Computer system
There is a need to provide a computer system capable of preventing a failure from propagating and recovering from the failure. VCPU# 0 through VCPU# 2 each operate different OS's. VCPU# 0 operates a management OS that manages the other OS's. When notified of bus error occurrence, a virtual CPU execution portion 201 operates only VCPU# 0 regardless of an execution sequence stored in schedule register A. VCPU# 0 reinitializes a bus where an error occurred.
1. A computer system comprising:
an interrupt controller to notify a bus error occurrence; and a multithreaded processor,
wherein the multithreaded processor includes:
a schedule register that settles a sequence of performing a plurality of virtual CPUs and stores data for virtual CPUs to be performed; and
the schedule register includes a sequence table that is used to cyclically perform virtual CPUs;
a virtual CPU execution portion that performs virtual CPUs according to a sequence settled by the schedule register;
wherein virtual CPUs operate different OS's and include a first virtual CPU that operates a management OS to manage other OS's;
wherein, when notified of bus error occurrence, the virtual CPU execution portion operates only the first virtual CPU and stops other VCPUs regardless of an execution sequence settled in the schedule register, and
wherein the first virtual CPU reinitializes a bus where an error occurred;
wherein the first virtual CPU requests the other virtual CPUs to reinitialize a device to be used after bus reinitialization and allows the virtual CPU execution portion to perform a virtual CPU according to a sequence settled in the schedule register, and
wherein the other virtual CPUs reinitialize devices to be used.
2. The computer system according to claim 1 ,
wherein the first virtual CPU changes the execution sequence before the bus reinitialization so as to operate only the first virtual CPU and a second virtual CPU using a device not coupled to a bus where an error occurred.
3. The computer system according to claim 2 ,
wherein the first virtual CPU requests a third virtual CPU after bus reinitialization, using a device coupled to a bus causing an error, to reinitialize a device to be used and restores the execution sequence to an original state, and
wherein the third virtual CPU reinitializes a device to be used.
4. A computer system comprising:
an interrupt controller to notify a bus error occurrence; and
a multithreaded processor,
wherein the multithreaded processor includes:
a plurality of schedule registers each of which settles a sequence of performing a plurality of virtual CPUs and stores data for virtual CPUs to be performed; and
each schedule register includes a sequence table that is used to cyclically perform virtual CPUs;
a virtual CPU execution portion that performs virtual CPUs according to a sequence settled by a selected one of the schedule registers;
wherein virtual CPUs operate different OS's and include a first virtual CPU that operates a management OS to manage other OS's;
wherein, before being notified of bus error occurrence, the virtual CPU execution portion selects a first schedule register that includes all virtual CPUs as virtual CPUs to be performed;
wherein, when notified of bus error occurrence, the virtual CPU execution portion selects a schedule register that does not include, as a virtual CPU to be performed, a second virtual CPU using a device coupled to a bus where the error occurred, and
wherein the first virtual CPU reinitializes a bus where an error occurred;
wherein, after bus reinitialization, the first virtual CPU requests the second virtual CPU to reinitialize a device to be used and reenables the first schedule register to be selected, and
wherein the second virtual CPU reinitializes a device to be used.