[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210726070030.GB9035@lst.de>
Date: Mon, 26 Jul 2021 09:00:30 +0200
From: Christoph Hellwig <hch@....de>
To: Atish Patra <atish.patra@....com>
Cc: linux-kernel@...r.kernel.org, Albert Ou <aou@...s.berkeley.edu>,
Christoph Hellwig <hch@....de>, devicetree@...r.kernel.org,
Dmitry Vyukov <dvyukov@...gle.com>,
Frank Rowand <frowand.list@...il.com>,
Guo Ren <guoren@...ux.alibaba.com>,
iommu@...ts.linux-foundation.org, linux-riscv@...ts.infradead.org,
Marek Szyprowski <m.szyprowski@...sung.com>,
Palmer Dabbelt <palmer@...belt.com>,
Paul Walmsley <paul.walmsley@...ive.com>,
Rob Herring <robh+dt@...nel.org>,
Robin Murphy <robin.murphy@....com>,
Tobias Klauser <tklauser@...tanz.ch>
Subject: Re: [RFC 3/5] dma-mapping: Enable global non-coherent pool support
for RISC-V
On Fri, Jul 23, 2021 at 02:40:29PM -0700, Atish Patra wrote:
> Currently, linux,dma-default is used to reserve a global non-coherent pool
> to allocate memory for dma operations. This can be useful for RISC-V as
> well as the ISA specification doesn't specify a method to modify PMA
> attributes or page table entries to define non-cacheable area yet.
> A non-cacheable memory window is an alternate options for vendors to
> support non-coherent devices.
Please explain why you do not want to use the simply non-cachable
window support using arch_dma_set_uncached as used by mips, niops2 and
xtensa.
> +static int __dma_init_global_coherent(phys_addr_t phys_addr, dma_addr_t device_addr, size_t size)
> {
> struct dma_coherent_mem *mem;
>
> - mem = dma_init_coherent_memory(phys_addr, phys_addr, size, true);
> + if (phys_addr == device_addr)
> + mem = dma_init_coherent_memory(phys_addr, device_addr, size, true);
> + else
> + mem = dma_init_coherent_memory(phys_addr, device_addr, size, false);
Nak. The phys_addr != device_addr support is goign away. This needs
to be filled in using dma-ranges property hanging of the struct device.
Powered by blists - more mailing lists