Dear All,
Please remove my e-mail adress from your list
E-mail: [email protected]
Thanks,
E van den Boom
----- Original Message -----
From: "JMGross" <[email protected]>
To: "MSPGCC mailing list," <[email protected]>
Sent: Monday, January 25, 2010 1:27 PM
Subject: Re: [Mspgcc-users] Possible preprocessor bug
Hi!
After you telling me that it works as expected for you, I tried some more
test cases. And yes, in your setup it works.
The difference is that the include file in question was placed differently.
If the included files are in the project folder, all works as expected.
If they are in the msp430/include folder, things do not work as expected. Or
maybe they do, but not as _I_ expected.
try the following:
add the lines
#include "msp430x54xx.h"
#define WDTCTL_ 0
#ifdef WDTCTL_
#warning redefinition warning expected
#define WDTCTL_ 1
#endif
right at the start of your projects "msp430xyyy.h" file and compile.
On my machine it gives
---------------------
c:/Programme/msp430/mspgcc_0812/bin/msp430-gcc -v -c -std=gnu99 -g -Os -Wall
-Wcast-align -Wcast-qual -Wimplicit -Wnested-externs -Wpointer-arith -Wredundant-decls
-Wreturn-type -Wshadow -Wstrict-
prototypes -Wswitch -Wunused -Wundef -Wunreachable-code -mmcu=msp430x5438 -Wa,-ahlms=obj/ez3.lst
-I. -I../EPOS_COMMON_3 ez3.c -o obj/ez3.o
Reading specs from
c:\Programme\msp430\mspgcc_0812\bin\..\lib\gcc-lib\msp430\3.2.3\specs
Configured with:
./configure --target=msp430 --prefix=c:/work/mspgcc/sf/packaging/build/installed
--disable-multilib --disable-libc --disable-libssp --disable-intl --disable-libiberty
--with-gcc --with-gnu-ld --with-gnu-as --
with-stabs --disable-shared --disable-threads --disable-win32-registry --disable-nls
--enable-languages=c,c++
Thread model: single
gcc version 3.2.3
c:\Programme\msp430\mspgcc_0812\bin\..\lib\gcc-lib\msp430\3.2.3\cc1.exe -lang-c
-v -I. -I../EPOS_COMMON_3 -iprefix
c:/Programme/msp430/mspgcc_0812/bin/..\lib/gcc-lib/msp430\3.2.3\ -D__GNUC__=3
-
D__GNUC_MINOR__=2 -D__GNUC_PATCHLEVEL__=3 -D__GXX_ABI_VERSION=102 -DMSP430 -D__MSP430__
-D__MSP430 -D__OPTIMIZE_SIZE__ -D__OPTIMIZE__ -D__STDC_HOSTED__=1 -DMSP430
_HAS_HW_MUL -D__MSP430_5438__ -D__MSP430X2__ -DMSP430_HAS_HWMUL -D__SIZE_TYPE__=unsigned
int -D__PTRDIFF_TYPE__=int -D__INT_MAX__=32767 ez3.c -quiet -dumpbase
ez3.c -
mmcu=msp430x5438 -g -Os -Wall -Wcast-align -Wcast-qual -Wimplicit -Wnested-externs
-Wpointer-arith -Wredundant-decls -Wreturn-type -Wshadow -Wstrict-prototypes
-Wswitch -Wunused -Wundef -
Wunreachable-code -std=gnu99 -version -o
C:\DOKUME~1\Grossi\LOKALE~1\Temp/ccuoFwwf.s
GNU CPP version 3.2.3 (cpplib) (GNU assembler syntax)
GNU C version 3.2.3 (msp430)
compiled by GNU C version 3.4.5 (mingw-vista special r3).
ignoring nonexistent directory
"c:/Programme/msp430/mspgcc_0812/msp430/sys-include"
ignoring nonexistent directory
"c:/work/mspgcc/sf/packaging/build/installed/include"
ignoring nonexistent directory
"c:/work/mspgcc/sf/packaging/build/installed/lib/gcc-lib/msp430/3.2.3/include"
ignoring nonexistent directory
"c:/work/mspgcc/sf/packaging/build/installed/lib/gcc-lib/msp430/3.2.3/../../../../msp430/sys-include"
ignoring nonexistent directory
"c:/work/mspgcc/sf/packaging/build/installed/lib/gcc-lib/msp430/3.2.3/../../../../msp430/include"
#include "..." search starts here:
#include <...> search starts here:
.
../EPOS_COMMON_3
c:/Programme/msp430/mspgcc_0812/lib/gcc-lib/msp430/3.2.3/include
c:/Programme/msp430/mspgcc_0812/msp430/include
End of search list.
In file included from ../EPOS_COMMON_3/system54xx.h:27,
from system.h:5,
from ez3.c:11:
c:/Programme/msp430/mspgcc_0812/msp430/include/msp430x54xx.h:6:2: warning:
#warning redefinition warning expected
c:\Programme\msp430\mspgcc_0812\bin\..\lib\gcc-lib\msp430\3.2.3\..\..\..\..\msp430\bin\as.exe
--traditional-format -mmcu=msp430x5438 -ahlms=obj/ez3.lst -o
obj/ez3.oC:\DOKUME~1\Grossi\LOKALE~1\Temp/ccuoFwwf.s 262 16 8 286 11e
obj/ez3.o------------------------------you see, the manually placed warning is printed, but the
redefinition is notdetected.It looks like the redefinition warning is suppressed when (msp)gcc
includesa file from its global include path, but not when the included file is inthe project folder
itself or one of the folders you added with the -Ioption.Since all global include files do an
#undef before redefining things, I'mnot sure whether this is an intentional behaviour. And even if
it is, itmakes working on the global include files more difficult - and a real painifyou're not
aware of this.As a side effect I noticed that mspgcc does not add dependencies to globalinclude
files into its generated dependency file, so if you change one ofthem, make will not remake the
object files which include it. Onceagain, maybe intended (it definitely speeds up dependency
checking) butstill a source of problems if you are working on the global includes.So question is:
is there an option to disable this, at least as long as oneis working on the global includes.I just
found a workaround: by manually adding the global include folder tothe search path (option
-Ic:/Programme/msp430/mspgcc_0812/msp430/include),warnings do come as expected. And I got a couple
morewarnings than expected.There are some more undiscovered redefinitions in theglobal includes: in
string.h, memset is defined twice (identical), thestructured register declaration in wdt_a.h
complains about a missingsemicolon and there was another double-definition in my
ownunified_clock_system.h that slipped detection. All of them not harmful butstill undetected
before.JMGross----- Ursprüngliche Nachricht -----Von: Przemek KlosowskiAn:
[email protected] am: 22 Jan 2010 23:23:40Betreff: Re: [Mspgcc-users] Possible
preprocessor bugJens-Michael,I didn't understand the problem. Perhaps you could explain it by
providingan example of a minimal set of .h files that demonstrate your problem byshowing what CPP
gets vs. what you expected.I for instance made three files, a.h, b.h and c.h:a.h:#define WDTCTL0
1b.h:#define WDTCTL0 2c,h:#include "a.h"#include "b.h"and my 3.2.3 mspgcc seems
to do the right
thing:[...]------------------------------------------------------------------------------Throughout
its 18-year history, RSA Conference consistently attracts theworld's best and brightest in the
field, creating opportunities forConferenceattendees to learn about information security's most
important issuesthroughinteractions with peers, luminaries and emerging and established
companies.http://p.sf.net/sfu/rsaconf-dev2dev_______________________________________________Mspgcc-users
mailing
[email protected]https://lists.sourceforge.net/lists/listinfo/mspgcc-users
