Stream: git-wasmtime

Topic: wasmtime / issue #10486 What is the reason for the `V128`...

Wasmtime GitHub notifications bot (Mar 29 2025 at 16:10):

Robbepop opened issue #10486:

When Wasmtime's CLI returns a v128 value, it is formatted as u128 value.
In Wasmi I have had a lot of thoughts about the formatting of v128 values and came to the conclusion that it is probably best to format it as hex-dec number with leading zeros. For example: 0x616E5F6E61635F6E61635F6E61635F6E

Was there a specific reason Wasmtime chose to format v128 as a simple u128?

Ideally Wasmi has the same formatting as Wasmtime (or more generally: Wasm runtimes share the same formatting), but at this point I find the u128 formatting to be very weird and thus I suppose I am missing something here.

Wasmtime GitHub notifications bot (Mar 29 2025 at 19:59):

abrown commented on issue #10486:

Yeah, I also prefer the hex format; that's what I think of when I'm thinking about lanes. Not sure why we used the u128 formatting. Convenience perhaps?

Wasmtime GitHub notifications bot (Mar 29 2025 at 20:26):

Robbepop edited issue #10486:

When Wasmtime's CLI returns a v128 value, it is formatted as u128 value.
In Wasmi I have made some thoughts about the formatting of v128 values and came to the conclusion that it is probably best to format it as hex-dec number with leading zeros. For example: 0x616E5F6E61635F6E61635F6E61635F6E

Was there a specific reason Wasmtime chose to format v128 as a simple u128?

Ideally Wasmi has the same formatting as Wasmtime (or more generally: Wasm runtimes share the same formatting), but at this point I find the u128 formatting to be very weird and thus I suppose I am missing something here.

Wasmtime GitHub notifications bot (Mar 29 2025 at 20:26):

Robbepop edited issue #10486:

When Wasmtime's CLI returns a v128 value, it is formatted as u128 value.
In Wasmi I have made some thoughts about the formatting of v128 values and came to the conclusion that it is probably best to format it as hex-dec number with leading zeros. For example: 0x006E006E61635F6E61635F6E61635F6E

Was there a specific reason Wasmtime chose to format v128 as a simple u128?

Ideally Wasmi has the same formatting as Wasmtime (or more generally: Wasm runtimes share the same formatting), but at this point I find the u128 formatting to be very weird and thus I suppose I am missing something here.

Wasmtime GitHub notifications bot (Mar 31 2025 at 14:34):

alexcrichton commented on issue #10486:

Yeah I think the reason for the current formatting is that prior to v128 everything printed as a number, so "why not also print as a number". I think it'd be reasonable to change as we don't guarantee the formatting at this time.

Last updated: Apr 18 2025 at 16:03 UTC