noexcept is default for destructors as of c++11. By throwing in reverselock's
destructor if it's lock has been tampered with, the likely result is
std::terminate being called. Indeed that happened before this change.
Once reverselock has taken another lock (its ctor didn't throw), it makes no
sense to try to grab or lock the parent lock. That is be broken/undefined
behavior depending on the parent lock's implementation, but it shouldn't cause
the reverselock to fail to re-lock when destroyed.
To avoid those problems, simply swap the parent lock's contents with a dummy
for the duration of the lock. That will ensure that any undefined behavior is
caught at the call-site rather than the reverse lock's destruction.
Barring a failed mutex unlock which would be indicative of a larger problem,
the destructor should now never throw.
Get rid of underscore prefixes for include guards.
This is because they may conflict with reserved names. Also took
this opportunity to make the guards in the zcash subdir a bit
more uniform by having them all conform to ZC_<...>_H_
This closes#1272
Remove Boost conditional compilation.
Boost is currently at version 106200 in the depends subsystem
so it is safe to remove code the refers to earlier versions.
This closes#2445
This is because they may conflict with reserved names. Also took
this opportunity to make the guards in the zcash subdir a bit
more uniform by having them all conform to ZC_<...>_H_
This closes#1272
Previously libsnark would try to build libgtest if /usr/src/gtest
exists on the build machine. This caused issues because the
version of libgtest in /usr/src isn't necessarily the same as
the one that the rest of zcash's gtests were built with.
Get the sec-hard tests to run correctly.
This fixes the way arguments were passed to security-check, and also
a typo in how BIND_NOW was being searched for in a list.
Also fix how symbol-check is invoked although that script isn't
currently used.
This fixes the way arguments were passed to security-check, and also
a typo in how BIND_NOW was being searched for in a list.
Also fix how symbol-check is invoked although that script isn't
currently used.
The new parameter is to satisfy the principle of least astonishment
by providing a sensible default for the maximum number of transparent
inputs to shield. If users do not configure -mempooltxinputlimit
it is possible for them to create transactions with hundreds of
inputs which suffer from mining delay, due to the current state of
the network where some miners have configured -mempooltxinputlimit
as a way to deal with the problem of quadratic hashing.
Fix various thread assertion errors caused during shutdown.
Cherry-picked from the following upstream PRs:
- bitcoin/bitcoin#6719
- bitcoin/bitcoin#6990
- bitcoin/bitcoin#8421
- Second commit only in this PR
- bitcoin/bitcoin#11006
I've cherry-picked the relevant commits, along with a note in each commit referring to the original Bitcoin commit ID (and the Zcash issue numbers where applicable). I've tested each issue with/without these patches applied.
Closes#2214, #2334, and #2554.
Move libsnark in-repo as a git subtree
This PR pulls in the libsnark subtree at the exact commit that we currently fetch via the depends system. To verify:
```
$ ./contrib/devtools/git-subtree-check.sh src/snark
src/snark in HEAD was last updated to upstream commit 9ada3f84ab484c57b2247c2f41091fd6a0916573 (tree c10a38c759)
src/snark in HEAD currently refers to tree 34e916d3f6
:100644 100644 427f4f4ce913e54da68b M Makefile
:040000 040000 42f29e42d1dd73536163 M src
FAIL: subtree directory tree doesn't match subtree commit tree
```
This shows that there are changes relative to what we currently use, due to the later commits in the PR. If we exclude them, we see that the code is identical:
```
$ git checkout 26a8f68ea8
$ ./contrib/devtools/git-subtree-check.sh src/snark
src/snark in HEAD was last updated to upstream commit 9ada3f84ab484c57b2247c2f41091fd6a0916573 (tree c10a38c759)
src/snark in HEAD currently refers to tree c10a38c759
GOOD
```
Closes#820.
