Messages in this thread | | | Date | Fri, 14 May 1999 12:04:40 +0200 | From | Pavel Machek <> | Subject | Re: PATCH: rewritten bdflush |
| |
Hi!
> > > There is one notable side effect: you don't have to run the user space > > > update anymore, bdflush wakes up every so often all by itself. I > > > > How about notebook users who have configured update to stop > > writing out buffers and spin down disks after some idle time? > > I think we might want to change the kernel's behaviour > to include the following things: > - don't flush if: > - there are less than X dirty buffers && > - we haven't recently read from the disk > - flush if: > - there are more than X dirty buffers || > - we issued a read from the disk more than > Y and less than Z seconds ago || > - the dirty buffer is FAR too old > (say, an hour or so)
This is policy. Policy does not belong in kernel. (I, for example, do not like hour limit - it is way too low. My atimes do not need to be written; system can run 2 hours on backup batery and will suspend to disk after that. Only chance for crash is kernel bug. And there are no bugs in stable kernel, are they?
[but I believe this is much better than current status]
Pavel
PS: So there's at least one system, which just can not crash. Why do you want to sync 1 hour old buffers to disk on it? Why do you want to sync 1 year old buffers to disk? You don't want.
-- The best software in life is free (not shareware)! Pavel GCM d? s-: !g p?:+ au- a--@ w+ v- C++@ UL+++ L++ N++ E++ W--- M- Y- R+
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |