-
Notifications
You must be signed in to change notification settings - Fork 35.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
wallet: Implement independent BDB parser #26606
Conversation
The following sections might be updated with supplementary metadata relevant to reviewers and maintainers. Code CoverageFor detailed information about the code coverage, see the test coverage report. ReviewsSee the guideline for information on the review process.
If your review is incorrectly listed, please react with 👎 to this comment and the bot will ignore it on the next update. ConflictsReviewers, this pull request conflicts with the following ones:
If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first. |
e18fdbf
to
f135c19
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Some review nibbles.
src/streams.h
Outdated
@@ -530,6 +530,14 @@ class AutoFile | |||
// | |||
// Stream subset | |||
// | |||
void seek(int64_t offset, int origin) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
1d093bb: Note for other reviewers: yes
src/streams.h
Outdated
if (!file) | ||
throw std::ios_base::failure("CAutoFile::read: file handle is nullptr"); | ||
if (fseek(file, offset, origin) != 0) | ||
throw std::ios_base::failure(feof(file) ? "CAutoFile::seek: end of file" : "CAutoFile::seek: fseek failed"); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
1d093bb: maybe log the exact error code.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's kind of annoying to do that, so I'll leave it as is. The other functions do the same.
// file COPYING or http://www.opensource.org/licenses/mit-license.php. | ||
|
||
#ifndef BITCOIN_WALLET_MIGRATE_H | ||
#define BITCOIN_WALLET_MIGRATE_H |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
1224f45: maybe call this file bdb_ro_wallet_db.h
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The intent is to have all of the migration stuff in these files, so I will leave it as is.
Concept ACK. Have you looked into a fuzz target for this parser? |
I haven't. If someone would like to write one, I'd be happy to add it. |
2d61534
to
f2dc015
Compare
Adding a fuzz target that compares this to the real thing seems useful. Given that this PR changes It seems a bit daunting to compare the low level DBD parsing code in this PR to the real thing. Having a fuzzer also reduces the need to be very thorough in that. I'm not sure how much confidence I have in the existing test coverage for dumpwallet. |
944a020
to
f108c6b
Compare
…l page sizes) This aims to complete our test framework BDB parser to reflect our read-only BDB parser in the wallet codebase. This could be useful both for making review of bitcoin#26606 easier and to also possibly improve our functional tests for the BDB parser by comparing with an alternative implementation.
BerkeleyRODatabase is intended for use after BDB is removed, so it needs to be able to read all of the records from a BDB file. Thus an independent deserializer for BDB data files is implemented in it. This deserializer is targeted towards the data files that Bitcoin Core creates so it does not fully support all of BDB's features (e.g. other database types, encryption, etc.).
Implements MakeBerkeleyRODatabase and adds DatabaseFormat::BERKELEY_RO so that MakeDatabase can use BerkeleyRO as the backend database.
…lets In order to ease the transition to not having BDB, make the dump tool use DatabaseFormmat::BERKELEY_RO when -withinternalbdb is set.
Byteswapped databases make it easier to test opening and deserializing other endian databases.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Re-ACK d51fbab
reACK d51fbab |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
reACK d51fbab
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
re-ACK d51fbab
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
ACK d51fbab
Convinced myself that this parser works reliably partly by looking at the original code, partly by extending the test framework's BDB parser to match this PR's logic (see #30125), which helped a lot in understanding the structure. Also tested with a few random signet legacy wallets that I still had around, though those were all quite small.
Btw, I couldn't manage to find or create a BDB wallet where the DELETE flag is set in a record header (would be nice to exercise that in a test, though it's hard to verify that it's really set without inspecting manually, or e.g. adding debug messages on deserialization), but the code to handle that is simple enough and looks correct.
error.original == "Page number mismatch" || | ||
error.original == "Bad btree level" || | ||
error.original == "Bad page size" || | ||
error.original == "File size is not a multiple of page size" || |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: isn't thrown in the parser (commented out), hence can be removed here
error.original == "File size is not a multiple of page size" || |
error.original == "Unexpected number of entries in outer database root page" || | ||
error.original == "Subdatabase has an unexpected name" || | ||
error.original == "Subdatabase page number has unexpected length" || | ||
error.original == "Unexpected inner database page type" || |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: isn't thrown anywhere, can be removed
error.original == "Unexpected inner database page type" || |
@@ -24,11 +24,14 @@ class ToolWalletTest(BitcoinTestFramework): | |||
def add_options(self, parser): | |||
self.add_wallet_options(parser) | |||
parser.add_argument("--bdbro", action="store_true", help="Use the BerkeleyRO internal parser when dumping a Berkeley DB wallet file") | |||
parser.add_argument("--swap-bdb-endian", action="store_true",help="When making Legacy BDB wallets, always make then byte swapped internally") |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
typo nit
parser.add_argument("--swap-bdb-endian", action="store_true",help="When making Legacy BDB wallets, always make then byte swapped internally") | |
parser.add_argument("--swap-bdb-endian", action="store_true",help="When making Legacy BDB wallets, always make them byte swapped internally") |
I wonder if it makes sense to collect the fuzz inputs to qa-assets, from testers that still have them. However, the fuzz CI does not compile with BDB, so that'll probably need to be adjusted first. |
…l page sizes) This aims to complete our test framework BDB parser to reflect our read-only BDB parser in the wallet codebase. This could be useful both for making review of bitcoin#26606 easier and to also possibly improve our functional tests for the BDB parser by comparing with an alternative implementation.
Ported to the CMake-based build system in hebasto#220. |
Split from #26596
This PR adds
BerkeleyRODatabase
which is an independent implementation of a BDB file parser. It provides read only access to a BDB file, and can therefore be used as a read only database backend for wallets. This will be used for dumping legacy wallet records and migrating legacy wallets without the need for BDB itself.Wallettool's
dump
command is changed to useBerkeleyRODatabase
instead ofBerkeleyDatabase
(andCWallet
itself) to demonstrate that this parser works and to test it against the existing wallettool functional tests.