330331 - Building with option --disable-xul failed

Status: RESOLVED FIXED

(Core Graveyard :: File Handling, defect)

Description

[Oleg Romashin (:romaxa)]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20060205 Debian/1.7.12-1.1
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20060205 Debian/1.7.12-1.1

Building mozilla form latest trunk sources with option --disable-xul  failed.


Reproducible: Always

Steps to Reproduce:
Steps to Reproduce:
1. cvs -z3 -d :pserver:[e-mail]:/cvsroot co
mozilla/client.mk
2. cp mozilla/minimo/config/mozconfig/linux_x86 mozilla/mozconfig
3. add option "--disable-xul" to mozconfig file
4. make -f client.mk checkout
autoconf && ./configure && make

Actual Results:  
Building failed with many errors related to disabled RDF component, and usriloader/ExternalHandler (that failed to build without RDF).


Expected Results:  
Building finished successefully

Will be fixed in next patch

Comment 1

[Oleg Romashin (:romaxa)]

Attachment: BUGFIX: patch for fixing of this bug

This is BUGFIX for this bug
Has been added additional option --disable-exthandler that depends from --disable-xul
Has been fixed sources by using ifdef MOZ_EXTHANDLER

Comment 2

[Doug Turner (:dougt)]

Comment on attachment 214900 [details] [diff] [review]
BUGFIX: patch for fixing of this bug

please use `cvs diff -u10` when attaching patches.

drop the extra spaces added in configure.in

In your disable exthandler, i think you also want the optional flag:

+[  --disable-exthandler           Disable Exthandler],
+    MOZ_EXTHANDLER=,
+    MOZ_EXTHANDLER=1)

Is there a way to instead ifdef out the RDF stuff from exthandler so that we might have a smaller patch?

Comment 3

[Benjamin Smedberg]

I don't think we particularly need or want the MOZ_EXTHANDLER makefile var/define, just re-use MOZ_XUL or wait for the embedded-profile work in bug 318041

Comment 4

[Oleg Romashin (:romaxa)]

(In reply to comment #3)
> I don't think we particularly need or want the MOZ_EXTHANDLER makefile
> var/define, just re-use MOZ_XUL or wait for the embedded-profile work in bug
> 318041
> 

This bug also appears, even with using updated configure.in file from your comment (318041)
In first "non official" patch for this bug, instead MOZ_EXTHANDLER was MOZ_XUL.
I can revert this feature back...

Comment 5

[Oleg Romashin (:romaxa)]

Attachment: Some missed ifdef's has been added to new patch

Sorry, but in any case, in previous patch has been missed some defines.
This patch contain additional ifdefs.

Comment 6

[Oleg Romashin (:romaxa)]

(In reply to comment #2)
> (From update of attachment 214900 [details] [diff] [review] [edit])
> please use `cvs diff -u10` when attaching patches.

Sorry but I have some problems with updating from CVS ("cvs diff" is longer operation on my connection), I hope it will be fixed in future ;)

Comment 7

[Benjamin Smedberg]

Comment on attachment 214913 [details] [diff] [review]
Some missed ifdef's has been added to new patch

A uriloader peer needs to review this; deferring to biesi.

Comment 8

[Christian :Biesinger (don't email me, ping me on IRC)]

Why disable exthandler?

Comment 9

[Benjamin Smedberg]

Because RDF is not available in the minimal profile.

Comment 10

[Christian :Biesinger (don't email me, ping me on IRC)]

exthandler does more than parse mimeTypes.rdf. Do you really not want any helper app dialogs?

Comment 11

[Oleg Romashin (:romaxa)]

Attachment: Patch for building exthandler without RDF (MOZ_XUL - disabled)

This patch should set needed ifdef's for building uriloader/exthandler without RDF, or with --disable-xul option

Comment 12

[Oleg Romashin (:romaxa)]

Attachment: Some ifdefs for building mozilla without XUL

This patch should be applied together with previous patch 215864,

Also it has been tested with next configuration:
./configure  --enable-application=minimo --prefix=/usr --enable-default-toolkit=gtk2 --enable-pango --enable-xft --disable-freetype2 --enable-reorder --disable-js-static-build --enable-embedding-tests --with-system-png --with-system-jpeg --with-system-zlib --with-system-bz2 --without-system-nspr --disable-xinerama --disable-svg --disable-jsd --disable-mathml --disable-pedantic --disable-gnomevfs --disable-gnomeui --disable-tests --disable-canvas --disable-xpcom-obsolete --disable-js-binary --without-readline --disable-installer --disable-javaxpcom --disable-mailnews --disable-ldap --disable-postscript --disable-xprint --disable-oji --disable-ctl --disable-accessibility --disable-view-source --enable-single-profile --disable-printing --enable-plaintext-editor-only --disable-composer --disable-xpcom-fastload --disable-negotiateauth --disable-universalchardet --disable-javaxpcom --disable-ldap-experimental --disable-updater --disable-profilesharing --disable-profilelocking --disable-necko-disk-cache --enable-necko-small-buffers --disable-xul --enable-chrome-format=flat '--enable-extensions=xml-rpc xmlextras webservices spatialnavigation' '--enable-protocols=about data http file res jar ftp' --disable-static --enable-shared --enable-optimize=-O0 --disable-strip --enable-debug --enable-logging --enable-long-long-warning --disable-arm-device --enable-elf-dynstr-gc

Comment 13

[Christian :Biesinger (don't email me, ping me on IRC)]

this bug is in the wrong product.

Comment 14

[Christian :Biesinger (don't email me, ping me on IRC)]

I don't really like that MOZ_XUL also means MOZ_RDF. can we add a MOZ_RDF define and test that instead?

(could you add -p to your diff flags, and use a bit more context?)

Comment 15

[Oleg Romashin (:romaxa)]

Attachment: Added MOZ_RDF , and --disable-rdf option

This patch should add MOZ_RDF option, that depends from MOZ_XUL

Comment 16

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 216108 [details] [diff] [review]
Added MOZ_RDF , and --disable-rdf option

ok, comments on the uriloader part:
@@ -509,33 +515,36 @@ nsExternalHelperAppService::~nsExternalH

the RDF service getting should probably move inside the second ifdef

 PRBool nsExternalHelperAppService::MIMETypeIsInDataSource(const char * aContentType)

you could ifdef out the entire function except the return false, as there won't be a datasource w/o RDF.

similar for GetMIMEInfoForMimeTypeFromDS and GetTypeFromDS

and InitDataSource should ABSOLUTELY NOT return success and set mDataSourceInitialized when there is not in fact a data source present.

Comment 17

[Oleg Romashin (:romaxa)]

Attachment: Added MOZ_RDF , and --disable-rdf option

Comment 18

[timeless]

Comment on attachment 224560 [details] [diff] [review]
Added MOZ_RDF , and --disable-rdf option

>Index: configure.in
>@@ -7055,26 +7044,44 @@ AC_SUBST(MOZ_CAIRO_LIBS)
> dnl ========================================================
> dnl disable xul
> dnl ========================================================
> MOZ_ARG_DISABLE_BOOL(xul,
> [  --disable-xul           Disable XUL],
>     MOZ_XUL= )
> if test "$MOZ_XUL"; then
>   AC_DEFINE(MOZ_XUL)
> else
>   dnl remove extensions that require XUL
>+  MOZ_RDF=
>+  ACCESSIBILITY=

I thought about this a lot. I don't like it. instead, i think that the ACCESSIBILITY= bit should be later...
>   MOZ_EXTENSIONS=`echo $MOZ_EXTENSIONS | sed -e 's/inspector//' -e 's/venkman//' -e 's/irc//' -e 's/tasks//'`
>+  
> fi
> 
> AC_SUBST(MOZ_XUL)
> 
> dnl ========================================================
>+dnl accessibility support on by default on all platforms 
>+dnl except OS X.
>+dnl ========================================================
>+MOZ_ARG_DISABLE_BOOL(accessibility,
>+[  --disable-accessibility Disable accessibility support (off by default on OS X)],
>+    ACCESSIBILITY=,
>+    ACCESSIBILITY=1 )
>+

instead of these nesting ifs, i'd do:
>+if test "$MOZ_XUL"; then
 |    dnl accessibility currently requires xul
 |    ACCESSIBILITY=
 |fi
>+if test "$ACCESSIBILITY"; then
>+    AC_DEFINE(ACCESSIBILITY)
>+fi
and remove this fi:
>+fi

The reason for this change is that if anyone is silly enough to actually test $ACCESSIBILITY later in this file, they'll get the right answer. I'd also rather apply the same approach to rdf.

Comment 19

[Benjamin Smedberg]

--disable-xul should never be used by itself but should be part of the --with-embedding-profile feature selections.

Comment 20

[Oleg Romashin (:romaxa)]

(In reply to comment #19)
> --disable-xul should never be used by itself but should be part of the
> --with-embedding-profile feature selections.
> 
I have file new bug againgst --with-embedding-profile=minimal 
340795

Comment 21

[Benjamin Smedberg]

Comment on attachment 224560 [details] [diff] [review]
Added MOZ_RDF , and --disable-rdf option

>Index: configure.in

> dnl ========================================================
> dnl disable xul
> dnl ========================================================
> MOZ_ARG_DISABLE_BOOL(xul,
> [  --disable-xul           Disable XUL],
>     MOZ_XUL= )
> if test "$MOZ_XUL"; then
>   AC_DEFINE(MOZ_XUL)
> else
>   dnl remove extensions that require XUL
>+  MOZ_RDF=
>+  ACCESSIBILITY=
>   MOZ_EXTENSIONS=`echo $MOZ_EXTENSIONS | sed -e 's/inspector//' -e 's/venkman//' -e 's/irc//' -e 's/tasks//'`
>+  
> fi

--disable-xul should not have side affects. Both of these lines disabling accessibility/inspector/venkman/irc/tasks should be removed.

>+if test "$MOZ_XUL"; then
>+if test "$ACCESSIBILITY"; then
>+    AC_DEFINE(ACCESSIBILITY)
>+fi
>+fi

Remove the MOZ_XUL check here as well. the --with-embedding-profile=minimal flag should be enough to disable accessibility (and rdf).

>Index: toolkit/library/nsStaticXULComponents.cpp

> #ifdef MOZ_XUL
> #define XULENABLED_MODULES                   \
>     MODULE(tkAutoCompleteModule)             \
>     MODULE(satchel)                          \
>-    MODULE(PKI)
>+    MODULE(PKI)				     \
>+    MODULE(nsRDFModule)                      \
>+    MODULE(nsXPIntlModule)                   \
>+    MODULE(nsWindowDataSourceModule)
> #else
> #define XULENABLED_MODULES
> #endif

You should probably add a separate RDF_MODULES ifdef MOZ_RDF

>Index: uriloader/exthandler/nsExternalHelperAppService.cpp

>   mDataSourceInitialized = PR_TRUE;

Didn't biesi say this was incorrect? He will need to review the uriloader changes.

>Index: xpfe/browser/src/nsBrowserInstance.cpp

>+#ifdef MOZ_RDF
> #include "nsIBrowserHistory.h"
>+#endif

Why? That header doesn't seem to have any RDF dependencies.

>Index: xpfe/components/build/nsModule.cpp

> #include "nsIGenericFactory.h"
> #include "nsICategoryManager.h"
>-#include "rdf.h"
> #include "nsXPIDLString.h"
>+
>+#ifdef MOZ_RDF
> #include "nsDirectoryViewer.h"
> #include "nsRDFCID.h"
>+#include "rdf.h"
>+#else
>+#include "nsNetUtil.h"
>+#endif

The "else" looks suspicious... why not just include nsNetUtil.h always?

Comment 22

[Oleg Romashin (:romaxa)]

Attachment: fixed rdf disable patch

Comment 23

[Oleg Romashin (:romaxa)]

Attachment: Updated patch according to changes in BUG 340795

Comment 24

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 224916 [details] [diff] [review]
Updated patch according to changes in BUG 340795

note: I didn't really review the changes outside of uriloader/, let me know if you want me to.

uriloader/exthandler/nsExternalHelperAppService.cpp

nsExternalHelperAppService::InitDataSource() should not return success when it doesn't do anything. why don't you ifdef out the entire function and just return NS_ERROR_NOT_AVAILABLE in the non-RDF case?

also, don't add trailing whitespace (on the line before return rv;)

   // can't do anything if we have no datasource...
+
   if (!mOverRideDataSource)

please don't add these empty lines (multiple places)

also, imo returning NS_ERROR_NOT_AVAILABLE in nsExternalHelperAppService::GetMIMEInfoForMimeTypeFromDS makes more sense

xpfe/components/directory/nsDirectoryViewer.h
wrong indentation:
+class nsHTTPIndex : 
+#ifdef MOZ_RDF
+		    public nsIHTTPIndex,
                     public nsIRDFDataSource,
+#endif

Comment 25

[Tom Marn]

(In reply to comment #24)
> (From update of attachment 224916 [details] [diff] [review] [edit])
> note: I didn't really review the changes outside of uriloader/, let me know if
> you want me to.
> 
> uriloader/exthandler/nsExternalHelperAppService.cpp
> 

After apply patches (id=224916) + update 340795, a got build error

c++  -fno-rtti -fno-exceptions -Wall -Wconversion -Wpointer-arith -Wcast-align -Woverloaded-virtual -Wsynth -Wno-ctor-dtor-privacy -Wno-non-virtual-dtor -Wno-long-long -pedantic -fshort-wchar -pthread -pipe  -DDEBUG -D_DEBUG -DDEBUG_tom -DTRACING -g -fno-inline -Os -fPIC -shared -Wl,-z,defs -Wl,-h,libdocshell.so -o libdocshell.so  nsDocShellModule.o    -lpthread           -Wl,--whole-archive ../base/libbasedocshell_s.a ../../uriloader/base/liburiloaderbase_s.a ../../uriloader/exthandler/libexthandler_s.a ../../uriloader/prefetch/libprefetch_s.a ../shistory/src/libshistory_s.a  -Wl,--no-whole-archive -L../../dist/bin -L../../dist/lib  -L../../dist/bin -lmozjs ../../dist/lib/libunicharutil_s.a -L../../dist/bin -lxpcom -lxpcom_core  -L../../dist/bin -L../../dist/lib -lplds4 -lplc4 -lnspr4 -lpthread -ldl   -Wl,-Bsymbolic -lgtk-x11-2.0 -latk-1.0 -lgdk-x11-2.0 -lgdk_pixbuf-2.0 -lm -lpangocairo-1.0 -lfontconfig -lXext -lXrender -lXinerama -lXi -lXrandr -lXcursor -lXfixes -lpango-1.0 -lcairo -lgmodule-2.0 -ldl -lX11 -lgobject-2.0 -lglib-2.0   -ldl -lm    
../../uriloader/exthandler/libexthandler_s.a(nsExternalHelperAppService.o): In function `nsExternalAppHandler::OnStartRequest(nsIRequest*, nsISupports*)':
/home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp:1627: undefined reference to `gNetStrings'


gNetStrings is from docshell component.

tom

Comment 26

[Tom Marn]

Attachment: minimo debug config file.

Comment 27

[Tom Marn]

(In reply to comment #25)
> (In reply to comment #24)

/home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp:1627
> undefined reference to `gNetStrings'
> 
> 

I looked for gNetStrings 

grep gNetStrings * -rni

netwerk/base/public/nsChannelProperties.h:62:#define NS_CHANNEL_PROP_CONTENT_LENGTH gNetStrings->kContentLength
netwerk/base/public/nsNetStrings.h:54:extern NS_HIDDEN_(nsNetStrings*) gNetStrings;
netwerk/base/src/nsNetStrings.cpp:40:NS_HIDDEN_(nsNetStrings*) gNetStrings;
netwerk/build/nsNetModule.cpp:590:    gNetStrings = new nsNetStrings();
netwerk/build/nsNetModule.cpp:591:    return gNetStrings ? NS_OK : NS_ERROR_OUT_OF_MEMORY;
netwerk/build/nsNetModule.cpp:608:    delete gNetStrings;
netwerk/build/nsNetModule.cpp:609:    gNetStrings = nsnull;

and found out that is only used in nsExternalHelperAppService.cpp 1627 as S_CHANNEL_PROP_CONTENT_LENGTH

After changing, just to overcome the problem:


--- mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp.orig    2006-06-14 10:13:52.000000000 +0200
+++ mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp 2006-06-14 10:28:59.000000000 +0200
@@ -1623,8 +1623,10 @@
   // Get content length
   nsCOMPtr<nsIPropertyBag2> props(do_QueryInterface(request, &rv));
   if (props) {
-    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
+/*    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
                                    &mContentLength.mValue);
+*/
+      rv =  NS_ERROR_FAILURE;
   }
   // If that failed, ask the channel
   if (NS_FAILED(rv) && aChannel) {



the following error:occured

There are still libxul dependecies :(

c++ -o GtkPromptService.o -c -fvisibility=hidden -DIMPL_XREAPI -DMOZILLA_INTERNAL_API -D_IMPL_NS_COM -DEXPORT_XPT_API -DEXPORT_XPTC_API -DEXPORT_XPTI_API -D_IMPL_NS_COM_OBSOLETE -D_IMPL_NS_GFX -D_IMPL_NS_WIDGET -DIMPL_XREAPI -DIMPL_NS_NET  -DOSTYPE=\"Linux2.6\" -DOSARCH=\"Linux\" -DBUILD_ID=0000000000 -D_IMPL_GTKMOZEMBED  -I../../../../dist/include/xpcom -I../../../../dist/include/string -I../../../../dist/include/docshell -I../../../../dist/include/necko -I../../../../dist/include/widget -I../../../../dist/include/dom -I../../../../dist/include/gfx -I../../../../dist/include/layout -I../../../../dist/include/uriloader -I../../../../dist/include/webbrwsr -I../../../../dist/include/shistory -I../../../../dist/include/embed_base -I../../../../dist/include/windowwatcher -I../../../../dist/include/profdirserviceprovider -I../../../../dist/include/xulapp -I../../../../dist/include   -I../../../../dist/include/gtkembedmoz -I../../../../dist/include/nspr    -I../../../../dist/sdk/include    -fPIC   -fno-rtti -fno-exceptions -Wall -Wconversion -Wpointer-arith -Wcast-align -Woverloaded-virtual -Wsynth -Wno-ctor-dtor-privacy -Wno-non-virtual-dtor -Wno-long-long -pedantic -fshort-wchar -pthread -pipe  -DDEBUG -D_DEBUG -DDEBUG_tom -DTRACING -g -fno-inline -Os  -I/usr/include/gtk-2.0 -I/usr/lib/gtk-2.0/include -I/usr/include/atk-1.0 -I/usr/include/cairo -I/usr/include/pango-1.0 -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include     -DMOZILLA_CLIENT -include ../../../../mozilla-config.h -Wp,-MD,.deps/GtkPromptService.pp /home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/embedding/browser/gtk/src/GtkPromptService.cpp
rm -f libgtkembedmoz.a
ar cr libgtkembedmoz.a gtkmozembedmarshal.o gtkmozembed2.o EmbedPrivate.o EmbedWindow.o EmbedProgress.o EmbedContentListener.o EmbedEventListener.o EmbedWindowCreator.o EmbedPrompter.o GtkPromptService.o  
ranlib libgtkembedmoz.a
/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/config/nsinstall -R -m 644 libgtkembedmoz.a ../../../../dist/lib
make[5]: Leaving directory `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/embedding/browser/gtk/src'
make[4]: Leaving directory `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/embedding/browser/gtk'
make[4]: Entering directory `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/toolkit/library'
make[4]: *** No rule to make target `../../toolkit/xre/libxulapp_s.a', needed by `libxul.so'.  Stop.


For my compile configure settings, please see attached minimo.config file.

Comment 28

[Oleg Romashin (:romaxa)]

(In reply to comment #27)
> (In reply to comment #25)
> > (In reply to comment #24)
> 
> /home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp:1627
> > undefined reference to `gNetStrings'
> > 
> > 
> 
> I looked for gNetStrings 
> 
> grep gNetStrings * -rni
> 
> netwerk/base/public/nsChannelProperties.h:62:#define
> NS_CHANNEL_PROP_CONTENT_LENGTH gNetStrings->kContentLength
> netwerk/base/public/nsNetStrings.h:54:extern NS_HIDDEN_(nsNetStrings*)
> gNetStrings;
> netwerk/base/src/nsNetStrings.cpp:40:NS_HIDDEN_(nsNetStrings*) gNetStrings;
> netwerk/build/nsNetModule.cpp:590:    gNetStrings = new nsNetStrings();
> netwerk/build/nsNetModule.cpp:591:    return gNetStrings ? NS_OK :
> NS_ERROR_OUT_OF_MEMORY;
> netwerk/build/nsNetModule.cpp:608:    delete gNetStrings;
> netwerk/build/nsNetModule.cpp:609:    gNetStrings = nsnull;
> 
> and found out that is only used in nsExternalHelperAppService.cpp 1627 as
> S_CHANNEL_PROP_CONTENT_LENGTH
> 
> After changing, just to overcome the problem:
> 
> 
> --- mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp.orig   
> 2006-06-14 10:13:52.000000000 +0200
> +++ mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp 2006-06-14
> 10:28:59.000000000 +0200
> @@ -1623,8 +1623,10 @@
>    // Get content length
>    nsCOMPtr<nsIPropertyBag2> props(do_QueryInterface(request, &rv));
>    if (props) {
> -    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
> +/*    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
>                                     &mContentLength.mValue);
> +*/
> +      rv =  NS_ERROR_FAILURE;
>    }
>    // If that failed, ask the channel
>    if (NS_FAILED(rv) && aChannel) {
> 
> 
> 
> the following error:occured
> 
> There are still libxul dependecies :(
> 

I'am wondered.... why you are trying to BUILD XUL based application (MINIMO) without XUL? and this patch should fix building minimal level for xulrunner application.....
 
Also I can't understand what is it MOZILLA_1_9_BRANCH ? could you check it with trunk?


> c++ -o GtkPromptService.o -c -fvisibility=hidden -DIMPL_XREAPI
> -DMOZILLA_INTERNAL_API -D_IMPL_NS_COM -DEXPORT_XPT_API -DEXPORT_XPTC_API
> -DEXPORT_XPTI_API -D_IMPL_NS_COM_OBSOLETE -D_IMPL_NS_GFX -D_IMPL_NS_WIDGET
> -DIMPL_XREAPI -DIMPL_NS_NET  -DOSTYPE=\"Linux2.6\" -DOSARCH=\"Linux\"
> -DBUILD_ID=0000000000 -D_IMPL_GTKMOZEMBED  -I../../../../dist/include/xpcom
> -I../../../../dist/include/string -I../../../../dist/include/docshell
> -I../../../../dist/include/necko -I../../../../dist/include/widget
> -I../../../../dist/include/dom -I../../../../dist/include/gfx
> -I../../../../dist/include/layout -I../../../../dist/include/uriloader
> -I../../../../dist/include/webbrwsr -I../../../../dist/include/shistory
> -I../../../../dist/include/embed_base -I../../../../dist/include/windowwatcher
> -I../../../../dist/include/profdirserviceprovider
> -I../../../../dist/include/xulapp -I../../../../dist/include  
> -I../../../../dist/include/gtkembedmoz -I../../../../dist/include/nspr   
> -I../../../../dist/sdk/include    -fPIC   -fno-rtti -fno-exceptions -Wall
> -Wconversion -Wpointer-arith -Wcast-align -Woverloaded-virtual -Wsynth
> -Wno-ctor-dtor-privacy -Wno-non-virtual-dtor -Wno-long-long -pedantic
> -fshort-wchar -pthread -pipe  -DDEBUG -D_DEBUG -DDEBUG_tom -DTRACING -g
> -fno-inline -Os  -I/usr/include/gtk-2.0 -I/usr/lib/gtk-2.0/include
> -I/usr/include/atk-1.0 -I/usr/include/cairo -I/usr/include/pango-1.0
> -I/usr/include/glib-2.0 -I/usr/lib/glib-2.0/include     -DMOZILLA_CLIENT
> -include ../../../../mozilla-config.h -Wp,-MD,.deps/GtkPromptService.pp
> /home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/embedding/browser/gtk/src/GtkPromptService.cpp
> rm -f libgtkembedmoz.a
> ar cr libgtkembedmoz.a gtkmozembedmarshal.o gtkmozembed2.o EmbedPrivate.o
> EmbedWindow.o EmbedProgress.o EmbedContentListener.o EmbedEventListener.o
> EmbedWindowCreator.o EmbedPrompter.o GtkPromptService.o  
> ranlib libgtkembedmoz.a
> /home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/config/nsinstall -R -m 644
> libgtkembedmoz.a ../../../../dist/lib
> make[5]: Leaving directory
> `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/embedding/browser/gtk/src'
> make[4]: Leaving directory
> `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/embedding/browser/gtk'
> make[4]: Entering directory
> `/home/tom/tests/build/MOZILLA_1_9_BRANCH/build_debug/toolkit/library'
> make[4]: *** No rule to make target `../../toolkit/xre/libxulapp_s.a', needed
> by `libxul.so'.  Stop.
> 
> 
> For my compile configure settings, please see attached minimo.config file.
> 

Comment 29

[Christian :Biesinger (don't email me, ping me on IRC)]

your analysis is wrong, see http://lxr.mozilla.org/seamonkey/source/netwerk/base/public/nsChannelProperties.h#61
it only uses gNetStrings if IMPL_NS_NET is defined, which should only be the case inside of netwerk/ or when libXUL is enabled.

Comment 30

[Oleg Romashin (:romaxa)]

Attachment: Fixes for Comment #24

Comment 31

[Tom Marn]

(In reply to comment #28)
> (In reply to comment #27)
> > (In reply to comment #25)
> > > (In reply to comment #24)
> > 
> > /home/tom/tests/build/MOZILLA_1_9_BRANCH/mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp:1627
> > > undefined reference to `gNetStrings'
> > > 
> > > 
> > 
> > I looked for gNetStrings 
> > 
> > grep gNetStrings * -rni
> > 
> > netwerk/base/public/nsChannelProperties.h:62:#define
> > NS_CHANNEL_PROP_CONTENT_LENGTH gNetStrings->kContentLength
> > netwerk/base/public/nsNetStrings.h:54:extern NS_HIDDEN_(nsNetStrings*)
> > gNetStrings;
> > netwerk/base/src/nsNetStrings.cpp:40:NS_HIDDEN_(nsNetStrings*) gNetStrings;
> > netwerk/build/nsNetModule.cpp:590:    gNetStrings = new nsNetStrings();
> > netwerk/build/nsNetModule.cpp:591:    return gNetStrings ? NS_OK :
> > NS_ERROR_OUT_OF_MEMORY;
> > netwerk/build/nsNetModule.cpp:608:    delete gNetStrings;
> > netwerk/build/nsNetModule.cpp:609:    gNetStrings = nsnull;
> > 
> > and found out that is only used in nsExternalHelperAppService.cpp 1627 as
> > S_CHANNEL_PROP_CONTENT_LENGTH
> > 
> > After changing, just to overcome the problem:
> > 
> > 
> > --- mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp.orig   
> > 2006-06-14 10:13:52.000000000 +0200
> > +++ mozilla/uriloader/exthandler/nsExternalHelperAppService.cpp 2006-06-14
> > 10:28:59.000000000 +0200
> > @@ -1623,8 +1623,10 @@
> >    // Get content length
> >    nsCOMPtr<nsIPropertyBag2> props(do_QueryInterface(request, &rv));
> >    if (props) {
> > -    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
> > +/*    rv = props->GetPropertyAsInt64(NS_CHANNEL_PROP_CONTENT_LENGTH,
> >                                     &mContentLength.mValue);
> > +*/
> > +      rv =  NS_ERROR_FAILURE;
> >    }
> >    // If that failed, ask the channel
> >    if (NS_FAILED(rv) && aChannel) {
> > 
> > 
> > 
> > the following error:occured
> > 
> > There are still libxul dependecies :(
> > 
> 
> I'am wondered.... why you are trying to BUILD XUL based application (MINIMO)
> without XUL? and this patch should fix building minimal level for xulrunner
> application.....
> 
Hi romaxa

I would just like to compile mozilla libraries (libgtkmozembed.lib) for an embedded system (MIPS) with minimal foot print (without xul, etc..) directly from cvs source. Full Minimo browser is not my target. TestGtkEmbed application would be just for testing purposes. This is not my first cross compiling of embedded mozilla. Also, I need shared libraries "--enable-shared".

I don't know if the XUL is just minimo related stuff.


> Also I can't understand what is it MOZILLA_1_9_BRANCH ? could you check it with
> trunk?

Sorry, don't bother with this tag. It is just my labeling from mozilla-config.h file where MOZILLA_LOCALE_VERSION "1.9a1".
Tom

Comment 32

[Benjamin Smedberg]

Tom Marn, bugzilla is not help/discussion forum. For help, please post to the mozilla.dev.embedding newsgroup.

Comment 33

[Tom Marn]

(In reply to comment #32)
> Tom Marn, bugzilla is not help/discussion forum. For help, please post to the
> mozilla.dev.embedding newsgroup.
> 

Sorry Benjamin, if previous post looks off-topic. But it isn't, I just answering   to the romaxa comment and problem that I have with --disable-xul parameter. And is not embedded related problem. I just point out the platform. Maybe I will solve the problem with your help.

Comment 34

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 225582 [details] [diff] [review]
Fixes for Comment #24 

guys, _please_ don't quote the entire comment you are replying to. delete the parts that you are not referring to. thanks.

what about:
>also, imo returning NS_ERROR_NOT_AVAILABLE in
>nsExternalHelperAppService::GetMIMEInfoForMimeTypeFromDS makes more sense

and you are still adding empty lines

Comment 35

[Oleg Romashin (:romaxa)]

Attachment: Update FIX for comment 34

Comment 36

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 225854 [details] [diff] [review]
Update FIX for comment 34

   // Can't do anything without a datasource
+
   if (!mOverRideDataSource)

I think I said a few times now that you should not add this empty line.

Comment 37

[Oleg Romashin (:romaxa)]

Attachment: UPdated to lates trunk, removed empty line

Comment 38

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 226111 [details] [diff] [review]
UPdated to lates trunk, removed empty line

r=biesi on the uriloader/ parts of this patch.

Comment 39

[Oleg Romashin (:romaxa)]

(In reply to comment #38)
> (From update of attachment 226111 [details] [diff] [review] [edit])
> r=biesi on the uriloader/ parts of this patch.
> 

What I should to do for applying this patch to trunk, and applying patch from bug 340795?

Comment 40

[Benjamin Smedberg]

Comment on attachment 226111 [details] [diff] [review]
UPdated to lates trunk, removed empty line

I asked timeless to check the interdiff of this patch and if it looks reasonable to coordinate landing this.

Comment 41

[timeless]

Comment on attachment 226111 [details] [diff] [review]
UPdated to lates trunk, removed empty line

the interdiff is fine.

we'll land these monday hopefully. if i can find a device that has my ssh keys anyway (they're on doppler which i can now access, but i don't actually trust most other devices, and doppler doesn't have real net access yet).

Comment 42

[Oleg Romashin (:romaxa)]

Attachment: Updated to latest trunk

Comment 43

[Oleg Romashin (:romaxa)]

Attachment: rev4 Updated to latest trunk

Comment 44

[Oleg Romashin (:romaxa)]

Attachment: Fixed review comments

Comment 45

[Christian :Biesinger (don't email me, ping me on IRC)]

Comment on attachment 233560 [details] [diff] [review]
Fixed review comments

r=biesi on the uriloader part

Comment 46

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 233560 [details] [diff] [review]
Fixed review comments

>Index: content/html/document/src/nsHTMLDocument.h
>+#ifdef MOZ_RDF
> #include "rdf.h"
> #include "nsRDFCID.h"
> #include "nsIRDFService.h"
>+#endif

Can't these just go away completely?

sr=bzbarsky either way.

Comment 47

[Oleg Romashin (:romaxa)]

Patch applied, and tested.
Closing