In article <1992Aug17.151311.29507@ods.com>, david@ods.com (David Engel) writes:
| william E Davidsen (davidsen@ariel.crd.GE.COM) wrote:

| :   Well, no, actually. Since the new version of the page is still valid
| : for all processes it would still be sharable. And it would not have to
| : be saved somewhere in the unlikely event that it gets paged out, because
| : the unmodified varsion is still perfectly valid.
| 
| No, it wouldn't be shareable.  When a process writes to a shared page,
| it gets its very own copy of it which is no longer shared.  There is a
| term for this that you probably already know, but others may not.  It's
| called Copy-On-Write.

  This is just what you don't want/need to do in this particular
circumstance. This is a very neat exception to the usual rules of shared
or not memory. This memory doesn't get set dirty when written by the
kernel (I'm saying that, what I mean is the user proc can't write it)
or copied on write because it is really still valid for all processes
which share it, and it doesn't need to be paged out, because the
original copy is also valid.

  The only case in which this rewriting of code can't be done is when an
unalligned address falls over a page boundary. In that case it stays
unfixed.

  I hope I made this clear this time around, the page is written by
{something other than the user process} and therefore it is equally
valid with or without the mod and can still be shared freely.

  The analogy for this is the jump table library, if you were to patch
on the fly, which systems in the past have done, the fixed version of a
procedure is added to the end of the lib, and then the jump table is
patched to call it. Processes sharing the jump table should not get
their own copy.

  I'm explaining these methods /not/saying we should use them. As long
as I don't have to use jump tables I can be happy, although I hope
people will distribute .a files so the end user can link them.
-- 
bill davidsen, GE Corp. R&D Center; Box 8; Schenectady NY 12345
    I admit that when I was in school I wrote COBOL. But I didn't compile.