linux-stable/arch/s390/kvm/diag.c
Dominik Dingel 3c038e6be0 KVM: async_pf: Async page fault support on s390
This patch enables async page faults for s390 kvm guests.
It provides the userspace API to enable and disable_wait this feature.
The disable_wait will enforce that the feature is off by waiting on it.
Also it includes the diagnose code, called by the guest to enable async page faults.

The async page faults will use an already existing guest interface for this
purpose, as described in "CP Programming Services (SC24-6084)".

Signed-off-by: Dominik Dingel <dingel@linux.vnet.ibm.com>
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
2014-01-30 13:11:02 +01:00

244 lines
6.8 KiB
C

/*
* handling diagnose instructions
*
* Copyright IBM Corp. 2008, 2011
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License (version 2 only)
* as published by the Free Software Foundation.
*
* Author(s): Carsten Otte <cotte@de.ibm.com>
* Christian Borntraeger <borntraeger@de.ibm.com>
*/
#include <linux/kvm.h>
#include <linux/kvm_host.h>
#include <asm/virtio-ccw.h>
#include "kvm-s390.h"
#include "trace.h"
#include "trace-s390.h"
#include "gaccess.h"
static int diag_release_pages(struct kvm_vcpu *vcpu)
{
unsigned long start, end;
unsigned long prefix = vcpu->arch.sie_block->prefix;
start = vcpu->run->s.regs.gprs[(vcpu->arch.sie_block->ipa & 0xf0) >> 4];
end = vcpu->run->s.regs.gprs[vcpu->arch.sie_block->ipa & 0xf] + 4096;
if (start & ~PAGE_MASK || end & ~PAGE_MASK || start > end
|| start < 2 * PAGE_SIZE)
return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
VCPU_EVENT(vcpu, 5, "diag release pages %lX %lX", start, end);
vcpu->stat.diagnose_10++;
/* we checked for start > end above */
if (end < prefix || start >= prefix + 2 * PAGE_SIZE) {
gmap_discard(start, end, vcpu->arch.gmap);
} else {
if (start < prefix)
gmap_discard(start, prefix, vcpu->arch.gmap);
if (end >= prefix)
gmap_discard(prefix + 2 * PAGE_SIZE,
end, vcpu->arch.gmap);
}
return 0;
}
static int __diag_page_ref_service(struct kvm_vcpu *vcpu)
{
struct prs_parm {
u16 code;
u16 subcode;
u16 parm_len;
u16 parm_version;
u64 token_addr;
u64 select_mask;
u64 compare_mask;
u64 zarch;
};
struct prs_parm parm;
int rc;
u16 rx = (vcpu->arch.sie_block->ipa & 0xf0) >> 4;
u16 ry = (vcpu->arch.sie_block->ipa & 0x0f);
unsigned long hva_token = KVM_HVA_ERR_BAD;
if (vcpu->run->s.regs.gprs[rx] & 7)
return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
if (copy_from_guest(vcpu, &parm, vcpu->run->s.regs.gprs[rx], sizeof(parm)))
return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
if (parm.parm_version != 2 || parm.parm_len < 5 || parm.code != 0x258)
return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
switch (parm.subcode) {
case 0: /* TOKEN */
if (vcpu->arch.pfault_token != KVM_S390_PFAULT_TOKEN_INVALID) {
/*
* If the pagefault handshake is already activated,
* the token must not be changed. We have to return
* decimal 8 instead, as mandated in SC24-6084.
*/
vcpu->run->s.regs.gprs[ry] = 8;
return 0;
}
if ((parm.compare_mask & parm.select_mask) != parm.compare_mask ||
parm.token_addr & 7 || parm.zarch != 0x8000000000000000ULL)
return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
hva_token = gfn_to_hva(vcpu->kvm, gpa_to_gfn(parm.token_addr));
if (kvm_is_error_hva(hva_token))
return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING);
vcpu->arch.pfault_token = parm.token_addr;
vcpu->arch.pfault_select = parm.select_mask;
vcpu->arch.pfault_compare = parm.compare_mask;
vcpu->run->s.regs.gprs[ry] = 0;
rc = 0;
break;
case 1: /*
* CANCEL
* Specification allows to let already pending tokens survive
* the cancel, therefore to reduce code complexity, we assume
* all outstanding tokens are already pending.
*/
if (parm.token_addr || parm.select_mask ||
parm.compare_mask || parm.zarch)
return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
vcpu->run->s.regs.gprs[ry] = 0;
/*
* If the pfault handling was not established or is already
* canceled SC24-6084 requests to return decimal 4.
*/
if (vcpu->arch.pfault_token == KVM_S390_PFAULT_TOKEN_INVALID)
vcpu->run->s.regs.gprs[ry] = 4;
else
vcpu->arch.pfault_token = KVM_S390_PFAULT_TOKEN_INVALID;
rc = 0;
break;
default:
rc = -EOPNOTSUPP;
break;
}
return rc;
}
static int __diag_time_slice_end(struct kvm_vcpu *vcpu)
{
VCPU_EVENT(vcpu, 5, "%s", "diag time slice end");
vcpu->stat.diagnose_44++;
kvm_vcpu_on_spin(vcpu);
return 0;
}
static int __diag_time_slice_end_directed(struct kvm_vcpu *vcpu)
{
struct kvm *kvm = vcpu->kvm;
struct kvm_vcpu *tcpu;
int tid;
int i;
tid = vcpu->run->s.regs.gprs[(vcpu->arch.sie_block->ipa & 0xf0) >> 4];
vcpu->stat.diagnose_9c++;
VCPU_EVENT(vcpu, 5, "diag time slice end directed to %d", tid);
if (tid == vcpu->vcpu_id)
return 0;
kvm_for_each_vcpu(i, tcpu, kvm)
if (tcpu->vcpu_id == tid) {
kvm_vcpu_yield_to(tcpu);
break;
}
return 0;
}
static int __diag_ipl_functions(struct kvm_vcpu *vcpu)
{
unsigned int reg = vcpu->arch.sie_block->ipa & 0xf;
unsigned long subcode = vcpu->run->s.regs.gprs[reg] & 0xffff;
VCPU_EVENT(vcpu, 5, "diag ipl functions, subcode %lx", subcode);
switch (subcode) {
case 3:
vcpu->run->s390_reset_flags = KVM_S390_RESET_CLEAR;
break;
case 4:
vcpu->run->s390_reset_flags = 0;
break;
default:
return -EOPNOTSUPP;
}
atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags);
vcpu->run->s390_reset_flags |= KVM_S390_RESET_SUBSYSTEM;
vcpu->run->s390_reset_flags |= KVM_S390_RESET_IPL;
vcpu->run->s390_reset_flags |= KVM_S390_RESET_CPU_INIT;
vcpu->run->exit_reason = KVM_EXIT_S390_RESET;
VCPU_EVENT(vcpu, 3, "requesting userspace resets %llx",
vcpu->run->s390_reset_flags);
trace_kvm_s390_request_resets(vcpu->run->s390_reset_flags);
return -EREMOTE;
}
static int __diag_virtio_hypercall(struct kvm_vcpu *vcpu)
{
int ret;
/* No virtio-ccw notification? Get out quickly. */
if (!vcpu->kvm->arch.css_support ||
(vcpu->run->s.regs.gprs[1] != KVM_S390_VIRTIO_CCW_NOTIFY))
return -EOPNOTSUPP;
/*
* The layout is as follows:
* - gpr 2 contains the subchannel id (passed as addr)
* - gpr 3 contains the virtqueue index (passed as datamatch)
* - gpr 4 contains the index on the bus (optionally)
*/
ret = kvm_io_bus_write_cookie(vcpu->kvm, KVM_VIRTIO_CCW_NOTIFY_BUS,
vcpu->run->s.regs.gprs[2] & 0xffffffff,
8, &vcpu->run->s.regs.gprs[3],
vcpu->run->s.regs.gprs[4]);
/*
* Return cookie in gpr 2, but don't overwrite the register if the
* diagnose will be handled by userspace.
*/
if (ret != -EOPNOTSUPP)
vcpu->run->s.regs.gprs[2] = ret;
/* kvm_io_bus_write_cookie returns -EOPNOTSUPP if it found no match. */
return ret < 0 ? ret : 0;
}
int kvm_s390_handle_diag(struct kvm_vcpu *vcpu)
{
int code = kvm_s390_get_base_disp_rs(vcpu) & 0xffff;
if (vcpu->arch.sie_block->gpsw.mask & PSW_MASK_PSTATE)
return kvm_s390_inject_program_int(vcpu, PGM_PRIVILEGED_OP);
trace_kvm_s390_handle_diag(vcpu, code);
switch (code) {
case 0x10:
return diag_release_pages(vcpu);
case 0x44:
return __diag_time_slice_end(vcpu);
case 0x9c:
return __diag_time_slice_end_directed(vcpu);
case 0x258:
return __diag_page_ref_service(vcpu);
case 0x308:
return __diag_ipl_functions(vcpu);
case 0x500:
return __diag_virtio_hypercall(vcpu);
default:
return -EOPNOTSUPP;
}
}