Cleanup #60
Cleanup #60
Conversation
… variables. This should be superceded in C++-17 with "optionals".
|
@trachten There are some compiler errors on your last commit here (869eccd). Also, do you use any compiler flags other than what is specified in |
|
Strange …. I’m just using llvm on a mac.
I’ve updated to fix the break (it compiles fine on my machines).
… On Jun 11, 2020, at 6:21 PM, Novak Boškov ***@***.***> wrote:
@trachten <https://github.com/trachten> There are some compiler errors <https://github.com/trachten/cpisync/runs/763207089> on your last commit here (869eccd <869eccd>).
Also, do you use any compiler flags other than what is specified in CmakeList.txt? Could you also share the compiler version you use with this project? I see you mention some compiler warnings in some of your commit messages and I'm just wondering if we use different compilers.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub <#60 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACM5LU6LCDSA7I57IXF6YJ3RWFKF7ANCNFSM4N2ZFMZA>.
---
Prof. Ari Trachtenberg ECE, Boston University
trachten@bu.edu http://people.bu.edu/trachten
|
| _commit_relocation_chain(relocStack); | ||
| itemsCount++; | ||
| return true; | ||
| } | ||
| } catch (overflow_error ignored) {} |
There was a problem hiding this comment.
This is within the cuckoo relocations loop.
In mostly full filters (say alpha>80) this exception is going to be thrown a lot. Once per node in the cuckoo graph. That might make our insertions slower.
Is it because of ignored that this will not be much slower? Is it because ignored short-circuits the exception handler traversal?
The two try-catchs above are different, they are called once per insertion.
There was a problem hiding this comment.
A good optimizer should optimize them out - see if you can do a quick profile to see whether it is really a problem?
There was a problem hiding this comment.
I tried it very naively:
int main() {
for (size_t ii=0; ii<4000; ii++) {
Cuckoo c = Cuckoo(7, 4, 1024, 500);
c.insert(DataObject(ZZ(Cuckoo::_rand(0, 1 << 12))));
}
}
And, well, Linux perf shows that both if and try-cache (with GCC -O3) spend about the same estimated number of cycles in insert. Plain time says the same...
No significant diference.
|
Great!
… On Jun 11, 2020, at 11:11 PM, 6/11/20, Novak Boškov ***@***.***> wrote:
@novakboskov commented on this pull request.
In src/Syncs/Cuckoo.cpp <#60 (comment)>:
> _commit_relocation_chain(relocStack);
itemsCount++;
return true;
- }
+ } catch (overflow_error ignored) {}
I tried it very naively:
int main() {
for (size_t ii=0; ii<4000; ii++) {
Cuckoo c = Cuckoo(7, 4, 1024, 500);
c.insert(DataObject(ZZ(Cuckoo::_rand(0, 1 << 12))));
}
}
And, well, Linux perf show both if and try-cache (with GCC -O3) spend about the same estimated number of cycles in insert. Plain time says the same.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub <#60 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACM5LU4RVV75I4IOECWIM33RWGMFJANCNFSM4N2ZFMZA>.
---
Ari Trachtenberg, Boston University
http://people.bu.edu/trachten mailto:trachten@bu.edu
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Random quote of the day:
A day without sunshine is like, night.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
Some more commits.