Bug 213419 – [JSC] Attempt to reduce timeout failures on Apple Watch Series 3

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Yusuke Suzuki]

[JSC] Attempt to reduce timeout failures on Apple Watch Series 3

[Yusuke Suzuki]

Created attachment 402350 [details]
Patch

[Mark Lam]

Comment on attachment 402350 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=402350&action=review

r=me

> JSTests/stress/memcpy-typed-loops.js:31
> -for (let i=0; i<10000000; ++i) noLoop(arr1, arr2)
> -for (let i=0; i<10000000; ++i) invalidStart(arr1, arr2)
> +for (let i=0; i<100000; ++i) noLoop(arr1, arr2)
> +for (let i=0; i<100000; ++i) invalidStart(arr1, arr2)

Instead of this, can we just do a $vm.useJIT() test and set the iteration count accordingly?  Since watch doesn't use JIT, we can go with an even smaller count than this.  Looks like there isn't a $vm.useJIT() but you can add one.

> JSTests/stress/set-iteration-oas.js:20
> +for (var i = 0; i < 1e4; ++i)

Ditto.

[Yusuke Suzuki]

Comment on attachment 402350 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=402350&action=review

>> JSTests/stress/memcpy-typed-loops.js:31
>> +for (let i=0; i<100000; ++i) invalidStart(arr1, arr2)
> 
> Instead of this, can we just do a $vm.useJIT() test and set the iteration count accordingly?  Since watch doesn't use JIT, we can go with an even smaller count than this.  Looks like there isn't a $vm.useJIT() but you can add one.

Fixed.

>> JSTests/stress/set-iteration-oas.js:20
>> +for (var i = 0; i < 1e4; ++i)
> 
> Ditto.

Nice. Fixed.

[Yusuke Suzuki]

Created attachment 402358 [details]
Patch for landing

[EWS]

Committed r263312: <https://trac.webkit.org/changeset/263312>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 402358 [details].

[Radar WebKit Bug Importer]

<rdar://problem/64553456>

[Radar WebKit Bug Importer]

<rdar://problem/64553458>

[Mark Lam]

<rdar://problem/64553456>