[patch, rfc] Don't try to build printcmd.c and valprint.c with -Wformat-nonliteral

Andrew Cagney ac131313@redhat.com
Sat Aug 16 17:49:00 GMT 2003


> Hello,
> 
> On these two I'm, for the moment, giving up (they are the only ones left).
> 
> printcmd.c: This is passing user input to printf as the format string. I think it needs to instead fully parse that format specifier.
> 
> valprint.c: This is running into local_hex_format et.al.  I think that needs an overhaul regardess - written in a way that is more host/target independant.
> 
> Comments, thoughts, anyone want to overhaul the language output format code?  The other option is to wind back the -Werror list, sigh.

I've checked this in.  Bug reports to follow.

Andrew


> 2003-08-10  Andrew Cagney  <cagney@redhat.com>
> 
> 	* Makefile.in (printcmd.o, valprint.o): Do not try to build with
> 	-Werror.  -Wformat-nonliteral problems.
> 
> Index: Makefile.in
> ===================================================================
> RCS file: /cvs/src/src/gdb/Makefile.in,v
> retrieving revision 1.424
> diff -u -r1.424 Makefile.in
> --- Makefile.in	8 Aug 2003 17:30:35 -0000	1.424
> +++ Makefile.in	10 Aug 2003 20:06:32 -0000
> @@ -1321,6 +1321,12 @@
>  	$(CC) -c $(INTERNAL_CFLAGS) $(TARGET_SYSTEM_ROOT_DEFINE) \
>  		-DBINDIR=\"$(bindir)\" $(srcdir)/main.c
>  
> +# FIXME: cagney/2003-08-10: Do not try to build "printcmd.c" with
> +# -Wformat-nonliteral.  It needs to be overhauled so that it doesn't
> +# pass input strings to host printf function.
> +printcmd.o: $(srcdir)/printcmd.c
> +	$(CC) -c $(INTERNAL_WARN_CFLAGS) $(NO_WERROR_CFLAGS) $(srcdir)/printcmd.c
> +
>  # FIXME: Procfs.o gets -Wformat errors because things like pid_t don't
>  # match output format strings.
>  procfs.o: $(srcdir)/procfs.c
> @@ -1337,6 +1343,12 @@
>  		$(TCL_CFLAGS) $(TK_CFLAGS) $(X11_CFLAGS) \
>  		$(GDBTK_CFLAGS) \
>  		$(srcdir)/v850ice.c
> +
> +# FIXME: cagney/2003-08-10: Do not try to build "valprint.c" with
> +# -Wformat-nonliteral.  It relies on local_hex_format et.al. and
> +# that's a mess.  It needs a serious overhaul.
> +valprint.o: $(srcdir)/valprint.c
> +	$(CC) -c $(INTERNAL_WARN_CFLAGS) $(NO_WERROR_CFLAGS) $(srcdir)/valprint.c
>  
>  # FIXME: z8k-tdep.c calls _initialize_gdbtypes().  Since that isn't
>  # declared -Wimplicit fails. It should be using the GDBARCH framework.




More information about the Gdb-patches mailing list