[PATCH] ahci: compile out msi/msix infrastructure

[PATCH] ahci: compile out msi/msix infrastructure

[Dan Williams]

Quoting Arnd:
    The AHCI driver is used for some on-chip devices that do not use PCI
    for probing, and it can be built even when CONFIG_PCI is disabled, but
    that now results in a build failure:

    ata/libahci.c: In function 'ahci_host_activate_multi_irqs':
    ata/libahci.c:2475:4: error: invalid use of undefined type 'struct msix_entry'
    ata/libahci.c:2475:21: error: dereferencing pointer to incomplete type 'struct msix_entry'

Add ifdef CONFIG_PCI_MSI infrastructure to compile out the mutli-msi and
multi-msix code.

Reported-by: Arnd Bergmann <arnd@arndb.de>
Reported-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Fixes: d684a90d38e2 ("ahci: per-port msix support")
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---
 drivers/ata/ahci.h    |   25 +++++++++++++++++++++++--
 drivers/ata/libahci.c |    7 +------
 2 files changed, 24 insertions(+), 8 deletions(-)

diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index 878470f9c3e2..e3d888d4b6d0 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -237,12 +237,18 @@ enum {
 	AHCI_HFLAG_DELAY_ENGINE		= (1 << 15), /* do not start engine on
 						        port start (wait until
 						        error-handling stage) */
-	AHCI_HFLAG_MULTI_MSI		= (1 << 16), /* multiple PCI MSIs */
 	AHCI_HFLAG_NO_DEVSLP		= (1 << 17), /* no device sleep */
 	AHCI_HFLAG_NO_FBS		= (1 << 18), /* no FBS */
 	AHCI_HFLAG_EDGE_IRQ		= (1 << 19), /* HOST_IRQ_STAT behaves as
 							Edge Triggered */
-	AHCI_HFLAG_MULTI_MSIX		= (1 << 20), /* per-port MSI-X */
+#ifdef CONFIG_PCI_MSI
+	AHCI_HFLAG_MULTI_MSI		= (1 << 20), /* multiple PCI MSIs */
+	AHCI_HFLAG_MULTI_MSIX		= (1 << 21), /* per-port MSI-X */
+#else
+	/* compile out MSI infrastructure */
+	AHCI_HFLAG_MULTI_MSI		= 0,
+	AHCI_HFLAG_MULTI_MSIX		= 0,
+#endif
 
 	/* ap->flags bits */
 
@@ -355,6 +361,21 @@ struct ahci_host_priv {
 	void			(*start_engine)(struct ata_port *ap);
 };
 
+#ifdef CONFIG_PCI_MSI
+static inline int ahci_irq_vector(struct ahci_host_priv *hpriv, int port)
+{
+	if (hpriv->flags & AHCI_HFLAG_MULTI_MSIX)
+		return hpriv->msix[i].vector;
+	else
+		return hpriv->irq + i;
+}
+#else
+static inline int ahci_irq_vector(struct ahci_host_priv *hpriv, int port)
+{
+	return hpriv->irq;
+}
+#endif
+
 extern int ahci_ignore_sss;
 
 extern struct device_attribute *ahci_shost_attrs[];
diff --git a/drivers/ata/libahci.c b/drivers/ata/libahci.c
index 1b6c7cc415bf..eda3cf2163bb 100644
--- a/drivers/ata/libahci.c
+++ b/drivers/ata/libahci.c
@@ -2469,12 +2469,7 @@ static int ahci_host_activate_multi_irqs(struct ata_host *host,
 	 */
 	for (i = 0; i < host->n_ports; i++) {
 		struct ahci_port_priv *pp = host->ports[i]->private_data;
-		int irq;
-
-		if (hpriv->flags & AHCI_HFLAG_MULTI_MSIX)
-			irq = hpriv->msix[i].vector;
-		else
-			irq = hpriv->irq + i;
+		int irq = ahci_irq_vector(hpriv, i);
 
 		/* Do not receive interrupts sent by dummy ports */
 		if (!pp) {

[PATCH] ahci: compile out msi/msix infrastructure

[Arnd Bergmann]

On Friday 04 December 2015 11:58:30 Dan Williams wrote:
> Quoting Arnd:
>     The AHCI driver is used for some on-chip devices that do not use PCI
>     for probing, and it can be built even when CONFIG_PCI is disabled, but
>     that now results in a build failure:
> 
>     ata/libahci.c: In function 'ahci_host_activate_multi_irqs':
>     ata/libahci.c:2475:4: error: invalid use of undefined type 'struct msix_entry'
>     ata/libahci.c:2475:21: error: dereferencing pointer to incomplete type 'struct msix_entry'
> 
> Add ifdef CONFIG_PCI_MSI infrastructure to compile out the mutli-msi and
> multi-msix code.
> 
> Reported-by: Arnd Bergmann <arnd@arndb.de>
> Reported-by: Paul Gortmaker <paul.gortmaker@windriver.com>
> Fixes: d684a90d38e2 ("ahci: per-port msix support")
> Signed-off-by: Dan Williams <dan.j.williams@intel.com>

Thanks a lot, looks good. I've verified that this fixes the build on the
configuration that failed earlier.

Tested-by: Arnd Bergmann <arnd@arndb.de>

[PATCH] ahci: compile out msi/msix infrastructure

[Arnd Bergmann]

On Friday 04 December 2015 11:58:30 Dan Williams wrote:
> +#ifdef CONFIG_PCI_MSI
> +static inline int ahci_irq_vector(struct ahci_host_priv *hpriv, int port)
> +{
> +       if (hpriv->flags & AHCI_HFLAG_MULTI_MSIX)
> +               return hpriv->msix[i].vector;
> +       else
> +               return hpriv->irq + i;
> +}
> 

My randconfig tests now noticed that 'i' is undeclared here. I suppose it
should be 'port' instead of 'i'?

	Arnd

[PATCH] ahci: compile out msi/msix infrastructure

[Arnd Bergmann]

On Saturday 05 December 2015 23:22:49 Arnd Bergmann wrote:
> On Friday 04 December 2015 11:58:30 Dan Williams wrote:
> > +#ifdef CONFIG_PCI_MSI
> > +static inline int ahci_irq_vector(struct ahci_host_priv *hpriv, int port)
> > +{
> > +       if (hpriv->flags & AHCI_HFLAG_MULTI_MSIX)
> > +               return hpriv->msix[i].vector;
> > +       else
> > +               return hpriv->irq + i;
> > +}
> > 
> 
> My randconfig tests now noticed that 'i' is undeclared here. I suppose it
> should be 'port' instead of 'i'?
> 

and one more:

In file included from /git/arm-soc/drivers/ata/ahci_platform.c:25:0:
/git/arm-soc/drivers/ata/ahci.h: In function 'ahci_irq_vector':
/git/arm-soc/drivers/ata/ahci.h:368:3: error: invalid use of undefined type 'struct msix_entry'
   return hpriv->msix[port].vector;
   ^
/git/arm-soc/drivers/ata/ahci.h:368:21: error: dereferencing pointer to incomplete type 'struct msix_entry'
   return hpriv->msix[port].vector;


diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
index c450b8a914eb..dfbd033706bb 100644
--- a/drivers/ata/ahci.h
+++ b/drivers/ata/ahci.h
@@ -37,6 +37,7 @@
 
 #include <linux/clk.h>
 #include <linux/libata.h>
+#include <linux/pci.h>
 #include <linux/phy/phy.h>
 #include <linux/regulator/consumer.h>
 

[PATCH] ahci: compile out msi/msix infrastructure

[Dan Williams]

On Sat, Dec 5, 2015 at 2:25 PM, Arnd Bergmann <arnd@arndb.de> wrote:
> On Saturday 05 December 2015 23:22:49 Arnd Bergmann wrote:
>> On Friday 04 December 2015 11:58:30 Dan Williams wrote:
>> > +#ifdef CONFIG_PCI_MSI
>> > +static inline int ahci_irq_vector(struct ahci_host_priv *hpriv, int port)
>> > +{
>> > +       if (hpriv->flags & AHCI_HFLAG_MULTI_MSIX)
>> > +               return hpriv->msix[i].vector;
>> > +       else
>> > +               return hpriv->irq + i;
>> > +}
>> >
>>
>> My randconfig tests now noticed that 'i' is undeclared here. I suppose it
>> should be 'port' instead of 'i'?
>>
>
> and one more:
>
> In file included from /git/arm-soc/drivers/ata/ahci_platform.c:25:0:
> /git/arm-soc/drivers/ata/ahci.h: In function 'ahci_irq_vector':
> /git/arm-soc/drivers/ata/ahci.h:368:3: error: invalid use of undefined type 'struct msix_entry'
>    return hpriv->msix[port].vector;
>    ^
> /git/arm-soc/drivers/ata/ahci.h:368:21: error: dereferencing pointer to incomplete type 'struct msix_entry'
>    return hpriv->msix[port].vector;
>
>
> diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h
> index c450b8a914eb..dfbd033706bb 100644
> --- a/drivers/ata/ahci.h
> +++ b/drivers/ata/ahci.h
> @@ -37,6 +37,7 @@
>
>  #include <linux/clk.h>
>  #include <linux/libata.h>
> +#include <linux/pci.h>
>  #include <linux/phy/phy.h>
>  #include <linux/regulator/consumer.h>
>

I'll resend, sloppy on my part.