[webkit-reviews] review requested: [Bug 171537] WebAssembly: running out of executable memory should throw OoM : [Attachment 313772] patch

[bugzilla-daemon at webkit.org]

JF Bastien <jfbastien at apple.com> has asked  for review:
Bug 171537: WebAssembly: running out of executable memory should throw OoM
https://bugs.webkit.org/show_bug.cgi?id=171537

Attachment 313772: patch

https://bugs.webkit.org/attachment.cgi?id=313772&action=review




--- Comment #3 from JF Bastien <jfbastien at apple.com> ---
Created attachment 313772

  --> https://bugs.webkit.org/attachment.cgi?id=313772&action=review

patch

I'm wondering if this is worth testing, and how. I'm open to suggestions!

I've been running a huge program with: JSC_webAssemblyBBQOptimizationLevel=1
JSC_useConcurrentGC=0 JSC_useConcurrentJIT=0 JSC_jitMemoryReservationSize=5000
to hit its import limit, and then higher for BBQ functions. Maybe I should
force llint by disabling DFG and FTL? Also, I'd like to make sure we don't leak
code that's associated with the Module and which was compiled before the
failure! Not all the failure points are tested, because that's kind of dynamic
(wasm->wasm entry points are tricky to hit). Tiering is also tricky to hit OoM
with.