Re: [RFC PATCH v2 2/3] accel: add dedicated minor for accelerator devices

From: Jeffrey Hugo <quic_jhugo@quicinc.com>
To: Oded Gabbay <ogabbay@kernel.org>,
	David Airlie <airlied@gmail.com>, Daniel Vetter <daniel@ffwll.ch>,
	Arnd Bergmann <arnd@arndb.de>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	<linux-kernel@vger.kernel.org>, <dri-devel@lists.freedesktop.org>,
	Jason Gunthorpe <jgg@nvidia.com>,
	John Hubbard <jhubbard@nvidia.com>,
	Alex Deucher <alexander.deucher@amd.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>,
	Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>,
	Jiho Chu <jiho.chu@samsung.com>,
	Christoph Hellwig <hch@infradead.org>,
	Thomas Zimmermann <tzimmermann@suse.de>,
	Kevin Hilman <khilman@baylibre.com>,
	Yuji Ishikawa <yuji2.ishikawa@toshiba.co.jp>,
	Maciej Kwapulinski <maciej.kwapulinski@linux.intel.com>,
	Jagan Teki <jagan@amarulasolutions.com>
Subject: Re: [RFC PATCH v2 2/3] accel: add dedicated minor for accelerator devices
Date: Wed, 2 Nov 2022 15:17:41 -0600	[thread overview]
Message-ID: <7654d9c0-c181-ae6b-96ee-349f20f24b18@quicinc.com> (raw)
In-Reply-To: <20221102203405.1797491-3-ogabbay@kernel.org>

On 11/2/2022 2:34 PM, Oded Gabbay wrote:
> @@ -24,16 +33,6 @@ static char *accel_devnode(struct device *dev, umode_t *mode)
> 
>   static CLASS_ATTR_STRING(accel_version, 0444, "accel 1.0.0 20221018");
> 
> -/**
> - * accel_sysfs_init - initialize sysfs helpers
> - *
> - * This is used to create the ACCEL class, which is the implicit parent of any
> - * other top-level ACCEL sysfs objects.
> - *
> - * You must call accel_sysfs_destroy() to release the allocated resources.
> - *
> - * Return: 0 on success, negative error code on failure.
> - */

Why are we removing this?

>   static int accel_sysfs_init(void)
>   {
>   	int err;
> @@ -54,11 +53,6 @@ static int accel_sysfs_init(void)
>   	return 0;
>   }
> 
> -/**
> - * accel_sysfs_destroy - destroys ACCEL class
> - *
> - * Destroy the ACCEL device class.
> - */

Again, why remove this?  Adding it in one patch than immediately 
removing it in the next patch seems wasteful.

>   static void accel_sysfs_destroy(void)
>   {
>   	if (IS_ERR_OR_NULL(accel_class))
> @@ -68,11 +62,185 @@ static void accel_sysfs_destroy(void)
>   	accel_class = NULL;
>   }
> 
> +static void accel_minor_release(struct drm_minor *minor)
> +{
> +	drm_dev_put(minor->dev);
> +}
> +
> +/**
> + * accel_open - open method for ACCEL file
> + * @inode: device inode
> + * @filp: file pointer.
> + *
> + * This function must be used by drivers as their &file_operations.open method.

Feels like it would be helpful to have an accel version of 
DEFINE_DRM_GEM_FOPS() which helps accel drivers to get this right

> + * It looks up the correct ACCEL device and instantiates all the per-file
> + * resources for it. It also calls the &drm_driver.open driver callback.
> + *
> + * Return: 0 on success or negative errno value on failure.
> + */
> +int accel_open(struct inode *inode, struct file *filp)
> +{
> +	struct drm_device *dev;
> +	struct drm_minor *minor;
> +	int retcode;
> +
> +	minor = accel_minor_acquire(iminor(inode));
> +	if (IS_ERR(minor))
> +		return PTR_ERR(minor);
> +
> +	dev = minor->dev;
> +
> +	atomic_fetch_inc(&dev->open_count);
> +
> +	/* share address_space across all char-devs of a single device */
> +	filp->f_mapping = dev->anon_inode->i_mapping;
> +
> +	retcode = drm_open_helper(filp, minor);
> +	if (retcode)
> +		goto err_undo;
> +
> +	return 0;
> +
> +err_undo:
> +	atomic_dec(&dev->open_count);
> +	accel_minor_release(minor);
> +	return retcode;
> +}
> +EXPORT_SYMBOL_GPL(accel_open);