Bug 31522

Summary: [Chromium] Allow building without SVG
Product: WebKit Reporter: Evan Martin <evan>
Component: SVGAssignee: Nobody <webkit-unassigned>
Status: RESOLVED FIXED    
Severity: Normal CC: jorlow, laszlo.gombos, mark
Priority: P2    
Version: 528+ (Nightly build)   
Hardware: PC   
OS: All   
Attachments:
Description Flags
wip patch
none
patch
none
patch jorlow: review+

Evan Martin
Reported 2009-11-15 17:09:25 PST
Created attachment 43257 [details] wip patch I don't use SVG in my day-to-day debugging and it's a bunch of extra code for the compiler, linker, etc. to move around. I measured it at around 14.5% of the resulting binary size of a Debug build. We should allow building without SVG.
Attachments
wip patch (3.72 KB, patch)
2009-11-15 17:09 PST, Evan Martin 		no flags
DetailsFormatted DiffDiff
patch (7.10 KB, patch)
2010-02-23 06:56 PST, Evan Martin 		no flags
DetailsFormatted DiffDiff
patch (7.72 KB, patch)
2010-02-23 07:04 PST, Evan Martin 		jorlow: review+
DetailsFormatted DiffDiff
Show Obsolete (2) View All Add attachment proposed patch, testcase, etc.
Laszlo Gombos
Comment 1 2010-02-22 20:50:21 PST
Evan, did you intended to put this patch up for review ? It seems to me that maybe you accidentally marked it for r-.
Evan Martin
Comment 2 2010-02-23 01:09:30 PST
No, the patch wasn't complete so I marked it so that it explicitly shouldn't be reviewed.
Evan Martin
Comment 3 2010-02-23 06:56:37 PST
Created attachment 49288 [details] patch
Evan Martin
Comment 4 2010-02-23 07:02:55 PST
Needs a changelog, arg.
Evan Martin
Comment 5 2010-02-23 07:04:43 PST
Created attachment 49292 [details] patch
Evan Martin
Comment 6 2010-02-23 07:06:06 PST
With this patch, my Chrome debug binary is 71mb smaller. Most of the cut fat is in DerivedSources which is the slowest file to build, too.
Evan Martin
Comment 7 2010-02-23 07:11:18 PST
+mark for gyp comments
Evan Martin
Comment 8 2010-02-23 07:11:57 PST
Chrome-side bits (not needed to land this patch, just for reference). http://codereview.chromium.org/655001
Jeremy Orlow
Comment 9 2010-02-23 07:12:44 PST
Comment on attachment 49292 [details] patch r=me This is probably worth a chromium-dev post. :-)
Evan Martin
Comment 10 2010-02-23 07:51:22 PST
r55152
Mark Mentovai
Comment 11 2010-02-23 08:11:47 PST
Comment on attachment 49292 [details] patch [ 'sources': [,] [...] >- # idl rule >+ # idl rules except for svg (added below) > '<@(webcore_bindings_idl_files)', > ], >+ 'conditions': [ >+ ['enable_svg!=0', { >+ 'sources': [ >+ '<@(webcore_bindings_idl_files)', >+ ], >+ }], >+ ], This says "add webcore_bindings_idl_files to sources, and if enable_svg, add them again." I doubt that's what you meant. GYP singleton processing will strip the extras so the above won't be harmful, but you should fix it. This section should be OK the way it was prior to this patch.
Evan Martin
Comment 12 2010-02-23 08:16:32 PST
Wow, how does this even compile!? It is supposed to reference a different variable (the svg-specific one).
Note You need to log in before you can comment on or make changes to this bug.