accel/kvm: Simplify few functions that can use global kvm_state

PATCH In review Philippe Mathieu-Daudé 3 participants 16 messages Jun 23, 2020 – Jul 5, 2020 KVM

lore

Share & tools

lore

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #0 lore raw

cover accel/kvm: Simplify few functions that can use global kvm_state

Following Paolo's idea on kvm_check_extension():
https://www.mail-archive.com/[email protected]/msg713794.html

CI:
https://travis-ci.org/github/philmd/qemu/builds/701213438

Philippe Mathieu-Daudé (7):
accel/kvm: Let kvm_check_extension use global KVM state accel/kvm: Simplify kvm_check_extension() accel/kvm: Simplify kvm_check_extension_list() accel/kvm: Simplify kvm_set_sigmask_len() target/i386/kvm: Simplify get_para_features() target/i386/kvm: Simplify kvm_get_mce_cap_supported() target/i386/kvm: Simplify kvm_get_supported_[feature]_msrs()

 include/sysemu/kvm.h         |  4 +-
 accel/kvm/kvm-all.c          | 76 +++++++++++++++----------------
 hw/hyperv/hyperv.c           |  2 +-
 hw/i386/kvm/clock.c          |  2 +-
 hw/i386/kvm/i8254.c          |  4 +-
 hw/i386/kvm/ioapic.c         |  2 +-
 hw/intc/arm_gic_kvm.c        |  2 +-
 hw/intc/openpic_kvm.c        |  2 +-
 hw/intc/xics_kvm.c           |  2 +-
 hw/s390x/s390-stattrib-kvm.c |  2 +-
 target/arm/kvm.c             | 13 +++---
 target/arm/kvm32.c           |  2 +-
 target/arm/kvm64.c           | 15 +++---
 target/i386/kvm.c            | 88 +++++++++++++++++-------------------
 target/mips/kvm.c            |  6 +--
 target/ppc/kvm.c             | 34 +++++++-------
 target/s390x/cpu_models.c    |  3 +-
 target/s390x/kvm.c           | 30 ++++++------
 18 files changed, 141 insertions(+), 148 deletions(-)

Signature · 1 line

2.21.3

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #1 lore raw

PATCH 1/7 accel/kvm: Let kvm_check_extension use global KVM state

As KVM supported extentions those should be the same for all VMs, it is safe to directly use the global kvm_state in kvm_check_extension().

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 accel/kvm/kvm-all.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

accel/kvm/kvm-all.c +1 -1

index f24d7da783..934a7d6b24 100644
@@ -913,7 +913,7 @@ int kvm_check_extension(KVMState *s, unsigned int extension)
 {
     int ret;
 
-    ret = kvm_ioctl(s, KVM_CHECK_EXTENSION, extension);
+    ret = kvm_ioctl(kvm_state, KVM_CHECK_EXTENSION, extension);
     if (ret < 0) {
         ret = 0;
     }

Signature · 1 line

2.21.3

Cornelia Huck

Reviewed-by on v1 1/7

2020-06-23 15:25 UTC #2 lore raw

Re: [PATCH 1/7] accel/kvm: Let kvm_check_extension use global KVM state

Quoted 15 lines · On Tue, 23 Jun 2020 12:50:46 +0200 Philippe Mathieu-Daudé <[email protected]>…

As KVM supported extentions those should be the same for all VMs, it is safe to directly use the global kvm_state in kvm_check_extension().
Suggested-by Paolo Bonzini <[email protected]>
Signed-off-by Philippe Mathieu-Daudé <[email protected]>
---
 accel/kvm/kvm-all.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
accel/kvm/kvm-all.c +0 -0
index f24d7da783..934a7d6b24 100644

accel/kvm/kvm-all.c int kvm_check_extension(KVMState *s, unsigned int extension) · line 919

Quoted 8 lines

accel/kvm/kvm-all.c +1 -1

@@ -913,7 +913,7 @@ int kvm_check_extension(KVMState *s, unsigned int extension)
 {
     int ret;
 
-    ret = kvm_ioctl(s, KVM_CHECK_EXTENSION, extension);
+    ret = kvm_ioctl(kvm_state, KVM_CHECK_EXTENSION, extension);
     if (ret < 0) {
         ret = 0;
     }

Reviewed-by Cornelia Huck <[email protected]>

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #3 lore raw

PATCH 2/7 accel/kvm: Simplify kvm_check_extension()

In reply to Philippe Mathieu-Daudé

In previous commit we let kvm_check_extension() use the global kvm_state. Since the KVMState* argument is now unused, drop it.

Convert callers with this Coccinelle script:

@@
expression kvm_state, extension;
@@
- kvm_check_extension(kvm_state, extension)
+ kvm_check_extension(extension)

Unused variables manually removed:
- CPUState* in hyperv_enabled()
- KVMState* in kvm_arm_get_max_vm_ipa_size()

Inspired-by Paolo Bonzini <[email protected]>

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 include/sysemu/kvm.h         |  2 +-
 accel/kvm/kvm-all.c          | 64 ++++++++++++++++++------------------
 hw/hyperv/hyperv.c           |  2 +-
 hw/i386/kvm/clock.c          |  2 +-
 hw/i386/kvm/i8254.c          |  4 +--
 hw/i386/kvm/ioapic.c         |  2 +-
 hw/intc/arm_gic_kvm.c        |  2 +-
 hw/intc/openpic_kvm.c        |  2 +-
 hw/intc/xics_kvm.c           |  2 +-
 hw/s390x/s390-stattrib-kvm.c |  2 +-
 target/arm/kvm.c             | 13 ++++----
 target/arm/kvm32.c           |  2 +-
 target/arm/kvm64.c           | 15 ++++-----
 target/i386/kvm.c            | 61 ++++++++++++++++------------------
 target/mips/kvm.c            |  4 +--
 target/ppc/kvm.c             | 34 +++++++++----------
 target/s390x/cpu_models.c    |  3 +-
 target/s390x/kvm.c           | 30 ++++++++---------
 18 files changed, 119 insertions(+), 127 deletions(-)

include/sysemu/kvm.h +1 -1

index b4174d941c..3662641c99 100644
@@ -436,7 +436,7 @@ void kvm_arch_update_guest_debug(CPUState *cpu, struct kvm_guest_debug *dbg);
 
 bool kvm_arch_stop_on_emulation_error(CPUState *cpu);
 
-int kvm_check_extension(KVMState *s, unsigned int extension);
+int kvm_check_extension(unsigned int extension);
 
 int kvm_vm_check_extension(KVMState *s, unsigned int extension);

accel/kvm/kvm-all.c +32 ~~-32~~

190 lines not on this page · raw message

hw/hyperv/hyperv.c +1 -1

index 844d00776d..92720efc3d 100644
@@ -605,7 +605,7 @@ static bool process_event_flags_userspace;
 int hyperv_set_event_flag_handler(uint32_t conn_id, EventNotifier *notifier)
 {
     if (!process_event_flags_userspace &&
-        !kvm_check_extension(kvm_state, KVM_CAP_HYPERV_EVENTFD)) {
+        !kvm_check_extension(KVM_CAP_HYPERV_EVENTFD)) {
         process_event_flags_userspace = true;
 
         warn_report("Hyper-V event signaling is not supported by this kernel; "

hw/i386/kvm/clock.c +1 -1

index 64283358f9..c0bfc69349 100644
@@ -166,7 +166,7 @@ static void kvmclock_vm_state_change(void *opaque, int running,
 {
     KVMClockState *s = opaque;
     CPUState *cpu;
-    int cap_clock_ctrl = kvm_check_extension(kvm_state, KVM_CAP_KVMCLOCK_CTRL);
+    int cap_clock_ctrl = kvm_check_extension(KVM_CAP_KVMCLOCK_CTRL);
     int ret;
 
     if (running) {

hw/i386/kvm/i8254.c +2 -2

index 876f5aa6fa..90532df071 100644
@@ -264,7 +264,7 @@ static void kvm_pit_realizefn(DeviceState *dev, Error **errp)
     };
     int ret;
 
-    if (kvm_check_extension(kvm_state, KVM_CAP_PIT2)) {
+    if (kvm_check_extension(KVM_CAP_PIT2)) {
         ret = kvm_vm_ioctl(kvm_state, KVM_CREATE_PIT2, &config);
     } else {
         ret = kvm_vm_ioctl(kvm_state, KVM_CREATE_PIT);

@@ -278,7 +278,7 @@ static void kvm_pit_realizefn(DeviceState *dev, Error **errp)
     case LOST_TICK_POLICY_DELAY:
         break; /* enabled by default */
     case LOST_TICK_POLICY_DISCARD:
-        if (kvm_check_extension(kvm_state, KVM_CAP_REINJECT_CONTROL)) {
+        if (kvm_check_extension(KVM_CAP_REINJECT_CONTROL)) {
             struct kvm_reinject_control control = { .pit_reinject = 0 };
 
             ret = kvm_vm_ioctl(kvm_state, KVM_REINJECT_CONTROL, &control);

hw/i386/kvm/ioapic.c +1 -1

index 4ba8e47251..718ed8ec6f 100644
@@ -25,7 +25,7 @@ void kvm_pc_setup_irq_routing(bool pci_enabled)
     KVMState *s = kvm_state;
     int i;
 
-    if (kvm_check_extension(s, KVM_CAP_IRQ_ROUTING)) {
+    if (kvm_check_extension(KVM_CAP_IRQ_ROUTING)) {
         for (i = 0; i < 8; ++i) {
             if (i == 2) {
                 continue;

hw/intc/arm_gic_kvm.c +1 -1

index d7df423a7a..b98437b265 100644
@@ -551,7 +551,7 @@ static void kvm_arm_gic_realize(DeviceState *dev, Error **errp)
                               KVM_DEV_ARM_VGIC_CTRL_INIT, NULL, true,
                               &error_abort);
         }
-    } else if (kvm_check_extension(kvm_state, KVM_CAP_DEVICE_CTRL)) {
+    } else if (kvm_check_extension(KVM_CAP_DEVICE_CTRL)) {
         error_setg_errno(errp, -ret, "error creating in-kernel VGIC");
         error_append_hint(errp,
                           "Perhaps the host CPU does not support GICv2?\n");

hw/intc/openpic_kvm.c +1 -1

index e4bf47d885..b02e914f5f 100644
@@ -203,7 +203,7 @@ static void kvm_openpic_realize(DeviceState *dev, Error **errp)
     struct kvm_create_device cd = {0};
     int ret, i;
 
-    if (!kvm_check_extension(s, KVM_CAP_DEVICE_CTRL)) {
+    if (!kvm_check_extension(KVM_CAP_DEVICE_CTRL)) {
         error_setg(errp, "Kernel is lacking Device Control API");
         return;
     }

hw/intc/xics_kvm.c +1 -1

index 8d6156578f..054648d16b 100644
@@ -365,7 +365,7 @@ int xics_kvm_connect(SpaprInterruptController *intc, uint32_t nr_servers,
         return 0;
     }
 
-    if (!kvm_enabled() || !kvm_check_extension(kvm_state, KVM_CAP_IRQ_XICS)) {
+    if (!kvm_enabled() || !kvm_check_extension(KVM_CAP_IRQ_XICS)) {
         error_setg(errp,
                    "KVM and IRQ_XICS capability must be present for in-kernel XICS");
         return -1;

hw/s390x/s390-stattrib-kvm.c +1 -1

index f89d8d9d16..77f536f15a 100644
@@ -22,7 +22,7 @@
 Object *kvm_s390_stattrib_create(void)
 {
     if (kvm_enabled() &&
-                kvm_check_extension(kvm_state, KVM_CAP_S390_CMMA_MIGRATION)) {
+                kvm_check_extension(KVM_CAP_S390_CMMA_MIGRATION)) {
         return object_new(TYPE_KVM_S390_STATTRIB);
     }
     return NULL;

target/arm/kvm.c +6 -7

index 7c672c78b8..dcdd01916b 100644
@@ -60,7 +60,7 @@ int kvm_arm_vcpu_finalize(CPUState *cs, int feature)
 
 void kvm_arm_init_serror_injection(CPUState *cs)
 {
-    cap_has_inject_serror_esr = kvm_check_extension(cs->kvm_state,
+    cap_has_inject_serror_esr = kvm_check_extension(
                                     KVM_CAP_ARM_INJECT_SERROR_ESR);
 }

@@ -210,15 +210,14 @@ void kvm_arm_add_vcpu_properties(Object *obj)
 
 bool kvm_arm_pmu_supported(void)
 {
-    return kvm_check_extension(kvm_state, KVM_CAP_ARM_PMU_V3);
+    return kvm_check_extension(KVM_CAP_ARM_PMU_V3);
 }
 
 int kvm_arm_get_max_vm_ipa_size(MachineState *ms)
 {
-    KVMState *s = KVM_STATE(ms->accelerator);
     int ret;
 
-    ret = kvm_check_extension(s, KVM_CAP_ARM_VM_IPA_SIZE);
+    ret = kvm_check_extension(KVM_CAP_ARM_VM_IPA_SIZE);
     return ret > 0 ? ret : 40;
 }

@@ -236,10 +235,10 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
      */
     kvm_halt_in_kernel_allowed = true;
 
-    cap_has_mp_state = kvm_check_extension(s, KVM_CAP_MP_STATE);
+    cap_has_mp_state = kvm_check_extension(KVM_CAP_MP_STATE);
 
     if (ms->smp.cpus > 256 &&
-        !kvm_check_extension(s, KVM_CAP_ARM_IRQ_LINE_LAYOUT_2)) {
+        !kvm_check_extension(KVM_CAP_ARM_IRQ_LINE_LAYOUT_2)) {
         error_report("Using more than 256 vcpus requires a host kernel "
                      "with KVM_CAP_ARM_IRQ_LINE_LAYOUT_2");
         ret = -EINVAL;

@@ -870,7 +869,7 @@ int kvm_arch_irqchip_create(KVMState *s)
     /* If we can create the VGIC using the newer device control API, we
      * let the device do this when it initializes itself, otherwise we
      * fall back to the old API */
-    return kvm_check_extension(s, KVM_CAP_DEVICE_CTRL);
+    return kvm_check_extension(KVM_CAP_DEVICE_CTRL);
 }
 
 int kvm_arm_vgic_probe(void)

target/arm/kvm32.c +1 -1

index 7b3a19e9ae..eff0176c8f 100644
@@ -221,7 +221,7 @@ int kvm_arch_init_vcpu(CPUState *cs)
     if (cpu->start_powered_off) {
         cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_POWER_OFF;
     }
-    if (kvm_check_extension(cs->kvm_state, KVM_CAP_ARM_PSCI_0_2)) {
+    if (kvm_check_extension(KVM_CAP_ARM_PSCI_0_2)) {
         cpu->psci_version = 2;
         cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_PSCI_0_2;
     }

target/arm/kvm64.c +7 -8

index 3dc494aaa7..49ef8ef15f 100644
@@ -84,14 +84,13 @@ GArray *hw_breakpoints, *hw_watchpoints;
  */
 static void kvm_arm_init_debug(CPUState *cs)
 {
-    have_guest_debug = kvm_check_extension(cs->kvm_state,
-                                           KVM_CAP_SET_GUEST_DEBUG);
+    have_guest_debug = kvm_check_extension(KVM_CAP_SET_GUEST_DEBUG);
 
-    max_hw_wps = kvm_check_extension(cs->kvm_state, KVM_CAP_GUEST_DEBUG_HW_WPS);
+    max_hw_wps = kvm_check_extension(KVM_CAP_GUEST_DEBUG_HW_WPS);
     hw_watchpoints = g_array_sized_new(true, true,
                                        sizeof(HWWatchpoint), max_hw_wps);
 
-    max_hw_bps = kvm_check_extension(cs->kvm_state, KVM_CAP_GUEST_DEBUG_HW_BPS);
+    max_hw_bps = kvm_check_extension(KVM_CAP_GUEST_DEBUG_HW_BPS);
     hw_breakpoints = g_array_sized_new(true, true,
                                        sizeof(HWBreakpoint), max_hw_bps);
     return;

@@ -654,12 +653,12 @@ bool kvm_arm_get_host_cpu_features(ARMHostCPUFeatures *ahcf)
 
 bool kvm_arm_aarch32_supported(void)
 {
-    return kvm_check_extension(kvm_state, KVM_CAP_ARM_EL1_32BIT);
+    return kvm_check_extension(KVM_CAP_ARM_EL1_32BIT);
 }
 
 bool kvm_arm_sve_supported(void)
 {
-    return kvm_check_extension(kvm_state, KVM_CAP_ARM_SVE);
+    return kvm_check_extension(KVM_CAP_ARM_SVE);
 }
 
 QEMU_BUILD_BUG_ON(KVM_ARM64_SVE_VQ_MIN != 1);

@@ -778,14 +777,14 @@ int kvm_arch_init_vcpu(CPUState *cs)
     if (cpu->start_powered_off) {
         cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_POWER_OFF;
     }
-    if (kvm_check_extension(cs->kvm_state, KVM_CAP_ARM_PSCI_0_2)) {
+    if (kvm_check_extension(KVM_CAP_ARM_PSCI_0_2)) {
         cpu->psci_version = 2;
         cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_PSCI_0_2;
     }
     if (!arm_feature(&cpu->env, ARM_FEATURE_AARCH64)) {
         cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_EL1_32BIT;
     }
-    if (!kvm_check_extension(cs->kvm_state, KVM_CAP_ARM_PMU_V3)) {
+    if (!kvm_check_extension(KVM_CAP_ARM_PMU_V3)) {
         cpu->has_pmu = false;
     }
     if (cpu->has_pmu) {

target/i386/kvm.c +28 ~~-33~~

200 lines not on this page · raw message

target/mips/kvm.c +2 -2

index 96cfa10cf2..0adfa70210 100644
@@ -50,8 +50,8 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
     /* MIPS has 128 signals */
     kvm_set_sigmask_len(s, 16);
 
-    kvm_mips_fpu_cap = kvm_check_extension(s, KVM_CAP_MIPS_FPU);
-    kvm_mips_msa_cap = kvm_check_extension(s, KVM_CAP_MIPS_MSA);
+    kvm_mips_fpu_cap = kvm_check_extension(KVM_CAP_MIPS_FPU);
+    kvm_mips_msa_cap = kvm_check_extension(KVM_CAP_MIPS_MSA);
 
     DPRINTF("%s\n", __func__);
     return 0;

target/ppc/kvm.c +17 ~~-17~~

index 2692f76130..ace44d9fc7 100644
@@ -110,24 +110,24 @@ static int kvmppc_get_dec_bits(void);
 
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
-    cap_interrupt_unset = kvm_check_extension(s, KVM_CAP_PPC_UNSET_IRQ);
-    cap_segstate = kvm_check_extension(s, KVM_CAP_PPC_SEGSTATE);
-    cap_booke_sregs = kvm_check_extension(s, KVM_CAP_PPC_BOOKE_SREGS);
+    cap_interrupt_unset = kvm_check_extension(KVM_CAP_PPC_UNSET_IRQ);
+    cap_segstate = kvm_check_extension(KVM_CAP_PPC_SEGSTATE);
+    cap_booke_sregs = kvm_check_extension(KVM_CAP_PPC_BOOKE_SREGS);
     cap_ppc_smt_possible = kvm_vm_check_extension(s, KVM_CAP_PPC_SMT_POSSIBLE);
-    cap_spapr_tce = kvm_check_extension(s, KVM_CAP_SPAPR_TCE);
-    cap_spapr_tce_64 = kvm_check_extension(s, KVM_CAP_SPAPR_TCE_64);
-    cap_spapr_multitce = kvm_check_extension(s, KVM_CAP_SPAPR_MULTITCE);
+    cap_spapr_tce = kvm_check_extension(KVM_CAP_SPAPR_TCE);
+    cap_spapr_tce_64 = kvm_check_extension(KVM_CAP_SPAPR_TCE_64);
+    cap_spapr_multitce = kvm_check_extension(KVM_CAP_SPAPR_MULTITCE);
     cap_spapr_vfio = kvm_vm_check_extension(s, KVM_CAP_SPAPR_TCE_VFIO);
-    cap_one_reg = kvm_check_extension(s, KVM_CAP_ONE_REG);
-    cap_hior = kvm_check_extension(s, KVM_CAP_PPC_HIOR);
-    cap_epr = kvm_check_extension(s, KVM_CAP_PPC_EPR);
-    cap_ppc_watchdog = kvm_check_extension(s, KVM_CAP_PPC_BOOKE_WATCHDOG);
+    cap_one_reg = kvm_check_extension(KVM_CAP_ONE_REG);
+    cap_hior = kvm_check_extension(KVM_CAP_PPC_HIOR);
+    cap_epr = kvm_check_extension(KVM_CAP_PPC_EPR);
+    cap_ppc_watchdog = kvm_check_extension(KVM_CAP_PPC_BOOKE_WATCHDOG);
     /*
      * Note: we don't set cap_papr here, because this capability is
      * only activated after this by kvmppc_set_papr()
      */
     cap_htab_fd = kvm_vm_check_extension(s, KVM_CAP_PPC_HTAB_FD);
-    cap_fixup_hcalls = kvm_check_extension(s, KVM_CAP_PPC_FIXUP_HCALL);
+    cap_fixup_hcalls = kvm_check_extension(KVM_CAP_PPC_FIXUP_HCALL);
     cap_ppc_smt = kvm_vm_check_extension(s, KVM_CAP_PPC_SMT);
     cap_htm = kvm_vm_check_extension(s, KVM_CAP_PPC_HTM);
     cap_mmu_radix = kvm_vm_check_extension(s, KVM_CAP_PPC_MMU_RADIX);

@@ -147,7 +147,7 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
      */
     cap_ppc_pvr_compat = false;
 
-    if (!kvm_check_extension(s, KVM_CAP_PPC_IRQ_LEVEL)) {
+    if (!kvm_check_extension(KVM_CAP_PPC_IRQ_LEVEL)) {
         error_report("KVM: Host kernel doesn't have level irq capability");
         exit(1);
     }

@@ -205,7 +205,7 @@ static int kvm_booke206_tlb_init(PowerPCCPU *cpu)
     int ret, i;
 
     if (!kvm_enabled() ||
-        !kvm_check_extension(cs->kvm_state, KVM_CAP_SW_TLB)) {
+        !kvm_check_extension(KVM_CAP_SW_TLB)) {
         return 0;
     }

@@ -246,7 +246,7 @@ static void kvm_get_smmu_info(struct kvm_ppc_smmu_info *info, Error **errp)
 
     assert(kvm_state != NULL);
 
-    if (!kvm_check_extension(kvm_state, KVM_CAP_PPC_GET_SMMU_INFO)) {
+    if (!kvm_check_extension(KVM_CAP_PPC_GET_SMMU_INFO)) {
         error_setg(errp, "KVM doesn't expose the MMU features it supports");
         error_append_hint(errp, "Consider switching to a newer KVM\n");
         return;

@@ -268,7 +268,7 @@ struct ppc_radix_page_info *kvm_get_radix_page_info(void)
     struct kvm_ppc_rmmu_info rmmu_info;
     int i;
 
-    if (!kvm_check_extension(s, KVM_CAP_PPC_MMU_RADIX)) {
+    if (!kvm_check_extension(KVM_CAP_PPC_MMU_RADIX)) {
         return NULL;
     }
     if (kvm_vm_ioctl(s, KVM_PPC_GET_RMMU_INFO, &rmmu_info)) {

@@ -2611,7 +2611,7 @@ int kvmppc_define_rtas_kernel_token(uint32_t token, const char *function)
         .token = token,
     };
 
-    if (!kvm_check_extension(kvm_state, KVM_CAP_PPC_RTAS)) {
+    if (!kvm_check_extension(KVM_CAP_PPC_RTAS)) {
         return -ENOENT;
     }

@@ -2828,7 +2828,7 @@ int kvm_handle_nmi(PowerPCCPU *cpu, struct kvm_run *run)
 
 int kvmppc_enable_hwrng(void)
 {
-    if (!kvm_enabled() || !kvm_check_extension(kvm_state, KVM_CAP_PPC_HWRNG)) {
+    if (!kvm_enabled() || !kvm_check_extension(KVM_CAP_PPC_HWRNG)) {
         return -1;
     }

target/s390x/cpu_models.c +1 -2

index 2fa609bffe..4b6185bb44 100644
@@ -222,8 +222,7 @@ bool s390_has_feat(S390Feat feat)
 #ifdef CONFIG_KVM
         if (kvm_enabled()) {
             if (feat == S390_FEAT_VECTOR) {
-                return kvm_check_extension(kvm_state,
-                                           KVM_CAP_S390_VECTOR_REGISTERS);
+                return kvm_check_extension(KVM_CAP_S390_VECTOR_REGISTERS);
             }
             if (feat == S390_FEAT_RUNTIME_INSTRUMENTATION) {
                 return kvm_s390_get_ri();

target/s390x/kvm.c +15 ~~-15~~

index f2f75d2a57..710f353fb0 100644
@@ -342,21 +342,21 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
     object_class_foreach(ccw_machine_class_foreach, TYPE_S390_CCW_MACHINE,
                          false, NULL);
 
-    if (!kvm_check_extension(kvm_state, KVM_CAP_DEVICE_CTRL)) {
+    if (!kvm_check_extension(KVM_CAP_DEVICE_CTRL)) {
         error_report("KVM is missing capability KVM_CAP_DEVICE_CTRL - "
                      "please use kernel 3.15 or newer");
         return -1;
     }
 
-    cap_sync_regs = kvm_check_extension(s, KVM_CAP_SYNC_REGS);
-    cap_async_pf = kvm_check_extension(s, KVM_CAP_ASYNC_PF);
-    cap_mem_op = kvm_check_extension(s, KVM_CAP_S390_MEM_OP);
-    cap_s390_irq = kvm_check_extension(s, KVM_CAP_S390_INJECT_IRQ);
-    cap_vcpu_resets = kvm_check_extension(s, KVM_CAP_S390_VCPU_RESETS);
-    cap_protected = kvm_check_extension(s, KVM_CAP_S390_PROTECTED);
+    cap_sync_regs = kvm_check_extension(KVM_CAP_SYNC_REGS);
+    cap_async_pf = kvm_check_extension(KVM_CAP_ASYNC_PF);
+    cap_mem_op = kvm_check_extension(KVM_CAP_S390_MEM_OP);
+    cap_s390_irq = kvm_check_extension(KVM_CAP_S390_INJECT_IRQ);
+    cap_vcpu_resets = kvm_check_extension(KVM_CAP_S390_VCPU_RESETS);
+    cap_protected = kvm_check_extension(KVM_CAP_S390_PROTECTED);
 
-    if (!kvm_check_extension(s, KVM_CAP_S390_GMAP)
-        || !kvm_check_extension(s, KVM_CAP_S390_COW)) {
+    if (!kvm_check_extension(KVM_CAP_S390_GMAP)
+        || !kvm_check_extension(KVM_CAP_S390_COW)) {
         phys_mem_set_alloc(legacy_s390_alloc);
     }

@@ -381,7 +381,7 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
      * newer machine types if KVM_CAP_S390_AIS_MIGRATION is available.
      */
     if (cpu_model_allowed() && kvm_kernel_irqchip_allowed() &&
-        kvm_check_extension(s, KVM_CAP_S390_AIS_MIGRATION)) {
+        kvm_check_extension(KVM_CAP_S390_AIS_MIGRATION)) {
         kvm_vm_enable_cap(s, KVM_CAP_S390_AIS, 0);
     }

@@ -1996,7 +1996,7 @@ void kvm_arch_init_irq_routing(KVMState *s)
      * are handled in-kernel, it is not true for s390 (yet); therefore, we
      * have to override the common code kvm_halt_in_kernel_allowed setting.
      */
-    if (kvm_check_extension(s, KVM_CAP_IRQ_ROUTING)) {
+    if (kvm_check_extension(KVM_CAP_IRQ_ROUTING)) {
         kvm_gsi_routing_allowed = true;
         kvm_halt_in_kernel_allowed = false;
     }

@@ -2015,7 +2015,7 @@ int kvm_s390_assign_subch_ioeventfd(EventNotifier *notifier, uint32_t sch,
     };
     trace_kvm_assign_subch_ioeventfd(kick.fd, kick.addr, assign,
                                      kick.datamatch);
-    if (!kvm_check_extension(kvm_state, KVM_CAP_IOEVENTFD)) {
+    if (!kvm_check_extension(KVM_CAP_IOEVENTFD)) {
         return -ENOSYS;
     }
     if (!assign) {

@@ -2082,7 +2082,7 @@ void kvm_s390_vcpu_interrupt_pre_save(S390CPU *cpu)
     CPUState *cs = CPU(cpu);
     int32_t bytes;
 
-    if (!kvm_check_extension(kvm_state, KVM_CAP_S390_IRQ_STATE)) {
+    if (!kvm_check_extension(KVM_CAP_S390_IRQ_STATE)) {
         return;
     }

@@ -2109,7 +2109,7 @@ int kvm_s390_vcpu_interrupt_post_load(S390CPU *cpu)
         return 0;
     }
 
-    if (!kvm_check_extension(kvm_state, KVM_CAP_S390_IRQ_STATE)) {
+    if (!kvm_check_extension(KVM_CAP_S390_IRQ_STATE)) {
         return -ENOSYS;
     }

@@ -2421,7 +2421,7 @@ void kvm_s390_get_host_cpu_model(S390CPUModel *model, Error **errp)
     }
 
     /* bpb needs kernel support for migration, VSIE and reset */
-    if (!kvm_check_extension(kvm_state, KVM_CAP_S390_BPB)) {
+    if (!kvm_check_extension(KVM_CAP_S390_BPB)) {
         clear_bit(S390_FEAT_BPB, model->features);
     }

Signature · 1 line

2.21.3

Cornelia Huck

Reviewed-by on v1 2/7

2020-06-23 15:44 UTC #4 lore raw

Re: [PATCH 2/7] accel/kvm: Simplify kvm_check_extension()

Quoted 12 lines · On Tue, 23 Jun 2020 12:50:47 +0200 Philippe Mathieu-Daudé <[email protected]>…

In previous commit we let kvm_check_extension() use the global kvm_state. Since the KVMState* argument is now unused, drop it.
Convert callers with this Coccinelle script:
@@
expression kvm_state, extension;
@@
- kvm_check_extension(kvm_state, extension)
+ kvm_check_extension(extension)

Quoted 5 lines

Unused variables manually removed:
- CPUState* in hyperv_enabled()
- KVMState* in kvm_arm_get_max_vm_ipa_size()
Inspired-by Paolo Bonzini <[email protected]>

🌟💡🌟

Quoted 13 lines

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 include/sysemu/kvm.h         |  2 +-
 accel/kvm/kvm-all.c          | 64 ++++++++++++++++++------------------
 hw/hyperv/hyperv.c           |  2 +-
 hw/i386/kvm/clock.c          |  2 +-
 hw/i386/kvm/i8254.c          |  4 +--
 hw/i386/kvm/ioapic.c         |  2 +-
 hw/intc/arm_gic_kvm.c        |  2 +-
 hw/intc/openpic_kvm.c        |  2 +-
 hw/intc/xics_kvm.c           |  2 +-
 hw/s390x/s390-stattrib-kvm.c |  2 +-
 target/arm/kvm.c             | 13 ++++----

Quoted 8 lines

 target/arm/kvm32.c           |  2 +-
 target/arm/kvm64.c           | 15 ++++-----
 target/i386/kvm.c            | 61 ++++++++++++++++------------------
 target/mips/kvm.c            |  4 +--
 target/ppc/kvm.c             | 34 +++++++++----------
 target/s390x/cpu_models.c    |  3 +-
 target/s390x/kvm.c           | 30 ++++++++---------
 18 files changed, 119 insertions(+), 127 deletions(-)

Reviewed-by Cornelia Huck <[email protected]>

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #5 lore raw

PATCH 3/7 accel/kvm: Simplify kvm_check_extension_list()

In reply to Philippe Mathieu-Daudé

The KVMState* argument is now unused, drop it.

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 accel/kvm/kvm-all.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

accel/kvm/kvm-all.c +3 -3

index b6b39b0e92..afd14492a0 100644
@@ -1088,7 +1088,7 @@ static int kvm_check_many_ioeventfds(void)
 }
 
 static const KVMCapabilityInfo *
-kvm_check_extension_list(KVMState *s, const KVMCapabilityInfo *list)
+kvm_check_extension_list(const KVMCapabilityInfo *list)
 {
     while (list->name) {
         if (!kvm_check_extension(list->value)) {

@@ -2104,10 +2104,10 @@ static int kvm_init(MachineState *ms)
         nc++;
     }
 
-    missing_cap = kvm_check_extension_list(s, kvm_required_capabilites);
+    missing_cap = kvm_check_extension_list(kvm_required_capabilites);
     if (!missing_cap) {
         missing_cap =
-            kvm_check_extension_list(s, kvm_arch_required_capabilities);
+            kvm_check_extension_list(kvm_arch_required_capabilities);
     }
     if (missing_cap) {
         ret = -EINVAL;

Signature · 1 line

2.21.3

Cornelia Huck

Reviewed-by on v1 3/7

2020-06-23 15:51 UTC #6 lore raw

Re: [PATCH 3/7] accel/kvm: Simplify kvm_check_extension_list()

Quoted 6 lines · On Tue, 23 Jun 2020 12:50:48 +0200 Philippe Mathieu-Daudé <[email protected]>…

The KVMState* argument is now unused, drop it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>
---
 accel/kvm/kvm-all.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

Reviewed-by Cornelia Huck <[email protected]>

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #7 lore raw

RFC PATCH 4/7 accel/kvm: Simplify kvm_set_sigmask_len()

In reply to Philippe Mathieu-Daudé

The sigmask_len is a property of the accelerator, not the VM. Simplify by directly using the global kvm_state, remove the unnecessary KVMState* argument.

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 include/sysemu/kvm.h | 2 +-
 accel/kvm/kvm-all.c  | 4 ++--
 target/mips/kvm.c    | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

include/sysemu/kvm.h +1 -1

index 3662641c99..44c1767a7f 100644
@@ -469,7 +469,7 @@ uint32_t kvm_arch_get_supported_cpuid(KVMState *env, uint32_t function,
 uint64_t kvm_arch_get_supported_msr_feature(KVMState *s, uint32_t index);
 
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len);
+void kvm_set_sigmask_len(unsigned int sigmask_len);
 
 #if !defined(CONFIG_USER_ONLY)
 int kvm_physical_memory_addr_from_host(KVMState *s, void *ram_addr,

accel/kvm/kvm-all.c +2 -2

index afd14492a0..7b3f76f23d 100644
@@ -2240,9 +2240,9 @@ err:
     return ret;
 }
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len)
+void kvm_set_sigmask_len(unsigned int sigmask_len)
 {
-    s->sigmask_len = sigmask_len;
+    kvm_state->sigmask_len = sigmask_len;
 }
 
 static void kvm_handle_io(uint16_t port, MemTxAttrs attrs, void *data, int direction,

target/mips/kvm.c +1 -1

index 0adfa70210..cc3e09bdef 100644
@@ -48,7 +48,7 @@ unsigned long kvm_arch_vcpu_id(CPUState *cs)
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     /* MIPS has 128 signals */
-    kvm_set_sigmask_len(s, 16);
+    kvm_set_sigmask_len(16);
 
     kvm_mips_fpu_cap = kvm_check_extension(KVM_CAP_MIPS_FPU);
     kvm_mips_msa_cap = kvm_check_extension(KVM_CAP_MIPS_MSA);

Signature · 1 line

2.21.3

Paolo Bonzini

maintainer

2020-07-05 07:12 UTC #8 lore raw

Re: [RFC PATCH 4/7] accel/kvm: Simplify kvm_set_sigmask_len()

Quoted 3 lines · Philippe Mathieu-Daudé

The sigmask_len is a property of the accelerator, not the VM. Simplify by directly using the global kvm_state, remove the unnecessary KVMState* argument.

This is not entirely true, if there were multiple KVMStates how would you know which one to read from? So it would have to be a global variable.

Paolo

Quoted 50 lines

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 include/sysemu/kvm.h | 2 +-
 accel/kvm/kvm-all.c  | 4 ++--
 target/mips/kvm.c    | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

include/sysemu/kvm.h +1 -1

index 3662641c99..44c1767a7f 100644
@@ -469,7 +469,7 @@ uint32_t kvm_arch_get_supported_cpuid(KVMState *env, uint32_t function,
 uint64_t kvm_arch_get_supported_msr_feature(KVMState *s, uint32_t index);
 
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len);
+void kvm_set_sigmask_len(unsigned int sigmask_len);
 
 #if !defined(CONFIG_USER_ONLY)
 int kvm_physical_memory_addr_from_host(KVMState *s, void *ram_addr,

accel/kvm/kvm-all.c +2 -2

index afd14492a0..7b3f76f23d 100644
@@ -2240,9 +2240,9 @@ err:
     return ret;
 }
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len)
+void kvm_set_sigmask_len(unsigned int sigmask_len)
 {
-    s->sigmask_len = sigmask_len;
+    kvm_state->sigmask_len = sigmask_len;
 }
 
 static void kvm_handle_io(uint16_t port, MemTxAttrs attrs, void *data, int direction,

target/mips/kvm.c +1 -1

index 0adfa70210..cc3e09bdef 100644
@@ -48,7 +48,7 @@ unsigned long kvm_arch_vcpu_id(CPUState *cs)
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     /* MIPS has 128 signals */
-    kvm_set_sigmask_len(s, 16);
+    kvm_set_sigmask_len(16);
 
     kvm_mips_fpu_cap = kvm_check_extension(KVM_CAP_MIPS_FPU);
     kvm_mips_msa_cap = kvm_check_extension(KVM_CAP_MIPS_MSA);

Philippe Mathieu-Daudé

author

2020-07-05 09:57 UTC #9 lore raw

Re: [RFC PATCH 4/7] accel/kvm: Simplify kvm_set_sigmask_len()

Quoted 7 lines · Paolo Bonzini

Quoted 3 lines · Philippe Mathieu-Daudé
The sigmask_len is a property of the accelerator, not the VM. Simplify by directly using the global kvm_state, remove the unnecessary KVMState* argument.
This is not entirely true, if there were multiple KVMStates how would you know which one to read from? So it would have to be a global variable.

Ah I guess I understand. Thanks for reviewing and queuing the rest!

Quoted 54 lines

Paolo

Quoted 50 lines

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 include/sysemu/kvm.h | 2 +-
 accel/kvm/kvm-all.c  | 4 ++--
 target/mips/kvm.c    | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

include/sysemu/kvm.h +1 -1

index 3662641c99..44c1767a7f 100644
@@ -469,7 +469,7 @@ uint32_t kvm_arch_get_supported_cpuid(KVMState *env, uint32_t function,
 uint64_t kvm_arch_get_supported_msr_feature(KVMState *s, uint32_t index);
 
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len);
+void kvm_set_sigmask_len(unsigned int sigmask_len);
 
 #if !defined(CONFIG_USER_ONLY)
 int kvm_physical_memory_addr_from_host(KVMState *s, void *ram_addr,

accel/kvm/kvm-all.c +2 -2

index afd14492a0..7b3f76f23d 100644
@@ -2240,9 +2240,9 @@ err:
     return ret;
 }
 
-void kvm_set_sigmask_len(KVMState *s, unsigned int sigmask_len)
+void kvm_set_sigmask_len(unsigned int sigmask_len)
 {
-    s->sigmask_len = sigmask_len;
+    kvm_state->sigmask_len = sigmask_len;
 }
 
 static void kvm_handle_io(uint16_t port, MemTxAttrs attrs, void *data, int direction,

target/mips/kvm.c +1 -1

index 0adfa70210..cc3e09bdef 100644
@@ -48,7 +48,7 @@ unsigned long kvm_arch_vcpu_id(CPUState *cs)
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     /* MIPS has 128 signals */
-    kvm_set_sigmask_len(s, 16);
+    kvm_set_sigmask_len(16);
 
     kvm_mips_fpu_cap = kvm_check_extension(KVM_CAP_MIPS_FPU);
     kvm_mips_msa_cap = kvm_check_extension(KVM_CAP_MIPS_MSA);

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #10 lore raw

PATCH 5/7 target/i386/kvm: Simplify get_para_features()

In reply to Philippe Mathieu-Daudé

The KVMState* argument is now unused, drop it.

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 target/i386/kvm.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

target/i386/kvm.c +2 -2

index 03df6ac3b4..19d3db657a 100644
@@ -289,7 +289,7 @@ static const struct kvm_para_features {
     { KVM_CAP_ASYNC_PF, KVM_FEATURE_ASYNC_PF },
 };
 
-static int get_para_features(KVMState *s)
+static int get_para_features(void)
 {
     int i, features = 0;

@@ -452,7 +452,7 @@ uint32_t kvm_arch_get_supported_cpuid(KVMState *s, uint32_t function,
 
     /* fallback for older kernels */
     if ((function == KVM_CPUID_FEATURES) && !found) {
-        ret = get_para_features(s);
+        ret = get_para_features();
     }
 
     return ret;

Signature · 1 line

2.21.3

Cornelia Huck

Reviewed-by on v1 5/7

2020-06-23 15:57 UTC #11 lore raw

Re: [PATCH 5/7] target/i386/kvm: Simplify get_para_features()

Quoted 6 lines · On Tue, 23 Jun 2020 12:50:50 +0200 Philippe Mathieu-Daudé <[email protected]>…

The KVMState* argument is now unused, drop it.
Signed-off-by Philippe Mathieu-Daudé <[email protected]>
---
 target/i386/kvm.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Reviewed-by Cornelia Huck <[email protected]>

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #12 lore raw

RFC PATCH 6/7 target/i386/kvm: Simplify kvm_get_mce_cap_supported()

In reply to Philippe Mathieu-Daudé

As the MCE supported capabilities should be the same for all VMs, it is safe to directly use the global kvm_state. Remove the unnecessary KVMState* argument.

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 target/i386/kvm.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

target/i386/kvm.c +3 -3

index 19d3db657a..626cb04d88 100644
@@ -536,7 +536,7 @@ uint64_t kvm_arch_get_supported_msr_feature(KVMState *s, uint32_t index)
     }
 }
 
-static int kvm_get_mce_cap_supported(KVMState *s, uint64_t *mce_cap,
+static int kvm_get_mce_cap_supported(uint64_t *mce_cap,
                                      int *max_banks)
 {
     int r;

@@ -544,7 +544,7 @@ static int kvm_get_mce_cap_supported(KVMState *s, uint64_t *mce_cap,
     r = kvm_check_extension(KVM_CAP_MCE);
     if (r > 0) {
         *max_banks = r;
-        return kvm_ioctl(s, KVM_X86_GET_MCE_CAP_SUPPORTED, mce_cap);
+        return kvm_ioctl(kvm_state, KVM_X86_GET_MCE_CAP_SUPPORTED, mce_cap);
     }
     return -ENOSYS;
 }

@@ -1707,7 +1707,7 @@ int kvm_arch_init_vcpu(CPUState *cs)
         int banks;
         int ret;
 
-        ret = kvm_get_mce_cap_supported(cs->kvm_state, &mcg_cap, &banks);
+        ret = kvm_get_mce_cap_supported(&mcg_cap, &banks);
         if (ret < 0) {
             fprintf(stderr, "kvm_get_mce_cap_supported: %s", strerror(-ret));
             return ret;

Signature · 1 line

2.21.3

Philippe Mathieu-Daudé

author

2020-06-23 10:51 UTC #13 lore raw

RFC PATCH 7/7 target/i386/kvm: Simplify kvm_get_supported_[feature]_msrs()

In reply to Philippe Mathieu-Daudé

As the MSR supported features should be the same for all VMs, it is safe to directly use the global kvm_state. Remove the unnecessary KVMState* argument.

Signed-off-by Philippe Mathieu-Daudé <[email protected]>

---
 target/i386/kvm.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)

target/i386/kvm.c +9 -8

index 626cb04d88..988ed3c238 100644
@@ -1887,7 +1887,7 @@ void kvm_arch_do_init_vcpu(X86CPU *cpu)
     }
 }
 
-static int kvm_get_supported_feature_msrs(KVMState *s)
+static int kvm_get_supported_feature_msrs(void)
 {
     int ret = 0;

@@ -1902,7 +1902,7 @@ static int kvm_get_supported_feature_msrs(KVMState *s)
     struct kvm_msr_list msr_list;
 
     msr_list.nmsrs = 0;
-    ret = kvm_ioctl(s, KVM_GET_MSR_FEATURE_INDEX_LIST, &msr_list);
+    ret = kvm_ioctl(kvm_state, KVM_GET_MSR_FEATURE_INDEX_LIST, &msr_list);
     if (ret < 0 && ret != -E2BIG) {
         error_report("Fetch KVM feature MSR list failed: %s",
             strerror(-ret));

@@ -1915,7 +1915,8 @@ static int kvm_get_supported_feature_msrs(KVMState *s)
                  msr_list.nmsrs * sizeof(msr_list.indices[0]));
 
     kvm_feature_msrs->nmsrs = msr_list.nmsrs;
-    ret = kvm_ioctl(s, KVM_GET_MSR_FEATURE_INDEX_LIST, kvm_feature_msrs);
+    ret = kvm_ioctl(kvm_state, KVM_GET_MSR_FEATURE_INDEX_LIST,
+                    kvm_feature_msrs);
 
     if (ret < 0) {
         error_report("Fetch KVM feature MSR list failed: %s",

@@ -1928,7 +1929,7 @@ static int kvm_get_supported_feature_msrs(KVMState *s)
     return 0;
 }
 
-static int kvm_get_supported_msrs(KVMState *s)
+static int kvm_get_supported_msrs(void)
 {
     int ret = 0;
     struct kvm_msr_list msr_list, *kvm_msr_list;

@@ -1938,7 +1939,7 @@ static int kvm_get_supported_msrs(KVMState *s)
      *  save/restore.
      */
     msr_list.nmsrs = 0;
-    ret = kvm_ioctl(s, KVM_GET_MSR_INDEX_LIST, &msr_list);
+    ret = kvm_ioctl(kvm_state, KVM_GET_MSR_INDEX_LIST, &msr_list);
     if (ret < 0 && ret != -E2BIG) {
         return ret;
     }

@@ -1951,7 +1952,7 @@ static int kvm_get_supported_msrs(KVMState *s)
                                           sizeof(msr_list.indices[0])));
 
     kvm_msr_list->nmsrs = msr_list.nmsrs;
-    ret = kvm_ioctl(s, KVM_GET_MSR_INDEX_LIST, kvm_msr_list);
+    ret = kvm_ioctl(kvm_state, KVM_GET_MSR_INDEX_LIST, kvm_msr_list);
     if (ret >= 0) {
         int i;

@@ -2107,12 +2108,12 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
         }
     }
 
-    ret = kvm_get_supported_msrs(s);
+    ret = kvm_get_supported_msrs();
     if (ret < 0) {
         return ret;
     }
 
-    kvm_get_supported_feature_msrs(s);
+    kvm_get_supported_feature_msrs();
 
     uname(&utsname);
     lm_capable_kernel = strcmp(utsname.machine, "x86_64") == 0;

Signature · 1 line

2.21.3

Philippe Mathieu-Daudé

author

2020-07-04 16:50 UTC #14 lore raw

In reply to Philippe Mathieu-Daudé

kind ping :)

Quoted 35 lines · Philippe Mathieu-Daudé

Following Paolo's idea on kvm_check_extension():
https://www.mail-archive.com/[email protected]/msg713794.html
CI:
https://travis-ci.org/github/philmd/qemu/builds/701213438
Philippe Mathieu-Daudé (7):
accel/kvm: Let kvm_check_extension use global KVM state accel/kvm: Simplify kvm_check_extension() accel/kvm: Simplify kvm_check_extension_list() accel/kvm: Simplify kvm_set_sigmask_len() target/i386/kvm: Simplify get_para_features() target/i386/kvm: Simplify kvm_get_mce_cap_supported() target/i386/kvm: Simplify kvm_get_supported_[feature]_msrs()
 include/sysemu/kvm.h         |  4 +-
 accel/kvm/kvm-all.c          | 76 +++++++++++++++----------------
 hw/hyperv/hyperv.c           |  2 +-
 hw/i386/kvm/clock.c          |  2 +-
 hw/i386/kvm/i8254.c          |  4 +-
 hw/i386/kvm/ioapic.c         |  2 +-
 hw/intc/arm_gic_kvm.c        |  2 +-
 hw/intc/openpic_kvm.c        |  2 +-
 hw/intc/xics_kvm.c           |  2 +-
 hw/s390x/s390-stattrib-kvm.c |  2 +-
 target/arm/kvm.c             | 13 +++---
 target/arm/kvm32.c           |  2 +-
 target/arm/kvm64.c           | 15 +++---
 target/i386/kvm.c            | 88 +++++++++++++++++-------------------
 target/mips/kvm.c            |  6 +--
 target/ppc/kvm.c             | 34 +++++++-------
 target/s390x/cpu_models.c    |  3 +-
 target/s390x/kvm.c           | 30 ++++++------
 18 files changed, 141 insertions(+), 148 deletions(-)

Paolo Bonzini

maintainer

2020-07-05 06:16 UTC #15 lore raw

Quoted 1 line · Philippe Mathieu-Daudé

kind ping :)

Queued all except 4.

Paolo

Quoted 37 lines · Philippe Mathieu-Daudé

Quoted 35 lines
Following Paolo's idea on kvm_check_extension():
https://www.mail-archive.com/[email protected]/msg713794.html
CI:
https://travis-ci.org/github/philmd/qemu/builds/701213438
Philippe Mathieu-Daudé (7):
accel/kvm: Let kvm_check_extension use global KVM state accel/kvm: Simplify kvm_check_extension() accel/kvm: Simplify kvm_check_extension_list() accel/kvm: Simplify kvm_set_sigmask_len() target/i386/kvm: Simplify get_para_features() target/i386/kvm: Simplify kvm_get_mce_cap_supported() target/i386/kvm: Simplify kvm_get_supported_[feature]_msrs()
 include/sysemu/kvm.h         |  4 +-
 accel/kvm/kvm-all.c          | 76 +++++++++++++++----------------
 hw/hyperv/hyperv.c           |  2 +-
 hw/i386/kvm/clock.c          |  2 +-
 hw/i386/kvm/i8254.c          |  4 +-
 hw/i386/kvm/ioapic.c         |  2 +-
 hw/intc/arm_gic_kvm.c        |  2 +-
 hw/intc/openpic_kvm.c        |  2 +-
 hw/intc/xics_kvm.c           |  2 +-
 hw/s390x/s390-stattrib-kvm.c |  2 +-
 target/arm/kvm.c             | 13 +++---
 target/arm/kvm32.c           |  2 +-
 target/arm/kvm64.c           | 15 +++---
 target/i386/kvm.c            | 88 +++++++++++++++++-------------------
 target/mips/kvm.c            |  6 +--
 target/ppc/kvm.c             | 34 +++++++-------
 target/s390x/cpu_models.c    |  3 +-
 target/s390x/kvm.c           | 30 ++++++------
 18 files changed, 141 insertions(+), 148 deletions(-)