[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0e55585d-d6d7-b9ac-aed5-8df56895e209@amd.com>
Date: Fri, 2 Jun 2023 13:43:40 +0200
From: Christian König <christian.koenig@....com>
To: Chia-I Wu <olvaffe@...il.com>, dri-devel@...ts.freedesktop.org
Cc: Alex Deucher <alexander.deucher@....com>,
"Pan, Xinhui" <Xinhui.Pan@....com>,
David Airlie <airlied@...il.com>,
Daniel Vetter <daniel@...ll.ch>,
Felix Kuehling <Felix.Kuehling@....com>,
Philip Yang <Philip.Yang@....com>,
Hawking Zhang <Hawking.Zhang@....com>,
Mukul Joshi <mukul.joshi@....com>,
Danijel Slivka <danijel.slivka@....com>,
Lang Yu <Lang.Yu@....com>,
Marek Olšák <marek.olsak@....com>,
Jammy Zhou <Jammy.Zhou@....com>, amd-gfx@...ts.freedesktop.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] amdgpu: validate offset_in_bo of drm_amdgpu_gem_va
Am 02.06.23 um 00:44 schrieb Chia-I Wu:
> This is motivated by OOB access in amdgpu_vm_update_range when
> offset_in_bo+map_size overflows.
>
> v2: keep the validations in amdgpu_vm_bo_map
> v3: add the validations to amdgpu_vm_bo_map/amdgpu_vm_bo_replace_map
> rather than to amdgpu_gem_va_ioctl
>
> Fixes: 9f7eb5367d00 ("drm/amdgpu: actually use the VM map parameters")
> Signed-off-by: Chia-I Wu <olvaffe@...il.com>
Reviewed-by: Christian König <christian.koenig@....com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 16 ++++++++--------
> 1 file changed, 8 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index 22f9a65ca0fc7..76d57bc7ac620 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -1434,14 +1434,14 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
> uint64_t eaddr;
>
> /* validate the parameters */
> - if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK ||
> - size == 0 || size & ~PAGE_MASK)
> + if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK || size & ~PAGE_MASK)
> + return -EINVAL;
> + if (saddr + size <= saddr || offset + size <= offset)
> return -EINVAL;
>
> /* make sure object fit at this offset */
> eaddr = saddr + size - 1;
> - if (saddr >= eaddr ||
> - (bo && offset + size > amdgpu_bo_size(bo)) ||
> + if ((bo && offset + size > amdgpu_bo_size(bo)) ||
> (eaddr >= adev->vm_manager.max_pfn << AMDGPU_GPU_PAGE_SHIFT))
> return -EINVAL;
>
> @@ -1500,14 +1500,14 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev,
> int r;
>
> /* validate the parameters */
> - if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK ||
> - size == 0 || size & ~PAGE_MASK)
> + if (saddr & ~PAGE_MASK || offset & ~PAGE_MASK || size & ~PAGE_MASK)
> + return -EINVAL;
> + if (saddr + size <= saddr || offset + size <= offset)
> return -EINVAL;
>
> /* make sure object fit at this offset */
> eaddr = saddr + size - 1;
> - if (saddr >= eaddr ||
> - (bo && offset + size > amdgpu_bo_size(bo)) ||
> + if ((bo && offset + size > amdgpu_bo_size(bo)) ||
> (eaddr >= adev->vm_manager.max_pfn << AMDGPU_GPU_PAGE_SHIFT))
> return -EINVAL;
>
Powered by blists - more mailing lists