rpms/qemu-kvm
7
0
Fork 0
Code Issues Pull Requests Releases Activity

* Tue Jul 08 2025 Miroslav Rezanina <mrezanin@redhat.com> - 9.1.0-25

Browse Source 
- kvm-s390x-Fix-leak-in-machine_set_loadparm.patch [RHEL-98554]
- kvm-hw-s390x-ccw-device-Fix-memory-leak-in-loadparm-sett.patch [RHEL-98554]
- kvm-amd_iommu-Rename-variable-mmio-to-mr_mmio.patch [RHEL-66202]
- kvm-amd_iommu-Add-support-for-pass-though-mode.patch [RHEL-66202]
- kvm-amd_iommu-Use-shared-memory-region-for-Interrupt-Rem.patch [RHEL-66202]
- kvm-amd_iommu-Send-notification-when-invalidate-interrup.patch [RHEL-66202]
- kvm-amd_iommu-Check-APIC-ID-255-for-XTSup.patch [RHEL-66202]
- kvm-io-Fix-partial-struct-copy-in-qio_dns_resolver_looku.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-setting-client-sockopts-i.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-success-and-failure-paths.patch [RHEL-67104]
- kvm-util-qemu-sockets-Add-support-for-keep-alive-flag-to.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-inet_parse-to-use-QemuOpt.patch [RHEL-67104]
- kvm-util-qemu-sockets-Introduce-inet-socket-options-cont.patch [RHEL-67104]
- kvm-tests-unit-test-util-sockets-fix-mem-leak-on-error-o.patch [RHEL-67104]
- kvm-target-i386-Expose-bits-related-to-SRSO-vulnerabilit.patch [RHEL-52649]
- kvm-target-i386-Add-PerfMonV2-feature-bit.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-CPU-model-for-Cache-property.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Rome-CPU-model-for-Cache-pro.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Milan-CPU-model-for-Cache-pr.patch [RHEL-52649]
- kvm-target-i386-Add-couple-of-feature-bits-in-CPUID_Fn80.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Genoa-for-Cache-property-per.patch [RHEL-52649]
- kvm-target-i386-Add-support-for-EPYC-Turin-model.patch [RHEL-52649]
- kvm-hw-i386-amd_iommu-Assign-pci-id-0x1419-for-the-AMD-I.patch [RHEL-70926]
- kvm-hw-i386-amd_iommu-Isolate-AMDVI-PCI-from-amd-iommu-d.patch [RHEL-70925]
- kvm-hw-i386-amd_iommu-Allow-migration-when-explicitly-cr.patch [RHEL-70925]
- kvm-Enable-amd-iommu-device.patch [RHEL-70925]
- kvm-include-qemu-compiler-add-QEMU_UNINITIALIZED-attribu.patch [RHEL-99888]
- kvm-hw-virtio-virtio-avoid-cost-of-ftrivial-auto-var-ini.patch [RHEL-99888]
- kvm-block-skip-automatic-zero-init-of-large-array-in-ioq.patch [RHEL-99888]
- kvm-chardev-char-fd-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-chardev-char-pty-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-chardev-char-socket-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-audio-ac97-skip-automatic-zero-init-of-large-arra.patch [RHEL-99888]
- kvm-hw-audio-cs4231a-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-hw-audio-es1370-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-audio-gus-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-audio-marvell_88w8618-skip-automatic-zero-init-of.patch [RHEL-99888]
- kvm-hw-audio-sb16-skip-automatic-zero-init-of-large-arra.patch [RHEL-99888]
- kvm-hw-audio-via-ac97-skip-automatic-zero-init-of-large-.patch [RHEL-99888]
- kvm-hw-char-sclpconsole-lm-skip-automatic-zero-init-of-l.patch [RHEL-99888]
- kvm-hw-dma-xlnx_csu_dma-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-display-vmware_vga-skip-automatic-zero-init-of-la.patch [RHEL-99888]
- kvm-hw-hyperv-syndbg-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-hw-misc-aspeed_hace-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-net-rtl8139-skip-automatic-zero-init-of-large-arr.patch [RHEL-99888]
- kvm-hw-net-tulip-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-net-virtio-net-skip-automatic-zero-init-of-large-.patch [RHEL-99888]
- kvm-hw-net-xgamc-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-nvme-ctrl-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-ppc-spapr_tpm_proxy-skip-automatic-zero-init-of-l.patch [RHEL-99888]
- kvm-hw-usb-hcd-ohci-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-scsi-lsi53c895a-skip-automatic-zero-init-of-large.patch [RHEL-99888]
- kvm-hw-scsi-megasas-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-ufs-lu-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-net-socket-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-net-stream-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-ui-vnc-Update-display-update-interval-when-VM-state-.patch [RHEL-100741]
- Resolves: RHEL-98554
  ([s390x][RHEL9.7.0][virtio_block] there would be memory leak with virtio_blk disks)
- Resolves: RHEL-66202
  ([AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu)
- Resolves: RHEL-67104
  (postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S')
- Resolves: RHEL-52649
  ([AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model)
- Resolves: RHEL-70926
  (Qemu/amd-iommu: Advertise a suitable device id)
- Resolves: RHEL-70925
  (Qemu/amd-iommu: Add ability to manually specify the AMDVI-PCI device)
- Resolves: RHEL-99888
  (-ftrivial-auto-var-init=zero reduced performance [rhel-9])
- Resolves: RHEL-100741
  (Video stuck after switchover phase when play one video during migration [rhel-9])
This commit is contained in:
Miroslav Rezanina 2025-07-08 02:53:14 -04:00
parent 0dc0c63dbf
commit bd7b93c113
58 changed files with 5256 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
kvm-Enable-amd-iommu-device.patch Normal file
View File

@ -0,0 +1,38 @@
From 0608561efc441f234d9aaf45f1867ffb5c43cffe Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Jun 2025 15:41:14 -0500
Subject: [PATCH 26/57] Enable amd-iommu device
RH-Author: John Allen <None>
RH-MergeRequest: 380: Add ability to manually specify the AMDVI-PCI device
RH-Jira: RHEL-70925
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/3] 852500a18275e14bcd94d598ccd0ee33b76578dc (johnalle/qemu-kvm-fork)
Now that the amdvi-pci device that amd-iommu creates can be specified
manually, amd-iommu device can be enabled.
JIRA: https://issues.redhat.com/browse/RHEL-70925
Upstream: RHEL ONLY
Signed-off-by: John Allen <johnalle@redhat.com>
---
configs/devices/x86_64-softmmu/x86_64-rh-devices.mak | 1 +
1 file changed, 1 insertion(+)
diff --git a/configs/devices/x86_64-softmmu/x86_64-rh-devices.mak b/configs/devices/x86_64-softmmu/x86_64-rh-devices.mak
index 3e5f693b62..2b15fdc2db 100644
--- a/configs/devices/x86_64-softmmu/x86_64-rh-devices.mak
+++ b/configs/devices/x86_64-softmmu/x86_64-rh-devices.mak
@@ -97,6 +97,7 @@ CONFIG_VIRTIO_MEM=y
CONFIG_VIRTIO_PCI=y
CONFIG_VIRTIO_VGA=y
CONFIG_VIRTIO_IOMMU=y
+CONFIG_AMD_IOMMU=y
CONFIG_VMMOUSE=y
CONFIG_VMPORT=y
CONFIG_VTD=y
--
2.39.3

141
kvm-amd_iommu-Add-support-for-pass-though-mode.patch Normal file
View File

@ -0,0 +1,141 @@
From 4114553452f7187283aefa001bc8342fc65b6b72 Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Dec 2024 15:06:48 -0600
Subject: [PATCH 04/57] amd_iommu: Add support for pass though mode
RH-Author: John Allen <None>
RH-MergeRequest: 303: Interrupt Remap support for emulated amd viommu
RH-Jira: RHEL-66202
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/5] 0434fefd554baf27fb9d93026af513c621f8cdb0 (johnalle/qemu-kvm-fork)
JIRA: https://issues.redhat.com/browse/RHEL-66202
commit c1f46999ef506d9854534560a94d02cf3cf9edd1
Author: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Fri Sep 27 12:29:10 2024 -0500
amd_iommu: Add support for pass though mode
Introduce 'nodma' shared memory region to support PT mode
so that for each device, we only create an alias to shared memory
region when DMA-remapping is disabled.
Reviewed-by: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Santosh Shukla <santosh.shukla@amd.com>
Message-Id: <20240927172913.121477-3-santosh.shukla@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: John Allen <john.allen@amd.com>
---
hw/i386/amd_iommu.c | 49 ++++++++++++++++++++++++++++++++++++---------
hw/i386/amd_iommu.h | 2 ++
2 files changed, 42 insertions(+), 9 deletions(-)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 148b5ee51d..567cb8adc9 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -60,8 +60,9 @@ struct AMDVIAddressSpace {
uint8_t bus_num; /* bus number */
uint8_t devfn; /* device function */
AMDVIState *iommu_state; /* AMDVI - one per machine */
- MemoryRegion root; /* AMDVI Root memory map region */
+ MemoryRegion root; /* AMDVI Root memory map region */
IOMMUMemoryRegion iommu; /* Device's address translation region */
+ MemoryRegion iommu_nodma; /* Alias of shared nodma memory region */
MemoryRegion iommu_ir; /* Device's interrupt remapping region */
AddressSpace as; /* device's corresponding address space */
};
@@ -1412,6 +1413,7 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
AMDVIState *s = opaque;
AMDVIAddressSpace **iommu_as, *amdvi_dev_as;
int bus_num = pci_bus_num(bus);
+ X86IOMMUState *x86_iommu = X86_IOMMU_DEVICE(s);
iommu_as = s->address_spaces[bus_num];
@@ -1436,13 +1438,13 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
* Memory region relationships looks like (Address range shows
* only lower 32 bits to make it short in length...):
*
- * |-----------------+-------------------+----------|
- * | Name | Address range | Priority |
- * |-----------------+-------------------+----------+
- * | amdvi_root | 00000000-ffffffff | 0 |
- * | amdvi_iommu | 00000000-ffffffff | 1 |
- * | amdvi_iommu_ir | fee00000-feefffff | 64 |
- * |-----------------+-------------------+----------|
+ * |--------------------+-------------------+----------|
+ * | Name | Address range | Priority |
+ * |--------------------+-------------------+----------+
+ * | amdvi-root | 00000000-ffffffff | 0 |
+ * | amdvi-iommu_nodma | 00000000-ffffffff | 0 |
+ * | amdvi-iommu_ir | fee00000-feefffff | 64 |
+ * |--------------------+-------------------+----------|
*/
memory_region_init_iommu(&amdvi_dev_as->iommu,
sizeof(amdvi_dev_as->iommu),
@@ -1461,7 +1463,25 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
64);
memory_region_add_subregion_overlap(&amdvi_dev_as->root, 0,
MEMORY_REGION(&amdvi_dev_as->iommu),
- 1);
+ 0);
+
+ /* Build the DMA Disabled alias to shared memory */
+ memory_region_init_alias(&amdvi_dev_as->iommu_nodma, OBJECT(s),
+ "amdvi-sys", &s->mr_sys, 0,
+ memory_region_size(&s->mr_sys));
+ memory_region_add_subregion_overlap(&amdvi_dev_as->root, 0,
+ &amdvi_dev_as->iommu_nodma,
+ 0);
+
+ if (!x86_iommu->pt_supported) {
+ memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, false);
+ memory_region_set_enabled(MEMORY_REGION(&amdvi_dev_as->iommu),
+ true);
+ } else {
+ memory_region_set_enabled(MEMORY_REGION(&amdvi_dev_as->iommu),
+ false);
+ memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, true);
+ }
}
return &iommu_as[devfn]->as;
}
@@ -1602,6 +1622,17 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
"amdvi-mmio", AMDVI_MMIO_SIZE);
memory_region_add_subregion(get_system_memory(), AMDVI_BASE_ADDR,
&s->mr_mmio);
+
+ /* Create the share memory regions by all devices */
+ memory_region_init(&s->mr_sys, OBJECT(s), "amdvi-sys", UINT64_MAX);
+
+ /* set up the DMA disabled memory region */
+ memory_region_init_alias(&s->mr_nodma, OBJECT(s),
+ "amdvi-nodma", get_system_memory(), 0,
+ memory_region_size(get_system_memory()));
+ memory_region_add_subregion_overlap(&s->mr_sys, 0,
+ &s->mr_nodma, 0);
+
pci_setup_iommu(bus, &amdvi_iommu_ops, s);
amdvi_init(s);
}
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index e5c2ae94f2..be417e51c4 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -354,6 +354,8 @@ struct AMDVIState {
uint32_t pprlog_tail; /* ppr log tail */
MemoryRegion mr_mmio; /* MMIO region */
+ MemoryRegion mr_sys;
+ MemoryRegion mr_nodma;
uint8_t mmior[AMDVI_MMIO_SIZE]; /* read/write MMIO */
uint8_t w1cmask[AMDVI_MMIO_SIZE]; /* read/write 1 clear mask */
uint8_t romask[AMDVI_MMIO_SIZE]; /* MMIO read/only mask */
--
2.39.3

66
kvm-amd_iommu-Check-APIC-ID-255-for-XTSup.patch Normal file
View File

@ -0,0 +1,66 @@
From 0397ebacdba6539147d9986255c3f81cbfdabf1e Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Dec 2024 15:07:03 -0600
Subject: [PATCH 07/57] amd_iommu: Check APIC ID > 255 for XTSup
RH-Author: John Allen <None>
RH-MergeRequest: 303: Interrupt Remap support for emulated amd viommu
RH-Jira: RHEL-66202
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [5/5] f39b3e3cdefc2b562f1ad2ef939a37bf404f355a (johnalle/qemu-kvm-fork)
JIRA: https://issues.redhat.com/browse/RHEL-66202
commit b12cb3819baf6d9ee8140d4dd6d36fa829e2c6d9
Author: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Fri Sep 27 12:29:13 2024 -0500
amd_iommu: Check APIC ID > 255 for XTSup
The XTSup mode enables x2APIC support for AMD IOMMU, which is needed
to support vcpu w/ APIC ID > 255.
Reviewed-by: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Santosh Shukla <santosh.shukla@amd.com>
Message-Id: <20240927172913.121477-6-santosh.shukla@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: John Allen <john.allen@amd.com>
---
hw/i386/amd_iommu.c | 11 +++++++++++
1 file changed, 11 insertions(+)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 82d76dfca9..d804656ea8 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -32,6 +32,7 @@
#include "trace.h"
#include "hw/i386/apic-msidef.h"
#include "hw/qdev-properties.h"
+#include "kvm/kvm_i386.h"
/* used AMD-Vi MMIO registers */
const char *amdvi_mmio_low[] = {
@@ -1651,6 +1652,16 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
memory_region_add_subregion_overlap(&s->mr_sys, AMDVI_INT_ADDR_FIRST,
&s->mr_ir, 1);
+ /* AMD IOMMU with x2APIC mode requires xtsup=on */
+ if (x86ms->apic_id_limit > 255 && !s->xtsup) {
+ error_report("AMD IOMMU with x2APIC confguration requires xtsup=on");
+ exit(EXIT_FAILURE);
+ }
+ if (s->xtsup && kvm_irqchip_is_split() && !kvm_enable_x2apic()) {
+ error_report("AMD IOMMU xtsup=on requires support on the KVM side");
+ exit(EXIT_FAILURE);
+ }
+
pci_setup_iommu(bus, &amdvi_iommu_ops, s);
amdvi_init(s);
}
--
2.39.3

94
kvm-amd_iommu-Rename-variable-mmio-to-mr_mmio.patch Normal file
View File

@ -0,0 +1,94 @@
From f733325d3d91576ae9f6e341faabc301542fc6c8 Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Dec 2024 15:06:44 -0600
Subject: [PATCH 03/57] amd_iommu: Rename variable mmio to mr_mmio
RH-Author: John Allen <None>
RH-MergeRequest: 303: Interrupt Remap support for emulated amd viommu
RH-Jira: RHEL-66202
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/5] 1996a48efb7210d4d1e0b929be2d115d672e1a02 (johnalle/qemu-kvm-fork)
JIRA: https://issues.redhat.com/browse/RHEL-66202
commit 2e6f051cfc58e69dcb392cd245d8f01b0c2e963f
Author: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Fri Sep 27 12:29:09 2024 -0500
amd_iommu: Rename variable mmio to mr_mmio
Rename the MMIO memory region variable 'mmio' to 'mr_mmio'
so to correctly name align with struct AMDVIState::variable type.
No functional change intended.
Reviewed-by: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Santosh Shukla <santosh.shukla@amd.com>
Message-Id: <20240927172913.121477-2-santosh.shukla@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: John Allen <john.allen@amd.com>
---
hw/i386/acpi-build.c | 4 ++--
hw/i386/amd_iommu.c | 6 +++---
hw/i386/amd_iommu.h | 2 +-
3 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
index 5d4bd2b710..032fb1f904 100644
--- a/hw/i386/acpi-build.c
+++ b/hw/i386/acpi-build.c
@@ -2397,7 +2397,7 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
/* Capability offset */
build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
/* IOMMU base address */
- build_append_int_noprefix(table_data, s->mmio.addr, 8);
+ build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
build_append_int_noprefix(table_data, 0, 2);
/* IOMMU info */
@@ -2432,7 +2432,7 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
/* Capability offset */
build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
/* IOMMU base address */
- build_append_int_noprefix(table_data, s->mmio.addr, 8);
+ build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
build_append_int_noprefix(table_data, 0, 2);
/* IOMMU info */
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 87643d2891..148b5ee51d 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -1598,10 +1598,10 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
x86ms->ioapic_as = amdvi_host_dma_iommu(bus, s, AMDVI_IOAPIC_SB_DEVID);
/* set up MMIO */
- memory_region_init_io(&s->mmio, OBJECT(s), &mmio_mem_ops, s, "amdvi-mmio",
- AMDVI_MMIO_SIZE);
+ memory_region_init_io(&s->mr_mmio, OBJECT(s), &mmio_mem_ops, s,
+ "amdvi-mmio", AMDVI_MMIO_SIZE);
memory_region_add_subregion(get_system_memory(), AMDVI_BASE_ADDR,
- &s->mmio);
+ &s->mr_mmio);
pci_setup_iommu(bus, &amdvi_iommu_ops, s);
amdvi_init(s);
}
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index 73619fe9ea..e5c2ae94f2 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -353,7 +353,7 @@ struct AMDVIState {
uint32_t pprlog_head; /* ppr log head */
uint32_t pprlog_tail; /* ppr log tail */
- MemoryRegion mmio; /* MMIO region */
+ MemoryRegion mr_mmio; /* MMIO region */
uint8_t mmior[AMDVI_MMIO_SIZE]; /* read/write MMIO */
uint8_t w1cmask[AMDVI_MMIO_SIZE]; /* read/write 1 clear mask */
uint8_t romask[AMDVI_MMIO_SIZE]; /* MMIO read/only mask */
--
2.39.3

81
kvm-amd_iommu-Send-notification-when-invalidate-interrup.patch Normal file
View File

@ -0,0 +1,81 @@
From 17ce6ac0d8edb04ba79bb39d3f695cd0506a9dc2 Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Dec 2024 15:06:59 -0600
Subject: [PATCH 06/57] amd_iommu: Send notification when invalidate interrupt
entry cache
RH-Author: John Allen <None>
RH-MergeRequest: 303: Interrupt Remap support for emulated amd viommu
RH-Jira: RHEL-66202
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [4/5] d57e8fb4e69f3c01d32673bf658aae5067d6b969 (johnalle/qemu-kvm-fork)
JIRA: https://issues.redhat.com/browse/RHEL-66202
commit f84aad4d718b83d2a4d90485992e5421430032e1
Author: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Fri Sep 27 12:29:12 2024 -0500
amd_iommu: Send notification when invalidate interrupt entry cache
In order to support AMD IOMMU interrupt remapping emulation with PCI
pass-through devices, QEMU needs to notify VFIO when guest IOMMU driver
updates and invalidate the guest interrupt remapping table (IRT), and
communicate information so that the host IOMMU driver can update
the shadowed interrupt remapping table in the host IOMMU.
Therefore, send notification when guest IOMMU emulates the IRT
invalidation commands.
Reviewed-by: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Santosh Shukla <santosh.shukla@amd.com>
Message-Id: <20240927172913.121477-5-santosh.shukla@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: John Allen <john.allen@amd.com>
---
hw/i386/amd_iommu.c | 12 ++++++++++++
1 file changed, 12 insertions(+)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 8fcf5eacb4..82d76dfca9 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -431,6 +431,12 @@ static void amdvi_complete_ppr(AMDVIState *s, uint64_t *cmd)
trace_amdvi_ppr_exec();
}
+static void amdvi_intremap_inval_notify_all(AMDVIState *s, bool global,
+ uint32_t index, uint32_t mask)
+{
+ x86_iommu_iec_notify_all(X86_IOMMU_DEVICE(s), global, index, mask);
+}
+
static void amdvi_inval_all(AMDVIState *s, uint64_t *cmd)
{
if (extract64(cmd[0], 0, 60) || cmd[1]) {
@@ -438,6 +444,9 @@ static void amdvi_inval_all(AMDVIState *s, uint64_t *cmd)
s->cmdbuf + s->cmdbuf_head);
}
+ /* Notify global invalidation */
+ amdvi_intremap_inval_notify_all(s, true, 0, 0);
+
amdvi_iotlb_reset(s);
trace_amdvi_all_inval();
}
@@ -486,6 +495,9 @@ static void amdvi_inval_inttable(AMDVIState *s, uint64_t *cmd)
return;
}
+ /* Notify global invalidation */
+ amdvi_intremap_inval_notify_all(s, true, 0, 0);
+
trace_amdvi_intr_inval();
}
--
2.39.3

105
kvm-amd_iommu-Use-shared-memory-region-for-Interrupt-Rem.patch Normal file
View File

@ -0,0 +1,105 @@
From 4859d41adfaae8933e074dcefdc81edd3832c914 Mon Sep 17 00:00:00 2001
From: John Allen <john.allen@amd.com>
Date: Wed, 11 Dec 2024 15:06:55 -0600
Subject: [PATCH 05/57] amd_iommu: Use shared memory region for Interrupt
Remapping
RH-Author: John Allen <None>
RH-MergeRequest: 303: Interrupt Remap support for emulated amd viommu
RH-Jira: RHEL-66202
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/5] 48c0513c80257bfbd12c2cf3bab2503bd95d0b1c (johnalle/qemu-kvm-fork)
JIRA: https://issues.redhat.com/browse/RHEL-66202
commit 9fc9dbac61ddde7d8df37e84c8e02cec249d3222
Author: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Fri Sep 27 12:29:11 2024 -0500
amd_iommu: Use shared memory region for Interrupt Remapping
Use shared memory region for interrupt remapping which can be
aliased by all devices.
Reviewed-by: Alejandro Jimenez <alejandro.j.jimenez@oracle.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: Santosh Shukla <santosh.shukla@amd.com>
Message-Id: <20240927172913.121477-4-santosh.shukla@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: John Allen <john.allen@amd.com>
---
hw/i386/amd_iommu.c | 22 ++++++++++++++--------
hw/i386/amd_iommu.h | 1 +
2 files changed, 15 insertions(+), 8 deletions(-)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 567cb8adc9..8fcf5eacb4 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -1443,7 +1443,7 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
* |--------------------+-------------------+----------+
* | amdvi-root | 00000000-ffffffff | 0 |
* | amdvi-iommu_nodma | 00000000-ffffffff | 0 |
- * | amdvi-iommu_ir | fee00000-feefffff | 64 |
+ * | amdvi-iommu_ir | fee00000-feefffff | 1 |
* |--------------------+-------------------+----------|
*/
memory_region_init_iommu(&amdvi_dev_as->iommu,
@@ -1454,13 +1454,6 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
memory_region_init(&amdvi_dev_as->root, OBJECT(s),
"amdvi_root", UINT64_MAX);
address_space_init(&amdvi_dev_as->as, &amdvi_dev_as->root, name);
- memory_region_init_io(&amdvi_dev_as->iommu_ir, OBJECT(s),
- &amdvi_ir_ops, s, "amd_iommu_ir",
- AMDVI_INT_ADDR_SIZE);
- memory_region_add_subregion_overlap(&amdvi_dev_as->root,
- AMDVI_INT_ADDR_FIRST,
- &amdvi_dev_as->iommu_ir,
- 64);
memory_region_add_subregion_overlap(&amdvi_dev_as->root, 0,
MEMORY_REGION(&amdvi_dev_as->iommu),
0);
@@ -1472,6 +1465,13 @@ static AddressSpace *amdvi_host_dma_iommu(PCIBus *bus, void *opaque, int devfn)
memory_region_add_subregion_overlap(&amdvi_dev_as->root, 0,
&amdvi_dev_as->iommu_nodma,
0);
+ /* Build the Interrupt Remapping alias to shared memory */
+ memory_region_init_alias(&amdvi_dev_as->iommu_ir, OBJECT(s),
+ "amdvi-ir", &s->mr_ir, 0,
+ memory_region_size(&s->mr_ir));
+ memory_region_add_subregion_overlap(MEMORY_REGION(&amdvi_dev_as->iommu),
+ AMDVI_INT_ADDR_FIRST,
+ &amdvi_dev_as->iommu_ir, 1);
if (!x86_iommu->pt_supported) {
memory_region_set_enabled(&amdvi_dev_as->iommu_nodma, false);
@@ -1633,6 +1633,12 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
memory_region_add_subregion_overlap(&s->mr_sys, 0,
&s->mr_nodma, 0);
+ /* set up the Interrupt Remapping memory region */
+ memory_region_init_io(&s->mr_ir, OBJECT(s), &amdvi_ir_ops,
+ s, "amdvi-ir", AMDVI_INT_ADDR_SIZE);
+ memory_region_add_subregion_overlap(&s->mr_sys, AMDVI_INT_ADDR_FIRST,
+ &s->mr_ir, 1);
+
pci_setup_iommu(bus, &amdvi_iommu_ops, s);
amdvi_init(s);
}
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index be417e51c4..e0dac4d9a9 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -356,6 +356,7 @@ struct AMDVIState {
MemoryRegion mr_mmio; /* MMIO region */
MemoryRegion mr_sys;
MemoryRegion mr_nodma;
+ MemoryRegion mr_ir;
uint8_t mmior[AMDVI_MMIO_SIZE]; /* read/write MMIO */
uint8_t w1cmask[AMDVI_MMIO_SIZE]; /* read/write 1 clear mask */
uint8_t romask[AMDVI_MMIO_SIZE]; /* MMIO read/only mask */
--
2.39.3

48
kvm-block-skip-automatic-zero-init-of-large-array-in-ioq.patch Normal file
View File

@ -0,0 +1,48 @@
From d38bdce712f572e1920e3344132ff6600d657de2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:41 +0100
Subject: [PATCH 29/57] block: skip automatic zero-init of large array in
ioq_submit
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/30] 301a08b3acdcd95634dec5dab1d96fcfe3abf3be (stefanha/centos-stream-qemu-kvm)
The 'ioq_submit' method has a struct array that is 8k in size.
Skip the automatic zero-init of this array to eliminate the
performance overhead in the I/O hot path.
The 'iocbs' array will selectively initialized when processing
the I/O data.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-4-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 83750c1da807c973b0b11d977d61df7e41122d03)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
block/linux-aio.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/linux-aio.c b/block/linux-aio.c
index e3b5ec9aba..26d9f086d2 100644
--- a/block/linux-aio.c
+++ b/block/linux-aio.c
@@ -291,7 +291,7 @@ static void ioq_submit(LinuxAioState *s)
{
int ret, len;
struct qemu_laiocb *aiocb;
- struct iocb *iocbs[MAX_EVENTS];
+ QEMU_UNINITIALIZED struct iocb *iocbs[MAX_EVENTS];
QSIMPLEQ_HEAD(, qemu_laiocb) completed;
do {
--
2.39.3

49
kvm-chardev-char-fd-skip-automatic-zero-init-of-large-ar.patch Normal file
View File

@ -0,0 +1,49 @@
From 1e8798a3adbbfc42167aaba0ee18175deac37193 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:42 +0100
Subject: [PATCH 30/57] chardev/char-fd: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [4/30] b16fe5c9af4756e1856cd330df02a1a09d9f33ea (stefanha/centos-stream-qemu-kvm)
The 'fd_chr_read' method has a 4k byte array used for copying
data between the socket and device. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O
hot path.
The 'buf' array will be fully initialized when reading data off
the network socket.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-5-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit a503bdc22b91869e3bf45522e36b122889465306)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
chardev/char-fd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/chardev/char-fd.c b/chardev/char-fd.c
index d2c4923359..8dd662c066 100644
--- a/chardev/char-fd.c
+++ b/chardev/char-fd.c
@@ -50,7 +50,7 @@ static gboolean fd_chr_read(QIOChannel *chan, GIOCondition cond, void *opaque)
Chardev *chr = CHARDEV(opaque);
FDChardev *s = FD_CHARDEV(opaque);
int len;
- uint8_t buf[CHR_READ_BUF_LEN];
+ QEMU_UNINITIALIZED uint8_t buf[CHR_READ_BUF_LEN];
ssize_t ret;
len = sizeof(buf);
--
2.39.3

49
kvm-chardev-char-pty-skip-automatic-zero-init-of-large-a.patch Normal file
View File

@ -0,0 +1,49 @@
From 74311b0ee8e211fccff211b975e4ae9236c063dc Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:43 +0100
Subject: [PATCH 31/57] chardev/char-pty: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [5/30] a3b8458c30f485551093f292c00c20b0e118df77 (stefanha/centos-stream-qemu-kvm)
The 'pty_chr_read' method has a 4k byte array used for copying
data between the PTY and device. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O
hot path.
The 'buf' array will be fully initialized when reading data off
the PTY.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-6-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 45bb7fb21c8d18294a9f92da99d01ab3c67c7df2)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
chardev/char-pty.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/chardev/char-pty.c b/chardev/char-pty.c
index cc2f7617fe..3319ad215d 100644
--- a/chardev/char-pty.c
+++ b/chardev/char-pty.c
@@ -152,7 +152,7 @@ static gboolean pty_chr_read(QIOChannel *chan, GIOCondition cond, void *opaque)
Chardev *chr = CHARDEV(opaque);
PtyChardev *s = PTY_CHARDEV(opaque);
gsize len;
- uint8_t buf[CHR_READ_BUF_LEN];
+ QEMU_UNINITIALIZED uint8_t buf[CHR_READ_BUF_LEN];
ssize_t ret;
len = sizeof(buf);
--
2.39.3

49
kvm-chardev-char-socket-skip-automatic-zero-init-of-larg.patch Normal file
View File

@ -0,0 +1,49 @@
From d56a8ce56f0de70ab2de266a80e25cf309e72fda Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:44 +0100
Subject: [PATCH 32/57] chardev/char-socket: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [6/30] 86a2ac03efa1838fb30931c38945ee77de9bbe06 (stefanha/centos-stream-qemu-kvm)
The 'tcp_chr_read' method has a 4k byte array used for copying
data between the socket and device. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O
hot path.
The 'buf' array will be fully initialized when reading data off
the network socket.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-7-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 9a23075cef1ac6e73a95a489ac72f41c573ceb9b)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
chardev/char-socket.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/chardev/char-socket.c b/chardev/char-socket.c
index 1ca9441b1b..99d644e89f 100644
--- a/chardev/char-socket.c
+++ b/chardev/char-socket.c
@@ -497,7 +497,7 @@ static gboolean tcp_chr_read(QIOChannel *chan, GIOCondition cond, void *opaque)
{
Chardev *chr = CHARDEV(opaque);
SocketChardev *s = SOCKET_CHARDEV(opaque);
- uint8_t buf[CHR_READ_BUF_LEN];
+ QEMU_UNINITIALIZED uint8_t buf[CHR_READ_BUF_LEN];
int len, size;
if ((s->state != TCP_CHARDEV_STATE_CONNECTED) ||
--
2.39.3

57
kvm-hw-audio-ac97-skip-automatic-zero-init-of-large-arra.patch Normal file
View File

@ -0,0 +1,57 @@
From 2018f62f2242d8d4a970d83ebef9b3c2bccf6fda Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:45 +0100
Subject: [PATCH 33/57] hw/audio/ac97: skip automatic zero-init of large arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [7/30] 4a6b59a9b9122d9f89e99b3e44df19e6d92ed941 (stefanha/centos-stream-qemu-kvm)
The 'read_audio' & 'write_audio' methods have a 4k byte array used
for copying data between the audio backend and device. Skip the
automatic zero-init of these arrays to eliminate the performance
overhead in the I/O hot path.
The 'tmpbuf' array will be fully initialized when reading data from
the audio backend and/or device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-8-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 2553d2d26a9d0f46386bf8c37d184567e5cede6c)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/ac97.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/audio/ac97.c b/hw/audio/ac97.c
index 3f0053f94d..681b5752a1 100644
--- a/hw/audio/ac97.c
+++ b/hw/audio/ac97.c
@@ -886,7 +886,7 @@ static void nabm_writel(void *opaque, uint32_t addr, uint32_t val)
static int write_audio(AC97LinkState *s, AC97BusMasterRegs *r,
int max, int *stop)
{
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
uint32_t addr = r->bd.addr;
uint32_t temp = r->picb << 1;
uint32_t written = 0;
@@ -959,7 +959,7 @@ static void write_bup(AC97LinkState *s, int elapsed)
static int read_audio(AC97LinkState *s, AC97BusMasterRegs *r,
int max, int *stop)
{
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
uint32_t addr = r->bd.addr;
uint32_t temp = r->picb << 1;
uint32_t nread = 0;
--
2.39.3

59
kvm-hw-audio-cs4231a-skip-automatic-zero-init-of-large-a.patch Normal file
View File

@ -0,0 +1,59 @@
From bd32bb22fb324a37b31ed9ac3387524f6f4ea5be Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:46 +0100
Subject: [PATCH 34/57] hw/audio/cs4231a: skip automatic zero-init of large
arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [8/30] 6c454bcc2927e49896c62718287fb9e4b37b3bb9 (stefanha/centos-stream-qemu-kvm)
The 'cs_write_audio' method has a pair of byte arrays, one 4k in size
and one 8k, which are used in converting audio samples. Skip the
automatic zero-init of these arrays to eliminate the performance
overhead in the I/O hot path.
The 'tmpbuf' array will be fully initialized when reading a block of
data from the guest. The 'linbuf' array will be fully initialized
when converting the audio samples.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-9-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit ca2cc0385d97cea66cd54ee42553f385c403d4a6)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/cs4231a.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/audio/cs4231a.c b/hw/audio/cs4231a.c
index 9ef57f042d..5c312642cc 100644
--- a/hw/audio/cs4231a.c
+++ b/hw/audio/cs4231a.c
@@ -528,7 +528,7 @@ static int cs_write_audio (CSState *s, int nchan, int dma_pos,
int dma_len, int len)
{
int temp, net;
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
IsaDmaClass *k = ISADMA_GET_CLASS(s->isa_dma);
temp = len;
@@ -547,7 +547,7 @@ static int cs_write_audio (CSState *s, int nchan, int dma_pos,
copied = k->read_memory(s->isa_dma, nchan, tmpbuf, dma_pos, to_copy);
if (s->tab) {
int i;
- int16_t linbuf[4096];
+ QEMU_UNINITIALIZED int16_t linbuf[4096];
for (i = 0; i < copied; ++i)
linbuf[i] = s->tab[tmpbuf[i]];
--
2.39.3

49
kvm-hw-audio-es1370-skip-automatic-zero-init-of-large-ar.patch Normal file
View File

@ -0,0 +1,49 @@
From cb12ddc6ed836091aa7724e2f77ab79cd9089cad Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:47 +0100
Subject: [PATCH 35/57] hw/audio/es1370: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [9/30] b992e4247d8d31dc09f9dc7671e7a532558174ec (stefanha/centos-stream-qemu-kvm)
The 'es1370_transfer_audio' method has a 4k byte array used for
copying data between the audio backend and device. Skip the automatic
zero-init of this array to eliminate the performance overhead in
the I/O hot path.
The 'tmpbuf' array will be fully initialized when reading data from
the audio backend and/or device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-10-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 8236e206084b832d1d7ec947a4798b818f4cdf1f)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/es1370.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/audio/es1370.c b/hw/audio/es1370.c
index 4ab61d3b9d..6aea934f54 100644
--- a/hw/audio/es1370.c
+++ b/hw/audio/es1370.c
@@ -604,7 +604,7 @@ static uint64_t es1370_read(void *opaque, hwaddr addr, unsigned size)
static void es1370_transfer_audio (ES1370State *s, struct chan *d, int loop_sel,
int max, bool *irq)
{
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
size_t to_transfer;
uint32_t addr = d->frame_addr;
int sc = d->scount & 0xffff;
--
2.39.3

48
kvm-hw-audio-gus-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,48 @@
From 9ad7091d82fd0577488f27ab54bb7851fe957020 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:48 +0100
Subject: [PATCH 36/57] hw/audio/gus: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [10/30] 366953d0417ac31e3060fdc327fe8dade3375bf0 (stefanha/centos-stream-qemu-kvm)
The 'GUS_read_DMA' method has a 4k byte array used for copying
data between the audio backend and device. Skip the automatic
zero-init of this array to eliminate the performance overhead in
the I/O hot path.
The 'tmpbuf' array will be fully initialized when reading data
from device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-11-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 2e438da4929018c62609381e1156aac0b2fe3de3)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/gus.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/audio/gus.c b/hw/audio/gus.c
index 4beb3fd74e..e8b0b85d44 100644
--- a/hw/audio/gus.c
+++ b/hw/audio/gus.c
@@ -183,7 +183,7 @@ static int GUS_read_DMA (void *opaque, int nchan, int dma_pos, int dma_len)
{
GUSState *s = opaque;
IsaDmaClass *k = ISADMA_GET_CLASS(s->isa_dma);
- char tmpbuf[4096];
+ QEMU_UNINITIALIZED char tmpbuf[4096];
int pos = dma_pos, mode, left = dma_len - dma_pos;
ldebug ("read DMA %#x %d\n", dma_pos, dma_len);
--
2.39.3

50
kvm-hw-audio-marvell_88w8618-skip-automatic-zero-init-of.patch Normal file
View File

@ -0,0 +1,50 @@
From 5cf61823cbe80b1ace2f5bdb9cc1971956425b98 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:49 +0100
Subject: [PATCH 37/57] hw/audio/marvell_88w8618: skip automatic zero-init of
large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [11/30] e09cdb76430552081168873dadfef1b5c8f74327 (stefanha/centos-stream-qemu-kvm)
The 'mv88w8618_audio_callback' method has a 4k byte array used for
copying data between the audio backend and device. Skip the automatic
zero-init of this array to eliminate the performance overhead in
the I/O hot path.
The 'buf' array will be fully initialized when reading data from
device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-12-berrange@redhat.com
[Fixed hw/audio/gus in commit message --Stefan]
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 5b6cd5c5df4229972d8a0fd9dd9a089a1644d6ba)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/marvell_88w8618.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/audio/marvell_88w8618.c b/hw/audio/marvell_88w8618.c
index cc285444bc..b7b4b27272 100644
--- a/hw/audio/marvell_88w8618.c
+++ b/hw/audio/marvell_88w8618.c
@@ -66,7 +66,7 @@ static void mv88w8618_audio_callback(void *opaque, int free_out, int free_in)
{
mv88w8618_audio_state *s = opaque;
int16_t *codec_buffer;
- int8_t buf[4096];
+ QEMU_UNINITIALIZED int8_t buf[4096];
int8_t *mem_buffer;
int pos, block_size;
--
2.39.3

48
kvm-hw-audio-sb16-skip-automatic-zero-init-of-large-arra.patch Normal file
View File

@ -0,0 +1,48 @@
From 0b4d59d75edd49ef99f0a82fbcbe360c5b48e4f8 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:50 +0100
Subject: [PATCH 38/57] hw/audio/sb16: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [12/30] 6475d67546bf04745636b317e965bcd89b6fb2d2 (stefanha/centos-stream-qemu-kvm)
The 'write_audio' method has a 4k byte array used for copying data
between the audio backend and device. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O hot
path.
The 'tmpbuf' array will be fully initialized when reading data from
device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-13-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 30c82f6657c1ee9fbb5473924b4d3273f214bd6f)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/sb16.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/audio/sb16.c b/hw/audio/sb16.c
index fd76e78d18..04c818ed3d 100644
--- a/hw/audio/sb16.c
+++ b/hw/audio/sb16.c
@@ -1181,7 +1181,7 @@ static int write_audio (SB16State *s, int nchan, int dma_pos,
IsaDma *isa_dma = nchan == s->dma ? s->isa_dma : s->isa_hdma;
IsaDmaClass *k = ISADMA_GET_CLASS(isa_dma);
int temp, net;
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
temp = len;
net = 0;
--
2.39.3

49
kvm-hw-audio-via-ac97-skip-automatic-zero-init-of-large-.patch Normal file
View File

@ -0,0 +1,49 @@
From 35332282ef8bd06f59206266006eff222ffe6bec Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:51 +0100
Subject: [PATCH 39/57] hw/audio/via-ac97: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [13/30] 6391a04b29fcbb8bcdbce2c6b786758fc34f0d71 (stefanha/centos-stream-qemu-kvm)
The 'out_cb' method has a 4k byte array used for copying data
between the audio backend and device. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O hot
path.
The 'tmpbuf' array will be fully initialized when reading data from
device memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-14-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit bb71d9fe1419f44529c91d1b09464718d157e647)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/audio/via-ac97.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/audio/via-ac97.c b/hw/audio/via-ac97.c
index 4c127a1def..e8fcf44e5d 100644
--- a/hw/audio/via-ac97.c
+++ b/hw/audio/via-ac97.c
@@ -175,7 +175,7 @@ static void out_cb(void *opaque, int avail)
ViaAC97SGDChannel *c = &s->aur;
int temp, to_copy, copied;
bool stop = false;
- uint8_t tmpbuf[4096];
+ QEMU_UNINITIALIZED uint8_t tmpbuf[4096];
if (c->stat & STAT_PAUSED) {
return;
--
2.39.3

49
kvm-hw-char-sclpconsole-lm-skip-automatic-zero-init-of-l.patch Normal file
View File

@ -0,0 +1,49 @@
From b0c16a93460c2dfe834a9f439d25dc833dfb7427 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:52 +0100
Subject: [PATCH 40/57] hw/char/sclpconsole-lm: skip automatic zero-init of
large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [14/30] 1491e0147a799ec523fa67fd49649722a07299e7 (stefanha/centos-stream-qemu-kvm)
The 'process_mdb' method has a 4k byte array used for copying data
between the guest and the chardev backend. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O hot
path.
The 'buffer' array will be selectively initialized when data is converted
between EBCDIC and ASCII.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-15-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 8b1dac1ad57082611419b0e2f347acd96115d25f)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/char/sclpconsole-lm.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/char/sclpconsole-lm.c b/hw/char/sclpconsole-lm.c
index 7719f438f6..19e64b92f6 100644
--- a/hw/char/sclpconsole-lm.c
+++ b/hw/char/sclpconsole-lm.c
@@ -214,7 +214,7 @@ static int process_mdb(SCLPEvent *event, MDBO *mdbo)
{
int rc;
int len;
- uint8_t buffer[SIZE_BUFFER];
+ QEMU_UNINITIALIZED uint8_t buffer[SIZE_BUFFER];
len = be16_to_cpu(mdbo->length);
len -= sizeof(mdbo->length) + sizeof(mdbo->type)
--
2.39.3

49
kvm-hw-display-vmware_vga-skip-automatic-zero-init-of-la.patch Normal file
View File

@ -0,0 +1,49 @@
From 7b5624efccf55184278c6f4924efc2141df460f0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:54 +0100
Subject: [PATCH 42/57] hw/display/vmware_vga: skip automatic zero-init of
large struct
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [16/30] 4aaf459d4356bf28164be742889b9a78d3656703 (stefanha/centos-stream-qemu-kvm)
The 'vmsvga_fifo_run' method has a struct which is a little over 20k
in size, used for holding image data for cursor changes. Skip the
automatic zero-init of this struct to eliminate the performance
overhead in the I/O hot path.
The cursor variable will be fully initialized only when processing
a cursor definition message from the guest.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-17-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 7048e70f391df76d009eecca25f8027858f9f304)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/display/vmware_vga.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/display/vmware_vga.c b/hw/display/vmware_vga.c
index 3db3ff98f7..69afe98a2f 100644
--- a/hw/display/vmware_vga.c
+++ b/hw/display/vmware_vga.c
@@ -618,7 +618,7 @@ static void vmsvga_fifo_run(struct vmsvga_state_s *s)
uint32_t cmd, colour;
int args, len, maxloop = 1024;
int x, y, dx, dy, width, height;
- struct vmsvga_cursor_definition_s cursor;
+ QEMU_UNINITIALIZED struct vmsvga_cursor_definition_s cursor;
uint32_t cmd_start;
len = vmsvga_fifo_length(s);
--
2.39.3

47
kvm-hw-dma-xlnx_csu_dma-skip-automatic-zero-init-of-larg.patch Normal file
View File

@ -0,0 +1,47 @@
From cd3500c9e248dbefb36273046e6eee44ee0d5cbe Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:53 +0100
Subject: [PATCH 41/57] hw/dma/xlnx_csu_dma: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [15/30] 063c88269c7d3bf07ae05aaf2d3d154e2016db81 (stefanha/centos-stream-qemu-kvm)
The 'xlnx_csu_dma_src_notify' method has a 4k byte array used for
copying DMA data. Skip the automatic zero-init of this array to
eliminate the performance overhead in the I/O hot path.
The 'buf' array will be fully initialized when data is copied.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-16-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit ce14f24611aa0469b464a9512e192b4fd51dca2b)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/dma/xlnx_csu_dma.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/dma/xlnx_csu_dma.c b/hw/dma/xlnx_csu_dma.c
index ae307482f2..9d1cccc5ca 100644
--- a/hw/dma/xlnx_csu_dma.c
+++ b/hw/dma/xlnx_csu_dma.c
@@ -287,7 +287,7 @@ static uint32_t xlnx_csu_dma_advance(XlnxCSUDMA *s, uint32_t len)
static void xlnx_csu_dma_src_notify(void *opaque)
{
XlnxCSUDMA *s = XLNX_CSU_DMA(opaque);
- unsigned char buf[4 * 1024];
+ QEMU_UNINITIALIZED unsigned char buf[4 * 1024];
size_t rlen = 0;
ptimer_transaction_begin(s->src_timer);
--
2.39.3

56
kvm-hw-hyperv-syndbg-skip-automatic-zero-init-of-large-a.patch Normal file
View File

@ -0,0 +1,56 @@
From a4673aab85958c60867b12c65cc3483d734bb6e0 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:55 +0100
Subject: [PATCH 43/57] hw/hyperv/syndbg: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [17/30] 5f71779c431128601baf46115fe65178532a3836 (stefanha/centos-stream-qemu-kvm)
The 'handle_recv_msg' method has a 4k byte array used for copying
data between the network socket and guest memory. Skip the automatic
zero-init of this array to eliminate the performance overhead in the
I/O hot path.
The 'data_buf' array will be fully initialized when data is read
off the network socket.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-18-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 5a1f614d0cd0bcc8e84e0b7ab6af63d56bd348a2)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Conflicts:
hw/hyperv/syndbg.c
Context conflict due to missing commit 3efb9d226221
("hw/hyperv/syndbg: common compilation unit") downstream. There is no
need to backport the commit because it's not a bug fix.
---
hw/hyperv/syndbg.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/hyperv/syndbg.c b/hw/hyperv/syndbg.c
index 065e12fb1e..c7c43c8009 100644
--- a/hw/hyperv/syndbg.c
+++ b/hw/hyperv/syndbg.c
@@ -188,7 +188,7 @@ static uint16_t handle_recv_msg(HvSynDbg *syndbg, uint64_t outgpa,
uint64_t timeout, uint32_t *retrieved_count)
{
uint16_t ret;
- uint8_t data_buf[TARGET_PAGE_SIZE - UDP_PKT_HEADER_SIZE];
+ QEMU_UNINITIALIZED uint8_t data_buf[TARGET_PAGE_SIZE - UDP_PKT_HEADER_SIZE];
hwaddr out_len;
void *out_data;
ssize_t recv_byte_count;
--
2.39.3

117
kvm-hw-i386-amd_iommu-Allow-migration-when-explicitly-cr.patch Normal file
View File

@ -0,0 +1,117 @@
From f1ff9d3b379697a2d4627e9529067195841d86a8 Mon Sep 17 00:00:00 2001
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Sun, 4 May 2025 17:04:05 +0000
Subject: [PATCH 25/57] hw/i386/amd_iommu: Allow migration when explicitly
create the AMDVI-PCI device
RH-Author: John Allen <None>
RH-MergeRequest: 380: Add ability to manually specify the AMDVI-PCI device
RH-Jira: RHEL-70925
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/3] a42b88116e608a79b6fae13ebe3709874f2a853f (johnalle/qemu-kvm-fork)
Add migration support for AMD IOMMU model by saving necessary AMDVIState
parameters for MMIO registers, device table, command buffer, and event
buffers.
Also change devtab_len type from size_t to uint64_t to avoid 32-bit build
issue.
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Message-Id: <20250504170405.12623-3-suravee.suthikulpanit@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit 28931c2e1591deb4bfaaf744fdc8813e96c230f1)
JIRA: https://issues.redhat.com/browse/RHEL-70925
Signed-off-by: John Allen <johnalle@redhat.com>
---
hw/i386/amd_iommu.c | 48 +++++++++++++++++++++++++++++++++++++++++++++
hw/i386/amd_iommu.h | 2 +-
2 files changed, 49 insertions(+), 1 deletion(-)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 6a5e76cfef..a34e0c5f59 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -1611,8 +1611,55 @@ static void amdvi_sysbus_reset(DeviceState *dev)
amdvi_init(s);
}
+static const VMStateDescription vmstate_amdvi_sysbus_migratable = {
+ .name = "amd-iommu",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .priority = MIG_PRI_IOMMU,
+ .fields = (VMStateField[]) {
+ /* Updated in amdvi_handle_control_write() */
+ VMSTATE_BOOL(enabled, AMDVIState),
+ VMSTATE_BOOL(ga_enabled, AMDVIState),
+ VMSTATE_BOOL(ats_enabled, AMDVIState),
+ VMSTATE_BOOL(cmdbuf_enabled, AMDVIState),
+ VMSTATE_BOOL(completion_wait_intr, AMDVIState),
+ VMSTATE_BOOL(evtlog_enabled, AMDVIState),
+ VMSTATE_BOOL(evtlog_intr, AMDVIState),
+ /* Updated in amdvi_handle_devtab_write() */
+ VMSTATE_UINT64(devtab, AMDVIState),
+ VMSTATE_UINT64(devtab_len, AMDVIState),
+ /* Updated in amdvi_handle_cmdbase_write() */
+ VMSTATE_UINT64(cmdbuf, AMDVIState),
+ VMSTATE_UINT64(cmdbuf_len, AMDVIState),
+ /* Updated in amdvi_handle_cmdhead_write() */
+ VMSTATE_UINT32(cmdbuf_head, AMDVIState),
+ /* Updated in amdvi_handle_cmdtail_write() */
+ VMSTATE_UINT32(cmdbuf_tail, AMDVIState),
+ /* Updated in amdvi_handle_evtbase_write() */
+ VMSTATE_UINT64(evtlog, AMDVIState),
+ VMSTATE_UINT32(evtlog_len, AMDVIState),
+ /* Updated in amdvi_handle_evthead_write() */
+ VMSTATE_UINT32(evtlog_head, AMDVIState),
+ /* Updated in amdvi_handle_evttail_write() */
+ VMSTATE_UINT32(evtlog_tail, AMDVIState),
+ /* Updated in amdvi_handle_pprbase_write() */
+ VMSTATE_UINT64(ppr_log, AMDVIState),
+ VMSTATE_UINT32(pprlog_len, AMDVIState),
+ /* Updated in amdvi_handle_pprhead_write() */
+ VMSTATE_UINT32(pprlog_head, AMDVIState),
+ /* Updated in amdvi_handle_tailhead_write() */
+ VMSTATE_UINT32(pprlog_tail, AMDVIState),
+ /* MMIO registers */
+ VMSTATE_UINT8_ARRAY(mmior, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_UINT8_ARRAY(romask, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_UINT8_ARRAY(w1cmask, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_END_OF_LIST()
+ }
+};
+
static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
{
+ DeviceClass *dc = (DeviceClass *) object_get_class(OBJECT(dev));
AMDVIState *s = AMD_IOMMU_DEVICE(dev);
MachineState *ms = MACHINE(qdev_get_machine());
PCMachineState *pcms = PC_MACHINE(ms);
@@ -1634,6 +1681,7 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
}
s->pci = AMD_IOMMU_PCI(pdev);
+ dc->vmsd = &vmstate_amdvi_sysbus_migratable;
} else {
s->pci = AMD_IOMMU_PCI(object_new(TYPE_AMD_IOMMU_PCI));
/* This device should take care of IOMMU PCI properties */
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index ece71ff0b6..741dd9a910 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -329,7 +329,7 @@ struct AMDVIState {
bool excl_enabled;
hwaddr devtab; /* base address device table */
- size_t devtab_len; /* device table length */
+ uint64_t devtab_len; /* device table length */
hwaddr cmdbuf; /* command buffer base address */
uint64_t cmdbuf_len; /* command buffer length */
--
2.39.3

57
kvm-hw-i386-amd_iommu-Assign-pci-id-0x1419-for-the-AMD-I.patch Normal file
View File

@ -0,0 +1,57 @@
From e611119b8b4e0712ab103628051d69ea84538719 Mon Sep 17 00:00:00 2001
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Tue, 25 Mar 2025 02:11:40 +0000
Subject: [PATCH 23/57] hw/i386/amd_iommu: Assign pci-id 0x1419 for the AMD
IOMMU device
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: John Allen <None>
RH-MergeRequest: 379: hw/i386/amd_iommu: Assign pci-id 0x1419 for the AMD IOMMU device
RH-Jira: RHEL-70926
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/1] 69d847f64543caf328da3e7663e7d2ebe53cd448 (johnalle/qemu-kvm-fork)
Currently, the QEMU-emulated AMD IOMMU device use PCI vendor id 0x1022
(AMD) with device id zero (undefined). Eventhough this does not cause any
functional issue for AMD IOMMU driver since it normally uses information
in the ACPI IVRS table to probe and initialize the device per
recommendation in the AMD IOMMU specification, the device id zero causes
the Windows Device Manager utility to show the device as an unknown device.
Since Windows only recognizes AMD IOMMU device with device id 0x1419 as
listed in the machine.inf file, modify the QEMU AMD IOMMU model to use
the id 0x1419 to avoid the issue. This advertise the IOMMU as the AMD
IOMMU device for Family 15h (Models 10h-1fh).
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Message-Id: <20250325021140.5676-1-suravee.suthikulpanit@amd.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Yan Vugenfirer <yvugenfi@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit 719255486df2fcbe1b8599786b37f4bb80272f1a)
JIRA: https://issues.redhat.com/browse/RHEL-70926
Signed-off-by: John Allen <johnalle@redhat.com>
---
hw/i386/amd_iommu.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index d804656ea8..59e1a01b7c 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -1714,6 +1714,7 @@ static void amdvi_pci_class_init(ObjectClass *klass, void *data)
PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
k->vendor_id = PCI_VENDOR_ID_AMD;
+ k->device_id = 0x1419;
k->class_id = 0x0806;
k->realize = amdvi_pci_realize;
--
2.39.3

267
kvm-hw-i386-amd_iommu-Isolate-AMDVI-PCI-from-amd-iommu-d.patch Normal file
View File

@ -0,0 +1,267 @@
From 5a697d0f66360acca8216f49c06dc9702231d470 Mon Sep 17 00:00:00 2001
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Date: Sun, 4 May 2025 17:04:04 +0000
Subject: [PATCH 24/57] hw/i386/amd_iommu: Isolate AMDVI-PCI from amd-iommu
device to allow full control over the PCI device creation
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: John Allen <None>
RH-MergeRequest: 380: Add ability to manually specify the AMDVI-PCI device
RH-Jira: RHEL-70925
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/3] 58254a72ba2d810b57c610462494f76691126521 (johnalle/qemu-kvm-fork)
Current amd-iommu model internally creates an AMDVI-PCI device. Here is
a snippet from info qtree:
bus: main-system-bus
type System
dev: amd-iommu, id ""
xtsup = false
pci-id = ""
intremap = "on"
device-iotlb = false
pt = true
...
dev: q35-pcihost, id ""
MCFG = -1 (0xffffffffffffffff)
pci-hole64-size = 34359738368 (32 GiB)
below-4g-mem-size = 134217728 (128 MiB)
above-4g-mem-size = 0 (0 B)
smm-ranges = true
x-pci-hole64-fix = true
x-config-reg-migration-enabled = true
bypass-iommu = false
bus: pcie.0
type PCIE
dev: AMDVI-PCI, id ""
addr = 01.0
romfile = ""
romsize = 4294967295 (0xffffffff)
rombar = -1 (0xffffffffffffffff)
multifunction = false
x-pcie-lnksta-dllla = true
x-pcie-extcap-init = true
failover_pair_id = ""
acpi-index = 0 (0x0)
x-pcie-err-unc-mask = true
x-pcie-ari-nextfn-1 = false
x-max-bounce-buffer-size = 4096 (4 KiB)
x-pcie-ext-tag = true
busnr = 0 (0x0)
class Class 0806, addr 00:01.0, pci id 1022:0000 (sub 1af4:1100)
...
This prohibits users from specifying the PCI topology for the amd-iommu device,
which becomes a problem when trying to support VM migration since it does not
guarantee the same enumeration of AMD IOMMU device.
Therefore, allow the 'AMDVI-PCI' device to optionally be pre-created and
associated with a 'amd-iommu' device via a new 'pci-id' parameter on the
latter.
For example:
-device AMDVI-PCI,id=iommupci0,bus=pcie.0,addr=0x05 \
-device amd-iommu,intremap=on,pt=on,xtsup=on,pci-id=iommupci0 \
For backward-compatibility, internally create the AMDVI-PCI device if not
specified on the CLI.
Co-developed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Message-Id: <20250504170405.12623-2-suravee.suthikulpanit@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
(cherry picked from commit f864a3235ea1d1d714b3cde2d9a810ea6344a7b5)
JIRA: https://issues.redhat.com/browse/RHEL-70925
Signed-off-by: John Allen <johnalle@redhat.com>
---
hw/i386/acpi-build.c | 8 +++----
hw/i386/amd_iommu.c | 53 ++++++++++++++++++++++++++------------------
hw/i386/amd_iommu.h | 3 ++-
3 files changed, 38 insertions(+), 26 deletions(-)
diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
index 032fb1f904..236261f8aa 100644
--- a/hw/i386/acpi-build.c
+++ b/hw/i386/acpi-build.c
@@ -2392,10 +2392,10 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
build_append_int_noprefix(table_data, ivhd_blob->len + 24, 2);
/* DeviceID */
build_append_int_noprefix(table_data,
- object_property_get_int(OBJECT(&s->pci), "addr",
+ object_property_get_int(OBJECT(s->pci), "addr",
&error_abort), 2);
/* Capability offset */
- build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
+ build_append_int_noprefix(table_data, s->pci->capab_offset, 2);
/* IOMMU base address */
build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
@@ -2427,10 +2427,10 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
build_append_int_noprefix(table_data, ivhd_blob->len + 40, 2);
/* DeviceID */
build_append_int_noprefix(table_data,
- object_property_get_int(OBJECT(&s->pci), "addr",
+ object_property_get_int(OBJECT(s->pci), "addr",
&error_abort), 2);
/* Capability offset */
- build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
+ build_append_int_noprefix(table_data, s->pci->capab_offset, 2);
/* IOMMU base address */
build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 59e1a01b7c..6a5e76cfef 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -167,11 +167,11 @@ static void amdvi_generate_msi_interrupt(AMDVIState *s)
{
MSIMessage msg = {};
MemTxAttrs attrs = {
- .requester_id = pci_requester_id(&s->pci.dev)
+ .requester_id = pci_requester_id(&s->pci->dev)
};
- if (msi_enabled(&s->pci.dev)) {
- msg = msi_get_message(&s->pci.dev, 0);
+ if (msi_enabled(&s->pci->dev)) {
+ msg = msi_get_message(&s->pci->dev, 0);
address_space_stl_le(&address_space_memory, msg.address, msg.data,
attrs, NULL);
}
@@ -239,7 +239,7 @@ static void amdvi_page_fault(AMDVIState *s, uint16_t devid,
info |= AMDVI_EVENT_IOPF_I | AMDVI_EVENT_IOPF;
amdvi_encode_event(evt, devid, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/*
@@ -256,7 +256,7 @@ static void amdvi_log_devtab_error(AMDVIState *s, uint16_t devid,
amdvi_encode_event(evt, devid, devtab, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/* log an event trying to access command buffer
@@ -269,7 +269,7 @@ static void amdvi_log_command_error(AMDVIState *s, hwaddr addr)
amdvi_encode_event(evt, 0, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/* log an illegal command event
@@ -310,7 +310,7 @@ static void amdvi_log_pagetab_error(AMDVIState *s, uint16_t devid,
info |= AMDVI_EVENT_PAGE_TAB_HW_ERROR;
amdvi_encode_event(evt, devid, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
@@ -1607,7 +1607,7 @@ static void amdvi_sysbus_reset(DeviceState *dev)
{
AMDVIState *s = AMD_IOMMU_DEVICE(dev);
- msi_reset(&s->pci.dev);
+ msi_reset(&s->pci->dev);
amdvi_init(s);
}
@@ -1619,14 +1619,32 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
X86MachineState *x86ms = X86_MACHINE(ms);
PCIBus *bus = pcms->pcibus;
- s->iotlb = g_hash_table_new_full(amdvi_uint64_hash,
- amdvi_uint64_equal, g_free, g_free);
+ if (s->pci_id) {
+ PCIDevice *pdev = NULL;
+ int ret = pci_qdev_find_device(s->pci_id, &pdev);
- /* This device should take care of IOMMU PCI properties */
- if (!qdev_realize(DEVICE(&s->pci), &bus->qbus, errp)) {
- return;
+ if (ret) {
+ error_report("Cannot find PCI device '%s'", s->pci_id);
+ return;
+ }
+
+ if (!object_dynamic_cast(OBJECT(pdev), TYPE_AMD_IOMMU_PCI)) {
+ error_report("Device '%s' must be an AMDVI-PCI device type", s->pci_id);
+ return;
+ }
+
+ s->pci = AMD_IOMMU_PCI(pdev);
+ } else {
+ s->pci = AMD_IOMMU_PCI(object_new(TYPE_AMD_IOMMU_PCI));
+ /* This device should take care of IOMMU PCI properties */
+ if (!qdev_realize(DEVICE(s->pci), &bus->qbus, errp)) {
+ return;
+ }
}
+ s->iotlb = g_hash_table_new_full(amdvi_uint64_hash,
+ amdvi_uint64_equal, g_free, g_free);
+
/* Pseudo address space under root PCI bus. */
x86ms->ioapic_as = amdvi_host_dma_iommu(bus, s, AMDVI_IOAPIC_SB_DEVID);
@@ -1668,6 +1686,7 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
static Property amdvi_properties[] = {
DEFINE_PROP_BOOL("xtsup", AMDVIState, xtsup, false),
+ DEFINE_PROP_STRING("pci-id", AMDVIState, pci_id),
DEFINE_PROP_END_OF_LIST(),
};
@@ -1676,13 +1695,6 @@ static const VMStateDescription vmstate_amdvi_sysbus = {
.unmigratable = 1
};
-static void amdvi_sysbus_instance_init(Object *klass)
-{
- AMDVIState *s = AMD_IOMMU_DEVICE(klass);
-
- object_initialize(&s->pci, sizeof(s->pci), TYPE_AMD_IOMMU_PCI);
-}
-
static void amdvi_sysbus_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
@@ -1704,7 +1716,6 @@ static const TypeInfo amdvi_sysbus = {
.name = TYPE_AMD_IOMMU_DEVICE,
.parent = TYPE_X86_IOMMU_DEVICE,
.instance_size = sizeof(AMDVIState),
- .instance_init = amdvi_sysbus_instance_init,
.class_init = amdvi_sysbus_class_init
};
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index e0dac4d9a9..ece71ff0b6 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -315,7 +315,8 @@ struct AMDVIPCIState {
struct AMDVIState {
X86IOMMUState iommu; /* IOMMU bus device */
- AMDVIPCIState pci; /* IOMMU PCI device */
+ AMDVIPCIState *pci; /* IOMMU PCI device */
+ char *pci_id; /* ID of AMDVI-PCI device, if user created */
uint32_t version;
--
2.39.3

57
kvm-hw-misc-aspeed_hace-skip-automatic-zero-init-of-larg.patch Normal file
View File

@ -0,0 +1,57 @@
From 0bfbd2c49c01ee77d3b5a21bf9fe675916cbf0ed Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:56 +0100
Subject: [PATCH 44/57] hw/misc/aspeed_hace: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [18/30] ec8510be6b23b26b3eecd6767e1deb0c0c50dd58 (stefanha/centos-stream-qemu-kvm)
The 'do_hash_operation' method has a 256 element iovec array used for
holding pointers to data that is to be hashed. Skip the automatic
zero-init of this array to eliminate the performance overhead in the
I/O hot path.
The 'iovec' array will be selectively initialized based on data that
needs to be hashed.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-19-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 6992c886838282f36b20deee44b666bbfc573a8f)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Conflicts:
hw/misc/aspeed_hace.c
Context conflict due to missing commit b9ccbe212e24
("hw/misc/aspeed_hace: Extract accumulation-mode hash execution into
helper function") downstream. The commit is not a bug fix, so there is
no need to backport it.
---
hw/misc/aspeed_hace.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/misc/aspeed_hace.c b/hw/misc/aspeed_hace.c
index c06c04ddc6..d2118f1864 100644
--- a/hw/misc/aspeed_hace.c
+++ b/hw/misc/aspeed_hace.c
@@ -188,7 +188,7 @@ static int gen_acc_mode_iov(AspeedHACEState *s, struct iovec *iov, int id,
static void do_hash_operation(AspeedHACEState *s, int algo, bool sg_mode,
bool acc_mode)
{
- struct iovec iov[ASPEED_HACE_MAX_SG];
+ QEMU_UNINITIALIZED struct iovec iov[ASPEED_HACE_MAX_SG];
g_autofree uint8_t *digest_buf = NULL;
size_t digest_len = 0;
int niov = 0;
--
2.39.3

48
kvm-hw-net-rtl8139-skip-automatic-zero-init-of-large-arr.patch Normal file
View File

@ -0,0 +1,48 @@
From cc173deaaa4d9dc6ad9188e0b03f46b7e64f26b2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:57 +0100
Subject: [PATCH 45/57] hw/net/rtl8139: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [19/30] 344c720aef2feb35f84fd4b21f2b1b31e5572286 (stefanha/centos-stream-qemu-kvm)
The 'rtl8139_transmit_one' method has a 8k byte array used for
copying data between guest and host. Skip the automatic zero-init
of this array to eliminate the performance overhead in the I/O
hot path.
The 'txbuffer' will be fully initialized when reading PCI DMA
buffers.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-20-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 3ccc6489dd4925ddd1f3066bd3751389169cd7aa)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/net/rtl8139.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/net/rtl8139.c b/hw/net/rtl8139.c
index f2fe057535..a2732bf1c1 100644
--- a/hw/net/rtl8139.c
+++ b/hw/net/rtl8139.c
@@ -1818,7 +1818,7 @@ static int rtl8139_transmit_one(RTL8139State *s, int descriptor)
PCIDevice *d = PCI_DEVICE(s);
int txsize = s->TxStatus[descriptor] & 0x1fff;
- uint8_t txbuffer[0x2000];
+ QEMU_UNINITIALIZED uint8_t txbuffer[0x2000];
DPRINTF("+++ transmit reading %d bytes from host memory at 0x%08x\n",
txsize, s->TxAddr[descriptor]);
--
2.39.3

47
kvm-hw-net-tulip-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,47 @@
From 400b5c8ae7f06a450ef91230343d7ce489142a38 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:58 +0100
Subject: [PATCH 46/57] hw/net/tulip: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [20/30] b3d29de8495c0ff40c26974673adefe4eb27a417 (stefanha/centos-stream-qemu-kvm)
The 'tulip_setup_frame' method has a 4k byte array used for copynig
DMA data from the device. Skip the automatic zero-init of this array
to eliminate the performance overhead in the I/O hot path.
The 'buf' array will be fully initialized when reading data from the
device.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-21-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit e1afd5ee6eb2954f4baf3c97820e4aaf7de97d2a)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/net/tulip.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/net/tulip.c b/hw/net/tulip.c
index 1f2ef20977..5cf2b96fbd 100644
--- a/hw/net/tulip.c
+++ b/hw/net/tulip.c
@@ -629,7 +629,7 @@ static void tulip_setup_filter_addr(TULIPState *s, uint8_t *buf, int n)
static void tulip_setup_frame(TULIPState *s,
struct tulip_descriptor *desc)
{
- uint8_t buf[4096];
+ QEMU_UNINITIALIZED uint8_t buf[4096];
int len = (desc->control >> TDES1_BUF1_SIZE_SHIFT) & TDES1_BUF1_SIZE_MASK;
int i;
--
2.39.3

54
kvm-hw-net-virtio-net-skip-automatic-zero-init-of-large-.patch Normal file
View File

@ -0,0 +1,54 @@
From 0925796a4537e20e033a675ebc8899e4580235f3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:36:59 +0100
Subject: [PATCH 47/57] hw/net/virtio-net: skip automatic zero-init of large
arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [21/30] 0450189a4c4c779b5a1850e9ea8278a5129c5f7f (stefanha/centos-stream-qemu-kvm)
The 'virtio_net_receive_rcu' method has three arrays with
VIRTQUEUE_MAX_SIZE elements, which are apprixmately 32k in
size used for copying data between guest and host. Skip the
automatic zero-init of these arrays to eliminate the
performance overhead in the I/O hot path.
The three arrays will be selectively initialized as required
when processing network buffers.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-22-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 21cf31c51a7aeff4270c9b30b37e019c536d54b2)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/net/virtio-net.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 3d2b2460ad..086ea20ea0 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -1895,9 +1895,9 @@ static ssize_t virtio_net_receive_rcu(NetClientState *nc, const uint8_t *buf,
VirtIONet *n = qemu_get_nic_opaque(nc);
VirtIONetQueue *q = virtio_net_get_subqueue(nc);
VirtIODevice *vdev = VIRTIO_DEVICE(n);
- VirtQueueElement *elems[VIRTQUEUE_MAX_SIZE];
- size_t lens[VIRTQUEUE_MAX_SIZE];
- struct iovec mhdr_sg[VIRTQUEUE_MAX_SIZE];
+ QEMU_UNINITIALIZED VirtQueueElement *elems[VIRTQUEUE_MAX_SIZE];
+ QEMU_UNINITIALIZED size_t lens[VIRTQUEUE_MAX_SIZE];
+ QEMU_UNINITIALIZED struct iovec mhdr_sg[VIRTQUEUE_MAX_SIZE];
struct virtio_net_hdr_v1_hash extra_hdr;
unsigned mhdr_cnt = 0;
size_t offset, i, guest_offset, j;
--
2.39.3

47
kvm-hw-net-xgamc-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,47 @@
From 34116b3a243f005938a30e9b38c6f47a62752c3e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:00 +0100
Subject: [PATCH 48/57] hw/net/xgamc: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [22/30] 63536d627705775c4bf72a511de3d68ec30ac7de (stefanha/centos-stream-qemu-kvm)
The 'xgmac_enet_send' method has a 8k byte array used for copying
data between guest and host. Skip the automatic zero-init of this
array to eliminate the performance overhead in the I/O hot path.
The 'frame' buffer will be fully initialized when reading guest
memory to fetch the data to send.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-23-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 8b723287b84a62bb5d1a7799ef0959ca8e6c293a)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/net/xgmac.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/net/xgmac.c b/hw/net/xgmac.c
index ffe3fc8dbe..eff8022aca 100644
--- a/hw/net/xgmac.c
+++ b/hw/net/xgmac.c
@@ -207,7 +207,7 @@ static void xgmac_enet_send(XgmacState *s)
struct desc bd;
int frame_size;
int len;
- uint8_t frame[8192];
+ QEMU_UNINITIALIZED uint8_t frame[8192];
uint8_t *ptr;
ptr = frame;
--
2.39.3

72
kvm-hw-nvme-ctrl-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,72 @@
From 3e0134b45828bf9a623a26ac41d5fbb3a8d2917b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:01 +0100
Subject: [PATCH 49/57] hw/nvme/ctrl: skip automatic zero-init of large arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [23/30] 57ce4361ffb307be4ea4d3edf9e0dac269d16908 (stefanha/centos-stream-qemu-kvm)
The 'nvme_map_sgl' method has a 256 element array used for copying
data from the device. Skip the automatic zero-init of this array
to eliminate the performance overhead in the I/O hot path.
The 'segment' array will be fully initialized when reading data from
the device.
The 'nme_changed_nslist' method has a 4k byte array that is manually
initialized with memset(). The compiler ought to be intelligent
enough to turn the memset() into a static initialization operation,
and thus not duplicate the automatic zero-init. Replacing memset()
with '{}' makes it unambiguous that the array is statically initialized.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Message-id: 20250610123709.835102-24-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 7eeb1d3acc175813ad3d5e824f26123e0992093a)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/nvme/ctrl.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/hw/nvme/ctrl.c b/hw/nvme/ctrl.c
index d451ee0d00..75d7f20801 100644
--- a/hw/nvme/ctrl.c
+++ b/hw/nvme/ctrl.c
@@ -1047,7 +1047,8 @@ static uint16_t nvme_map_sgl(NvmeCtrl *n, NvmeSg *sg, NvmeSglDescriptor sgl,
*/
#define SEG_CHUNK_SIZE 256
- NvmeSglDescriptor segment[SEG_CHUNK_SIZE], *sgld, *last_sgld;
+ QEMU_UNINITIALIZED NvmeSglDescriptor segment[SEG_CHUNK_SIZE];
+ NvmeSglDescriptor *sgld, *last_sgld;
uint64_t nsgld;
uint32_t seg_len;
uint16_t status;
@@ -5029,7 +5030,7 @@ static uint16_t nvme_error_info(NvmeCtrl *n, uint8_t rae, uint32_t buf_len,
static uint16_t nvme_changed_nslist(NvmeCtrl *n, uint8_t rae, uint32_t buf_len,
uint64_t off, NvmeRequest *req)
{
- uint32_t nslist[1024];
+ uint32_t nslist[1024] = {};
uint32_t trans_len;
int i = 0;
uint32_t nsid;
@@ -5039,7 +5040,6 @@ static uint16_t nvme_changed_nslist(NvmeCtrl *n, uint8_t rae, uint32_t buf_len,
return NVME_INVALID_FIELD | NVME_DNR;
}
- memset(nslist, 0x0, sizeof(nslist));
trans_len = MIN(sizeof(nslist) - off, buf_len);
while ((nsid = find_first_bit(n->changed_nsids, NVME_CHANGED_NSID_SIZE)) !=
--
2.39.3

52
kvm-hw-ppc-spapr_tpm_proxy-skip-automatic-zero-init-of-l.patch Normal file
View File

@ -0,0 +1,52 @@
From 4c3fe6e7b88c58713c0c499d4bf0658a055ee52e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:03 +0100
Subject: [PATCH 50/57] hw/ppc/spapr_tpm_proxy: skip automatic zero-init of
large arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [24/30] 8d963380c64a33a27adc99738b42b52864229111 (stefanha/centos-stream-qemu-kvm)
The 'tpm_execute' method has a pair of 4k arrays used for copying
data between guest and host. Skip the automatic zero-init of these
arrays to eliminate the performance overhead in the I/O hot path.
The two arrays will be fully initialized when reading data from
guest memory or reading data from the proxy FD.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-26-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 5dd9087fff74b5672526cad254e76f790fb35c7a)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/ppc/spapr_tpm_proxy.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/ppc/spapr_tpm_proxy.c b/hw/ppc/spapr_tpm_proxy.c
index e10af35a18..88833d9e2e 100644
--- a/hw/ppc/spapr_tpm_proxy.c
+++ b/hw/ppc/spapr_tpm_proxy.c
@@ -41,8 +41,8 @@ static ssize_t tpm_execute(SpaprTpmProxy *tpm_proxy, target_ulong *args)
target_ulong data_in_size = args[2];
uint64_t data_out = ppc64_phys_to_real(args[3]);
target_ulong data_out_size = args[4];
- uint8_t buf_in[TPM_SPAPR_BUFSIZE];
- uint8_t buf_out[TPM_SPAPR_BUFSIZE];
+ QEMU_UNINITIALIZED uint8_t buf_in[TPM_SPAPR_BUFSIZE];
+ QEMU_UNINITIALIZED uint8_t buf_out[TPM_SPAPR_BUFSIZE];
ssize_t ret;
trace_spapr_tpm_execute(data_in, data_in_size, data_out, data_out_size);
--
2.39.3

47
kvm-hw-s390x-ccw-device-Fix-memory-leak-in-loadparm-sett.patch Normal file
View File

@ -0,0 +1,47 @@
From b25bbfcad4a3df94555f6b5f238910314a5d17ea Mon Sep 17 00:00:00 2001
From: Kevin Wolf <kwolf@redhat.com>
Date: Wed, 25 Jun 2025 10:27:51 +0200
Subject: [PATCH 02/57] hw/s390x/ccw-device: Fix memory leak in loadparm setter
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Thomas Huth <thuth@redhat.com>
RH-MergeRequest: 387: s390x: Fix memory leaks related to loadparm [rhel-9]
RH-Jira: RHEL-98554
RH-Acked-by: Cédric Le Goater <clg@redhat.com>
RH-Acked-by: Kevin Wolf <kwolf@redhat.com>
RH-Commit: [2/2] d85cf8b3c93ede47b51c4aa1336dc54f58b8cc3f (thuth/qemu-kvm-cs)
Commit bdf12f2a fixed the setter for the "loadparm" machine property,
which gets a string from a visitor, passes it to s390_ipl_fmt_loadparm()
and then forgot to free it. It left another instance of the same problem
unfixed in the "loadparm" device property. Fix it.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250625082751.24896-1-kwolf@redhat.com>
Reviewed-by: Eric Farman <farman@linux.ibm.com>
Reviewed-by: Halil Pasic <pasic@linux.ibm.com>
Tested-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
(cherry picked from commit 78e3781541209b3dcd6f4bb66adf3a3e504b88a4)
---
hw/s390x/ccw-device.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/s390x/ccw-device.c b/hw/s390x/ccw-device.c
index 30f2fb486f..63e937401e 100644
--- a/hw/s390x/ccw-device.c
+++ b/hw/s390x/ccw-device.c
@@ -57,7 +57,7 @@ static void ccw_device_set_loadparm(Object *obj, Visitor *v,
Error **errp)
{
CcwDevice *dev = CCW_DEVICE(obj);
- char *val;
+ g_autofree char *val = NULL;
int index;
index = object_property_get_int(obj, "bootindex", NULL);
--
2.39.3

49
kvm-hw-scsi-lsi53c895a-skip-automatic-zero-init-of-large.patch Normal file
View File

@ -0,0 +1,49 @@
From 45884bfad1f14585407a04eff9230a75bc5095fa Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:05 +0100
Subject: [PATCH 52/57] hw/scsi/lsi53c895a: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [26/30] 235884d43fcb3e49b320e36faa631a3656d07de6 (stefanha/centos-stream-qemu-kvm)
The 'lsi_memcpy' method has a 4k byte array used for copying data
to/from the device. Skip the automatic zero-init of this array to
eliminate the performance overhead in the I/O hot path.
The 'buf' array will be fully initialized when data is copied.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-28-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 55243edf42ee87bce9f36ca251f3ab9cda1563e4)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/scsi/lsi53c895a.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/scsi/lsi53c895a.c b/hw/scsi/lsi53c895a.c
index f1935e5328..f165705f8a 100644
--- a/hw/scsi/lsi53c895a.c
+++ b/hw/scsi/lsi53c895a.c
@@ -1112,7 +1112,7 @@ bad:
static void lsi_memcpy(LSIState *s, uint32_t dest, uint32_t src, int count)
{
int n;
- uint8_t buf[LSI_BUF_SIZE];
+ QEMU_UNINITIALIZED uint8_t buf[LSI_BUF_SIZE];
trace_lsi_memcpy(dest, src, count);
while (count) {
--
2.39.3

73
kvm-hw-scsi-megasas-skip-automatic-zero-init-of-large-ar.patch Normal file
View File

@ -0,0 +1,73 @@
From 9f76103e90ce8406bc5bbda72a7314b82e56652e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:06 +0100
Subject: [PATCH 53/57] hw/scsi/megasas: skip automatic zero-init of large
arrays
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [27/30] b3a3f466fd03c64c665c52e26079b03def376f48 (stefanha/centos-stream-qemu-kvm)
The 'megasas_dcmd_pd_get_list' and 'megasas_dcmd_get_properties'
methods have 4k structs used for copying data from the device.
Skip the automatic zero-init of this array to eliminate the
performance overhead in the I/O hot path.
The 'info' structs are manually initialized with memset(). The
compiler ought to be intelligent enough to turn the memset()
into a static initialization operation, and thus not duplicate
the automatic zero-init. Replacing memset() with '{}' makes it
unambiguous that the arrays are statically initialized.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-29-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit ca0559e2350c618048f7caf80cb79c1259e7cfd2)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/scsi/megasas.c | 7 ++-----
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/hw/scsi/megasas.c b/hw/scsi/megasas.c
index 2d0c607177..91b65accbc 100644
--- a/hw/scsi/megasas.c
+++ b/hw/scsi/megasas.c
@@ -981,13 +981,11 @@ static int megasas_event_wait(MegasasState *s, MegasasCmd *cmd)
static int megasas_dcmd_pd_get_list(MegasasState *s, MegasasCmd *cmd)
{
- struct mfi_pd_list info;
- size_t dcmd_size = sizeof(info);
+ struct mfi_pd_list info = {};
BusChild *kid;
uint32_t offset, dcmd_limit, num_pd_disks = 0, max_pd_disks;
dma_addr_t residual;
- memset(&info, 0, dcmd_size);
offset = 8;
dcmd_limit = offset + sizeof(struct mfi_pd_address);
if (cmd->iov_size < dcmd_limit) {
@@ -1429,11 +1427,10 @@ static int megasas_dcmd_cfg_read(MegasasState *s, MegasasCmd *cmd)
static int megasas_dcmd_get_properties(MegasasState *s, MegasasCmd *cmd)
{
- struct mfi_ctrl_props info;
+ struct mfi_ctrl_props info = {};
size_t dcmd_size = sizeof(info);
dma_addr_t residual;
- memset(&info, 0x0, dcmd_size);
if (cmd->iov_size < dcmd_size) {
trace_megasas_dcmd_invalid_xfer_len(cmd->index, cmd->iov_size,
dcmd_size);
--
2.39.3

50
kvm-hw-ufs-lu-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,50 @@
From 3a0ae5a2f873fc7062262efc24a5403233988f5f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:07 +0100
Subject: [PATCH 54/57] hw/ufs/lu: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [28/30] 62e7c83d15143387f6d6b366c8ec46b312d05577 (stefanha/centos-stream-qemu-kvm)
The 'ufs_emulate_scsi_cmd' method has a 4k byte array used for
copying data from the device. Skip the automatic zero-init of
this array to eliminate the performance overhead in the I/O hot
path.
The 'outbuf' array will be fully initialized when data is copied
from the guest.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-30-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 7708e298180550eac262c1fd742e6e80c711a5d8)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/ufs/lu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/ufs/lu.c b/hw/ufs/lu.c
index 81bfff9b4e..caad82dcc4 100644
--- a/hw/ufs/lu.c
+++ b/hw/ufs/lu.c
@@ -194,7 +194,7 @@ static int ufs_emulate_wlun_inquiry(UfsRequest *req, uint8_t *outbuf,
static UfsReqResult ufs_emulate_scsi_cmd(UfsLu *lu, UfsRequest *req)
{
uint8_t lun = lu->lun;
- uint8_t outbuf[4096];
+ QEMU_UNINITIALIZED uint8_t outbuf[4096];
uint8_t sense_buf[UFS_SENSE_SIZE];
uint8_t scsi_status;
int len = 0;
--
2.39.3

50
kvm-hw-usb-hcd-ohci-skip-automatic-zero-init-of-large-ar.patch Normal file
View File

@ -0,0 +1,50 @@
From 6d4761010ea4dc218a1623513f410fc2d1cfc832 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:04 +0100
Subject: [PATCH 51/57] hw/usb/hcd-ohci: skip automatic zero-init of large
array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [25/30] 721dd97d384fb755c4a6a00cfc3d867e43f25b0b (stefanha/centos-stream-qemu-kvm)
The 'ohci_service_iso_td' method has a 8k byte array used for copying
data between guest and host. Skip the automatic zero-init of this
array to eliminate the performance overhead in the I/O hot path.
The 'buf' array will be fully initialized when reading data from guest
memory.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-27-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 14997d521d1cd0bb36c902ef1032f0d3f2a3c912)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/usb/hcd-ohci.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/hw/usb/hcd-ohci.c b/hw/usb/hcd-ohci.c
index 71b54914d3..72a9f9f474 100644
--- a/hw/usb/hcd-ohci.c
+++ b/hw/usb/hcd-ohci.c
@@ -577,7 +577,7 @@ static int ohci_service_iso_td(OHCIState *ohci, struct ohci_ed *ed)
USBDevice *dev;
USBEndpoint *ep;
USBPacket *pkt;
- uint8_t buf[8192];
+ QEMU_UNINITIALIZED uint8_t buf[8192];
bool int_req;
struct ohci_iso_td iso_td;
uint32_t addr;
--
2.39.3

73
kvm-hw-virtio-virtio-avoid-cost-of-ftrivial-auto-var-ini.patch Normal file
View File

@ -0,0 +1,73 @@
From 4727c044a09fb8c4fb6d667f26eb55bb6de7554d Mon Sep 17 00:00:00 2001
From: Stefan Hajnoczi <stefanha@redhat.com>
Date: Tue, 10 Jun 2025 13:36:40 +0100
Subject: [PATCH 28/57] hw/virtio/virtio: avoid cost of -ftrivial-auto-var-init
in hot path
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/30] 1c2cc6292deaaac068f4514439703c22c9ccb300 (stefanha/centos-stream-qemu-kvm)
Since commit 7ff9ff039380 ("meson: mitigate against use of uninitialize
stack for exploits") the -ftrivial-auto-var-init=zero compiler option is
used to zero local variables. While this reduces security risks
associated with uninitialized stack data, it introduced a measurable
bottleneck in the virtqueue_split_pop() and virtqueue_packed_pop()
functions.
These virtqueue functions are in the hot path. They are called for each
element (request) that is popped from a VIRTIO device's virtqueue. Using
__attribute__((uninitialized)) on large stack variables in these
functions improves fio randread bs=4k iodepth=64 performance from 304k
to 332k IOPS (+9%).
This issue was found using perf-top(1). virtqueue_split_pop() was one of
the top CPU consumers and the "annotate" feature showed that the memory
zeroing instructions at the beginning of the functions were hot.
Fixes: 7ff9ff039380 ("meson: mitigate against use of uninitialize stack for exploits")
Cc: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 20250610123709.835102-3-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit ba2868ce091cd4abe4be6de4b7e44b3be303b352)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
hw/virtio/virtio.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 10f24a58dd..7f7b178a50 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1680,8 +1680,8 @@ static void *virtqueue_split_pop(VirtQueue *vq, size_t sz)
VirtIODevice *vdev = vq->vdev;
VirtQueueElement *elem = NULL;
unsigned out_num, in_num, elem_entries;
- hwaddr addr[VIRTQUEUE_MAX_SIZE];
- struct iovec iov[VIRTQUEUE_MAX_SIZE];
+ hwaddr QEMU_UNINITIALIZED addr[VIRTQUEUE_MAX_SIZE];
+ struct iovec QEMU_UNINITIALIZED iov[VIRTQUEUE_MAX_SIZE];
VRingDesc desc;
int rc;
@@ -1826,8 +1826,8 @@ static void *virtqueue_packed_pop(VirtQueue *vq, size_t sz)
VirtIODevice *vdev = vq->vdev;
VirtQueueElement *elem = NULL;
unsigned out_num, in_num, elem_entries;
- hwaddr addr[VIRTQUEUE_MAX_SIZE];
- struct iovec iov[VIRTQUEUE_MAX_SIZE];
+ hwaddr QEMU_UNINITIALIZED addr[VIRTQUEUE_MAX_SIZE];
+ struct iovec QEMU_UNINITIALIZED iov[VIRTQUEUE_MAX_SIZE];
VRingPackedDesc desc;
uint16_t id;
int rc;
--
2.39.3

80
kvm-include-qemu-compiler-add-QEMU_UNINITIALIZED-attribu.patch Normal file
View File

@ -0,0 +1,80 @@
From cf92fd8487195ac45bfbdad15168eaec70f3aaa9 Mon Sep 17 00:00:00 2001
From: Stefan Hajnoczi <stefanha@redhat.com>
Date: Tue, 10 Jun 2025 13:36:39 +0100
Subject: [PATCH 27/57] include/qemu/compiler: add QEMU_UNINITIALIZED attribute
macro
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/30] 43c2412d318b6d8e0dcb0b37340640a9d90c3188 (stefanha/centos-stream-qemu-kvm)
The QEMU_UNINITIALIZED macro is to be used to skip the default compiler
variable initialization done by -ftrivial-auto-var-init=zero.
Use this in cases where there a method in the device I/O path (or other
important hot paths), that has large variables on the stack. A rule of
thumb is that "large" means a method with 4kb data in the local stack
frame. Any variables which are KB in size, should be annotated with this
attribute, to pre-emptively eliminate any potential overhead from the
compiler zero'ing memory.
Given that this turns off a security hardening feature, when using this
to flag variables, it is important that the code is double-checked to
ensure there is no possible use of uninitialized data in the method.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-id: 20250610123709.835102-2-berrange@redhat.com
[DB: split off patch & rewrite guidance on when to use the annotation]
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit c653b67d1863b7ebfa67f7c9f4aec209d7b5ced5)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Conflicts:
include/qemu/compiler.h
Context conflict due to clang Thread Safety Analysis macros.
---
include/qemu/compiler.h | 20 ++++++++++++++++++++
1 file changed, 20 insertions(+)
diff --git a/include/qemu/compiler.h b/include/qemu/compiler.h
index c06954ccb4..cc193d5b82 100644
--- a/include/qemu/compiler.h
+++ b/include/qemu/compiler.h
@@ -212,6 +212,26 @@
# define QEMU_USED
#endif
+/*
+ * Disable -ftrivial-auto-var-init on a local variable.
+ *
+ * Use this in cases where there a method in the device I/O path (or other
+ * important hot paths), that has large variables on the stack. A rule of
+ * thumb is that "large" means a method with 4kb data in the local stack
+ * frame. Any variables which are KB in size, should be annotated with this
+ * attribute, to pre-emptively eliminate any potential overhead from the
+ * compiler's implicit zero'ing of memory.
+ *
+ * Given that this turns off a security hardening feature, when using this
+ * to flag variables, it is important that the code is double-checked to
+ * ensure there is no possible use of uninitialized data in the method.
+ */
+#if __has_attribute(uninitialized)
+# define QEMU_UNINITIALIZED __attribute__((uninitialized))
+#else
+# define QEMU_UNINITIALIZED
+#endif
+
/*
* Ugly CPP trick that is like "defined FOO", but also works in C
* code. Useful to replace #ifdef with "if" statements; assumes
--
2.39.3

73
kvm-io-Fix-partial-struct-copy-in-qio_dns_resolver_looku.patch Normal file
View File

@ -0,0 +1,73 @@
From 4545870823aea92b18a7e747b686b666d08006a4 Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:30 +0200
Subject: [PATCH 08/57] io: Fix partial struct copy in
qio_dns_resolver_lookup_sync_inet()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/7] 92c8b3e63c22a3ca6e5adc76cac1a9f812034912 (JurajMarcin/centos-src-qemu-kvm)
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, but this flag is not copied together
with other options in qio_dns_resolver_lookup_sync_inet().
This patch fixes this issue and also prevents future ones by copying the
entire structure first and only then overriding a few attributes that
need to be different.
Fixes: aec21d31756c (qapi: Add InetSocketAddress member keep-alive)
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 0dc051aa85e1bd68d5c5110fa8af69204e6dbd3d)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
io/dns-resolver.c | 21 +++++----------------
1 file changed, 5 insertions(+), 16 deletions(-)
diff --git a/io/dns-resolver.c b/io/dns-resolver.c
index 53b0e8407a..3712438f82 100644
--- a/io/dns-resolver.c
+++ b/io/dns-resolver.c
@@ -111,22 +111,11 @@ static int qio_dns_resolver_lookup_sync_inet(QIODNSResolver *resolver,
uaddr, INET6_ADDRSTRLEN, uport, 32,
NI_NUMERICHOST | NI_NUMERICSERV);
- newaddr->u.inet = (InetSocketAddress){
- .host = g_strdup(uaddr),
- .port = g_strdup(uport),
- .has_numeric = true,
- .numeric = true,
- .has_to = iaddr->has_to,
- .to = iaddr->to,
- .has_ipv4 = iaddr->has_ipv4,
- .ipv4 = iaddr->ipv4,
- .has_ipv6 = iaddr->has_ipv6,
- .ipv6 = iaddr->ipv6,
-#ifdef HAVE_IPPROTO_MPTCP
- .has_mptcp = iaddr->has_mptcp,
- .mptcp = iaddr->mptcp,
-#endif
- };
+ newaddr->u.inet = *iaddr;
+ newaddr->u.inet.host = g_strdup(uaddr),
+ newaddr->u.inet.port = g_strdup(uport),
+ newaddr->u.inet.has_numeric = true,
+ newaddr->u.inet.numeric = true,
(*addrs)[i] = newaddr;
}
--
2.39.3

49
kvm-net-socket-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,49 @@
From 4b9a1a9154467fd65ac2a0a26959d3342d8fcd49 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:08 +0100
Subject: [PATCH 55/57] net/socket: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [29/30] 645ad4d138d1222ea9bd1b2ac3b84d9ff83e2fa2 (stefanha/centos-stream-qemu-kvm)
The 'net_socket_send' method has a 68k byte array used for copying
data between guest and host. Skip the automatic zero-init of this
array to eliminate the performance overhead in the I/O hot path.
The 'buf1' array will be fully initialized when reading data off
the network socket.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-31-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 751b0e79f1e0e7f88fad2fe2f22595ad03d78859)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
net/socket.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/socket.c b/net/socket.c
index 8e3702e1f3..784dda686f 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -157,7 +157,7 @@ static void net_socket_send(void *opaque)
NetSocketState *s = opaque;
int size;
int ret;
- uint8_t buf1[NET_BUFSIZE];
+ QEMU_UNINITIALIZED uint8_t buf1[NET_BUFSIZE];
const uint8_t *buf;
size = recv(s->fd, buf1, sizeof(buf1), 0);
--
2.39.3

49
kvm-net-stream-skip-automatic-zero-init-of-large-array.patch Normal file
View File

@ -0,0 +1,49 @@
From 94310a4168257297e52058d5d6aea4a2d06630c6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Daniel=20P=2E=20Berrang=C3=A9?= <berrange@redhat.com>
Date: Tue, 10 Jun 2025 13:37:09 +0100
Subject: [PATCH 56/57] net/stream: skip automatic zero-init of large array
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
RH-MergeRequest: 382: Solve -ftrivial-auto-var-init performance regression with QEMU_UNINITIALIZED
RH-Jira: RHEL-99888
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [30/30] 9dfec5c0e6358e3557bf58d66eee8e4ba6e93621 (stefanha/centos-stream-qemu-kvm)
The 'net_stream_send' method has a 68k byte array used for copying
data between guest and host. Skip the automatic zero-init of this
array to eliminate the performance overhead in the I/O hot path.
The 'buf1' array will be fully initialized when reading data off
the network socket.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Message-id: 20250610123709.835102-32-berrange@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
(cherry picked from commit 837b87c4c5ba9ac7a255133c6642b8d578272a70)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
---
net/stream.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/stream.c b/net/stream.c
index 97e6ec6679..12384ffee5 100644
--- a/net/stream.c
+++ b/net/stream.c
@@ -148,7 +148,7 @@ static gboolean net_stream_send(QIOChannel *ioc,
NetStreamState *s = data;
int size;
int ret;
- char buf1[NET_BUFSIZE];
+ QEMU_UNINITIALIZED char buf1[NET_BUFSIZE];
const char *buf;
size = qio_channel_read(s->ioc, buf1, sizeof(buf1), NULL);
--
2.39.3

60
kvm-s390x-Fix-leak-in-machine_set_loadparm.patch Normal file
View File

@ -0,0 +1,60 @@
From 4f627e0ae8efb96380070b6a8d50e88c71f40477 Mon Sep 17 00:00:00 2001
From: Fabiano Rosas <farosas@suse.de>
Date: Fri, 9 May 2025 14:49:38 -0300
Subject: [PATCH 01/57] s390x: Fix leak in machine_set_loadparm
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Thomas Huth <thuth@redhat.com>
RH-MergeRequest: 387: s390x: Fix memory leaks related to loadparm [rhel-9]
RH-Jira: RHEL-98554
RH-Acked-by: Cédric Le Goater <clg@redhat.com>
RH-Acked-by: Kevin Wolf <kwolf@redhat.com>
RH-Commit: [1/2] dadf5b9e187a644e0a8a8c565b1b913ef7f4dcc8 (thuth/qemu-kvm-cs)
ASAN spotted a leaking string in machine_set_loadparm():
Direct leak of 9 byte(s) in 1 object(s) allocated from:
#0 0x560ffb5bb379 in malloc ../projects/compiler-rt/lib/asan/asan_malloc_linux.cpp:69:3
#1 0x7f1aca926518 in g_malloc ../glib/gmem.c:106
#2 0x7f1aca94113e in g_strdup ../glib/gstrfuncs.c:364
#3 0x560ffc8afbf9 in qobject_input_type_str ../qapi/qobject-input-visitor.c:542:12
#4 0x560ffc8a80ff in visit_type_str ../qapi/qapi-visit-core.c:349:10
#5 0x560ffbe6053a in machine_set_loadparm ../hw/s390x/s390-virtio-ccw.c:802:10
#6 0x560ffc0c5e52 in object_property_set ../qom/object.c:1450:5
#7 0x560ffc0d4175 in object_property_set_qobject ../qom/qom-qobject.c:28:10
#8 0x560ffc0c6004 in object_property_set_str ../qom/object.c:1458:15
#9 0x560ffbe2ae60 in update_machine_ipl_properties ../hw/s390x/ipl.c:569:9
#10 0x560ffbe2aa65 in s390_ipl_update_diag308 ../hw/s390x/ipl.c:594:5
#11 0x560ffbdee132 in handle_diag_308 ../target/s390x/diag.c:147:9
#12 0x560ffbebb956 in helper_diag ../target/s390x/tcg/misc_helper.c:137:9
#13 0x7f1a3c51c730 (/memfd:tcg-jit (deleted)+0x39730)
Cc: qemu-stable@nongnu.org
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Message-ID: <20250509174938.25935-1-farosas@suse.de>
Fixes: 1fd396e3228 ("s390x: Register TYPE_S390_CCW_MACHINE properties as class properties")
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
(cherry picked from commit bdf12f2a56bf3f13c52eb51f0a994bbfe40706b2)
---
hw/s390x/s390-virtio-ccw.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/hw/s390x/s390-virtio-ccw.c b/hw/s390x/s390-virtio-ccw.c
index 77a1bde71e..fc18ab575f 100644
--- a/hw/s390x/s390-virtio-ccw.c
+++ b/hw/s390x/s390-virtio-ccw.c
@@ -782,6 +782,7 @@ static void machine_set_loadparm(Object *obj, Visitor *v,
}
s390_ipl_fmt_loadparm(ms->loadparm, val, errp);
+ g_free(val);
}
static void ccw_machine_class_init(ObjectClass *oc, void *data)
--
2.39.3

105
kvm-target-i386-Add-PerfMonV2-feature-bit.patch Normal file
View File

@ -0,0 +1,105 @@
From 1587da0703e72cca8325a20b709280b8df85d066 Mon Sep 17 00:00:00 2001
From: Sandipan Das <sandipan.das@amd.com>
Date: Thu, 24 Oct 2024 17:18:21 -0500
Subject: [PATCH 16/57] target/i386: Add PerfMonV2 feature bit
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/8] ec365cf4ac558c6c83f7a957e8df937cb6fbfa27 (johnalle/qemu-kvm-fork)
CPUID leaf 0x80000022, i.e. ExtPerfMonAndDbg, advertises new performance
monitoring features for AMD processors. Bit 0 of EAX indicates support
for Performance Monitoring Version 2 (PerfMonV2) features. If found to
be set during PMU initialization, the EBX bits can be used to determine
the number of available counters for different PMUs. It also denotes the
availability of global control and status registers.
Add the required CPUID feature word and feature bit to allow guests to
make use of the PerfMonV2 features.
Signed-off-by: Sandipan Das <sandipan.das@amd.com>
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/a96f00ee2637674c63c61e9fc4dee343ea818053.1729807947.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit 209b0ac12074341d0093985eb9ad3e7edb252ce5)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 26 ++++++++++++++++++++++++++
target/i386/cpu.h | 4 ++++
2 files changed, 30 insertions(+)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 53069a460c..4546369836 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -1246,6 +1246,22 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
.tcg_features = 0,
.unmigratable_flags = 0,
},
+ [FEAT_8000_0022_EAX] = {
+ .type = CPUID_FEATURE_WORD,
+ .feat_names = {
+ "perfmon-v2", NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ NULL, NULL, NULL, NULL,
+ },
+ .cpuid = { .eax = 0x80000022, .reg = R_EAX, },
+ .tcg_features = 0,
+ .unmigratable_flags = 0,
+ },
[FEAT_XSAVE] = {
.type = CPUID_FEATURE_WORD,
.feat_names = {
@@ -7096,6 +7112,16 @@ void cpu_x86_cpuid(CPUX86State *env, uint32_t index, uint32_t count,
*edx = 0;
}
break;
+ case 0x80000022:
+ *eax = *ebx = *ecx = *edx = 0;
+ /* AMD Extended Performance Monitoring and Debug */
+ if (kvm_enabled() && cpu->enable_pmu &&
+ (env->features[FEAT_8000_0022_EAX] & CPUID_8000_0022_EAX_PERFMON_V2)) {
+ *eax |= CPUID_8000_0022_EAX_PERFMON_V2;
+ *ebx |= kvm_arch_get_supported_cpuid(cs->kvm_state, index, count,
+ R_EBX) & 0xf;
+ }
+ break;
case 0xC0000000:
*eax = env->cpuid_xlevel2;
*ebx = 0;
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 9a16239b8e..cf92a4972c 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -638,6 +638,7 @@ typedef enum FeatureWord {
FEAT_8000_0007_EDX, /* CPUID[8000_0007].EDX */
FEAT_8000_0008_EBX, /* CPUID[8000_0008].EBX */
FEAT_8000_0021_EAX, /* CPUID[8000_0021].EAX */
+ FEAT_8000_0022_EAX, /* CPUID[8000_0022].EAX */
FEAT_C000_0001_EDX, /* CPUID[C000_0001].EDX */
FEAT_KVM, /* CPUID[4000_0001].EAX (KVM_CPUID_FEATURES) */
FEAT_KVM_HINTS, /* CPUID[4000_0001].EDX */
@@ -1044,6 +1045,9 @@ uint64_t x86_cpu_get_supported_feature_word(X86CPU *cpu, FeatureWord w);
/* Not vulnerable to SRSO at the user-kernel boundary */
#define CPUID_8000_0021_EAX_SRSO_USER_KERNEL_NO (1U << 30)
+/* Performance Monitoring Version 2 */
+#define CPUID_8000_0022_EAX_PERFMON_V2 (1U << 0)
+
#define CPUID_XSAVE_XSAVEOPT (1U << 0)
#define CPUID_XSAVE_XSAVEC (1U << 1)
#define CPUID_XSAVE_XGETBV1 (1U << 2)
--
2.39.3

83
kvm-target-i386-Add-couple-of-feature-bits-in-CPUID_Fn80.patch Normal file
View File

@ -0,0 +1,83 @@
From 79ac76edecdbbe253ad42385730aac18cdc40bd7 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Fri, 20 Jun 2025 14:54:53 -0500
Subject: [PATCH 20/57] target/i386: Add couple of feature bits in
CPUID_Fn80000021_EAX
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [6/8] c6507eb24fcef271fdd6a234d2c255ef38c4e691 (johnalle/qemu-kvm-fork)
Add CPUID bit indicates that a WRMSR to MSR_FS_BASE, MSR_GS_BASE, or
MSR_KERNEL_GS_BASE is non-serializing amd PREFETCHI that the
cates
support for IC prefetch.
CPUID_Fn80000021_EAX
Bit Feature description
20 Indicates support for IC prefetch.
1 FsGsKernelGsBaseNonSerializing.
WRMSR to FS_BASE, GS_BASE and KernelGSbase are
serializing.
Link: https://www.amd.com/content/dam/amd/en/documents/epyc-technical-docs/programmer-references/57238.zip
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/a5f6283a59579b09ac345b3f21ecb3b3b2d92451.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit dfd5b456108a75588ab094358ba5754787146d3d)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 4 ++--
target/i386/cpu.h | 4 ++++
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 7d48c51767..2218071fca 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -1233,12 +1233,12 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
[FEAT_8000_0021_EAX] = {
.type = CPUID_FEATURE_WORD,
.feat_names = {
- "no-nested-data-bp", NULL, "lfence-always-serializing", NULL,
+ "no-nested-data-bp", "fs-gs-base-ns", "lfence-always-serializing", NULL,
NULL, NULL, "null-sel-clr-base", NULL,
"auto-ibrs", NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
- NULL, NULL, NULL, NULL,
+ "prefetchi", NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
"ibpb-brtype", "srso-no", "srso-user-kernel-no", NULL,
},
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index cf92a4972c..e513e5f62d 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1030,12 +1030,16 @@ uint64_t x86_cpu_get_supported_feature_word(X86CPU *cpu, FeatureWord w);
/* Processor ignores nested data breakpoints */
#define CPUID_8000_0021_EAX_NO_NESTED_DATA_BP (1U << 0)
+/* WRMSR to FS_BASE, GS_BASE, or KERNEL_GS_BASE is non-serializing */
+#define CPUID_8000_0021_EAX_FS_GS_BASE_NS (1U << 1)
/* LFENCE is always serializing */
#define CPUID_8000_0021_EAX_LFENCE_ALWAYS_SERIALIZING (1U << 2)
/* Null Selector Clears Base */
#define CPUID_8000_0021_EAX_NULL_SEL_CLR_BASE (1U << 6)
/* Automatic IBRS */
#define CPUID_8000_0021_EAX_AUTO_IBRS (1U << 8)
+/* Indicates support for IC prefetch */
+#define CPUID_8000_0021_EAX_PREFETCHI (1U << 20)
/* Selective Branch Predictor Barrier */
#define CPUID_8000_0021_EAX_SBPB (1U << 27)
/* IBPB includes branch type prediction flushing */
--
2.39.3

200
kvm-target-i386-Add-support-for-EPYC-Turin-model.patch Normal file
View File

@ -0,0 +1,200 @@
From e0b59a57883faac254cd75cc243fed784ad4975b Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 8 May 2025 14:58:04 -0500
Subject: [PATCH 22/57] target/i386: Add support for EPYC-Turin model
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [8/8] 42e90c7fc6bf858f98ae3a6be35d70b824a7d6bf (johnalle/qemu-kvm-fork)
Add the support for AMD EPYC zen 5 processors (EPYC-Turin).
Add the following new feature bits on top of the feature bits from
the previous generation EPYC models.
movdiri : Move Doubleword as Direct Store Instruction
movdir64b : Move 64 Bytes as Direct Store Instruction
avx512-vp2intersect : AVX512 Vector Pair Intersection to a Pair
of Mask Register
avx-vnni : AVX VNNI Instruction
prefetchi : Indicates support for IC prefetch
sbpb : Selective Branch Predictor Barrier
ibpb-brtype : IBPB includes branch type prediction flushing
srso-user-kernel-no : Not vulnerable to SRSO at the user-kernel boundary
Link: https://www.amd.com/content/dam/amd/en/documents/epyc-technical-docs/programmer-references/57238.zip
Link: https://www.amd.com/content/dam/amd/en/documents/corporate/cr/speculative-return-stack-overflow-whitepaper.pdf
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/b4fa7708a0e1453d2e9b8ec3dc881feb92eeca0b.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit 3771a4daa273ba17cb27309984413790d1df5651)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 138 ++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 138 insertions(+)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 2bc2d41259..fdfa183f4d 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -2651,6 +2651,61 @@ static const CPUCaches epyc_genoa_v2_cache_info = {
.share_level = CPU_TOPO_LEVEL_DIE,
},
};
+
+static const CPUCaches epyc_turin_cache_info = {
+ .l1d_cache = &(CPUCacheInfo) {
+ .type = DATA_CACHE,
+ .level = 1,
+ .size = 48 * KiB,
+ .line_size = 64,
+ .associativity = 12,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l1i_cache = &(CPUCacheInfo) {
+ .type = INSTRUCTION_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l2_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 2,
+ .size = 1 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 1024,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .inclusive = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l3_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 3,
+ .size = 32 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 32768,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .no_invd_sharing = true,
+ .complex_indexing = false,
+ .share_level = CPU_TOPO_LEVEL_DIE,
+ },
+};
+
/* The following VMX features are not supported by KVM and are left out in the
* CPU definitions:
*
@@ -5644,6 +5699,89 @@ static const X86CPUDefinition builtin_x86_defs[] = {
{ /* end of list */ }
}
},
+ {
+ .name = "EPYC-Turin",
+ .level = 0xd,
+ .vendor = CPUID_VENDOR_AMD,
+ .family = 26,
+ .model = 0,
+ .stepping = 0,
+ .features[FEAT_1_ECX] =
+ CPUID_EXT_RDRAND | CPUID_EXT_F16C | CPUID_EXT_AVX |
+ CPUID_EXT_XSAVE | CPUID_EXT_AES | CPUID_EXT_POPCNT |
+ CPUID_EXT_MOVBE | CPUID_EXT_SSE42 | CPUID_EXT_SSE41 |
+ CPUID_EXT_PCID | CPUID_EXT_CX16 | CPUID_EXT_FMA |
+ CPUID_EXT_SSSE3 | CPUID_EXT_MONITOR | CPUID_EXT_PCLMULQDQ |
+ CPUID_EXT_SSE3,
+ .features[FEAT_1_EDX] =
+ CPUID_SSE2 | CPUID_SSE | CPUID_FXSR | CPUID_MMX | CPUID_CLFLUSH |
+ CPUID_PSE36 | CPUID_PAT | CPUID_CMOV | CPUID_MCA | CPUID_PGE |
+ CPUID_MTRR | CPUID_SEP | CPUID_APIC | CPUID_CX8 | CPUID_MCE |
+ CPUID_PAE | CPUID_MSR | CPUID_TSC | CPUID_PSE | CPUID_DE |
+ CPUID_VME | CPUID_FP87,
+ .features[FEAT_6_EAX] =
+ CPUID_6_EAX_ARAT,
+ .features[FEAT_7_0_EBX] =
+ CPUID_7_0_EBX_FSGSBASE | CPUID_7_0_EBX_BMI1 | CPUID_7_0_EBX_AVX2 |
+ CPUID_7_0_EBX_SMEP | CPUID_7_0_EBX_BMI2 | CPUID_7_0_EBX_ERMS |
+ CPUID_7_0_EBX_INVPCID | CPUID_7_0_EBX_AVX512F |
+ CPUID_7_0_EBX_AVX512DQ | CPUID_7_0_EBX_RDSEED | CPUID_7_0_EBX_ADX |
+ CPUID_7_0_EBX_SMAP | CPUID_7_0_EBX_AVX512IFMA |
+ CPUID_7_0_EBX_CLFLUSHOPT | CPUID_7_0_EBX_CLWB |
+ CPUID_7_0_EBX_AVX512CD | CPUID_7_0_EBX_SHA_NI |
+ CPUID_7_0_EBX_AVX512BW | CPUID_7_0_EBX_AVX512VL,
+ .features[FEAT_7_0_ECX] =
+ CPUID_7_0_ECX_AVX512_VBMI | CPUID_7_0_ECX_UMIP | CPUID_7_0_ECX_PKU |
+ CPUID_7_0_ECX_AVX512_VBMI2 | CPUID_7_0_ECX_GFNI |
+ CPUID_7_0_ECX_VAES | CPUID_7_0_ECX_VPCLMULQDQ |
+ CPUID_7_0_ECX_AVX512VNNI | CPUID_7_0_ECX_AVX512BITALG |
+ CPUID_7_0_ECX_AVX512_VPOPCNTDQ | CPUID_7_0_ECX_LA57 |
+ CPUID_7_0_ECX_RDPID | CPUID_7_0_ECX_MOVDIRI |
+ CPUID_7_0_ECX_MOVDIR64B,
+ .features[FEAT_7_0_EDX] =
+ CPUID_7_0_EDX_FSRM | CPUID_7_0_EDX_AVX512_VP2INTERSECT,
+ .features[FEAT_7_1_EAX] =
+ CPUID_7_1_EAX_AVX_VNNI | CPUID_7_1_EAX_AVX512_BF16,
+ .features[FEAT_8000_0001_ECX] =
+ CPUID_EXT3_OSVW | CPUID_EXT3_3DNOWPREFETCH |
+ CPUID_EXT3_MISALIGNSSE | CPUID_EXT3_SSE4A | CPUID_EXT3_ABM |
+ CPUID_EXT3_CR8LEG | CPUID_EXT3_SVM | CPUID_EXT3_LAHF_LM |
+ CPUID_EXT3_TOPOEXT | CPUID_EXT3_PERFCORE,
+ .features[FEAT_8000_0001_EDX] =
+ CPUID_EXT2_LM | CPUID_EXT2_RDTSCP | CPUID_EXT2_PDPE1GB |
+ CPUID_EXT2_FFXSR | CPUID_EXT2_MMXEXT | CPUID_EXT2_NX |
+ CPUID_EXT2_SYSCALL,
+ .features[FEAT_8000_0007_EBX] =
+ CPUID_8000_0007_EBX_OVERFLOW_RECOV | CPUID_8000_0007_EBX_SUCCOR,
+ .features[FEAT_8000_0008_EBX] =
+ CPUID_8000_0008_EBX_CLZERO | CPUID_8000_0008_EBX_XSAVEERPTR |
+ CPUID_8000_0008_EBX_WBNOINVD | CPUID_8000_0008_EBX_IBPB |
+ CPUID_8000_0008_EBX_IBRS | CPUID_8000_0008_EBX_STIBP |
+ CPUID_8000_0008_EBX_STIBP_ALWAYS_ON |
+ CPUID_8000_0008_EBX_AMD_SSBD | CPUID_8000_0008_EBX_AMD_PSFD,
+ .features[FEAT_8000_0021_EAX] =
+ CPUID_8000_0021_EAX_NO_NESTED_DATA_BP |
+ CPUID_8000_0021_EAX_FS_GS_BASE_NS |
+ CPUID_8000_0021_EAX_LFENCE_ALWAYS_SERIALIZING |
+ CPUID_8000_0021_EAX_NULL_SEL_CLR_BASE |
+ CPUID_8000_0021_EAX_AUTO_IBRS | CPUID_8000_0021_EAX_PREFETCHI |
+ CPUID_8000_0021_EAX_SBPB | CPUID_8000_0021_EAX_IBPB_BRTYPE |
+ CPUID_8000_0021_EAX_SRSO_USER_KERNEL_NO,
+ .features[FEAT_8000_0022_EAX] =
+ CPUID_8000_0022_EAX_PERFMON_V2,
+ .features[FEAT_XSAVE] =
+ CPUID_XSAVE_XSAVEOPT | CPUID_XSAVE_XSAVEC |
+ CPUID_XSAVE_XGETBV1 | CPUID_XSAVE_XSAVES,
+ .features[FEAT_SVM] =
+ CPUID_SVM_NPT | CPUID_SVM_LBRV | CPUID_SVM_NRIPSAVE |
+ CPUID_SVM_TSCSCALE | CPUID_SVM_VMCBCLEAN | CPUID_SVM_FLUSHASID |
+ CPUID_SVM_PAUSEFILTER | CPUID_SVM_PFTHRESHOLD |
+ CPUID_SVM_V_VMSAVE_VMLOAD | CPUID_SVM_VGIF |
+ CPUID_SVM_VNMI | CPUID_SVM_SVME_ADDR_CHK,
+ .xlevel = 0x80000022,
+ .model_id = "AMD EPYC-Turin Processor",
+ .cache_info = &epyc_turin_cache_info,
+ },
};
/*
--
2.39.3

84
kvm-target-i386-Expose-bits-related-to-SRSO-vulnerabilit.patch Normal file
View File

@ -0,0 +1,84 @@
From 1d667a354613385b1552fdbae91799882776f908 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 24 Oct 2024 17:18:23 -0500
Subject: [PATCH 15/57] target/i386: Expose bits related to SRSO vulnerability
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [1/8] 9a6f4126ab023269e8afb3537aaa94ae60228382 (johnalle/qemu-kvm-fork)
Add following bits related Speculative Return Stack Overflow (SRSO).
Guests can make use of these bits if supported.
These bits are reported via CPUID Fn8000_0021_EAX.
===================================================================
Bit Feature Description
===================================================================
27 SBPB Indicates support for the Selective Branch Predictor Barrier.
28 IBPB_BRTYPE MSR_PRED_CMD[IBPB] flushes all branch type predictions.
29 SRSO_NO Not vulnerable to SRSO.
30 SRSO_USER_KERNEL_NO Not vulnerable to SRSO at the user-kernel boundary.
===================================================================
Link: https://www.amd.com/content/dam/amd/en/documents/corporate/cr/speculative-return-stack-overflow-whitepaper.pdf
Link: https://www.amd.com/content/dam/amd/en/documents/epyc-technical-docs/programmer-references/57238.zip
Signed-off-by: Babu Moger <babu.moger@amd.com>
Link: https://lore.kernel.org/r/dadbd70c38f4e165418d193918a3747bd715c5f4.1729807947.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit 2ec282b8eaaddf5c136f7566b5f61d80288a2065)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 2 +-
target/i386/cpu.h | 14 +++++++++++---
2 files changed, 12 insertions(+), 4 deletions(-)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 0a955b1c45..53069a460c 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -1240,7 +1240,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL,
- NULL, NULL, NULL, NULL,
+ "ibpb-brtype", "srso-no", "srso-user-kernel-no", NULL,
},
.cpuid = { .eax = 0x80000021, .reg = R_EAX, },
.tcg_features = 0,
diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index 4da9ed5930..9a16239b8e 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1028,13 +1028,21 @@ uint64_t x86_cpu_get_supported_feature_word(X86CPU *cpu, FeatureWord w);
#define CPUID_8000_0008_EBX_AMD_PSFD (1U << 28)
/* Processor ignores nested data breakpoints */
-#define CPUID_8000_0021_EAX_No_NESTED_DATA_BP (1U << 0)
+#define CPUID_8000_0021_EAX_NO_NESTED_DATA_BP (1U << 0)
/* LFENCE is always serializing */
#define CPUID_8000_0021_EAX_LFENCE_ALWAYS_SERIALIZING (1U << 2)
/* Null Selector Clears Base */
-#define CPUID_8000_0021_EAX_NULL_SEL_CLR_BASE (1U << 6)
+#define CPUID_8000_0021_EAX_NULL_SEL_CLR_BASE (1U << 6)
/* Automatic IBRS */
-#define CPUID_8000_0021_EAX_AUTO_IBRS (1U << 8)
+#define CPUID_8000_0021_EAX_AUTO_IBRS (1U << 8)
+/* Selective Branch Predictor Barrier */
+#define CPUID_8000_0021_EAX_SBPB (1U << 27)
+/* IBPB includes branch type prediction flushing */
+#define CPUID_8000_0021_EAX_IBPB_BRTYPE (1U << 28)
+/* Not vulnerable to Speculative Return Stack Overflow */
+#define CPUID_8000_0021_EAX_SRSO_NO (1U << 29)
+/* Not vulnerable to SRSO at the user-kernel boundary */
+#define CPUID_8000_0021_EAX_SRSO_USER_KERNEL_NO (1U << 30)
#define CPUID_XSAVE_XSAVEOPT (1U << 0)
#define CPUID_XSAVE_XSAVEC (1U << 1)
--
2.39.3

147
kvm-target-i386-Update-EPYC-CPU-model-for-Cache-property.patch Normal file
View File

@ -0,0 +1,147 @@
From 4091d13096918dfff5f3a292b43e613ea888ddc1 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 8 May 2025 14:57:59 -0500
Subject: [PATCH 17/57] target/i386: Update EPYC CPU model for Cache property,
RAS, SVM feature bits
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/8] afc52d066ad5f66732b0bb04142e210c52896708 (johnalle/qemu-kvm-fork)
Found that some of the cache properties are not set correctly for EPYC models.
l1d_cache.no_invd_sharing should not be true.
l1i_cache.no_invd_sharing should not be true.
L2.self_init should be true.
L2.inclusive should be true.
L3.inclusive should not be true.
L3.no_invd_sharing should be true.
Fix the cache properties.
Also add the missing RAS and SVM features bits on AMD
EPYC CPU models. The SVM feature bits are used in nested guests.
succor : Software uncorrectable error containment and recovery capability.
overflow-recov : MCA overflow recovery support.
lbrv : LBR virtualization
tsc-scale : MSR based TSC rate control
vmcb-clean : VMCB clean bits
flushbyasid : Flush by ASID
pause-filter : Pause intercept filter
pfthreshold : PAUSE filter threshold
v-vmsave-vmload : Virtualized VMLOAD and VMSAVE
vgif : Virtualized GIF
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/515941861700d7066186c9600bc5d96a1741ef0c.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit 397db937e85d7b9f5a6f0b30764786cef09d1ff3)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 73 insertions(+)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 4546369836..32c575f63b 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -2166,6 +2166,60 @@ static CPUCaches epyc_v4_cache_info = {
},
};
+static CPUCaches epyc_v5_cache_info = {
+ .l1d_cache = &(CPUCacheInfo) {
+ .type = DATA_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l1i_cache = &(CPUCacheInfo) {
+ .type = INSTRUCTION_CACHE,
+ .level = 1,
+ .size = 64 * KiB,
+ .line_size = 64,
+ .associativity = 4,
+ .partitions = 1,
+ .sets = 256,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l2_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 2,
+ .size = 512 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 1024,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .inclusive = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l3_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 3,
+ .size = 8 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 8192,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .no_invd_sharing = true,
+ .complex_indexing = false,
+ .share_level = CPU_TOPO_LEVEL_DIE,
+ },
+};
+
static const CPUCaches epyc_rome_cache_info = {
.l1d_cache = &(CPUCacheInfo) {
.type = DATA_CACHE,
@@ -5059,6 +5113,25 @@ static const X86CPUDefinition builtin_x86_defs[] = {
},
.cache_info = &epyc_v4_cache_info
},
+ {
+ .version = 5,
+ .props = (PropValue[]) {
+ { "overflow-recov", "on" },
+ { "succor", "on" },
+ { "lbrv", "on" },
+ { "tsc-scale", "on" },
+ { "vmcb-clean", "on" },
+ { "flushbyasid", "on" },
+ { "pause-filter", "on" },
+ { "pfthreshold", "on" },
+ { "v-vmsave-vmload", "on" },
+ { "vgif", "on" },
+ { "model-id",
+ "AMD EPYC-v5 Processor" },
+ { /* end of list */ }
+ },
+ .cache_info = &epyc_v5_cache_info
+ },
{ /* end of list */ }
}
},
--
2.39.3

167
kvm-target-i386-Update-EPYC-Genoa-for-Cache-property-per.patch Normal file
View File

@ -0,0 +1,167 @@
From 768e39f40b394eb4524a83857b86e8f7497f4414 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 8 May 2025 14:58:03 -0500
Subject: [PATCH 21/57] target/i386: Update EPYC-Genoa for Cache property,
perfmon-v2, RAS and SVM feature bits
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [7/8] b144233a1115385a1f792c4454f4511173f753d8 (johnalle/qemu-kvm-fork)
Found that some of the cache properties are not set correctly for EPYC models.
l1d_cache.no_invd_sharing should not be true.
l1i_cache.no_invd_sharing should not be true.
L2.self_init should be true.
L2.inclusive should be true.
L3.inclusive should not be true.
L3.no_invd_sharing should be true.
Fix these cache properties.
Also add the missing RAS and SVM features bits on AMD EPYC-Genoa model.
The SVM feature bits are used in nested guests.
perfmon-v2 : Allow guests to make use of the PerfMonV2 features.
succor : Software uncorrectable error containment and recovery capability.
overflow-recov : MCA overflow recovery support.
lbrv : LBR virtualization
tsc-scale : MSR based TSC rate control
vmcb-clean : VMCB clean bits
flushbyasid : Flush by ASID
pause-filter : Pause intercept filter
pfthreshold : PAUSE filter threshold
v-vmsave-vmload: Virtualized VMLOAD and VMSAVE
vgif : Virtualized GIF
fs-gs-base-ns : WRMSR to {FS,GS,KERNEL_GS}_BASE is non-serializing
The feature details are available in APM listed below [1].
[1] AMD64 Architecture Programmer's Manual Volume 2: System Programming
Publication # 24593 Revision 3.41.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=206537
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/afe3f05d4116124fd5795f28fc23d7b396140313.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit abc92cc8488b5dbcc403b5be24d8092180605101)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 80 ++++++++++++++++++++++++++++++++++++++++++++++-
1 file changed, 79 insertions(+), 1 deletion(-)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 2218071fca..2bc2d41259 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -2598,6 +2598,59 @@ static const CPUCaches epyc_genoa_cache_info = {
},
};
+static const CPUCaches epyc_genoa_v2_cache_info = {
+ .l1d_cache = &(CPUCacheInfo) {
+ .type = DATA_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l1i_cache = &(CPUCacheInfo) {
+ .type = INSTRUCTION_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l2_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 2,
+ .size = 1 * MiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 2048,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .inclusive = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l3_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 3,
+ .size = 32 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 32768,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .no_invd_sharing = true,
+ .complex_indexing = false,
+ .share_level = CPU_TOPO_LEVEL_DIE,
+ },
+};
/* The following VMX features are not supported by KVM and are left out in the
* CPU definitions:
*
@@ -5530,7 +5583,7 @@ static const X86CPUDefinition builtin_x86_defs[] = {
CPUID_8000_0008_EBX_STIBP_ALWAYS_ON |
CPUID_8000_0008_EBX_AMD_SSBD | CPUID_8000_0008_EBX_AMD_PSFD,
.features[FEAT_8000_0021_EAX] =
- CPUID_8000_0021_EAX_No_NESTED_DATA_BP |
+ CPUID_8000_0021_EAX_NO_NESTED_DATA_BP |
CPUID_8000_0021_EAX_LFENCE_ALWAYS_SERIALIZING |
CPUID_8000_0021_EAX_NULL_SEL_CLR_BASE |
CPUID_8000_0021_EAX_AUTO_IBRS,
@@ -5565,6 +5618,31 @@ static const X86CPUDefinition builtin_x86_defs[] = {
.xlevel = 0x80000022,
.model_id = "AMD EPYC-Genoa Processor",
.cache_info = &epyc_genoa_cache_info,
+ .versions = (X86CPUVersionDefinition[]) {
+ { .version = 1 },
+ {
+ .version = 2,
+ .props = (PropValue[]) {
+ { "overflow-recov", "on" },
+ { "succor", "on" },
+ { "lbrv", "on" },
+ { "tsc-scale", "on" },
+ { "vmcb-clean", "on" },
+ { "flushbyasid", "on" },
+ { "pause-filter", "on" },
+ { "pfthreshold", "on" },
+ { "v-vmsave-vmload", "on" },
+ { "vgif", "on" },
+ { "fs-gs-base-ns", "on" },
+ { "perfmon-v2", "on" },
+ { "model-id",
+ "AMD EPYC-Genoa-v2 Processor" },
+ { /* end of list */ }
+ },
+ .cache_info = &epyc_genoa_v2_cache_info
+ },
+ { /* end of list */ }
+ }
},
};
--
2.39.3

146
kvm-target-i386-Update-EPYC-Milan-CPU-model-for-Cache-pr.patch Normal file
View File

@ -0,0 +1,146 @@
From a2cd6a5aac0ba2bbb50d2ff22b83c8b9d7761028 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 8 May 2025 14:58:01 -0500
Subject: [PATCH 19/57] target/i386: Update EPYC-Milan CPU model for Cache
property, RAS, SVM feature bits
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [5/8] e9e34ade25cb7be05d40745e1d074c0356d1923f (johnalle/qemu-kvm-fork)
Found that some of the cache properties are not set correctly for EPYC models.
l1d_cache.no_invd_sharing should not be true.
l1i_cache.no_invd_sharing should not be true.
L2.self_init should be true.
L2.inclusive should be true.
L3.inclusive should not be true.
L3.no_invd_sharing should be true.
Fix these cache properties.
Also add the missing RAS and SVM features bits on AMD EPYC-Milan model.
The SVM feature bits are used in nested guests.
succor : Software uncorrectable error containment and recovery capability.
overflow-recov : MCA overflow recovery support.
lbrv : LBR virtualization
tsc-scale : MSR based TSC rate control
vmcb-clean : VMCB clean bits
flushbyasid : Flush by ASID
pause-filter : Pause intercept filter
pfthreshold : PAUSE filter threshold
v-vmsave-vmload : Virtualized VMLOAD and VMSAVE
vgif : Virtualized GIF
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/c619c0e09a9d5d496819ed48d69181d65f416891.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit fc014d9ba5b26b27401e0e88a4e1ef827c68fe64)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 73 insertions(+)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index a73b5bfca4..7d48c51767 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -2490,6 +2490,60 @@ static const CPUCaches epyc_milan_v2_cache_info = {
},
};
+static const CPUCaches epyc_milan_v3_cache_info = {
+ .l1d_cache = &(CPUCacheInfo) {
+ .type = DATA_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l1i_cache = &(CPUCacheInfo) {
+ .type = INSTRUCTION_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l2_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 2,
+ .size = 512 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 1024,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .inclusive = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l3_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 3,
+ .size = 32 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 32768,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .no_invd_sharing = true,
+ .complex_indexing = false,
+ .share_level = CPU_TOPO_LEVEL_DIE,
+ },
+};
+
static const CPUCaches epyc_genoa_cache_info = {
.l1d_cache = &(CPUCacheInfo) {
.type = DATA_CACHE,
@@ -5418,6 +5472,25 @@ static const X86CPUDefinition builtin_x86_defs[] = {
},
.cache_info = &epyc_milan_v2_cache_info
},
+ {
+ .version = 3,
+ .props = (PropValue[]) {
+ { "overflow-recov", "on" },
+ { "succor", "on" },
+ { "lbrv", "on" },
+ { "tsc-scale", "on" },
+ { "vmcb-clean", "on" },
+ { "flushbyasid", "on" },
+ { "pause-filter", "on" },
+ { "pfthreshold", "on" },
+ { "v-vmsave-vmload", "on" },
+ { "vgif", "on" },
+ { "model-id",
+ "AMD EPYC-Milan-v3 Processor" },
+ { /* end of list */ }
+ },
+ .cache_info = &epyc_milan_v3_cache_info
+ },
{ /* end of list */ }
}
},
--
2.39.3

147
kvm-target-i386-Update-EPYC-Rome-CPU-model-for-Cache-pro.patch Normal file
View File

@ -0,0 +1,147 @@
From dc86ee01fb27b174871ff8be9095ed1a20513772 Mon Sep 17 00:00:00 2001
From: Babu Moger <babu.moger@amd.com>
Date: Thu, 8 May 2025 14:58:00 -0500
Subject: [PATCH 18/57] target/i386: Update EPYC-Rome CPU model for Cache
property, RAS, SVM feature bits
RH-Author: John Allen <None>
RH-MergeRequest: 378: Update EPYC Models and Feature Bits
RH-Jira: RHEL-52649
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [4/8] f23618215eee3c54d9ba52c5b74f5d574c522649 (johnalle/qemu-kvm-fork)
Found that some of the cache properties are not set correctly for EPYC models.
l1d_cache.no_invd_sharing should not be true.
l1i_cache.no_invd_sharing should not be true.
L2.self_init should be true.
L2.inclusive should be true.
L3.inclusive should not be true.
L3.no_invd_sharing should be true.
Fix these cache properties.
Also add the missing RAS and SVM features bits on AMD EPYC-Rome. The SVM
feature bits are used in nested guests.
succor : Software uncorrectable error containment and recovery capability.
overflow-recov : MCA overflow recovery support.
lbrv : LBR virtualization
tsc-scale : MSR based TSC rate control
vmcb-clean : VMCB clean bits
flushbyasid : Flush by ASID
pause-filter : Pause intercept filter
pfthreshold : PAUSE filter threshold
v-vmsave-vmload : Virtualized VMLOAD and VMSAVE
vgif : Virtualized GIF
Signed-off-by: Babu Moger <babu.moger@amd.com>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Link: https://lore.kernel.org/r/8265af72057b84c99ac3a02a5487e32759cc69b1.1746734284.git.babu.moger@amd.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
(cherry picked from commit 83d940e9700527ff080416ce2fa52ee1f4771d72)
JIRA: https://issues.redhat.com/browse/RHEL-52649
Signed-off-by: John Allen <johnalle@redhat.com>
---
target/i386/cpu.c | 73 +++++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 73 insertions(+)
diff --git a/target/i386/cpu.c b/target/i386/cpu.c
index 32c575f63b..a73b5bfca4 100644
--- a/target/i386/cpu.c
+++ b/target/i386/cpu.c
@@ -2328,6 +2328,60 @@ static const CPUCaches epyc_rome_v3_cache_info = {
},
};
+static const CPUCaches epyc_rome_v5_cache_info = {
+ .l1d_cache = &(CPUCacheInfo) {
+ .type = DATA_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l1i_cache = &(CPUCacheInfo) {
+ .type = INSTRUCTION_CACHE,
+ .level = 1,
+ .size = 32 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 64,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l2_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 2,
+ .size = 512 * KiB,
+ .line_size = 64,
+ .associativity = 8,
+ .partitions = 1,
+ .sets = 1024,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .inclusive = true,
+ .share_level = CPU_TOPO_LEVEL_CORE,
+ },
+ .l3_cache = &(CPUCacheInfo) {
+ .type = UNIFIED_CACHE,
+ .level = 3,
+ .size = 16 * MiB,
+ .line_size = 64,
+ .associativity = 16,
+ .partitions = 1,
+ .sets = 16384,
+ .lines_per_tag = 1,
+ .self_init = true,
+ .no_invd_sharing = true,
+ .complex_indexing = false,
+ .share_level = CPU_TOPO_LEVEL_DIE,
+ },
+};
+
static const CPUCaches epyc_milan_cache_info = {
.l1d_cache = &(CPUCacheInfo) {
.type = DATA_CACHE,
@@ -5270,6 +5324,25 @@ static const X86CPUDefinition builtin_x86_defs[] = {
{ /* end of list */ }
},
},
+ {
+ .version = 5,
+ .props = (PropValue[]) {
+ { "overflow-recov", "on" },
+ { "succor", "on" },
+ { "lbrv", "on" },
+ { "tsc-scale", "on" },
+ { "vmcb-clean", "on" },
+ { "flushbyasid", "on" },
+ { "pause-filter", "on" },
+ { "pfthreshold", "on" },
+ { "v-vmsave-vmload", "on" },
+ { "vgif", "on" },
+ { "model-id",
+ "AMD EPYC-Rome-v5 Processor" },
+ { /* end of list */ }
+ },
+ .cache_info = &epyc_rome_v5_cache_info
+ },
{ /* end of list */ }
}
},
--
2.39.3

53
kvm-tests-unit-test-util-sockets-fix-mem-leak-on-error-o.patch Normal file
View File

@ -0,0 +1,53 @@
From 83f09a8c65e1fef416e39d9b0a4ead14ed00601e Mon Sep 17 00:00:00 2001
From: Matheus Tavares Bernardino <matheus.bernardino@oss.qualcomm.com>
Date: Mon, 26 May 2025 10:20:55 -0700
Subject: [PATCH 14/57] tests/unit/test-util-sockets: fix mem-leak on error
object
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [7/7] 31c74f784a26812e5c5898efacda9b4069874ed7 (JurajMarcin/centos-src-qemu-kvm)
The test fails with --enable-asan as the error struct is never freed.
In the case where the test expects a success but it fails, let's also
report the error for debugging (it will be freed internally).
Fixes 316e8ee8d6 ("util/qemu-sockets: Refactor inet_parse() to use QemuOpts")
Signed-off-by: Matheus Tavares Bernardino <matheus.bernardino@oss.qualcomm.com>
Reviewed-by: Juraj Marcin <jmarcin@redhat.com>
Message-ID: <518d94c7db20060b2a086cf55ee9bffab992a907.1748280011.git.matheus.bernardino@oss.qualcomm.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
(cherry picked from commit 5c54a367265ec19ed94a535cd15d178c16b8cae0)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
tests/unit/test-util-sockets.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
index 8492f4d68f..ee66d727c3 100644
--- a/tests/unit/test-util-sockets.c
+++ b/tests/unit/test-util-sockets.c
@@ -341,8 +341,12 @@ static void inet_parse_test_helper(const char *str,
int rc = inet_parse(&addr, str, &error);
if (success) {
+ if (error) {
+ error_report_err(error);
+ }
g_assert_cmpint(rc, ==, 0);
} else {
+ error_free(error);
g_assert_cmpint(rc, <, 0);
}
if (exp_addr != NULL) {
--
2.39.3

97
kvm-ui-vnc-Update-display-update-interval-when-VM-state-.patch Normal file
View File

@ -0,0 +1,97 @@
From a69b8d66fb515cd55cef2fcaa626c350d761f1a9 Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 17:16:13 +0200
Subject: [PATCH 57/57] ui/vnc: Update display update interval when VM state
changes to RUNNING
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 385: ui/vnc: Update display update interval when VM state changes to RUNNING
RH-Jira: RHEL-100741
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Marc-André Lureau <marcandre.lureau@redhat.com>
RH-Commit: [1/1] 30dd4790a607d646465c18d621073df997e8850b (JurajMarcin/centos-src-qemu-kvm)
If a virtual machine is paused for an extended period time, for example,
due to an incoming migration, there are also no changes on the screen.
VNC in such case increases the display update interval by
VNC_REFRESH_INTERVAL_INC (50 ms). The update interval can then grow up
to VNC_REFRESH_INTERVAL_MAX (3000 ms).
When the machine resumes, it can then take up to 3 seconds for the first
display update. Furthermore, the update interval is then halved with
each display update with changes on the screen. If there are moving
elements on the screen, such as a video, this can be perceived as
freezing and stuttering for few seconds before the movement is smooth
again.
This patch resolves this issue, by adding a listener to VM state changes
and changing the update interval when the VM state changes to RUNNING.
The update_displaychangelistener() function updates the internal timer,
and the display is refreshed immediately if the timer is expired.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Link: https://lore.kernel.org/r/20250521151616.3951178-1-jmarcin@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
(cherry picked from commit 0310d594d98b39f9dde79b87fd8b0ad16e7c5459)
JIRA: https://issues.redhat.com/browse/RHEL-100741
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
ui/vnc.c | 12 ++++++++++++
ui/vnc.h | 2 ++
2 files changed, 14 insertions(+)
diff --git a/ui/vnc.c b/ui/vnc.c
index 5057ec8680..4afc925a18 100644
--- a/ui/vnc.c
+++ b/ui/vnc.c
@@ -3386,6 +3386,16 @@ static const DisplayChangeListenerOps dcl_ops = {
.dpy_cursor_define = vnc_dpy_cursor_define,
};
+static void vmstate_change_handler(void *opaque, bool running, RunState state)
+{
+ VncDisplay *vd = opaque;
+
+ if (state != RUN_STATE_RUNNING) {
+ return;
+ }
+ update_displaychangelistener(&vd->dcl, VNC_REFRESH_INTERVAL_BASE);
+}
+
void vnc_display_init(const char *id, Error **errp)
{
VncDisplay *vd;
@@ -3422,6 +3432,8 @@ void vnc_display_init(const char *id, Error **errp)
vd->dcl.ops = &dcl_ops;
register_displaychangelistener(&vd->dcl);
vd->kbd = qkbd_state_init(vd->dcl.con);
+ vd->vmstate_handler_entry = qemu_add_vm_change_state_handler(
+ &vmstate_change_handler, vd);
}
diff --git a/ui/vnc.h b/ui/vnc.h
index e5fa2efa3e..e9da707dbc 100644
--- a/ui/vnc.h
+++ b/ui/vnc.h
@@ -186,6 +186,8 @@ struct VncDisplay
#endif
AudioState *audio_state;
+
+ VMChangeStateEntry *vmstate_handler_entry;
};
typedef struct VncTight {
--
2.39.3

86
kvm-util-qemu-sockets-Add-support-for-keep-alive-flag-to.patch Normal file
View File

@ -0,0 +1,86 @@
From 644b9e34d8c764598e663eb983e2d6eca4ed2510 Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:33 +0200
Subject: [PATCH 11/57] util/qemu-sockets: Add support for keep-alive flag to
passive sockets
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [4/7] a8ec1996262b2bb657b8fe2e72c9045faee0a64c (JurajMarcin/centos-src-qemu-kvm)
Commit aec21d3175 (qapi: Add InetSocketAddress member keep-alive)
introduces the keep-alive flag, which enables the SO_KEEPALIVE socket
option, but only on client-side sockets. However, this option is also
useful for server-side sockets, so they can check if a client is still
reachable or drop the connection otherwise.
This patch enables the SO_KEEPALIVE socket option on passive server-side
sockets if the keep-alive flag is enabled. This socket option is then
inherited by active server-side sockets communicating with connected
clients.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 00064705ed1f3943d3634be25da434466c87e7d5)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
qapi/sockets.json | 4 ++--
util/qemu-sockets.c | 9 +++------
2 files changed, 5 insertions(+), 8 deletions(-)
diff --git a/qapi/sockets.json b/qapi/sockets.json
index 6a95023315..62797cd027 100644
--- a/qapi/sockets.json
+++ b/qapi/sockets.json
@@ -56,8 +56,8 @@
# @ipv6: whether to accept IPv6 addresses, default try both IPv4 and
# IPv6
#
-# @keep-alive: enable keep-alive when connecting to this socket. Not
-# supported for passive sockets. (Since 4.2)
+# @keep-alive: enable keep-alive when connecting to/listening on this socket.
+# (Since 4.2, not supported for listening sockets until 10.1)
#
# @mptcp: enable multi-path TCP. (Since 6.1)
#
diff --git a/util/qemu-sockets.c b/util/qemu-sockets.c
index 631d0c4023..8fc1f86145 100644
--- a/util/qemu-sockets.c
+++ b/util/qemu-sockets.c
@@ -236,12 +236,6 @@ static int inet_listen_saddr(InetSocketAddress *saddr,
int saved_errno = 0;
bool socket_created = false;
- if (saddr->keep_alive) {
- error_setg(errp, "keep-alive option is not supported for passive "
- "sockets");
- return -1;
- }
-
memset(&ai,0, sizeof(ai));
ai.ai_flags = AI_PASSIVE;
if (saddr->has_numeric && saddr->numeric) {
@@ -349,6 +343,9 @@ static int inet_listen_saddr(InetSocketAddress *saddr,
goto fail;
}
/* We have a listening socket */
+ if (inet_set_sockopts(slisten, saddr, errp) < 0) {
+ goto fail;
+ }
freeaddrinfo(res);
return slisten;
}
--
2.39.3

314
kvm-util-qemu-sockets-Introduce-inet-socket-options-cont.patch Normal file
View File

@ -0,0 +1,314 @@
From 3e9458cd71f909474c1dd051f43fd3fbef8d53fd Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:35 +0200
Subject: [PATCH 13/57] util/qemu-sockets: Introduce inet socket options
controlling TCP keep-alive
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [6/7] 3861e7874d5952c53a5020c123b3b2e632149008 (JurajMarcin/centos-src-qemu-kvm)
With the default TCP stack configuration, it could be even 2 hours
before the connection times out due to the other side not being
reachable. However, in some cases, the application needs to be aware of
a connection issue much sooner.
This is the case, for example, for postcopy live migration. If there is
no traffic from the migration destination guest (server-side) to the
migration source guest (client-side), the destination keeps waiting for
pages indefinitely and does not switch to the postcopy-paused state.
This can happen, for example, if the destination QEMU instance is
started with the '-S' command line option and the machine is not started
yet, or if the machine is idle and produces no new page faults for
not-yet-migrated pages.
This patch introduces new inet socket parameters that control count,
idle period, and interval of TCP keep-alive packets before the
connection is considered broken. These parameters are available on
systems where the respective TCP socket options are defined, that
includes Linux, Windows, macOS, but not OpenBSD. Additionally, macOS
defines TCP_KEEPIDLE as TCP_KEEPALIVE instead, so the patch supplies its
own definition.
The default value for all is 0, which means the system configuration is
used.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 1bd4237cb1095d71c16afad3ce93b4a1e453173e)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
meson.build | 30 +++++++++++++
qapi/sockets.json | 19 ++++++++
tests/unit/test-util-sockets.c | 39 +++++++++++++++++
util/qemu-sockets.c | 80 ++++++++++++++++++++++++++++++++++
4 files changed, 168 insertions(+)
diff --git a/meson.build b/meson.build
index 5bb2b757c3..c4539b66c5 100644
--- a/meson.build
+++ b/meson.build
@@ -2581,6 +2581,36 @@ config_host_data.set('HAVE_OPTRESET',
cc.has_header_symbol('getopt.h', 'optreset'))
config_host_data.set('HAVE_IPPROTO_MPTCP',
cc.has_header_symbol('netinet/in.h', 'IPPROTO_MPTCP'))
+config_host_data.set('HAVE_TCP_KEEPCNT',
+ cc.has_header_symbol('netinet/tcp.h', 'TCP_KEEPCNT') or
+ cc.compiles('''
+ #include <ws2tcpip.h>
+ #ifndef TCP_KEEPCNT
+ #error
+ #endif
+ int main(void) { return 0; }''',
+ name: 'Win32 TCP_KEEPCNT'))
+# On Darwin TCP_KEEPIDLE is available under different name, TCP_KEEPALIVE.
+# https://github.com/apple/darwin-xnu/blob/xnu-4570.1.46/bsd/man/man4/tcp.4#L172
+config_host_data.set('HAVE_TCP_KEEPIDLE',
+ cc.has_header_symbol('netinet/tcp.h', 'TCP_KEEPIDLE') or
+ cc.has_header_symbol('netinet/tcp.h', 'TCP_KEEPALIVE') or
+ cc.compiles('''
+ #include <ws2tcpip.h>
+ #ifndef TCP_KEEPIDLE
+ #error
+ #endif
+ int main(void) { return 0; }''',
+ name: 'Win32 TCP_KEEPIDLE'))
+config_host_data.set('HAVE_TCP_KEEPINTVL',
+ cc.has_header_symbol('netinet/tcp.h', 'TCP_KEEPINTVL') or
+ cc.compiles('''
+ #include <ws2tcpip.h>
+ #ifndef TCP_KEEPINTVL
+ #error
+ #endif
+ int main(void) { return 0; }''',
+ name: 'Win32 TCP_KEEPINTVL'))
# has_member
config_host_data.set('HAVE_SIGEV_NOTIFY_THREAD_ID',
diff --git a/qapi/sockets.json b/qapi/sockets.json
index 62797cd027..f9f559daba 100644
--- a/qapi/sockets.json
+++ b/qapi/sockets.json
@@ -59,6 +59,22 @@
# @keep-alive: enable keep-alive when connecting to/listening on this socket.
# (Since 4.2, not supported for listening sockets until 10.1)
#
+# @keep-alive-count: number of keep-alive packets sent before the connection is
+# closed. Only supported for TCP sockets on systems where TCP_KEEPCNT
+# socket option is defined (this includes Linux, Windows, macOS, FreeBSD,
+# but not OpenBSD). When set to 0, system setting is used. (Since 10.1)
+#
+# @keep-alive-idle: time in seconds the connection needs to be idle before
+# sending a keepalive packet. Only supported for TCP sockets on systems
+# where TCP_KEEPIDLE socket option is defined (this includes Linux,
+# Windows, macOS, FreeBSD, but not OpenBSD). When set to 0, system setting
+# is used. (Since 10.1)
+#
+# @keep-alive-interval: time in seconds between keep-alive packets. Only
+# supported for TCP sockets on systems where TCP_KEEPINTVL is defined (this
+# includes Linux, Windows, macOS, FreeBSD, but not OpenBSD). When set to
+# 0, system setting is used. (Since 10.1)
+#
# @mptcp: enable multi-path TCP. (Since 6.1)
#
# Since: 1.3
@@ -71,6 +87,9 @@
'*ipv4': 'bool',
'*ipv6': 'bool',
'*keep-alive': 'bool',
+ '*keep-alive-count': { 'type': 'uint32', 'if': 'HAVE_TCP_KEEPCNT' },
+ '*keep-alive-idle': { 'type': 'uint32', 'if': 'HAVE_TCP_KEEPIDLE' },
+ '*keep-alive-interval': { 'type': 'uint32', 'if': 'HAVE_TCP_KEEPINTVL' },
'*mptcp': { 'type': 'bool', 'if': 'HAVE_IPPROTO_MPTCP' } } }
##
diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
index 9e39b92e7c..8492f4d68f 100644
--- a/tests/unit/test-util-sockets.c
+++ b/tests/unit/test-util-sockets.c
@@ -359,6 +359,24 @@ static void inet_parse_test_helper(const char *str,
g_assert_cmpint(addr.ipv6, ==, exp_addr->ipv6);
g_assert_cmpint(addr.has_keep_alive, ==, exp_addr->has_keep_alive);
g_assert_cmpint(addr.keep_alive, ==, exp_addr->keep_alive);
+#ifdef HAVE_TCP_KEEPCNT
+ g_assert_cmpint(addr.has_keep_alive_count, ==,
+ exp_addr->has_keep_alive_count);
+ g_assert_cmpint(addr.keep_alive_count, ==,
+ exp_addr->keep_alive_count);
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ g_assert_cmpint(addr.has_keep_alive_idle, ==,
+ exp_addr->has_keep_alive_idle);
+ g_assert_cmpint(addr.keep_alive_idle, ==,
+ exp_addr->keep_alive_idle);
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ g_assert_cmpint(addr.has_keep_alive_interval, ==,
+ exp_addr->has_keep_alive_interval);
+ g_assert_cmpint(addr.keep_alive_interval, ==,
+ exp_addr->keep_alive_interval);
+#endif
#ifdef HAVE_IPPROTO_MPTCP
g_assert_cmpint(addr.has_mptcp, ==, exp_addr->has_mptcp);
g_assert_cmpint(addr.mptcp, ==, exp_addr->mptcp);
@@ -460,6 +478,18 @@ static void test_inet_parse_all_options_good(void)
.ipv6 = true,
.has_keep_alive = true,
.keep_alive = true,
+#ifdef HAVE_TCP_KEEPCNT
+ .has_keep_alive_count = true,
+ .keep_alive_count = 10,
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ .has_keep_alive_idle = true,
+ .keep_alive_idle = 60,
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ .has_keep_alive_interval = true,
+ .keep_alive_interval = 30,
+#endif
#ifdef HAVE_IPPROTO_MPTCP
.has_mptcp = true,
.mptcp = false,
@@ -467,6 +497,15 @@ static void test_inet_parse_all_options_good(void)
};
inet_parse_test_helper(
"[::1]:5000,numeric=on,to=5006,ipv4=off,ipv6=on,keep-alive=on"
+#ifdef HAVE_TCP_KEEPCNT
+ ",keep-alive-count=10"
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ ",keep-alive-idle=60"
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ ",keep-alive-interval=30"
+#endif
#ifdef HAVE_IPPROTO_MPTCP
",mptcp=off"
#endif
diff --git a/util/qemu-sockets.c b/util/qemu-sockets.c
index 8017124c74..ef0a137bd2 100644
--- a/util/qemu-sockets.c
+++ b/util/qemu-sockets.c
@@ -45,6 +45,14 @@
# define AI_NUMERICSERV 0
#endif
+/*
+ * On macOS TCP_KEEPIDLE is available under a different name, TCP_KEEPALIVE.
+ * https://github.com/apple/darwin-xnu/blob/xnu-4570.1.46/bsd/man/man4/tcp.4#L172
+ */
+#if defined(TCP_KEEPALIVE) && !defined(TCP_KEEPIDLE)
+# define TCP_KEEPIDLE TCP_KEEPALIVE
+#endif
+
static int inet_getport(struct addrinfo *e)
{
@@ -218,6 +226,42 @@ static int inet_set_sockopts(int sock, InetSocketAddress *saddr, Error **errp)
"Unable to set keep-alive option on socket");
return -1;
}
+#ifdef HAVE_TCP_KEEPCNT
+ if (saddr->has_keep_alive_count && saddr->keep_alive_count) {
+ int keep_count = saddr->keep_alive_count;
+ ret = setsockopt(sock, IPPROTO_TCP, TCP_KEEPCNT, &keep_count,
+ sizeof(keep_count));
+ if (ret < 0) {
+ error_setg_errno(errp, errno,
+ "Unable to set TCP keep-alive count option on socket");
+ return -1;
+ }
+ }
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ if (saddr->has_keep_alive_idle && saddr->keep_alive_idle) {
+ int keep_idle = saddr->keep_alive_idle;
+ ret = setsockopt(sock, IPPROTO_TCP, TCP_KEEPIDLE, &keep_idle,
+ sizeof(keep_idle));
+ if (ret < 0) {
+ error_setg_errno(errp, errno,
+ "Unable to set TCP keep-alive idle option on socket");
+ return -1;
+ }
+ }
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ if (saddr->has_keep_alive_interval && saddr->keep_alive_interval) {
+ int keep_interval = saddr->keep_alive_interval;
+ ret = setsockopt(sock, IPPROTO_TCP, TCP_KEEPINTVL, &keep_interval,
+ sizeof(keep_interval));
+ if (ret < 0) {
+ error_setg_errno(errp, errno,
+ "Unable to set TCP keep-alive interval option on socket");
+ return -1;
+ }
+ }
+#endif
}
return 0;
}
@@ -631,6 +675,24 @@ static QemuOptsList inet_opts = {
.name = "keep-alive",
.type = QEMU_OPT_BOOL,
},
+#ifdef HAVE_TCP_KEEPCNT
+ {
+ .name = "keep-alive-count",
+ .type = QEMU_OPT_NUMBER,
+ },
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ {
+ .name = "keep-alive-idle",
+ .type = QEMU_OPT_NUMBER,
+ },
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ {
+ .name = "keep-alive-interval",
+ .type = QEMU_OPT_NUMBER,
+ },
+#endif
#ifdef HAVE_IPPROTO_MPTCP
{
.name = "mptcp",
@@ -696,6 +758,24 @@ int inet_parse(InetSocketAddress *addr, const char *str, Error **errp)
addr->has_keep_alive = true;
addr->keep_alive = qemu_opt_get_bool(opts, "keep-alive", false);
}
+#ifdef HAVE_TCP_KEEPCNT
+ if (qemu_opt_find(opts, "keep-alive-count")) {
+ addr->has_keep_alive_count = true;
+ addr->keep_alive_count = qemu_opt_get_number(opts, "keep-alive-count", 0);
+ }
+#endif
+#ifdef HAVE_TCP_KEEPIDLE
+ if (qemu_opt_find(opts, "keep-alive-idle")) {
+ addr->has_keep_alive_idle = true;
+ addr->keep_alive_idle = qemu_opt_get_number(opts, "keep-alive-idle", 0);
+ }
+#endif
+#ifdef HAVE_TCP_KEEPINTVL
+ if (qemu_opt_find(opts, "keep-alive-interval")) {
+ addr->has_keep_alive_interval = true;
+ addr->keep_alive_interval = qemu_opt_get_number(opts, "keep-alive-interval", 0);
+ }
+#endif
#ifdef HAVE_IPPROTO_MPTCP
if (qemu_opt_find(opts, "mptcp")) {
addr->has_mptcp = true;
--
2.39.3

461
kvm-util-qemu-sockets-Refactor-inet_parse-to-use-QemuOpt.patch Normal file
View File

@ -0,0 +1,461 @@
From d2155b6fe1f200a588dd5e95d7587e96109da989 Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:34 +0200
Subject: [PATCH 12/57] util/qemu-sockets: Refactor inet_parse() to use
QemuOpts
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [5/7] 7fd97ed112c6259928d94595c2ae23fe1208621e (JurajMarcin/centos-src-qemu-kvm)
Currently, the inet address parser cannot handle multiple options where
one is prefixed with the name of the other. For example, with the
'keep-alive-idle' option added, the current parser cannot parse
'127.0.0.1:5000,keep-alive-idle=60,keep-alive' correctly. Instead, it
fails with "error parsing 'keep-alive' flag '-idle=60,keep-alive'".
To resolve these issues, this patch rewrites the inet address parsing
using the QemuOpts parser, which the inet_parse_flag() function tries to
mimic. This new parser supports all previously supported options and on
top of that the 'numeric' flag is now also supported. The only
difference is, the new parser produces an error if an unknown option is
passed, instead of silently ignoring it.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 316e8ee8d614f049bfae697570a5e62af450491c)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
tests/unit/test-util-sockets.c | 196 +++++++++++++++++++++++++++++++++
util/qemu-sockets.c | 158 +++++++++++++-------------
2 files changed, 270 insertions(+), 84 deletions(-)
diff --git a/tests/unit/test-util-sockets.c b/tests/unit/test-util-sockets.c
index 4c9dd0b271..9e39b92e7c 100644
--- a/tests/unit/test-util-sockets.c
+++ b/tests/unit/test-util-sockets.c
@@ -332,6 +332,177 @@ static void test_socket_unix_abstract(void)
#endif /* CONFIG_LINUX */
+static void inet_parse_test_helper(const char *str,
+ InetSocketAddress *exp_addr, bool success)
+{
+ InetSocketAddress addr;
+ Error *error = NULL;
+
+ int rc = inet_parse(&addr, str, &error);
+
+ if (success) {
+ g_assert_cmpint(rc, ==, 0);
+ } else {
+ g_assert_cmpint(rc, <, 0);
+ }
+ if (exp_addr != NULL) {
+ g_assert_cmpstr(addr.host, ==, exp_addr->host);
+ g_assert_cmpstr(addr.port, ==, exp_addr->port);
+ /* Own members: */
+ g_assert_cmpint(addr.has_numeric, ==, exp_addr->has_numeric);
+ g_assert_cmpint(addr.numeric, ==, exp_addr->numeric);
+ g_assert_cmpint(addr.has_to, ==, exp_addr->has_to);
+ g_assert_cmpint(addr.to, ==, exp_addr->to);
+ g_assert_cmpint(addr.has_ipv4, ==, exp_addr->has_ipv4);
+ g_assert_cmpint(addr.ipv4, ==, exp_addr->ipv4);
+ g_assert_cmpint(addr.has_ipv6, ==, exp_addr->has_ipv6);
+ g_assert_cmpint(addr.ipv6, ==, exp_addr->ipv6);
+ g_assert_cmpint(addr.has_keep_alive, ==, exp_addr->has_keep_alive);
+ g_assert_cmpint(addr.keep_alive, ==, exp_addr->keep_alive);
+#ifdef HAVE_IPPROTO_MPTCP
+ g_assert_cmpint(addr.has_mptcp, ==, exp_addr->has_mptcp);
+ g_assert_cmpint(addr.mptcp, ==, exp_addr->mptcp);
+#endif
+ }
+
+ g_free(addr.host);
+ g_free(addr.port);
+}
+
+static void test_inet_parse_nohost_good(void)
+{
+ char host[] = "";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ };
+ inet_parse_test_helper(":5000", &exp_addr, true);
+}
+
+static void test_inet_parse_empty_bad(void)
+{
+ inet_parse_test_helper("", NULL, false);
+}
+
+static void test_inet_parse_only_colon_bad(void)
+{
+ inet_parse_test_helper(":", NULL, false);
+}
+
+static void test_inet_parse_ipv4_good(void)
+{
+ char host[] = "127.0.0.1";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ };
+ inet_parse_test_helper("127.0.0.1:5000", &exp_addr, true);
+}
+
+static void test_inet_parse_ipv4_noport_bad(void)
+{
+ inet_parse_test_helper("127.0.0.1", NULL, false);
+}
+
+static void test_inet_parse_ipv6_good(void)
+{
+ char host[] = "::1";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ };
+ inet_parse_test_helper("[::1]:5000", &exp_addr, true);
+}
+
+static void test_inet_parse_ipv6_noend_bad(void)
+{
+ inet_parse_test_helper("[::1", NULL, false);
+}
+
+static void test_inet_parse_ipv6_noport_bad(void)
+{
+ inet_parse_test_helper("[::1]:", NULL, false);
+}
+
+static void test_inet_parse_ipv6_empty_bad(void)
+{
+ inet_parse_test_helper("[]:5000", NULL, false);
+}
+
+static void test_inet_parse_hostname_good(void)
+{
+ char host[] = "localhost";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ };
+ inet_parse_test_helper("localhost:5000", &exp_addr, true);
+}
+
+static void test_inet_parse_all_options_good(void)
+{
+ char host[] = "::1";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ .has_numeric = true,
+ .numeric = true,
+ .has_to = true,
+ .to = 5006,
+ .has_ipv4 = true,
+ .ipv4 = false,
+ .has_ipv6 = true,
+ .ipv6 = true,
+ .has_keep_alive = true,
+ .keep_alive = true,
+#ifdef HAVE_IPPROTO_MPTCP
+ .has_mptcp = true,
+ .mptcp = false,
+#endif
+ };
+ inet_parse_test_helper(
+ "[::1]:5000,numeric=on,to=5006,ipv4=off,ipv6=on,keep-alive=on"
+#ifdef HAVE_IPPROTO_MPTCP
+ ",mptcp=off"
+#endif
+ , &exp_addr, true);
+}
+
+static void test_inet_parse_all_implicit_bool_good(void)
+{
+ char host[] = "::1";
+ char port[] = "5000";
+ InetSocketAddress exp_addr = {
+ .host = host,
+ .port = port,
+ .has_numeric = true,
+ .numeric = true,
+ .has_to = true,
+ .to = 5006,
+ .has_ipv4 = true,
+ .ipv4 = true,
+ .has_ipv6 = true,
+ .ipv6 = true,
+ .has_keep_alive = true,
+ .keep_alive = true,
+#ifdef HAVE_IPPROTO_MPTCP
+ .has_mptcp = true,
+ .mptcp = true,
+#endif
+ };
+ inet_parse_test_helper(
+ "[::1]:5000,numeric,to=5006,ipv4,ipv6,keep-alive"
+#ifdef HAVE_IPPROTO_MPTCP
+ ",mptcp"
+#endif
+ , &exp_addr, true);
+}
+
int main(int argc, char **argv)
{
bool has_ipv4, has_ipv6;
@@ -377,6 +548,31 @@ int main(int argc, char **argv)
test_socket_unix_abstract);
#endif
+ g_test_add_func("/util/socket/inet-parse/nohost-good",
+ test_inet_parse_nohost_good);
+ g_test_add_func("/util/socket/inet-parse/empty-bad",
+ test_inet_parse_empty_bad);
+ g_test_add_func("/util/socket/inet-parse/only-colon-bad",
+ test_inet_parse_only_colon_bad);
+ g_test_add_func("/util/socket/inet-parse/ipv4-good",
+ test_inet_parse_ipv4_good);
+ g_test_add_func("/util/socket/inet-parse/ipv4-noport-bad",
+ test_inet_parse_ipv4_noport_bad);
+ g_test_add_func("/util/socket/inet-parse/ipv6-good",
+ test_inet_parse_ipv6_good);
+ g_test_add_func("/util/socket/inet-parse/ipv6-noend-bad",
+ test_inet_parse_ipv6_noend_bad);
+ g_test_add_func("/util/socket/inet-parse/ipv6-noport-bad",
+ test_inet_parse_ipv6_noport_bad);
+ g_test_add_func("/util/socket/inet-parse/ipv6-empty-bad",
+ test_inet_parse_ipv6_empty_bad);
+ g_test_add_func("/util/socket/inet-parse/hostname-good",
+ test_inet_parse_hostname_good);
+ g_test_add_func("/util/socket/inet-parse/all-options-good",
+ test_inet_parse_all_options_good);
+ g_test_add_func("/util/socket/inet-parse/all-bare-bool-good",
+ test_inet_parse_all_implicit_bool_good);
+
end:
return g_test_run();
}
diff --git a/util/qemu-sockets.c b/util/qemu-sockets.c
index 8fc1f86145..8017124c74 100644
--- a/util/qemu-sockets.c
+++ b/util/qemu-sockets.c
@@ -30,6 +30,7 @@
#include "qapi/qobject-input-visitor.h"
#include "qapi/qobject-output-visitor.h"
#include "qemu/cutils.h"
+#include "qemu/option.h"
#include "trace.h"
#ifndef AI_ADDRCONFIG
@@ -601,115 +602,104 @@ err:
return -1;
}
-/* compatibility wrapper */
-static int inet_parse_flag(const char *flagname, const char *optstr, bool *val,
- Error **errp)
-{
- char *end;
- size_t len;
-
- end = strstr(optstr, ",");
- if (end) {
- if (end[1] == ',') { /* Reject 'ipv6=on,,foo' */
- error_setg(errp, "error parsing '%s' flag '%s'", flagname, optstr);
- return -1;
- }
- len = end - optstr;
- } else {
- len = strlen(optstr);
- }
- if (len == 0 || (len == 3 && strncmp(optstr, "=on", len) == 0)) {
- *val = true;
- } else if (len == 4 && strncmp(optstr, "=off", len) == 0) {
- *val = false;
- } else {
- error_setg(errp, "error parsing '%s' flag '%s'", flagname, optstr);
- return -1;
- }
- return 0;
-}
+static QemuOptsList inet_opts = {
+ .name = "InetSocketAddress",
+ .head = QTAILQ_HEAD_INITIALIZER(inet_opts.head),
+ .implied_opt_name = "addr",
+ .desc = {
+ {
+ .name = "addr",
+ .type = QEMU_OPT_STRING,
+ },
+ {
+ .name = "numeric",
+ .type = QEMU_OPT_BOOL,
+ },
+ {
+ .name = "to",
+ .type = QEMU_OPT_NUMBER,
+ },
+ {
+ .name = "ipv4",
+ .type = QEMU_OPT_BOOL,
+ },
+ {
+ .name = "ipv6",
+ .type = QEMU_OPT_BOOL,
+ },
+ {
+ .name = "keep-alive",
+ .type = QEMU_OPT_BOOL,
+ },
+#ifdef HAVE_IPPROTO_MPTCP
+ {
+ .name = "mptcp",
+ .type = QEMU_OPT_BOOL,
+ },
+#endif
+ { /* end of list */ }
+ },
+};
int inet_parse(InetSocketAddress *addr, const char *str, Error **errp)
{
- const char *optstr, *h;
- char host[65];
- char port[33];
- int to;
- int pos;
- char *begin;
-
+ QemuOpts *opts = qemu_opts_parse(&inet_opts, str, true, errp);
+ if (!opts) {
+ return -1;
+ }
memset(addr, 0, sizeof(*addr));
/* parse address */
- if (str[0] == ':') {
- /* no host given */
- host[0] = '\0';
- if (sscanf(str, ":%32[^,]%n", port, &pos) != 1) {
- error_setg(errp, "error parsing port in address '%s'", str);
- return -1;
- }
- } else if (str[0] == '[') {
+ const char *addr_str = qemu_opt_get(opts, "addr");
+ if (!addr_str) {
+ error_setg(errp, "error parsing address ''");
+ return -1;
+ }
+ if (str[0] == '[') {
/* IPv6 addr */
- if (sscanf(str, "[%64[^]]]:%32[^,]%n", host, port, &pos) != 2) {
- error_setg(errp, "error parsing IPv6 address '%s'", str);
+ const char *ip_end = strstr(addr_str, "]:");
+ if (!ip_end || ip_end - addr_str < 2 || strlen(ip_end) < 3) {
+ error_setg(errp, "error parsing IPv6 address '%s'", addr_str);
return -1;
}
+ addr->host = g_strndup(addr_str + 1, ip_end - addr_str - 1);
+ addr->port = g_strdup(ip_end + 2);
} else {
- /* hostname or IPv4 addr */
- if (sscanf(str, "%64[^:]:%32[^,]%n", host, port, &pos) != 2) {
- error_setg(errp, "error parsing address '%s'", str);
+ /* no host, hostname or IPv4 addr */
+ const char *port = strchr(addr_str, ':');
+ if (!port || strlen(port) < 2) {
+ error_setg(errp, "error parsing address '%s'", addr_str);
return -1;
}
+ addr->host = g_strndup(addr_str, port - addr_str);
+ addr->port = g_strdup(port + 1);
}
- addr->host = g_strdup(host);
- addr->port = g_strdup(port);
-
/* parse options */
- optstr = str + pos;
- h = strstr(optstr, ",to=");
- if (h) {
- h += 4;
- if (sscanf(h, "%d%n", &to, &pos) != 1 ||
- (h[pos] != '\0' && h[pos] != ',')) {
- error_setg(errp, "error parsing to= argument");
- return -1;
- }
+ if (qemu_opt_find(opts, "numeric")) {
+ addr->has_numeric = true,
+ addr->numeric = qemu_opt_get_bool(opts, "numeric", false);
+ }
+ if (qemu_opt_find(opts, "to")) {
addr->has_to = true;
- addr->to = to;
+ addr->to = qemu_opt_get_number(opts, "to", 0);
}
- begin = strstr(optstr, ",ipv4");
- if (begin) {
- if (inet_parse_flag("ipv4", begin + 5, &addr->ipv4, errp) < 0) {
- return -1;
- }
+ if (qemu_opt_find(opts, "ipv4")) {
addr->has_ipv4 = true;
+ addr->ipv4 = qemu_opt_get_bool(opts, "ipv4", false);
}
- begin = strstr(optstr, ",ipv6");
- if (begin) {
- if (inet_parse_flag("ipv6", begin + 5, &addr->ipv6, errp) < 0) {
- return -1;
- }
+ if (qemu_opt_find(opts, "ipv6")) {
addr->has_ipv6 = true;
+ addr->ipv6 = qemu_opt_get_bool(opts, "ipv6", false);
}
- begin = strstr(optstr, ",keep-alive");
- if (begin) {
- if (inet_parse_flag("keep-alive", begin + strlen(",keep-alive"),
- &addr->keep_alive, errp) < 0)
- {
- return -1;
- }
+ if (qemu_opt_find(opts, "keep-alive")) {
addr->has_keep_alive = true;
+ addr->keep_alive = qemu_opt_get_bool(opts, "keep-alive", false);
}
#ifdef HAVE_IPPROTO_MPTCP
- begin = strstr(optstr, ",mptcp");
- if (begin) {
- if (inet_parse_flag("mptcp", begin + strlen(",mptcp"),
- &addr->mptcp, errp) < 0)
- {
- return -1;
- }
+ if (qemu_opt_find(opts, "mptcp")) {
addr->has_mptcp = true;
+ addr->mptcp = qemu_opt_get_bool(opts, "mptcp", 0);
}
#endif
return 0;
--
2.39.3

83
kvm-util-qemu-sockets-Refactor-setting-client-sockopts-i.patch Normal file
View File

@ -0,0 +1,83 @@
From cc7fbd3aabe3be1e2966472a151dc618be02ac4c Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:31 +0200
Subject: [PATCH 09/57] util/qemu-sockets: Refactor setting client sockopts
into a separate function
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [2/7] b9b258d9a21a31ead1dd4488f1f55f1728fa8b41 (JurajMarcin/centos-src-qemu-kvm)
This is done in preparation for enabling the SO_KEEPALIVE support for
server sockets and adding settings for more TCP keep-alive socket
options.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit b8b5278aca78be4a1c2e7cbb11c6be176f63706d)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
util/qemu-sockets.c | 29 +++++++++++++++++++----------
1 file changed, 19 insertions(+), 10 deletions(-)
diff --git a/util/qemu-sockets.c b/util/qemu-sockets.c
index 60c44b2b56..2c0e4883ce 100644
--- a/util/qemu-sockets.c
+++ b/util/qemu-sockets.c
@@ -205,6 +205,22 @@ static int try_bind(int socket, InetSocketAddress *saddr, struct addrinfo *e)
#endif
}
+static int inet_set_sockopts(int sock, InetSocketAddress *saddr, Error **errp)
+{
+ if (saddr->keep_alive) {
+ int keep_alive = 1;
+ int ret = setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE,
+ &keep_alive, sizeof(keep_alive));
+
+ if (ret < 0) {
+ error_setg_errno(errp, errno,
+ "Unable to set keep-alive option on socket");
+ return -1;
+ }
+ }
+ return 0;
+}
+
static int inet_listen_saddr(InetSocketAddress *saddr,
int port_offset,
int num,
@@ -476,16 +492,9 @@ int inet_connect_saddr(InetSocketAddress *saddr, Error **errp)
return sock;
}
- if (saddr->keep_alive) {
- int val = 1;
- int ret = setsockopt(sock, SOL_SOCKET, SO_KEEPALIVE,
- &val, sizeof(val));
-
- if (ret < 0) {
- error_setg_errno(errp, errno, "Unable to set KEEPALIVE");
- close(sock);
- return -1;
- }
+ if (inet_set_sockopts(sock, saddr, errp) < 0) {
+ close(sock);
+ return -1;
}
return sock;
--
2.39.3

141
kvm-util-qemu-sockets-Refactor-success-and-failure-paths.patch Normal file
View File

@ -0,0 +1,141 @@
From 17af9176a1fff5af0d4c1bf1beb52228f7fc324d Mon Sep 17 00:00:00 2001
From: Juraj Marcin <jmarcin@redhat.com>
Date: Wed, 21 May 2025 15:52:32 +0200
Subject: [PATCH 10/57] util/qemu-sockets: Refactor success and failure paths
in inet_listen_saddr()
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
RH-Author: Juraj Marcin <None>
RH-MergeRequest: 369: util/qemu-sockets: Introduce inet socket options controlling TCP keep-alive
RH-Jira: RHEL-67104
RH-Acked-by: Peter Xu <peterx@redhat.com>
RH-Acked-by: Miroslav Rezanina <mrezanin@redhat.com>
RH-Commit: [3/7] 5d5f0e704cdf3454a3c402ea7b130aa42705eb9f (JurajMarcin/centos-src-qemu-kvm)
To get a listening socket, we need to first create a socket, try binding
it to a certain port, and lastly starting listening to it. Each of these
operations can fail due to various reasons, one of them being that the
requested address/port is already in use. In such case, the function
tries the same process with a new port number.
This patch refactors the port number loop, so the success path is no
longer buried inside the 'if' statements in the middle of the loop. Now,
the success path is not nested and ends at the end of the iteration
after successful socket creation, binding, and listening. In case any of
the operations fails, it either continues to the next iteration (and the
next port) or jumps out of the loop to handle the error and exits the
function.
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 911e0f2c6e2d00c985affa75ec188c8edcf480f2)
JIRA: https://issues.redhat.com/browse/RHEL-67104
Signed-off-by: Juraj Marcin <jmarcin@redhat.com>
---
util/qemu-sockets.c | 51 ++++++++++++++++++++++++---------------------
1 file changed, 27 insertions(+), 24 deletions(-)
diff --git a/util/qemu-sockets.c b/util/qemu-sockets.c
index 2c0e4883ce..631d0c4023 100644
--- a/util/qemu-sockets.c
+++ b/util/qemu-sockets.c
@@ -303,11 +303,20 @@ static int inet_listen_saddr(InetSocketAddress *saddr,
port_min = inet_getport(e);
port_max = saddr->has_to ? saddr->to + port_offset : port_min;
for (p = port_min; p <= port_max; p++) {
+ if (slisten >= 0) {
+ /*
+ * We have a socket we tried with the previous port. It cannot
+ * be rebound, we need to close it and create a new one.
+ */
+ close(slisten);
+ slisten = -1;
+ }
inet_setport(e, p);
slisten = create_fast_reuse_socket(e);
if (slisten < 0) {
- /* First time we expect we might fail to create the socket
+ /*
+ * First time we expect we might fail to create the socket
* eg if 'e' has AF_INET6 but ipv6 kmod is not loaded.
* Later iterations should always succeed if first iteration
* worked though, so treat that as fatal.
@@ -317,40 +326,38 @@ static int inet_listen_saddr(InetSocketAddress *saddr,
} else {
error_setg_errno(errp, errno,
"Failed to recreate failed listening socket");
- goto listen_failed;
+ goto fail;
}
}
socket_created = true;
rc = try_bind(slisten, saddr, e);
if (rc < 0) {
- if (errno != EADDRINUSE) {
- error_setg_errno(errp, errno, "Failed to bind socket");
- goto listen_failed;
- }
- } else {
- if (!listen(slisten, num)) {
- goto listen_ok;
+ if (errno == EADDRINUSE) {
+ /* This port is already used, try the next one */
+ continue;
}
- if (errno != EADDRINUSE) {
- error_setg_errno(errp, errno, "Failed to listen on socket");
- goto listen_failed;
+ error_setg_errno(errp, errno, "Failed to bind socket");
+ goto fail;
+ }
+ if (listen(slisten, num)) {
+ if (errno == EADDRINUSE) {
+ /* This port is already used, try the next one */
+ continue;
}
+ error_setg_errno(errp, errno, "Failed to listen on socket");
+ goto fail;
}
- /* Someone else managed to bind to the same port and beat us
- * to listen on it! Socket semantics does not allow us to
- * recover from this situation, so we need to recreate the
- * socket to allow bind attempts for subsequent ports:
- */
- close(slisten);
- slisten = -1;
+ /* We have a listening socket */
+ freeaddrinfo(res);
+ return slisten;
}
}
error_setg_errno(errp, errno,
socket_created ?
"Failed to find an available port" :
"Failed to create a socket");
-listen_failed:
+fail:
saved_errno = errno;
if (slisten >= 0) {
close(slisten);
@@ -358,10 +365,6 @@ listen_failed:
freeaddrinfo(res);
errno = saved_errno;
return -1;
-
-listen_ok:
- freeaddrinfo(res);
- return slisten;
}
#ifdef _WIN32
--
2.39.3

191
qemu-kvm.spec
View File

@ -149,7 +149,7 @@ Obsoletes: %{name}-block-ssh <= %{epoch}:%{version} \
Summary: QEMU is a machine emulator and virtualizer
Name: qemu-kvm
Version: 9.1.0
Release: 24%{?rcrel}%{?dist}%{?cc_suffix}
Release: 25%{?rcrel}%{?dist}%{?cc_suffix}
# Epoch because we pushed a qemu-1.0 package. AIUI this can't ever be dropped
# Epoch 15 used for RHEL 8
# Epoch 17 used for RHEL 9 (due to release versioning offset in RHEL 8.5)
@ -621,6 +621,120 @@ Patch218: kvm-tests-Add-iotest-mirror-sparse-for-recent-patches.patch
# For RHEL-82906 - --migrate-disks-detect-zeroes doesn't take effect for disk migration [rhel-9.7]
# For RHEL-83015 - Disk size of target raw image is full allocated when doing mirror with default discard value [rhel-9.7]
Patch219: kvm-mirror-Reduce-I-O-when-destination-is-detect-zeroes-.patch
# For RHEL-98554 - [s390x][RHEL9.7.0][virtio_block] there would be memory leak with virtio_blk disks
Patch220: kvm-s390x-Fix-leak-in-machine_set_loadparm.patch
# For RHEL-98554 - [s390x][RHEL9.7.0][virtio_block] there would be memory leak with virtio_blk disks
Patch221: kvm-hw-s390x-ccw-device-Fix-memory-leak-in-loadparm-sett.patch
# For RHEL-66202 - [AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu
Patch222: kvm-amd_iommu-Rename-variable-mmio-to-mr_mmio.patch
# For RHEL-66202 - [AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu
Patch223: kvm-amd_iommu-Add-support-for-pass-though-mode.patch
# For RHEL-66202 - [AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu
Patch224: kvm-amd_iommu-Use-shared-memory-region-for-Interrupt-Rem.patch
# For RHEL-66202 - [AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu
Patch225: kvm-amd_iommu-Send-notification-when-invalidate-interrup.patch
# For RHEL-66202 - [AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu
Patch226: kvm-amd_iommu-Check-APIC-ID-255-for-XTSup.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch227: kvm-io-Fix-partial-struct-copy-in-qio_dns_resolver_looku.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch228: kvm-util-qemu-sockets-Refactor-setting-client-sockopts-i.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch229: kvm-util-qemu-sockets-Refactor-success-and-failure-paths.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch230: kvm-util-qemu-sockets-Add-support-for-keep-alive-flag-to.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch231: kvm-util-qemu-sockets-Refactor-inet_parse-to-use-QemuOpt.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch232: kvm-util-qemu-sockets-Introduce-inet-socket-options-cont.patch
# For RHEL-67104 - postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S'
Patch233: kvm-tests-unit-test-util-sockets-fix-mem-leak-on-error-o.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch234: kvm-target-i386-Expose-bits-related-to-SRSO-vulnerabilit.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch235: kvm-target-i386-Add-PerfMonV2-feature-bit.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch236: kvm-target-i386-Update-EPYC-CPU-model-for-Cache-property.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch237: kvm-target-i386-Update-EPYC-Rome-CPU-model-for-Cache-pro.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch238: kvm-target-i386-Update-EPYC-Milan-CPU-model-for-Cache-pr.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch239: kvm-target-i386-Add-couple-of-feature-bits-in-CPUID_Fn80.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch240: kvm-target-i386-Update-EPYC-Genoa-for-Cache-property-per.patch
# For RHEL-52649 - [AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model
Patch241: kvm-target-i386-Add-support-for-EPYC-Turin-model.patch
# For RHEL-70926 - Qemu/amd-iommu: Advertise a suitable device id
Patch242: kvm-hw-i386-amd_iommu-Assign-pci-id-0x1419-for-the-AMD-I.patch
# For RHEL-70925 - Qemu/amd-iommu: Add ability to manually specify the AMDVI-PCI device
Patch243: kvm-hw-i386-amd_iommu-Isolate-AMDVI-PCI-from-amd-iommu-d.patch
# For RHEL-70925 - Qemu/amd-iommu: Add ability to manually specify the AMDVI-PCI device
Patch244: kvm-hw-i386-amd_iommu-Allow-migration-when-explicitly-cr.patch
# For RHEL-70925 - Qemu/amd-iommu: Add ability to manually specify the AMDVI-PCI device
Patch245: kvm-Enable-amd-iommu-device.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch246: kvm-include-qemu-compiler-add-QEMU_UNINITIALIZED-attribu.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch247: kvm-hw-virtio-virtio-avoid-cost-of-ftrivial-auto-var-ini.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch248: kvm-block-skip-automatic-zero-init-of-large-array-in-ioq.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch249: kvm-chardev-char-fd-skip-automatic-zero-init-of-large-ar.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch250: kvm-chardev-char-pty-skip-automatic-zero-init-of-large-a.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch251: kvm-chardev-char-socket-skip-automatic-zero-init-of-larg.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch252: kvm-hw-audio-ac97-skip-automatic-zero-init-of-large-arra.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch253: kvm-hw-audio-cs4231a-skip-automatic-zero-init-of-large-a.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch254: kvm-hw-audio-es1370-skip-automatic-zero-init-of-large-ar.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch255: kvm-hw-audio-gus-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch256: kvm-hw-audio-marvell_88w8618-skip-automatic-zero-init-of.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch257: kvm-hw-audio-sb16-skip-automatic-zero-init-of-large-arra.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch258: kvm-hw-audio-via-ac97-skip-automatic-zero-init-of-large-.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch259: kvm-hw-char-sclpconsole-lm-skip-automatic-zero-init-of-l.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch260: kvm-hw-dma-xlnx_csu_dma-skip-automatic-zero-init-of-larg.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch261: kvm-hw-display-vmware_vga-skip-automatic-zero-init-of-la.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch262: kvm-hw-hyperv-syndbg-skip-automatic-zero-init-of-large-a.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch263: kvm-hw-misc-aspeed_hace-skip-automatic-zero-init-of-larg.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch264: kvm-hw-net-rtl8139-skip-automatic-zero-init-of-large-arr.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch265: kvm-hw-net-tulip-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch266: kvm-hw-net-virtio-net-skip-automatic-zero-init-of-large-.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch267: kvm-hw-net-xgamc-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch268: kvm-hw-nvme-ctrl-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch269: kvm-hw-ppc-spapr_tpm_proxy-skip-automatic-zero-init-of-l.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch270: kvm-hw-usb-hcd-ohci-skip-automatic-zero-init-of-large-ar.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch271: kvm-hw-scsi-lsi53c895a-skip-automatic-zero-init-of-large.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch272: kvm-hw-scsi-megasas-skip-automatic-zero-init-of-large-ar.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch273: kvm-hw-ufs-lu-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch274: kvm-net-socket-skip-automatic-zero-init-of-large-array.patch
# For RHEL-99888 - -ftrivial-auto-var-init=zero reduced performance [rhel-9]
Patch275: kvm-net-stream-skip-automatic-zero-init-of-large-array.patch
# For RHEL-100741 - Video stuck after switchover phase when play one video during migration [rhel-9]
Patch276: kvm-ui-vnc-Update-display-update-interval-when-VM-state-.patch
%if %{have_clang}
BuildRequires: clang
@ -1696,6 +1810,81 @@ useradd -r -u 107 -g qemu -G kvm -d / -s /sbin/nologin \
%endif
%changelog
* Tue Jul 08 2025 Miroslav Rezanina <mrezanin@redhat.com> - 9.1.0-25
- kvm-s390x-Fix-leak-in-machine_set_loadparm.patch [RHEL-98554]
- kvm-hw-s390x-ccw-device-Fix-memory-leak-in-loadparm-sett.patch [RHEL-98554]
- kvm-amd_iommu-Rename-variable-mmio-to-mr_mmio.patch [RHEL-66202]
- kvm-amd_iommu-Add-support-for-pass-though-mode.patch [RHEL-66202]
- kvm-amd_iommu-Use-shared-memory-region-for-Interrupt-Rem.patch [RHEL-66202]
- kvm-amd_iommu-Send-notification-when-invalidate-interrup.patch [RHEL-66202]
- kvm-amd_iommu-Check-APIC-ID-255-for-XTSup.patch [RHEL-66202]
- kvm-io-Fix-partial-struct-copy-in-qio_dns_resolver_looku.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-setting-client-sockopts-i.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-success-and-failure-paths.patch [RHEL-67104]
- kvm-util-qemu-sockets-Add-support-for-keep-alive-flag-to.patch [RHEL-67104]
- kvm-util-qemu-sockets-Refactor-inet_parse-to-use-QemuOpt.patch [RHEL-67104]
- kvm-util-qemu-sockets-Introduce-inet-socket-options-cont.patch [RHEL-67104]
- kvm-tests-unit-test-util-sockets-fix-mem-leak-on-error-o.patch [RHEL-67104]
- kvm-target-i386-Expose-bits-related-to-SRSO-vulnerabilit.patch [RHEL-52649]
- kvm-target-i386-Add-PerfMonV2-feature-bit.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-CPU-model-for-Cache-property.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Rome-CPU-model-for-Cache-pro.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Milan-CPU-model-for-Cache-pr.patch [RHEL-52649]
- kvm-target-i386-Add-couple-of-feature-bits-in-CPUID_Fn80.patch [RHEL-52649]
- kvm-target-i386-Update-EPYC-Genoa-for-Cache-property-per.patch [RHEL-52649]
- kvm-target-i386-Add-support-for-EPYC-Turin-model.patch [RHEL-52649]
- kvm-hw-i386-amd_iommu-Assign-pci-id-0x1419-for-the-AMD-I.patch [RHEL-70926]
- kvm-hw-i386-amd_iommu-Isolate-AMDVI-PCI-from-amd-iommu-d.patch [RHEL-70925]
- kvm-hw-i386-amd_iommu-Allow-migration-when-explicitly-cr.patch [RHEL-70925]
- kvm-Enable-amd-iommu-device.patch [RHEL-70925]
- kvm-include-qemu-compiler-add-QEMU_UNINITIALIZED-attribu.patch [RHEL-99888]
- kvm-hw-virtio-virtio-avoid-cost-of-ftrivial-auto-var-ini.patch [RHEL-99888]
- kvm-block-skip-automatic-zero-init-of-large-array-in-ioq.patch [RHEL-99888]
- kvm-chardev-char-fd-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-chardev-char-pty-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-chardev-char-socket-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-audio-ac97-skip-automatic-zero-init-of-large-arra.patch [RHEL-99888]
- kvm-hw-audio-cs4231a-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-hw-audio-es1370-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-audio-gus-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-audio-marvell_88w8618-skip-automatic-zero-init-of.patch [RHEL-99888]
- kvm-hw-audio-sb16-skip-automatic-zero-init-of-large-arra.patch [RHEL-99888]
- kvm-hw-audio-via-ac97-skip-automatic-zero-init-of-large-.patch [RHEL-99888]
- kvm-hw-char-sclpconsole-lm-skip-automatic-zero-init-of-l.patch [RHEL-99888]
- kvm-hw-dma-xlnx_csu_dma-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-display-vmware_vga-skip-automatic-zero-init-of-la.patch [RHEL-99888]
- kvm-hw-hyperv-syndbg-skip-automatic-zero-init-of-large-a.patch [RHEL-99888]
- kvm-hw-misc-aspeed_hace-skip-automatic-zero-init-of-larg.patch [RHEL-99888]
- kvm-hw-net-rtl8139-skip-automatic-zero-init-of-large-arr.patch [RHEL-99888]
- kvm-hw-net-tulip-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-net-virtio-net-skip-automatic-zero-init-of-large-.patch [RHEL-99888]
- kvm-hw-net-xgamc-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-nvme-ctrl-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-hw-ppc-spapr_tpm_proxy-skip-automatic-zero-init-of-l.patch [RHEL-99888]
- kvm-hw-usb-hcd-ohci-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-scsi-lsi53c895a-skip-automatic-zero-init-of-large.patch [RHEL-99888]
- kvm-hw-scsi-megasas-skip-automatic-zero-init-of-large-ar.patch [RHEL-99888]
- kvm-hw-ufs-lu-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-net-socket-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-net-stream-skip-automatic-zero-init-of-large-array.patch [RHEL-99888]
- kvm-ui-vnc-Update-display-update-interval-when-VM-state-.patch [RHEL-100741]
- Resolves: RHEL-98554
([s390x][RHEL9.7.0][virtio_block] there would be memory leak with virtio_blk disks)
- Resolves: RHEL-66202
([AMDSERVER 9.6 Feature] qemu: Interrupt Remap support for emulated amd viommu)
- Resolves: RHEL-67104
(postcopy on the destination host can't switch into pause status under the network issue if boot VM with '-S')
- Resolves: RHEL-52649
([AMDSERVER 9.6 Feature] Turin: Qemu EPYC-Turin Model)
- Resolves: RHEL-70926
(Qemu/amd-iommu: Advertise a suitable device id)
- Resolves: RHEL-70925
(Qemu/amd-iommu: Add ability to manually specify the AMDVI-PCI device)
- Resolves: RHEL-99888
(-ftrivial-auto-var-init=zero reduced performance [rhel-9])
- Resolves: RHEL-100741
(Video stuck after switchover phase when play one video during migration [rhel-9])
* Mon Jun 16 2025 Jon Maloy <jmaloy@redhat.com> - 9.1.0-24
- kvm-s390x-pci-add-support-for-guests-that-request-direct.patch [RHEL-11430]
- kvm-s390x-pci-indicate-QEMU-supports-relaxed-translation.patch [RHEL-11430]