[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1163029281.28571.767.camel@localhost.localdomain>
Date: Thu, 09 Nov 2006 10:41:21 +1100
From: Benjamin Herrenschmidt <benh@...nel.crashing.org>
To: Russell King <rmk+lkml@....linux.org.uk>
Cc: Muli Ben-Yehuda <muli@...ibm.com>,
linux-input@...ey.karlin.mff.cuni.cz,
Linux Kernel list <linux-kernel@...r.kernel.org>,
"David S. Miller" <davem@...emloft.net>,
Paul Mackerras <paulus@...ba.org>,
Anton Blanchard <anton@...ba.org>, Greg KH <greg@...ah.com>
Subject: Re: DMA APIs gumble grumble
On Wed, 2006-11-08 at 22:56 +0000, Russell King wrote:
> On Wed, Nov 08, 2006 at 07:47:33PM +1100, Benjamin Herrenschmidt wrote:
> > Yes, I need multiple dma_ops for powerpc too
>
> Ditto for ARM.
Ok, so there is some interest in having the dma_ops in struct device
beyond powerpc.
I'll put together today a patch doing:
- add #include <asm/device.h> to linux/device.h
- add a device.h file in asm/* that does:
struct dev_sysdata {
};
- add a struct dev_sysdata sysdata; field to struct device
That patch alone is 0 overhead and allows archs to start adding things.
I'll then modify my pending patches for 2.6.20 to use that instead of my
current device_ext thing.
Is that ok with everybody for 2.6.20 ?
Then, we can do, in no special order:
- on x86, put the acpi data in there and remove firmware_data from
struct device
- on x86, m32, frv, put the dma_coherent_mem pointer in there too and
remove it from struct device
- you can use it on ARM to put your dma_operations pointer as I'm
doing in for powerpc
- x86 can do the same
etc...
Cheers,
Ben.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists