Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Mon, 27 Sep 2021 15:30:47 +0100
From:   Christoph Hellwig <hch@infradead.org>
To:     Jason Gunthorpe <jgg@ziepe.ca>
Cc:     Cai Huoqing <caihuoqing@baidu.com>,
        Mustafa Ismail <mustafa.ismail@intel.com>,
        Shiraz Saleem <shiraz.saleem@intel.com>,
        Doug Ledford <dledford@redhat.com>, linux-rdma@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] RDMA/irdma: Use dma_alloc_coherent() instead of
 kmalloc/dma_map_single()
Message-ID: <YVHVlzvu58wJlA0t@infradead.org>
References: <20210926061124.335-1-caihuoqing@baidu.com>
 <20210927120235.GB3544071@ziepe.ca>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20210927120235.GB3544071@ziepe.ca>
Precedence: bulk

On Mon, Sep 27, 2021 at 09:02:35AM -0300, Jason Gunthorpe wrote:
> This I'm not sure about, I see lots of calls to dma_sync_single_* for
> this memory and it is not unconditionally true that using coherent
> memory is better than doing the cache flushes. It depends very much
> on the access pattern.
> 
> At the very least if you convert to coherent memory I expect to see
> the sync's removed too..

In general coherent memory actually is worse for not cache coherent
architectures when you hav chance, an should mkae no difference for
cache coherent ones.  So I'd like to see numbers here instead of a
claim.