Linux CXL
 help / color / mirror / Atom feed
From: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
To: Dan Williams <dan.j.williams@intel.com>
Cc: <linux-cxl@vger.kernel.org>, <ben.widawsky@intel.com>,
	<ira.weiny@intel.com>, <alison.schofield@intel.com>,
	<vishal.l.verma@intel.com>
Subject: Re: [PATCH 09/14] cxl/mem: Fix CXL DVSEC Range Sizing
Date: Wed, 18 May 2022 17:40:47 +0100	[thread overview]
Message-ID: <20220518174047.00007711@Huawei.com> (raw)
In-Reply-To: <165237930521.3832067.16931437806464317011.stgit@dwillia2-desk3.amr.corp.intel.com>

On Thu, 12 May 2022 11:15:05 -0700
Dan Williams <dan.j.williams@intel.com> wrote:

> Per CXL 2.0 Section 8.1.3.8.4 "DVSEC CXL Range 1 Base Low" there is no
> way to specify decode sizes smaller than 256M. Fix cxl_dvsec_ranges()
> and cxl_hdm_decode_init() to account for that default decode range.

This is effectively the same as the discussion on patch 14.

My reading of the spec suggests that size can be 0 and that would mean
no access is passed on to the hardware.  It's a rather odd corner case
and would mean the device would only work with HDM decoders and might
well fail some compliance tests (I haven't checked)

It's not a corner case I care about... 

> Note, that this means that any BIOS implementation that sets mem_enable,
> but not HDM Decoder Capability enable will cause the driver to fail to
> attach. A later change validates the DVSEC ranges against platform CXL
> decode (CXL CFMWS) to make a decision about overriding the default DVSEC
> Range configuration.

> 
> Fixes: 560f78559006 ("cxl/pci: Retrieve CXL DVSEC memory info")
> Signed-off-by: Dan Williams <dan.j.williams@intel.com>
> ---
>  drivers/cxl/core/pci.c |   10 +++++++---
>  drivers/cxl/mem.c      |   10 +---------
>  2 files changed, 8 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/cxl/core/pci.c b/drivers/cxl/core/pci.c
> index f3e59f8b6621..f1c0677a4f52 100644
> --- a/drivers/cxl/core/pci.c
> +++ b/drivers/cxl/core/pci.c
> @@ -236,7 +236,12 @@ int cxl_dvsec_ranges(struct cxl_dev_state *cxlds,
>  		if (rc)
>  			return rc;
>  
> -		size = (u64)temp << 32;
> +		/*
> +		 * Per CXL 2.0 Section 8.1.3.8.4 "DVSEC CXL Range 1 Base
> +		 * Low", the minimum decode size is 256MB
> +		 */
> +		size = SZ_256M;

This is not how I read the spec.  The match is base <= Addr < base + size.
If size == 0 then there are no matches as base = base + size and so the
right condition isn't met.  Maybe I'm missing something though...


> +		size |= (u64)temp << 32;
>  
>  		rc = pci_read_config_dword(
>  			pdev, d + CXL_DVSEC_RANGE_SIZE_LOW(i), &temp);
> @@ -264,8 +269,7 @@ int cxl_dvsec_ranges(struct cxl_dev_state *cxlds,
>  			.end = base + size - 1
>  		};
>  
> -		if (size)
> -			ranges++;
> +		ranges++;
>  	}
>  
>  	info->ranges = ranges;
> diff --git a/drivers/cxl/mem.c b/drivers/cxl/mem.c
> index 902d1f6e189e..af4a88d3c5fa 100644
> --- a/drivers/cxl/mem.c
> +++ b/drivers/cxl/mem.c
> @@ -84,15 +84,7 @@ __mock bool cxl_hdm_decode_init(struct cxl_dev_state *cxlds,
>  			    CXL_HDM_DECODER_CTRL_OFFSET);
>  	global_enable = global_ctrl & CXL_HDM_DECODER_ENABLE;
>  
> -	/*
> -	 * Per CXL 2.0 Section 8.1.3.8.3 and 8.1.3.8.4 DVSEC CXL Range 1 Base
> -	 * [High,Low] when HDM operation is enabled the range register values
> -	 * are ignored by the device, but the spec also recommends matching the
> -	 * DVSEC Range 1,2 to HDM Decoder Range 0,1. So, non-zero info->ranges
> -	 * are expected even though Linux does not require or maintain that
> -	 * match.
> -	 */
> -	if (!global_enable && info->mem_enabled && info->ranges)
> +	if (!global_enable && info->mem_enabled)
>  		goto out;
>  
>  	retval = true;
> 


  reply	other threads:[~2022-05-18 16:41 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-12 18:14 [PATCH 00/14] cxl: Fix "mem_enable" handling Dan Williams
2022-05-12 18:14 ` [PATCH 01/14] cxl/mem: Drop mem_enabled check from wait_for_media() Dan Williams
2022-05-18 17:21   ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 02/14] cxl/pci: Consolidate wait_for_media() and wait_for_media_ready() Dan Williams
2022-05-18 17:22   ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 03/14] cxl/pci: Drop wait_for_valid() from cxl_await_media_ready() Dan Williams
2022-05-18 17:22   ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 04/14] cxl/mem: Fix cxl_mem_probe() error exit Dan Williams
2022-05-18 17:23   ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 05/14] cxl/mem: Validate port connectivity before dvsec ranges Dan Williams
2022-05-18 16:13   ` Jonathan Cameron
2022-05-18 16:41     ` Dan Williams
2022-05-18 17:21       ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 06/14] cxl/pci: Move cxl_await_media_ready() to the core Dan Williams
2022-05-18 16:21   ` Jonathan Cameron
2022-05-18 16:37     ` Dan Williams
2022-05-18 17:20       ` Jonathan Cameron
2022-05-18 18:22         ` Dan Williams
2022-05-12 18:14 ` [PATCH 07/14] cxl/mem: Consolidate CXL DVSEC Range enumeration in " Dan Williams
2022-05-18 16:31   ` Jonathan Cameron
2022-05-18 16:52     ` Dan Williams
2022-05-18 17:24       ` Jonathan Cameron
2022-05-12 18:14 ` [PATCH 08/14] cxl/mem: Skip range enumeration if mem_enable clear Dan Williams
2022-05-18 17:25   ` Jonathan Cameron
2022-05-12 18:15 ` [PATCH 09/14] cxl/mem: Fix CXL DVSEC Range Sizing Dan Williams
2022-05-18 16:40   ` Jonathan Cameron [this message]
2022-05-18 17:06     ` Dan Williams
2022-05-12 18:15 ` [PATCH 10/14] cxl/mem: Merge cxl_dvsec_ranges() and cxl_hdm_decode_init() Dan Williams
2022-05-12 18:15 ` [PATCH 11/14] cxl/pci: Drop @info argument to cxl_hdm_decode_init() Dan Williams
2022-05-18 16:45   ` Jonathan Cameron
2022-05-12 18:15 ` [PATCH 12/14] cxl/port: Move endpoint HDM Decoder Capability init to port driver Dan Williams
2022-05-18 16:50   ` Jonathan Cameron
2022-05-12 18:15 ` [PATCH 13/14] cxl/port: Reuse 'struct cxl_hdm' context for hdm init Dan Williams
2022-05-18 16:50   ` Jonathan Cameron
2022-05-12 18:15 ` [PATCH 14/14] cxl/port: Enable HDM Capability after validating DVSEC Ranges Dan Williams
2022-05-16 18:41   ` Ariel.Sibley
2022-05-16 18:52     ` Dan Williams
2022-05-16 19:31       ` Ariel.Sibley
2022-05-16 20:07         ` Dan Williams
2022-05-18  0:38   ` [PATCH v2 " Dan Williams
2022-05-18  2:07     ` Ariel.Sibley
2022-05-18  2:44       ` Dan Williams
2022-05-18 15:33         ` Jonathan Cameron
2022-05-18 17:17     ` Jonathan Cameron
2022-05-18 18:00       ` Dan Williams
2022-05-18  0:50 ` [PATCH 00/14] cxl: Fix "mem_enable" handling Ira Weiny

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=20220518174047.00007711@Huawei.com \
    --to=jonathan.cameron@huawei.com \
    --cc=alison.schofield@intel.com \
    --cc=ben.widawsky@intel.com \
    --cc=dan.j.williams@intel.com \
    --cc=ira.weiny@intel.com \
    --cc=linux-cxl@vger.kernel.org \
    --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