From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ville =?iso-8859-1?Q?Syrj=E4l=E4?= Subject: Re: radeon, apertures & memory mapping Date: Mon, 14 Mar 2005 02:08:35 +0200 Message-ID: <20050314000835.GA20393@sci.fi> References: <1110705646.14684.126.camel@gaston> <20050313103936.GA11002@sci.fi> <1110715499.14684.132.camel@gaston> <9e473391050313081937cde207@mail.gmail.com> <1110750553.5787.155.camel@gaston> <9e47339105031314101c89e50e@mail.gmail.com> <1110752401.19810.177.camel@gaston> <9e47339105031315002a444f00@mail.gmail.com> <20050313232735.GA19781@sci.fi> <1110757699.5787.196.camel@gaston> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline In-Reply-To: <1110757699.5787.196.camel@gaston> Sender: dri-devel-admin@lists.sourceforge.net Errors-To: dri-devel-admin@lists.sourceforge.net List-Unsubscribe: , List-Id: List-Post: List-Help: List-Subscribe: , List-Archive: Content-Type: text/plain; charset="iso-8859-1" To: Benjamin Herrenschmidt Cc: Jon Smirl , Jon Smirl , dri-devel@lists.sourceforge.net, Linux Fbdev development list , xorg@lists.freedesktop.org On Mon, Mar 14, 2005 at 10:48:19AM +1100, Benjamin Herrenschmidt wrote: >=20 > > > That shouldn't matter the page brought in would be for a speculativ= e > > > read and never accessed. It should just fall out of the cache and n= ot > > > be written back. There is only one cachable mapping. In this model > > > writes are always followed by a flush before telling the GPU to acc= ess > > > the memory that has just been written. > >=20 > > What about this scenario? > >=20 > > Speculative read -> AGP master writes new data -> CPU has invalid dat= a in=20 > > cache :( >=20 > First, we must be very careful with AGP master writes. I don't know if > we do a lot of them currently, but I know a collection of north bridges > that do not support them. I don't think "normal" drivers do them at all. I did experiment with=20 DirectFB at one point and had it place all offscreen surfaces to AGP=20 memory. It worked really well on my hardware (G400 + VIA KT133=20 northbridge). I also tried it with PCI transfers and that too worked but=20 was naturally slower. I'd like to make DirectFB use AGP again since 32MB=20 of video memory isn't always enough. > (Which is interesting, that means that if we want to copy something out > of video memory, we can't write it to AGP memory and then read it, we > need to actually do the blit from the CPU, good to know for our memory > manager. That also means that we have a problem if the video memory > isn't entirely accessible by the CPU ...) What about PCI master writes? Are there bridges that don't support even=20 those? > That's something we should probably think about doing properly: Have a > list of AGP "issues" (errata ?) bits that are communicated by the AGP > host driver to the DRM. >=20 > At least all the early Apple AGP bridges don't do writes, and I remembe= r > we have trouble with a few x86 ones as well. There are also issues when > a single AGP burst crosses a page boundary, and other things like that. :( --=20 Ville Syrj=E4l=E4 syrjala@sci.fi http://www.sci.fi/~syrjala/ ------------------------------------------------------- SF email is sponsored by - The IT Product Guide Read honest & candid reviews on hundreds of IT Products from real users. Discover which products truly live up to the hype. Start reading now. http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click --