2003-05-26 04:47:59

by Mika Penttilä

[permalink] [raw]
Subject: Re: [patch] cache flush bug in mm/filemap.c (all kernels >= 2.5.30(at least))

I don't think the flush_dcache_page thing is done almost anywhere in the
block/driver level right now. And we shouldn't be doing io reads to
pagecache pages with user mappings anyway normally. direct-io is a
different thing.

--Mika


David S. Miller wrote:

>On Fri, 2003-05-23 at 11:34, Russell King wrote:
>
>
>>So no, I don't think it is a device driver issue at all.
>>
>>DaveM?
>>
>>
>
>Oh yes, this part is. If you don't ensure this, everything
>breaks.
>
>At the end of an I/O operation, say to a page cache page, that
>data ought to be visible equally to a userspace vs. a kernel
>space mapping to that page.
>
>For example, this is why we use language about "cpu visibility" in the
>DMA api documentation and not "kernel cpu visibility" :-) And because
>PIO transfers are basically pseudo-DMA they need to make the same exact
>guarentees.
>
>If you've been living in a world where you didn't think this is
>necessary, I certainly feel bad for you :-)
>
>
>



2003-05-26 04:56:22

by David Miller

[permalink] [raw]
Subject: Re: [patch] cache flush bug in mm/filemap.c (all kernels >= 2.5.30(at least))

From: **UNKNOWN CHARSET** <[email protected]>
Date: Mon, 26 May 2003 08:07:10 +0300

I don't think the flush_dcache_page thing is done almost anywhere in the
block/driver level right now.

It isn't and it shouldn't :-)

And we shouldn't be doing io reads to pagecache pages with user
mappings anyway normally. direct-io is a different thing.

We are talking about the case where we are bringing in the
data for the first time, on the page cache lookup miss.