Bug 186013 – [Baseline] Merge JITPropertyAccess, JITArithmetic, JITOpcodes, and JITCall to JIT

• Status: NEW
• Component: New Bugs

[Yusuke Suzuki]

[Baseline] Merge JITPropertyAccess, JITArithmetic, JITOpcodes, and JITCall to JIT

[Yusuke Suzuki]

Created attachment 341398 [details]
Patch

[Yusuke Suzuki]

Created attachment 341507 [details]
Patch

[Yusuke Suzuki]

Rebaselined. Ready for review.

[Saam Barati]

Comment on attachment 341507 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=341507&action=review

> Source/JavaScriptCore/ChangeLog:8
> +        It is hard to check whether the given functionality is implemented in 32bit / 64bit,

Do you really believe that this is hard to do?

Does this patch do anything besides copy old implementation into this new file?

[Yusuke Suzuki]

(In reply to Saam Barati from comment #4)
> Comment on attachment 341507 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=341507&action=review
> 
> > Source/JavaScriptCore/ChangeLog:8
> > +        It is hard to check whether the given functionality is implemented in 32bit / 64bit,
> 
> Do you really believe that this is hard to do?
> 

Yes, for example, we have bunch of #if USE(JSVALUE64) / USE(JSVALUE32_64) in each file. Small ifdefs are consolidated into JIT.cpp. And large ifdef gaurds are removed and code is moved into JIT32_64.cpp / JIT64.cpp. Small ifdefs in functions are typically doing the same things (like, defining registers), they can be later removed easily in a batch styled manner.
And this code structure is completely the same to DFG, it is easy to find code.

> Does this patch do anything besides copy old implementation into this new
> file?

Yes, this patch just copies the implementations into these files.

[Alex Christensen]

Comment on attachment 341507 [details]
Patch

This has been requesting review for more than one year.  If this is still needed, please rebase and re-request review.