Date: Thu, 5 Mar 1998 17:10:05 -0500 (EST) From: "Ron G. Minnich" <rminnich@Sarnoff.COM> To: Terry Lambert <tlambert@primenet.com> Cc: toor@dyson.iquest.net, hackers@FreeBSD.ORG Subject: Re: vm architecture of freebsd. Message-ID: <Pine.SUN.3.91.980305170133.4194C-100000@terra> In-Reply-To: <199803051932.MAA02034@usr06.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Here's a specific and I hope interesting question for this list.
suppose we have:
msync with MS_INVALIDATE
msync WITHOUT MS_INVALIDATE
can the vm system communicate this request to the underlying VFS as two
different types of requests? I am pretty sure it can not. This is 3.0-SMP
by the way.
in vm_map_clean:
if (current->protection & VM_PROT_WRITE) {
vm_object_page_clean(object,
OFF_TO_IDX(offset),
OFF_TO_IDX(offset + size),
(syncio||invalidate)?1:0, TRUE);
so syncio and invalidate mean ...
void
vm_object_page_clean(object, start, end, syncio, lockflag)
vm_object_t object;
vm_pindex_t start;
vm_pindex_t end;
boolean_t syncio;
boolean_t lockflag;
so at some point invalidate is used to tell the vnodes underneath to do sync
io. But the info as to invalidate is basically lost. This is the type of
thing that can make life hard: information of use is disappearing as you walk
down the function call tree. It would be nice to keep that info there,
right down to the bottom of the food chain.
thanks
ron
Ron Minnich |Java: an operating-system-independent,
rminnich@sarnoff.com |architecture-independent programming language
(609)-734-3120 |for Windows/95 and Windows/NT on the Pentium
ftp://ftp.sarnoff.com/pub/mnfs/www/docs/cluster.html
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.SUN.3.91.980305170133.4194C-100000>