[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f0ab2aa5-ca23-fa40-6147-43fc1f385bc8@arm.com>
Date: Thu, 28 Apr 2022 15:56:24 +0100
From: Robin Murphy <robin.murphy@....com>
To: Christoph Hellwig <hch@...radead.org>
Cc: Tianyu Lan <ltykernel@...il.com>, m.szyprowski@...sung.com,
michael.h.kelley@...rosoft.com, kys@...rosoft.com,
parri.andrea@...il.com, thomas.lendacky@....com,
wei.liu@...nel.org, Andi Kleen <ak@...ux.intel.com>,
Tianyu Lan <Tianyu.Lan@...rosoft.com>,
linux-hyperv@...r.kernel.org, konrad.wilk@...cle.com,
linux-kernel@...r.kernel.org, kirill.shutemov@...el.com,
iommu@...ts.linux-foundation.org, andi.kleen@...el.com,
brijesh.singh@....com, vkuznets@...hat.com, hch@....de
Subject: Re: [RFC PATCH 1/2] swiotlb: Split up single swiotlb lock
On 2022-04-28 15:45, Christoph Hellwig wrote:
> On Thu, Apr 28, 2022 at 03:44:36PM +0100, Robin Murphy wrote:
>> Rather than introduce this extra level of allocator complexity, how about
>> just dividing up the initial SWIOTLB allocation into multiple io_tlb_mem
>> instances?
>
> Yeah. We're almost done removing all knowledge of swiotlb from drivers,
> so the very last thing I want is an interface that allows a driver to
> allocate a per-device buffer.
FWIW I'd already started thinking about having a distinct io_tlb_mem for
non-coherent devices where vaddr is made non-cacheable to avoid the
hassle of keeping the arch_dma_sync_* calls lined up, so I'm certainly
in favour of bringing in a bit more flexibility at this level :)
Robin.
Powered by blists - more mailing lists