Bug 200632 - [WHLSL] TypeNamer can be simplified by replacing BaseTypeNameNode with uniqued AST::UnnamedTypes
Summary: [WHLSL] TypeNamer can be simplified by replacing BaseTypeNameNode with unique...
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Sam Weinig
URL:
Keywords: InRadar
Depends on:
Blocks:
 
Reported: 2019-08-12 10:42 PDT by Sam Weinig
Modified: 2019-08-26 11:59 PDT (History)
6 users (show)

See Also:

Attachments
Patch (27.40 KB, patch)
2019-08-12 11:00 PDT, Sam Weinig 		no flags Details | Formatted Diff | Diff
Patch (27.40 KB, patch)
2019-08-13 09:37 PDT, Sam Weinig 		no flags Details | Formatted Diff | Diff
Patch (25.82 KB, patch)
2019-08-17 13:58 PDT, Sam Weinig 		no flags Details | Formatted Diff | Diff
Patch (25.73 KB, patch)
2019-08-24 19:38 PDT, Sam Weinig 		no flags Details | Formatted Diff | Diff
Show Obsolete (3) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Sam Weinig 2019-08-12 10:42:07 PDT 
[WHLSL] TypeNamer can be simplified by replacing BaseTypeNameNode with uniqued AST::UnnamedTypes
Comment 1 Sam Weinig 2019-08-12 11:00:52 PDT 
Created attachment 376080 [details]
Patch
Comment 2 Sam Weinig 2019-08-13 09:37:13 PDT 
Created attachment 376175 [details]
Patch
Comment 3 Sam Weinig 2019-08-16 20:11:45 PDT 
Saam, any thoughts on this?
Comment 4 Myles C. Maxfield 2019-08-16 21:04:16 PDT 
Comment on attachment 376175 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=376175&action=review

> Source/WebCore/ChangeLog:11
> +        There is no longer a reason to keep a parallel tree of the UnnamedType-like objects
> +        BaseTypeNameNodes. Instead, we can store a single HashMap mapping from UnnamedTypeKeys
> +        to MangledTypeName, and use the the UnnamedType stored in the UnnamedTypeKey while 
> +        emitting the metal code.        

What's the motivation for this?
Comment 5 Sam Weinig 2019-08-16 22:30:45 PDT 
(In reply to Myles C. Maxfield from comment #4)
> Comment on attachment 376175 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=376175&action=review
> 
> > Source/WebCore/ChangeLog:11
> > +        There is no longer a reason to keep a parallel tree of the UnnamedType-like objects
> > +        BaseTypeNameNodes. Instead, we can store a single HashMap mapping from UnnamedTypeKeys
> > +        to MangledTypeName, and use the the UnnamedType stored in the UnnamedTypeKey while 
> > +        emitting the metal code.        
> 
> What's the motivation for this?

Simplifying the code. Removing unnecessary indirection.
Comment 6 Sam Weinig 2019-08-17 13:58:34 PDT 
Created attachment 376613 [details]
Patch
Comment 7 Sam Weinig 2019-08-24 19:38:14 PDT 
Created attachment 377219 [details]
Patch
Comment 8 Saam Barati 2019-08-26 11:26:12 PDT 
Comment on attachment 377219 [details]
Patch

Nice. r=me
Comment 9 WebKit Commit Bot 2019-08-26 11:58:07 PDT 
Comment on attachment 377219 [details]
Patch

Clearing flags on attachment: 377219

Committed r249109: <https://trac.webkit.org/changeset/249109>
Comment 10 WebKit Commit Bot 2019-08-26 11:58:09 PDT 
All reviewed patches have been landed.  Closing bug.
Comment 11 Radar WebKit Bug Importer 2019-08-26 11:59:17 PDT 
<rdar://problem/54716312>