should grow_memory ever trap?

[lukewagner]

AstSemantics.md does not describe grow_memory as trapping. However, GrowMemory traps in two cases and there is a test for this. I think the trap and test predate grow_memory returning -1 on error and, now that it does, it'd be better to always return -1 instead of trapping. That way the caller doesn't have to check for two separate error conditions.

[titzer]

On Mon, Sep 5, 2016 at 11:17 PM, Luke Wagner notifications@github.com
wrote:

AstSemantics.md does not describe grow_memory
https://github.com/WebAssembly/design/blob/master/AstSemantics.md#resizing
as trapping. However, GrowMemory traps in two cases
https://github.com/WebAssembly/spec/blob/master/ml-proto/spec/eval.ml#L288
and there is a test
https://github.com/WebAssembly/spec/blob/master/ml-proto/test/memory_trap.wast#L36
for this. I think the trap and test predate grow_memory returning -1 on
error and, now that it does, it'd be better to always return -1 instead of
trapping. That way the caller doesn't have to check for two separate error
conditions.

Agreed.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#334, or mute the thread
https://github.com/notifications/unsubscribe-auth/ALnq1IPSsmWB7yMaa9AwpBZjq4r5z7jqks5qnIcFgaJpZM4J1Uhr
.

[rossberg]

This is a duplicate of #307. I fixed it a while ago in #308, but only on the 0xc branch, because the patch did not apply cleanly on master. I'll give it another shot to merge it back.

[rossberg]

OK, done.

[lukewagner]

Oops, thanks.