[Webkit-unassigned] [Bug 14743] New: FE/win: FE chars draws as white box (Tofu) after viewing some pages

[bugzilla-daemon at webkit.org]

http://bugs.webkit.org/show_bug.cgi?id=14743

           Summary: FE/win: FE chars draws as white box (Tofu) after viewing
                    some pages
           Product: WebKit
           Version: 522+ (nightly)
          Platform: PC
        OS/Version: Windows XP
            Status: UNCONFIRMED
          Severity: Normal
          Priority: P4
         Component: Platform
        AssignedTo: webkit-unassigned at lists.webkit.org
        ReportedBy: 808caaa4.8ce9.9cd6c799e9f6 at gmail.com


related: bug#14601

After using Safari/WebKit for a while, white boxes appeared at the position of
FE chars. This time, is not encoding problem, since some part of FE strings
are drawed correctly.

At that time, langFontLink->MapFont() returns E_INVALID, and 
FE font mapping does not succeeded. It's not for invalid argument.
Even arguments intact, this error may occure.

Today some pages depends on MapFont(), and cousumes mlang's internal cache.
With IDE disassembly pane, MapFont() seems have very small font cache entries,
probably only 16, << if we call MapFont with 2nd argument nonzero. >>
If all of entries occupied once, they seems simply return with error
for every new custom fonts creation after that.
// Applies for mlang.dll, v6.0.2900.2180, deployed with XPSP2.
// By MS DLLDB online, WindowsServer2003 SP's mlang is newer, but not tested.
// Cache entry have internal refcount for each, and
// with ReleaseFont(), cache entry is reused (as documented in MSDN.)

See attached example again, some middle of body is rendered.
It's explicit-specified to 'Lucida Grande' part, 
it's cached previously at the time of drawing UI.

I don't know what to do about this concisely,
or whether we should use MapFont() with specify 0 in argument#2,
or whether we should cache by the MapFont's param rather than MapFont's result,
once suspected at bug#14601.


-- 
Configure bugmail: http://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.