[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-arm-kernel
Subject:    Re: ioremap non-contiguous phys to contiguous virt?
From:       Marc Singer <elf () buici ! com>
Date:       2004-10-19 20:21:12
Message-ID: 20041019202112.GA27283 () buici ! com
[Download RAW message or body]

On Tue, Oct 19, 2004 at 09:50:27AM -0400, Nicolas Pitre wrote:
> On Mon, 18 Oct 2004, Marc Singer wrote:
> 
> > On Mon, Oct 18, 2004 at 05:43:40PM -0400, Nicolas Pitre wrote:
> > > Anyway, even if your combined flash chips are mapped contiguous, the MTD 
> > > code will still detect 2 different MTD devices anyway, right? 
> 
> [replying to myself]: Not exactly.  They end up in the same MTD logical 
> device.  My brain was out last night it seems.
> 
> > That is what the cfi_ code does.  The cfi_intelext_point code will
> > 'point' every chip and then return a single memory region for all of
> > them.  It works correctly when the chips are all contiguously mapped.
> > However, it is buggy because even though it may find that the chips
> > are discontiguous, it doesn't recognize that fact and disable the
> > point functions.
> 
> Then it's a bug.

Right.  I may post a fix to it.  While it is a bug, it is also kinda
stupid to do what I did.  The extra 8MiB of pages means that, say
about 2000?, descriptors are allocated that point to nothing.

-------------------------------------------------------------------
Subscription options: http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm-kernel
FAQ:       http://www.arm.linux.org.uk/armlinux/mlfaq.php
Etiquette: http://www.arm.linux.org.uk/armlinux/mletiquette.php
[prev in list] [next in list] [prev in thread] [next in thread]