| Summary: | Add some comments to describe the DFG UseKind representations. | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | WebKit | Reporter: | Mark Lam <mark.lam> | ||||||||||||
| Component: | JavaScriptCore | Assignee: | Mark Lam <mark.lam> | ||||||||||||
| Status: | RESOLVED FIXED | ||||||||||||||
| Severity: | Normal | CC: | fpizlo, ggaren, mmirman, msaboff | ||||||||||||
| Priority: | P2 | ||||||||||||||
| Version: | 528+ (Nightly build) | ||||||||||||||
| Hardware: | Unspecified | ||||||||||||||
| OS: | Unspecified | ||||||||||||||
| Attachments: |
|
||||||||||||||
|
Description
Mark Lam
2014-11-20 14:09:54 PST
Created attachment 241980 [details]
the patch.
Comment on attachment 241980 [details] the patch. View in context: https://bugs.webkit.org/attachment.cgi?id=241980&action=review > Source/JavaScriptCore/dfg/DFGUseKind.h:38 > UntypedUse, Which case is UntypedUse? (Seems like it should be JSValue, right?) > Source/JavaScriptCore/dfg/DFGUseKind.h:43 > + // in a FP register. "an" Created attachment 241989 [details]
patch 2: addressed Geoff's feedback.
Comment on attachment 241989 [details] patch 2: addressed Geoff's feedback. View in context: https://bugs.webkit.org/attachment.cgi?id=241989&action=review > Source/JavaScriptCore/dfg/DFGUseKind.h:54 > UntypedUse, This could be dangerous; I believe that UntypedUse is expected to be zero. Comment on attachment 241989 [details]
patch 2: addressed Geoff's feedback.
Please arrange to have UntypedUse be the first thing, as before.
Created attachment 241992 [details]
patch 3: UntypedUse is at enum 0 again, with extra comments.
Created attachment 241995 [details]
patch 4: don't like the comment about UntypedUse in last patch. Updated.
Comment on attachment 241995 [details]
patch 4: don't like the comment about UntypedUse in last patch. Updated.
You could also have moved all of the other JS representations up with UntypedUse. That would make the most logical sense to me.
Created attachment 241996 [details]
patch 5: moved al the JSValue representation UseKinds up nelow UntypedUse as Filip suggested.
Thanks for the review. Landed in r176425: <http://trac.webkit.org/r176425>. |