roachtest: fix tpchvec perf

[yuzefovich]

This commit fixes a problem with "test case struct" resolution that was
introduced recently when we added smithcmp config to tpchvec test.
The root of the issue is that testRun function was made a part of the
tpchVecTestCase interface which doesn't really makes sense.

Release note: None

[cockroach-teamcity]

This change is 

[yuzefovich]

See #47401 (comment)

[asubiotto]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @asubiotto and @yuzefovich)

---

pkg/cmd/roachtest/tpchvec.go, line 140 at r1 (raw file):

	c *cluster,
	version crdbVersion,
	vectorizeOptions []bool,

I think this solution is confusing because we have the option to either use the passed in variable or call b.vectorizeOptions and the other testRun implementation ignores the arguments. The problem here is that there are no commonalities between the test types so putting this into an interface doesn't work cleanly. Instead, I think we should get rid of testRun and pass in a function to runTPCHVec one of which is the original and the other which is the new placeholder test.

[yuzefovich]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @asubiotto)

---

pkg/cmd/roachtest/tpchvec.go, line 140 at r1 (raw file):

Previously, asubiotto (Alfonso Subiotto Marqués) wrote…

I think this solution is confusing because we have the option to either use the passed in variable or call b.vectorizeOptions and the other testRun implementation ignores the arguments. The problem here is that there are no commonalities between the test types so putting this into an interface doesn't work cleanly. Instead, I think we should get rid of testRun and pass in a function to runTPCHVec one of which is the original and the other which is the new placeholder test.

Good point, done.

[asubiotto]

Reviewable status: complete! 1 of 0 LGTMs obtained

[yuzefovich]

TFTR!

bors r+

[yuzefovich]

bors r+

[craig]

Build succeeded

• GitHub CI (Cockroach)