Expand the MSI/MSI-X API to address some deficiencies in the MSI-X support.

- First off, device drivers really do need to know if they are allocating
  MSI or MSI-X messages.  MSI requires allocating powerof2() messages for
  example where MSI-X does not.  To address this, split out the MSI-X
  support from pci_msi_count() and pci_alloc_msi() into new driver-visible
  functions pci_msix_count() and pci_alloc_msix().  As a result,
  pci_msi_count() now just returns a count of the max supported MSI
  messages for the device, and pci_alloc_msi() only tries to allocate MSI
  messages.  To get a count of the max supported MSI-X messages, use
  pci_msix_count().  To allocate MSI-X messages, use pci_alloc_msix().
  pci_release_msi() still handles both MSI and MSI-X messages, however.
  As a result of this change, drivers using the existing API will only
  use MSI messages and will no longer try to use MSI-X messages.
- Because MSI-X allows for each message to have its own data and address
  values (and thus does not require all of the messages to have their
  MD vectors allocated as a group), some devices allow for "sparse" use
  of MSI-X message slots.  For example, if a device supports 8 messages
  but the OS is only able to allocate 2 messages, the device may make the
  best use of 2 IRQs if it enables the messages at slots 1 and 4 rather
  than default of using the first N slots (or indicies) at 1 and 2.  To
  support this, add a new pci_remap_msix() function that a driver may call
  after a successful pci_alloc_msix() (but before allocating any of the
  SYS_RES_IRQ resources) to allow the allocated IRQ resources to be
  assigned to different message indices.  For example, from the earlier
  example, after pci_alloc_msix() returned a value of 2, the driver would
  call pci_remap_msix() passing in array of integers { 1, 4 } as the
  new message indices to use.  The rid's for the SYS_RES_IRQ resources
  will always match the message indices.  Thus, after the call to
  pci_remap_msix() the driver would be able to access the first message
  in slot 1 at SYS_RES_IRQ rid 1, and the second message at slot 4 at
  SYS_RES_IRQ rid 4.  Note that the message slots/indices are 1-based
  rather than 0-based so that they will always correspond to the rid
  values (SYS_RES_IRQ rid 0 is reserved for the legacy INTx interrupt).
  To support this API, a new PCIB_REMAP_MSIX() method was added to the
  pcib interface to change the message index for a single IRQ.

Tested by:	scottl

5 files changed, 37 insertions, 0 deletions

diff --git a/sys/amd64/amd64/mptable_pci.c b/sys/amd64/amd64/mptable_pci.c
index 851b827..d8ed069 100644
--- a/sys/amd64/amd64/mptable_pci.c
+++ b/sys/amd64/amd64/mptable_pci.c
@@ -120,6 +120,7 @@ static device_method_t mptable_hostb_methods[] = {
 	DEVMETHOD(pcib_alloc_msi,	mptable_hostb_alloc_msi),
 	DEVMETHOD(pcib_release_msi,	pcib_release_msi),
 	DEVMETHOD(pcib_alloc_msix,	mptable_hostb_alloc_msix),
+	DEVMETHOD(pcib_remap_msix,	pcib_remap_msix),
 	DEVMETHOD(pcib_release_msix,	pcib_release_msix),
 
 	{ 0, 0 }
@@ -176,6 +177,7 @@ static device_method_t mptable_pcib_pci_methods[] = {
 	DEVMETHOD(pcib_alloc_msi,	pcib_alloc_msi),
 	DEVMETHOD(pcib_release_msi,	pcib_release_msi),
 	DEVMETHOD(pcib_alloc_msix,	pcib_alloc_msix),
+	DEVMETHOD(pcib_remap_msix,	pcib_remap_msix),
 	DEVMETHOD(pcib_release_msix,	pcib_release_msix),
 
 	{0, 0}
diff --git a/sys/amd64/amd64/msi.c b/sys/amd64/amd64/msi.c
index 7be702a..321078c 100644
--- a/sys/amd64/amd64/msi.c
+++ b/sys/amd64/amd64/msi.c
@@ -480,6 +480,30 @@ msix_alloc(device_t dev, int index, int *irq, int *new)
 }
 
 int
+msix_remap(int index, int irq)
+{
+	struct msi_intsrc *msi;
+
+	sx_xlock(&msi_sx);
+	msi = (struct msi_intsrc *)intr_lookup_source(irq);
+	if (msi == NULL) {
+		sx_xunlock(&msi_sx);
+		return (ENOENT);
+	}
+
+	/* Make sure this is an MSI-X message. */
+	if (!msi->msi_msix) {
+		sx_xunlock(&msi_sx);
+		return (EINVAL);
+	}
+
+	KASSERT(msi->msi_dev != NULL, ("unowned message"));
+	msi->msi_index = index;
+	sx_xunlock(&msi_sx);
+	return (0);
+}
+
+int
 msix_release(int irq)
 {
 	struct msi_intsrc *msi;
diff --git a/sys/amd64/amd64/nexus.c b/sys/amd64/amd64/nexus.c
index c06f629..01e9f42 100644
--- a/sys/amd64/amd64/nexus.c
+++ b/sys/amd64/amd64/nexus.c
@@ -105,6 +105,7 @@ static void nexus_delete_resource(device_t, device_t, int, int);
 static	int nexus_alloc_msi(device_t pcib, device_t dev, int count, int maxcount, int *irqs);
 static	int nexus_release_msi(device_t pcib, device_t dev, int count, int *irqs);
 static	int nexus_alloc_msix(device_t pcib, device_t dev, int index, int *irq);
+static	int nexus_remap_msix(device_t pcib, device_t dev, int index, int irq);
 static	int nexus_release_msix(device_t pcib, device_t dev, int irq);
 
 static device_method_t nexus_methods[] = {
@@ -135,6 +136,7 @@ static device_method_t nexus_methods[] = {
 	DEVMETHOD(pcib_alloc_msi,	nexus_alloc_msi),
 	DEVMETHOD(pcib_release_msi,	nexus_release_msi),
 	DEVMETHOD(pcib_alloc_msix,	nexus_alloc_msix),
+	DEVMETHOD(pcib_remap_msix,	nexus_remap_msix),
 	DEVMETHOD(pcib_release_msix,	nexus_release_msix),
 
 	{ 0, 0 }
@@ -510,6 +512,13 @@ nexus_alloc_msix(device_t pcib, device_t dev, int index, int *irq)
 }
 
 static int
+nexus_remap_msix(device_t pcib, device_t dev, int index, int irq)
+{
+
+	return (msix_remap(index, irq));
+}
+
+static int
 nexus_release_msix(device_t pcib, device_t dev, int irq)
 {
 
diff --git a/sys/amd64/include/intr_machdep.h b/sys/amd64/include/intr_machdep.h
index 1dca25b..c0054b1 100644
--- a/sys/amd64/include/intr_machdep.h
+++ b/sys/amd64/include/intr_machdep.h
@@ -152,6 +152,7 @@ int	msi_alloc(device_t dev, int count, int maxcount, int *irqs, int *newirq,
 void	msi_init(void);
 int	msi_release(int *irqs, int count);
 int	msix_alloc(device_t dev, int index, int *irq, int *new);
+int	msix_remap(int index, int irq);
 int	msix_release(int irq);
 
 #endif	/* !LOCORE */
diff --git a/sys/amd64/pci/pci_bus.c b/sys/amd64/pci/pci_bus.c
index 5384831..37a4c6a 100644
--- a/sys/amd64/pci/pci_bus.c
+++ b/sys/amd64/pci/pci_bus.c
@@ -347,6 +347,7 @@ static device_method_t legacy_pcib_methods[] = {
 	DEVMETHOD(pcib_alloc_msi,	legacy_pcib_alloc_msi),
 	DEVMETHOD(pcib_release_msi,	pcib_release_msi),
 	DEVMETHOD(pcib_alloc_msix,	legacy_pcib_alloc_msix),
+	DEVMETHOD(pcib_remap_msix,	pcib_remap_msix),
 	DEVMETHOD(pcib_release_msix,	pcib_release_msix),
 
 	{ 0, 0 }