> From: Johannes Erdfelt <[EMAIL PROTECTED]>
> Date: Thu, 9 May 2002 18:49:08 -0400
> We use clear_bit and set_bit because it's atomic and makes sure we don't
> race with the HC updating the same data.
Check out the following e-mail, which I carefuly stowed away.
JE was included on the list of recipients.
------------------------------------------------------
From: Linus Torvalds <[EMAIL PROTECTED]>
To: Pete Zaitcev <[EMAIL PROTECTED]>
cc: Johannes Erdfelt <[EMAIL PROTECTED]>
Subject: Re: About atomicity of set_bit
In-Reply-To: <[EMAIL PROTECTED]>
Message-ID: <[EMAIL PROTECTED]>
On Mon, 29 Oct 2001, Pete Zaitcev wrote:
>
> Do you really think that set_bit is atomic against DMA masters?
> JE said you did, see below. I do not think it is right.
Heh. You're right. It _is_ atomic on my regular boxes, but not on UP. The
SMP-atomicity automatically gives DMA master atomicity on x86 too, but it
is certainly the case that that is not guaranteed on other architectures
anyway, even when they are SMP (although most sane architectures will work
in SMP).
I guess the right thing to do is to make up a new operation that is
DMA-safe on UP too..
Linus
_______________________________________________________________
Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: [EMAIL PROTECTED]
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
