Received-SPF: pass (google.com: best guess record for domain of linux-crypto-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20200127165646.19806-1-andrew.smirnov@gmail.com>
 <20200127165646.19806-2-andrew.smirnov@gmail.com> <VI1PR0402MB3485FF5402B8C0FFF48FBF2298030@VI1PR0402MB3485.eurprd04.prod.outlook.com>
In-Reply-To: <VI1PR0402MB3485FF5402B8C0FFF48FBF2298030@VI1PR0402MB3485.eurprd04.prod.outlook.com>
From:   Andrey Smirnov <andrew.smirnov@gmail.com>
Date:   Sun, 15 Mar 2020 21:14:49 -0700
Message-ID: <CAHQ1cqGZP5RKTsc4+jikyPVggt-mGViRtKNvyOx9FGkYW9pgmg@mail.gmail.com>
Subject: Re: [PATCH v7 1/9] crypto: caam - allocate RNG instantiation
 descriptor with GFP_DMA
To:     Horia Geanta <horia.geanta@nxp.com>
Cc:     "linux-crypto@vger.kernel.org" <linux-crypto@vger.kernel.org>,
        Chris Healy <cphealy@gmail.com>,
        Lucas Stach <l.stach@pengutronix.de>,
        Herbert Xu <herbert@gondor.apana.org.au>,
        Iuliana Prodan <iuliana.prodan@nxp.com>,
        dl-linux-imx <linux-imx@nxp.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-crypto-owner@vger.kernel.org
Precedence: bulk

On Tue, Feb 4, 2020 at 6:08 AM Horia Geanta <horia.geanta@nxp.com> wrote:
>
> On 1/27/2020 6:57 PM, Andrey Smirnov wrote:
> > Be consistent with the rest of the codebase and use GFP_DMA when
> > allocating memory for a CAAM JR descriptor.
> >
> Please use GFP_DMA32 instead.
> Device is not limited to less than 32 bits of addressing
> in any of its incarnations.
>
> s/GFP_DMA/GFP_DMA32 should be performed throughout caam driver.
> (But of course, I wouldn't include this change in current patch series).
>

Hmm, I am triggering
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/mm/slub.c?h=v5.6-rc6#n1721
by using GFP_DMA32. AFAICT, GFP_DMA32 can't be used in SLUB/SLAB
allocated memory:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/mm/internal.h?h=v5.6-rc6#n32

I'll stick with GFP_DMA for now, unless you have a different preference.

Thanks,
Andrey Smirnov