5.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Gerd Bayer <gbayer@xxxxxxxxxxxxx>
[ Upstream commit ab42fcb511fd9d241bbab7cc3ca04e34e9fc0666 ]
On a PCI adapter that provides up to 8 MSI interrupt sources the s390
implementation of PCI interrupts rejected to accommodate them, although
the underlying hardware is able to support that.
For MSI-X it is sufficient to allocate a single irq_desc per msi_desc,
but for MSI multiple irq descriptors are attached to and controlled by
a single msi descriptor. Add the appropriate loops to maintain multiple
irq descriptors and tie/untie them to/from the appropriate AIBV bit, if
a device driver allocates more than 1 MSI interrupt.
Common PCI code passes on requests to allocate a number of interrupt
vectors based on the device drivers' demand and the PCI functions'
capabilities. However, the root-complex of s390 systems support just a
limited number of interrupt vectors per PCI function.
Produce a kernel log message to inform about any architecture-specific
capping that might be done.
With this change, we had a PCI adapter successfully raising
interrupts to its device driver via all 8 sources.
Fixes: a384c8924a8b ("s390/PCI: Fix single MSI only check")
Signed-off-by: Gerd Bayer <gbayer@xxxxxxxxxxxxx>
Reviewed-by: Niklas Schnelle <schnelle@xxxxxxxxxxxxx>
Signed-off-by: Vasily Gorbik <gor@xxxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
arch/s390/pci/pci_irq.c | 62 ++++++++++++++++++++++++++++-------------
1 file changed, 42 insertions(+), 20 deletions(-)
diff --git a/arch/s390/pci/pci_irq.c b/arch/s390/pci/pci_irq.c
index 39c3c29f0d1d3..4a1dfce1a5cd2 100644
--- a/arch/s390/pci/pci_irq.c
+++ b/arch/s390/pci/pci_irq.c
@@ -292,8 +292,8 @@ static int __alloc_airq(struct zpci_dev *zdev, int msi_vecs,
int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
{
+ unsigned int hwirq, msi_vecs, irqs_per_msi, i, cpu;
struct zpci_dev *zdev = to_zpci(pdev);
- unsigned int hwirq, msi_vecs, cpu;
struct msi_desc *msi;
struct msi_msg msg;
unsigned long bit;
@@ -303,30 +303,46 @@ int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
zdev->aisb = -1UL;
zdev->msi_first_bit = -1U;
- if (type == PCI_CAP_ID_MSI && nvec > 1)
- return 1;
msi_vecs = min_t(unsigned int, nvec, zdev->max_msi);
+ if (msi_vecs < nvec) {
+ pr_info("%s requested %d irqs, allocate system limit of %d",
+ pci_name(pdev), nvec, zdev->max_msi);
+ }
rc = __alloc_airq(zdev, msi_vecs, &bit);
if (rc < 0)
return rc;
- /* Request MSI interrupts */
+ /*
+ * Request MSI interrupts:
+ * When using MSI, nvec_used interrupt sources and their irq
+ * descriptors are controlled through one msi descriptor.
+ * Thus the outer loop over msi descriptors shall run only once,
+ * while two inner loops iterate over the interrupt vectors.
+ * When using MSI-X, each interrupt vector/irq descriptor
+ * is bound to exactly one msi descriptor (nvec_used is one).
+ * So the inner loops are executed once, while the outer iterates
+ * over the MSI-X descriptors.
+ */
hwirq = bit;
msi_for_each_desc(msi, &pdev->dev, MSI_DESC_NOTASSOCIATED) {
- rc = -EIO;
if (hwirq - bit >= msi_vecs)
break;
- irq = __irq_alloc_descs(-1, 0, 1, 0, THIS_MODULE,
- (irq_delivery == DIRECTED) ?
- msi->affinity : NULL);
+ irqs_per_msi = min_t(unsigned int, msi_vecs, msi->nvec_used);
+ irq = __irq_alloc_descs(-1, 0, irqs_per_msi, 0, THIS_MODULE,
+ (irq_delivery == DIRECTED) ?
+ msi->affinity : NULL);
if (irq < 0)
return -ENOMEM;
- rc = irq_set_msi_desc(irq, msi);
- if (rc)
- return rc;
- irq_set_chip_and_handler(irq, &zpci_irq_chip,
- handle_percpu_irq);
+
+ for (i = 0; i < irqs_per_msi; i++) {
+ rc = irq_set_msi_desc_off(irq, i, msi);
+ if (rc)
+ return rc;
+ irq_set_chip_and_handler(irq + i, &zpci_irq_chip,
+ handle_percpu_irq);
+ }
+
msg.data = hwirq - bit;
if (irq_delivery == DIRECTED) {
if (msi->affinity)
@@ -339,31 +355,35 @@ int arch_setup_msi_irqs(struct pci_dev *pdev, int nvec, int type)
msg.address_lo |= (cpu_addr << 8);
for_each_possible_cpu(cpu) {
- airq_iv_set_data(zpci_ibv[cpu], hwirq, irq);
+ for (i = 0; i < irqs_per_msi; i++)
+ airq_iv_set_data(zpci_ibv[cpu],
+ hwirq + i, irq + i);
}
} else {
msg.address_lo = zdev->msi_addr & 0xffffffff;
- airq_iv_set_data(zdev->aibv, hwirq, irq);
+ for (i = 0; i < irqs_per_msi; i++)
+ airq_iv_set_data(zdev->aibv, hwirq + i, irq + i);
}
msg.address_hi = zdev->msi_addr >> 32;
pci_write_msi_msg(irq, &msg);
- hwirq++;
+ hwirq += irqs_per_msi;
}
zdev->msi_first_bit = bit;
- zdev->msi_nr_irqs = msi_vecs;
+ zdev->msi_nr_irqs = hwirq - bit;
rc = zpci_set_irq(zdev);
if (rc)
return rc;
- return (msi_vecs == nvec) ? 0 : msi_vecs;
+ return (zdev->msi_nr_irqs == nvec) ? 0 : zdev->msi_nr_irqs;
}
void arch_teardown_msi_irqs(struct pci_dev *pdev)
{
struct zpci_dev *zdev = to_zpci(pdev);
struct msi_desc *msi;
+ unsigned int i;
int rc;
/* Disable interrupts */
@@ -373,8 +393,10 @@ void arch_teardown_msi_irqs(struct pci_dev *pdev)
/* Release MSI interrupts */
msi_for_each_desc(msi, &pdev->dev, MSI_DESC_ASSOCIATED) {
- irq_set_msi_desc(msi->irq, NULL);
- irq_free_desc(msi->irq);
+ for (i = 0; i < msi->nvec_used; i++) {
+ irq_set_msi_desc(msi->irq + i, NULL);
+ irq_free_desc(msi->irq + i);
+ }
msi->msg.address_lo = 0;
msi->msg.address_hi = 0;
msi->msg.data = 0;
--
2.43.0
