SIGSEGV on fclose()
Harald Grossauer
I have a strange problem getting a SIGSEGV from a fclose(). It does not
happen for every set of data but I have a set of data where it happens
always.
The function (code see below) should write an image given as a red[],
green[] and blue[] array into a bitmap file. The output is the following:
<code>
Starting writeimg()
Before fclose()
MPI process rank 0 (n0, p27563) caught a SIGSEGV.
----------------------------------------------------------------------------
-
One of the processes started by mpirun has exited with a nonzero exit
code. This typically indicates that the process finished in error.
If your process did not finish in error, be sure to include a "return
0" or "exit(0)" in your C code before exiting the application.
PID 27563 failed on node n0 with exit status 1.
----------------------------------------------------------------------------
-
</code>
As you can see it is an MPI app. I have a "master" process which does all
the file I/O, spreads the data between several "child" processes, collects
the processed data and writes it back to the file. The strange thing is,
that if I replace a line "vabs = vx*vx+vy*vy;" in the processing routine of
the child processes with the correct "vabs = sqrt(vx*vx+vy*vy);" the SIGSEGV
does not occur!
Ok, here is the code:
int writeimg(const unsigned char* red, const unsigned char* green, const
unsigned char* blue, const int width, const int height, const char* name)
{
int i,j;
char bitmapHeader[60];
*(short*)&bitmapHeader[0] = 0x4d42; // BM
*(int*)&bitmapHeader[2] = 0; // filesize in byte, may be 0 for uncompressed
*(int*)&bitmapHeader[6] = 0; // Reserved
*(int*)&bitmapHeader[10] = 56; // Offset of image data
*(int*)&bitmapHeader[14] = 40; // size of bitmapinfoheader structure
*(int*)&bitmapHeader[18] = width; // width
*(int*)&bitmapHeader[22] = height; // height
*(short*)&bitmapHeader[26] = 1; // planes
*(short*)&bitmapHeader[28] = 24; // bitcount
*(int*)&bitmapHeader[30] = 0; // compression
*(int*)&bitmapHeader[34] = 0; // SizeImage (may be set to zero for
uncompressed images)
*(int*)&bitmapHeader[38] = 0;
*(int*)&bitmapHeader[42] = 0;
*(int*)&bitmapHeader[46] = 0;
*(int*)&bitmapHeader[50] = 0;
FILE* file;
if(!(file = fopen(name, "w"))) {cout << name << ": Error opening file\n";
return 1;}
if(fwrite(bitmapHeader, 54, 1, file) != 1) {cout << name << ": Error
writing bitmap header\n"; fclose(file); return 1;}
int noPadBytes = (4-(width*3)%4)%4;
int resr = 0; int resg = 0; int resb = 0;
for ( j = 0; j < height; j++)
{
for(i = 0; i < width && resr!=EOF && resg!=EOF && resb!=EOF ; i++)
{
resb = fputc(blue[j*width+i], file);
resg = fputc(green[j*width+i], file);
resr = fputc(red[j*width+i], file);
}
if(resr==EOF || resg==EOF || resb==EOF){cout << name << ": Error writing
image data\n"; fclose(file); return 1;}
if(noPadBytes)
{
for(i=0;i<noPadBytes;i++)
{
fputc(0, file);
}
}
fflush(file);
}
cout << "Before fclose()\n";
fclose(file);
cout << "After fclose()\n";
return 0;
} // end writeimg()
Kevin Easton
> Hi,
>
> I have a strange problem getting a SIGSEGV from a fclose(). It does not
> happen for every set of data but I have a set of data where it happens
> always.
>
> The function (code see below) should write an image given as a red[],
> green[] and blue[] array into a bitmap file. The output is the following:
The code you pasted looks fine to me. I suspect you have a bug
elsewhere in your code that is corrupting the heap, causing a memory
allocation or (more likely) free() call by fclose() to crash.
Try running your code with Valgrind.
- Kevin.
Kasper Dupont
>
> Hi,
>
> I have a strange problem getting a SIGSEGV from a fclose(). It does not
> happen for every set of data but I have a set of data where it happens
> always.
Probably related to this:
http://www.daimi.au.dk/~kasperd/comp.os.linux.development.faq.html#SIGSEGV
--
Kasper Dupont -- der bruger for meget tid på usenet.
For sending spam use mailto:aaa...@daimi.au.dk
Hvem er fjenden i Aalborg?