So I'm having some bizarre issues that I can't quite figure out how to deal with. I think there have been a few ways that this has presented, but the one that consistently happens is at the very beginning of my project I have a line that looks like this:
pc.uifname = malloc(strlen(uifname)+1);
strcpy(pc.uifname, uifname);
These are the first two lines in load_ui(), which is one of the very first functions I'm running. What's happening is that the malloc call isn't "taking". It's not being skipped, because the execution pauses there if I set a breakpoint, but malloc is only allocating like 5 bytes to the array, rather than 27. Setting a watch expression for strlen(uifname)+1 gives me the correct number, setting int len = strlen(uifname)+1 gives me the right number, but malloc is returning a much smaller value for some reason.
And it gets worse. If I insert the line "int len = strlen(uifname)+1" before the malloc call, the malloc call works fine. No change in that line, just assigned a random variable beforehand. However, somewhere way later in the program, I get a GPF (at a call to InsertListItem). Update: After removing the int len = strlen(uifname)+1 line, the malloc call works again, but the GPF is still there.
Here's all the code that runs before load_ui() is called:
int main (int argc, char *argv[])
{
//Generate locks, thread safe variables
CmtNewLock(NULL, OPT_TL_PROCESS_EVENTS_WHILE_WAITING, &lock_pb);
CmtNewLock(NULL, OPT_TL_PROCESS_EVENTS_WHILE_WAITING, &lock_pp);
CmtNewLock(NULL, OPT_TL_PROCESS_EVENTS_WHILE_WAITING, &lock_DAQ);
CmtNewLock(NULL, OPT_TL_PROCESS_EVENTS_WHILE_WAITING, &lock_plot);
InitializeQuitUpdateStatus();
InitializeQuitIdle();
InitializeDoubleQuitIdle();
InitializeStatus();
idle_thread = NULL;
update_thread = NULL;
if (InitCVIRTE (0, argv, 0) == 0)
return -1; /* out of memory */
//Main panel
char *uifname = "Magnetometer Controller.uir"; // So it's easy to change this if we need to.
if(load_ui(uifname)) // This function loads the UI and creates the structs.
return -1;
This has happened before, but usually I start trying to debug it and after some time it goes away. It's happened on multiple machines with the same presentation, so it's not likely to be bad physical memory. I've restarted LabWindows a few times. At the first call in load_ui, the resource tracker was showing 1 block of memory, 5 bytes, allocated, plus the thread locks and thread safe variables.
Any ideas on what this could be?
