ASLR is not currently working for the bitcoin-cli.exe binary. This is
due to it not having a .reloc section, which is stripped by default by
the mingw-w64 ld we use for gitian builds. A good summary of issues with
ld and mingw-w64 is available in this thread:
https://sourceware.org/bugzilla/show_bug.cgi?id=19011.
All other Windows binaries that we distribute (bitcoind, bitcoin-qt,
bitcoin-wallet, bitcoin-tx and test_bitcoin) do not suffer this issue,
and currently having working ASLR. This is due to them exporting
(inadvertent or not) libsecp256k1 symbols, and, as a result, the .reloc
section is not stripped by ld.
This change is a temporary workaround, also the same one described here:
https://www.kb.cert.org/vuls/id/307144/, that causes main() to be
exported. Exporting a symbol will mean that the .reloc section is not
stripped, and ASLR will function correctly.
Github-Pull: #18702
Rebased-From: 315a4d36f716341a38bc4e4de8630b3246d27dbc
fad691cafe083743a26f434488990f060ae4ac45 rpc: Make verifychain default values static, not depend on global args (MarcoFalke)
Pull request description:
This fixes several issues:
* The documentation is not compile-time static and depends on run-time arguments, making it impossible to host it on a static resource like a website or pdf. See also a similar change in the wallet rpc code: #18499
* The same call (relying on default values) will run different code on different machines, depending on the command line args that were used to start the server. This might lead to hard-to-debug-remote issues.
This is a small behaviour change, and I will add release notes.
ACKs for top commit:
theStack:
ACK fad691cafe
promag:
Code review ACK fad691cafe083743a26f434488990f060ae4ac45.
Tree-SHA512: 1c7a253ff0ec13a973b10d3777b71c70954ded5805b65a3ab06317327014de4cd0601d71d30c6ce89a581722c150cb5567acc1bd3e0c789cb51bab6ef0dcfc4a
7fcdec0f326f7fb547ec9c651871842cf518ea38 Remove PID file at the very end (Hennadii Stepanov)
Pull request description:
While reproducing the bug from #18517, I've noticed that the `bitcoind.pid` file has already been removed when the `bitcoind` hangs.
This PR makes `Shutdown()` keep the `bitcoind.pid` file available until the end.
ACKs for top commit:
MarcoFalke:
ACK 7fcdec0f326f7fb547ec9c651871842cf518ea38
emilengler:
utACK 7fcdec0f326f7fb547ec9c651871842cf518ea38
promag:
Code review ACK 7fcdec0f326f7fb547ec9c651871842cf518ea38.
theStack:
Code review ACK 7fcdec0f326f7fb547ec9c651871842cf518ea38
Tree-SHA512: 9732ef34e137dbee70a06d922b316b8ea7b9a1c959cf8861b6940cd789336dc19ee468a4c3a28d95d1458076a48270c676b0ff27fec30cf57eced6ddab0a2a9b
fa1da3d4bfc0511a89f5b19d5a4d89e55ff7ccde test: Add basic addr relay test (MarcoFalke)
fa1793c1c44a3f75a09f9c636467b8274c541bdd net: Pass connman const when relaying address (MarcoFalke)
fa47a0b003f53708b6d5df1ed4e7f8a7c68aa3ac net: Make addr relay mockable (MarcoFalke)
Pull request description:
As usual:
* Switch to std::chrono time to be type-safe and mockable
* Add basic test that relies on mocktime to add code coverage
ACKs for top commit:
naumenkogs:
utACK fa1da3d
promag:
ACK fa1da3d4bfc0511a89f5b19d5a4d89e55ff7ccde (fabe56e44b6f683e24e37246a7a8851190947cb3 before https://github.com/bitcoin/bitcoin/pull/18454#issuecomment-607866453), fa5bf23d527a450e72c2bf13d013e5393b664ca3 was dropped since last review.
Tree-SHA512: 0552bf8fcbe375baa3cab62acd8c23b2994efa47daff818ad1116d0ffaa0b9e520dc1bca2bbc68369b25584e85e54861fe6fd0968de4f503b95439c099df9bd7
b1d24d1d031a2b2ce67bf846bafa1c3a499b7553 Reorder the test instructions by number (Pieter Wuille)
c2ccadc26a04358b11539097c1aadb8d11b85c21 Merge and generalize case 3 and case 6 (Pieter Wuille)
402ad5aaca9509d45d861d77eb6431d6e1944f91 Only run sanity check once at the end (Pieter Wuille)
eda8309bfc6a8c94f0b7c076d1cccc86c1011cbc Assert immediately rather than caching failure (Pieter Wuille)
55608455cbed4234f26f62ed9ff500fe5dbc21c4 Make a fuzzer-based copy of the prevector randomized test (Pieter Wuille)
Pull request description:
The current prevector test effectively randomly generates a number of operations to perform on a prevector and a normal vector, and checks consistency between the two.
By converting this into a fuzzer the operations can be targetted rather than random.
ACKs for top commit:
MarcoFalke:
ACK b1d24d1d031a2b2ce67bf846bafa1c3a499b7553 🍬
Tree-SHA512: 2b5c62abcd5fee94f42db03400531484d98c59e7f4308e0e683c61aabcd9ce42f85c5d058d2d5e7f8221124f71d2112b6a5f3c80e5d0fdae265a70647747e92f
cdfb8e7afa7648405dd6b957f47b1c7ab566a076 tests: Add fuzzing harness for HTTPRequest, libevent's evhttp and related functions (practicalswift)
Pull request description:
Add fuzzing harness for `HTTPRequest`, `libevent`'s `evhttp` and related functions.
ACKs for top commit:
laanwj:
ACK cdfb8e7afa7648405dd6b957f47b1c7ab566a076
Tree-SHA512: da481afed5eb3232d3f3d0583094e56050e6234223dfcb356d8567fe0616336eb1b78c5e6821325fc9767e385e5dfaf3c96f0d35ffdb67f18d74f9a9a9464e24
7777e3624fabe4718675b2be8b088697b7ad4d0d scripted-diff: Replace strCommand with msg_type (MarcoFalke)
Pull request description:
Receiving a message is not a command, but simply a message of some type
ACKs for top commit:
promag:
ACK 7777e3624fabe4718675b2be8b088697b7ad4d0d.
naumenkogs:
ACK 7777e36
practicalswift:
ACK 7777e3624fabe4718675b2be8b088697b7ad4d0d -- I've always thought the `strCommand` name is confusing :)
theStack:
ACK 7777e36
Tree-SHA512: 662bac579064c621191916274314b85111cfb4df488f00893ceb16def1c47af4b2a0f34cd7349722099b5a9d23160edb8eb999841f1d64af3e0da02e4870b4bf
283bd72156959f420f13acc7a34e513ca3446025 tests: Add coverage of {,Incremental}DynamicUsage(const std::set<X, Y>& s) to existing fuzzer (practicalswift)
bf76000493082da05bf7258a5038e16fa76cd143 tests: Add fuzzing harness for classes/functions in cuckoocache.h (practicalswift)
57890b2555ca347373109052f6789c23f46bc594 tests: Add fuzzing harness for classes/functions in checkqueue.h (practicalswift)
2df5701e902effa93834d9520690cbaca7e504f3 tests: Add coverage of GetVirtualTransactionSize(...) to existing fuzzer (practicalswift)
7b9a2dc86426926038b2f49d3d4ce4cb64dcd14b tests: Add fuzzing harness for AdditionOverflow(...) (practicalswift)
44fb2a596b4a1aa70253c4145c35be6de68da22a tests: Add fuzzing harness for FeeFilterRounder (practicalswift)
Pull request description:
Includes:
```
tests: Add fuzzing harness for FeeFilterRounder
tests: Add fuzzing harness for classes/functions in checkqueue.h
tests: Add fuzzing harness for classes/functions in cuckoocache.h
tests: Add coverage of {,Incremental}DynamicUsage(const std::set<X, Y>& s) to existing fuzzer
tests: Add coverage of GetVirtualTransactionSize(...) to existing fuzzer
tests: Add fuzzing harness for AdditionOverflow(...)
```
See [`doc/fuzzing.md`](https://github.com/bitcoin/bitcoin/blob/master/doc/fuzzing.md) for information on how to fuzz Bitcoin Core.
ACKs for top commit:
MarcoFalke:
ACK 283bd72156959f420f13acc7a34e513ca3446025
Tree-SHA512: 2361edfb5c47741b22d9fb996836c5250c5a26bc5e956039ea6a0c55ba2d36c78f241d66f85bc02f5b85b9b83d5fde56a5c4702b9d1b7ac4a9a3ae391ca79eaa
Use TestingSetup fixture to fix unregister_all_during_call test not calling
UnregisterBackgroundSignalScheduler, which could trigger an assert in
RegisterBackgroundSignalScheduler when called in later tests
Failure reported by fanquake <fanquake@gmail.com>
https://github.com/bitcoin/bitcoin/pull/18551#issuecomment-610974251
The previous code for MainSignalsInstance::Clear would decrement the reference
count of every interface, including ones that were already Unregister()ed but
still being executed.
fa1a92224dd78de817d15bcda35a8310254e1a54 rpc: Avoid initialization-order-fiasco on static CRPCCommand tables (MarcoFalke)
Pull request description:
Currently the fiasco is only theoretical because all content of the table are compile-time constants. However, the fiasco materializes should they ever become run-time constants (e.g. #18531).
ACKs for top commit:
promag:
ACK fa1a92224dd78de817d15bcda35a8310254e1a54.
practicalswift:
ACK fa1a92224dd78de817d15bcda35a8310254e1a54 -- fiasco bad :)
Tree-SHA512: cccadb0ad56194599b74f04264d74c34fa865958580a850efc6474bbdc56f30cadce6b2e9a6ad5472ff46c3f4c793366acd8090fad409a45b25d961f2d89da19
7a2ecf16df938dd95d3130a46082def7a02338eb Wallet: Change IsMine check in CWallet::DelAddressBook from assert to failure (Luke Dashjr)
2952c46b923042f2de801f319e03ed5c4c4eb735 Wallet: Replace CAddressBookData.name with GetLabel() method (Luke Dashjr)
d7092c392e10889cd7a080b3d22ed6446a59b87a QA: Test that change doesn't turn into non-change when spent in an avoid-reuse wallet (Luke Dashjr)
Pull request description:
Follow-up to #18192, not strictly necessary for 0.20
ACKs for top commit:
MarcoFalke:
re-ACK 7a2ecf16df, only change is adding an assert_equal in the test 🔰
jnewbery:
utACK 7a2ecf16df938dd95d3130a46082def7a02338eb
Tree-SHA512: e0933ee40f705b751697dc27249e1868ed4874254b174ebdd0a7150125d8c818402e66df2371718c7eeb90e67ee2317215fb260aa9b9d7b9b45ee436de2988ff
b5795a788639305bab86a8b3f6b75d6ce81be083 Wallet: Add warning comments and assert to CWallet::DelAddressBook (Luke Dashjr)
6d2905f57aaeb3ec3b63d31043f7673ca10003f2 Wallet: Avoid unnecessary/redundant m_address_book lookups (Luke Dashjr)
c751d886f499257627b308b11ffaa51c22db6cc0 Wallet: Avoid treating change-in-the-addressbook as non-change everywhere (Luke Dashjr)
8e64b8c84bcbd63caea06f3af087af1f0609eaf5 Wallet: New FindAddressBookEntry method to filter out change entries (and skip ->second everywhere) (Luke Dashjr)
65b6bdc2b164343ec3cc3d32a0297daff9e24fec Wallet: Add CAddressBookData::IsChange which returns true iff label has never been set (Luke Dashjr)
144b2f85da4d51bf7d72b987888ddcaf5b429eed Wallet: Require usage of new CAddressBookData::setLabel to change label (Luke Dashjr)
b86cd155f6f661052042048aa7cfc2a397afe4f7 scripted-diff: Wallet: Rename mapAddressBook to m_address_book (Luke Dashjr)
Pull request description:
In many places, our code assumes that presence in the address book indicates a non-change key, and absence of an entry in mapAddressBook indicates change.
This no longer holds true after #13756 (first released in 0.19) since it added a "used" DestData populated even for change addresses. Only avoid-reuse wallets should be affected by this issue.
Thankfully, populating DestData does not write a label to the database, so we can retroactively fix this (so long as the user didn't see the change address and manually assign it a real label).
Fixing it is accomplished by:
* Adding a new bool to CAddressBookData to track if the label has ever been assigned, either by loading one from the database, or by assigning one at runtime.
* `CAddressBookData::IsChange` and `CWallet::FindAddressBookEntry` are new methods to assist in excluding change from code that doesn't expect to see them.
* For safety in merging, `CAddressBookData::name` has been made read-only (the actual data is stored in `m_label`, a new private member, and can be changed only with `setLabel` which updates the `m_change` flag), and `mapAddressBook` has been renamed to `m_address_book` (to force old code to be rebased to compile).
A final commit also does some minor optimisation, avoiding redundant lookups in `m_address_book` when we already have a pointer to the `CAddressBookData`.
ACKs for top commit:
ryanofsky:
Code review ACK b5795a788639305bab86a8b3f6b75d6ce81be083. Pretty clever and nicely implemented fix!
jonatack:
ACK b5795a788639305bab86a8b3f6b75d6ce81be083 nice improvements -- code review, built/ran tests rebased on current master ff53433fe4ed06893d7c4 and tested manually with rpc/cli
jnewbery:
Good fix. utACK b5795a788.
Tree-SHA512: 40525185a0bcc1723f602243c269499ec86ecb298fecb5ef24d626bbdd5e3efece86cdb1084ad7eebf7eeaf251db4a6e056bcd25bc8457b417fcbb53d032ebf0
7b8e15728d1ad058a4b7d7569fd5d5ba6806ca28 rpc: Fix rpcRunLater race in walletpassphrase (João Barbosa)
Pull request description:
Release locks before calling `rpcRunLater`.
Quick explanation: `rpcRunLater` leads to `event_free` which calls `event_del` which can wait for the event callback to finish if it's already running and that callback will try to lock wallet mutex - which is already locked in http thread.
Fixes#14995 , fixes#18482. Best reviewed with whitespace changes hidden.
ACKs for top commit:
MarcoFalke:
ACK 7b8e15728d, only tested that this avoids the node freezing. Did not look at how libevent works or how the deadlock happens or if this breaks other stuff. 📞
ryanofsky:
Code review ACK 7b8e15728d1ad058a4b7d7569fd5d5ba6806ca28. Just updated comment since last review
Tree-SHA512: 17874a2fa7b0e164fb0d7ee4cb7d59650275b8c03476fb291d60af8b758495457660d3912623fb26259fefe84aeba21c0a9e0c6467982ba511f19344ed5413ab
cd3b1569d9ad8e24d3a222aff74e0c254baadf79 Correctly compute redeemScript from witnessScript for signrawtransaction (Andrew Chow)
Pull request description:
`ParsePrevouts` uses `GetScriptForWitness` on the given witnessScript to find the corresponding redeemScript. This is incorrect when the witnessScript is either a P2PK or P2PKH script as it returns the corresponding P2WPK script instead of turning the witnessScript into a P2WSH script. Instead this should make the script a `WitnessV0ScriptHash` destination and get the script for that.
Test cases are also added. These will fail on master with a `redeemScript does not correspond to witnessScript`
Reported on [Bitcointalk](https://bitcointalk.org/index.php?topic=5236818.0)
ACKs for top commit:
MarcoFalke:
weak ACK cd3b1569d9, only checked that the test fails without the code change 🚰
instagibbs:
utACK cd3b1569d9
Tree-SHA512: afac671dbb52ce88bfb4a9ca3dd6065427ad52c9778d0549ad40e9286778f308adad24fb3b3c3089545d7f88c57c53d41224fd7a4bb207550eff2fe06600118f
d6815a2313158862d448733954a73520f223deb6 refactor: drop boost::signals2 in validationinterface (Russell Yanofsky)
Pull request description:
Stop using boost::signals2 internally in validationinterface. Replace with std::list and Add/Remove/Clear/Iterate helper functions.
Motivation for change is to reduce dependencies and avoid issues happening with boost versions before 1.59: https://github.com/bitcoin/bitcoin/issues/18517, https://github.com/bitcoin/bitcoin/pull/18471
ACKs for top commit:
MarcoFalke:
ACK d6815a2313158862d448733954a73520f223deb6
laanwj:
ACK d6815a2313158862d448733954a73520f223deb6
hebasto:
re-ACK d6815a2313158862d448733954a73520f223deb6
promag:
ACK d6815a2313158862d448733954a73520f223deb6.
Tree-SHA512: 4fc0f14a8446e8616cc142af6c3d36815f3254525d30348ba8e4d4bc74c249a5a8c9bc119bdd1be7ebd7abe0b784bc0c5551a3e156a766890cb2fdd891a95919
0eeb0468e7debb1dbe38242769207d22ed52c1df net: Hardcoded seeds update for 0.20 (Wladimir J. van der Laan)
Pull request description:
Update hardcoded seeds from http://bitcoin.sipa.be/seeds.txt.gz,
according to release process.
Output from makeseeds.py:
```
IPv4 IPv6 Onion Pass
1364173 244127 2454 Initial
1364173 244127 2454 Skip entries with invalid address
1129552 213117 2345 After removing duplicates
1129548 213117 2345 Skip entries from suspicious hosts
338216 191944 2249 Enforce minimal number of blocks
336851 188993 2189 Require service bit 1
6998 1520 150 Require minimum uptime
5682 1290 89 Require a known and recent user agent
5622 1279 89 Filter out hosts with multiple bitcoin ports
512 146 89 Look up ASNs and limit results per ASN and per net
```
Top commit has no ACKs.
Tree-SHA512: ce1c2cda18dd5bd22586a5283a0877f3bd890437cc29dc1d85452ba4a4d28032f591c8b37f3329e8e649556cf83750b6949a068fad76d1773853d93014609da0
9e071b00898aedd9632f105a22d976dc6dbc84b1 test: remove rapidcheck integration and tests (fanquake)
Pull request description:
Whilst the property tests are interesting, ultimately [rapidcheck](https://github.com/emil-e/rapidcheck) integration in this repository has not gained much traction. We have a limited number of tests, and they are rarely (if ever) run. Have discussed this with Chris Stewart.
ACKs for top commit:
practicalswift:
ACK 9e071b00898aedd9632f105a22d976dc6dbc84b1
Tree-SHA512: d0c12af3163382eee8413da420c63e39265a7b700709a05d518445832d45e049aed9508e32524db5228fe3ac114609a00b7bb890be047c07032e44a5ef4611e9
