PPC updates
This time around I also included a fix for the shmem compile bug, as my tree wouldn't compile otherwise.
The following changes since commit cb93bbdd7db92e50ff5e60a346b23df68acae46b:
Blue Swirl (1):
Fix OpenBSD linker warning
are available in the git repository at:
git://repo.or.cz/qemu/agraf.git ppc-next
Alexander Graf (4):
Fix compile of ivshmem on 32 bit hosts
KVM: PPC: Add level based interrupt logic
PPC: Qdev'ify e500 pci
PPC: Make e500 pci byte swap config data
hw/ivshmem.c | 6 +- hw/ppc.c | 11 +++++ hw/ppce500_pci.c | 106 ++++++++++++++++++++++++++++++++++--------------- target-ppc/kvm.c | 37 ++++++++++++++++- target-ppc/kvm_ppc.h | 13 ++++++ 5 files changed, 135 insertions(+), 38 deletions(-)
PATCH 1/4 Fix compile of ivshmem on 32 bit hosts
The compile of ivshmem.c on my ppc system breaks with the following error:
ivshmem.c: In function ‘check_shm_size’:
ivshmem.c:356: error: format ‘%ld’ expects type ‘long int’, but argument
4 has type ‘__off64_t’
This patch adds an explicit cast, making printf happy.
--- hw/ivshmem.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-)
hw/ivshmem.c +3 -3
@@ -351,9 +351,9 @@ static int check_shm_size(IVShmemState *s, int fd) {fstat (fd, &buf);if (s->ivshmem_size > buf.st_size) { -fprintf (stderr,"IVSHMEM ERROR: Requested memory size greater" ); -fprintf (stderr," than shared object size (%" PRIu64" > %ld)\n" , - s->ivshmem_size, buf.st_size); +fprintf (stderr,"IVSHMEM ERROR: Requested memory size greater than " +"shared object size (%" PRIu64" > %" PRIu64")\n" , + s->ivshmem_size, (uint64_t )buf.st_size);return -1 ; }else {return 0 ;
Signature · 1 line
1.6.0.2
PATCH 2/4 KVM: PPC: Add level based interrupt logic
In reply to Alexander GrafKVM on PowerPC used to have completely broken interrupt logic. Usually, interrupts work by having a PIC that pulls a line up/down, so the CPU knows that an interrupt is active. This line stays active until some action is done to the PIC to release the line.
On KVM for PPC, we just checked if there was an interrupt pending and pulled a line in the kernel module. We never released it though, hoping that kernel space would just declare an interrupt as released when injected - which is wrong.
To fix this, we need to completely redesign the interrupt injection logic. Whenever an interrupt line gets triggered, we need to notify kernel space that the line is up. Whenever it gets released, we do the same. This way we can assure that the interrupt state is always known to kernel space.
This fixes random stalls in KVM guests on PowerPC that were waiting for an interrupt while everyone else thought they received it already.
--- hw/ppc.c | 11 +++++++++++ target-ppc/kvm.c | 37 +++++++++++++++++++++++++++++++++++-- target-ppc/kvm_ppc.h | 13 +++++++++++++ 3 files changed, 59 insertions(+), 2 deletions(-)
hw/ppc.c +11 -0
@@ -28,6 +28,8 @@#include "nvram.h" #include "qemu-log.h" #include "loader.h" +#include "kvm.h" +#include "kvm_ppc.h" //#define PPC_DEBUG_IRQ //#define PPC_DEBUG_TB
@@ -50,6 +52,8 @@ static void cpu_ppc_tb_start (CPUState *env);static void ppc_set_irq (CPUState *env,int n_IRQ,int level) { +unsigned int old_pending = env->pending_interrupts; +if (level) { env->pending_interrupts |=1 << n_IRQ;cpu_interrupt (env,CPU_INTERRUPT_HARD );
@@ -58,6 +62,13 @@ static void ppc_set_irq (CPUState *env, int n_IRQ, int level)if (env->pending_interrupts ==0 )cpu_reset_interrupt (env,CPU_INTERRUPT_HARD ); } + +if (old_pending != env->pending_interrupts) { +#ifdef CONFIG_KVM +kvmppc_set_interrupt (env, n_IRQ, level); +#endif + } +LOG_IRQ ("%s: %p n_IRQ %d level %d => pending %08" PRIx32"req %08x\n" , __func__, env, n_IRQ, level, env->pending_interrupts, env->interrupt_request);
target-ppc/kvm.c +35 -2
@@ -37,6 +37,9 @@ do { } while (0)#endif +static int cap_interrupt_unset = false; +static int cap_interrupt_level = false; +/* XXX We have a race condition where we actually have a level triggered * interrupt, but the infrastructure can't expose that yet, so the guest * takes but ignores it, goes to sleep and never gets notified that there's
@@ -55,6 +58,18 @@ static void kvm_kick_env(void *env)int kvm_arch_init (KVMState *s,int smp_cpus) { +#ifdef KVM_CAP_PPC_UNSET_IRQ + cap_interrupt_unset =kvm_check_extension (s,KVM_CAP_PPC_UNSET_IRQ ); +#endif +#ifdef KVM_CAP_PPC_IRQ_LEVEL + cap_interrupt_level =kvm_check_extension (s,KVM_CAP_PPC_IRQ_LEVEL ); +#endif + +if (!cap_interrupt_level) { +fprintf (stderr,"KVM: Couldn't find level irq capability. Expect the " +"VM to stall at times!\n" ); + } +return 0 ; }
@@ -178,6 +193,23 @@ int kvm_arch_get_registers(CPUState *env)return 0 ; } +int kvmppc_set_interrupt (CPUState *env,int irq,int level) +{ +unsigned virq = level ?KVM_INTERRUPT_SET_LEVEL :KVM_INTERRUPT_UNSET ; + +if (irq !=PPC_INTERRUPT_EXT ) { +return 0 ; + } + +if (!kvm_enabled () || !cap_interrupt_unset || !cap_interrupt_level) { +return 0 ; + } + +kvm_vcpu_ioctl (env,KVM_INTERRUPT , &virq); + +return 0 ; +} +#if defined(TARGET_PPCEMB )#define PPC_INPUT_INT PPC40x_INPUT_INT#elif defined(TARGET_PPC64 )
@@ -193,7 +225,8 @@ int kvm_arch_pre_run(CPUState *env, struct kvm_run *run)/* PowerPC Qemu tracks the various core input pins (interrupt, critical * interrupt, reset, etc) in PPC-specific env->irq_input_state. */ -if (run->ready_for_interrupt_injection && +if (!cap_interrupt_level && + run->ready_for_interrupt_injection && (env->interrupt_request &CPU_INTERRUPT_HARD ) && (env->irq_input_state & (1 <<PPC_INPUT_INT ))) {
@@ -201,7 +234,7 @@ int kvm_arch_pre_run(CPUState *env, struct kvm_run *run)* future KVM could cache it in-kernel to avoid a heavyweight exit * when reading the UIC. */ - irq =-1U ; + irq =KVM_INTERRUPT_SET ;dprintf ("injected interrupt %d\n" , irq); r =kvm_vcpu_ioctl (env,KVM_INTERRUPT , &irq);
target-ppc/kvm_ppc.h +13 -0
@@ -16,5 +16,18 @@ int kvmppc_read_host_property(const char *node_path, const char *prop,uint32_t kvmppc_get_tbfreq (void );int kvmppc_get_hypercall (CPUState *env,uint8_t *buf,int buf_len); +int kvmppc_set_interrupt (CPUState *env,int irq,int level); + +#ifndef KVM_INTERRUPT_SET +#define KVM_INTERRUPT_SET -1 +#endif + +#ifndef KVM_INTERRUPT_UNSET +#define KVM_INTERRUPT_UNSET -2 +#endif + +#ifndef KVM_INTERRUPT_SET_LEVEL +#define KVM_INTERRUPT_SET_LEVEL -3 +#endif #endif /* __KVM_PPC_H__ */
Signature · 1 line
1.6.0.2
PATCH 3/4 PPC: Qdev'ify e500 pci
In reply to Alexander GrafThe e500 PCI controller isn't qdev'ified yet. This leads to severe issues when running with -drive.
To be able to use a virtio disk with an e500 VM, let's convert the PCI controller over to qdev.
--- hw/ppce500_pci.c | 106 +++++++++++++++++++++++++++++++++++++----------------- 1 files changed, 73 insertions(+), 33 deletions(-)
hw/ppce500_pci.c +73 -33
@@ -73,11 +73,11 @@ struct pci_inbound { };struct PPCE500PCIState { +PCIHostState pci_state;struct pci_outbound pob[PPCE500_PCI_NR_POBS ];struct pci_inbound pib[PPCE500_PCI_NR_PIBS ];uint32_t gasket_time; -PCIHostState pci_state; -PCIDevice *pci_dev; +uint64_t base_addr; };typedef struct PPCE500PCIState PPCE500PCIState ;
@@ -221,7 +221,7 @@ static void ppce500_pci_save(QEMUFile *f, void *opaque)PPCE500PCIState *controller = opaque;int i; -pci_device_save (controller->pci_dev, f); +/* pci_device_save(controller->pci_dev, f); */ for (i =0 ; i <PPCE500_PCI_NR_POBS ; i++) {qemu_put_be32s (f, &controller->pob[i].potar);
@@ -247,7 +247,7 @@ static int ppce500_pci_load(QEMUFile *f, void *opaque, int version_id)if (version_id !=1 )return -EINVAL ; -pci_device_load (controller->pci_dev, f); +/* pci_device_load(controller->pci_dev, f); */ for (i =0 ; i <PPCE500_PCI_NR_POBS ; i++) {qemu_get_be32s (f, &controller->pob[i].potar);
@@ -269,55 +269,95 @@ static int ppce500_pci_load(QEMUFile *f, void *opaque, int version_id)PCIBus *ppce500_pci_init (qemu_irq pci_irqs[4 ],target_phys_addr_t registers) { -PPCE500PCIState *controller; +DeviceState *dev; +PCIBus *b; +PCIHostState *h; +PPCE500PCIState *s;PCIDevice *d; -int index;static int ppce500_pci_id; - controller =qemu_mallocz (sizeof (PPCE500PCIState)); + dev =qdev_create (NULL ,"e500-pcihost" ); + h =FROM_SYSBUS (PCIHostState,sysbus_from_qdev (dev)); + s =DO_UPCAST (PPCE500PCIState, pci_state, h); + +qdev_prop_set_uint64 (dev,"base_addr" , registers); + b =pci_register_bus (&s->pci_state.busdev.qdev,NULL , mpc85xx_pci_set_irq, + mpc85xx_pci_map_irq, pci_irqs,PCI_DEVFN (0x11 ,0 ),4 ); + + s->pci_state.bus = b; +qdev_init_nofail (dev); + d =pci_create_simple (b,0 ,"e500-host-bridge" ); + +/* XXX load/save code not tested. */ +register_savevm (&d->qdev,"ppce500_pci" , ppce500_pci_id++, +1 , ppce500_pci_save, ppce500_pci_load, s); - controller->pci_state.bus =pci_register_bus (NULL ,"pci" , - mpc85xx_pci_set_irq, - mpc85xx_pci_map_irq, - pci_irqs,PCI_DEVFN (0x11 ,0 ), -4 ); - d =pci_register_device (controller->pci_state.bus, -"host bridge" ,sizeof (PCIDevice), -0 ,NULL ,NULL ); +return b; +} -pci_config_set_vendor_id (d->config,PCI_VENDOR_ID_FREESCALE ); -pci_config_set_device_id (d->config,PCI_DEVICE_ID_MPC8533E ); -pci_config_set_class (d->config,PCI_CLASS_PROCESSOR_POWERPC ); +static int e500_pcihost_initfn (SysBusDevice *dev) +{ +PCIHostState *h; +PPCE500PCIState *s; +target_phys_addr_t registers; +int index; - controller->pci_dev = d; + h =FROM_SYSBUS (PCIHostState,sysbus_from_qdev (dev)); + s =DO_UPCAST (PPCE500PCIState, pci_state, h); + registers = (target_phys_addr_t )s->base_addr;/* CFGADDR */ - index =pci_host_conf_register_mmio (&controller->pci_state,0 ); + index =pci_host_conf_register_mmio (&s->pci_state,0 );if (index <0 ) - goto free; +return -1 ;cpu_register_physical_memory (registers +PCIE500_CFGADDR ,4 , index);/* CFGDATA */ - index =pci_host_data_register_mmio (&controller->pci_state,0 ); + index =pci_host_data_register_mmio (&s->pci_state,0 );if (index <0 ) - goto free; +return -1 ;cpu_register_physical_memory (registers +PCIE500_CFGDATA ,4 , index); index =cpu_register_io_memory (e500_pci_reg_read, - e500_pci_reg_write, controller); + e500_pci_reg_write, s);if (index <0 ) - goto free; +return -1 ;cpu_register_physical_memory (registers +PCIE500_REG_BASE ,PCIE500_REG_SIZE , index); +return 0 ; +} -/* XXX load/save code not tested. */ -register_savevm (&d->qdev,"ppce500_pci" , ppce500_pci_id++, -1 , ppce500_pci_save, ppce500_pci_load, controller); +static int e500_host_bridge_initfn (PCIDevice *dev) +{ +pci_config_set_vendor_id (dev->config,PCI_VENDOR_ID_FREESCALE ); +pci_config_set_device_id (dev->config,PCI_DEVICE_ID_MPC8533E ); +pci_config_set_class (dev->config,PCI_CLASS_PROCESSOR_POWERPC ); + +return 0 ; +} + +static PCIDeviceInfo e500_host_bridge_info = { + .qdev.name ="e500-host-bridge" , + .qdev.desc ="Host bridge" , + .qdev.size =sizeof (PCIDevice), + .qdev.no_user =1 , + .init = e500_host_bridge_initfn, +}; -return controller->pci_state.bus; +static SysBusDeviceInfo e500_pcihost_info = { + .init = e500_pcihost_initfn, + .qdev.name ="e500-pcihost" , + .qdev.size =sizeof (PPCE500PCIState), + .qdev.no_user =1 , + .qdev.props = (Property []) { +DEFINE_PROP_UINT64 ("base_addr" , PPCE500PCIState, base_addr,0 ), +DEFINE_PROP_END_OF_LIST (), + } +}; -free: -printf ("%s error\n" , __func__); -qemu_free (controller); -return NULL ; +static void e500_pci_register (void ) +{ +sysbus_register_withprop (&e500_pcihost_info); +pci_qdev_register (&e500_host_bridge_info); } +device_init (e500_pci_register);
Signature · 1 line
1.6.0.2
PATCH 4/4 PPC: Make e500 pci byte swap config data
In reply to Alexander GrafThe config data field on the e500 pci controller is in little endian, so we need to enable byte swap there.
--- hw/ppce500_pci.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
hw/ppce500_pci.c +1 -1
@@ -313,7 +313,7 @@ static int e500_pcihost_initfn(SysBusDevice *dev)cpu_register_physical_memory (registers +PCIE500_CFGADDR ,4 , index);/* CFGDATA */ - index =pci_host_data_register_mmio (&s->pci_state,0 ); + index =pci_host_data_register_mmio (&s->pci_state,1 );if (index <0 )return -1 ;cpu_register_physical_memory (registers +PCIE500_CFGDATA ,4 , index);
Signature · 1 line
1.6.0.2