[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20210414133750.000042b1@Huawei.com>
Date: Wed, 14 Apr 2021 13:37:50 +0100
From: Jonathan Cameron <Jonathan.Cameron@...wei.com>
To: Ben Widawsky <ben.widawsky@...el.com>
CC: <linux-cxl@...r.kernel.org>, <linux-pci@...r.kernel.org>,
<linux-acpi@...r.kernel.org>, <ira.weiny@...el.com>,
<vishal.l.verma@...el.com>, <alison.schofield@...el.com>,
<dan.j.williams@...el.com>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] cxl/mem: Clarify UAPI documentation
On Tue, 13 Apr 2021 07:32:47 -0700
Ben Widawsky <ben.widawsky@...el.com> wrote:
> Signed-off-by: Ben Widawsky <ben.widawsky@...el.com>
Sensible update.
Acked-by: Jonathan Cameron <Jonathan.Cameron@...wei.com>
> ---
> include/uapi/linux/cxl_mem.h | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/include/uapi/linux/cxl_mem.h b/include/uapi/linux/cxl_mem.h
> index f6e8a005b113..8dd516ddb098 100644
> --- a/include/uapi/linux/cxl_mem.h
> +++ b/include/uapi/linux/cxl_mem.h
> @@ -81,6 +81,13 @@ static const struct {
> * - @size_in = -1
> * - @size_out = 0
> *
> + * Commands which have a variable length input likely have a minimum length.
> + * This is not enforced by the UAPI, and therefore might look like the command
> + * succeeded when sending too small of an input payload. Caution should be taken
> + * by checking the @cxl_send_command.retval for such cases. For commands with a
> + * variable length output, the caller is free to consume as little or as much as
> + * they want.
> + *
> * See struct cxl_mem_query_commands.
> */
> struct cxl_command_info {
Powered by blists - more mailing lists