Re: Why do flush page cache twice when change TT's cache attribute

[Scott Fang <donnie.fang@xxxxxxxxx>]

yes, flush page twice cause system slow in some CPU.
Meanwhile, set_memory_uc/wc doesn't only do flush page cache, but also change page table attribute, so your solution seems to have some problem.

2012/3/20 Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>

On Tue, Mar 20, 2012 at 10:15:02AM +0800, Scott Fang wrote:
> Can I do the optimization like:
>
>     if (ttm->caching_state == tt_cached)
>  -        drm_clflush_pages(ttm->pages, ttm->num_pages);
> +        for (i = 0; i < ttm->num_pages; ++i)
> +           if (PageHighMem(ttm->pages[i]))
> +                drm_clflush_pages(&ttm->pages[i], 1);
>
> only do flush cache when high memory and leave the linear memory flush in
> function set_memory_uc/wc?

So what are you trying to solve? I mean one way to fix this
is to do:

>
> 2012/3/20 Jerome Glisse <j.glisse@xxxxxxxxx>
>
> > On Mon, 2012-03-19 at 23:11 +0800, Scott Fang wrote:
> > > In function ttm_tt_set_caching
> > > ,,,,,,,
> > >
> > >     if (ttm->caching_state == tt_cached)
> > >         drm_clflush_pages(ttm->pages, ttm->num_pages);

             goto out;

> > >
> > >     for (i = 0; i < ttm->num_pages; ++i) {
> > >         cur_page = ttm->pages[i];
> > >         if (likely(cur_page != NULL)) {
> > >             ret = ttm_tt_set_page_caching(cur_page,
> > >                               ttm->caching_state,
> > >                               c_state);
> > >             if (unlikely(ret != 0))
> > >                 goto out_err;
> > >         }
> > >     }
> >

out:

> > >     ttm->caching_state = c_state;
> > >
> > >     return 0;

Is the problem with calling page change twice making the machine slow?

_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/dri-devel