Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Zero copy between ISR, kernel and User

2 views
Skip to first unread message

Wink Saville

unread,
Sep 27, 2006, 12:10:07 AM9/27/06
to
Hello,

I would like to allow the transferring of data between ISR's, kernel and
user code, without requiring copying. I envision allocating buffers in
the kernel and then mapping them so that they appear at the same
addresses to all code, and never being swapped out of memory.

Is this feasible for all supported Linux architectures and is there
existing code that someone could point me towards?

Regards,

Wink Saville

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Stephen Hemminger

unread,
Sep 27, 2006, 3:20:10 PM9/27/06
to
On Tue, 26 Sep 2006 21:01:45 -0700
Wink Saville <wi...@saville.com> wrote:

> Hello,
>
> I would like to allow the transferring of data between ISR's, kernel and
> user code, without requiring copying. I envision allocating buffers in
> the kernel and then mapping them so that they appear at the same
> addresses to all code, and never being swapped out of memory.
>
> Is this feasible for all supported Linux architectures and is there
> existing code that someone could point me towards?
>
> Regards,
>
> Wink Saville
>

Your better off having application mmap a device, then transfer
the data to there. Something like AF_PACKET.

--
Stephen Hemminger <shemm...@osdl.org>

Wink Saville

unread,
Sep 28, 2006, 10:50:10 AM9/28/06
to
Stephen Hemminger wrote:
> On Tue, 26 Sep 2006 21:01:45 -0700
> Wink Saville <wi...@saville.com> wrote:
>
>
>> Hello,
>>
>> I would like to allow the transferring of data between ISR's, kernel and
>> user code, without requiring copying. I envision allocating buffers in
>> the kernel and then mapping them so that they appear at the same
>> addresses to all code, and never being swapped out of memory.
>>
>> Is this feasible for all supported Linux architectures and is there
>> existing code that someone could point me towards?
>>
>> Regards,
>>
>> Wink Saville
>>
>>
>
> Your better off having application mmap a device, then transfer
> the data to there. Something like AF_PACKET.
>
>
Is there some reason a kernel module can't mmap first, for instance I
assume display drivers might do that? One of the reasons I need it to
done in the kernel first is that data could come from the device or
other entities before the application is running.

But I will study AF_PACKET handling.

Thanks,

Wink

0 new messages