Mail Archives: djgpp-workers/1998/07/28/04:07:35
On Mon, 27 Jul 1998, Martin Str|mberg wrote:
> And sync on my machine (Linux) doesn't take several seconds. Less than
> one, actually, and it's not idle.
One second is also a lot. You don't want to waste that much on each
fork/exec, do you?
> Eh! Of course you can distinguish between read and write cache. It's
> rather necessary to do so: you have to flush the write buffers if
> you're rebooting. Perhaps I didn't understand what you meant?
What I meant is that the cache doesn't maintain separate ``read'' and
``write'' buffers, AFAIK. It maintains them in a single pool. Which
is a sensible thing to do, since a buffer that was read could be
written to after that.
> How about "smartdrv/s"?
You mean `smartdrv /c', right? (/s only shows the status, IIRC.)
AFAIK, /c does the same as our `_flush_disk_cache' does. And the
problem is not to find a way to do this with SmartDrv, it's how to do
it with *any* type of cache.
- Raw text -