From: Robert Richter <rrichter@amd.com>
To: Alison Schofield <alison.schofield@intel.com>,
Vishal Verma <vishal.l.verma@intel.com>,
Ira Weiny <ira.weiny@intel.com>,
Dan Williams <dan.j.williams@intel.com>,
Jonathan Cameron <Jonathan.Cameron@huawei.com>,
Dave Jiang <dave.jiang@intel.com>,
Davidlohr Bueso <dave@stgolabs.net>
Cc: <linux-cxl@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
Gregory Price <gourry@gourry.net>,
"Fabio M. De Francesco" <fabio.m.de.francesco@linux.intel.com>,
Terry Bowman <terry.bowman@amd.com>,
Robert Richter <rrichter@amd.com>
Subject: [PATCH v5 03/14] cxl/pci: Add comments to cxl_hdm_decode_init()
Date: Mon, 28 Apr 2025 23:43:06 +0200 [thread overview]
Message-ID: <20250428214318.1682212-4-rrichter@amd.com> (raw)
In-Reply-To: <20250428214318.1682212-1-rrichter@amd.com>
There are various configuration cases of HDM decoder registers causing
different code paths. Add comments to cxl_hdm_decode_init() to better
explain them.
Signed-off-by: Robert Richter <rrichter@amd.com>
Reviewed-by: Gregory Price <gourry@gourry.net>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Alison Schofield <alison.schofield@intel.com>
Reviewed-by: Fabio M. De Francesco <fabio.m.de.francesco@linux.intel.com>
Tested-by: Gregory Price <gourry@gourry.net>
---
drivers/cxl/core/pci.c | 31 ++++++++++++++++++++++---------
1 file changed, 22 insertions(+), 9 deletions(-)
diff --git a/drivers/cxl/core/pci.c b/drivers/cxl/core/pci.c
index 6386e84e51a4..80bfd8ca4f8d 100644
--- a/drivers/cxl/core/pci.c
+++ b/drivers/cxl/core/pci.c
@@ -416,9 +416,19 @@ int cxl_hdm_decode_init(struct cxl_dev_state *cxlds, struct cxl_hdm *cxlhdm,
if (global_ctrl & CXL_HDM_DECODER_ENABLE || (!hdm && info->mem_enabled))
return devm_cxl_enable_mem(&port->dev, cxlds);
+ /*
+ * If the HDM Decoder Capability does not exist and DVSEC was
+ * not setup, the DVSEC based emulation cannot be used.
+ */
if (!hdm)
return -ENODEV;
+ /* The HDM Decoder Capability exists but is globally disabled. */
+
+ /*
+ * If the DVSEC CXL Range registers are not enabled, just
+ * enable and use the HDM Decoder Capability registers.
+ */
if (!info->mem_enabled) {
rc = devm_cxl_enable_hdm(&port->dev, cxlhdm);
if (rc)
@@ -427,6 +437,18 @@ int cxl_hdm_decode_init(struct cxl_dev_state *cxlds, struct cxl_hdm *cxlhdm,
return devm_cxl_enable_mem(&port->dev, cxlds);
}
+ /*
+ * 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. Check if at least one DVSEC range is enabled and allowed by
+ * the platform. That is, the DVSEC range must be covered by a locked
+ * platform window (CFMWS). Fail otherwise as the endpoint's decoders
+ * cannot be used.
+ */
+
root = to_cxl_port(port->dev.parent);
while (!is_cxl_root(root) && is_cxl_port(root->dev.parent))
root = to_cxl_port(root->dev.parent);
@@ -454,15 +476,6 @@ int cxl_hdm_decode_init(struct cxl_dev_state *cxlds, struct cxl_hdm *cxlhdm,
return -ENXIO;
}
- /*
- * 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 at least one DVSEC range is enabled and allowed, skip HDM
- * Decoder Capability Enable.
- */
return 0;
}
EXPORT_SYMBOL_NS_GPL(cxl_hdm_decode_init, "CXL");
--
2.39.5
next prev parent reply other threads:[~2025-04-28 21:43 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-28 21:43 [PATCH v5 00/14] cxl: Address translation support, part 1: Cleanups and refactoring Robert Richter
2025-04-28 21:43 ` [PATCH v5 01/14] cxl: Remove else after return Robert Richter
2025-04-29 1:35 ` Alison Schofield
2025-04-28 21:43 ` [PATCH v5 02/14] cxl/pci: Moving code in cxl_hdm_decode_init() Robert Richter
2025-04-28 22:07 ` Dave Jiang
2025-04-29 15:17 ` Jonathan Cameron
2025-04-28 21:43 ` Robert Richter [this message]
2025-04-28 22:08 ` [PATCH v5 03/14] cxl/pci: Add comments to cxl_hdm_decode_init() Dave Jiang
2025-04-28 21:43 ` [PATCH v5 04/14] cxl: Introduce parent_port_of() helper Robert Richter
2025-04-28 21:43 ` [PATCH v5 05/14] cxl/region: Rename function to cxl_port_pick_region_decoder() Robert Richter
2025-04-28 22:24 ` Dave Jiang
2025-04-29 15:31 ` Jonathan Cameron
2025-05-07 15:15 ` Jonathan Cameron
2025-05-05 17:20 ` Fabio M. De Francesco
2025-04-28 21:43 ` [PATCH v5 06/14] cxl/region: Avoid duplicate call of cxl_port_pick_region_decoder() Robert Richter
2025-04-28 22:28 ` Dave Jiang
2025-04-29 15:34 ` Jonathan Cameron
2025-05-05 17:51 ` Fabio M. De Francesco
2025-04-28 21:43 ` [PATCH v5 07/14] cxl/region: Move find_cxl_root() to cxl_add_to_region() Robert Richter
2025-04-28 21:43 ` [PATCH v5 08/14] cxl/port: Replace put_cxl_root() by a cleanup helper Robert Richter
2025-04-28 22:27 ` Gregory Price
2025-04-28 23:03 ` Dave Jiang
2025-04-29 15:38 ` Jonathan Cameron
2025-04-29 15:38 ` Jonathan Cameron
2025-05-05 17:55 ` Fabio M. De Francesco
2025-04-28 21:43 ` [PATCH v5 09/14] cxl/region: Factor out code to find the root decoder Robert Richter
2025-04-28 21:43 ` [PATCH v5 10/14] cxl/region: Factor out code to find a root decoder's region Robert Richter
2025-04-28 21:43 ` [PATCH v5 11/14] cxl/region: Add function to find a port's switch decoder by range Robert Richter
2025-05-05 18:18 ` Fabio M. De Francesco
2025-04-28 21:43 ` [PATCH v5 12/14] cxl/region: Add a dev_warn() on registration failure Robert Richter
2025-04-28 21:43 ` [PATCH v5 13/14] cxl/region: Add a dev_err() on missing target list entries Robert Richter
2025-04-28 21:43 ` [PATCH v5 14/14] cxl: Add a dev_dbg() when a decoder was added to a port Robert Richter
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=20250428214318.1682212-4-rrichter@amd.com \
--to=rrichter@amd.com \
--cc=Jonathan.Cameron@huawei.com \
--cc=alison.schofield@intel.com \
--cc=dan.j.williams@intel.com \
--cc=dave.jiang@intel.com \
--cc=dave@stgolabs.net \
--cc=fabio.m.de.francesco@linux.intel.com \
--cc=gourry@gourry.net \
--cc=ira.weiny@intel.com \
--cc=linux-cxl@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=terry.bowman@amd.com \
--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