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;
MIME-Version: 1.0
References: <20200708164936.9340-1-nsaenzjulienne@suse.de>
In-Reply-To: <20200708164936.9340-1-nsaenzjulienne@suse.de>
From:   Amit Pundir <amit.pundir@linaro.org>
Date:   Tue, 21 Jul 2020 12:51:50 +0530
Message-ID: <CAMi1Hd35tRM=cnmzwX=SDgu-OoXi1Xj+twFkoULaVZBbTpe6sw@mail.gmail.com>
Subject: Re: [PATCH] dma-pool: Do not allocate pool memory from CMA
To:     Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Cc:     Christoph Hellwig <hch@lst.de>,
        Marek Szyprowski <m.szyprowski@samsung.com>,
        Robin Murphy <robin.murphy@arm.com>,
        David Rientjes <rientjes@google.com>,
        linux-rpi-kernel@lists.infradead.org, jeremy.linton@arm.com,
        iommu@lists.linux-foundation.org,
        lkml <linux-kernel@vger.kernel.org>,
        John Stultz <john.stultz@linaro.org>,
        Sumit Semwal <sumit.semwal@linaro.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, 8 Jul 2020 at 22:43, Nicolas Saenz Julienne
<nsaenzjulienne@suse.de> wrote:
>
> There is no guarantee to CMA's placement, so allocating a zone specific
> atomic pool from CMA might return memory from a completely different
> memory zone. So stop using it.
>
> Fixes: c84dc6e68a1d ("dma-pool: add additional coherent pools to map to gfp mask")

Hi Nicolas,

I see a boot regression with this commit d9765e41d8e9 "dma-pool:
Do not allocate pool memory from CMA" on my Xiaomi Poco F1
(Qcom sdm845) phone running v5.8-rc6. I can't boot past the
bootloader splash screen with this patch.

Phone boots fine if I revert this patch. I carry only one out of tree
dts patch https://lkml.org/lkml/2020/6/25/52. And since this is a stock
phone, I don't have access to serial/dmesg logs until I boot to AOSP
(adb) shell.

Any thoughts as to what might be going wrong here? I'd be happy to
help debug things. For what it's worth, I don't see this regression on
other two sdm845 devices (db845c and Pixel 3) I tested on.

Regards,
Amit Pundir

> Reported-by: Jeremy Linton <jeremy.linton@arm.com>
> Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
> ---
>
> An more costly alternative would be adding an option to
> dma_alloc_from_contiguous() so it fails when the allocation doesn't fall
> in a specific zone.
>
>  kernel/dma/pool.c | 11 ++---------
>  1 file changed, 2 insertions(+), 9 deletions(-)
>
> diff --git a/kernel/dma/pool.c b/kernel/dma/pool.c
> index 8cfa01243ed2..4bc1c46ae6ef 100644
> --- a/kernel/dma/pool.c
> +++ b/kernel/dma/pool.c
> @@ -6,7 +6,6 @@
>  #include <linux/debugfs.h>
>  #include <linux/dma-direct.h>
>  #include <linux/dma-noncoherent.h>
> -#include <linux/dma-contiguous.h>
>  #include <linux/init.h>
>  #include <linux/genalloc.h>
>  #include <linux/set_memory.h>
> @@ -69,12 +68,7 @@ static int atomic_pool_expand(struct gen_pool *pool, size_t pool_size,
>
>         do {
>                 pool_size = 1 << (PAGE_SHIFT + order);
> -
> -               if (dev_get_cma_area(NULL))
> -                       page = dma_alloc_from_contiguous(NULL, 1 << order,
> -                                                        order, false);
> -               else
> -                       page = alloc_pages(gfp, order);
> +               page = alloc_pages(gfp, order);
>         } while (!page && order-- > 0);
>         if (!page)
>                 goto out;
> @@ -118,8 +112,7 @@ static int atomic_pool_expand(struct gen_pool *pool, size_t pool_size,
>         dma_common_free_remap(addr, pool_size);
>  #endif
>  free_page: __maybe_unused
> -       if (!dma_release_from_contiguous(NULL, page, 1 << order))
> -               __free_pages(page, order);
> +       __free_pages(page, order);
>  out:
>         return ret;
>  }
> --
> 2.27.0
>