[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Thread skyline construction through stencil interpretation (issue 58
From: |
dak |
Subject: |
Re: Thread skyline construction through stencil interpretation (issue 581960043 by address@hidden) |
Date: |
Mon, 27 Apr 2020 13:31:55 -0700 |
That's the kind of stuff that warrants compiling a few scores with
-ddebug-skylines and looking rather closely for kinks.
https://codereview.appspot.com/581960043/diff/561710044/lily/stencil-integral.cc
File lily/stencil-integral.cc (right):
https://codereview.appspot.com/581960043/diff/561710044/lily/stencil-integral.cc#newcode245
lily/stencil-integral.cc:245: for (vsize i = 0; i < 8; i += 2)
On 2020/04/27 19:56:59, hahnjo wrote:
> hm, can we avoid hardcoding the size twice - or even get rid of it
completely? I
> think you can write
> Offset points[] = { ... };
> and
> i < sizeof(points) / sizeof(points[0])
Offset points[2][] = { ... };
for (auto &i : points)
skyline->add_segment (transform, i[0], i[1]);
May want to be auto i, I am not yet all too acquainted with C++11
hotness.
https://codereview.appspot.com/581960043/diff/561710044/lily/stencil-integral.cc#newcode253
lily/stencil-integral.cc:253: vector<Offset> points;
Seems unused.
https://codereview.appspot.com/581960043/