kicad-developers team mailing list archive
-
kicad-developers team
-
Mailing list archive
-
Message #31033
Re: [PATCH] Fix some performance issues
On 9/26/2017 3:56 PM, Camille 019 wrote:
> Hi,
>
> Ok, I'll keep the patchset synchronized with the master branch, for
> future inclusion. When will the feature freeze take effect ?
I was hoping it would be by the end of September but obviously that is
not going to happen. I have my symbol library table work done. I just
have to do some more testing. I plan on merging next week since I will
be available that weekend to fix any issues as this will be a disruptive
change. At that time, I will make a last call before the feature freeze
so hopefully no later than the end of October until the stable 5 branch.
>
> I agree my patchset has no real impact on the performance. I mainly
> wanted to test clang-tidy on a consistent source code base. I also think
> that keeping automated checks empty of warning can help to improve the
> overall code quality.
>
> Regards,
>
> --
> Camille
>
>
> Le lun. 25 sept. 2017 à 16:21, Wayne Stambaugh <stambaughw@xxxxxxxxx> a
> écrit :
>> On 9/25/2017 4:07 AM, Maciej Sumiński wrote:
>>
>> Hi Camille, I think the attached patches are sensible, even though
>> I hope most compilers apply optimizations to avoid performance
>> penalties whenever possible.
>>
>> I agree although I'm not sure about the compiler optimization. A quick
>> look at the patches and I'm not sure the performance gains will be
>> that noticeable. Most of the changes are in code that is called once
>> on demand in response to a user request as opposed to the drawing or
>> drc code where large numbers of calls happen frequently. That being
>> said, we really should be doing a better job of not passing entire
>> objects on the stack unless there is a good reason to do so.
>>
>> If the attached patches are autogenerated, then I would kindly ask
>> you to generate them when we reach the feature freeze stage. There
>> are a few branches that await to be merged and I am not sure
>> whether your patches would not create too many conflicts. During
>> the feature freeze we anticipate the code changes will not be as
>> disruptive as they are now.
>>
>> I second this. There are quite a few outstanding patches that will
>> likely have enough conflicts so I would rather not pile more on top of
>> that.
>>
>> Regards, Orson On 09/23/2017 12:17 PM, Camille 019 wrote:
>>
>> Hi all, I recently play with clang-tidy on the KiCad source
>> code. Here is a set of 2 patch which fix the
>> unnecessary-copy-initialization<https://clang.llvm.org/extra/clang-tidy/checks/performance-unnecessary-copy-initialization.html>
>> and
>> unnecessary-value-param<https://clang.llvm.org/extra/clang-tidy/checks/performance-unnecessary-value-param.html>
>> checks. I will run more performance checks in the future. The
>> next two: -
>> for-range-copy<https://clang.llvm.org/extra/clang-tidy/checks/performance-for-range-copy.html>
>> -
>> type-promotion-in-math-fn<https://clang.llvm.org/extra/clang-tidy/checks/performance-type-promotion-in-math-fn.html>
>> Best regards, -- Camille
>> _______________________________________________ Mailing list:
>> https://launchpad.net/~kicad-developers Post to :
>> kicad-developers@xxxxxxxxxxxxxxxxxxx
>> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx> Unsubscribe :
>> https://launchpad.net/~kicad-developers More help :
>> https://help.launchpad.net/ListHelp
>>
>> _______________________________________________ Mailing list:
>> https://launchpad.net/~kicad-developers Post to :
>> kicad-developers@xxxxxxxxxxxxxxxxxxx
>> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx> Unsubscribe :
>> https://launchpad.net/~kicad-developers More help :
>> https://help.launchpad.net/ListHelp
>>
>> _______________________________________________ Mailing list:
>> https://launchpad.net/~kicad-developers Post to :
>> kicad-developers@xxxxxxxxxxxxxxxxxxx
>> <mailto:kicad-developers@xxxxxxxxxxxxxxxxxxx> Unsubscribe :
>> https://launchpad.net/~kicad-developers More help :
>> https://help.launchpad.net/ListHelp
References