[PATCH 2/2] uacce: add uacce module

Greg Kroah-Hartman gregkh at linuxfoundation.org
Fri Aug 16 00:12:25 AEST 2019


On Wed, Aug 14, 2019 at 05:34:25PM +0800, Zhangfei Gao wrote:
> diff --git a/include/uapi/misc/uacce.h b/include/uapi/misc/uacce.h
> new file mode 100644
> index 0000000..44a0a5d
> --- /dev/null
> +++ b/include/uapi/misc/uacce.h
> @@ -0,0 +1,44 @@
> +/* SPDX-License-Identifier: GPL-2.0-or-later */
> +#ifndef _UAPIUUACCE_H
> +#define _UAPIUUACCE_H
> +
> +#include <linux/types.h>
> +#include <linux/ioctl.h>
> +
> +#define UACCE_CLASS_NAME	"uacce"

Why is this in a uapi file?

> +#define UACCE_DEV_ATTRS		"attrs"

Same here.

> +#define UACCE_CMD_SHARE_SVAS	_IO('W', 0)
> +#define UACCE_CMD_START		_IO('W', 1)
> +#define UACCE_CMD_GET_SS_DMA	_IOR('W', 2, unsigned long)
> +
> +/**
> + * UACCE Device Attributes:
> + *
> + * NOIOMMU: the device has no IOMMU support
> + *	can do share sva, but no map to the dev
> + * PASID: the device has IOMMU which support PASID setting
> + *	can do share sva, mapped to dev per process
> + * FAULT_FROM_DEV: the device has IOMMU which can do page fault request
> + *	no need for share sva, should be used with PASID
> + * SVA: full function device
> + * SHARE_DOMAIN: no PASID, can do share sva only for one process and the kernel
> + */
> +#define UACCE_DEV_NOIOMMU		(1 << 0)
> +#define UACCE_DEV_PASID			(1 << 1)
> +#define UACCE_DEV_FAULT_FROM_DEV	(1 << 2)
> +#define UACCE_DEV_SVA		(UACCE_DEV_PASID | UACCE_DEV_FAULT_FROM_DEV)
> +#define UACCE_DEV_SHARE_DOMAIN	(0)
> +
> +#define UACCE_API_VER_NOIOMMU_SUBFIX	"_noiommu"
> +
> +#define UACCE_QFR_NA ((unsigned long)-1)
> +enum uacce_qfrt {
> +	UACCE_QFRT_MMIO = 0,	/* device mmio region */
> +	UACCE_QFRT_DKO,		/* device kernel-only */
> +	UACCE_QFRT_DUS,		/* device user share */
> +	UACCE_QFRT_SS,		/* static share memory */
> +	UACCE_QFRT_MAX,

These enums need to be explicitly set, as per the documentation,
otherwise they could be messed up when dealing with odd compilers.

> +};
> +#define UACCE_QFRT_INVALID UACCE_QFRT_MAX

Why not just use INVALID instead of MAX?

thanks,

greg k-h


More information about the Linux-accelerators mailing list