Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:5 as permitted sender) client-ip=2620:137:e000::3:5;
Date:   Mon, 2 Oct 2023 08:04:24 +0200
From:   Christoph Hellwig <hch@lst.de>
To:     Jerry Snitselaar <jsnitsel@redhat.com>
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Hannes Reinecke <hare@suse.de>,
        Chris Leech <cleech@redhat.com>,
        Christoph Hellwig <hch@lst.de>,
        Rasesh Mody <rmody@marvell.com>,
        Ariel Elior <aelior@marvell.com>,
        Sudarsana Kalluru <skalluru@marvell.com>,
        Manish Chopra <manishc@marvell.com>,
        Nilesh Javali <njavali@marvell.com>,
        Manish Rangankar <mrangankar@marvell.com>,
        John Meneghini <jmeneghi@redhat.com>,
        Lee Duncan <lduncan@suse.com>,
        Mike Christie <michael.christie@oracle.com>,
        Hannes Reinecke <hare@kernel.org>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/3] cnic,bnx2,bnx2x: use UIO_MEM_DMA_COHERENT
Message-ID: <20231002060424.GA781@lst.de>
References: <20230929170023.1020032-1-cleech@redhat.com> <20230929170023.1020032-4-cleech@redhat.com> <2023093055-gotten-astronomy-a98b@gregkh> <ZRhmqBRNUB3AfLv/@rhel-developer-toolbox> <2023093002-unlighted-ragged-c6e1@gregkh> <e0360d8f-6d36-4178-9069-d633d9b7031d@suse.de> <2023100114-flatware-mourner-3fed@gregkh> <7pq4ptas5wpcxd3v4p7iwvgoj7vrpta6aqfppqmuoccpk4mg5t@fwxm3apjkez3>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <7pq4ptas5wpcxd3v4p7iwvgoj7vrpta6aqfppqmuoccpk4mg5t@fwxm3apjkez3>
User-Agent: Mutt/1.5.17 (2007-11-01)
Precedence: bulk

On Sun, Oct 01, 2023 at 07:22:36AM -0700, Jerry Snitselaar wrote:
> Changes last year to the dma-mapping api to no longer allow __GFP_COMP,
> in particular these two (from the e529d3507a93 dma-mapping pull for
> 6.2):

That's complete BS.  The driver was broken since day 1 and always
ignored the DMA API requirement to never try to grab the page from the
dma coherent allocation because you generally speaking can't.  It just
happened to accidentally work the trivial dma coherent allocator that
is used on x86.