Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Matthew Brost <matthew.brost@intel.com>
To: Brian Welty <brian.welty@intel.com>
Cc: <intel-xe@lists.freedesktop.org>
Subject: Re: [PATCH] drm/xe: Use distinct error code in VM_CREATE for device fault mode checks
Date: Wed, 24 Apr 2024 03:20:34 +0000	[thread overview]
Message-ID: <Zih6go8i/0CJ9TIU@DUT025-TGLU.fm.intel.com> (raw)
In-Reply-To: <20240423234707.6587-1-brian.welty@intel.com>

On Tue, Apr 23, 2024 at 04:47:07PM -0700, Brian Welty wrote:
> vm_create_ioctl doesn't allow creating new VM in fault_mode while existing
> VMs are in non-fault mode, and vice-versa.
> This is not necessarily the user doing anything wrong if others are sharing
> the system. Instead of returning -EINVAL which suggests the user provided a
> bad argument, return distinct error code of -EBUSY for this class of errors
> so the user knows they can retry the vm_create when system is idle.
> 

We (Thomas, myself, UMD teams) have been discussing error codes for VM
bind and exec IOCTLs but haven't considered the error codes in the
create IOCTLs. We likely need to do a complete scrub of these IOCTLs
too.

Looking at xe_vm_create_ioctl, after this change it looks correct to me.
I think if we all agree on this change (and other error codes) we should
also update the kerenl doc with all the possible returns.

It looks like:
-EINVAL: Various bad input
-EBUSY: VM in fault mode or non-fault mode present
-ENOMEM: kmalloc or device memory allocation failed

The only change I could possibly see is for extensions return
-EOPNOTSUPP.

Let's get a few others input here.

Matt

> Signed-off-by: Brian Welty <brian.welty@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_vm.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
> index 85d6f359142d..e8628a982db6 100644
> --- a/drivers/gpu/drm/xe/xe_vm.c
> +++ b/drivers/gpu/drm/xe/xe_vm.c
> @@ -1840,16 +1840,16 @@ int xe_vm_create_ioctl(struct drm_device *dev, void *data,
>  			 args->flags & DRM_XE_VM_CREATE_FLAG_FAULT_MODE))
>  		return -EINVAL;
>  
> +	if (XE_IOCTL_DBG(xe, args->extensions))
> +		return -EINVAL;
> +
>  	if (XE_IOCTL_DBG(xe, args->flags & DRM_XE_VM_CREATE_FLAG_FAULT_MODE &&
>  			 xe_device_in_non_fault_mode(xe)))
> -		return -EINVAL;
> +		return -EBUSY;
>  
>  	if (XE_IOCTL_DBG(xe, !(args->flags & DRM_XE_VM_CREATE_FLAG_FAULT_MODE) &&
>  			 xe_device_in_fault_mode(xe)))
> -		return -EINVAL;
> -
> -	if (XE_IOCTL_DBG(xe, args->extensions))
> -		return -EINVAL;
> +		return -EBUSY;
>  
>  	if (args->flags & DRM_XE_VM_CREATE_FLAG_SCRATCH_PAGE)
>  		flags |= XE_VM_FLAG_SCRATCH_PAGE;
> -- 
> 2.43.0
> 

  parent reply	other threads:[~2024-04-24  3:21 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-23 23:47 [PATCH] drm/xe: Use distinct error code in VM_CREATE for device fault mode checks Brian Welty
2024-04-23 23:55 ` ✓ CI.Patch_applied: success for " Patchwork
2024-04-23 23:55 ` ✓ CI.checkpatch: " Patchwork
2024-04-23 23:56 ` ✓ CI.KUnit: " Patchwork
2024-04-24  0:16 ` ✓ CI.Build: " Patchwork
2024-04-24  0:19 ` ✓ CI.Hooks: " Patchwork
2024-04-24  0:20 ` ✓ CI.checksparse: " Patchwork
2024-04-24  0:50 ` ✓ CI.BAT: " Patchwork
2024-04-24  3:20 ` Matthew Brost [this message]
2024-05-14 14:32 ` [PATCH] " Ghimiray, Himal Prasad
2024-05-17  4:18 ` Matthew Brost

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=Zih6go8i/0CJ9TIU@DUT025-TGLU.fm.intel.com \
    --to=matthew.brost@intel.com \
    --cc=brian.welty@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    /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