swarm-support
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: MacOS X


From: Bill Northcott
Subject: Re: MacOS X
Date: Thu, 2 Jan 2003 10:20:05 +1100

David wrote on 02/01/2003 05:21:34 AM:

> After doing this it gives me an error at line 115 of init.c,
> init.c:115: only 1 arg to macro 'DEBUG_PRINTF' (2 expected)  The line
> in question does, indeed, only have one argument :-)

> This lead me to an #error on line 34 of sendmsg.c when it claimed that
> MFRAME_STACK_STRUCT was undefined.  I defined it in mframe-generic.h,
> so the only reason this should be happening is if something is either
> not being included correctly or my understanding of how this compiles
> is fubar.

> I can bypass this trivially (and I think I know how to fix it, just
> need to analyze the code), but then that runs me into an error:

> objc -I/Users/raven/Documents/Projects/swarm/libobjc/objc
> - -I/Users/raven/Documents/Projects/swarm/libobjc -DBUILDING_LIBOBJC
> - -Wall -Werror -g -O2 -c NXConstStr.m -MT NXConstStr.lo -MD -MP -MF
> .deps/NXConstStr.TPlo  -fno-common -DPIC -o .libs/NXConstStr.lo
> cpp-precomp: could not open 'NXConstStr.lo'

> Anyways, just thought I would make a status report.

This is a problem with the Apple precompiler cpp-precomp.  It does not 
like the GNU technique used in libobjc of compiling a code fragment, using 
it in a couple of compiles and then deleting and recompiling the same code 
fragment for use in the next compile.  I suspect it would be impossible to 
remove this from the Swarm code as long as it is based on GNU objc.  The 
fix is to add -no-cpp-precomp to the CFLAGS thus:

../../../configure \
CC='cc -fgnu-runtime -no-cpp-precomp' \
--prefix=/Users/billn/Public/Swarm/swarm/macosx/swarmx/build/dst/usr \
--with-tcldir=/usr/local/lib \
--without-jdkdir \
--with-hdf5dir=/usr/local

I could commit that configure command to the cvs, because it will soon 
make its way into the PB project.  Do you think that is a good idea? or 
would it be better spooled into the autogen stuff.

Many thanks for your report.  I did the trivial fix of adding a second 
rubbish argument but did not commit it to the cvs because it was obviously 
wrong.  This is my only current uncommitted patch.  If you can see the 
right fix please let me know, and I will commit it.

I currently break here:
cc -fgnu-runtime -no-cpp-precomp -DHAVE_CONFIG_H -I. 
-I/Users/billn/Public/Swarm/swarm/libobjc -I. -I./objc 
-I/Users/billn/Public/Swarm/swarm/libobjc/objc 
-I/Users/billn/Public/Swarm/swarm/libobjc -DBUILDING_LIBOBJC -Wall -Werror 
-g -O2 -c /Users/billn/Public/Swarm/swarm/libobjc/mframe.m -MT mframe.lo 
-MD -MP -MF .deps/mframe.TPlo  -fno-common -DPIC -o .libs/mframe.lo
cc1obj: warnings being treated as errors
/Users/billn/Public/Swarm/swarm/libobjc/mframe.m: In function 
`mframe_build_signature':
/Users/billn/Public/Swarm/swarm/libobjc/mframe.m:146: warning: int format, 
long unsigned int arg (arg 5)

I am not quite understanding 'arg 5' there only seem to be four, or does 
it mean the function prototype has five?

Bill Northcott

                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]