Make the block db open RAII style by calling it in the BlockManager
constructor.
Before this change the block tree db was needlessly re-opened during
startup when loading a completed snapshot. Improve this by letting the
block manager open it on construction. This also simplifies the test
code a bit.
The change was initially motivated to make it easier for users of the
kernel library to instantiate a BlockManager that may be used to read
data from disk without loading the block index into a cache.
This commit is done in preparation for the next commit. Here, the block
tree options are moved to the blockmanager options and the block tree is
instantiated through a helper method of the BlockManager, which is
removed again in the next commit.
Co-authored-by: MarcoFalke <*~=`'#}+{/-|&$^_@721217.xyz>
The check for whether the block tree db has been wiped before calling
NeedsRedownload() is confusing. The boolean is set in case of a reindex.
It was originally introduced to guard NeedsRedownload in case of a
reindex in #21009. However NeedsRedownload already returns early if the
chain's tip is not loaded. Since that is the case during a reindex, the
pre-check is redundant.
This avoids having to rely on implicit casts when passing them to the
various functions allocating the caches.
This also ensures that if the requested amount of db_cache does not fit
in a size_t, it is clamped to the maximum value of a size_t.
Also take this opportunity to make the total amounts of cache in the
chainstate manager a size_t too.
They are not related to the txdb, so a better place for them is the
new kernel and node cache file. Re-use the default amount of kernel
cache for the default node cache.
Carrying non-kernel related fields in the cache sizes for the indexes is
confusing for kernel library users. The cache sizes also are set
currently with magic numbers in bitcoin-chainstate. The comments for the
cache size calculations are also not completely clear.
Solve these things by moving the kernel-specific cache size fields to
their own struct.
This slightly changes the way the cache is allocated if the txindex
and/or blockfilterindex is used. Since they are now given precedence
over the block tree db cache, this results in a bit less cache being
allocated to the block tree db, coinsdb and coins caches. The effect is
negligible though, i.e. cache sizes with default dbcache reported
through the logs are:
master:
Cache configuration:
* Using 2.0 MiB for block index database
* Using 56.0 MiB for transaction index database
* Using 49.0 MiB for basic block filter index database
* Using 8.0 MiB for chain state database
* Using 335.0 MiB for in-memory UTXO set (plus up to 286.1 MiB of unused mempool space)
this branch:
Cache configuration:
* Using 2.0 MiB for block index database
* Using 56.2 MiB for transaction index database
* Using 49.2 MiB for basic block filter index database
* Using 8.0 MiB for chain state database
* Using 334.5 MiB for in-memory UTXO set (plus up to 286.1 MiB of unused mempool space)
The helpers are used in the following commits to increase the safety of
conversions during cache size calculations.
Co-authored-by: Ryan Ofsky <ryan@ofsky.org>
Co-authored-by: stickies-v <stickies-v@protonmail.com>
(total_cache / 4) + (1 << 23) is at least 8 MiB and nMaxCoinsDBCache is
also 8 MiB, so the minimum between the two will always be
nMaxCoinsDBCache. This is just a simplification and not changing the
result of the calculation.
Co-authored-by: Ryan Ofsky <ryan@ofsky.org>
It is always applied in the same way, no matter how the txindex is
setup. This was no longer accurate after 8181db8, where their
initialization was made independent.
This was missed during the original PR switching the nHashType argument
to int32_t in SignatureHash in bc52cda1f3.
The problem was discovered after running into a linker error when
attempting to link this code as a static library using the header as a
declaration with a riscv32 bare metal toolchain. The compiler would
error with:
/opt/riscv-ilp32/lib/gcc/riscv32-unknown-elf/13.2.0/../../../../riscv32-unknown-elf/bin/ld: build_kernel_riscv/src/libbitcoin_consensus.a(interpreter.cpp.o): in function `GenericTransactionSignatureChecker<CTransaction>::CheckECDSASignature(std::vector<unsigned char, std::allocator<unsigned char> > const&, std::vector<unsigned char, std::allocator<unsigned char> > const&, CScript const&, SigVersion) const':
/home/user/bitcoin/build_kernel_riscv/./script/interpreter.cpp:2043:(.text._ZNK34GenericTransactionSignatureCheckerI12CTransactionE19CheckECDSASignatureERKSt6vectorIhSaIhEES6_RK7CScript10SigVersion[_ZNK34GenericTransactionSignatureCheckerI12CTransactionE19CheckECDSASignatureERKSt6vectorIhSaIhEES6_RK7CScript10SigVersion]+0xee): undefined reference to `uint256 SignatureHash<CTransaction>(CScript const&, CTransaction const&, unsigned int, int, long long const&, SigVersion, PrecomputedTransactionData const*)'
The duplicate checks are repeated early in the contextual checks of
ProcessNewBlock. If duplicate blocks are detected much of their
validation is skipped. Depending on the constitution of the block,
validating the merkle root of the block is part of the more intensive
workload when validating a block. This could be an argument for moving
the pre-checks into block processing. In net_processing this would have
a smaller effect however, since the block mutation check, which also
validates the merkle root, is done before.
A side effect of this change is that a duplicate block is persisted
again on disk even when pruning is activated. This is similar to the
behaviour with getblockfrompeer. Add a release note for this change in
behaviour.
Testing spamming a node with valid, but duplicate unrequested blocks
seems to exhaust a CPU thread, but does not seem to significantly impact
keeping up with the tip. The benefits of adding these checks to
net_processing are questionable, especially since there are other ways
to trigger the more CPU-intensive checks without submitting a duplicate
block. Since these DOS concerns apply even less to the RPC interface,
which does not have banning mechanics built in, remove them too.
---
With the introduction of a mining ipc interface and the potential future
introduction of a kernel library API it becomes increasingly important
to offer common behaviour between them. An example of this is
ProcessNewBlock, which is used by ipc, rpc, net_processing and
(potentially) the kernel library. Having divergent behaviour on
suggested pre-checks and checks for these functions is confusing to both
developers and users and is a maintenance burden.
The rpc interface for ProcessNewBlock (submitblock) currently pre-checks
if the block has a coinbase transaction and whether it has been
processed before. While the current example binary for how to use the
kernel library, bitcoin-chainstate, imitates these checks, the other
interfaces do not.
ProcessNewBlock fails if an invalid duplicate block is passed in through
its call to AcceptBlock and AcceptBlockHeader. The failure in
AcceptBlockHeader makes AcceptBlock return early. This makes the
pre-check in submitblock redundant.
---
With the introduction of a mining ipc interface and the potential future
introduction of a kernel library API it becomes increasingly important
to offer common behaviour between them. An example of this is
ProcessNewBlock, which is used by ipc, rpc, net_processing and
(potentially) the kernel library. Having divergent behaviour on
suggested pre-checks and checks for these functions is confusing to both
developers and users and is a maintenance burden.
The rpc interface for ProcessNewBlock (submitblock) currently pre-checks
if the block has a coinbase transaction and whether it has been
processed before. While the current example binary for how to use the
kernel library, bitcoin-chainstate, imitates these checks, the other
interfaces do not.
The coinbase check is repeated again early during ProcessNewBlock.
Pre-checking it may also shadow more fundamental problems with a block.
In most cases the block header is checked first, before validating the
transactions. Checking the coinbase first therefore masks potential
issues with the header. Fix this by removing the pre-check.
The pre-check was likely introduced on top of
ada0caa165 to fix UB in
GetWitnessCommitmentIndex in case a block's transactions are empty. This
code path could only be reached because of the call to
UpdateUncommittedBlockStructures in submitblock, but cannot be reached
through net_processing.
Add some functional test cases to cover the previous conditions that
lead to a "Block does not start with a coinbase" json rpc error being
returned.
---
With the introduction of a mining ipc interface and the potential future
introduction of a kernel library API it becomes increasingly important
to offer common behaviour between them. An example of this is
ProcessNewBlock, which is used by ipc, rpc, net_processing and
(potentially) the kernel library. Having divergent behaviour on
suggested pre-checks and checks for these functions is confusing to both
developers and users and is a maintenance burden.
The rpc interface for ProcessNewBlock (submitblock) currently pre-checks
if the block has a coinbase transaction and whether it has been
processed before. While the current example binary for how to use the
kernel library, bitcoin-chainstate, imitates these checks, the other
interfaces do not.
This ensures the options are applied consistently from contexts where
they might not pass through the args manager, such as in some tests, or
when used through the kernel library.
This is similar to the patch applied in 09ef322acc.
The *_RECENT_CONSENSUS_CHANGE variants in the validation result
enumerations were always unused. They seem to have been kept around
speculatively for a soft fork after segwit, however they were never used
for taproot either. This points at them not having a clear purpose.
Based on the original pull requests' comments their usage was never
entirely clear:
https://github.com/bitcoin/bitcoin/pull/11639#issuecomment-370234133https://github.com/bitcoin/bitcoin/pull/15141#discussion_r271039747
Since they are part of the validation interface and need to exposed by
the kernel library keeping them around may also be confusing to future
users of the library.
The `addPackageTxs` method of the `BlockAssembler` currently has access
to two mempool variables, as an argument and as a member. Clean this up
and clarify that they both are the same mempool instance by removing the
argument and instead only using the member variable in the method.
Co-Authored-By: Anthony Towns <aj@erisian.com.au>
Co-authored-by: stickies-v <stickies-v@protonmail.com>
The chainstate caches are currently re-balanced on startup
even in the non-assumeutxo case, leading to the database being
needlessly re-opened and its cache re-allocated.
Similar to `InitCoinsCache` and `m_coinstip_cache_size_bytes` the
`m_coinsdb_cache_size_bytes` should be set in `InitCoinsDB`.
Together with only conservatively setting the cache values when a
assumeutxo chainstate is present, this allows for skipping the cache
re-balance during initialization in the normal non-assumeutxo case.
It is bad, because it is both printed for non-GUI users and does not
convey additional information.
Co-authored-by: Martin Zumsande <mzumsande@gmail.com>
The for loop has been a long standing source of confusion and bugs, both
because its purpose is not clearly documented and because the body of
the for loop contains a lot of logic.
Co-Authored-By: Ryan Ofsky <ryan@ofsky.org>
This makes it clearer which state is being mutated by the function and
facilitates getting rid of the for loop in the following commit. Move
creation of the required options into the function too, such that the
function takes fewer arguments and is more self-contained.
Co-Authored-By: Ryan Ofsky <ryan@ofsky.org>
The thread handle is never used by the ChainstateManager, so move it out
and into the node context. Users of the kernel library now no longer
have to manually join the thread when destructing the ChainstateManager.
Since the move to cmake, the kernel static library that is installed
after a cmake --install build is unusable. It lacks symbols for the
internal libraries, besides those defined in the kernel library target.
This is because cmake, unlike the libtool archiver, does not combine
multiple static libraries into a single static library on installation.
This is likely an intentional design choice, since there were a bunch of
caveats to the way libtool calculated these libraries.
Fix this problem by installing all the required libraries. The user must
then link all of them along with the bitcoin kernel library.
Makes it friendlier for potential future users of the kernel library if
they do not store the headers in a std::vector, but can guarantee
contiguous memory.
Makes it friendlier for potential future users of the kernel library if
they do not store the headers in a std::vector, but can guarantee
contiguous memory.
Makes it friendlier for potential future users of the kernel library if
they do not store the headers in a std::vector, but can guarantee
contiguous memory.
The globals setup by the function calls when creating a new kernel
context only need to be setup once. Calling them multiple times may be
wasteful and has no apparent benefit.
Besides kernel users potentially creating multiple contexts, this change
may also be useful for tests creating multiple setups.
Co-authored-by: stickies-v <stickies-v@protonmail.com>
Instead of constructing a new class every time a file operation is done,
construct them once for each of the undo and block file when a new
BlockManager is created.
In future, this might make it easier to introduce an abstract block
store.
The body of the fuzz test should ideally be a pure function. If data is
persisted in the cache over many iterations, and there is a crash,
reproducing it from the input might be difficult.
Adds more testing options for creating an environment without networking
and a validation interface. This is useful for improving the performance
of the utxo snapshot fuzz test, which constructs a new TestingSetup on
each iteration.
In future, users of the kernel library might run multiple chainstates in
parallel, or create and destroy multiple chainstates over the lifetime
of a process. Having static, mutable variables could lead to state
inconsistencies in these scenarios.
In future, users of the kernel library might run multiple chainstates in
parallel, or create and destroy multiple chainstates over the lifetime
of a process. Having static, mutable variables could lead to state
inconsistencies in these scenarios.
Move its ownership to the ChainstateManager class.
Next to simplifying usage of the kernel library by no longer requiring
manual setup of the cache prior to using validation code, it also slims
down the amount of memory allocated by BasicTestingSetup.
Use this opportunity to make SignatureCache RAII styled
Co-authored-by: Ryan Ofsky <ryan@ofsky.org>
Move its ownership to the ChainstateManager class.
Next to simplifying usage of the kernel library by no longer requiring
manual setup of the cache prior to using validation code, it also slims
down the amount of memory allocated by BasicTestingSetup.
Before this change continuing a reindex without the -reindex flag set
would leave the block and coins db intact, but discard the data of the
optional indexes. While not a bug per se, wiping the data again is
wasteful, both in terms of having to write it again, and potentially
leading to longer startup times.
When initially running a reindex, both the block index and any further
activated indexes are wiped. On an index's Init(), both the best block
stored by the index and the chain's tip are null. An index's m_synced
member is therefore true. This means that it will process blocks through
validation events while the reindex is running.
Currently, if the reindex is continued without the user re-specifying
the reindex flag, the block index is preserved but further index data is
wiped. This leads to the stored best block being null, but the chain tip
existing. The m_synced member will be set to false. The index will not
process blocks through the validation interface, but instead use the
background sync once the reindex is completed.
If the index is preserved (this change) after a restart its best block
may potentially match the chain tip. The m_synced member will be set to
true and the index can process validation events during the rest of the
reindex.
It does not control any actual logic and the log message as well as the
comment are obsolete, since no database initialization takes place there
anymore. Log messages indicating when indexes and chainstate databases
are loaded exist in other places.
Reverts a bug introduced in b47bd95920
"kernel: De-globalize fReindex". The change leads to a GUI user being
prompted to re-index on a chainstate loading failure more than once as
well as the node actually not reindexing if the user chooses to. Fix
this by setting the reindexing option instead of the atomic, which can
be safely re-used to indicate that a reindex should be attempted.
The bug specifically is caused by the chainman, and thus the blockman
and its m_reindexing atomic being destroyed on every iteration of
the for loop.
The reindex option for ChainstateLoadOptions is currently also set in a
confusing way. By using the reindex atomic, it is not obvious in which
scenario it is true or false.
The atomic is controlled by both the user passing the -reindex option,
the user chosing to reindex if something went wrong during chainstate
loading when running the gui, and by reading the reindexing flag from
the block tree database in LoadBlockIndexDB. In practice this read is
done through the chainstate module's CompleteChainstateInitialization's
call to LoadBlockIndex. Since this is only done after the reindex option
is set already, it does not have an effect on it.
Make this clear by using the reindex option from the blockman opts which
is only controlled by the user.
Move HexStr and HexDigit functions from util to crypto. The crypto library does
not actually use these functions, but the consensus library does. The consensus
and util libraries not allowed to depend on each other, but are allowed to
depend on the cryto library, so the crypto library is a reasonable put these.
The consensus library uses HexStr and HexDigit in script.cpp, transaction.cpp,
and uint256.cpp.
The util library does not use HexStr but does use HexDigit in strencodings.cpp
to parse integers.
fReindex is one of the last remaining globals exposed by the kernel
library, so move it into the blockstorage class to reduce the amount of
global mutable state and make the kernel library a bit less awkward to
use.
The current usage of ScheduleBatchPriority is not transparent. Once the
thread scheduling is changed, it remains unchanged for the remainder of
the thread's lifetime. So move the call from `ImportBlocks` to the init
code where it is clearer that its effect lasts for the entire lifetime
of the thread.
Users of the kernel library might not expect `ImportBlocks` to have an
influence on the thread it is called in. Particularly since it is only a
compile time option and cannot be controlled at runtime. With this patch
users of the kernel library can now choose their own scheduling policy.
The key module's functionality is not used by the kernel library, but
currently kernel users are still required to initialize the key module's
`secp256k1_context_sign` global as part of the `kernel::Context` through
`ECC_Start`.
The extra `bilingual_str` argument of the fatal error notifications and
`node::AbortNode()` is often unused and when used usually contains the
same string as the message argument. It also seems to be confusing,
since it is not consistently used for errors requiring user action. For
example some assumeutxo fatal errors require the user to do something,
but are not translated.
So simplify the fatal error and abort node interfaces by only passing a
translated string. This slightly changes the fatal errors displayed to
the user.
Also de-duplicate the abort error log since it is repeated in noui.cpp.
By defining a virtual interface class for the scheduler client, users of
the kernel can now define their own event consuming infrastructure,
without having to spawn threads or rely on the scheduler design.
Removing CScheduler also allows removing the thread and
exception modules from the kernel library.
This is done in preparation for the next two commits, where the
CMainSignals are de-globalized.
This avoids adding new constructor arguments to the ChainstateManager
and CTxMemPool classes over the next two commits.
This could also allow future tests that are only interested in the
internal behaviour of the classes to forgo instantiating the signals.
-BEGIN VERIFY SCRIPT-
regex_string='^(?!//).*(AC_APPLE_UNIVERSAL_BUILD|BOOST_PROCESS_USE_STD_FS|CHAR_EQUALS_INT8|CLIENT_VERSION_BUILD|CLIENT_VERSION_IS_RELEASE|CLIENT_VERSION_MAJOR|CLIENT_VERSION_MINOR|COPYRIGHT_HOLDERS|COPYRIGHT_HOLDERS_FINAL|COPYRIGHT_HOLDERS_SUBSTITUTION|COPYRIGHT_YEAR|ENABLE_ARM_SHANI|ENABLE_AVX2|ENABLE_EXTERNAL_SIGNER|ENABLE_SSE41|ENABLE_TRACING|ENABLE_WALLET|ENABLE_X86_SHANI|ENABLE_ZMQ|HAVE_BOOST|HAVE_BUILTIN_CLZL|HAVE_BUILTIN_CLZLL|HAVE_BYTESWAP_H|HAVE_CLMUL|HAVE_CONSENSUS_LIB|HAVE_CXX20|HAVE_DECL_BE16TOH|HAVE_DECL_BE32TOH|HAVE_DECL_BE64TOH|HAVE_DECL_BSWAP_16|HAVE_DECL_BSWAP_32|HAVE_DECL_BSWAP_64|HAVE_DECL_FORK|HAVE_DECL_FREEIFADDRS|HAVE_DECL_GETIFADDRS|HAVE_DECL_HTOBE16|HAVE_DECL_HTOBE32|HAVE_DECL_HTOBE64|HAVE_DECL_HTOLE16|HAVE_DECL_HTOLE32|HAVE_DECL_HTOLE64|HAVE_DECL_LE16TOH|HAVE_DECL_LE32TOH|HAVE_DECL_LE64TOH|HAVE_DECL_PIPE2|HAVE_DECL_SETSID|HAVE_DECL_STRERROR_R|HAVE_DEFAULT_VISIBILITY_ATTRIBUTE|HAVE_DLFCN_H|HAVE_DLLEXPORT_ATTRIBUTE|HAVE_ENDIAN_H|HAVE_EVHTTP_CONNECTION_GET_PEER_CONST_CHAR|HAVE_FDATASYNC|HAVE_GETENTROPY_RAND|HAVE_GETRANDOM|HAVE_GMTIME_R|HAVE_INTTYPES_H|HAVE_LIBADVAPI32|HAVE_LIBCOMCTL32|HAVE_LIBCOMDLG32|HAVE_LIBGDI32|HAVE_LIBIPHLPAPI|HAVE_LIBKERNEL32|HAVE_LIBOLE32|HAVE_LIBOLEAUT32|HAVE_LIBSHELL32|HAVE_LIBSHLWAPI|HAVE_LIBUSER32|HAVE_LIBUUID|HAVE_LIBWINMM|HAVE_LIBWS2_32|HAVE_MALLOC_INFO|HAVE_MALLOPT_ARENA_MAX|HAVE_MINIUPNPC_MINIUPNPC_H|HAVE_MINIUPNPC_UPNPCOMMANDS_H|HAVE_MINIUPNPC_UPNPERRORS_H|HAVE_NATPMP_H|HAVE_O_CLOEXEC|HAVE_POSIX_FALLOCATE|HAVE_PTHREAD|HAVE_PTHREAD_PRIO_INHERIT|HAVE_STDINT_H|HAVE_STDIO_H|HAVE_STDLIB_H|HAVE_STRERROR_R|HAVE_STRINGS_H|HAVE_STRING_H|HAVE_STRONG_GETAUXVAL|HAVE_SYSCTL|HAVE_SYSCTL_ARND|HAVE_SYSTEM|HAVE_SYS_ENDIAN_H|HAVE_SYS_PRCTL_H|HAVE_SYS_RESOURCES_H|HAVE_SYS_SELECT_H|HAVE_SYS_STAT_H|HAVE_SYS_SYSCTL_H|HAVE_SYS_TYPES_H|HAVE_SYS_VMMETER_H|HAVE_THREAD_LOCAL|HAVE_TIMINGSAFE_BCMP|HAVE_UNISTD_H|HAVE_VM_VM_PARAM_H|LT_OBJDIR|PACKAGE_BUGREPORT|PACKAGE_NAME|PACKAGE_STRING|PACKAGE_TARNAME|PACKAGE_URL|PACKAGE_VERSION|PTHREAD_CREATE_JOINABLE|QT_QPA_PLATFORM_ANDROID|QT_QPA_PLATFORM_COCOA|QT_QPA_PLATFORM_MINIMAL|QT_QPA_PLATFORM_WINDOWS|QT_QPA_PLATFORM_XCB|QT_STATICPLUGIN|STDC_HEADERS|STRERROR_R_CHAR_P|USE_ASM|USE_BDB|USE_DBUS|USE_NATPMP|USE_QRCODE|USE_SQLITE|USE_UPNP|_FILE_OFFSET_BITS|_LARGE_FILES)'
exclusion_files=":(exclude)src/minisketch :(exclude)src/crc32c :(exclude)src/secp256k1 :(exclude)src/crypto/sha256_arm_shani.cpp :(exclude)src/crypto/sha256_avx2.cpp :(exclude)src/crypto/sha256_sse41.cpp :(exclude)src/crypto/sha256_x86_shani.cpp"
git grep --perl-regexp --files-with-matches "$regex_string" -- '*.cpp' $exclusion_files | xargs git grep -L "bitcoin-config.h" | while read -r file; do line_number=$(awk -v my_file="$file" '/\/\/ file COPYING or https?:\/\/www.opensource.org\/licenses\/mit-license.php\./ {line = NR} /^\/\// && NR == line + 1 {while(getline && /^\/\//) line = NR} END {print line+1}' "$file"); sed -i "${line_number}i\\\\n\#if defined(HAVE_CONFIG_H)\\n#include <config/bitcoin-config.h>\\n\#endif" "$file"; done;
git grep --perl-regexp --files-with-matches "$regex_string" -- '*.h' $exclusion_files | xargs git grep -L "bitcoin-config.h" | while read -r file; do sed -i "/#define.*_H/a \\\\n\#if defined(HAVE_CONFIG_H)\\n#include <config/bitcoin-config.h>\\n\#endif" "$file"; done;
for file in $(git grep --files-with-matches 'bitcoin-config.h' -- '*.cpp' '*.h' $exclusion_files); do if ! grep -q --perl-regexp "$regex_string" $file; then sed -i '/HAVE_CONFIG_H/{N;N;N;d;}' $file; fi; done;
-END VERIFY SCRIPT-
The first command creates a regular expression for matching all bitcoin-config.h symbols in the following form: ^(?!//).*(AC_APPLE_UNIVERSAL_BUILD|BOOST_PROCESS_USE_STD_FS|...|_LARGE_FILES). It was generated with:
./autogen.sh && printf '^(?!//).*(%s)' $(awk '/^#undef/ {print $2}' src/config/bitcoin-config.h.in | paste -sd "|" -)
The second command holds a list of files and directories that should not be processed. These include subtree directories as well as some crypto files that already get their symbols through the makefile.
The third command checks for missing bitcoin-config headers in .cpp files and adds the header if it is missing.
The fourth command checks for missing bitcoin-config headers in .h files and adds the header if it is missing.
The fifth command checks for unneeded bitcoin-config headers in sources files and removes the header if it is unneeded.
make check runs a bunch of other subtree tests that exercise code that
is hardly ever changed and have a comparatively long runtime. There
seems to be no target for running just the unit tests, so add one.
We assume to be using python3, so don't check for it. This removes a
type error on the line `from io import BytesIO as StringIO`.
Specify the encoding as "utf8" when opening a file.
This has the goal of prohibiting users from accidentally creating
runtime failures, e.g. by interacting with iterator_to with a copied
entry.
CTxMemPoolEntry is already implicitly not move-constructable. So be
explicit about this and use a std::list to collect the values in the
policy_estimator fuzz test instead of a std::vector.
Co-authored-by: Anthony Towns <aj@erisian.com.au>
Currently the shutdown function resets the kernel before the
chainman and scheduler. Invert this order by resetting the kernel
last, since they might rely on the kernel.
This commit makes compat.h no longer a required include for users of the
libbitcoinkernel. Including compat.h imports a bunch of
platform-specific definitions.
This commit is part of the libbitcoinkernel project, namely its stage 1
step 3: Decouple most non-consensus headers from libbitcoinkernel.
Move functions requiring the netaddress.h include out of
libbitcoinkernel source files.
The netaddress.h file contains many non-consensus related definitions
and should thus not be part of the libbitcoinkernel. This commit makes
netaddress.h no longer a required include for users of the
libbitcoinkernel.
This commit is part of the libbitcoinkernel project, namely its stage 1
step 3: Decouple most non-consensus headers from libbitcoinkernel.
The protocol.h file contains many non-consensus related definitions and
should thus not be part of the libbitcoinkernel. This commit makes
protocol.h no longer a required include for users of the
libbitcoinkernel.
This commit is part of the libbitcoinkernel project, namely its stage 1
step 3: Decouple most non-consensus headers from libbitcoinkernel.
Co-Authored-By: Cory Fields <cory-nospam-@coryfields.com>
By returning an error code if either `FlushUndoFile` or `FlushBlockFile`
fail, the caller now has to explicitly handle block undo file flushing
errors. Before this change such errors were non-explicitly ignored
without a clear rationale.
Besides the call to `FlushUndoFile` in `FlushBlockFile`, ignore its
return code at its call site in `WriteUndoDataForBlock`. There, a failed
flush of the undo data should not be indicative of a failed write.
Add [[nodiscard]] annotations to `FlushUndoFile` such that its return
value is not just ignored in the future.
By returning an error code if `FlushBlockFile` fails, the caller now has
to explicitly handle block file flushing errors. Before this change
such errors were non-explicitly ignored without a clear rationale.
Prior to this patch `FlushBlockFile` may have failed silently in
`Chainstate::FlushStateToDisk`. Improve this with a log line. Also add a
TODO comment to flesh out whether returning early in the case of an
error is appropriate or not. Returning early might be appropriate to
prohibit `WriteBlockIndexDB` from writing a block index entry that does
not refer to a fully flushed block.
Besides `Chainstate::FlushStateToDisk`, `FlushBlockFile` is also called
by `FindBlockPos`. Don't change the abort behavior there, since we don't
want to fail the function if the flushing of already written blocks
fails. Instead, just document it.
A false return value indicates a fatal error (disk space being too low),
so make sure we always consume this error code.
This commit is part of an ongoing process for making the handling of
fatal errors more transparent and easier to understand.
The context of this commit is an effort to decouple the dbwrapper header
file from leveldb includes. To this end, the includes are moved to the
dbwrapper implementation file. This is done as part of the kernel
project to reduce the number of required includes for users of the
kernel.
Keep the generic serialization in the header, while moving
leveldb-specifics to the implementation file.
Since CharCast is no longer needed in the header, move it to the
implementation file.
The context of this commit is an effort to decouple the dbwrapper header
file from leveldb includes. To this end, the includes are moved to the
dbwrapper implementation file. This is done as part of the kernel
project to reduce the number of required includes for users of the
kernel.
Make it a static function in dbwrapper.cpp, since it is not used
elsewhere and when left in the header, would expose a leveldb type.
The context of this commit is an effort to decouple the dbwrapper header
file from leveldb includes. To this end, the includes are moved to the
dbwrapper implementation file. This is done as part of the kernel
project to reduce the number of required includes for users of the
kernel.
Keep the generic serialization in the header, while moving
leveldb-specifics to the implementation file.
The context of this commit is an effort to decouple the dbwrapper header
file from leveldb includes. To this end, the includes are moved to the
dbwrapper implementation file. This is done as part of the kernel
project to reduce the number of required includes for users of the
kernel.
TheCharlatan