97264 - [xlib] libgfx_xlib.so doesn't link when xprint is enabled

Status: VERIFIED FIXED

(Core Graveyard :: Printing: Xprint, defect, P3)

Description

[cls]

nsFontMetricsXlib.o: In function `nsFontXlib::LoadFont(void)':
nsFontMetricsXlib.o(.text+0x2898): undefined reference to `XpGetContext'
nsFontMetricsXlib.o: In function `GetFontNames(char const *, nsFontNodeArrayXlib
*)':
nsFontMetricsXlib.o(.text+0x4c27): undefined reference to `XpGetContext'
collect2: ld returned 1 exit status

Comment 1

[Roland Mainz]

Swapping QA <--> Owner

Comment 2

[Roland Mainz]

Accepting bug.
Fix follows in a 1-2 hours, my machine is still building...

Comment 3

[Roland Mainz]

cls:
What's the symbol to |ifdef| stuff in Makefile.in when StaticBuild is active ?

Comment 4

[cls]

Why?  This bug has nothing to do with the static build.  BUILD_STATIC_LIBS.

As you pointed out on irc, adding EXTRA_DSO_LDOPTS += $(MOZ_XPRINT_LDFLAGS)
inside the MOZ_ENABLE_XPRINT ifdef fixes the problem.

Comment 5

[Roland Mainz]

cls:
I thougth it may be usefull to kill the |_IMPL_NS_XPRINT| thingies which are
(AFAIK) only required for StaticBuild and link Xprint module against
nsDrawingSurfaceXlib.o (which wastes ~8k on SPARC platforms) if the build is a
StaticBuild ...

Anyway ... I'll file the simple patch now and kill |_IMPL_NS_XPRINT| in a later
bug, OK ?

Comment 6

[Roland Mainz]

Attachment: Patch for 2001-08-27-08-trunk

Comment 7

[Roland Mainz]

Filed patch (one fix, one whitespace-->tab change).

Requesting r= and a= for {0.9.4,trunk}

Comment 8

[Roland Mainz]

cls:
I do not understand one detail:
Why does it work on my Solaris/WS+gcc and SuSE_Linux/gcc builds and fails on
your Linux box... where's the difference ?

Comment 9

[cls]

r=cls

Comment 10

[Roland Mainz]

Requesting a= for {0.9.4,trunk}

Comment 11

[Christopher Blizzard (:blizzard)]

a=blizzard on behalf of drivers for 0.9.4

Comment 12

[Roland Mainz]

cls:
Wanna check this "in", please ?

Comment 13

[cls]

Patch checked in.

Comment 14

[Masaki Katakai]

I'm now verifying the Xprt bugs, however, I'm not sure how to
verify this bug.

Roland,

could you give the exact instruction how to verify this problem,
how to build in static?

Comment 15

[Roland Mainz]

katakai wrote:
> could you give the exact instruction how to verify this problem,

Well... if you build with Xlib toolkit and the build doesn't stop with the error
listed in the initial bug report then it works... :)

Marking VERIFIED myself... :)