OOM errors

[jfbastien]

Forking from #901 (comment)

Should we call out out-of-memory errors separately, or do they cause Compile / Link errors? This seems similar to @gahaas' clarification w.r.t. stack error in #875.

[flagxor]

We should do:
GrowMemory stays as is.
Everything else is up to the embedder, see Andreas' other verbiage.

[jfbastien]

As a clarification: I think OOM on WebAssembly.validate should be an OOM exception, not a validation failure: we may never be able to compile the module, but it's not fundamentally invalid. Pretty much the only valid usage of validate is to do feature detection, it would be odd to "lie" on OOM.

Again, this is a corner-case.

[titzer]

What JF says sounds reasonable to me.

…

On Sun, Dec 18, 2016 at 10:32 PM, JF Bastien ***@***.***> wrote: As a clarification: I think OOM on WebAssembly.validate should be an OOM exception, not a validation failure: we may never be able to compile the module, but it's not fundamentally invalid. Pretty much the only valid usage of validate is to do feature detection, it would be odd to "lie" on OOM. Again, this is a corner-case. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#903 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ALnq1JMYNpA0HPXtuIyziZpvKn2ENoJgks5rJaZxgaJpZM4LL-9P> .

[lukewagner]

Agreed; that also matches what we've done.

[sunfishcode]

This was fixed in #1059.