From: Felix Kuehling <felix.kuehling@amd.com>
To: "Christian König" <ckoenig.leichtzumerken@gmail.com>,
alex.sierra@amd.com, amd-gfx@lists.freedesktop.org
Subject: Re: [PATCH] drm/amdgpu: stop allocating PDs/PTs with the eviction lock held
Date: Fri, 14 Feb 2020 16:12:34 -0500 [thread overview]
Message-ID: <7f0c6e3e-be75-3fa8-a5e7-d8ae6bf0dba5@amd.com> (raw)
In-Reply-To: <20200212151426.14197-1-christian.koenig@amd.com>
Now you allow eviction of page tables while you allocate page tables.
Isn't the whole point of the eviction lock to prevent page table
evictions while manipulating page tables?
Or does this only apply to PTE invalidations which never allocated
memory? Is that the only case that doesn't reserve page tables?
Regards,
Felix
On 2020-02-12 10:14 a.m., Christian König wrote:
> We need to make sure to not allocate PDs/PTs while holding
> the eviction lock or otherwise we will run into lock inversion
> in the MM as soon as we enable the MMU notifier.
>
> Signed-off-by: Christian König <christian.koenig@amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 31 +++++++++++++++++++++-----
> 1 file changed, 25 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index 77c400675b79..e7ab0c1e2793 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -897,27 +897,42 @@ static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
> struct amdgpu_vm_pt *entry = cursor->entry;
> struct amdgpu_bo_param bp;
> struct amdgpu_bo *pt;
> + bool need_entries;
> int r;
>
> - if (cursor->level < AMDGPU_VM_PTB && !entry->entries) {
> + need_entries = cursor->level < AMDGPU_VM_PTB && !entry->entries;
> + if (!need_entries && entry->base.bo)
> + return 0;
> +
> + /* We need to make sure that we don't allocate PDs/PTs while holding the
> + * eviction lock or we run into lock recursion in the MM.
> + */
> + amdgpu_vm_eviction_unlock(vm);
> +
> + if (need_entries) {
> unsigned num_entries;
>
> num_entries = amdgpu_vm_num_entries(adev, cursor->level);
> entry->entries = kvmalloc_array(num_entries,
> sizeof(*entry->entries),
> GFP_KERNEL | __GFP_ZERO);
> - if (!entry->entries)
> - return -ENOMEM;
> + if (!entry->entries) {
> + r = -ENOMEM;
> + goto error_lock;
> + }
> }
>
> - if (entry->base.bo)
> - return 0;
> + if (entry->base.bo) {
> + r = 0;
> + goto error_lock;
> + }
>
> amdgpu_vm_bo_param(adev, vm, cursor->level, direct, &bp);
>
> r = amdgpu_bo_create(adev, &bp, &pt);
> + amdgpu_vm_eviction_lock(vm);
> if (r)
> - return r;
> + goto error_free_pt;
>
> /* Keep a reference to the root directory to avoid
> * freeing them up in the wrong order.
> @@ -936,6 +951,10 @@ static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
> amdgpu_bo_unref(&pt);
> entry->base.bo = NULL;
> return r;
> +
> +error_lock:
> + amdgpu_vm_eviction_lock(vm);
> + return r;
> }
>
> /**
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
next prev parent reply other threads:[~2020-02-14 21:12 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-12 15:14 [PATCH] drm/amdgpu: stop allocating PDs/PTs with the eviction lock held Christian König
2020-02-14 21:12 ` Felix Kuehling [this message]
2020-02-27 15:28 ` Christian König
2020-02-27 18:43 ` Felix Kuehling
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=7f0c6e3e-be75-3fa8-a5e7-d8ae6bf0dba5@amd.com \
--to=felix.kuehling@amd.com \
--cc=alex.sierra@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=ckoenig.leichtzumerken@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).