Add pvpanic device to deal with guest panic event
v21 cover Add pvpanic device to deal with guest panic event
This series introduces a new simulated device, pvpanic, to notify qemu when guest panic event happens.
Along with this series, there are two patches to add seabios ACPI driver and kernel ACPI driver for the device, respectively.
Tested with:
- qemu(kvm)/qemu(tcg)
- qemu piix/q35
- default ioport/custom ioport
Changes from v20:
- rebase
- suppress pvpanic device with all machine types older than pc 1.5
- split fw_cfg header fix off into a seperate patch
Christian Borntraeger (1):
Wire up disabled wait a panicked event on s390
Hu Tao (7):
add a new runstate: RUN_STATE_GUEST_PANICKED add a new qevent: QEVENT_GUEST_PANICKED introduce a new qom device to deal with panicked event
fw_cfg: add required header files
pvpanic: pass configurable ioport to seabios
pvpanic: add document of pvpanic
pvpanic: create pvpanic by default for machine 1.5
QMP/qmp-events.txt | 14 ++++ default-configs/i386-softmmu.mak | 1 + default-configs/x86_64-softmmu.mak | 1 + docs/specs/pvpanic.txt | 39 ++++++++++ hw/i386/pc_piix.c | 12 ++++ hw/i386/pc_q35.c | 7 ++ hw/misc/Makefile.objs | 2 + hw/misc/pvpanic.c | 144 +++++++++++++++++++++++++++++++++++++ hw/nvram/fw_cfg.c | 8 ++- include/hw/i386/pc.h | 3 + include/hw/nvram/fw_cfg.h | 7 ++ include/monitor/monitor.h | 1 + include/sysemu/sysemu.h | 1 + monitor.c | 1 + qapi-schema.json | 5 +- qmp.c | 3 +- target-s390x/kvm.c | 17 ++++- vl.c | 13 +++- 18 files changed, 270 insertions(+), 9 deletions(-) create mode 100644 docs/specs/pvpanic.txt create mode 100644 hw/misc/pvpanic.c
Signature · 1 line
1.8.1.4
PATCH v21 1/8 add a new runstate: RUN_STATE_GUEST_PANICKED
The guest will be in this state when it is panicked.
--- include/sysemu/sysemu.h | 1 + qapi-schema.json | 5 ++++- qmp.c | 3 +-- vl.c | 13 +++++++++++-- 4 files changed, 17 insertions(+), 5 deletions(-)
include/sysemu/sysemu.h +1 -0
@@ -22,6 +22,7 @@ int qemu_uuid_parse(const char *str, uint8_t *uuid);bool runstate_check (RunState state);void runstate_set (RunState new_state);int runstate_is_running (void ); +bool runstate_needs_reset (void );typedef struct vm_change_state_entry VMChangeStateEntry ;typedef void VMChangeStateHandler (void *opaque,int running,RunState state);
qapi-schema.json +4 -1
@@ -174,11 +174,14 @@ # @suspended: guest is suspended (ACPI S3) # # @watchdog: the watchdog action is configured to pause and has been triggered +# +# @guest-panicked: guest has been panicked as a result of guest OS panic ## { 'enum': 'RunState', 'data': [ 'debug', 'inmigrate', 'internal-error', 'io-error', 'paused', 'postmigrate', 'prelaunch', 'finish-migrate', 'restore-vm', - 'running', 'save-vm', 'shutdown', 'suspended', 'watchdog' ] } + 'running', 'save-vm', 'shutdown', 'suspended', 'watchdog', + 'guest-panicked' ] } ## # @SnapshotInfo
qmp.c +1 -2
@@ -149,8 +149,7 @@ void qmp_cont(Error **errp) {Error *local_err =NULL ; -if (runstate_check (RUN_STATE_INTERNAL_ERROR ) || -runstate_check (RUN_STATE_SHUTDOWN )) { +if (runstate_needs_reset ()) {error_set (errp,QERR_RESET_REQUIRED );return ; }else if (runstate_check (RUN_STATE_SUSPENDED )) {
vl.c +11 -2
@@ -594,6 +594,7 @@ static const RunStateTransition runstate_transitions_def[] = { { RUN_STATE_RUNNING, RUN_STATE_SAVE_VM }, { RUN_STATE_RUNNING, RUN_STATE_SHUTDOWN }, { RUN_STATE_RUNNING, RUN_STATE_WATCHDOG }, + { RUN_STATE_RUNNING, RUN_STATE_GUEST_PANICKED }, { RUN_STATE_SAVE_VM, RUN_STATE_RUNNING },
@@ -608,6 +609,8 @@ static const RunStateTransition runstate_transitions_def[] = { { RUN_STATE_WATCHDOG, RUN_STATE_RUNNING }, { RUN_STATE_WATCHDOG, RUN_STATE_FINISH_MIGRATE }, + { RUN_STATE_GUEST_PANICKED, RUN_STATE_PAUSED }, + { RUN_STATE_MAX, RUN_STATE_MAX }, };
@@ -649,6 +652,13 @@ int runstate_is_running(void)return runstate_check (RUN_STATE_RUNNING ); } +bool runstate_needs_reset (void ) +{ +return runstate_check (RUN_STATE_INTERNAL_ERROR ) || +runstate_check (RUN_STATE_SHUTDOWN ) || +runstate_check (RUN_STATE_GUEST_PANICKED ); +} +StatusInfo *qmp_query_status (Error **errp) {StatusInfo *info =g_malloc0 (sizeof (*info));
@@ -1966,8 +1976,7 @@ static bool main_loop_should_exit(void)cpu_synchronize_all_states ();qemu_system_reset (VMRESET_REPORT );resume_all_vcpus (); -if (runstate_check (RUN_STATE_INTERNAL_ERROR ) || -runstate_check (RUN_STATE_SHUTDOWN )) { +if (runstate_needs_reset ()) {runstate_set (RUN_STATE_PAUSED ); } }
Signature · 1 line
1.8.1.4
PATCH v21 2/8 add a new qevent: QEVENT_GUEST_PANICKED
In reply to Hu TaoThis event will be emited when qemu detects guest panic.
--- QMP/qmp-events.txt | 14 ++++++++++++++ include/monitor/monitor.h | 1 + monitor.c | 1 + 3 files changed, 16 insertions(+)
QMP/qmp-events.txt +14 -0
@@ -446,3 +446,17 @@ Example: Note: If action is "reset", "shutdown", or "pause" the WATCHDOG event is followed respectively by the RESET, SHUTDOWN, or STOP events. + +GUEST_PANICKED +-------------- + +Emitted when guest OS panic is detected. + +Data: + +- "action": Action that has been taken (json-string, currently always "pause"). + +Example: + +{ "event": "GUEST_PANICKED", + "data": { "action": "pause" } }
include/monitor/monitor.h +1 -0
@@ -46,6 +46,7 @@ typedef enum MonitorEvent {QEVENT_WAKEUP ,QEVENT_BALLOON_CHANGE ,QEVENT_SPICE_MIGRATE_COMPLETED , + QEVENT_GUEST_PANICKED,/* Add to 'monitor_event_names' array in monitor.c when * defining new events here */
monitor.c +1 -0
@@ -496,6 +496,7 @@ static const char *monitor_event_names[] = { [QEVENT_WAKEUP] = "WAKEUP", [QEVENT_BALLOON_CHANGE] = "BALLOON_CHANGE", [QEVENT_SPICE_MIGRATE_COMPLETED] = "SPICE_MIGRATE_COMPLETED", + [QEVENT_GUEST_PANICKED ] ="GUEST_PANICKED" , };QEMU_BUILD_BUG_ON (ARRAY_SIZE (monitor_event_names) !=QEVENT_MAX )
Signature · 1 line
1.8.1.4
PATCH v21 4/8 fw_cfg: add required header files
In reply to Hu TaoIf fw_cfg.h is included alone, gcc gives error messages like these:
error: unknown type name ‘uint32_t’
error: unknown type name ‘size_t’
error: unknown type name ‘hwaddr’
...
--- include/hw/nvram/fw_cfg.h | 7 +++++++ 1 file changed, 7 insertions(+)
include/hw/nvram/fw_cfg.h +7 -0
@@ -1,6 +1,13 @@#ifndef FW_CFG_H #define FW_CFG_H +#ifndef NO_QEMU_PROTOS +#include <stdint.h> +#include <stddef.h> + +#include "exec/hwaddr.h" +#endif +#define FW_CFG_SIGNATURE 0x00#define FW_CFG_ID 0x01#define FW_CFG_UUID 0x02
Signature · 1 line
1.8.1.4
PATCH v21 5/8 pvpanic: pass configurable ioport to seabios
In reply to Hu TaoThis lets seabios patch the corresponding SSDT entry.
Also add fw_cfg object to /machine/fw_cfg so we can reference it elsewhere.
--- hw/misc/pvpanic.c | 14 ++++++++++++++ hw/nvram/fw_cfg.c | 8 +++++++- 2 files changed, 21 insertions(+), 1 deletion(-)
hw/misc/pvpanic.c +14 -0
@@ -18,6 +18,8 @@#include "sysemu/sysemu.h" #include "qemu/log.h" +#include "hw/nvram/fw_cfg.h" +/* The bit of supported pv event */ #define PVPANIC_F_PANICKED 0
@@ -86,10 +88,22 @@ static const MemoryRegionOps pvpanic_ops = {static int pvpanic_isa_initfn (ISADevice *dev) {PVPanicState *s =ISA_PVPANIC_DEVICE (dev); +static bool port_configured; +void *fw_cfg;memory_region_init_io (&s->io, &pvpanic_ops, s,"pvpanic" ,1 );isa_register_ioport (dev, &s->io, s->ioport); +if (!port_configured) { + fw_cfg =object_resolve_path ("/machine/fw_cfg" ,NULL ); +if (fw_cfg) { +fw_cfg_add_file (fw_cfg,"etc/pvpanic-port" , +g_memdup (&s->ioport,sizeof (s->ioport)), +sizeof (s->ioport)); + port_configured = true; + } + } +return 0 ; }
hw/nvram/fw_cfg.c +7 -1
@@ -489,11 +489,17 @@ FWCfgState *fw_cfg_init(uint32_t ctl_port, uint32_t data_port, dev =qdev_create (NULL ,"fw_cfg" );qdev_prop_set_uint32 (dev,"ctl_iobase" , ctl_port);qdev_prop_set_uint32 (dev,"data_iobase" , data_port); -qdev_init_nofail (dev); d =SYS_BUS_DEVICE (dev); s =DO_UPCAST (FWCfgState, busdev.qdev, dev); +if (!object_resolve_path ("/machine/fw_cfg" ,NULL )) { +object_property_add_child (qdev_get_machine (),"fw_cfg" ,OBJECT (s), +NULL ); + } + +qdev_init_nofail (dev); +if (ctl_addr) {sysbus_mmio_map (d,0 , ctl_addr); }
Signature · 1 line
1.8.1.4
PATCH v21 3/8 introduce a new qom device to deal with panicked event
In reply to Hu Taopvpanic device is used to send guest panic event from guest to qemu.
When guest panic happens, pvpanic device driver will write a event number to IO port 0x505(which is the IO port occupied by pvpanic device, by default). On receiving the event, pvpanic device will pause guest cpu(s), and send a qmp event QEVENT_GUEST_PANICKED.
--- default-configs/i386-softmmu.mak | 1 + default-configs/x86_64-softmmu.mak | 1 + hw/misc/Makefile.objs | 2 + hw/misc/pvpanic.c | 123 +++++++++++++++++++++++++++++++++++++ 4 files changed, 127 insertions(+) create mode 100644 hw/misc/pvpanic.c
default-configs/i386-softmmu.mak +1 -0
@@ -44,3 +44,4 @@ CONFIG_LPC_ICH9=y CONFIG_PCI_Q35=y CONFIG_APIC=y CONFIG_IOAPIC=y +CONFIG_PVPANIC=y
default-configs/x86_64-softmmu.mak +1 -0
@@ -44,3 +44,4 @@ CONFIG_LPC_ICH9=y CONFIG_PCI_Q35=y CONFIG_APIC=y CONFIG_IOAPIC=y +CONFIG_PVPANIC=y
hw/misc/Makefile.objs +2 -0
@@ -39,3 +39,5 @@ obj-$(CONFIG_OMAP) += omap_tap.o obj-$(CONFIG_PXA2XX) += pxa2xx_pcmcia.o obj-$(CONFIG_SLAVIO) += slavio_misc.o obj-$(CONFIG_ZYNQ) += zynq_slcr.o + +obj-$(CONFIG_PVPANIC) += pvpanic.o
hw/misc/pvpanic.c new file +123 -0
123 lines not on this page · raw message
Signature · 1 line
1.8.1.4
PATCH v21 7/8 pvpanic: create pvpanic by default for machine 1.5
In reply to Hu Tao--- hw/i386/pc_piix.c | 12 ++++++++++++ hw/i386/pc_q35.c | 7 +++++++ hw/misc/pvpanic.c | 7 +++++++ include/hw/i386/pc.h | 3 +++ 4 files changed, 29 insertions(+)
hw/i386/pc_piix.c +12 -0
@@ -55,6 +55,8 @@ static const int ide_iobase[MAX_IDE_BUS] = { 0x1f0, 0x170 };static const int ide_iobase2[MAX_IDE_BUS ] = {0x3f6 ,0x376 };static const int ide_irq[MAX_IDE_BUS ] = {14 ,15 }; +static bool has_pvpanic = true; +/* PC hardware initialisation */ static void pc_init1(MemoryRegion *system_memory, MemoryRegion *system_io,
@@ -218,6 +220,10 @@ static void pc_init1(MemoryRegion *system_memory,if (pci_enabled) {pc_pci_device_init (pci_bus); } + +if (has_pvpanic) { +pvpanic_init (isa_bus); + } }static void pc_init_pci (QEMUMachineInitArgs *args)
@@ -238,6 +244,7 @@ static void pc_init_pci(QEMUMachineInitArgs *args)static void pc_init_pci_1_4 (QEMUMachineInitArgs *args) { pc_sysfw_flash_vs_rom_bug_compatible = true; + has_pvpanic = false;pc_init_pci (args); }
@@ -245,6 +252,7 @@ static void pc_init_pci_1_3(QEMUMachineInitArgs *args) {enable_compat_apic_id_mode (); pc_sysfw_flash_vs_rom_bug_compatible = true; + has_pvpanic = false;pc_init_pci (args); }
@@ -254,6 +262,7 @@ static void pc_init_pci_1_2(QEMUMachineInitArgs *args)disable_kvm_pv_eoi ();enable_compat_apic_id_mode (); pc_sysfw_flash_vs_rom_bug_compatible = true; + has_pvpanic = false;pc_init_pci (args); }
@@ -262,6 +271,7 @@ static void pc_init_pci_1_0(QEMUMachineInitArgs *args) {disable_kvm_pv_eoi ();enable_compat_apic_id_mode (); + has_pvpanic = false;pc_init_pci (args); }
@@ -274,6 +284,7 @@ static void pc_init_pci_no_kvmclock(QEMUMachineInitArgs *args)const char *kernel_cmdline = args->kernel_cmdline;const char *initrd_filename = args->initrd_filename;const char *boot_device = args->boot_device; + has_pvpanic = false;disable_kvm_pv_eoi ();enable_compat_apic_id_mode (); pc_init1(get_system_memory(),
@@ -291,6 +302,7 @@ static void pc_init_isa(QEMUMachineInitArgs *args)const char *kernel_cmdline = args->kernel_cmdline;const char *initrd_filename = args->initrd_filename;const char *boot_device = args->boot_device; + has_pvpanic = false;if (cpu_model ==NULL ) cpu_model ="486" ;disable_kvm_pv_eoi ();
hw/i386/pc_q35.c +7 -0
@@ -45,6 +45,8 @@/* ICH9 AHCI has 6 ports */ #define MAX_SATA_PORTS 6 +static bool has_pvpanic = true; +/* PC hardware initialisation */ static void pc_q35_init (QEMUMachineInitArgs *args) {
@@ -194,11 +196,16 @@ static void pc_q35_init(QEMUMachineInitArgs *args)if (pci_enabled) {pc_pci_device_init (host_bus); } + +if (has_pvpanic) { +pvpanic_init (isa_bus); + } }static void pc_q35_init_1_4 (QEMUMachineInitArgs *args) { pc_sysfw_flash_vs_rom_bug_compatible = true; + has_pvpanic = false;pc_q35_init (args); }
hw/misc/pvpanic.c +7 -0
@@ -19,6 +19,7 @@#include "qemu/log.h" #include "hw/nvram/fw_cfg.h" +#include "hw/i386/pc.h" /* The bit of supported pv event */ #define PVPANIC_F_PANICKED 0
@@ -107,6 +108,12 @@ static int pvpanic_isa_initfn(ISADevice *dev)return 0 ; } +int pvpanic_init (ISABus *bus) +{ +isa_create_simple (bus,TYPE_ISA_PVPANIC_DEVICE ); +return 0 ; +} +static Property pvpanic_isa_properties[] = {DEFINE_PROP_UINT16 ("ioport" , PVPanicState, ioport,0x505 ),DEFINE_PROP_END_OF_LIST (),
include/hw/i386/pc.h +3 -0
@@ -171,6 +171,9 @@ static inline bool isa_ne2000_init(ISABus *bus, int base, int irq, NICInfo *nd)extern bool pc_sysfw_flash_vs_rom_bug_compatible;void pc_system_firmware_init (MemoryRegion *rom_memory); +/* pvpanic.c */ +int pvpanic_init (ISABus *bus); +/* e820 types */ #define E820_RAM 1#define E820_RESERVED 2
Signature · 1 line
1.8.1.4
PATCH v21 6/8 pvpanic: add document of pvpanic
In reply to Hu Tao--- docs/specs/pvpanic.txt | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) create mode 100644 docs/specs/pvpanic.txt
docs/specs/pvpanic.txt new file +39 -0
@@ -0,0 +1,39 @@ +PVPANIC DEVICE +============== + +pvpanic device is a simulated ISA device, through which a guest panic +event is sent to qemu, and a QMP event is generated. This allows +management apps (e.g. libvirt) to be notified and respond to the event. + +The management app has the option of waiting for GUEST_PANICKED events, +and/or polling for guest-panicked RunState, to learn when the pvpanic +device has fired a panic event. + +ISA Interface +------------- + +pvpanic exposes a single I/O port, by default 0x505. On read, the bits +recognized by the device are set. Software should ignore bits it doesn't +recognize. On write, the bits not recognized by the device are ignored. +Software should set only bits both itself and the device recognize. +Currently, only bit 0 is recognized, setting it indicates a guest panic +has happened. + +ACPI Interface +-------------- + +pvpanic device is defined with ACPI ID "QEMU0001". Custom methods: + +RDPT: To determine whether guest panic notification is supported. +Arguments: None +Return: Returns a byte, bit 0 set to indicate guest panic + notification is supported. Other bits are reserved and + should be ignored. + +WRPT: To send a guest panic event +Arguments: Arg0 is a byte, with bit 0 set to indicate guest panic has + happened. Other bits are reserved and should be cleared. +Return: None + +The ACPI device will automatically refer to the right port in case it +is modified.
Signature · 1 line
1.8.1.4
PATCH v21 8/8 Wire up disabled wait a panicked event on s390
In reply to Hu TaoFrom: Christian Borntraeger <[email protected]>
On s390 the disabled wait state indicates a state of attention. For example Linux uses that state after a panic. Lets put the system into panicked state.
An alternative implementation would be to state disabled-wait <address> instead of pause in the action field. (e.g. z/OS, z/VM and other classic OSes use the address of the disabled wait to indicate an error code).
--- target-s390x/kvm.c | 17 ++++++++++++++--- 1 file changed, 14 insertions(+), 3 deletions(-)
target-s390x/kvm.c +14 -3
@@ -34,6 +34,8 @@#include "sysemu/kvm.h" #include "cpu.h" #include "sysemu/device_tree.h" +#include "qapi/qmp/qjson.h" +#include "monitor/monitor.h" /* #define DEBUG_KVM */
@@ -705,9 +707,18 @@ static int handle_intercept(S390CPU *cpu) r =handle_instruction (cpu, run);break ;case ICPT_WAITPSW : -if (s390_del_running_cpu (cpu) ==0 && -is_special_wait_psw (cs)) { -qemu_system_shutdown_request (); +/* disabled wait, since enabled wait is handled in kernel */ +if (s390_del_running_cpu (cpu) ==0 ) { +if (is_special_wait_psw (cs)) { +qemu_system_shutdown_request (); + }else { +QObject *data; + + data =qobject_from_jsonf ("{ 'action': %s }" ,"pause" ); +monitor_protocol_event (QEVENT_GUEST_PANICKED , data); +qobject_decref (data); +vm_stop (RUN_STATE_GUEST_PANICKED ); + } } r =EXCP_HALTED ;break ;
Signature · 1 line
1.8.1.4
Re: [Qemu-devel] [PATCH v21 1/8] add a new runstate: RUN_STATE_GUEST_PANICKED
In reply to Hu TaoQuoted 4 lines · Hu Tao
The guest will be in this state when it is panicked.
Signed-off-by Wen Congyang <[email protected]>Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 2/8] add a new qevent: QEVENT_GUEST_PANICKED
In reply to Hu TaoQuoted 4 lines · Hu Tao
This event will be emited when qemu detects guest panic.
Signed-off-by Wen Congyang <[email protected]>Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 3/8] introduce a new qom device to deal with panicked event
In reply to Hu TaoQuoted 9 lines · Hu Tao
pvpanic device is used to send guest panic event from guest to qemu.
When guest panic happens, pvpanic device driver will write a event number to IO port 0x505(which is the IO port occupied by pvpanic device, by default). On receiving the event, pvpanic device will pause guest cpu(s), and send a qmp event
QEVENT_GUEST_PANICKED.Signed-off-by Wen Congyang <[email protected]>Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 4/8] fw_cfg: add required header files
In reply to Hu TaoQuoted 8 lines · Hu Tao
If
fw_cfg.his included alone, gcc gives error messages like these:error: unknown type name ‘uint32_t’
error: unknown type name ‘size_t’
error: unknown type name ‘hwaddr’
...Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 5/8] pvpanic: pass configurable ioport to seabios
In reply to Hu TaoQuoted 6 lines · Hu Tao
This lets seabios patch the corresponding SSDT entry.
Also add
fw_cfgobject to /machine/fw_cfg so we can reference it elsewhere.Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 6/8] pvpanic: add document of pvpanic
In reply to Hu TaoQuoted 1 line · Hu Tao
Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 7/8] pvpanic: create pvpanic by default for machine 1.5
In reply to Hu TaoQuoted 2 lines · Hu Tao
Signed-off-by Paolo Bonzini <[email protected]>Signed-off-by Hu Tao <[email protected]>
Re: [Qemu-devel] [PATCH v21 8/8] Wire up disabled wait a panicked event on s390
In reply to Hu TaoQuoted 6 lines · Hu Tao
From: Christian Borntraeger <[email protected]>
On s390 the disabled wait state indicates a state of attention. For example Linux uses that state after a panic. Lets put the system into panicked state.
Quoted 6 lines
An alternative implementation would be to state disabled-wait <address> instead of pause in the action field. (e.g. z/OS, z/VM and other classic OSes use the address of the disabled wait to indicate an error code).
Signed-off-by Christian Borntraeger <[email protected]>
I'm not qualified to review this one. Thanks for your patience!
Anthony Liguori
AppliedApplied. Thanks.
Regards,
Anthony Liguori