Re: [RFC] [PATCH 6/6] Kprobes: Remove breakpoints from the copied pages

[Hugh Dickins <hugh@veritas.com>]

On Tue, 9 May 2006, Prasanna S Panchamukhi wrote:
> This patch removes the breakpoints if the pages read from the page
> cache contains breakpoints. If the pages containing the breakpoints
> is copied from the page cache, the copied image would also contain
> breakpoints in them. This could be a major problem for tools like
> tripwire etc and cause security concerns, hence must be prevented.
> This patch hooks up the actor routine, checks if the executable was
> a probed executable using the file inode and then replaces the
> breakpoints with the original opcodes in the copied image.

You've done a nice job of making the code look like kernel code
throughout, it's a much tidier patchset than many.

With that said... it looks to me like one of the scariest and
most inappropriate sets I can remember.  Getting the kernel to
connive in presenting an incoherent view of its pagecache:
I don't think we'd ever want that.

There's all kinds of things that could be said about the details
(your locking is often insufficient, for example); but there's a
lot going on, and it doesn't seem worth going through this line
by line, when the whole concept seems so unwelcome.

You've a big task to convince people that this is something the
Linux kernel will want: and perhaps you'll succeed - good luck.

But please approach what you're trying to do from userspace:
you can patch the binaries from there if you wish (but not on
my system, thanks).  Or perhaps you can patch it all into the
kernel via kprobes itself, but I wouldn't recommend it.

Hugh