Date: Wed, 17 Feb 1999 14:46:35 -0500 From: "AARON MARKS" <amarks@sarnoff.com> To: freebsd-hackers@FreeBSD.ORG Cc: Matthew Dillon <dillon@apollo.backplane.com> Subject: Re: Memory-Based VFS Question Message-ID: <36CB1C9B.77725962@sarnoff.com> References: <199902171908.MAA22366@usr07.primenet.com>
next in thread | previous in thread | raw e-mail | index | archive | help
I appologize for the ignorance, but:
Terry Lambert wrote:
>
> The generic code assumes a vnode as a backing store. If he does not
> have a vnode as a backing store, he can not use the generic code.
What do you mean using a vnode as a backing store? A file in my MFS has
a vnode allocated to it.
>
> Basically, if he could use it, since that's the default that's
> inherited from the default ops vector, it'd already be working,
> and he wouldn't be asking the question.
>
> He's asking, so ipso facto, he's not using a vnode as backing store.
>
> This makes sense, since he said he's writing a new MFS; he's probably
> managing the page allocation himself from the KVA space.
I think so. I allocate blocks of 8k memory and grow as necessary.
>
> To answer the question: pattern your code after the code in the file
> /sys/vm/device_pager.c's dev_pager_getpages()/dev_pager_putpages()
> code, and you should be OK.
Great. I'll take a look-see.
I have another question, though: I mapped my _{get,put}pages functions
(they're empty, just print to console and return 0) to the vop functions
but when I do a system file copy (via cp), the OS never calls my
{get,put}pages functions.
Thanks again,
-A.
--
Aaron J. Marks Communications and Computing Systems Lab
Assoc. Member Tech Staff Advanced Networks and Computation Group
amarks@sarnoff.com Sarnoff Corporation
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?36CB1C9B.77725962>