Enable some optimization flags in Release

[isaacbrodsky]

I wasn't able to enable a few of the flags from that PR due to compiler differences and the flags not being available in the compiler I tested with. I turned on interprocedural optimization by default, and added -O2 -finline-functions.

From my quick test, interprocedural optimization provides a small benefit in most benchmarks, and a more noticeable benefit in benchmarkKRing. The other compiler flags (-O2...) seemed to have a fairly small effect.

Before:

$ ./bin/benchmarkH3Api
	-- geoToH3: 0.697000 microseconds per iteration (10000 iterations)
	-- h3ToGeo: 0.406500 microseconds per iteration (10000 iterations)
	-- h3ToGeoBoundary: 2.181400 microseconds per iteration (10000 iterations)
$ ./bin/benchmarkH3Line
	-- h3LineNear: 14.893200 microseconds per iteration (10000 iterations)
	-- h3LineFar: 660.081000 microseconds per iteration (1000 iterations)
$ ./bin/benchmarkH3SetToLinkedGeo
	-- h3SetToLinkedGeoRing2: 62.637000 microseconds per iteration (10000 iterations)
	-- h3SetToLinkedGeoDonut: 23.926700 microseconds per iteration (10000 iterations)
	-- h3SetToLinkedGeoNestedDonuts: 97.380100 microseconds per iteration (10000 iterations)
$ ./bin/benchmarkKRing
	-- kRing10: 8.973500 microseconds per iteration (10000 iterations)
	-- kRing20: 31.355800 microseconds per iteration (10000 iterations)
	-- kRing30: 63.647000 microseconds per iteration (10000 iterations)
	-- kRing40: 108.356200 microseconds per iteration (10000 iterations)
	-- kRingPentagon10: 203.794000 microseconds per iteration (500 iterations)
	-- kRingPentagon20: 1782.148000 microseconds per iteration (500 iterations)
	-- kRingPentagon30: 5789.460000 microseconds per iteration (50 iterations)
	-- kRingPentagon40: 13681.900000 microseconds per iteration (10 iterations)
$ ./bin/benchmarkPolyfill
	-- polyfillSF: 1237.274000 microseconds per iteration (500 iterations)
	-- polyfillAlameda: 1704.760000 microseconds per iteration (500 iterations)
	-- polyfillSouthernExpansion: 55909.100000 microseconds per iteration (10 iterations)
$ ./bin/benchmarkPolygon
	-- pointInsideGeofenceSmall: 0.027360 microseconds per iteration (100000 iterations)
	-- pointInsideGeofenceLarge: 0.233560 microseconds per iteration (100000 iterations)
	-- bboxFromGeofenceSmall: 0.023000 microseconds per iteration (100000 iterations)
	-- bboxFromGeofenceLarge: 0.292130 microseconds per iteration (100000 iterations)

After:

$ ./bin/benchmarkH3Api
	-- geoToH3: 0.595500 microseconds per iteration (10000 iterations)
	-- h3ToGeo: 0.385000 microseconds per iteration (10000 iterations)
	-- h3ToGeoBoundary: 2.120800 microseconds per iteration (10000 iterations)
$ ./bin/benchmarkH3Line
	-- h3LineNear: 12.015000 microseconds per iteration (10000 iterations)
	-- h3LineFar: 540.725000 microseconds per iteration (1000 iterations)
$ ./bin/benchmarkH3SetToLinkedGeo
	-- h3SetToLinkedGeoRing2: 59.331000 microseconds per iteration (10000 iterations)
	-- h3SetToLinkedGeoDonut: 23.104300 microseconds per iteration (10000 iterations)
	-- h3SetToLinkedGeoNestedDonuts: 92.236200 microseconds per iteration (10000 iterations)
$ ./bin/benchmarkKRing
	-- kRing10: 6.175500 microseconds per iteration (10000 iterations)
	-- kRing20: 19.483500 microseconds per iteration (10000 iterations)
	-- kRing30: 41.906900 microseconds per iteration (10000 iterations)
	-- kRing40: 73.102700 microseconds per iteration (10000 iterations)
	-- kRingPentagon10: 171.604000 microseconds per iteration (500 iterations)
	-- kRingPentagon20: 1515.862000 microseconds per iteration (500 iterations)
	-- kRingPentagon30: 4967.980000 microseconds per iteration (50 iterations)
	-- kRingPentagon40: 11967.700000 microseconds per iteration (10 iterations)
$ ./bin/benchmarkPolyfill
	-- polyfillSF: 1169.436000 microseconds per iteration (500 iterations)
	-- polyfillAlameda: 1551.546000 microseconds per iteration (500 iterations)
	-- polyfillSouthernExpansion: 47067.600000 microseconds per iteration (10 iterations)
$ ./bin/benchmarkPolygon
	-- pointInsideGeofenceSmall: 0.034600 microseconds per iteration (100000 iterations)
	-- pointInsideGeofenceLarge: 0.342790 microseconds per iteration (100000 iterations)
	-- bboxFromGeofenceSmall: 0.028570 microseconds per iteration (100000 iterations)
	-- bboxFromGeofenceLarge: 0.353520 microseconds per iteration (100000 iterations)

Builds on #326. cc @Komzpa

[isaacbrodsky]

Looks like I'm getting the same failure in CI. From some quick searches, seems like it might be related to clang-tidy somehow?

[coveralls]

Coverage remained the same at 99.133% when pulling 5a677e4 on isaacbrodsky:inline-optimization into cd34251 on uber:master.

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you all sign our Contributor License Agreement before we can accept your contribution.
2 out of 3 committers have signed the CLA.

✅ Komzpa
✅ isaacbrodsky
❌ Isaac Brodsky

---

Isaac Brodsky seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}