Re: [RFC PATCH] x86: explicit call to mmiotrace in do_page_fault()

[Pekka Paalanen <pq@iki.fi>]

On Sat, 9 Feb 2008 20:56:07 +0200
"Pekka Enberg" <penberg@cs.helsinki.fi> wrote:

> Hi Pekka,
> 
> On Feb 9, 2008 8:23 PM, Pekka Paalanen <pq@iki.fi> wrote:
> > By default it reserves 32 MB/cpu of relay buffers during module init.
> > It's this big based on experimentation with the nvidia proprietary driver.
> > There is no other way than unloading the module to free the buffers, or
> > even to clear them. This means that every time you start the user space
> > logger application, it gets old data first. Also poking the module
> > parameters via /sys is not allowed.
> 
> So how is mmiotrace different from, say, blktrace that has an ioctl()
> to setup/teardown the relay channels (for mmiotrace it can be a
> "status" file that does the setup/teardown when written some
> appropriate values)?

No difference. I just have to do it at some point. I can't work on this
every day and when it comes to kernel, I'm a complete newbie. So things
will take time. If people want that I fix this before mmiotrace goes into
mainline, then so be it.

blktrace, thanks for the hint, I should take a look at that.

-- 
Pekka Paalanen
http://www.iki.fi/pq/