247629 - (PATCH)mozilla-thunderbird fails to build on 64-bit solaris 9 with compile error under freetype2 tree

Status: RESOLVED FIXED

(Core Graveyard :: GFX: Gtk, defect)

Description

[Sunil]

User-Agent:       Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.6) Gecko/20040424 Galeon/1.3.14
Build Identifier: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.6) Gecko/20040424 Galeon/1.3.14

I get compile error when I try to build mozilla-thunderbird-0.7. The
/private/portage/var/tmp/portage/mozilla-thunderbird-0.7/work/mozilla/gfx/src/freetype
 directory fails to compile with errors like this:

-----------------------
....
Z_DLL_SUFFIX=".so" -DXP_UNIX=1 -DUNIX_ASYNC_DNS=1 -DJS_THREADSAFE=1
-DNS_PRINT_PREVIEW=1 -DNS_PRINTING=1 -DMOZILLA_LOCALE_VERSION="1.7"
-DMOZILLA_REGION_VERSION="1.7" -DMOZILLA_SKIN_VERSION="1.5" -D_MOZILLA_CONFIG_H_
-DMOZILLA_CLIENT nsFreeType.cpp
CC: Warning: -xO1 overwrites previously set optimization level of -O3
"nsFreeType.cpp", line 101: Error: Cannot cast from int(*)(FT_FaceRec_*)* to int.
"nsFreeType.cpp", line 102: Error: Cannot cast from int(*)(FT_LibraryRec_*)* to int.
"nsFreeType.cpp", line 103: Error: Cannot cast from int(*)(FT_GlyphRec_*)* to int.
...
----------------------
The attched patch resolves the issue with function pointer offsets being
integer. The patch makes it long. I am assuming that since its the same code
base, mozilla 64-bit must suffer from same bug....

Reproducible: Always
Steps to Reproduce:
1. compile mozilla-thunderbird-0.7
2.
3.

Actual Results:  
compile errors

Expected Results:  
should compile

Comment 1

[Sunil]

Attachment: use long for the function pointer offset instead of int...duh?

Comment 2

[Andrew Schultz]

gtk

Comment 3

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 151192 [details] [diff] [review]
use long for the function pointer offset instead of int...duh?

Requesting approval for simple build bustage fix on 64-bit systems.

Comment 4

[chris hofmann]

Comment on attachment 151192 [details] [diff] [review]
use long for the function pointer offset instead of int...duh?

a=chofmann

Comment 5

[Boris Zbarsky [:bzbarsky]]

Patch checked in for 1.8b2.  Sunil, thanks for the patch!  In general, for
future reference, you want to request reviews on a patch to make sure it
actually gets noticed by someone... :(

Comment 6

[Boris Zbarsky [:bzbarsky]]

*** Bug 310338 has been marked as a duplicate of this bug. ***

Comment 7

[Boris Zbarsky [:bzbarsky]]

So bug 310338 is about 1.7 branch.  Do we want this patch on those branches?