Core Improvements and Bugfixes

[SmartArray]

Core Improvements and Bugfixes

This PR fixes several bugs that were detected by fixing the functional tests. Those fixes include for instance

• Dandelion bugfixes (mempool bugs)
• RPC bugfixes (generatetoaddress, difficulties)
• Fee improvements
• Introduction of easypow parameter for RegTest
• Several refactorings
• Fixed testnet issues
• Bumped protocol versions
• Proper handling of WTXID protocol command

[JaredTate]

ACK. Awesome work @SmartArray! This seems to finally have eliminated the crash bug with dandelion. Going commit by commit it all looks good. The client compiles, runs, and works as expected. One question I had was the commit about DEFAULT_MIN_RELAY_TX_FEE = 100000. Thats 1/100 of a DGB. Should we make it 1/10?

[gto90]

ACK

Excellent work on this @SmartArray ! It is absolutely refreshing to be so much closer to having the functional test suite fully operational. I'm looking forward to review #90 !

[barrystyle]

So it was this legacy code that was causing a transaction to effectively be removed twice?
9be9458

However this identical code appears in the 0.17 branch as well:
https://github.com/DigiByte-Core/digibyte/blob/develop/src/txmempool.cpp#L526-L530

[barrystyle]

Receive this on feature/8.22.0 with pull request #89 applied (note that digibyted itself is fine).

[barrystyle]

Receive this on feature/8.22.0 with pull request #89 applied (note that digibyted itself is fine).

If you remove the additional primitives/block.cpp added to the Makefile.am:

diff --git a/src/Makefile.am b/src/Makefile.am
index 9ba20eadcc..42f90ee164 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -668,6 +668,7 @@ libdigibyte_cli_a_SOURCES = \
   compat/stdin.h \
   compat/stdin.cpp \
   rpc/client.cpp \
+  primitives/block.cpp \
   $(DIGIBYTE_CORE_H)

and instead make this change to Makefile.am:

diff --git a/src/Makefile.am b/src/Makefile.am
index 9ba20eadc..40573435c 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -724,6 +724,8 @@ endif
 digibyte_cli_LDADD = \
   $(LIBDIGIBYTE_CLI) \
   $(LIBUNIVALUE) \
+  $(LIBDIGIBYTE_COMMON) \
+  $(LIBDIGIBYTE_CONSENSUS) \
   $(LIBDIGIBYTE_UTIL) \
   $(LIBDIGIBYTE_CRYPTO)

It will compile correctly with no symbol errors.

[barrystyle]

Upon running unit tests:

root@loderunner:~/digibyte/src/test# ./test_digibyte 
Running 470 test cases...
test_digibyte: test/util/setup_common.cpp:226: TestChain100Setup::TestChain100Setup(): Assertion `m_node.chainman->ActiveChain().Tip()->GetBlockHash().ToString() == "1b14e04a06ce040bf88b9de8f4c2c4bb38c125acefcbdc2fc841d18bb02cdff6"' failed.
unknown location(0): fatal error: in "blockfilter_index_tests/blockfilter_index_initial_sync": signal: SIGABRT (application abort requested)
test/blockfilter_index_tests.cpp(105): last checkpoint: "blockfilter_index_initial_sync" fixture ctor
test_digibyte: util/system.cpp:654: void ArgsManager::AddArg(const string&, const string&, unsigned int, const OptionsCategory&): Assertion `ret.second' failed.

[SmartArray]

Thank you @barrystyle for testing this.

Yes, I think this is the correct way to deal with the linking error.

There were several issues regarding the mempool. And yes, the code you highlighted was absolutely in there before. That would explain why reorgs caused the client to misbehaving on occasion.

Again, thank you for checking this 🙂

I will apply and check the fix soon!

[SmartArray]

Upon running unit tests:

root@loderunner:~/digibyte/src/test# ./test_digibyte 

Running 470 test cases...

test_digibyte: test/util/setup_common.cpp:226: TestChain100Setup::TestChain100Setup(): Assertion `m_node.chainman->ActiveChain().Tip()->GetBlockHash().ToString() == "1b14e04a06ce040bf88b9de8f4c2c4bb38c125acefcbdc2fc841d18bb02cdff6"' failed.

unknown location(0): fatal error: in "blockfilter_index_tests/blockfilter_index_initial_sync": signal: SIGABRT (application abort requested)

test/blockfilter_index_tests.cpp(105): last checkpoint: "blockfilter_index_initial_sync" fixture ctor

test_digibyte: util/system.cpp:654: void ArgsManager::AddArg(const string&, const string&, unsigned int, const OptionsCategory&): Assertion `ret.second' failed.

Found that too. This is because I had to change some network params

[SmartArray]

Hey guys, please try to ./autogen.sh, ./configure and make once more! I have integrated the change as proposed by @barrystyle

[barrystyle]

ACK Happy with these changes

[ycagel]

@barrystyle Thanks for the review. Are you able to do PR 90?

[gto90]

Very strong work @SmartArray ! Thank you for all your hard effort here.

[barrystyle]

@barrystyle Thanks for the review. Are you able to do PR 90?

can do.