Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Christian König" <christian.koenig@amd.com>
To: "Thomas Hellström" <thomas.hellstrom@linux.intel.com>,
	intel-xe@lists.freedesktop.org
Cc: Matthew Brost <matthew.brost@intel.com>, dri-devel@lists.freedesktop.org
Subject: Re: [RFC PATCH] drm/ttm: Forward -ENOSPC to drivers requesting it
Date: Tue, 3 Sep 2024 17:14:14 +0200	[thread overview]
Message-ID: <60e8310d-5f8d-41d9-b889-58bc7cbaa395@amd.com> (raw)
In-Reply-To: <20240903133849.17119-1-thomas.hellstrom@linux.intel.com>

Am 03.09.24 um 15:38 schrieb Thomas Hellström:
> Some user-space APIs distinguison between graphics memory OOMs and
> system (host) memory OOMs. To aid UMDs in determining the type of
> OOM, allow forwarding the ENOSPC from resource managers to drivers
> on calls to ttm_bo_validate().
>
> Cc: Christian König <christian.koenig@amd.com>
> Cc: Matthew Brost <matthew.brost@intel.com>
> Cc: dri-devel@lists.freedesktop.org
> Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>

Ah yes that was on my TODO list as well.

> ---
>   drivers/gpu/drm/ttm/ttm_bo.c | 2 +-
>   include/drm/ttm/ttm_bo.h     | 3 +++
>   2 files changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
> index dd867b5e744c..d9a320dc8130 100644
> --- a/drivers/gpu/drm/ttm/ttm_bo.c
> +++ b/drivers/gpu/drm/ttm/ttm_bo.c
> @@ -836,7 +836,7 @@ int ttm_bo_validate(struct ttm_buffer_object *bo,
>   	} while (ret && force_space);
>   
>   	/* For backward compatibility with userspace */
> -	if (ret == -ENOSPC)
> +	if (ret == -ENOSPC && !ctx->forward_enospc)

Mhm, couldn't we put that into the bdev? I would rather like to keep the 
UAPI consistent at least per driver.

Christian.

>   		return -ENOMEM;
>   
>   	/*
> diff --git a/include/drm/ttm/ttm_bo.h b/include/drm/ttm/ttm_bo.h
> index 5804408815be..d3e12318d336 100644
> --- a/include/drm/ttm/ttm_bo.h
> +++ b/include/drm/ttm/ttm_bo.h
> @@ -174,6 +174,8 @@ struct ttm_bo_kmap_obj {
>    * BOs share the same reservation object.
>    * @force_alloc: Don't check the memory account during suspend or CPU page
>    * faults. Should only be used by TTM internally.
> + * @forward_enospc: Don't translate -ENOSPC errors from resource managers to
> + * -ENOMEM, but forward them to the driver.
>    * @resv: Reservation object to allow reserved evictions with.
>    * @bytes_moved: Statistics on how many bytes have been moved.
>    *
> @@ -185,6 +187,7 @@ struct ttm_operation_ctx {
>   	bool no_wait_gpu;
>   	bool gfp_retry_mayfail;
>   	bool allow_res_evict;
> +	bool forward_enospc;
>   	bool force_alloc;
>   	struct dma_resv *resv;
>   	uint64_t bytes_moved;


  reply	other threads:[~2024-09-03 15:14 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-09-03 13:38 [RFC PATCH] drm/ttm: Forward -ENOSPC to drivers requesting it Thomas Hellström
2024-09-03 15:14 ` Christian König [this message]
2024-09-03 15:22   ` Thomas Hellström
2024-09-03 15:24     ` Christian König
2024-09-03 18:41 ` ✓ CI.Patch_applied: success for " Patchwork
2024-09-03 18:41 ` ✓ CI.checkpatch: " Patchwork
2024-09-03 18:42 ` ✓ CI.KUnit: " Patchwork
2024-09-03 18:54 ` ✓ CI.Build: " Patchwork
2024-09-03 18:56 ` ✓ CI.Hooks: " Patchwork
2024-09-03 18:58 ` ✗ CI.checksparse: warning " Patchwork
2024-09-03 19:17 ` ✓ CI.BAT: success " Patchwork
2024-09-03 20:47 ` ✗ CI.FULL: failure " Patchwork

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=60e8310d-5f8d-41d9-b889-58bc7cbaa395@amd.com \
    --to=christian.koenig@amd.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.brost@intel.com \
    --cc=thomas.hellstrom@linux.intel.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