Re: [PATCH+discussion] symlink recursion

David Mosberger (davidm@napali.hpl.hp.com)
Wed, 19 Jun 2002 12:14:01 -0700


>>>>> On Wed, 19 Jun 2002 11:55:23 -0700 (PDT), Linus Torvalds <torvalds@transmeta.com> said:

Linus> Yes. But did you look at the stack frames of those things?
Linus> It's something like 16 bytes for ext2_follow_link (it just
Linus> calls directly back to the VFS layer), 20 bytes for
Linus> vfs_follow_link(), and 56 for link_path_walk.

Linux> ...

Linus> But there are other numbers, like performance (sometimes
Linus> linearizing recursion loses, sometimes it wins), or somebody
Linus> doing the math on ia-64 and showing that the 100 bytes/level
Linus> on x86 is actually more like 2kB on ia-64 and totally
Linus> unacceptable.

Just to avoid starting false rumours: on ia-64, I see the following
(2.4.18, with gcc3.1):

- ext2_follow_link(): 16 bytes/frame
- vfs_follow_link(): 56 bytes/frame
- link_path_walk(): 128 bytes/frame
--------------------- ---------------
total: 200 bytes/frame

Just about in line with what you'd expect given that registers are 64 bits.

--david
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/