Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <b07c93bc7cb71a32091794cd97f7c702c34539da.camel@redhat.com>
Subject: Re: [PATCH 3/4] drm/nouveau: stop using is_swiotlb_active
From:   Lyude Paul <lyude@redhat.com>
To:     Christoph Hellwig <hch@lst.de>, Juergen Gross <jgross@suse.com>,
        Stefano Stabellini <sstabellini@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        Dave Hansen <dave.hansen@linux.intel.com>, x86@kernel.org,
        "H. Peter Anvin" <hpa@zytor.com>, Ben Skeggs <bskeggs@redhat.com>,
        Karol Herbst <kherbst@redhat.com>
Cc:     xen-devel@lists.xenproject.org, iommu@lists.linux.dev,
        linux-kernel@vger.kernel.org, nouveau@lists.freedesktop.org
Date:   Thu, 18 May 2023 16:30:49 -0400
In-Reply-To: <20230518134253.909623-4-hch@lst.de>
References: <20230518134253.909623-1-hch@lst.de>
         <20230518134253.909623-4-hch@lst.de>
Organization: Red Hat Inc.
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4 (3.44.4-3.fc36) 
MIME-Version: 1.0
Precedence: bulk

Reviewed-by: Lyude Paul <lyude@redhat.com>

Thanks for getting to this!

On Thu, 2023-05-18 at 15:42 +0200, Christoph Hellwig wrote:
> Drivers have no business looking into dma-mapping internals and check
> what backend is used.  Unfortunstely the DRM core is still broken and
> tries to do plain page allocations instead of using DMA API allocators
> by default and uses various bandaids on when to use dma_alloc_coherent.
>=20
> Switch nouveau to use the same (broken) scheme as amdgpu and radeon
> to remove the last driver user of is_swiotlb_active.
>=20
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
>  drivers/gpu/drm/nouveau/nouveau_ttm.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
>=20
> diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouv=
eau/nouveau_ttm.c
> index 1469a88910e45d..486f39f31a38df 100644
> --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
> +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
> @@ -24,9 +24,9 @@
>   */
> =20
>  #include <linux/limits.h>
> -#include <linux/swiotlb.h>
> =20
>  #include <drm/ttm/ttm_range_manager.h>
> +#include <drm/drm_cache.h>
> =20
>  #include "nouveau_drv.h"
>  #include "nouveau_gem.h"
> @@ -265,7 +265,6 @@ nouveau_ttm_init(struct nouveau_drm *drm)
>  	struct nvkm_pci *pci =3D device->pci;
>  	struct nvif_mmu *mmu =3D &drm->client.mmu;
>  	struct drm_device *dev =3D drm->dev;
> -	bool need_swiotlb =3D false;
>  	int typei, ret;
> =20
>  	ret =3D nouveau_ttm_init_host(drm, 0);
> @@ -300,13 +299,10 @@ nouveau_ttm_init(struct nouveau_drm *drm)
>  		drm->agp.cma =3D pci->agp.cma;
>  	}
> =20
> -#if IS_ENABLED(CONFIG_SWIOTLB) && IS_ENABLED(CONFIG_X86)
> -	need_swiotlb =3D is_swiotlb_active(dev->dev);
> -#endif
> -
>  	ret =3D ttm_device_init(&drm->ttm.bdev, &nouveau_bo_driver, drm->dev->d=
ev,
>  				  dev->anon_inode->i_mapping,
> -				  dev->vma_offset_manager, need_swiotlb,
> +				  dev->vma_offset_manager,
> +				  drm_need_swiotlb(drm->client.mmu.dmabits),
>  				  drm->client.mmu.dmabits <=3D 32);
>  	if (ret) {
>  		NV_ERROR(drm, "error initialising bo driver, %d\n", ret);

--=20
Cheers,
 Lyude Paul (she/her)
 Software Engineer at Red Hat