Some refactorings to improve wildcard performance on debug CTS builds
authorEric Werness <ewerness@nvidia.com>
Thu, 10 Nov 2022 22:23:15 +0000 (14:23 -0800)
committerPiotr Byszewski <piotr.byszewski@mobica.com>
Mon, 19 Dec 2022 10:31:33 +0000 (10:31 +0000)
commit7d2e0aaa3fb70f383f8bd8801de5de6e9a181aa1
tree7e4d3ef7875e300015e904a88b57ef206bcec823
parent1ceeb714e7ac23ced36d7f71cd48e46125b020ec
Some refactorings to improve wildcard performance on debug CTS builds

Measuring a VS2019 debug build running "--deqp-case=dEQP-VK.*faketestname" these changes
take the runtime from 800 seconds to 240 seconds

Tests run locally:

 --deqp-case=dEQP-VK.*faketestname and confirmed zero tests run and improved performance with debug build

Following commands run and confirmed same count of tests before and after the change:

Wildcard without dot

 --deqp-case=dEQP-VK.ray_tracing_pipeline*rgen*

Wildcard with dot

 --deqp-case=dEQP-VK.info.*

Wildcard at start and end

 --deqp-case=*peer_memory*

Components: Framework

VK-GL-CTS issue: 4149

Change-Id: I91348b3e0824014897393b6ee07864a4b786186f
framework/common/tcuCommandLine.cpp