Programming Forums > Application Development > C

deanosrs · May 11th, 2006, 6:42 AM

Hi guys - remember me from that pointer crisis!!!?

Got a much smaller problem this time. For some reason gcc on linux (but strangely enough not on windows/cygwin) is refusing to recognise a call to the pow function even though i have included math.h .

These are the relevant lines of code:

(Toggle Plain Text)

#include <math.h>
...
b = pow(2, (c));

b and c are declared as ints, whereas my research has told me they should strictly be doubles - but even when they are declared as doubles and I put 2.0 instead of 2, I still get the exact same error message, which is:

(Toggle Plain Text)

/tmp/ccK1ydZK.o(.text+0x91d): In function `genbitbinary':
: undefined reference to `pow'

Any help greatly appreciated - this coursework is finally due in tomorrow and can't find anyone who knows what's going on here!

DaWei · May 11th, 2006, 7:28 AM

Perhaps you could peer into your suspect copy of math.h and see if it mentions 'pow'. Seems silly, I know.

Narue · May 11th, 2006, 8:47 AM

You need to link with the math library:

(Toggle Plain Text)

gcc prog.c -lm

deanosrs · May 11th, 2006, 9:25 AM

Mmmm... well I can't control how it's compiled so looks like I'm going to have to write myself a pow function. Is there anywhere i can get the one that should be in math.h?

Infinite Recursion · May 11th, 2006, 9:32 AM

(post deleted)

Disregard, I failed to realize it was C and not C++. Thanks DaWei

DaWei · May 11th, 2006, 9:32 AM

Narue is implying that it's a link error, not a compile error. I don't use gcc, so I didn't recognize that. A compile error means that it wasn't declared or defined in a place the compiler could find it. That would be your source code or your included headers. A link error means that actual code to fulfill the promise of the declaration wasn't found. That means a library or object file (or whatever they may be called your particular system) wasn't found. You have to make sure that you have the library and that it's in a place the linker knows to look. It's the partner to having to have an include file and the compiler having to know where to find it.

This is the C forum, IR. I have no idea how he's actually writing and compiling his code.

OP, you need to distinguish between declarations and code/definitions. It's confusing you. Header files say something is/will be available before the process is finished, they don't necessarily provide it to the compiler. The compiler only wants to know how to reference it. The linker needs the Real Thang.

InfoGeek · May 11th, 2006, 10:52 AM

Quote:

... have to write myself a pow function...

It quite simple if you're working with only ints.

(Toggle Plain Text)

int power(int base,int exponent)
{
   int i,ret=base;
   for(i=1;i<exponent;++i)
      ret *= base;
   return ret;
}

May 11th, 2006, 6:42 AM	#1
deanosrs Programmer Join Date: Apr 2006 Location: Bristol, UK Posts: 31 Rep Power: 0	include math.h not working for pow function Hi guys - remember me from that pointer crisis!!!? Got a much smaller problem this time. For some reason gcc on linux (but strangely enough not on windows/cygwin) is refusing to recognise a call to the pow function even though i have included math.h . These are the relevant lines of code: (Toggle Plain Text) #include <math.h> ... b = pow(2, (c)); #include <math.h> ... b = pow(2, (c)); b and c are declared as ints, whereas my research has told me they should strictly be doubles - but even when they are declared as doubles and I put 2.0 instead of 2, I still get the exact same error message, which is: (Toggle Plain Text) /tmp/ccK1ydZK.o(.text+0x91d): In function `genbitbinary': : undefined reference to `pow' /tmp/ccK1ydZK.o(.text+0x91d): In function `genbitbinary': : undefined reference to `pow' Any help greatly appreciated - this coursework is finally due in tomorrow and can't find anyone who knows what's going on here!

May 11th, 2006, 7:28 AM	#2
DaWei Resident Grouch Join Date: Jun 2005 Posts: 6,453 Rep Power: 10	Perhaps you could peer into your suspect copy of math.h and see if it mentions 'pow'. Seems silly, I know. __________________ Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code. Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

May 11th, 2006, 8:47 AM	#3
Narue Professional Programmer Join Date: Sep 2005 Posts: 419 Rep Power: 4	You need to link with the math library: (Toggle Plain Text) gcc prog.c -lm gcc prog.c -lm __________________ Even if the voices aren't real, they have some pretty good ideas.

May 11th, 2006, 9:32 AM	#5
Infinite Recursion Programming Guru Join Date: Jul 2004 Location: United States Posts: 3,473 Rep Power: 8	(post deleted) Disregard, I failed to realize it was C and not C++. Thanks DaWei __________________ http://jasonpowers.net "There are a thousand hacking at the branches of evil to one who is striking at the root."

May 11th, 2006, 9:32 AM	#6
DaWei Resident Grouch Join Date: Jun 2005 Posts: 6,453 Rep Power: 10	Narue is implying that it's a link error, not a compile error. I don't use gcc, so I didn't recognize that. A compile error means that it wasn't declared or defined in a place the compiler could find it. That would be your source code or your included headers. A link error means that actual code to fulfill the promise of the declaration wasn't found. That means a library or object file (or whatever they may be called your particular system) wasn't found. You have to make sure that you have the library and that it's in a place the linker knows to look. It's the partner to having to have an include file and the compiler having to know where to find it. This is the C forum, IR. I have no idea how he's actually writing and compiling his code. OP, you need to distinguish between declarations and code/definitions. It's confusing you. Header files say something is/will be available before the process is finished, they don't necessarily provide it to the compiler. The compiler only wants to know how to reference it. The linker needs the Real Thang. __________________ Abstraction doesn't make it impossible to write bad code; it makes it possible to write superior code. Contributor's Corner: Grumpy on C++ Exceptions DaWei on Pointers

May 11th, 2006, 9:25 AM	#4
deanosrs Programmer Join Date: Apr 2006 Location: Bristol, UK Posts: 31 Rep Power: 0	Mmmm... well I can't control how it's compiled so looks like I'm going to have to write myself a pow function. Is there anywhere i can get the one that should be in math.h?

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Programming Forums > Application Development > C

include math.h not working for pow function