Dear Ilya Yanok,
Hi Marek,
[sorry for copying, forget to CC the list]
On Sat, Jun 16, 2012 at 2:29 AM, Marek Vasut marek.vasut@gmail.com wrote:
Hm, can't we just punch a hole in the MMU table at runtime instead of preallocating it like this?
It's allocated at runtime now, do you mean allocate it on demand? Good point, Probably we can malloc big enough block and make it uncached directly from dma_alloc_coherent(). Is it what you suggest?
Kind of ... I mean rather insert an entry into MMU table at runtime that says "this region is uncached". But that'd need some hack in the mallocator now that I think about it. It might not be as simple as I thought at first.
On the other hand, most MMUs allow you to allocate stuff with 4k density, which should be ok.
Also, what is this for? Can we not simply flush/invalidate the caches?
flush/invalidate can be racy for some hardware. Sometimes we need to write some field to DMA descriptor and then read another one. And because one cannot flush/invalidate individual bytes write/flush can destroy the field updated by hardware. (Well, we can invalidate/read before write/flush but that introduces a race).
But that's shitty hardware. Do you really need to do it? Where? I fixed similar issue in fec_mxc.c recently.
Regards, Ilya.
Best regards, Marek Vasut