qemu-kvm/kvm-spapr-pci-Consolidate-de-allocation-of-MSIs.patch
Danilo C. L. de Paula 200e3560ab * Tue Sep 10 2019 Danilo Cesar Lemes de Paula <ddepaula@redhat.com> - 4.1.0-9.el8
- kvm-migration-always-initialise-ram_counters-for-a-new-m.patch [bz#1734316]
- kvm-migration-add-qemu_file_update_transfer-interface.patch [bz#1734316]
- kvm-migration-add-speed-limit-for-multifd-migration.patch [bz#1734316]
- kvm-migration-update-ram_counters-for-multifd-sync-packe.patch [bz#1734316]
- kvm-spapr-pci-Consolidate-de-allocation-of-MSIs.patch [bz#1750200]
- kvm-spapr-pci-Free-MSIs-during-reset.patch [bz#1750200]
- Resolves: bz#1734316
  (multifd migration does not honour speed limits, consumes entire bandwidth of NIC)
- Resolves: bz#1750200
  ([RHEL8.1][QEMU4.1]boot up guest with vf device,then system_reset guest,error prompt(qemu-kvm: Can't allocate MSIs for device 2800: IRQ 4904 is not free))
2019-09-10 19:56:34 +01:00

104 lines
3.7 KiB
Diff

From b27062f4b3ddf47dea926026e5511f15d5b31320 Mon Sep 17 00:00:00 2001
From: David Gibson <dgibson@redhat.com>
Date: Tue, 10 Sep 2019 07:04:27 +0100
Subject: [PATCH 5/6] spapr/pci: Consolidate de-allocation of MSIs
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: David Gibson <dgibson@redhat.com>
Message-id: <20190910070428.28628-2-dgibson@redhat.com>
Patchwork-id: 90362
O-Subject: [RHEL-AV-8.1 qemu-kvm PATCH 1/2] spapr/pci: Consolidate de-allocation of MSIs
Bugzilla: 1750200
RH-Acked-by: Thomas Huth <thuth@redhat.com>
RH-Acked-by: Philippe Mathieu-Daudé <philmd@redhat.com>
RH-Acked-by: Laurent Vivier <lvivier@redhat.com>
From: Greg Kurz <groug@kaod.org>
When freeing MSIs, we need to:
- remove them from the machine's MSI bitmap
- remove them from the IC backend
- remove them from the PHB's MSI cache
This is currently open coded in two places in rtas_ibm_change_msi(),
and we're about to need this in spapr_phb_reset() as well. Instead of
duplicating this code again, make it a destroy function for the PHB's
MSI cache. Removing an MSI device from the cache will call the destroy
function internally.
Signed-off-by: Greg Kurz <groug@kaod.org>
Message-Id: <156415227855.1064338.5657793835271464648.stgit@bahia.lan>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
(cherry picked from commit 078eb6b05b7f962e43d8bc376e0b96cdd550c17a)
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1750200
Signed-off-by: David Gibson <dgibson@redhat.com>
Signed-off-by: Danilo C. L. de Paula <ddepaula@redhat.com>
---
hw/ppc/spapr_pci.c | 24 +++++++++++++++---------
1 file changed, 15 insertions(+), 9 deletions(-)
diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c
index 9003fe9..1ffcfae 100644
--- a/hw/ppc/spapr_pci.c
+++ b/hw/ppc/spapr_pci.c
@@ -336,10 +336,6 @@ static void rtas_ibm_change_msi(PowerPCCPU *cpu, SpaprMachineState *spapr,
return;
}
- if (!smc->legacy_irq_allocation) {
- spapr_irq_msi_free(spapr, msi->first_irq, msi->num);
- }
- spapr_irq_free(spapr, msi->first_irq, msi->num);
if (msi_present(pdev)) {
spapr_msi_setmsg(pdev, 0, false, 0, 0);
}
@@ -409,10 +405,6 @@ static void rtas_ibm_change_msi(PowerPCCPU *cpu, SpaprMachineState *spapr,
/* Release previous MSIs */
if (msi) {
- if (!smc->legacy_irq_allocation) {
- spapr_irq_msi_free(spapr, msi->first_irq, msi->num);
- }
- spapr_irq_free(spapr, msi->first_irq, msi->num);
g_hash_table_remove(phb->msi, &config_addr);
}
@@ -1806,6 +1798,19 @@ static void spapr_phb_unrealize(DeviceState *dev, Error **errp)
memory_region_del_subregion(get_system_memory(), &sphb->mem32window);
}
+static void spapr_phb_destroy_msi(gpointer opaque)
+{
+ SpaprMachineState *spapr = SPAPR_MACHINE(qdev_get_machine());
+ SpaprMachineClass *smc = SPAPR_MACHINE_GET_CLASS(spapr);
+ spapr_pci_msi *msi = opaque;
+
+ if (!smc->legacy_irq_allocation) {
+ spapr_irq_msi_free(spapr, msi->first_irq, msi->num);
+ }
+ spapr_irq_free(spapr, msi->first_irq, msi->num);
+ g_free(msi);
+}
+
static void spapr_phb_realize(DeviceState *dev, Error **errp)
{
/* We don't use SPAPR_MACHINE() in order to exit gracefully if the user
@@ -2017,7 +2022,8 @@ static void spapr_phb_realize(DeviceState *dev, Error **errp)
spapr_tce_get_iommu(tcet));
}
- sphb->msi = g_hash_table_new_full(g_int_hash, g_int_equal, g_free, g_free);
+ sphb->msi = g_hash_table_new_full(g_int_hash, g_int_equal, g_free,
+ spapr_phb_destroy_msi);
return;
unrealize:
--
1.8.3.1