86368 - nsFontMetricsGTK::PickASizeAndLoad Symbol font disable and MathML-enabled build crashes"

Status: RESOLVED FIXED

(Core :: MathML, defect)

Description

[Roland Mainz]

2001-06-16-08-trunk build with "configure --with-xprint --enable-mathml
--enable-svg --enable-xsl --with-extensions --disable-ldap --enable-optimize
--enable-nspr-autoconf" using Sun Workshop 6 Update 2 EarlyAccess2 crashes on my
Solaris 7 UltraSPARC machine when I try to view a MathML page.

Same sources build with Xlib-toolkit (--enable-toolkit)xlib) work fine and
printing rocks (using Xprint, --with-xprint)...

dbx backtrace from coredump:
-- snip --
detected a multithreaded program
t@1 (l@1) terminated by signal BUS (invalid address alignment)
Current function is nsFontMetricsGTK::PickASizeAndLoad
 2365       if (FONT_HAS_GLYPH(font->mMap, aChar)) {
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) where
current thread: t@1
=>[1] nsFontMetricsGTK::PickASizeAndLoad(this = 0x724060, aStretch = 0x9d9d10,
aCharSet = 0xfd3f7c10, aChar = 770U, aName = 0x9da420
"adobe-symbol-adobe-fontspecific"), line 2365 in "nsFontMetricsGTK.cpp"
  [2] nsFontMetricsGTK::SearchNode(this = 0x724060, aNode = 0x9da408, aChar =
770U), line 2716 in "nsFontMetricsGTK.cpp"
  [3] nsFontMetricsGTK::FindAnyFont(this = 0x724060, aChar = 770U), line 3480 in
"nsFontMetricsGTK.cpp"
  [4] nsFontMetricsGTK::FindFont(this = 0x724060, aChar = 770U), line 3682 in
"nsFontMetricsGTK.cpp"
  [5] nsRenderingContextGTK::GetWidth(this = 0xa79db8, aString = 0xffbea284,
aLength = 1U, aWidth = -14290324, aFontID = (nil)), line 1316 in
"nsRenderingContextGTK.cpp"
  [6] nsTextFrame::MeasureText(this = ???, aPresContext = ???, aReflowState =
STRUCT, aTx = CLASS, aLb = ???, aTs = STRUCT, aTextData = STRUCT) (optimized),
at 0xfc1b5f64 (line ~4566) in "nsTextFrame.cpp"
  [7] nsTextFrame::Reflow(this = ???, aPresContext = ???, aMetrics = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc1b69bc (line ~5048) in
"nsTextFrame.cpp"
  [8] nsContainerFrame::ReflowChild(this = ???, aKidFrame = ???, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aX = ???, aY = ???, aFlags =
???, aStatus = ???) (optimized), at 0xfc12c538 (line ~723) in
"nsContainerFrame.cpp"
  [9] nsMathMLContainerFrame::ReflowChild(this = ???, aKidFrame = ???,
aPresContext = ???, aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???)
(optimized), at 0xfc37e874 (line ~190) in "nsMathMLContainerFrame.h"
  [10] nsMathMLContainerFrame::ReflowTokenFor(aFrame = ???, aPresContext = ???,
aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at
0xfc37d33c (line ~1122) in "nsMathMLContainerFrame.cpp"
  [11] nsMathMLmoFrame::Reflow(this = ???, aPresContext = ???, aDesiredSize =
STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc393d30 (line
~805) in "nsMathMLmoFrame.cpp"
  [12] nsContainerFrame::ReflowChild(this = ???, aKidFrame = ???, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aX = ???, aY = ???, aFlags =
???, aStatus = ???) (optimized), at 0xfc12c538 (line ~723) in
"nsContainerFrame.cpp"
  [13] nsMathMLContainerFrame::ReflowChild(this = ???, aKidFrame = ???,
aPresContext = ???, aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???)
(optimized), at 0xfc37e874 (line ~190) in "nsMathMLContainerFrame.h"
  [14] nsMathMLContainerFrame::Reflow(this = ???, aPresContext = ???,
aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at
0xfc37d754 (line ~1278) in "nsMathMLContainerFrame.cpp"
  [15] nsContainerFrame::ReflowChild(this = ???, aKidFrame = ???, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aX = ???, aY = ???, aFlags =
???, aStatus = ???) (optimized), at 0xfc12c538 (line ~723) in
"nsContainerFrame.cpp"
  [16] nsMathMLContainerFrame::ReflowChild(this = ???, aKidFrame = ???,
aPresContext = ???, aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???)
(optimized), at 0xfc37e874 (line ~190) in "nsMathMLContainerFrame.h"
  [17] nsMathMLContainerFrame::Reflow(this = ???, aPresContext = ???,
aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at
0xfc37d754 (line ~1278) in "nsMathMLContainerFrame.cpp"
  [18] nsLineLayout::ReflowFrame(this = ???, aFrame = ???, aNextRCFrame = ???,
aReflowStatus = ???, aMetrics = ???, aPushedFrame = ???) (optimized), at
0xfc16ed38 (line ~955) in "nsLineLayout.cpp"
  [19] nsInlineFrame::ReflowInlineFrame(this = ???, aPresContext = ???,
aReflowState = STRUCT, irs = STRUCT, aFrame = ???, aStatus = ???) (optimized),
at 0xfc1682b4 (line ~612) in "nsInlineFrame.cpp"
  [20] nsInlineFrame::ReflowFrames(this = ???, aPresContext = ???, aReflowState
= STRUCT, irs = STRUCT, aMetrics = STRUCT, aStatus = ???) (optimized), at
0xfc167dcc (line ~427) in "nsInlineFrame.cpp"
  [21] nsInlineFrame::Reflow(this = ???, aPresContext = ???, aMetrics = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc167b8c (line ~338) in
"nsInlineFrame.cpp"
  [22] nsLineLayout::ReflowFrame(this = ???, aFrame = ???, aNextRCFrame = ???,
aReflowStatus = ???, aMetrics = ???, aPushedFrame = ???) (optimized), at
0xfc16ed38 (line ~955) in "nsLineLayout.cpp"
  [23] nsBlockFrame::ReflowInlineFrame(this = ???, aState = CLASS, aLineLayout =
CLASS, aLine = ???, aFrame = ???, aLineReflowStatus = ???) (optimized), at
0xfc11b634 (line ~3611) in "nsBlockFrame.cpp"
  [24] nsBlockFrame::DoReflowInlineFrames(this = ???, aState = CLASS,
aLineLayout = CLASS, aLine = ???, aKeepReflowGoing = ???, aLineReflowStatus =
???, aUpdateMaximumWidth = ???, aDamageDirtyArea = ???) (optimized), at
0xfc11b384 (line ~3421) in "nsBlockFrame.cpp"
  [25] nsBlockFrame::DoReflowInlineFramesAuto(this = ???, aState = CLASS, aLine
= ???, aKeepReflowGoing = ???, aLineReflowStatus = ???, aUpdateMaximumWidth =
???, aDamageDirtyArea = ???) (optimized), at 0xfc11b178 (line ~3271) in
"nsBlockFrame.cpp"
  [26] nsBlockFrame::ReflowInlineFrames(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???, aDamageDirtyArea = ???, aUpdateMaximumWidth = ???)
(optimized), at 0xfc11afd8 (line ~3228) in "nsBlockFrame.cpp"
  [27] nsBlockFrame::ReflowLine(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???, aDamageDirtyArea = ???) (optimized), at 0xfc119970 (line
~2380) in "nsBlockFrame.cpp"
  [28] nsBlockFrame::ReflowDirtyLines(this = ???, aState = CLASS) (optimized),
at 0xfc119004 (line ~2174) in "nsBlockFrame.cpp"
  [29] nsBlockFrame::Reflow(this = ???, aPresContext = ???, aMetrics = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc116d70 (line ~1031) in
"nsBlockFrame.cpp"
  [30] nsBlockReflowContext::DoReflowBlock(this = ???, aReflowState = STRUCT,
aReason = ???, aFrame = ???, aSpace = STRUCT, aApplyTopMargin = ???,
aPrevBottomMargin = ???, aIsAdjacentWithTop = ???, aComputedOffsets = STRUCT,
aFrameReflowStatus = ???) (optimized), at 0xfc124414 (line ~571) in
"nsBlockReflowContext.cpp"
  [31] nsBlockReflowContext::ReflowBlock(this = ???, aFrame = ???, aSpace =
STRUCT, aApplyTopMargin = ???, aPrevBottomMargin = ???, aIsAdjacentWithTop =
???, aComputedOffsets = STRUCT, aFrameReflowStatus = ???) (optimized), at
0xfc123eb4 (line ~336) in "nsBlockReflowContext.cpp"
  [32] nsBlockFrame::ReflowBlockFrame(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???) (optimized), at 0xfc11a9ac (line ~3181) in
"nsBlockFrame.cpp"
  [33] nsBlockFrame::ReflowLine(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???, aDamageDirtyArea = ???) (optimized), at 0xfc1195c0 (line
~2380) in "nsBlockFrame.cpp"
  [34] nsBlockFrame::ReflowDirtyLines(this = ???, aState = CLASS) (optimized),
at 0xfc119004 (line ~2174) in "nsBlockFrame.cpp"
  [35] nsBlockFrame::Reflow(this = ???, aPresContext = ???, aMetrics = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc116d70 (line ~1031) in
"nsBlockFrame.cpp"
  [36] nsBlockReflowContext::DoReflowBlock(this = ???, aReflowState = STRUCT,
aReason = ???, aFrame = ???, aSpace = STRUCT, aApplyTopMargin = ???,
aPrevBottomMargin = ???, aIsAdjacentWithTop = ???, aComputedOffsets = STRUCT,
aFrameReflowStatus = ???) (optimized), at 0xfc124414 (line ~571) in
"nsBlockReflowContext.cpp"
  [37] nsBlockReflowContext::ReflowBlock(this = ???, aFrame = ???, aSpace =
STRUCT, aApplyTopMargin = ???, aPrevBottomMargin = ???, aIsAdjacentWithTop =
???, aComputedOffsets = STRUCT, aFrameReflowStatus = ???) (optimized), at
0xfc123eb4 (line ~336) in "nsBlockReflowContext.cpp"
  [38] nsBlockFrame::ReflowBlockFrame(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???) (optimized), at 0xfc11a9ac (line ~3181) in
"nsBlockFrame.cpp"
  [39] nsBlockFrame::ReflowLine(this = ???, aState = CLASS, aLine = ???,
aKeepReflowGoing = ???, aDamageDirtyArea = ???) (optimized), at 0xfc1195c0 (line
~2380) in "nsBlockFrame.cpp"
  [40] nsBlockFrame::ReflowDirtyLines(this = ???, aState = CLASS) (optimized),
at 0xfc119004 (line ~2174) in "nsBlockFrame.cpp"
  [41] nsBlockFrame::Reflow(this = ???, aPresContext = ???, aMetrics = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc116d70 (line ~1031) in
"nsBlockFrame.cpp"
  [42] nsContainerFrame::ReflowChild(this = ???, aKidFrame = ???, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aX = ???, aY = ???, aFlags =
???, aStatus = ???) (optimized), at 0xfc12c538 (line ~723) in
"nsContainerFrame.cpp"
  [43] CanvasFrame::Reflow(this = ???, aPresContext = ???, aDesiredSize =
STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc14e264 (line
~569) in "nsHTMLFrame.cpp"
  [44] nsBoxToBlockAdaptor::Reflow(this = ???, aState = CLASS, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aStatus = ???, aX = ???, aY =
???, aWidth = ???, aHeight = ???, aMoveFrame = ???) (optimized), at 0xfc2cc0a0
(line ~864) in "nsBoxToBlockAdaptor.cpp"
  [45] nsBoxToBlockAdaptor::DoLayout(this = ???, aState = CLASS) (optimized), at
0xfc2cb990 (line ~515) in "nsBoxToBlockAdaptor.cpp"
  [46] nsBox::Layout(this = ???, aState = CLASS) (optimized), at 0xfc2c6940
(line ~983) in "nsBox.cpp"
  [47] nsScrollBoxFrame::DoLayout(this = ???, aState = CLASS) (optimized), at
0xfc297a50 (line ~377) in "nsScrollBoxFrame.cpp"
  [48] nsBox::Layout(this = ???, aState = CLASS) (optimized), at 0xfc2c6940
(line ~983) in "nsBox.cpp"
  [49] nsContainerBox::LayoutChildAt(aState = CLASS, aBox = ???, aRect = STRUCT)
(optimized), at 0xfc2d0114 (line ~591) in "nsContainerBox.cpp"
  [50] nsGfxScrollFrameInner::LayoutBox(this = ???, aState = CLASS, aBox = ???,
aRect = STRUCT) (optimized), at 0xfc1c311c (line ~1038) in
"nsGfxScrollFrame.cpp"
  [51] nsGfxScrollFrameInner::Layout(this = ???, aState = CLASS) (optimized), at
0xfc1c338c (line ~1142) in "nsGfxScrollFrame.cpp"
  [52] nsGfxScrollFrame::DoLayout(this = ???, aState = CLASS) (optimized), at
0xfc1c3150 (line ~1046) in "nsGfxScrollFrame.cpp"
  [53] nsBox::Layout(this = ???, aState = CLASS) (optimized), at 0xfc2c6940
(line ~983) in "nsBox.cpp"
  [54] nsBoxFrame::Reflow(this = ???, aPresContext = ???, aDesiredSize = STRUCT,
aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc2e6614 (line ~778) in
"nsBoxFrame.cpp"
  [55] nsGfxScrollFrame::Reflow(this = ???, aPresContext = ???, aDesiredSize =
STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc1c24ec (line
~735) in "nsGfxScrollFrame.cpp"
  [56] nsContainerFrame::ReflowChild(this = ???, aKidFrame = ???, aPresContext =
???, aDesiredSize = STRUCT, aReflowState = STRUCT, aX = ???, aY = ???, aFlags =
???, aStatus = ???) (optimized), at 0xfc12c538 (line ~723) in
"nsContainerFrame.cpp"
  [57] ViewportFrame::Reflow(this = ???, aPresContext = ???, aDesiredSize =
STRUCT, aReflowState = STRUCT, aStatus = ???) (optimized), at 0xfc1c04a4 (line
~536) in "nsViewportFrame.cpp"
  [58] PresShell::InitialReflow(this = ???, aWidth = ???, aHeight = ???)
(optimized), at 0xfc18ba20 (line ~2664) in "nsPresShell.cpp"
  [59] nsXMLContentSink::StartLayout(this = ???) (optimized), at 0xfcb26348
(line ~1597) in "nsXMLContentSink.cpp"
  [60] nsXMLContentSink::DidBuildModel(this = ???, aQualityLevel = ???)
(optimized), at 0xfcb21a94 (line ~297) in "nsXMLContentSink.cpp"
  [61] CWellFormedDTD::DidBuildModel(this = ???, anErrorCode = ???, aNotifySink
= ???, aParser = ???, aSink = ???) (optimized), at 0xfd4ec3f0 (line ~296) in
"nsWellFormedDTD.cpp"
  [62] nsParser::DidBuildModel(this = ???, anErrorCode = ???) (optimized), at
0xfd4e0b24 (line ~1445) in "nsParser.cpp"
  [63] nsParser::ResumeParse(this = ???, allowIteration = ???, aIsFinalChunk =
???) (optimized), at 0xfd4e169c (line ~1958) in "nsParser.cpp"
  [64] nsParser::OnStopRequest(this = ???, request = ???, aContext = ???, status
= ???) (optimized), at 0xfd4e23b0 (line ~2383) in "nsParser.cpp"
  [65] nsDocumentOpenInfo::OnStopRequest(this = ???, request = ???, aCtxt = ???,
aStatus = ???) (optimized), at 0xfd56bd94 (line ~252) in "nsURILoader.cpp"
  [66] nsStreamListenerTee::OnStopRequest(this = ???, request = ???, context =
???, status = ???) (optimized), at 0xfd792460 (line ~24) in
"nsStreamListenerTee.cpp"
  [67] nsHttpChannel::OnStopRequest(this = ???, request = ???, ctxt = ???,
status = ???) (optimized), at 0xfd7d8458 (line ~2100) in "nsHttpChannel.cpp"
  [68] nsOnStopRequestEvent::HandleEvent(this = ???) (optimized), at 0xfd76aa88
(line ~161) in "nsRequestObserverProxy.cpp"
  [69] nsARequestObserverEvent::HandlePLEvent(plev = ???) (optimized), at
0xfd76921c (line ~63) in "nsRequestObserverProxy.cpp"
  [70] PL_HandleEvent(self = ???) (optimized), at 0xff243880 (line ~590) in
"plevent.c"
  [71] PL_ProcessPendingEvents(self = ???) (optimized), at 0xff2436c8 (line
~520) in "plevent.c"
  [72] nsEventQueueImpl::ProcessPendingEvents(this = ???) (optimized), at
0xff2459c0 (line ~374) in "nsEventQueue.cpp"
  [73] event_processor_callback(data = ???, source = ???, condition = ???)
(optimized), at 0xfda264f0 (line ~168) in "nsAppShell.cpp"
  [74] our_gdk_io_invoke(source = ???, condition = ???, data = ???) (optimized),
at 0xfda2607c (line ~60) in "nsAppShell.cpp"
dbx: warning: can't find file
"/home/gisburn/package-builds/glib/glib-1.2.8/objdir/giounix.lo"
dbx: warning: see `help finding-files'
  [75] g_io_unix_dispatch(0x23ad58, 0xffbef058, 0x23a150, 0x0, 0x0, 0xffbeefc0),
at 0xfea32dc8
dbx: warning: can't find file
"/home/gisburn/package-builds/glib/glib-1.2.8/objdir/gmain.lo"
  [76] g_main_dispatch(0xffbef058, 0x11e130, 0x1, 0x23a150, 0xff3df650, 0x378),
at 0xfea36dc8
  [77] g_main_iterate(0x1, 0x1, 0xfe531f28, 0xfda26b48, 0xff3df650, 0x19), at
0xfea37bcc
  [78] g_main_run(0x23a198, 0x23a198, 0xff245c44, 0x5, 0xf2d20, 0x0), at
0xfea37f64
dbx: warning: can't find file
"/home/gisburn/package-builds/gtk+/gtk+-1.2.8/objdir/gtk/gtkmain.lo"
  [79] gtk_main(0xf2d20, 0xff248198, 0x0, 0x80004000, 0x0, 0xfda9cc50), at
0xfed560a0
  [80] nsAppShell::Run(this = ???) (optimized), at 0xfda26b48 (line ~365) in
"nsAppShell.cpp"
  [81] nsAppShellService::Run(this = ???) (optimized), at 0xfe487298 (line ~417)
in "nsAppShellService.cpp"
  [82] main1(argc = ???, argv = ???, nativeApp = ???) (optimized), at 0x177cc
(line ~1149) in "nsAppRunner.cpp"
  [83] main(argc = ???, argv = ???) (optimized), at 0x18204 (line ~1449) in
"nsAppRunner.cpp"
-- snip --

console output:
-- snip --
Enabling Quirk StyleSheet
Note: verifyreflow is disabled
Note: styleverifytree is disabled
Note: frameverifytree is disabled
Disabling Quirk StyleSheet
Start reading in bookmarks.html
Finished reading in bookmarks.html  (33127 microseconds)
WEBSHELL+ = 3
Enabling Quirk StyleSheet
Enabling Quirk StyleSheet
Opening file cookperm.txt failed
blank
Document http://www.mozilla.org/ loaded successfully
WEBSHELL+ = 4
Disabling Quirk StyleSheet
Enabling Quirk StyleSheet
/
Document http://www.mozilla.org/projects/mathml/ loaded successfully
Loading resource:/res/fonts/mathfontSymbol.properties ... Done
Bus Error - core dumped
-- snip --

Comment 1

[rbs]

Strange that it works on one platform and not the other. Did you try to disable 
'--enable-optimize' just to see what happens?

Comment 2

[Roland Mainz]

rbs:
I removed the "-O" option from config/autoconf.mk and cleaned/rebuild the whole
mozilla/gfx/src/ dir.

I am pretty sure that this is not a compiler bug... Sun Workshop 6 Update 2 EA
is _very_ good. The h*ll knows why Sun calls this release "EarlyAccess"... :-))

BTW: gcc implements a "hack" to force the SPARC-mandatory address alignment for
data, Sun Workshop does not implemenent such a hack (because this hack eats a
_large_ amount of CPU time in some cases).

But I don't think that this is the cause of the problem. Xlib-toolkit simply
uses a completely different code (mozilla/gfx/src/xlib/) than GTK+-toolkit
(mozilla/gfx/src/gtk/)...

Comment 3

[kill this account]

what is the value of font->mMap at the crash ?

Comment 4

[Roland Mainz]

-- snip --
Opening file cookperm.txt failed
blank
Document http://www.mozilla.org/ loaded successfully
WEBSHELL+ = 4
Disabling Quirk StyleSheet
Loading resource:/res/fonts/mathfontSymbol.properties ... Done
t@1 (l@1) signal BUS (invalid address alignment) in
nsFontMetricsGTK::PickASizeAndLoad at line 2365 in file "nsFontMetricsGTK.cpp"
 2365       if (FONT_HAS_GLYPH(font->mMap, aChar)) {
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) print font->mMap 
font->mMap = 0xfd3e4b13
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) print font      
font = 0x97ba58
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) print aChar
aChar = 770U
(/opt/SUNWspro/bin/../WS6U2/bin/sparcv9/dbx) print this
this = 0xa43408
-- snip --

Comment 5

[rbs]

re-assigning to bstell as this may probably be caused be some recent changes
that were not reflected in the MathML code.

Comment 6

[kill this account]

I'll rebuild with --enable-mathml and take a look

Comment 7

[kill this account]

Attachment: patch; correct data cast from hash

Comment 8

[kill this account]

For the font scaling info I need the nsFontCharSetMap* instead of the
nsFontCharSetMap.mInfo.

I missed changin this spot when I checked in the patch for bug 74889.

Comment 9

[kill this account]

For the font scaling info I need the nsFontCharSetMap* instead of the
nsFontCharSetMap.mInfo*.

I missed changing this spot when I checked in the patch for bug 74889.

Comment 10

[Roland Mainz]

Works for me (except that I am missing MathML fonts, but that's another issue.
At least GTK+-toolkit now produces the same output as Xlib-toolkit... :-)

r=[e-mail]

----

pocemit:
Wanna pull-over patch in bug 74889 and this "add-on" to Xlib-toolkit, please ?

Comment 11

[kill this account]

this crash only happens if you have mathml fonts.

see: http://www.mozilla.org/projects/mathml/fonts/

(http://support.wolfram.com/Systems/Unix/FE/Fonts/Type1.html)
(http://www.ibiblio.org/Linux/HOWTO/Font-HOWTO-4.html#ss4.2)
(Redhat users can just use chkfontpath. the format is chkfontpath --add 
directory)

Comment 12

[Roland Mainz]

I am on Solaris, and yes... I have the wolfram fonts in my fontpath:
% xset q | fgrep "fonts/misc"
-- snip --
/usr/openwin/lib/X11/fonts/F3/,/usr/openwin/lib/X11/fonts/F3bitmaps/,/usr/openwin/lib/X11/fonts/Type1/,/usr/openwin/lib/X11/fonts/Speedo/,/usr/openwin/lib/X11/fonts/misc/,/usr/openwin/lib/X11/fonts/75dpi/,/usr/openwin/lib/X11/fonts/100dpi/,/usr/openwin/lib/X11/fonts/TrueType,/opt/SUNWdtpcv/fonts,/usr/openwin/lib/locale/th_TH/X11/fonts/TrueType/,/usr/openwin/lib/locale/sun_eu_greek/X11/fonts/TrueType/,/usr/openwin/lib/locale/iso_8859_9/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_8/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_8/X11/fonts/TrueType/,/usr/openwin/lib/locale/iso_8859_7/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_5/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_4/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_2/X11/fonts/Type1/,/usr/openwin/lib/locale/iso_8859_15/X11/fonts/TrueType/,/usr/openwin/lib/locale/iso_8859_13/X11/fonts/Type1/,/usr/openwin/lib/locale/ar/X11/fonts/TrueType/,/usr/openwin/lib/locale/KOI8-R/X11/fonts/TrueType/,/usr/j2se/jre/lib/fonts/,/home/gisburn/mathml_fonts/wolfram,/home/gisburn/mathml_fonts/tex/cmpsfont/pfb
-- snip --

But the output still looks weired. That's discussed in bug 87285 ("MathML
"torture test" looks weired in Xlib-toolkit and Xprint")...

Comment 13

[kill this account]

this will also crash for symbol fonts

see http://lxr.mozilla.org/seamonkey/source/gfx/src/gtk/nsFontMetricsGTK.cpp#516

518 static nsFontCharSetMap gSpecialCharSetMap[] =
519 {
520   { "symbol-adobe-fontspecific", &FLG_NONE, &AdobeSymbol  },
521 
522 #ifdef MOZ_MATHML
523   { "cmex10-adobe-fontspecific", &FLG_NONE, &CMCMEX  },
524   { "cmsy10-adobe-fontspecific", &FLG_NONE, &CMCMSY  },

Comment 14

[Roland Mainz]

Brian Stell:
Wanna post a screenshot (viewing
http://www.mozilla.org/projects/mathml/demo/texvsmml.xml - the middle part
(10/11/12/13/14/15/16) of the table is the interesting one...) of your build,
please ? I'd like to compare your and my box...

Comment 15

[Frank Tang]

It looks like symbol font is busted by the same bug. See my new attachement. 
Viewing without the patch will show a question mark. view with the patch will 
show a hebrew alef.  There are other symbol glyph missing without this fix. Some 
of the symble glyph could be displayed by CJK font thought. (for example 
α)

Comment 16

[Frank Tang]

Attachment: test case to hit symbol font code.

Comment 17

[kill this account]

Attachment: mathml image

Comment 18

[kill this account]

Without the patch http://bugzilla.mozilla.org/showattachment.cgi?attach_id=40219
causes moz to crash.

Comment 19

[Roland Mainz]

Brian, thanks for the image.
Seems my problem was to get rid of all other crap extra fonts from my
/home/gisburn/mathml_fonts dir, pull the Wolfram fonts again and _only_ add
those one - no other fonts... :-)

Comment 20

[rbs]

*** Bug 87076 has been marked as a duplicate of this bug. ***

Comment 21

[Frank Tang]

I want to point out this bug is not ONLY affect mathml. The same bug also impact 
non MathML build while loading a Symbol. The behavior is random, could crash or 
make symbol font disable. use the 2nd attachement to verify that. 
Change the subject from "MathML-enabled build crashes in 
nsFontMetricsGTK::PickASizeAndLoad()" to 
"nsFontMetricsGTK::PickASizeAndLoad cause Symbol font disable and MathML-enabled 
build crashes"

Comment 22

[brentmh]

Since this can crash non-MathML browsers too, and we since have a fix in hand,
can I make a desparate plea to get this into 0.9.2?

Thanks

Comment 23

[Christopher Blizzard (:blizzard)]

Use NS_STATIC_CAST() and you will have an sr=blizzard.

Comment 24

[kill this account]

Attachment: patch using NS_STATIC_CAST

Comment 25

[Roland Mainz]

s/rolan/roland/ or use my nickname if you like it short... :-)

Comment 26

[Frank Tang]

we should land this into trunk. I think we should also check in to branch. 

Comment 27

[Frank Tang]

check into trunk now.

Comment 28

[kill this account]

mark fixed

Comment 29

[Frank Tang]

This also cause a crash on HP build- see bug 90758

Comment 30

[kill this account]

there is a bug in the patch:

3718       if (western_font) {
3719         NS_ASSERTION(western_font->SupportsChar(aChar), "font supposed to 
support this char");
3720         return font;
3721       }

this should return western_font not font

Comment 31

[Roland Mainz]

Attachment: patch for 2001-07-27-08-trunk

Comment 32

[Roland Mainz]

Filed patch for GTK+-port and Xlib-port.

Requesting r=/sr= and a= ...

Comment 33

[Christopher Blizzard (:blizzard)]

sr=blizzard

Comment 34

[kill this account]

patch checked in