diff mbox series

[v2,1/2] cxl/acpi: Add the Host Bridge base address to CXL port objects

Message ID e841b0283edcc281ff31e98e4d3512be3a131c6a.1623800340.git.alison.schofield@intel.com
State New
Headers show
Series CXL ACPI tables for object creation | expand

Commit Message

Alison Schofield June 16, 2021, 12:20 a.m. UTC
The base address for the Host Bridge port component registers is located
in the CXL Host Bridge Structure (CHBS) of the ACPI CXL Early Discovery
Table (CEDT). Retrieve the CHBS for each Host Bridge (ACPI0016 device)
and include that base address in the port object.

Co-developed-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Signed-off-by: Alison Schofield <alison.schofield@intel.com>
---
 drivers/cxl/acpi.c | 105 ++++++++++++++++++++++++++++++++++++++++++---
 1 file changed, 99 insertions(+), 6 deletions(-)

Comments

Jonathan Cameron June 16, 2021, 4:13 p.m. UTC | #1
On Tue, 15 Jun 2021 17:20:38 -0700
Alison Schofield <alison.schofield@intel.com> wrote:

> The base address for the Host Bridge port component registers is located
> in the CXL Host Bridge Structure (CHBS) of the ACPI CXL Early Discovery
> Table (CEDT). Retrieve the CHBS for each Host Bridge (ACPI0016 device)
> and include that base address in the port object.
> 
> Co-developed-by: Vishal Verma <vishal.l.verma@intel.com>
> Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
> Signed-off-by: Alison Schofield <alison.schofield@intel.com>

Hi Alison,

A few small suggestions from me.

> ---
>  drivers/cxl/acpi.c | 105 ++++++++++++++++++++++++++++++++++++++++++---
>  1 file changed, 99 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/cxl/acpi.c b/drivers/cxl/acpi.c
> index be357eea552c..b6d9cd45428c 100644
> --- a/drivers/cxl/acpi.c
> +++ b/drivers/cxl/acpi.c
> @@ -8,6 +8,61 @@
>  #include <linux/pci.h>
>  #include "cxl.h"
>  
> +static struct acpi_table_header *cedt_table;
> +
> +static struct acpi_cedt_chbs *cxl_acpi_match_chbs(struct device *dev, u32 uid)
> +{
> +	struct acpi_cedt_chbs *chbs, *chbs_match = NULL;
> +	acpi_size len, cur = 0;
> +	void *cedt_base;
> +	int rc = 0;
> +
> +	len = cedt_table->length - sizeof(*cedt_table);
> +	cedt_base = cedt_table + 1;
> +
> +	while (cur < len) {
> +		struct acpi_cedt_header *c = cedt_base + cur;
> +
> +		if (c->type != ACPI_CEDT_TYPE_CHBS) {
> +			cur += c->length;
> +			continue;
> +		}
> +
> +		chbs = cedt_base + cur;
> +
> +		if (chbs->header.length < sizeof(*chbs)) {
> +			dev_err(dev, "Invalid CHBS header length: %u\n",
> +				chbs->header.length);
> +			rc = -EINVAL;

As below, direct return would be more obvious to my eyes.

> +			break;
> +		}
> +
> +		if (chbs->uid == uid && !chbs_match) {
> +			chbs_match = chbs;
> +			cur += c->length;
> +			continue;
> +		}
> +
> +		if (chbs->uid == uid && chbs_match) {
> +			dev_err(dev, "Duplicate CHBS UIDs %u\n", uid);

Do we actually care, or should we just drop out on first match?
I don't think think there is any obligation to catch broken tables.

> +			rc = -EINVAL;

Direct return might be easier to follow.
			return ERR_PTR(-EINVAL);

> +			break;
> +		}

Maybe more readable as (your option is fine if you prefer it).

		if (chbs->uuid != uid) {
			cur += c->length;
			continue;
		}

		if (chbs_match) {
			dev_err(dev, "D...");
			return ERR_PTR(-EINVAL);
		}

		chbs_match = chbs;


> +		cur += c->length;
> +	}
> +	if (!chbs_match)
> +		rc = -EINVAL;

		return ERR_PTR(-EINVAL);

> +	if (rc)
> +		return ERR_PTR(rc);
> +
> +	return chbs_match;
> +}
> +
> +static resource_size_t get_chbcr(struct acpi_cedt_chbs *chbs)
> +{
> +	return IS_ERR(chbs) ? CXL_RESOURCE_NONE : chbs->base;
> +}
> +
>  struct cxl_walk_context {
>  	struct device *dev;
>  	struct pci_bus *root;
> @@ -50,6 +105,21 @@ static int match_add_root_ports(struct pci_dev *pdev, void *data)
>  	return 0;
>  }
>  
> +static struct cxl_dport *find_dport_by_dev(struct cxl_port *port, struct device *dev)
> +{
> +	struct cxl_dport *dport;
> +
> +	device_lock(&port->dev);
> +	list_for_each_entry(dport, &port->dports, list)
> +		if (dport->dport == dev) {
> +			device_unlock(&port->dev);
> +			return dport;
> +		}
> +
> +	device_unlock(&port->dev);
> +	return NULL;
> +}
> +
>  static struct acpi_device *to_cxl_host_bridge(struct device *dev)
>  {
>  	struct acpi_device *adev = to_acpi_device(dev);
> @@ -71,6 +141,7 @@ static int add_host_bridge_uport(struct device *match, void *arg)
>  	struct acpi_pci_root *pci_root;
>  	struct cxl_walk_context ctx;
>  	struct cxl_decoder *cxld;
> +	struct cxl_dport *dport;
>  	struct cxl_port *port;
>  
>  	if (!bridge)
> @@ -80,8 +151,15 @@ static int add_host_bridge_uport(struct device *match, void *arg)
>  	if (!pci_root)
>  		return -ENXIO;
>  
> -	/* TODO: fold in CEDT.CHBS retrieval */
> -	port = devm_cxl_add_port(host, match, CXL_RESOURCE_NONE, root_port);
> +	dport = find_dport_by_dev(root_port, match);
> +	if (!dport) {
> +		dev_dbg(host, "host bridge expected and not found\n");
> +		return -ENODEV;
> +	}
> +
> +	port = devm_cxl_add_port(host, match, dport->component_reg_phys,
> +				 root_port);
> +

Nitpick, no blank line before error handling block.

>  	if (IS_ERR(port))
>  		return PTR_ERR(port);
>  	dev_dbg(host, "%s: add: %s\n", dev_name(match), dev_name(&port->dev));
> @@ -120,6 +198,7 @@ static int add_host_bridge_dport(struct device *match, void *arg)
>  	int rc;
>  	acpi_status status;
>  	unsigned long long uid;
> +	struct acpi_cedt_chbs *chbs;
>  	struct cxl_port *root_port = arg;
>  	struct device *host = root_port->dev.parent;
>  	struct acpi_device *bridge = to_cxl_host_bridge(match);
> @@ -135,7 +214,12 @@ static int add_host_bridge_dport(struct device *match, void *arg)
>  		return -ENODEV;
>  	}
>  
> -	rc = cxl_add_dport(root_port, match, uid, CXL_RESOURCE_NONE);
> +	chbs = cxl_acpi_match_chbs(host, uid);
> +	if (IS_ERR(chbs))
> +		dev_dbg(host, "No CHBS found for Host Bridge: %s\n",
> +			dev_name(match));
> +
> +	rc = cxl_add_dport(root_port, match, uid, get_chbcr(chbs));
>  	if (rc) {
>  		dev_err(host, "failed to add downstream port: %s\n",
>  			dev_name(match));
> @@ -148,6 +232,7 @@ static int add_host_bridge_dport(struct device *match, void *arg)
>  static int cxl_acpi_probe(struct platform_device *pdev)
>  {
>  	int rc;
> +	acpi_status status;
>  	struct cxl_port *root_port;
>  	struct device *host = &pdev->dev;
>  	struct acpi_device *adev = ACPI_COMPANION(host);
> @@ -157,17 +242,25 @@ static int cxl_acpi_probe(struct platform_device *pdev)
>  		return PTR_ERR(root_port);
>  	dev_dbg(host, "add: %s\n", dev_name(&root_port->dev));
>  
> +	status = acpi_get_table(ACPI_SIG_CEDT, 0, &cedt_table);
> +	if (ACPI_FAILURE(status))
> +		return -ENXIO;
> +
>  	rc = bus_for_each_dev(adev->dev.bus, NULL, root_port,
>  			      add_host_bridge_dport);
>  	if (rc)
> -		return rc;
> +		goto out;
>  
>  	/*
>  	 * Root level scanned with host-bridge as dports, now scan host-bridges
>  	 * for their role as CXL uports to their CXL-capable PCIe Root Ports.
>  	 */
> -	return bus_for_each_dev(adev->dev.bus, NULL, root_port,
> -				add_host_bridge_uport);
> +	rc = bus_for_each_dev(adev->dev.bus, NULL, root_port,
> +			      add_host_bridge_uport);
> +
> +out:
> +	acpi_put_table(cedt_table);
> +	return rc;
>  }
>  
>  static const struct acpi_device_id cxl_acpi_ids[] = {
Alison Schofield June 16, 2021, 11:16 p.m. UTC | #2
Thanks for the review Jonathan -

On Wed, Jun 16, 2021 at 05:13:40PM +0100, Jonathan Cameron wrote:
> On Tue, 15 Jun 2021 17:20:38 -0700
> Alison Schofield <alison.schofield@intel.com> wrote:
> 
> > The base address for the Host Bridge port component registers is located
> > in the CXL Host Bridge Structure (CHBS) of the ACPI CXL Early Discovery
> > Table (CEDT). Retrieve the CHBS for each Host Bridge (ACPI0016 device)
> > and include that base address in the port object.
> > 
> > Co-developed-by: Vishal Verma <vishal.l.verma@intel.com>
> > Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
> > Signed-off-by: Alison Schofield <alison.schofield@intel.com>
> 
> Hi Alison,
> 
> A few small suggestions from me.
> 
> > ---
> >  drivers/cxl/acpi.c | 105 ++++++++++++++++++++++++++++++++++++++++++---
> >  1 file changed, 99 insertions(+), 6 deletions(-)
> > 
> > diff --git a/drivers/cxl/acpi.c b/drivers/cxl/acpi.c
> > index be357eea552c..b6d9cd45428c 100644
> > --- a/drivers/cxl/acpi.c
> > +++ b/drivers/cxl/acpi.c
> > @@ -8,6 +8,61 @@
> >  #include <linux/pci.h>
> >  #include "cxl.h"
> >  
> > +static struct acpi_table_header *cedt_table;
> > +
> > +static struct acpi_cedt_chbs *cxl_acpi_match_chbs(struct device *dev, u32 uid)
> > +{
> > +	struct acpi_cedt_chbs *chbs, *chbs_match = NULL;
> > +	acpi_size len, cur = 0;
> > +	void *cedt_base;
> > +	int rc = 0;
> > +
> > +	len = cedt_table->length - sizeof(*cedt_table);
> > +	cedt_base = cedt_table + 1;
> > +
> > +	while (cur < len) {
> > +		struct acpi_cedt_header *c = cedt_base + cur;
> > +
> > +		if (c->type != ACPI_CEDT_TYPE_CHBS) {
> > +			cur += c->length;
> > +			continue;
> > +		}
> > +
> > +		chbs = cedt_base + cur;
> > +
> > +		if (chbs->header.length < sizeof(*chbs)) {
> > +			dev_err(dev, "Invalid CHBS header length: %u\n",
> > +				chbs->header.length);
> > +			rc = -EINVAL;
> 
> As below, direct return would be more obvious to my eyes.
> 

Well....I decided to warn & continue on this case. See the updated flow
in v3.

> > +			break;
> > +		}
> > +
> > +		if (chbs->uid == uid && !chbs_match) {
> > +			chbs_match = chbs;
> > +			cur += c->length;
> > +			continue;
> > +		}
> > +
> > +		if (chbs->uid == uid && chbs_match) {
> > +			dev_err(dev, "Duplicate CHBS UIDs %u\n", uid);
> 
> Do we actually care, or should we just drop out on first match?
> I don't think think there is any obligation to catch broken tables.
> 

Agree on the obligation part, but if things go wrong, this would be
nice to know. I left it in as a dev warn once. If you think that's 
too strong - let me know.


> > +			rc = -EINVAL;
> 
> Direct return might be easier to follow.
> 			return ERR_PTR(-EINVAL);
> 
> > +			break;
> > +		}
> 
> Maybe more readable as (your option is fine if you prefer it).
> 
> 		if (chbs->uuid != uid) {
> 			cur += c->length;
> 			continue;
> 		}
> 
> 		if (chbs_match) {
> 			dev_err(dev, "D...");
> 			return ERR_PTR(-EINVAL);
> 		}
> 
> 		chbs_match = chbs;
> 
> 

Thanks, I reworked the flow along these lines.


snip
> > +
> > +	port = devm_cxl_add_port(host, match, dport->component_reg_phys,
> > +				 root_port);
> > +
> 
> Nitpick, no blank line before error handling block.
> 

Got it. Thanks!

snip
Dan Williams June 16, 2021, 11:50 p.m. UTC | #3
On Wed, Jun 16, 2021 at 4:20 PM Alison Schofield
<alison.schofield@intel.com> wrote:
>

>

> Thanks for the review Jonathan -

>

> On Wed, Jun 16, 2021 at 05:13:40PM +0100, Jonathan Cameron wrote:

> > On Tue, 15 Jun 2021 17:20:38 -0700

> > Alison Schofield <alison.schofield@intel.com> wrote:

> >

> > > The base address for the Host Bridge port component registers is located

> > > in the CXL Host Bridge Structure (CHBS) of the ACPI CXL Early Discovery

> > > Table (CEDT). Retrieve the CHBS for each Host Bridge (ACPI0016 device)

> > > and include that base address in the port object.

> > >

> > > Co-developed-by: Vishal Verma <vishal.l.verma@intel.com>

> > > Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>

> > > Signed-off-by: Alison Schofield <alison.schofield@intel.com>

> >

> > Hi Alison,

> >

> > A few small suggestions from me.

> >

> > > ---

> > >  drivers/cxl/acpi.c | 105 ++++++++++++++++++++++++++++++++++++++++++---

> > >  1 file changed, 99 insertions(+), 6 deletions(-)

> > >

> > > diff --git a/drivers/cxl/acpi.c b/drivers/cxl/acpi.c

> > > index be357eea552c..b6d9cd45428c 100644

> > > --- a/drivers/cxl/acpi.c

> > > +++ b/drivers/cxl/acpi.c

> > > @@ -8,6 +8,61 @@

> > >  #include <linux/pci.h>

> > >  #include "cxl.h"

> > >

> > > +static struct acpi_table_header *cedt_table;

> > > +

> > > +static struct acpi_cedt_chbs *cxl_acpi_match_chbs(struct device *dev, u32 uid)

> > > +{

> > > +   struct acpi_cedt_chbs *chbs, *chbs_match = NULL;

> > > +   acpi_size len, cur = 0;

> > > +   void *cedt_base;

> > > +   int rc = 0;

> > > +

> > > +   len = cedt_table->length - sizeof(*cedt_table);

> > > +   cedt_base = cedt_table + 1;

> > > +

> > > +   while (cur < len) {

> > > +           struct acpi_cedt_header *c = cedt_base + cur;

> > > +

> > > +           if (c->type != ACPI_CEDT_TYPE_CHBS) {

> > > +                   cur += c->length;

> > > +                   continue;

> > > +           }

> > > +

> > > +           chbs = cedt_base + cur;

> > > +

> > > +           if (chbs->header.length < sizeof(*chbs)) {

> > > +                   dev_err(dev, "Invalid CHBS header length: %u\n",

> > > +                           chbs->header.length);

> > > +                   rc = -EINVAL;

> >

> > As below, direct return would be more obvious to my eyes.

> >

>

> Well....I decided to warn & continue on this case. See the updated flow

> in v3.

>

> > > +                   break;

> > > +           }

> > > +

> > > +           if (chbs->uid == uid && !chbs_match) {

> > > +                   chbs_match = chbs;

> > > +                   cur += c->length;

> > > +                   continue;

> > > +           }

> > > +

> > > +           if (chbs->uid == uid && chbs_match) {

> > > +                   dev_err(dev, "Duplicate CHBS UIDs %u\n", uid);

> >

> > Do we actually care, or should we just drop out on first match?

> > I don't think think there is any obligation to catch broken tables.

> >

>

> Agree on the obligation part, but if things go wrong, this would be

> nice to know. I left it in as a dev warn once. If you think that's

> too strong - let me know.


I do think the driver should care about duplicate UID, but only if
"version, base, or length" mismatch. If the BIOS gives us ambiguous
answers about where the registers are located, the user should be
warned that the driver might be picking the "wrong" one by accident.
If they are identical, the BIOS is being repetitive, but no real harm
that the driver would care about. A dev_warn_once() sounds good as the
first duplicate should be sufficient to say something fishy is afoot,
but it's not an error. The warn_once will also re-trigger when / if
the module is reloaded.
diff mbox series

Patch

diff --git a/drivers/cxl/acpi.c b/drivers/cxl/acpi.c
index be357eea552c..b6d9cd45428c 100644
--- a/drivers/cxl/acpi.c
+++ b/drivers/cxl/acpi.c
@@ -8,6 +8,61 @@ 
 #include <linux/pci.h>
 #include "cxl.h"
 
+static struct acpi_table_header *cedt_table;
+
+static struct acpi_cedt_chbs *cxl_acpi_match_chbs(struct device *dev, u32 uid)
+{
+	struct acpi_cedt_chbs *chbs, *chbs_match = NULL;
+	acpi_size len, cur = 0;
+	void *cedt_base;
+	int rc = 0;
+
+	len = cedt_table->length - sizeof(*cedt_table);
+	cedt_base = cedt_table + 1;
+
+	while (cur < len) {
+		struct acpi_cedt_header *c = cedt_base + cur;
+
+		if (c->type != ACPI_CEDT_TYPE_CHBS) {
+			cur += c->length;
+			continue;
+		}
+
+		chbs = cedt_base + cur;
+
+		if (chbs->header.length < sizeof(*chbs)) {
+			dev_err(dev, "Invalid CHBS header length: %u\n",
+				chbs->header.length);
+			rc = -EINVAL;
+			break;
+		}
+
+		if (chbs->uid == uid && !chbs_match) {
+			chbs_match = chbs;
+			cur += c->length;
+			continue;
+		}
+
+		if (chbs->uid == uid && chbs_match) {
+			dev_err(dev, "Duplicate CHBS UIDs %u\n", uid);
+			rc = -EINVAL;
+			break;
+		}
+		cur += c->length;
+	}
+	if (!chbs_match)
+		rc = -EINVAL;
+	if (rc)
+		return ERR_PTR(rc);
+
+	return chbs_match;
+}
+
+static resource_size_t get_chbcr(struct acpi_cedt_chbs *chbs)
+{
+	return IS_ERR(chbs) ? CXL_RESOURCE_NONE : chbs->base;
+}
+
 struct cxl_walk_context {
 	struct device *dev;
 	struct pci_bus *root;
@@ -50,6 +105,21 @@  static int match_add_root_ports(struct pci_dev *pdev, void *data)
 	return 0;
 }
 
+static struct cxl_dport *find_dport_by_dev(struct cxl_port *port, struct device *dev)
+{
+	struct cxl_dport *dport;
+
+	device_lock(&port->dev);
+	list_for_each_entry(dport, &port->dports, list)
+		if (dport->dport == dev) {
+			device_unlock(&port->dev);
+			return dport;
+		}
+
+	device_unlock(&port->dev);
+	return NULL;
+}
+
 static struct acpi_device *to_cxl_host_bridge(struct device *dev)
 {
 	struct acpi_device *adev = to_acpi_device(dev);
@@ -71,6 +141,7 @@  static int add_host_bridge_uport(struct device *match, void *arg)
 	struct acpi_pci_root *pci_root;
 	struct cxl_walk_context ctx;
 	struct cxl_decoder *cxld;
+	struct cxl_dport *dport;
 	struct cxl_port *port;
 
 	if (!bridge)
@@ -80,8 +151,15 @@  static int add_host_bridge_uport(struct device *match, void *arg)
 	if (!pci_root)
 		return -ENXIO;
 
-	/* TODO: fold in CEDT.CHBS retrieval */
-	port = devm_cxl_add_port(host, match, CXL_RESOURCE_NONE, root_port);
+	dport = find_dport_by_dev(root_port, match);
+	if (!dport) {
+		dev_dbg(host, "host bridge expected and not found\n");
+		return -ENODEV;
+	}
+
+	port = devm_cxl_add_port(host, match, dport->component_reg_phys,
+				 root_port);
+
 	if (IS_ERR(port))
 		return PTR_ERR(port);
 	dev_dbg(host, "%s: add: %s\n", dev_name(match), dev_name(&port->dev));
@@ -120,6 +198,7 @@  static int add_host_bridge_dport(struct device *match, void *arg)
 	int rc;
 	acpi_status status;
 	unsigned long long uid;
+	struct acpi_cedt_chbs *chbs;
 	struct cxl_port *root_port = arg;
 	struct device *host = root_port->dev.parent;
 	struct acpi_device *bridge = to_cxl_host_bridge(match);
@@ -135,7 +214,12 @@  static int add_host_bridge_dport(struct device *match, void *arg)
 		return -ENODEV;
 	}
 
-	rc = cxl_add_dport(root_port, match, uid, CXL_RESOURCE_NONE);
+	chbs = cxl_acpi_match_chbs(host, uid);
+	if (IS_ERR(chbs))
+		dev_dbg(host, "No CHBS found for Host Bridge: %s\n",
+			dev_name(match));
+
+	rc = cxl_add_dport(root_port, match, uid, get_chbcr(chbs));
 	if (rc) {
 		dev_err(host, "failed to add downstream port: %s\n",
 			dev_name(match));
@@ -148,6 +232,7 @@  static int add_host_bridge_dport(struct device *match, void *arg)
 static int cxl_acpi_probe(struct platform_device *pdev)
 {
 	int rc;
+	acpi_status status;
 	struct cxl_port *root_port;
 	struct device *host = &pdev->dev;
 	struct acpi_device *adev = ACPI_COMPANION(host);
@@ -157,17 +242,25 @@  static int cxl_acpi_probe(struct platform_device *pdev)
 		return PTR_ERR(root_port);
 	dev_dbg(host, "add: %s\n", dev_name(&root_port->dev));
 
+	status = acpi_get_table(ACPI_SIG_CEDT, 0, &cedt_table);
+	if (ACPI_FAILURE(status))
+		return -ENXIO;
+
 	rc = bus_for_each_dev(adev->dev.bus, NULL, root_port,
 			      add_host_bridge_dport);
 	if (rc)
-		return rc;
+		goto out;
 
 	/*
 	 * Root level scanned with host-bridge as dports, now scan host-bridges
 	 * for their role as CXL uports to their CXL-capable PCIe Root Ports.
 	 */
-	return bus_for_each_dev(adev->dev.bus, NULL, root_port,
-				add_host_bridge_uport);
+	rc = bus_for_each_dev(adev->dev.bus, NULL, root_port,
+			      add_host_bridge_uport);
+
+out:
+	acpi_put_table(cedt_table);
+	return rc;
 }
 
 static const struct acpi_device_id cxl_acpi_ids[] = {