Shared Memory

[Lennart Sorensen]

On Tue, Feb 01, 2005 at 10:27:51AM -0500, Henry Spencer wrote:
> While I haven't looked at the Linux code lately, disk scheduling (the
> original technical buzzphrase for this) has been done in Unix kernels for
> decades; this is not some new experimental idea. 

But Linux doesn't have all the features that unix has had for decades
yet.  They all get added slowly a bit at a time. :)

> As Lennart notes, since drives got smart enough to do their own bad-sector
> remapping, the kernel doesn't have completely reliable knowledge of the
> order of sectors.  And for some time now, its knowledge of the drive
> geometry has been an approximation to the truth at best.  But it can still
> do a respectable job... and it can maintain much longer queues than most
> of the drives supported until recently. 

It would make sense to help out in software of course.  And I know my
drives don't all have 255 heads and 63 sectors per track.  As if. :)

> Readahead too has been a standard feature of Unixoid systems for decades.

I suspect most drives also implement some read ahead caching using their
onboard cache.  I haven't read any actual specs on how the cache on
hardisks is actually managed.  Not even sure the drive makers publish
that information.

Lennart Sorensen
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml