[Qemu-devel] [PULL 27/32] spapr: Move interrupt allocator to xics

[Alexander Graf <agraf@suse.de>]

From: Alexey Kardashevskiy <aik@ozlabs.ru>

The current allocator returns IRQ numbers from a pool and does not
support IRQs reuse in any form as it did not keep track of what it
previously returned, it only keeps the last returned IRQ. Some use
cases such as PCI hot(un)plug may require IRQ release and reallocation.

This moves an allocator from SPAPR to XICS.

This switches IRQ users to use new API.

This uses LSI/MSI flags to know if interrupt is allocated.

The interrupt release function will be posted as a separate patch.

Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Alexander Graf <agraf@suse.de>
---
 hw/intc/xics.c         | 88 ++++++++++++++++++++++++++++++++++++++++++++++++++
 hw/ppc/spapr.c         | 67 --------------------------------------
 hw/ppc/spapr_events.c  |  2 +-
 hw/ppc/spapr_pci.c     |  6 ++--
 hw/ppc/spapr_vio.c     |  2 +-
 include/hw/ppc/spapr.h | 10 ------
 include/hw/ppc/xics.h  |  2 ++
 trace-events           |  4 +++
 8 files changed, 99 insertions(+), 82 deletions(-)

diff --git a/hw/intc/xics.c b/hw/intc/xics.c
index 634101a..6cd980a 100644
--- a/hw/intc/xics.c
+++ b/hw/intc/xics.c
@@ -687,6 +687,94 @@ void xics_set_irq_type(XICSState *icp, int irq, bool lsi)
     ics_set_irq_type(ics, irq - ics->offset, lsi);
 }
 
+#define ICS_IRQ_FREE(ics, srcno)   \
+    (!((ics)->irqs[(srcno)].flags & (XICS_FLAGS_IRQ_MASK)))
+
+static int ics_find_free_block(ICSState *ics, int num, int alignnum)
+{
+    int first, i;
+
+    for (first = 0; first < ics->nr_irqs; first += alignnum) {
+        if (num > (ics->nr_irqs - first)) {
+            return -1;
+        }
+        for (i = first; i < first + num; ++i) {
+            if (!ICS_IRQ_FREE(ics, i)) {
+                break;
+            }
+        }
+        if (i == (first + num)) {
+            return first;
+        }
+    }
+
+    return -1;
+}
+
+int xics_alloc(XICSState *icp, int src, int irq_hint, bool lsi)
+{
+    ICSState *ics = &icp->ics[src];
+    int irq;
+
+    if (irq_hint) {
+        assert(src == xics_find_source(icp, irq_hint));
+        if (!ICS_IRQ_FREE(ics, irq_hint - ics->offset)) {
+            trace_xics_alloc_failed_hint(src, irq_hint);
+            return -1;
+        }
+        irq = irq_hint;
+    } else {
+        irq = ics_find_free_block(ics, 1, 1);
+        if (irq < 0) {
+            trace_xics_alloc_failed_no_left(src);
+            return -1;
+        }
+        irq += ics->offset;
+    }
+
+    ics_set_irq_type(ics, irq - ics->offset, lsi);
+    trace_xics_alloc(src, irq);
+
+    return irq;
+}
+
+/*
+ * Allocate block of consequtive IRQs, returns a number of the first.
+ * If align==true, aligns the first IRQ number to num.
+ */
+int xics_alloc_block(XICSState *icp, int src, int num, bool lsi, bool align)
+{
+    int i, first = -1;
+    ICSState *ics = &icp->ics[src];
+
+    assert(src == 0);
+    /*
+     * MSIMesage::data is used for storing VIRQ so
+     * it has to be aligned to num to support multiple
+     * MSI vectors. MSI-X is not affected by this.
+     * The hint is used for the first IRQ, the rest should
+     * be allocated continuously.
+     */
+    if (align) {
+        assert((num == 1) || (num == 2) || (num == 4) ||
+               (num == 8) || (num == 16) || (num == 32));
+        first = ics_find_free_block(ics, num, num);
+    } else {
+        first = ics_find_free_block(ics, num, 1);
+    }
+
+    if (first >= 0) {
+        for (i = first; i < first + num; ++i) {
+            ics_set_irq_type(ics, i, lsi);
+        }
+    }
+    first += ics->offset;
+
+    trace_xics_alloc_block(src, first, num, lsi, align);
+
+    return first;
+}
+
 /*
  * Guest interfaces
  */
diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
index 8d14f6b..ea9bda9 100644
--- a/hw/ppc/spapr.c
+++ b/hw/ppc/spapr.c
@@ -104,73 +104,6 @@ struct sPAPRMachineState {
 
 sPAPREnvironment *spapr;
 
-int spapr_allocate_irq(int hint, bool lsi)
-{
-    int irq;
-
-    if (hint) {
-        irq = hint;
-        if (hint >= spapr->next_irq) {
-            spapr->next_irq = hint + 1;
-        }
-        /* FIXME: we should probably check for collisions somehow */
-    } else {
-        irq = spapr->next_irq++;
-    }
-
-    /* Configure irq type */
-    if (!xics_get_qirq(spapr->icp, irq)) {
-        return 0;
-    }
-
-    xics_set_irq_type(spapr->icp, irq, lsi);
-
-    return irq;
-}
-
-/*
- * Allocate block of consequtive IRQs, returns a number of the first.
- * If msi==true, aligns the first IRQ number to num.
- */
-int spapr_allocate_irq_block(int num, bool lsi, bool msi)
-{
-    int first = -1;
-    int i, hint = 0;
-
-    /*
-     * MSIMesage::data is used for storing VIRQ so
-     * it has to be aligned to num to support multiple
-     * MSI vectors. MSI-X is not affected by this.
-     * The hint is used for the first IRQ, the rest should
-     * be allocated continuously.
-     */
-    if (msi) {
-        assert((num == 1) || (num == 2) || (num == 4) ||
-               (num == 8) || (num == 16) || (num == 32));
-        hint = (spapr->next_irq + num - 1) & ~(num - 1);
-    }
-
-    for (i = 0; i < num; ++i) {
-        int irq;
-
-        irq = spapr_allocate_irq(hint, lsi);
-        if (!irq) {
-            return -1;
-        }
-
-        if (0 == i) {
-            first = irq;
-            hint = 0;
-        }
-
-        /* If the above doesn't create a consecutive block then that's
-         * an internal bug */
-        assert(irq == (first + i));
-    }
-
-    return first;
-}
-
 static XICSState *try_create_xics(const char *type, int nr_servers,
                                   int nr_irqs)
 {
diff --git a/hw/ppc/spapr_events.c b/hw/ppc/spapr_events.c
index 5ce96a7..1b6157d 100644
--- a/hw/ppc/spapr_events.c
+++ b/hw/ppc/spapr_events.c
@@ -314,7 +314,7 @@ static void check_exception(PowerPCCPU *cpu, sPAPREnvironment *spapr,
 
 void spapr_events_init(sPAPREnvironment *spapr)
 {
-    spapr->epow_irq = spapr_allocate_msi(0);
+    spapr->epow_irq = xics_alloc(spapr->icp, 0, 0, false);
     spapr->epow_notifier.notify = spapr_powerdown_req;
     qemu_register_powerdown_notifier(&spapr->epow_notifier);
     spapr_rtas_register(RTAS_CHECK_EXCEPTION, "check-exception",
diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c
index d1e3e0f..d115fca 100644
--- a/hw/ppc/spapr_pci.c
+++ b/hw/ppc/spapr_pci.c
@@ -360,8 +360,8 @@ static void rtas_ibm_change_msi(PowerPCCPU *cpu, sPAPREnvironment *spapr,
 
     /* There is no cached config, allocate MSIs */
     if (!phb->msi_table[ndev].nvec) {
-        irq = spapr_allocate_irq_block(req_num, false,
-                                       ret_intr_type == RTAS_TYPE_MSI);
+        irq = xics_alloc_block(spapr->icp, 0, req_num, false,
+                               ret_intr_type == RTAS_TYPE_MSI);
         if (irq < 0) {
             error_report("Cannot allocate MSIs for device#%d", ndev);
             rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
@@ -634,7 +634,7 @@ static void spapr_phb_realize(DeviceState *dev, Error **errp)
     for (i = 0; i < PCI_NUM_PINS; i++) {
         uint32_t irq;
 
-        irq = spapr_allocate_lsi(0);
+        irq = xics_alloc_block(spapr->icp, 0, 1, true, false);
         if (!irq) {
             error_setg(errp, "spapr_allocate_lsi failed");
             return;
diff --git a/hw/ppc/spapr_vio.c b/hw/ppc/spapr_vio.c
index 8b765c6..dc9e46a 100644
--- a/hw/ppc/spapr_vio.c
+++ b/hw/ppc/spapr_vio.c
@@ -449,7 +449,7 @@ static int spapr_vio_busdev_init(DeviceState *qdev)
         dev->qdev.id = id;
     }
 
-    dev->irq = spapr_allocate_msi(dev->irq);
+    dev->irq = xics_alloc(spapr->icp, 0, dev->irq, false);
     if (!dev->irq) {
         return -1;
     }
diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
index 3b6ccd3..6b2f21c 100644
--- a/include/hw/ppc/spapr.h
+++ b/include/hw/ppc/spapr.h
@@ -339,16 +339,6 @@ target_ulong spapr_hypercall(PowerPCCPU *cpu, target_ulong opcode,
 int spapr_allocate_irq(int hint, bool lsi);
 int spapr_allocate_irq_block(int num, bool lsi, bool msi);
 
-static inline int spapr_allocate_msi(int hint)
-{
-    return spapr_allocate_irq(hint, false);
-}
-
-static inline int spapr_allocate_lsi(int hint)
-{
-    return spapr_allocate_irq(hint, true);
-}
-
 /* RTAS return codes */
 #define RTAS_OUT_SUCCESS            0
 #define RTAS_OUT_NO_ERRORS_FOUND    1
diff --git a/include/hw/ppc/xics.h b/include/hw/ppc/xics.h
index 2891599..30b1441 100644
--- a/include/hw/ppc/xics.h
+++ b/include/hw/ppc/xics.h
@@ -160,6 +160,8 @@ struct ICSIRQState {
 
 qemu_irq xics_get_qirq(XICSState *icp, int irq);
 void xics_set_irq_type(XICSState *icp, int irq, bool lsi);
+int xics_alloc(XICSState *icp, int src, int irq_hint, bool lsi);
+int xics_alloc_block(XICSState *icp, int src, int num, bool lsi, bool align);
 
 void xics_cpu_setup(XICSState *icp, PowerPCCPU *cpu);
 
diff --git a/trace-events b/trace-events
index ba01ad5..2a46e6c 100644
--- a/trace-events
+++ b/trace-events
@@ -1188,6 +1188,10 @@ xics_set_irq_lsi(int srcno, int nr) "set_irq_lsi: srcno %d [irq %#x]"
 xics_ics_write_xive(int nr, int srcno, int server, uint8_t priority) "ics_write_xive: irq %#x [src %d] server %#x prio %#x"
 xics_ics_reject(int nr, int srcno) "reject irq %#x [src %d]"
 xics_ics_eoi(int nr) "ics_eoi: irq %#x"
+xics_alloc(int src, int irq) "source#%d, irq %d"
+xics_alloc_failed_hint(int src, int irq) "source#%d, irq %d is already in use"
+xics_alloc_failed_no_left(int src) "source#%d, no irq left"
+xics_alloc_block(int src, int first, int num, bool lsi, int align) "source#%d, first irq %d, %d irqs, lsi=%d, alignnum %d"
 
 # hw/ppc/spapr.c
 spapr_cas_failed(unsigned long n) "DT diff buffer is too small: %ld bytes"
-- 
1.8.1.4