Block pipeline #3727
Block pipeline #3727
Conversation
| else | ||
| { | ||
| // If there is no previous block, the balance must already be 0 and therefore cannot be a send | ||
| if (block->previous ().is_zero ()) |
There was a problem hiding this comment.
I do not know if this check is complete or makes sense.
At this point, the block could be even a receive, since at this point we wouldn't know that a receive doesn't have a corresponding send block.
It coudl also be a change block. I am not sure what this check gives us.
| #include <nano/secure/ledger.hpp> | ||
| #include <nano/secure/store.hpp> | ||
|
|
||
| void nano::block_pipeline::block_position_filter::sink (std::shared_ptr<nano::block> block, std::shared_ptr<nano::block> previous) |
There was a problem hiding this comment.
This filter mentions that it checks something to do with epoch going backwards.
I do not understand what the means.
However, I see no reference to any epoch blocks, is that a sign that something is missing?
Also I see no reference to work difficulties and epoch v2 is about work difficulty.
This simply seems to check that legacy blocks never follows a state block.
nano/lib/stats.cpp
Outdated
| break; | ||
| case nano::stat::detail::determinate_signer_filter_reject: | ||
| res = "determinate_signer_filter_reject"; | ||
| break; |
There was a problem hiding this comment.
We could add a default case with an debug_assert(0) to catch any missed cases.
There was a problem hiding this comment.
If we turned on warning to error conversion for missing case statements the CI could catch it at compile time.
clemahieu
force-pushed
the
block_pipeline
branch
6 times, most recently
from
47bf309 to
9c7ea8a
Compare
clemahieu
force-pushed
the
block_pipeline
branch
from
16951be to
5e78000
Compare
…essing steps in to small components focused on a specific part of processing blocks.
clemahieu
force-pushed
the
block_pipeline
branch
from
5e78000 to
66de849
Compare
# Conflicts: # nano/lib/stats.cpp # nano/lib/stats.hpp # nano/node/blockprocessor.cpp # submodules/cpptoml # submodules/diskhash # submodules/flatbuffers # submodules/gtest # submodules/lmdb # submodules/miniupnp # submodules/phc-winner-argon2
… may miss a trigger notification while being processed by the unchecked_map thread.
…ck sequence. This test stops block processing so no blocks will be inserted which means a sequence of block would have gaps that the pipeline would reject for gaps.
clemahieu
force-pushed
the
block_pipeline
branch
from
3686907 to
32dea05
Compare
…ying it with .flush ().
# Conflicts: # nano/lib/stats.cpp # nano/lib/stats.hpp
# Conflicts: # nano/secure/ledger.cpp # nano/secure/ledger.hpp
clemahieu
force-pushed
the
block_pipeline
branch
from
1f608a3 to
39b7d7e
Compare
clemahieu
force-pushed
the
block_pipeline
branch
from
39b7d7e to
065845e
Compare
Copy accounts locally within signature verification.
clemahieu
force-pushed
the
block_pipeline
branch
from
da886a1 to
80205a0
Compare
…te logic path or tracking a separate container.
clemahieu
force-pushed
the
block_pipeline
branch
from
214fed2 to
d374ccf
Compare
…ng it separately in epoch_restrictions_filter.
This is a refactor of the block processor which breaks the block processing steps in to small components focused on a specific part of processing blocks.