Does pc-lint catch this type of problem?

[Brian]

I am new to lint and have been experimenting with splint. An
initialization problem I was caught on recently was not found by
splint. Would pc-lint catch this problem:

/* test program for splint */
/* body of func1 commented out to recreate lack of initialization */
static int x[3];
void func1( void);
int func2(void);
void func3(int *);

void func1( void)
{
/* when commented out,
x[0]= 0;
x[1]= 1;
x[2]= 2;
x[] is not initialized */
}

int func2(void)
{
int y;
func3(&x[0]);
y= x[0] + x[1] + x[2];
return y;
}

void func3(int * u)
{
/* some filter code */
u[2]= u[1];
u[1]= u[0];
}

Splint 3.1.1 --- 12 April 2003
Finished checking --- no warnings

Thanks,

[Tom St Denis]

Brian wrote:

I am new to lint and have been experimenting with splint. An
initialization problem I was caught on recently was not found by
splint. Would pc-lint catch this problem:

/* test program for splint */
/* body of func1 commented out to recreate lack of initialization */
static int x[3];
x[] == {0, 0, 0}

It is initialized.

void func1( void);
int func2(void);
void func3(int *);

void func1( void)
{
/* when commented out,
x[0]= 0;
x[1]= 1;
x[2]= 2;
x[] is not initialized */
}

int func2(void)
{
int y;
func3(&x[0]);
You could write that as x instead of &x[0].
y= x[0] + x[1] + x[2];
This is valid because x[] == {0, 0, 0}
return y;
}

void func3(int * u)
{
/* some filter code */
u[2]= u[1];
u[1]= u[0];
}

Splint 3.1.1 --- 12 April 2003
Finished checking --- no warnings

The code is valid. Also from the point of view of just looking at the
functiosn blindly func3() is always valid (it doesn't use the globals).

Tom

[Brian]

Tom St Denis <to*@securescience.net> wrote in message news:<ka****************@news04.bloor.is.net.cable .rogers.com>...

Brian wrote:

I am new to lint and have been experimenting with splint. An
initialization problem I was caught on recently was not found by
splint. Would pc-lint catch this problem:

/* test program for splint */
/* body of func1 commented out to recreate lack of initialization */
static int x[3];

x[] == {0, 0, 0}

It is initialized.

void func1( void);
int func2(void);
void func3(int *);

void func1( void)
{
/* when commented out,
x[0]= 0;
x[1]= 1;
x[2]= 2;
x[] is not initialized */
}

int func2(void)
{
int y;
func3(&x[0]);

You could write that as x instead of &x[0].

y= x[0] + x[1] + x[2];

This is valid because x[] == {0, 0, 0}

return y;
}

void func3(int * u)
{
/* some filter code */
u[2]= u[1];
u[1]= u[0];
}

Splint 3.1.1 --- 12 April 2003
Finished checking --- no warnings

The code is valid. Also from the point of view of just looking at the
functiosn blindly func3() is always valid (it doesn't use the globals).

Tom

Ah! your right. I wonder why my embedded DSP code had garbage in x
until I explicitly initialized it. Maybe the boot loader isn't
initializing static data ?
In any case thanks for your help.