[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20240827072946.GB13080@lst.de>
Date: Tue, 27 Aug 2024 09:29:46 +0200
From: Christoph Hellwig <hch@....de>
To: Segher Boessenkool <segher@...nel.crashing.org>
Cc: LEROY Christophe <christophe.leroy2@...soprasteria.com>,
Christoph Hellwig <hch@....de>,
Christian Lamparter <christian.lamparter@....uni-stuttgart.de>,
Benjamin Herrenschmidt <benh@...nel.crashing.org>,
Paul Mackerras <paulus@...ba.org>,
Michael Ellerman <mpe@...erman.id.au>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"linuxppc-dev@...ts.ozlabs.org" <linuxppc-dev@...ts.ozlabs.org>,
Stan Johnson <userm57@...oo.com>,
Finn Thain <fthain@...ux-m68k.org>
Subject: Re: [PATCH v2] powerpc: warn on emulation of dcbz instruction in
kernel mode
On Sat, Aug 24, 2024 at 12:17:57PM -0500, Segher Boessenkool wrote:
> > Are these functions also used on DMA coherent memory ?
>
> Most won't show up high on most profiles, heh. Which you already
> can see from the problem not being attacked yet: if it was so obviously
> a problem, some people would have wanted to do something about it :-)
Most drivers try to avoid coherent allocations in the fast path if
they can. Another good option for Christians problem would be
to switch the the dmaengine driver to use dma_alloc_pages - it doesn't
actually need uncached memory as far as I can, dma_alloc_coherent is
just the only API we had to allocate guaranteed DMAable memory for most
of Linux's existence.
Powered by blists - more mailing lists