> I noticed that a few lines that referenced the vm_offset member of the
> vm_area_struct structure in fbmem.c were giving me errors when i
> attempted to compile kernel 2.3.25. Here is my bugfix patch (which works
> on my machine with ver limited testing) that turns all references to the
> vm_offset member of the aforementioned structure in the code to the
> vm_pgoff member of the structure.
>
> Ian Baird
> <ibaird@umr.edu>
<flame>
Gentlemen, could we _stop_ this "I don't understand what I've done, but
compiler is happy now" sort of patches? Trivial check would show that
vm_pgoff is not in bytes, it's in pages (AFAICS it's a part of LFS patch).
Result of this voodoo debugging is the code that passes compiler without
warnings but is badly broken. If you don't understand what happens - take
time to figure it out. Sheesh... Shutting the compiler up != fixing. It
hides the problem and makes it very hard to fix - I've seen similar beasts
that stayed around for 3-4 years.
</flame>
Sorry for that rant, but... I've seen it one too many times.
> --- linux-2.3.25/drivers/video/fbmem.c Thu Oct 28 16:34:46 1999
> +++ linux/drivers/video/fbmem.c Tue Nov 2 20:16:22 1999
> @@ -505,9 +505,9 @@
> len = (start & ~PAGE_MASK)+fix.smem_len;
> start &= PAGE_MASK;
> len = (len+~PAGE_MASK) & PAGE_MASK;
+PAGE_CACHE_SIZE-1) >> PAGE_CACHE_SHIFT;
> - if (vma->vm_offset >= len) {
> + if (vma->vm_pgoff >= len) {
> /* memory mapped io */
> - vma->vm_offset -= len;
> + vma->vm_pgoff -= len;
> fb->fb_get_var(&var, PROC_CONSOLE(info), info);
> if (var.accel_flags)
> return -EINVAL;
and so on...
Al "/* shut gcc up */ is a shootable offense" Viro
-- All that blue light from Orthanc at night? That was Saruman, trying to moderate news.admin.palantir-abuse.sightings. Mike Andrews in the Monastery
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/