[jkotas]

Fixes #21032 and #20958

[jkotas]

Now that the Span indexers are recognized properly again, we are hitting assert during crossgen that sneaked in in the meantime:

Assert failure(PID 2976 [0x00000ba0], Thread: 3816 [0x0ee8]): Assertion failed '((regMask & emitThisGCrefRegs) && (ins == INS_add)) || ((regMask & emitThisByrefRegs) && (ins == INS_add || ins == INS_sub))' in 'System.Globalization.CharUnicodeInfo:InternalGetNumericValue(int):double' (IL size 175)

[stephentoub]

Thanks, @jkotas. Do we have any sense for what kind of throughput impact https://github.com/dotnet/coreclr/issues/21032 has on various workloads? I'm wondering if it's worth trying to get this change in for Preview 1, since I believe (but could be wrong) that coreclr was already snapshotted for the release.

[AndyAyersMS]

Wish I had seen that CC on the original PR.

LGTM. Thanks.

[AndyAyersMS]

On very simple loops indexing via span it was around 20%...

[CarolEidt]

Thanks for handling this. I was just starting to dig into this, and it would have taken me quite some time to come up to speed on the signature building.