The whole string argument here is being included in the test name, and we are truncating test names at 200 characters in our report URLs, so each case is going to have the same report URL. The full test name is still included in our raw data, it just means that we can't distinguish them in our generated reports and graphs as we won't know which test case the report is displaying the data for. Either we need to change how we do our report URL truncation (maybe for >200 characters we put a small hash at the end), or modify this benchmark to not take the full string as a direct argument, but rather a string that describes the type of string being processed which it can lookup
In the Perf_BigInteger benchmark, we have some arguments with long strings, with one being 20,000 characters long: https://github.com/dotnet/performance/blob/11fe47276ef6947a84a4f8a5dde8e2f3366c15c5/src/benchmarks/micro/libraries/System.Runtime.Numerics/Perf.BigInteger.cs#L19-L21
The whole string argument here is being included in the test name, and we are truncating test names at 200 characters in our report URLs, so each case is going to have the same report URL. The full test name is still included in our raw data, it just means that we can't distinguish them in our generated reports and graphs as we won't know which test case the report is displaying the data for. Either we need to change how we do our report URL truncation (maybe for >200 characters we put a small hash at the end), or modify this benchmark to not take the full string as a direct argument, but rather a string that describes the type of string being processed which it can lookup