsGCZBlob(File::IOFile& file)
|
|
{
|
|
const u64 position = file.Tell();
|
|
if (!file.Seek(0, File::SeekOrigin::Begin))
|
|
return false;
|
|
CompressedBlobHeader header;
|
|
bool is_gcz = file.ReadArray(&header, 1) && header.magic_cookie == GCZ_MAGIC;
|
|
file.Seek(position, File::SeekOrigin::Begin);
|
|
return is_gcz;
|
|
}
|
|
|
|
} // namespace DiscIO
|