[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: gc issues
From: |
Mikael Djurfeldt |
Subject: |
Re: gc issues |
Date: |
19 Sep 2000 14:24:40 +0200 |
User-agent: |
Gnus/5.0807 (Gnus v5.8.7) Emacs/20.7 |
Dirk Herrmann <address@hidden> writes:
> Still, am I right saying that due to the multiprossor issue we _need_
> locking to obtain a new cell?
(Again, I still haven't read this thread, so I might be saying
irrelevant things.)
We are not going to lock a mutex for each cell allocation. That would
mean substantial overhead. The major reason for introducing the
clusters into the GC was to be able to avoid locking a mutex when
allocating.
I'm not sure exactly which problem seems to force us to lock a mutex,
but if it is the memory coherence problem, we'll simply only support
multiprocessor machines where memory coherence is guaranteed. This
holds for the i386 platform.
- Re: gc issues, (continued)
- Re: gc issues, Michael Livshin, 2000/09/13
- Re: gc issues, Michael Livshin, 2000/09/14
- Re: gc issues, Jim Blandy, 2000/09/14
- Re: gc issues, Michael Livshin, 2000/09/14
- Re: gc issues, Dirk Herrmann, 2000/09/14
- Re: gc issues, Michael Livshin, 2000/09/16
- Re: gc issues, Dirk Herrmann, 2000/09/19
- Re: gc issues, Michael Livshin, 2000/09/19
- Re: gc issues, Mikael Djurfeldt, 2000/09/19
- Re: gc issues, Michael Livshin, 2000/09/19
- Re: gc issues,
Mikael Djurfeldt <=
- Re: gc issues, Mikael Djurfeldt, 2000/09/19