Linux MIPS Architecture development
 help / color / mirror / Atom feed
From: Ralf Baechle <ralf@linux-mips.org>
To: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Cc: linux-mips@linux-mips.org
Subject: Re: ZONE_DMA on MIPS
Date: Mon, 19 Mar 2007 22:20:31 +0000	[thread overview]
Message-ID: <20070319222031.GB8707@linux-mips.org> (raw)
In-Reply-To: <20070320.013608.103777227.anemo@mba.ocn.ne.jp>

On Tue, Mar 20, 2007 at 01:36:08AM +0900, Atsushi Nemoto wrote:

> On Mon, 19 Mar 2007 15:48:21 +0000, Ralf Baechle <ralf@linux-mips.org> wrote:
> > > Are there any other platforms requires special DMA zone?
> > 
> > Qemu supports more or less the full PC braind^Wheritage, including the
> > good old too-old floppy controller.  IP22 supports Indigo 2 systems
> > which have EISA support, so we only want ZONE_DMA if EISA is enabled.
> > For a bunch of other systems ZONE_DMA may be required to support b0rked
> > PCI cards that only support like 31-bit DMA addresses or even less.
> 
> Hmm... So do you think making ZONE_DMA customizable for each platform
> (or user configurable) would have some sense?

It's probably reasonable to do something like:

config GENERIC_ISA_DMA
	bool
	select ZONE_DMA

I don't think we should expose such deep technical details to the Kconfig
user.

> For these legacy(?) PCI cards, we can check if it works or not by
> pci_set_dma_mask(), at least.

The Linux approach to handling such broken cards is not very gentle.  A
card that can't live with normal allocations, will always use bounce
ZONE_NORMAL buffers.

Obviously the message to the designers of PCI cards is to get their stuff
done right and not rely on OS kludgery to fix hardware shortcomings.

And while at it, we may want to think about ZONE_DMA32.  At least the
BCM1250 needs it for configs with more than 1GB memory due to it's
sparse memory map.

  Ralf

  reply	other threads:[~2007-03-19 22:22 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-03-19 15:09 ZONE_DMA on MIPS Atsushi Nemoto
2007-03-19 15:32 ` Sergei Shtylyov
2007-03-19 15:47   ` Atsushi Nemoto
2007-03-19 15:50     ` Sergei Shtylyov
2007-03-19 15:48 ` Ralf Baechle
2007-03-19 16:36   ` Atsushi Nemoto
2007-03-19 22:20     ` Ralf Baechle [this message]
2007-03-20 14:10       ` Atsushi Nemoto
2007-03-20 14:48         ` Ralf Baechle

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20070319222031.GB8707@linux-mips.org \
    --to=ralf@linux-mips.org \
    --cc=anemo@mba.ocn.ne.jp \
    --cc=linux-mips@linux-mips.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox