Subject: Re: MAXBSIZE Was Re: bread() returns EINVAL during vfs mount
To: Yevgeny Binder <yev@freeshell.org>
From: Jonathan Stone <jonathan@Pescadero.dsg.stanford.edu>
List: tech-kern
Date: 09/06/2005 11:56:45
In message <421E82CF-ED77-4C6E-AFEB-AA01984F8A87@freeshell.org>,
Yevgeny Binder writes:
>On Aug 27, 2005, at 11:13 AM, Thor Lancelot Simon wrote:
>
>> You can hand device drivers I/O requests of up to MAXPHYS in size (in
>> practice, you can hand them larger requests, too, but don't do that).
>> See the FFS code for examples.
>
>I'm curious: Is MAXPHYS ever less than MAXBSIZE?
From (at least) 4.3BSD nearly 20 years ago, through (at least) 1998,
that'd be a bug which would trigger really nasty behaviour. (I did
run or very nearly ran, into that with the pmax port, back when pmaxes
used non-MI SCSI drivers).
Personally, I think it'd still be a bug, tho these days I don't know
who could say for sure.