On Tue, Aug 26, 2003 at 07:18:55PM -0700, Michael Snyder wrote:
Daniel Jacobowitz wrote:
This patch doesn't do anything particularly important. I just moved some
code from stop_wait_callback out to a new function. I thought I'd need it
for my next patch; I turned out not to, but it's still cleaner this way.
Also fixes the two small problems I asked Jeff about earlier today - an
extra call to lin_lwp_thread_alive and a missing delete_thread.
Is this OK?
There's a bit more here than code movement -- the new code is not
identical to the old, even allowing for the jjohnstn changes.
If you'll say a word or two about the differences, I expect I'll approve
them.
The differences are exactly those two. There were two copies of the
code which called delete_lwp, and one of them was missing
delete_thread; so I collapsed them together. And there was a block
which checked lin_lwp_thread_alive, now gone.
Oh, I changed the text of the first error message from "exited" to
"vanished" so that we could tell from the logs which case was used.