Re: strange error on gcc 4.1.0

Re: strange error on gcc 4.1.0

[Jack Howarth]

Here is another posting which describes the problem a little clearer...

http://lists.apple.com/archives/darwin-development/2003/Jun/msg00000.html

So gcc needs to be fixed to not misparse -bundle, when it is the first
argument to gcc, as being the "-b Machine" target option.
     Shouldn't the fix be fairly simple? Just requiring that the
code that recognizes the -b flag require a trailing space after
the -b flag...otherwise it passes over it so that -bundle could be
recognized?
              Jack

Re: strange error on gcc 4.1.0

[Mike Stump]

On Saturday, July 30, 2005, at 09:09 AM, Andrew Pinski wrote:
> But I don't know what happened to it or came up from the discussion.

Looks like no response to Geoff's last comment.  If someone wants to 
address his stated concerns and resubmit it...  otherwise, -Wl,-bundle 
should do the trick.

Re: strange error on gcc 4.1.0

[Jack Howarth]

Thanks. It applies with offsets to the current 4.0 branch. Hopefully we can get
it in both branches soon. Currently everyone on Darwin is just hacking their
Makefiles around the problem.
          Jack

Re: strange error on gcc 4.1.0

[Andrew Pinski]

On Jul 30, 2005, at 11:52 AM, Jack Howarth wrote:

>    I can't find a patch either using google but I did find the  
> following
> posting which describes the problem and a workaround...
>
> http://www.mail-archive.com/fink-devel@lists.sourceforge.net/ 
> msg10604.html

I finally found the patch:
http://gcc.gnu.org/ml/gcc-patches/2002-12/msg00655.html
http://gcc.gnu.org/ml/gcc-patches/2002-10/msg01961.html

But I don't know what happened to it or came up from the discussion.

-- Pinski

Re: strange error on gcc 4.1.0

[Jack Howarth]

   I can't find a patch either using google but I did find the following
posting which describes the problem and a workaround...

http://www.mail-archive.com/fink-devel@lists.sourceforge.net/msg10604.html

                Jack

Re: strange error on gcc 4.1.0

[Andrew Pinski]

> 
> Andreas,
>      Are you sure gnu gcc shouldn't support this? I see -bundle under the
> Machine Dependent Options under Darwin options which are not marked as
> APPLE-ONLY. I understood that to mean that it should be present in the
> gnu gcc compiler (like -all_load which is in the same list and is 
> supported). Fortunately MacOS X 10.4 now supports a more linux-like dyload
> that can use standard dynamic libraries .so files instead of just bundles
> for shared library modules.

It is a bug in the driver,  see PR 21366: <http://gcc.gnu.org/PR21366>.
There was a patch posted about this too but I cannot find it.

Thanks,
Andrew Pinski

Re: strange error on gcc 4.1.0

[Jack Howarth]

Andreas,
     Are you sure gnu gcc shouldn't support this? I see -bundle under the
Machine Dependent Options under Darwin options which are not marked as
APPLE-ONLY. I understood that to mean that it should be present in the
gnu gcc compiler (like -all_load which is in the same list and is 
supported). Fortunately MacOS X 10.4 now supports a more linux-like dyload
that can use standard dynamic libraries .so files instead of just bundles
for shared library modules.
              Jack

Re: strange error on gcc 4.1.0

[Andreas Schwab]

howarth@bromo.msbb.uc.edu (Jack Howarth) writes:

>    I am trying to build xplor-nih on MacOS X 10.4 using gcc 4.1.0
> branch's gcc and g++ instead of Apple's and am running into a really
> strange error. I find the following linkage fails...
>
> g++-4 -bundle -flat_namespace -undefined suppress _xplorWrap.o libswigpy-xplor.dylib -o _xplorWrap.so -L/Users/howarth/Desktop/xplor-nih-2.11.2.1/bin.Darwin_8/ -lcommon -lnmrPot -lintVar -lvmd -lpy  -lswigpy-xplor \

$ info gcc 'Target Options'
[...]
`-b MACHINE'
     The argument MACHINE specifies the target machine for compilation.
[...]
 The `-V' and `-b' options work by running the
`<machine>-gcc-<version>' executable, so there's no real reason to use
them if you can just run that directly.

> Shouldn't the stock gcc 4.1.0 support the -bundle flag?

It's an Apple-specific flag.

Andreas.

-- 
Andreas Schwab, SuSE Labs, schwab@suse.de
SuSE Linux Products GmbH, Maxfeldstraße 5, 90409 Nürnberg, Germany
Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

strange error on gcc 4.1.0

[Jack Howarth]

   I am trying to build xplor-nih on MacOS X 10.4 using gcc 4.1.0
branch's gcc and g++ instead of Apple's and am running into a really
strange error. I find the following linkage fails...

g++-4 -bundle -flat_namespace -undefined suppress _xplorWrap.o libswigpy-xplor.dylib -o _xplorWrap.so -L/Users/howarth/Desktop/xplor-nih-2.11.2.1/bin.Darwin_8/ -lcommon -lnmrPot -lintVar -lvmd -lpy  -lswigpy-xplor \
             -lcrypto -L/Users/howarth/Desktop/xplor-nih-2.11.2.1/bin.Darwin_8/
g++-4: couldn't run 'undle-gcc-4.1.0': No such file or directory

I can't find any evidence of a undle-gcc-4.1.0 file in my build directory
for the current gcc main trunk cvs from which I built the installed
gcc 4.1.0 compiler. Shouldn't the stock gcc 4.1.0 support the -bundle
flag? Any idea why this flag isn't working?
             Jack