From: Arjan van de Ven <arjan@infradead.org>
To: Jeff Garzik <jeff@garzik.org>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>,
Ingo Molnar <mingo@elte.hu>, Thomas Gleixner <tglx@linutronix.de>,
linux-kernel <linux-kernel@vger.kernel.org>,
"H. Peter Anvin" <hpa@zytor.com>,
"David S. Miller" <davem@davemloft.net>,
Linus Torvalds <torvalds@linux-foundation.org>
Subject: Re: [patch] x86, voyager: fix ioremap_nocache()
Date: Mon, 28 Apr 2008 07:10:06 -0700 [thread overview]
Message-ID: <20080428071006.5ce27533@laptopd505.fenrus.org> (raw)
In-Reply-To: <4815009C.2010809@garzik.org>
On Sun, 27 Apr 2008 18:39:24 -0400
Jeff Garzik <jeff@garzik.org> wrote:
> James Bottomley wrote:
> > Here's another piece of the x86 API that's designed to be cached.
> > The dma_declare_coherent_memory() usually represents behind bridge
> > memory that's fully participatory in the coherence model.
> >
> > Making it uncached damages the utility of this memory because doing
> > cacheline sized burst cycles when needed to it is far faster than
> > individual byte/word/quad writes.
> >
> > Signed-off-by: James Bottomley
> > <James.Bottomley@HansenPartnership.com>
> >
> > ---
> >
> > diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c
> > index 388b113..df83ffd 100644
> > --- a/arch/x86/kernel/pci-dma.c
> > +++ b/arch/x86/kernel/pci-dma.c
> > @@ -214,7 +214,7 @@ int dma_declare_coherent_memory(struct device
> > *dev, dma_addr_t bus_addr,
> > /* FIXME: this routine just ignores
> > DMA_MEMORY_INCLUDES_CHILDREN */
> > - mem_base = ioremap(bus_addr, size);
> > + mem_base = ioremap_cache(bus_addr, size);
> > if (!mem_base)
> > goto out;
this patch patch is likely broken on x86; or rather, anyone who uses it is...
thinking you can find cache coherent memory on a PCI or similar bus that is actually
cachable... keep dreaming. (for now; there's talk about extending PCI)
> I would rather change drivers to use ioremap_nocache(), and leave the
> API as-is.
>
> Isn't there Yet More Breakage in lib/iomap.c, given these new
> semantics?
>
> if (flags & IORESOURCE_MEM) {
> if (flags & IORESOURCE_CACHEABLE)
> return ioremap(start, len);
> return ioremap_nocache(start, len);
> }
>
> Any driver using pci_iomap() (libata, and others) is affected.
only if you map ROM's. Anything but ROMs you cannot set IORESOURCE_CACHEABLE
on... since PCI MMIO memory isn't cache coherent per se. (it's cache coherent on
x86 by virtue of being uncachable ;-)
Thankfully Linux doesn't do that.
> I disagree with this semantics change. A number of code places _and
> drivers_ GET IT RIGHT, and these are all broken now?
Can you list one that gets it actually right ?
(cachable pretty much means: "the cpu is the only one changing it AND there is no side effect of
reading or writing")
--
If you want to reach me at my work email, use arjan@linux.intel.com
For development, discussion and tips for power savings,
visit http://www.lesswatts.org
next prev parent reply other threads:[~2008-04-28 14:10 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-27 20:51 Breakage caused by unreviewed patch in x86 tree James Bottomley
2008-04-27 20:53 ` David Miller
2008-04-27 21:48 ` [patch] x86, voyager: fix ioremap_nocache() Ingo Molnar
2008-04-27 22:05 ` James Bottomley
2008-04-27 22:36 ` Willy Tarreau
2008-04-27 22:41 ` Ingo Molnar
2008-04-27 23:18 ` Ingo Molnar
2008-04-27 23:31 ` David Miller
2008-04-28 0:31 ` Rik van Riel
2008-04-28 0:45 ` Al Viro
2008-04-28 0:52 ` H. Peter Anvin
2008-04-28 9:01 ` Alan Cox
2008-04-28 9:17 ` David Miller
2008-04-28 9:48 ` Adrian Bunk
2008-04-28 11:50 ` Ingo Molnar
2008-04-28 6:10 ` Christoph Hellwig
2008-04-28 16:55 ` H. Peter Anvin
2008-04-27 22:34 ` James Bottomley
2008-04-27 22:39 ` Jeff Garzik
2008-04-27 22:44 ` H. Peter Anvin
2008-04-27 22:46 ` David Miller
2008-04-27 22:52 ` H. Peter Anvin
2008-04-27 22:58 ` David Miller
2008-04-27 23:04 ` H. Peter Anvin
2008-04-30 20:35 ` Eric W. Biederman
2008-04-27 23:34 ` Jeff Garzik
2008-04-27 23:39 ` H. Peter Anvin
2008-04-27 22:53 ` Jeff Garzik
2008-04-27 22:56 ` H. Peter Anvin
2008-04-27 22:59 ` David Miller
2008-04-27 23:02 ` Jeff Garzik
2008-04-27 23:14 ` Arjan van de Ven
2008-04-27 23:01 ` Arjan van de Ven
2008-04-30 21:44 ` James Bottomley
2008-04-30 22:39 ` H. Peter Anvin
2008-04-27 23:01 ` Thomas Gleixner
2008-04-28 14:10 ` Arjan van de Ven [this message]
2008-04-28 14:29 ` James Bottomley
2008-04-28 15:07 ` Arjan van de Ven
2008-04-28 19:59 ` H. Peter Anvin
2008-04-27 22:00 ` Breakage caused by unreviewed patch in x86 tree H. Peter Anvin
2008-04-27 22:10 ` James Bottomley
2008-04-27 22:13 ` H. Peter Anvin
2008-04-27 22:18 ` James Bottomley
2008-04-27 22:31 ` H. Peter Anvin
2008-04-27 22:58 ` Arjan van de Ven
2008-04-27 23:00 ` David Miller
2008-04-27 23:07 ` Arjan van de Ven
2008-04-27 23:03 ` James Bottomley
2008-04-27 23:11 ` Arjan van de Ven
2008-04-27 23:17 ` H. Peter Anvin
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=20080428071006.5ce27533@laptopd505.fenrus.org \
--to=arjan@infradead.org \
--cc=James.Bottomley@HansenPartnership.com \
--cc=davem@davemloft.net \
--cc=hpa@zytor.com \
--cc=jeff@garzik.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.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
all inboxes | Powered by JetHome®