5db506ba59 tests: Add option --valgrind to run nodes under valgrind in the functional tests (practicalswift)
Pull request description:
What is better than fixing bugs? Fixing entire bug classes of course! :)
Add option `--valgrind` to run the functional tests under Valgrind.
Regular functional testing under Valgrind would have caught many of the uninitialized reads we've seen historically.
Let's kill this bug class once and for all: let's never use an uninitialized value ever again. Or at least not one that would be triggered by running the functional tests! :)
My hope is that this addition will make it super-easy to run the functional tests under Valgrind and thus increase the probability of people making use of it :)
Hopefully `test/functional/test_runner.py --valgrind` will become a natural part of the pre-release QA process.
**Usage:**
```
$ test/functional/test_runner.py --help
…
--valgrind run nodes under the valgrind memory error detector:
expect at least a ~10x slowdown, valgrind 3.14 or
later required
```
**Live demo:**
First, let's re-introduce a memory bug by reverting the recent P2P uninitialized read bug fix from PR #17624 ("net: Fix an uninitialized read in ProcessMessage(…, "tx", …) when receiving a transaction we already have").
```
$ git diff
diff --git a/src/consensus/validation.h b/src/consensus/validation.h
index 3401eb64c..940adea33 100644
--- a/src/consensus/validation.h
+++ b/src/consensus/validation.h
@@ -114,7 +114,7 @@ inline ValidationState::~ValidationState() {};
class TxValidationState : public ValidationState {
private:
- TxValidationResult m_result = TxValidationResult::TX_RESULT_UNSET;
+ TxValidationResult m_result;
public:
bool Invalid(TxValidationResult result,
const std::string &reject_reason="",
```
Second, let's test as normal without Valgrind:
```
$ test/functional/p2p_segwit.py -l INFO
2019-11-28T09:30:42.810000Z TestFramework (INFO): Initializing test directory /tmp/bitcoin_func_test__fc8q3qo
…
2019-11-28T09:31:57.187000Z TestFramework (INFO): Subtest: test_non_standard_witness_blinding (Segwit active = True)
…
2019-11-28T09:32:08.265000Z TestFramework (INFO): Tests successful
```
Third, let's test with `--valgrind` and see if the test fail (as we expect) when the unitialized value is used:
```
$ test/functional/p2p_segwit.py -l INFO --valgrind
2019-11-28T09:32:33.018000Z TestFramework (INFO): Initializing test directory /tmp/bitcoin_func_test_gtjecx2l
…
2019-11-28T09:40:36.702000Z TestFramework (INFO): Subtest: test_non_standard_witness_blinding (Segwit active = True)
2019-11-28T09:40:37.813000Z TestFramework (ERROR): Assertion failed
ConnectionRefusedError: [Errno 111] Connection refused
```
ACKs for top commit:
MarcoFalke:
ACK 5db506ba59
jonatack:
ACK 5db506ba59
Tree-SHA512: 2eaecacf4da166febad88b2a8ee6d7ac2bcd38d4c1892ca39516b6343e8f8c8814edf5eaf14c90f11a069a0389d24f0713076112ac284de987e72fc5f6cc3795
773d4572a4 Mark PSBTs spending unspendable outputs as invalid in analysis (Andrew Chow)
638e40cb60 Have a PSBTAnalysis state that indicates invalid PSBT (Andrew Chow)
Pull request description:
When analyzing an unspendable PSBT, report that it is unspendable and exit analysis early.
ACKs for top commit:
Sjors:
ACK 773d457
instagibbs:
After some thought ACK 773d4572a4
Tree-SHA512: 99b0cb2fa1ea37593fc65a20effe881639d69ddeeecf5197bc87bc7f2220cbeb40f1d429d517e4d27f2e9fb563a00cd845d2b4b1ce05246a75a6cb56fb9b0ba5
597d10ceb9 tests: Add fuzzing harness for various functions consuming only integrals (practicalswift)
575383b3e1 tests: Add corpora suppression (FUZZERS_MISSING_CORPORA) for fuzzers missing in https://github.com/bitcoin-core/qa-assets/tree/master/fuzz_seed_corpus (practicalswift)
Pull request description:
Add fuzzing harness for various functions consuming only integrals.
**Testing this PR**
Run:
```
$ CC=clang CXX=clang++ ./configure --enable-fuzz \
--with-sanitizers=address,fuzzer,undefined
$ make
$ src/test/fuzz/integer
```
Top commit has no ACKs.
Tree-SHA512: f0ccbd63671636f8e661385b682e16ad287fef8f92e7f91327ee2093afc36fcd424e1646fe90279388e28a760bcc795766eb80cf6375e0f873efff37fc7e2393
d5766f223f tests: Add corpora suppression (FUZZERS_MISSING_CORPORA) for fuzzers missing in https://github.com/bitcoin-core/qa-assets/tree/master/fuzz_seed_corpus (practicalswift)
e75ecb91c7 tests: Add fuzzing harness for various CTxOut related functions (practicalswift)
ce935292c0 tests: Add fuzzing harness for various CTxIn related functions (practicalswift)
Pull request description:
Add fuzzing harness for various `CTx{In,Out}` related functions.
**Testing this PR**
Run:
```
$ CC=clang CXX=clang++ ./configure --enable-fuzz --with-sanitizers=address,fuzzer,undefined
$ make
$ src/test/fuzz/tx_in
…
$ src/test/fuzz/tx_out
…
# And to to quickly verify that the relevant code regions are triggered, that the
# fuzzing throughput seems reasonable, etc.
$ contrib/devtools/test_fuzzing_harnesses.sh '^tx_'
```
`test_fuzzing_harnesses.sh` can be found in PR #17000.
Top commit has no ACKs.
Tree-SHA512: f1374307a2581ebc3968d012ea2438061bbb84ece068e584fae9750669a6cd003723dde14db88e77c9579281ecd4eaa2a7ff0614f253d8c075e6dd16dd2e68d5
dddd09eb33 test: Wait until mempool is loaded in wallet_abandonconflict (MarcoFalke)
Pull request description:
This might or might not fix intermittent issues such as https://ci.appveyor.com/project/DrahtBot/bitcoin/builds/28724018#L4091
I believe the mempool was not loaded fully after the restart, in which case it was not dumped either on the next restart. Thus, the previous mempool was attempted to be loaded a second time, which succeeded and contained the txs.
ACKs for top commit:
laanwj:
ACK dddd09eb33
Tree-SHA512: ab7061f946b5e5388f825dddceadb125f5197b24af3a7fcf1e700235d106a323419a56bfb4d84a2e27442e0de63e540c623b704343d83a98deaab3c02fcbdcbe
In accordance with #17314, Removing noisy fields from -getinfo. Fields removed: protocolversion, walletversion and keypoololdest. In addition to changing bitcoin-cli -getinfo, there is another change to test/functional/interface_bitcoin_cli.py. This change deletes tests that utilize removed -getinfo calls.
fabd71076c ci: Print free disk space (MarcoFalke)
fad9fdbea5 test: Properly deserialize integers in little-endian (MarcoFalke)
fa94fc10c8 ci: Run functional tests on s390x (MarcoFalke)
Pull request description:
Top commit has no ACKs.
Tree-SHA512: 98ba77eb56f283131fdaeb393fda86cc308f1bf9781e1e0e5736b8d616528dc8ff2e494d55ba107c138083025c66a59e382fcfa9962d4349a5fd6cbbc52484c3
02afb0c550 Fix origfee return for bumpfee with feerate arg (Gregory Sanders)
Pull request description:
fixes https://github.com/bitcoin/bitcoin/issues/17642 and adds a simple test that would have caught it
ACKs for top commit:
achow101:
ACK 02afb0c550
Tree-SHA512: 303e392e05407f204dffe360689b5bb5dc77fd462dd0e489bc0b6c8f94f89ab7fe2bd8cb47e4dc6dc5c23a619826d15f3bf6b02b2c8e96402fbb51953c462e2d
eadd1304c8 tests: Add a test for funding with sufficient preset inputs and subtractFeeFromOutputs (Andrew Chow)
ff330badd4 Default to bnb_used = false as there are many cases where BnB is not used (Andrew Chow)
Pull request description:
#17290 introduced a bug where, when we had preset inputs that covered the amount being sent and subtractFeeFrromOutputs was being used, transaction funding would result in a `Fee exceeds maximum configured by -maxtxfee` error. This was happening because we weren't setting `bnb_used = false` when the preset inputs were used as it should have been. This resulted in a too high fee because the change would go to fees accidentally.
Apparently this particular case doesn't have a test, so I've added one as well.
ACKs for top commit:
Sjors:
ACK eadd130. I can't get this new test to fail on macOS (without this PR). It passes whether or not I compile with `--enable-debug`. It does fail on Ubuntu. Yay undefined behavior... Anyway, it's a useful test.
fanquake:
ACK eadd1304c8
instagibbs:
utACK eadd1304c8
Tree-SHA512: 7286c321f78666eea558cc591174630d210263594df41cab1065417510591ee514ade0e1d0cec8af09a785757da68de82592b013e8fe8d4966cec3254368706e
b84e776fd1 wallet_importmulti: use addresses of the same type as being imported (Andrew Chow)
Pull request description:
When constructing an import from the solving data of an address, make sure that the original address is the same type as the one that will be imported.
See also: https://github.com/bitcoin/bitcoin/pull/17374#issuecomment-550036931
Part of #17261
ACKs for top commit:
Sjors:
Code review ACK b84e776
meshcollider:
Tested re-ACK b84e776fd1
Tree-SHA512: 53c49c63af8cbade0116a62beddc77df1a411d8ed76571c3053f6aff096f41a5325421a188bab3dcacfda69bb28fdff6ba921ddd80f29c4abbadb3b58fda884c
ea3c7e585c test: Remove libssl-dev packages from CI scripts (Wladimir J. van der Laan)
7ea55264b9 test: remove lsan suppression for libcrypto (Wladimir J. van der Laan)
2d7066527a build: remove libcrypto as internal dependency in libbitcoinconsensus.pc (Wladimir J. van der Laan)
278751ea11 doc: Remove ssl as a required dependency from build-unix (Wladimir J. van der Laan)
Pull request description:
Some doc and build cleanups following #17265.
I intentionally left the libssl-dev install in `gitian-win-signer.yml`, as it's necessary for the ossl signer.
ACKs for top commit:
MarcoFalke:
ACK ea3c7e585c 🗯
jamesob:
ACK ea3c7e585c
practicalswift:
ACK ea3c7e585c - nice!
fanquake:
ACK ea3c7e585c - thanks.
Tree-SHA512: 67ea35bdd6d6e512d69e6734713534c88cae033a2ed695677ea15c3e3d5ff570374e342775c88e60877fa43a19047853e7b2a433e2c9a4349a5c423726a7457e
1c23ea5fe6 test: fix bitcoind already running warnings on macOS (fanquake)
Pull request description:
On macOS, `pidof` installed via brew returns b'' rather than None.
Account for this, to remove spurious warnings from the test_runner.
ACKs for top commit:
laanwj:
ACK 1c23ea5fe6
Tree-SHA512: 640f4323d4105eac5c7abb52daf80486d5d3b4a074720490ceeb97c3dd8d73a3de9a988d2550f1e2076c620bb10d452b2959d8b723d2ee64f499878909824e31
fa2ec9f451 fuzz: Bump timeout in test_runner to accomodate for slow arm64 CPUs (MarcoFalke)
fa6e01b2f3 ci: Use clang-8 for fuzzing to run on aarch64 ci systems (MarcoFalke)
Pull request description:
Ubuntu bionic clang is clang version 6, which does not come with libfuzzer. So the ci system breaks down when run on aarch64.
Fix that by using clang-8
For reference, the previous error on my ci system was:
```
/usr/bin/ld: cannot find /usr/lib/llvm-6.0/lib/clang/6.0.0/lib/linux/libclang_rt.fuzzer-aarch64.a: No such file or directory
ACKs for top commit:
laanwj:
ACK fa2ec9f451
Tree-SHA512: 4954dbc36c444d1ae145290115eea6291753c9810c92003ab8d75433c3fe3bfee439d3a99dc394418275527157a8b89f04038c8b16e08c69ec9ded50fb869e70
