Printable View
I have recently (last week) restarted programming in C (gcc using Cygwin) after many years of Delphi programming. I am busy looking through code for embedded applications, which I must maintain & develop.
I understand the flow of most of it, but there is one reference in the makefile that I cannot understand, and no amount of googling has answered the question.
What does this mean:
I know what #include is, and what it does ...Code:!include
Thanks for your patience.
#include doesn't go into makefiles...
Are you talking about a makefile or a source file? Can you maybe post the entire line (and the ones before and after)?
QuantumPete
Erm - that's why I am confused. :) The file is definitely the makefile. The compiler used for this code is Borland BCC, v5.02, I believe. And it does compile, as well as make.Quote:
Originally Posted by QuantumPete
Here's a few lines:
And there's an equally confusing exclaimation mark later on in the same code snippet:Code:!include $(MAGIC_ROOT)\config.ini
#CODE_SEGMENT = 0x4000
#CODE_SEGMENT = 0x6100
CODE_SEGMENT = 0x2cca
#CODE_SEGMENT = 0x2a74
#CODE_SEGMENT = 0x2b23
ROOT_NAME = APP
SIGAPPLIB = $(AM)\LIB\am_call.lib \
$(AM)\LIB\pinapi.lib \
$(OS)\lib\os_call.lib \
$(AM)\LIB\sgl_call.lib \
$(CTRY)\LIB\sfl_call.lib
.
.
.
.
Code:# executable without debug
$(EXE_FILE).rom : $(COMMON_LIST) $(SIGAPPLIB) $(APP_OBJ_LIST)
$(TLINK) $(LINKOPT) @&&!
$(APP_OBJ_LIST), $(EXE_FILE).rom, $(EXE_FILE).map, $(SIGAPPLIB) $(CC_LIB)
!
$(HDRTOOL) $(EXE_FILE).rom $(ROOT_NAME).ini +c
That is a makefile (or something like a makefile). It then includes a config.ini from some "magic" place.
The principle is the same as for #include - but since # in makefiles means a comment, you couldn't use #include without breaking old makefiles (which may have #include as a comment somewhere).
--
Mats
You learn something new every day.Quote:
Originally Posted by matsp
Thank you very much.Quote:
Originally Posted by matsp
Another stupid newbie question ...
I'm used to Pascal/Delphi, which works well with strings.
I have written a function to return to me the hardware platform my app is running on. The function declaration looks like this:
Somewhere else, I'd like to call this function to find out what I have got. So I do this:Code:char* INIT_GetHWType(void)
{
char* HWType;
do stuff
do more stuff
return HWType;
}
This gives me a warning : Warning ..\source\init.c 45: Nonportable pointer comparison in function INIT_StartApp()Code:{
WORD wTermStatus;
WORD wBankResult;
BYTE szPin[5];
Storage_Init();
if (INIT_GetHWType == "xyz")
{
//Set display size options here
}
etc., etc., etc.
So I try the following:
And this gives a compiler error.Code:{
WORD wTermStatus;
WORD wBankResult;
BYTE szPin[5];
char* HWType;
Storage_Init();
HWType = INIT_GetHWType;
if ( HWType== "xyz")
{
//Set display size options here
}
etc., etc., etc.
Error ..\source\init.c 46: Cannot convert 'char * (*)()' to 'char *' in function INIT_StartApp()
OK, so I've googled this, and vaguely understand what the problem is. What I do not understand is: How do I assign the result of a function (which returns a string) to a variable that contains a string?
In Pascal, I would do something like the following:
How do I do the same or similar in C?Code:
function MyFunction : string;
begin
do stuff
result := 'stuff';
end;
procedure Callfunction
var
astring : string;
begin
astring := MyFunction;
end;
You compare stringes with strcmp (INIT_GetHWType, "xyz")
The == doesn't work on strings (pointers to char)
And by INIT_StartApp() you mean INIT_GetHWType()??
Actually, thisassigns a the INIT_GetHWType function address to HWType. The compiler moans because HWType is a pointer to char, and INIT_GetHWType is a pointer to a function that returns a char *. My guess is that you wantCode:HWType = INIT_GetHWType;
Then of course, you also need to listen to C_ntua's comments on comparing strings.Code:HWType = INIT_GetHWType();
--
Mats
:) Thanks. I have found a similar problem with integers ... so your answer makes perfect sense.
It's probably your "pascal" history that plays tricks. In Pascal, the name of a function or procedure on it's own, with no parenthesis is a valid way to call that function or procedure. In C and C++ you ALWAYS need the parenthesis, even if there's no parameters.Quote:
Originally Posted by TimL
I once had (what I thought was) a call to a function in some code, and I couldn't figure out why it wasn't doing what I expected. The compiler didn't give any warnings or anything - it turned out to be that I had writteninstead ofCode:func;
Since there was no result or parameters, the compiler did just do what I told it: take the address of this function, and don't do anything with it. Which of course wasn't quite what I meant, but the compiler did exactly what I told it to do.Code:func();
--
Mats
:D That sounds like what's been happening to me.
Thanks for all your help and explainations, people.
Yet another stupid question ...
What does the double colon do in the following function declaration? I have looked at the Borland C++ language reference, and it talks about the "scope access operator", but I don't understand how that applies here ...
Code:bool O_Comms::Init (PORT_TYPE port, COMM_PARAM_STRUCT* psComParam)
That has nothing to do with C. That is C++ and has to do with things like classes and namespaces.