Sprinter/mame
2
0
Fork 0
mirror of https://github.com/holub/mame synced 2025-07-06 02:18:09 +03:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

util/chd.cpp: Hacked around GitHub #13029 (issues creating delta CHDs).

Browse Source 
The code depends on getting away with trying to read at least one hunk
past the end of the parent CHD.  Previously it just ignored any errors
from reading the parent CHD, but adding error checking broke it.

This hack just ignores that specific error.  Also cleaned up some gross
flow control and made more things const.

The code should obviously be fixed properly, but that's non-trivial.
This gets it to a state no worse than it was in 0.270.
This commit is contained in:
Vas Crabb 2025-02-27 03:59:14 +11:00
parent e6989d494a
commit 65c263749e
1 changed files with 41 additions and 35 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

76
src/lib/util/chd.cpp
View File

@ -3047,12 +3047,11 @@ std::error_condition chd_file_compressor::compress_continue(double &progress, do
while (m_read_queue_offset < m_logicalbytes && osd_work_queue_items(m_read_queue) < 2)
{
// see if we have enough free work items to read the next half of a buffer
uint32_t startitem = m_read_queue_offset / hunk_bytes();
uint32_t enditem = startitem + WORK_BUFFER_HUNKS / 2;
uint32_t curitem;
for (curitem = startitem; curitem < enditem; curitem++)
if (m_work_item[curitem % WORK_BUFFER_HUNKS].m_status != WS_READY)
break;
uint32_t const startitem = m_read_queue_offset / hunk_bytes();
uint32_t const enditem = startitem + WORK_BUFFER_HUNKS / 2;
uint32_t curitem = startitem;
while ((curitem < enditem) && (m_work_item[curitem % WORK_BUFFER_HUNKS].m_status == WS_READY))
++curitem;
// if it's not all clear, defer
if (curitem != enditem)
@ -3076,20 +3075,24 @@ std::error_condition chd_file_compressor::compress_continue(double &progress, do
work_item &item = m_work_item[m_write_hunk % WORK_BUFFER_HUNKS];
// free any OSD work item
if (item.m_osd != nullptr)
if (item.m_osd)
{
osd_work_item_release(item.m_osd);
item.m_osd = nullptr;
item.m_osd = nullptr;
}
if (m_walking_parent)
{
// for parent walking, just add to the hashmap
uint32_t uph = hunk_bytes() / unit_bytes();
uint32_t const uph = hunk_bytes() / unit_bytes();
uint32_t units = uph;
if (item.m_hunknum == hunk_count() - 1 || !compressed())
units = 1;
for (uint32_t unit = 0; unit < units; unit++)
{
if (m_parent_map.find(item.m_hash[unit].m_crc16, item.m_hash[unit].m_sha1) == hashmap::NOT_FOUND)
m_parent_map.add(item.m_hunknum * uph + unit, item.m_hash[unit].m_crc16, item.m_hash[unit].m_sha1);
}
}
else if (!compressed())
{
@ -3105,7 +3108,7 @@ std::error_condition chd_file_compressor::compress_continue(double &progress, do
if (!err && codec == CHD_CODEC_NONE) // TODO: report error?
m_total_out += m_hunkbytes;
}
else do
else
{
// for compressing, process the result
@ -3114,25 +3117,24 @@ std::error_condition chd_file_compressor::compress_continue(double &progress, do
if (selfhunk != hashmap::NOT_FOUND)
{
hunk_copy_from_self(item.m_hunknum, selfhunk);
break;
}
// if not, see if it's in the parent map
if (m_parent)
else
{
uint64_t parentunit = m_parent_map.find(item.m_hash[0].m_crc16, item.m_hash[0].m_sha1);
// if not, see if it's in the parent map
uint64_t const parentunit = m_parent ? m_parent_map.find(item.m_hash[0].m_crc16, item.m_hash[0].m_sha1) : hashmap::NOT_FOUND;
if (parentunit != hashmap::NOT_FOUND)
{
hunk_copy_from_parent(item.m_hunknum, parentunit);
break;
}
else
{
// otherwise, append it compressed and add to the self map
hunk_write_compressed(item.m_hunknum, item.m_compression, item.m_compressed, item.m_complen, item.m_hash[0].m_crc16);
m_total_out += item.m_complen;
m_current_map.add(item.m_hunknum, item.m_hash[0].m_crc16, item.m_hash[0].m_sha1);
}
}
// otherwise, append it compressed and add to the self map
hunk_write_compressed(item.m_hunknum, item.m_compression, item.m_compressed, item.m_complen, item.m_hash[0].m_crc16);
m_total_out += item.m_complen;
m_current_map.add(item.m_hunknum, item.m_hash[0].m_crc16, item.m_hash[0].m_sha1);
} while (false);
}
// reset the item and advance
item.m_status = WS_READY;
@ -3141,19 +3143,18 @@ std::error_condition chd_file_compressor::compress_continue(double &progress, do
// if we hit the end, finalize
if (m_write_hunk == m_hunkcount)
{
// if this is just walking the parent, reset and get ready for compression
if (m_walking_parent)
{
// if this is just walking the parent, reset and get ready for compression
m_walking_parent = false;
m_read_queue_offset = m_read_done_offset = 0;
m_write_hunk = 0;
for (auto & elem : m_work_item)
for (auto &elem : m_work_item)
elem.m_status = WS_READY;
}
// wait for all reads to finish and if we're compressed, write the final SHA1 and map
else
{
// wait for all reads to finish and if we're compressed, write the final SHA1 and map
osd_work_queue_wait(m_read_queue, 30 * osd_ticks_per_second());
if (!compressed())
return std::error_condition();
@ -3306,29 +3307,34 @@ void chd_file_compressor::async_read()
return;
// determine parameters for the read
uint32_t work_buffer_bytes = WORK_BUFFER_HUNKS * hunk_bytes();
uint32_t const work_buffer_bytes = WORK_BUFFER_HUNKS * hunk_bytes();
uint32_t numbytes = work_buffer_bytes / 2;
if (m_read_done_offset + numbytes > logical_bytes())
if ((m_read_done_offset + numbytes) > logical_bytes())
numbytes = logical_bytes() - m_read_done_offset;
uint8_t *const dest = &m_work_buffer[0] + (m_read_done_offset % work_buffer_bytes);
assert((&m_work_buffer[0] == dest) || (&m_work_buffer[work_buffer_bytes / 2] == dest));
assert(!(m_read_done_offset % hunk_bytes()));
uint64_t const end_offset = m_read_done_offset + numbytes;
// catch any exceptions coming out of here
try
{
// do the read
uint8_t *dest = &m_work_buffer[0] + (m_read_done_offset % work_buffer_bytes);
assert(dest == &m_work_buffer[0] || dest == &m_work_buffer[work_buffer_bytes / 2]);
uint64_t end_offset = m_read_done_offset + numbytes;
if (m_walking_parent)
{
// if walking the parent, read in hunks from the parent CHD
uint64_t curoffs = m_read_done_offset;
uint8_t *curdest = dest;
for (uint64_t curoffs = m_read_done_offset; curoffs < end_offset + 1; curoffs += hunk_bytes())
uint32_t curhunk = m_read_done_offset / hunk_bytes();
while (curoffs < end_offset + 1)
{
std::error_condition err = m_parent->read_hunk(curoffs / hunk_bytes(), curdest);
if (err)
std::error_condition err = m_parent->read_hunk(curhunk, curdest);
if (err && (error::HUNK_OUT_OF_RANGE != err)) // FIXME: fix the code so it doesn't depend on trying to read past the end of the parent CHD
throw err;
curoffs += hunk_bytes();
curdest += hunk_bytes();
++curhunk;
}
}
else