On Tuesday 21 August 2012, Catalin Marinas wrote:
quoted
quoted
+asmlinkage long sys_mmap(unsigned long addr, unsigned long len,
+ unsigned long prot, unsigned long flags,
+ unsigned long fd, off_t off)
+{
+ if (offset_in_page(off) != 0)
+ return -EINVAL;
+
+ return sys_mmap_pgoff(addr, len, prot, flags, fd, off >> PAGE_SHIFT);
+}
I think
#define sys_mmap sys_mmap_pgoff
There are slightly different semantics with the last argument of
sys_mmap() which takes a byte offset. The sys_mmap_pgoff() function
takes the offset shifted by PAGE_SHIFT (which is the same as sys_mmap2).
Looking at the other architectures, it makes sense to use a generic
sys_mmap() implementation similar to the one above (or the ia-64, seems
to be the most complete).
Why that? The generic sys_mmap_pgoff was specifically added so new architectures
could just use that instead of having their own wrappers, see f8b72560.
Arnd