[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240305140543.GB3659@lst.de>
Date: Tue, 5 Mar 2024 15:05:43 +0100
From: Christoph Hellwig <hch@....de>
To: Robin Murphy <robin.murphy@....com>
Cc: Petr Tesařík <petr@...arici.cz>,
Michael Kelley <mhklinux@...look.com>,
Christoph Hellwig <hch@....de>, Will Deacon <will@...nel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Petr Tesarik <petr.tesarik1@...wei-partners.com>,
"kernel-team@...roid.com" <kernel-team@...roid.com>,
"iommu@...ts.linux.dev" <iommu@...ts.linux.dev>,
Marek Szyprowski <m.szyprowski@...sung.com>,
Dexuan Cui <decui@...rosoft.com>,
Nicolin Chen <nicolinc@...dia.com>
Subject: Re: [PATCH v5 6/6] swiotlb: Remove pointless stride adjustment for
allocations >= PAGE_SIZE
On Mon, Mar 04, 2024 at 01:37:56PM +0000, Robin Murphy wrote:
> FWIW I'm also starting to wonder about getting rid of the alloc_size
> argument and just have SWIOTLB round the end address up to alloc_align_mask
> itself as part of all these calculations. Seems like it could potentially
> end up a little simpler, maybe?
Yes, that does sound simpler and most importantly harder to get wrong..
Powered by blists - more mailing lists