Linux CXL
 help / color / mirror / Atom feed
From: Dan Williams <dan.j.williams@intel.com>
To: Vishal Verma <vishal.l.verma@intel.com>, <linux-cxl@vger.kernel.org>
Cc: Gregory Price <gregory.price@memverge.com>,
	Jonathan Cameron <Jonathan.Cameron@huawei.com>,
	Davidlohr Bueso <dave@stgolabs.net>,
	"Dan Williams" <dan.j.williams@intel.com>,
	Vishal Verma <vishal.l.verma@intel.com>, <nvdimm@lists.linux.dev>
Subject: RE: [PATCH ndctl 5/7] cxl/region: determine region type based on root decoder capability
Date: Tue, 7 Feb 2023 21:55:54 -0800	[thread overview]
Message-ID: <63e3396a17860_e3dae294d8@dwillia2-xfh.jf.intel.com.notmuch> (raw)
In-Reply-To: <20230120-vv-volatile-regions-v1-5-b42b21ee8d0b@intel.com>

Vishal Verma wrote:
> In the common case, root decoders are expected to be either pmem
> capable, or volatile capable, but not necessarily both simultaneously.
> If a decoder only has one of pmem or volatile capabilities,
> cxl-create-region should just infer the type of the region (pmem
> or ram) based on this capability.
> 
> Maintain the default behavior of cxl-create-region to choose type=pmem,
> but only as a fallback if the selected root decoder has multiple
> capabilities. If it is only capable of either pmem, or ram, then infer
> region type from this without requiring it to be specified explicitly.
> 
> Cc: Dan Williams <dan.j.williams@intel.com>
> Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
> ---
>  Documentation/cxl/cxl-create-region.txt |  3 ++-
>  cxl/region.c                            | 27 +++++++++++++++++++++++++++
>  2 files changed, 29 insertions(+), 1 deletion(-)
> 
> diff --git a/Documentation/cxl/cxl-create-region.txt b/Documentation/cxl/cxl-create-region.txt
> index ada0e52..f11a412 100644
> --- a/Documentation/cxl/cxl-create-region.txt
> +++ b/Documentation/cxl/cxl-create-region.txt
> @@ -75,7 +75,8 @@ include::bus-option.txt[]
>  
>  -t::
>  --type=::
> -	Specify the region type - 'pmem' or 'ram'. Defaults to 'pmem'.
> +	Specify the region type - 'pmem' or 'ram'. Default to root decoder
> +	capability, and if that is ambiguous, default to 'pmem'.
>  
>  -U::
>  --uuid=::
> diff --git a/cxl/region.c b/cxl/region.c
> index 9079b2d..1c8ccc7 100644
> --- a/cxl/region.c
> +++ b/cxl/region.c
> @@ -448,6 +448,31 @@ static int validate_decoder(struct cxl_decoder *decoder,
>  	return 0;
>  }
>  
> +static void set_type_from_decoder(struct cxl_ctx *ctx, struct parsed_params *p)
> +{
> +	int num_cap = 0;
> +
> +	/* if param.type was explicitly specified, nothing to do here */
> +	if (param.type)
> +		return;
> +
> +	/*
> +	 * if the root decoder only has one type of capability, default
> +	 * to that mode for the region.
> +	 */
> +	if (cxl_decoder_is_pmem_capable(p->root_decoder))
> +		num_cap++;
> +	if (cxl_decoder_is_volatile_capable(p->root_decoder))
> +		num_cap++;
> +
> +	if (num_cap == 1) {
> +		if (cxl_decoder_is_volatile_capable(p->root_decoder))
> +			p->mode = CXL_DECODER_MODE_RAM;
> +		else if (cxl_decoder_is_pmem_capable(p->root_decoder))
> +			p->mode = CXL_DECODER_MODE_PMEM;
> +	}

Is @num_cap needed? I.e. if this just does:

    if (cxl_decoder_is_volatile_capable(p->root_decoder))
    	p->mode = CXL_DECODER_MODE_RAM;
    if (cxl_decoder_is_pmem_capable(p->root_decoder))
    	p->mode = CXL_DECODER_MODE_PMEM;

...then it matches the changelog of defaulting to pmem if both types are
set, and otherwise the single capability dominates.

  parent reply	other threads:[~2023-02-08  5:56 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-02-07 19:16 [PATCH ndctl 0/7] cxl: add support for listing and creating volatile regions Vishal Verma
2023-02-07 19:16 ` [PATCH ndctl 1/7] cxl/region: skip region_actions for region creation Vishal Verma
2023-02-07 22:07   ` Fan Ni
2023-02-08  0:19     ` Verma, Vishal L
2023-02-08  3:45   ` Ira Weiny
2023-02-08  5:41   ` Dan Williams
2023-02-07 19:16 ` [PATCH ndctl 2/7] cxl: add a type attribute to region listings Vishal Verma
2023-02-08  3:46   ` Ira Weiny
2023-02-08  5:47   ` Dan Williams
2023-02-08  6:10     ` Verma, Vishal L
2023-02-10  0:47   ` Fan Ni
2023-02-07 19:16 ` [PATCH ndctl 3/7] cxl: add core plumbing for creation of ram regions Vishal Verma
2023-02-08  3:55   ` Ira Weiny
2023-02-08  6:23     ` Verma, Vishal L
2023-02-08 22:07       ` Ira Weiny
2023-02-08  5:49   ` Dan Williams
2023-02-10  1:04   ` Fan Ni
2023-02-10  1:10     ` Verma, Vishal L
2023-02-10  1:15       ` Fan Ni
2023-02-07 19:16 ` [PATCH ndctl 4/7] cxl/region: accept user-supplied UUIDs for pmem regions Vishal Verma
2023-02-08  3:56   ` Ira Weiny
2023-02-08  5:51   ` Dan Williams
2023-02-07 19:16 ` [PATCH ndctl 5/7] cxl/region: determine region type based on root decoder capability Vishal Verma
2023-02-08  4:07   ` Ira Weiny
2023-02-08  6:34     ` Verma, Vishal L
2023-02-08 22:09       ` Ira Weiny
2023-02-08  5:55   ` Dan Williams [this message]
2023-02-08  6:36     ` Verma, Vishal L
2023-02-07 19:16 ` [PATCH ndctl 6/7] cxl/list: Include regions in the verbose listing Vishal Verma
2023-02-08  4:08   ` Ira Weiny
2023-02-07 19:16 ` [PATCH ndctl 7/7] cxl/list: Enumerate device-dax properties for regions Vishal Verma
2023-02-08  4:15   ` Ira Weiny
2023-02-08  6:00   ` Dan Williams
2023-02-08  6:48     ` Verma, Vishal L

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=63e3396a17860_e3dae294d8@dwillia2-xfh.jf.intel.com.notmuch \
    --to=dan.j.williams@intel.com \
    --cc=Jonathan.Cameron@huawei.com \
    --cc=dave@stgolabs.net \
    --cc=gregory.price@memverge.com \
    --cc=linux-cxl@vger.kernel.org \
    --cc=nvdimm@lists.linux.dev \
    --cc=vishal.l.verma@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