Hello, Several large extensions were merged in VFIO recently: migration support with dirty tracking, support for different host IOMMU backend devices, multifd support, etc. This adds up to the previous extensions: vfio-platform, AP, CCW. The result is that VFIO is now a subsystem of over +16,000 lines of code : QEMU 2.0 : 3988 total QEMU 2.12 : 9336 total QEMU 6.2 : 12715 total QEMU 8.2 : 13690 total QEMU 9.2 : 15680 total QEMU 10.0 : 16607 total Organization is weak, naming inconsistent, the vfio-common.h header file and common.c are quite messy. It's time to address the technical debt before adding new features. This proposal reorganizes some of the VFIO files to isolate features, introduces new files and renames services to better reflect the namespace they belong to. This is code reshuffling and there are no intentional functional changes. If more could be done, please propose ! Timing seems right. I hope we can merge this when the QEMU 10.1 cycle starts and then address the two large series waiting : live update and vfio-user. Thanks, C. Cédric Le Goater (32): vfio: Move vfio_mig_active() into migration.c vfio: Rename vfio_reset_bytes_transferred() vfio: Introduce a new header file for external migration services vfio: Make vfio_un/block_multiple_devices_migration() static vfio: Make vfio_viommu_preset() static vfio: Introduce a new header file for internal migration services vfio: Introduce a new header file for VFIOdisplay declarations vfio: Move VFIOHostDMAWindow definition into spapr.c vfio: Introduce a new header file for VFIOIOMMUFD declarations vfio: Introduce new files for VFIORegion definitions and declarations vfio: Introduce a new header file for VFIOcontainer declarations vfio: Make vfio_group_list static vfio: Move VFIOAddressSpace helpers into container-base.c vfio: Move Host IOMMU type declarations into their respective files vfio: Introduce a new header file for helper services vfio: Move vfio_get_info_dma_avail() into helpers.c vfio: Move vfio_kvm_device_add/del_fd() to helpers.c vfio: Move vfio_get_device_info() to helpers.c vfio: Introduce a new file for VFIODevice definitions vfio: Introduce new files for CPR definitions and declarations vfio: Move vfio_kvm_device_fd() into helpers.c vfio: Move vfio_device_list into device.c vfio: Move vfio_de/attach_device() into device.c vfio: Introduce new files for dirty tracking definitions and declarations vfio: Move vfio_set_migration_error() into migration.c vfio: Rename vfio-common.h to vfio-device.h vfio: Rename VFIODevice related services vfio: Make vfio_devices_query_dirty_bitmap() static vfio: Rename VFIOContainer related services vfio: Rename VFIO dirty tracking services vfio: Introduce vfio_dirty_tracking_un/register() routines vfio: Rename RAM discard related services hw/vfio/cpr.h | 15 + hw/vfio/dirty-tracking.h | 19 + hw/vfio/display.h | 42 ++ hw/vfio/helpers.h | 35 ++ hw/vfio/iommufd.h | 29 + hw/vfio/migration-multifd.h | 2 +- hw/vfio/migration.h | 73 +++ hw/vfio/pci.h | 3 +- include/hw/s390x/vfio-ccw.h | 2 +- include/hw/vfio/vfio-common.h | 346 ------------ include/hw/vfio/vfio-container-base.h | 5 + include/hw/vfio/vfio-container.h | 35 ++ include/hw/vfio/vfio-device.h | 148 ++++++ include/hw/vfio/vfio-migration.h | 17 + include/hw/vfio/vfio-platform.h | 4 +- include/hw/vfio/vfio-region.h | 47 ++ backends/iommufd.c | 2 +- hw/core/sysbus-fdt.c | 1 + hw/ppc/spapr_pci_vfio.c | 6 +- hw/s390x/s390-pci-vfio.c | 3 +- hw/vfio/ap.c | 10 +- hw/vfio/ccw.c | 20 +- hw/vfio/container-base.c | 69 +++ hw/vfio/container.c | 123 ++--- hw/vfio/cpr.c | 3 +- hw/vfio/device.c | 371 +++++++++++++ hw/vfio/{common.c => dirty-tracking.c} | 382 ++------------ hw/vfio/display.c | 4 +- hw/vfio/helpers.c | 702 +++---------------------- hw/vfio/igd.c | 8 +- hw/vfio/iommufd.c | 25 +- hw/vfio/migration-multifd.c | 4 +- hw/vfio/migration.c | 112 +++- hw/vfio/pci-quirks.c | 1 + hw/vfio/pci.c | 51 +- hw/vfio/platform.c | 11 +- hw/vfio/region.c | 396 ++++++++++++++ hw/vfio/spapr.c | 10 +- migration/target.c | 4 +- hw/vfio/meson.build | 8 +- hw/vfio/trace-events | 36 +- 41 files changed, 1711 insertions(+), 1473 deletions(-) create mode 100644 hw/vfio/cpr.h create mode 100644 hw/vfio/dirty-tracking.h create mode 100644 hw/vfio/display.h create mode 100644 hw/vfio/helpers.h create mode 100644 hw/vfio/iommufd.h create mode 100644 hw/vfio/migration.h delete mode 100644 include/hw/vfio/vfio-common.h create mode 100644 include/hw/vfio/vfio-container.h create mode 100644 include/hw/vfio/vfio-device.h create mode 100644 include/hw/vfio/vfio-migration.h create mode 100644 include/hw/vfio/vfio-region.h create mode 100644 hw/vfio/device.c rename hw/vfio/{common.c => dirty-tracking.c} (81%) create mode 100644 hw/vfio/region.c -- 2.48.1 vfio_mig_active() is part of the VFIO migration API. Move the definitions where VFIO migration is implemented. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 16 ---------------- hw/vfio/migration.c | 16 ++++++++++++++++ 2 files changed, 16 insertions(+), 16 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; * Device state interfaces */ -bool vfio_mig_active(void) -{ - VFIODevice *vbasedev; - - if (QLIST_EMPTY(&vfio_device_list)) { - return false; - } - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->migration_blocker) { - return false; - } - } - return true; -} - static Error *multiple_devices_migration_blocker; /* diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ void vfio_mig_add_bytes_transferred(unsigned long val) qatomic_add(&bytes_transferred, val); } +bool vfio_mig_active(void) +{ + VFIODevice *vbasedev; + + if (QLIST_EMPTY(&vfio_device_list)) { + return false; + } + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->migration_blocker) { + return false; + } + } + return true; +} + /* * Return true when either migration initialized or blocker registered. * Currently only return false when adding blocker fails which will -- 2.48.1 Enforce a 'vfio_mig_' prefix for the VFIO migration API to better reflect the namespace these routines belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 +- hw/vfio/migration.c | 2 +- migration/target.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); int64_t vfio_mig_bytes_transferred(void); -void vfio_reset_bytes_transferred(void); +void vfio_mig_reset_bytes_transferred(void); void vfio_mig_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ int64_t vfio_mig_bytes_transferred(void) return MIN(qatomic_read(&bytes_transferred), INT64_MAX); } -void vfio_reset_bytes_transferred(void) +void vfio_mig_reset_bytes_transferred(void) { qatomic_set(&bytes_transferred, 0); } diff --git a/migration/target.c b/migration/target.c index XXXXXXX..XXXXXXX 100644 --- a/migration/target.c +++ b/migration/target.c @@ -XXX,XX +XXX,XX @@ void migration_populate_vfio_info(MigrationInfo *info) void migration_reset_vfio_bytes_transferred(void) { - vfio_reset_bytes_transferred(); + vfio_mig_reset_bytes_transferred(); } #else void migration_populate_vfio_info(MigrationInfo *info) -- 2.48.1 The migration core subsytem makes uses of the VFIO migration API to collect statistics on the number of bytes transferred. These services are declared in "hw/vfio/vfio-common.h" which also contains VFIO internal declarations. Move the migration declarations into a new header file "hw/vfio/vfio-migration.h" to reduce the exposure of VFIO internals. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 4 ---- include/hw/vfio/vfio-migration.h | 17 +++++++++++++++++ hw/vfio/migration-multifd.c | 1 + hw/vfio/migration.c | 1 + migration/target.c | 2 +- 5 files changed, 20 insertions(+), 5 deletions(-) create mode 100644 include/hw/vfio/vfio-migration.h diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_mig_active(void); int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); -int64_t vfio_mig_bytes_transferred(void); -void vfio_mig_reset_bytes_transferred(void); -void vfio_mig_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/include/hw/vfio/vfio-migration.h b/include/hw/vfio/vfio-migration.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-migration.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO migration interface + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_MIGRATION_H +#define HW_VFIO_VFIO_MIGRATION_H + +bool vfio_mig_active(void); +int64_t vfio_mig_bytes_transferred(void); +void vfio_mig_reset_bytes_transferred(void); +void vfio_mig_add_bytes_transferred(unsigned long val); + +#endif /* HW_VFIO_VFIO_MIGRATION_H */ diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "qapi/error.h" #include "qemu/bswap.h" diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include "system/runstate.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "migration/savevm.h" #include "migration/vmstate.h" diff --git a/migration/target.c b/migration/target.c index XXXXXXX..XXXXXXX 100644 --- a/migration/target.c +++ b/migration/target.c @@ -XXX,XX +XXX,XX @@ #include CONFIG_DEVICES #ifdef CONFIG_VFIO -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #endif #ifdef CONFIG_VFIO -- 2.48.1 Both of these routines are only used in file "migration.c". Move them there. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/common.c | 62 ----------------------------------- hw/vfio/migration.c | 62 +++++++++++++++++++++++++++++++++++ 3 files changed, 62 insertions(+), 64 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); -void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "qapi/error.h" #include "migration/misc.h" -#include "migration/blocker.h" #include "migration/qemu-file.h" #include "system/tcg.h" #include "system/tpm.h" @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; * Device state interfaces */ -static Error *multiple_devices_migration_blocker; - -/* - * Multiple devices migration is allowed only if all devices support P2P - * migration. Single device migration is allowed regardless of P2P migration - * support. - */ -static bool vfio_multiple_devices_migration_is_supported(void) -{ - VFIODevice *vbasedev; - unsigned int device_num = 0; - bool all_support_p2p = true; - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->migration) { - device_num++; - - if (!(vbasedev->migration->mig_flags & VFIO_MIGRATION_P2P)) { - all_support_p2p = false; - } - } - } - - return all_support_p2p || device_num <= 1; -} - -int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp) -{ - int ret; - - if (vfio_multiple_devices_migration_is_supported()) { - return 0; - } - - if (vbasedev->enable_migration == ON_OFF_AUTO_ON) { - error_setg(errp, "Multiple VFIO devices migration is supported only if " - "all of them support P2P migration"); - return -EINVAL; - } - - if (multiple_devices_migration_blocker) { - return 0; - } - - error_setg(&multiple_devices_migration_blocker, - "Multiple VFIO devices migration is supported only if all of " - "them support P2P migration"); - ret = migrate_add_blocker_normal(&multiple_devices_migration_blocker, errp); - - return ret; -} - -void vfio_unblock_multiple_devices_migration(void) -{ - if (!multiple_devices_migration_blocker || - !vfio_multiple_devices_migration_is_supported()) { - return; - } - - migrate_del_blocker(&multiple_devices_migration_blocker); -} bool vfio_viommu_preset(VFIODevice *vbasedev) { diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ static int vfio_migration_init(VFIODevice *vbasedev) return 0; } +static Error *multiple_devices_migration_blocker; + +/* + * Multiple devices migration is allowed only if all devices support P2P + * migration. Single device migration is allowed regardless of P2P migration + * support. + */ +static bool vfio_multiple_devices_migration_is_supported(void) +{ + VFIODevice *vbasedev; + unsigned int device_num = 0; + bool all_support_p2p = true; + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->migration) { + device_num++; + + if (!(vbasedev->migration->mig_flags & VFIO_MIGRATION_P2P)) { + all_support_p2p = false; + } + } + } + + return all_support_p2p || device_num <= 1; +} + +static int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp) +{ + int ret; + + if (vfio_multiple_devices_migration_is_supported()) { + return 0; + } + + if (vbasedev->enable_migration == ON_OFF_AUTO_ON) { + error_setg(errp, "Multiple VFIO devices migration is supported only if " + "all of them support P2P migration"); + return -EINVAL; + } + + if (multiple_devices_migration_blocker) { + return 0; + } + + error_setg(&multiple_devices_migration_blocker, + "Multiple VFIO devices migration is supported only if all of " + "them support P2P migration"); + ret = migrate_add_blocker_normal(&multiple_devices_migration_blocker, errp); + + return ret; +} + +static void vfio_unblock_multiple_devices_migration(void) +{ + if (!multiple_devices_migration_blocker || + !vfio_multiple_devices_migration_is_supported()) { + return; + } + + migrate_del_blocker(&multiple_devices_migration_blocker); +} + static void vfio_migration_deinit(VFIODevice *vbasedev) { VFIOMigration *migration = vbasedev->migration; -- 2.48.1 This routine is only used in file "migration.c". Move it there. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 6 ------ hw/vfio/migration.c | 5 +++++ 3 files changed, 5 insertions(+), 7 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_viommu_preset(VFIODevice *vbasedev); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; * Device state interfaces */ - -bool vfio_viommu_preset(VFIODevice *vbasedev) -{ - return vbasedev->bcontainer->space->as != &address_space_memory; -} - static void vfio_set_migration_error(int ret) { if (migration_is_running()) { diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ bool vfio_mig_active(void) return true; } +static bool vfio_viommu_preset(VFIODevice *vbasedev) +{ + return vbasedev->bcontainer->space->as != &address_space_memory; +} + /* * Return true when either migration initialized or blocker registered. * Currently only return false when adding blocker fails which will -- 2.48.1 Gather all VFIO migration related declarations into "migration.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/migration.h | 72 +++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 54 +------------------------- hw/vfio/common.c | 17 +-------- hw/vfio/iommufd.c | 1 + hw/vfio/migration-multifd.c | 1 + hw/vfio/migration.c | 17 +++++++++ hw/vfio/pci.c | 1 + 7 files changed, 94 insertions(+), 69 deletions(-) create mode 100644 hw/vfio/migration.h diff --git a/hw/vfio/migration.h b/hw/vfio/migration.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/migration.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO migration + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_MIGRATION_H +#define HW_VFIO_MIGRATION_H + +#ifdef CONFIG_LINUX +#include <linux/vfio.h> +#endif + +#include "qemu/notify.h" + +/* + * Flags to be used as unique delimiters for VFIO devices in the migration + * stream. These flags are composed as: + * 0xffffffff => MSB 32-bit all 1s + * 0xef10 => Magic ID, represents emulated (virtual) function IO + * 0x0000 => 16-bits reserved for flags + * + * The beginning of state information is marked by _DEV_CONFIG_STATE, + * _DEV_SETUP_STATE, or _DEV_DATA_STATE, respectively. The end of a + * certain state information is marked by _END_OF_STATE. + */ +#define VFIO_MIG_FLAG_END_OF_STATE (0xffffffffef100001ULL) +#define VFIO_MIG_FLAG_DEV_CONFIG_STATE (0xffffffffef100002ULL) +#define VFIO_MIG_FLAG_DEV_SETUP_STATE (0xffffffffef100003ULL) +#define VFIO_MIG_FLAG_DEV_DATA_STATE (0xffffffffef100004ULL) +#define VFIO_MIG_FLAG_DEV_INIT_DATA_SENT (0xffffffffef100005ULL) + +typedef struct VFIODevice VFIODevice; +typedef struct VFIOMultifd VFIOMultifd; + +typedef struct VFIOMigration { + struct VFIODevice *vbasedev; + VMChangeStateEntry *vm_state; + NotifierWithReturn migration_state; + uint32_t device_state; + int data_fd; + void *data_buffer; + size_t data_buffer_size; + uint64_t mig_flags; + uint64_t precopy_init_size; + uint64_t precopy_dirty_size; + bool multifd_transfer; + VFIOMultifd *multifd; + bool initial_data_sent; + + bool event_save_iterate_started; + bool event_precopy_empty_hit; +} VFIOMigration; + + +bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); +void vfio_migration_exit(VFIODevice *vbasedev); +bool vfio_device_state_is_running(VFIODevice *vbasedev); +bool vfio_device_state_is_precopy(VFIODevice *vbasedev); +int vfio_save_device_config_state(QEMUFile *f, void *opaque, Error **errp); +int vfio_load_device_config_state(QEMUFile *f, void *opaque); + +#ifdef CONFIG_LINUX +int vfio_migration_set_state(VFIODevice *vbasedev, + enum vfio_device_mig_state new_state, + enum vfio_device_mig_state recover_state, + Error **errp); +#endif + +#endif /* HW_VFIO_MIGRATION_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "qemu/queue.h" -#include "qemu/notify.h" #include "ui/console.h" #include "hw/display/ramfb.h" #ifdef CONFIG_LINUX @@ -XXX,XX +XXX,XX @@ #define VFIO_MSG_PREFIX "vfio %s: " -/* - * Flags to be used as unique delimiters for VFIO devices in the migration - * stream. These flags are composed as: - * 0xffffffff => MSB 32-bit all 1s - * 0xef10 => Magic ID, represents emulated (virtual) function IO - * 0x0000 => 16-bits reserved for flags - * - * The beginning of state information is marked by _DEV_CONFIG_STATE, - * _DEV_SETUP_STATE, or _DEV_DATA_STATE, respectively. The end of a - * certain state information is marked by _END_OF_STATE. - */ -#define VFIO_MIG_FLAG_END_OF_STATE (0xffffffffef100001ULL) -#define VFIO_MIG_FLAG_DEV_CONFIG_STATE (0xffffffffef100002ULL) -#define VFIO_MIG_FLAG_DEV_SETUP_STATE (0xffffffffef100003ULL) -#define VFIO_MIG_FLAG_DEV_DATA_STATE (0xffffffffef100004ULL) -#define VFIO_MIG_FLAG_DEV_INIT_DATA_SENT (0xffffffffef100005ULL) - enum { VFIO_DEVICE_TYPE_PCI = 0, VFIO_DEVICE_TYPE_PLATFORM = 1, @@ -XXX,XX +XXX,XX @@ typedef struct VFIORegion { uint8_t nr; /* cache the region number for debug */ } VFIORegion; -typedef struct VFIOMultifd VFIOMultifd; - -typedef struct VFIOMigration { - struct VFIODevice *vbasedev; - VMChangeStateEntry *vm_state; - NotifierWithReturn migration_state; - uint32_t device_state; - int data_fd; - void *data_buffer; - size_t data_buffer_size; - uint64_t mig_flags; - uint64_t precopy_init_size; - uint64_t precopy_dirty_size; - bool multifd_transfer; - VFIOMultifd *multifd; - bool initial_data_sent; - - bool event_save_iterate_started; - bool event_precopy_empty_hit; -} VFIOMigration; - struct VFIOGroup; typedef struct VFIOContainer { @@ -XXX,XX +XXX,XX @@ typedef struct VFIOIOMMUFDContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); typedef struct VFIODeviceOps VFIODeviceOps; +typedef struct VFIOMigration VFIOMigration; typedef struct VFIODevice { QLIST_ENTRY(VFIODevice) next; @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_device_state_is_running(VFIODevice *vbasedev); -bool vfio_device_state_is_precopy(VFIODevice *vbasedev); - -int vfio_save_device_config_state(QEMUFile *f, void *opaque, Error **errp); -int vfio_load_device_config_state(QEMUFile *f, void *opaque); - #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info); @@ -XXX,XX +XXX,XX @@ struct vfio_info_cap_header * vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); - -int vfio_migration_set_state(VFIODevice *vbasedev, - enum vfio_device_mig_state new_state, - enum vfio_device_mig_state recover_state, - Error **errp); #endif -bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); -void vfio_migration_exit(VFIODevice *vbasedev); - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); bool vfio_devices_all_dirty_tracking_started( const VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "migration/qemu-file.h" #include "system/tcg.h" #include "system/tpm.h" +#include "migration.h" VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); @@ -XXX,XX +XXX,XX @@ static void vfio_set_migration_error(int ret) } } -bool vfio_device_state_is_running(VFIODevice *vbasedev) -{ - VFIOMigration *migration = vbasedev->migration; - - return migration->device_state == VFIO_DEVICE_STATE_RUNNING || - migration->device_state == VFIO_DEVICE_STATE_RUNNING_P2P; -} - -bool vfio_device_state_is_precopy(VFIODevice *vbasedev) -{ - VFIOMigration *migration = vbasedev->migration; - - return migration->device_state == VFIO_DEVICE_STATE_PRE_COPY || - migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; -} - static bool vfio_devices_all_device_dirty_tracking_started( const VFIOContainerBase *bcontainer) { diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/cutils.h" #include "qemu/chardev_open.h" #include "pci.h" +#include "migration.h" static int iommufd_cdev_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ #include "io/channel-buffer.h" #include "migration/qemu-file.h" #include "migration-multifd.h" +#include "migration.h" #include "trace.h" #define VFIO_DEVICE_STATE_CONFIG_STATE (1) diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "trace.h" #include "hw/hw.h" +#include "migration.h" /* * This is an arbitrary size based on migration of mlx5 devices, where typically @@ -XXX,XX +XXX,XX @@ void vfio_migration_exit(VFIODevice *vbasedev) migrate_del_blocker(&vbasedev->migration_blocker); } + +bool vfio_device_state_is_running(VFIODevice *vbasedev) +{ + VFIOMigration *migration = vbasedev->migration; + + return migration->device_state == VFIO_DEVICE_STATE_RUNNING || + migration->device_state == VFIO_DEVICE_STATE_RUNNING_P2P; +} + +bool vfio_device_state_is_precopy(VFIODevice *vbasedev) +{ + VFIOMigration *migration = vbasedev->migration; + + return migration->device_state == VFIO_DEVICE_STATE_PRE_COPY || + migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; +} diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "migration/blocker.h" #include "migration/qemu-file.h" #include "system/iommufd.h" +#include "migration.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" -- 2.48.1 Gather all VFIOdisplay related declarations into "display.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/display.h | 41 +++++++++++++++++++++++++++++++++++ hw/vfio/pci.h | 1 + include/hw/vfio/vfio-common.h | 28 ------------------------ hw/vfio/display.c | 2 +- 4 files changed, 43 insertions(+), 29 deletions(-) create mode 100644 hw/vfio/display.h diff --git a/hw/vfio/display.h b/hw/vfio/display.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/display.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO display + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_DISPLAY_H +#define HW_VFIO_DISPLAY_H + +#include "ui/console.h" +#include "hw/display/ramfb.h" + +typedef struct VFIODMABuf { + QemuDmaBuf *buf; + uint32_t pos_x, pos_y, pos_updates; + uint32_t hot_x, hot_y, hot_updates; + int dmabuf_id; + QTAILQ_ENTRY(VFIODMABuf) next; +} VFIODMABuf; + +typedef struct VFIODisplay { + QemuConsole *con; + RAMFBState *ramfb; + struct vfio_region_info *edid_info; + struct vfio_region_gfx_edid *edid_regs; + uint8_t *edid_blob; + QEMUTimer *edid_link_timer; + struct { + VFIORegion buffer; + DisplaySurface *surface; + } region; + struct { + QTAILQ_HEAD(, VFIODMABuf) bufs; + VFIODMABuf *primary; + VFIODMABuf *cursor; + } dmabuf; +} VFIODisplay; + +#endif /* HW_VFIO_DISPLAY_H */ diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -XXX,XX +XXX,XX @@ #include "qemu/timer.h" #include "qom/object.h" #include "system/kvm.h" +#include "display.h" #define PCI_ANY_ID (~0) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "qemu/queue.h" -#include "ui/console.h" -#include "hw/display/ramfb.h" #ifdef CONFIG_LINUX #include <linux/vfio.h> #endif @@ -XXX,XX +XXX,XX @@ typedef struct VFIOGroup { #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" -typedef struct VFIODMABuf { - QemuDmaBuf *buf; - uint32_t pos_x, pos_y, pos_updates; - uint32_t hot_x, hot_y, hot_updates; - int dmabuf_id; - QTAILQ_ENTRY(VFIODMABuf) next; -} VFIODMABuf; - -typedef struct VFIODisplay { - QemuConsole *con; - RAMFBState *ramfb; - struct vfio_region_info *edid_info; - struct vfio_region_gfx_edid *edid_regs; - uint8_t *edid_blob; - QEMUTimer *edid_link_timer; - struct { - VFIORegion buffer; - DisplaySurface *surface; - } region; - struct { - QTAILQ_HEAD(, VFIODMABuf) bufs; - VFIODMABuf *primary; - VFIODMABuf *cursor; - } dmabuf; -} VFIODisplay; - VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); void vfio_address_space_insert(VFIOAddressSpace *space, diff --git a/hw/vfio/display.c b/hw/vfio/display.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/display.c +++ b/hw/vfio/display.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "hw/display/edid.h" -#include "ui/console.h" #include "qapi/error.h" #include "pci.h" +#include "display.h" #include "trace.h" #ifndef DRM_PLANE_TYPE_PRIMARY -- 2.48.1 VFIOHostDMAWindow is only used in file "spapr.c". Move it there. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 7 ------- hw/vfio/spapr.c | 7 +++++++ 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); -typedef struct VFIOHostDMAWindow { - hwaddr min_iova; - hwaddr max_iova; - uint64_t iova_pgsizes; - QLIST_ENTRY(VFIOHostDMAWindow) hostwin_next; -} VFIOHostDMAWindow; - typedef struct IOMMUFDBackend IOMMUFDBackend; typedef struct VFIOIOASHwpt { diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "trace.h" +typedef struct VFIOHostDMAWindow { + hwaddr min_iova; + hwaddr max_iova; + uint64_t iova_pgsizes; + QLIST_ENTRY(VFIOHostDMAWindow) hostwin_next; +} VFIOHostDMAWindow; + typedef struct VFIOSpaprContainer { VFIOContainer container; MemoryListener prereg_listener; -- 2.48.1 Gather all VFIOIOMMUFD related declarations into "iommufd.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/iommufd.h | 29 +++++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 21 +++------------------ hw/vfio/iommufd.c | 1 + 3 files changed, 33 insertions(+), 18 deletions(-) create mode 100644 hw/vfio/iommufd.h diff --git a/hw/vfio/iommufd.h b/hw/vfio/iommufd.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/iommufd.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO iommufd + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_IOMMUFD_H +#define HW_VFIO_IOMMUFD_H + + +typedef struct VFIOIOASHwpt { + uint32_t hwpt_id; + uint32_t hwpt_flags; + QLIST_HEAD(, VFIODevice) device_list; + QLIST_ENTRY(VFIOIOASHwpt) next; +} VFIOIOASHwpt; + +typedef struct VFIOIOMMUFDContainer { + VFIOContainerBase bcontainer; + IOMMUFDBackend *be; + uint32_t ioas_id; + QLIST_HEAD(, VFIOIOASHwpt) hwpt_list; +} VFIOIOMMUFDContainer; + +OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); + +#endif /* HW_VFIO_IOMMUFD_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); -typedef struct IOMMUFDBackend IOMMUFDBackend; - -typedef struct VFIOIOASHwpt { - uint32_t hwpt_id; - uint32_t hwpt_flags; - QLIST_HEAD(, VFIODevice) device_list; - QLIST_ENTRY(VFIOIOASHwpt) next; -} VFIOIOASHwpt; - -typedef struct VFIOIOMMUFDContainer { - VFIOContainerBase bcontainer; - IOMMUFDBackend *be; - uint32_t ioas_id; - QLIST_HEAD(, VFIOIOASHwpt) hwpt_list; -} VFIOIOMMUFDContainer; - -OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); - typedef struct VFIODeviceOps VFIODeviceOps; typedef struct VFIOMigration VFIOMigration; +typedef struct IOMMUFDBackend IOMMUFDBackend; +typedef struct VFIOIOASHwpt VFIOIOASHwpt; + typedef struct VFIODevice { QLIST_ENTRY(VFIODevice) next; QLIST_ENTRY(VFIODevice) container_next; diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/chardev_open.h" #include "pci.h" #include "migration.h" +#include "iommufd.h" static int iommufd_cdev_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) -- 2.48.1 Gather all VFIORegion related declarations and definitions into their own files to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". These declarations are made available externally (for "sysbus-fdt.c"). This is for 'vfio-platform' devices which have been deprecated and will be removed in QEMU 10.2. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/display.h | 1 + include/hw/vfio/vfio-common.h | 32 +-- include/hw/vfio/vfio-platform.h | 2 + include/hw/vfio/vfio-region.h | 47 ++++ hw/core/sysbus-fdt.c | 1 + hw/vfio/helpers.c | 363 ----------------------------- hw/vfio/pci-quirks.c | 1 + hw/vfio/pci.c | 1 + hw/vfio/platform.c | 1 + hw/vfio/region.c | 395 ++++++++++++++++++++++++++++++++ hw/vfio/meson.build | 1 + hw/vfio/trace-events | 16 +- 12 files changed, 460 insertions(+), 401 deletions(-) create mode 100644 include/hw/vfio/vfio-region.h create mode 100644 hw/vfio/region.c diff --git a/hw/vfio/display.h b/hw/vfio/display.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/display.h +++ b/hw/vfio/display.h @@ -XXX,XX +XXX,XX @@ #include "ui/console.h" #include "hw/display/ramfb.h" +#include "hw/vfio/vfio-region.h" typedef struct VFIODMABuf { QemuDmaBuf *buf; diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ enum { VFIO_DEVICE_TYPE_CCW = 2, VFIO_DEVICE_TYPE_AP = 3, }; - -typedef struct VFIOMmap { - MemoryRegion mem; - void *mmap; - off_t offset; - size_t size; -} VFIOMmap; - -typedef struct VFIORegion { - struct VFIODevice *vbasedev; - off_t fd_offset; /* offset of region within device fd */ - MemoryRegion *mem; /* slow, read/write access */ - size_t size; - uint32_t flags; /* VFIO region flags (rd/wr/mmap) */ - uint32_t nr_mmaps; - VFIOMmap *mmaps; - uint8_t nr; /* cache the region number for debug */ -} VFIORegion; - struct VFIOGroup; typedef struct VFIOContainer { @@ -XXX,XX +XXX,XX @@ void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); -void vfio_region_write(void *opaque, hwaddr addr, - uint64_t data, unsigned size); -uint64_t vfio_region_read(void *opaque, - hwaddr addr, unsigned size); -int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, - int index, const char *name); -int vfio_region_mmap(VFIORegion *region); -void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled); -void vfio_region_unmap(VFIORegion *region); -void vfio_region_exit(VFIORegion *region); -void vfio_region_finalize(VFIORegion *region); + void vfio_reset_handler(void *opaque); struct vfio_device_info *vfio_get_device_info(int fd); bool vfio_device_is_mdev(VFIODevice *vbasedev); @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp); bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); -extern const MemoryRegionOps vfio_region_ops; typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIOGroupList vfio_group_list; diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-platform.h +++ b/include/hw/vfio/vfio-platform.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOINTp { /* function type for user side eventfd handler */ typedef void (*eventfd_user_side_handler_t)(VFIOINTp *intp); +typedef struct VFIORegion VFIORegion; + struct VFIOPlatformDevice { SysBusDevice sbdev; VFIODevice vbasedev; /* not a QOM object */ diff --git a/include/hw/vfio/vfio-region.h b/include/hw/vfio/vfio-region.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-region.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO region + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_REGION_H +#define HW_VFIO_REGION_H + +#include "exec/memory.h" + +typedef struct VFIODevice VFIODevice; + +typedef struct VFIOMmap { + MemoryRegion mem; + void *mmap; + off_t offset; + size_t size; +} VFIOMmap; + +typedef struct VFIORegion { + struct VFIODevice *vbasedev; + off_t fd_offset; /* offset of region within device fd */ + MemoryRegion *mem; /* slow, read/write access */ + size_t size; + uint32_t flags; /* VFIO region flags (rd/wr/mmap) */ + uint32_t nr_mmaps; + VFIOMmap *mmaps; + uint8_t nr; /* cache the region number for debug */ +} VFIORegion; + + +void vfio_region_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size); +uint64_t vfio_region_read(void *opaque, + hwaddr addr, unsigned size); +int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, + int index, const char *name); +int vfio_region_mmap(VFIORegion *region); +void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled); +void vfio_region_unmap(VFIORegion *region); +void vfio_region_exit(VFIORegion *region); +void vfio_region_finalize(VFIORegion *region); + +#endif /* HW_VFIO_REGION_H */ diff --git a/hw/core/sysbus-fdt.c b/hw/core/sysbus-fdt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/sysbus-fdt.c +++ b/hw/core/sysbus-fdt.c @@ -XXX,XX +XXX,XX @@ #include "hw/vfio/vfio-platform.h" #include "hw/vfio/vfio-calxeda-xgmac.h" #include "hw/vfio/vfio-amd-xgbe.h" +#include "hw/vfio/vfio-region.h" #include "hw/display/ramfb.h" #include "hw/uefi/var-service-api.h" #include "hw/arm/fdt.h" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, return false; } -/* - * IO Port/MMIO - Beware of the endians, VFIO is always little endian - */ -void vfio_region_write(void *opaque, hwaddr addr, - uint64_t data, unsigned size) -{ - VFIORegion *region = opaque; - VFIODevice *vbasedev = region->vbasedev; - union { - uint8_t byte; - uint16_t word; - uint32_t dword; - uint64_t qword; - } buf; - - switch (size) { - case 1: - buf.byte = data; - break; - case 2: - buf.word = cpu_to_le16(data); - break; - case 4: - buf.dword = cpu_to_le32(data); - break; - case 8: - buf.qword = cpu_to_le64(data); - break; - default: - hw_error("vfio: unsupported write size, %u bytes", size); - break; - } - - if (pwrite(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { - error_report("%s(%s:region%d+0x%"HWADDR_PRIx", 0x%"PRIx64 - ",%d) failed: %m", - __func__, vbasedev->name, region->nr, - addr, data, size); - } - - trace_vfio_region_write(vbasedev->name, region->nr, addr, data, size); - - /* - * A read or write to a BAR always signals an INTx EOI. This will - * do nothing if not pending (including not in INTx mode). We assume - * that a BAR access is in response to an interrupt and that BAR - * accesses will service the interrupt. Unfortunately, we don't know - * which access will service the interrupt, so we're potentially - * getting quite a few host interrupts per guest interrupt. - */ - vbasedev->ops->vfio_eoi(vbasedev); -} - -uint64_t vfio_region_read(void *opaque, - hwaddr addr, unsigned size) -{ - VFIORegion *region = opaque; - VFIODevice *vbasedev = region->vbasedev; - union { - uint8_t byte; - uint16_t word; - uint32_t dword; - uint64_t qword; - } buf; - uint64_t data = 0; - - if (pread(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { - error_report("%s(%s:region%d+0x%"HWADDR_PRIx", %d) failed: %m", - __func__, vbasedev->name, region->nr, - addr, size); - return (uint64_t)-1; - } - switch (size) { - case 1: - data = buf.byte; - break; - case 2: - data = le16_to_cpu(buf.word); - break; - case 4: - data = le32_to_cpu(buf.dword); - break; - case 8: - data = le64_to_cpu(buf.qword); - break; - default: - hw_error("vfio: unsupported read size, %u bytes", size); - break; - } - - trace_vfio_region_read(vbasedev->name, region->nr, addr, size, data); - - /* Same as write above */ - vbasedev->ops->vfio_eoi(vbasedev); - - return data; -} - -const MemoryRegionOps vfio_region_ops = { - .read = vfio_region_read, - .write = vfio_region_write, - .endianness = DEVICE_LITTLE_ENDIAN, - .valid = { - .min_access_size = 1, - .max_access_size = 8, - }, - .impl = { - .min_access_size = 1, - .max_access_size = 8, - }, -}; - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size) { vbmap->pages = REAL_HOST_PAGE_ALIGN(size) / qemu_real_host_page_size(); @@ -XXX,XX +XXX,XX @@ vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id) return vfio_get_cap((void *)info, info->cap_offset, id); } -static int vfio_setup_region_sparse_mmaps(VFIORegion *region, - struct vfio_region_info *info) -{ - struct vfio_info_cap_header *hdr; - struct vfio_region_info_cap_sparse_mmap *sparse; - int i, j; - - hdr = vfio_get_region_info_cap(info, VFIO_REGION_INFO_CAP_SPARSE_MMAP); - if (!hdr) { - return -ENODEV; - } - - sparse = container_of(hdr, struct vfio_region_info_cap_sparse_mmap, header); - - trace_vfio_region_sparse_mmap_header(region->vbasedev->name, - region->nr, sparse->nr_areas); - - region->mmaps = g_new0(VFIOMmap, sparse->nr_areas); - - for (i = 0, j = 0; i < sparse->nr_areas; i++) { - if (sparse->areas[i].size) { - trace_vfio_region_sparse_mmap_entry(i, sparse->areas[i].offset, - sparse->areas[i].offset + - sparse->areas[i].size - 1); - region->mmaps[j].offset = sparse->areas[i].offset; - region->mmaps[j].size = sparse->areas[i].size; - j++; - } - } - - region->nr_mmaps = j; - region->mmaps = g_realloc(region->mmaps, j * sizeof(VFIOMmap)); - - return 0; -} - -int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, - int index, const char *name) -{ - g_autofree struct vfio_region_info *info = NULL; - int ret; - - ret = vfio_get_region_info(vbasedev, index, &info); - if (ret) { - return ret; - } - - region->vbasedev = vbasedev; - region->flags = info->flags; - region->size = info->size; - region->fd_offset = info->offset; - region->nr = index; - - if (region->size) { - region->mem = g_new0(MemoryRegion, 1); - memory_region_init_io(region->mem, obj, &vfio_region_ops, - region, name, region->size); - - if (!vbasedev->no_mmap && - region->flags & VFIO_REGION_INFO_FLAG_MMAP) { - - ret = vfio_setup_region_sparse_mmaps(region, info); - - if (ret) { - region->nr_mmaps = 1; - region->mmaps = g_new0(VFIOMmap, region->nr_mmaps); - region->mmaps[0].offset = 0; - region->mmaps[0].size = region->size; - } - } - } - - trace_vfio_region_setup(vbasedev->name, index, name, - region->flags, region->fd_offset, region->size); - return 0; -} - -static void vfio_subregion_unmap(VFIORegion *region, int index) -{ - trace_vfio_region_unmap(memory_region_name(&region->mmaps[index].mem), - region->mmaps[index].offset, - region->mmaps[index].offset + - region->mmaps[index].size - 1); - memory_region_del_subregion(region->mem, &region->mmaps[index].mem); - munmap(region->mmaps[index].mmap, region->mmaps[index].size); - object_unparent(OBJECT(&region->mmaps[index].mem)); - region->mmaps[index].mmap = NULL; -} - -int vfio_region_mmap(VFIORegion *region) -{ - int i, ret, prot = 0; - char *name; - - if (!region->mem) { - return 0; - } - - prot |= region->flags & VFIO_REGION_INFO_FLAG_READ ? PROT_READ : 0; - prot |= region->flags & VFIO_REGION_INFO_FLAG_WRITE ? PROT_WRITE : 0; - - for (i = 0; i < region->nr_mmaps; i++) { - size_t align = MIN(1ULL << ctz64(region->mmaps[i].size), 1 * GiB); - void *map_base, *map_align; - - /* - * Align the mmap for more efficient mapping in the kernel. Ideally - * we'd know the PMD and PUD mapping sizes to use as discrete alignment - * intervals, but we don't. As of Linux v6.12, the largest PUD size - * supporting huge pfnmap is 1GiB (ARCH_SUPPORTS_PUD_PFNMAP is only set - * on x86_64). Align by power-of-two size, capped at 1GiB. - * - * NB. qemu_memalign() and friends actually allocate memory, whereas - * the region size here can exceed host memory, therefore we manually - * create an oversized anonymous mapping and clean it up for alignment. - */ - map_base = mmap(0, region->mmaps[i].size + align, PROT_NONE, - MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); - if (map_base == MAP_FAILED) { - ret = -errno; - goto no_mmap; - } - - map_align = (void *)ROUND_UP((uintptr_t)map_base, (uintptr_t)align); - munmap(map_base, map_align - map_base); - munmap(map_align + region->mmaps[i].size, - align - (map_align - map_base)); - - region->mmaps[i].mmap = mmap(map_align, region->mmaps[i].size, prot, - MAP_SHARED | MAP_FIXED, - region->vbasedev->fd, - region->fd_offset + - region->mmaps[i].offset); - if (region->mmaps[i].mmap == MAP_FAILED) { - ret = -errno; - goto no_mmap; - } - - name = g_strdup_printf("%s mmaps[%d]", - memory_region_name(region->mem), i); - memory_region_init_ram_device_ptr(&region->mmaps[i].mem, - memory_region_owner(region->mem), - name, region->mmaps[i].size, - region->mmaps[i].mmap); - g_free(name); - memory_region_add_subregion(region->mem, region->mmaps[i].offset, - &region->mmaps[i].mem); - - trace_vfio_region_mmap(memory_region_name(&region->mmaps[i].mem), - region->mmaps[i].offset, - region->mmaps[i].offset + - region->mmaps[i].size - 1); - } - - return 0; - -no_mmap: - trace_vfio_region_mmap_fault(memory_region_name(region->mem), i, - region->fd_offset + region->mmaps[i].offset, - region->fd_offset + region->mmaps[i].offset + - region->mmaps[i].size - 1, ret); - - region->mmaps[i].mmap = NULL; - - for (i--; i >= 0; i--) { - vfio_subregion_unmap(region, i); - } - - return ret; -} - -void vfio_region_unmap(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - vfio_subregion_unmap(region, i); - } - } -} - -void vfio_region_exit(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - memory_region_del_subregion(region->mem, &region->mmaps[i].mem); - } - } - - trace_vfio_region_exit(region->vbasedev->name, region->nr); -} - -void vfio_region_finalize(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - munmap(region->mmaps[i].mmap, region->mmaps[i].size); - object_unparent(OBJECT(&region->mmaps[i].mem)); - } - } - - object_unparent(OBJECT(region->mem)); - - g_free(region->mem); - g_free(region->mmaps); - - trace_vfio_region_finalize(region->vbasedev->name, region->nr); - - region->mem = NULL; - region->mmaps = NULL; - region->nr_mmaps = 0; - region->size = 0; - region->flags = 0; - region->nr = 0; -} - -void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - memory_region_set_enabled(&region->mmaps[i].mem, enabled); - } - } - - trace_vfio_region_mmaps_set_enabled(memory_region_name(region->mem), - enabled); -} - int vfio_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info) { diff --git a/hw/vfio/pci-quirks.c b/hw/vfio/pci-quirks.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci-quirks.c +++ b/hw/vfio/pci-quirks.c @@ -XXX,XX +XXX,XX @@ #include "hw/qdev-properties.h" #include "pci.h" #include "pci-quirks.h" +#include "hw/vfio/vfio-region.h" #include "trace.h" /* diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "migration/qemu-file.h" #include "system/iommufd.h" #include "migration.h" +#include "hw/vfio/vfio-region.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -XXX,XX +XXX,XX @@ #include "hw/platform-bus.h" #include "hw/qdev-properties.h" #include "system/kvm.h" +#include "hw/vfio/vfio-region.h" /* * Functions used whatever the injection method diff --git a/hw/vfio/region.c b/hw/vfio/region.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ +/* + * low level and IOMMU backend agnostic helpers used by VFIO devices, + * related to regions, interrupts, capabilities + * + * Copyright Red Hat, Inc. 2012 + * + * Authors: + * Alex Williamson <alex.williamson@redhat.com> + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + * + * Based on qemu-kvm device-assignment: + * Adapted for KVM by Qumranet. + * Copyright (c) 2007, Neocleus, Alex Novik (alex@neocleus.com) + * Copyright (c) 2007, Neocleus, Guy Zana (guy@neocleus.com) + * Copyright (C) 2008, Qumranet, Amit Shah (amit.shah@qumranet.com) + * Copyright (C) 2008, Red Hat, Amit Shah (amit.shah@redhat.com) + * Copyright (C) 2008, IBM, Muli Ben-Yehuda (muli@il.ibm.com) + */ + +#include "qemu/osdep.h" +#include <sys/ioctl.h> + +#include "hw/vfio/vfio-common.h" +#include "hw/vfio/pci.h" +#include "hw/hw.h" +#include "trace.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/units.h" +#include "monitor/monitor.h" + +/* + * IO Port/MMIO - Beware of the endians, VFIO is always little endian + */ +void vfio_region_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size) +{ + VFIORegion *region = opaque; + VFIODevice *vbasedev = region->vbasedev; + union { + uint8_t byte; + uint16_t word; + uint32_t dword; + uint64_t qword; + } buf; + + switch (size) { + case 1: + buf.byte = data; + break; + case 2: + buf.word = cpu_to_le16(data); + break; + case 4: + buf.dword = cpu_to_le32(data); + break; + case 8: + buf.qword = cpu_to_le64(data); + break; + default: + hw_error("vfio: unsupported write size, %u bytes", size); + break; + } + + if (pwrite(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { + error_report("%s(%s:region%d+0x%"HWADDR_PRIx", 0x%"PRIx64 + ",%d) failed: %m", + __func__, vbasedev->name, region->nr, + addr, data, size); + } + + trace_vfio_region_write(vbasedev->name, region->nr, addr, data, size); + + /* + * A read or write to a BAR always signals an INTx EOI. This will + * do nothing if not pending (including not in INTx mode). We assume + * that a BAR access is in response to an interrupt and that BAR + * accesses will service the interrupt. Unfortunately, we don't know + * which access will service the interrupt, so we're potentially + * getting quite a few host interrupts per guest interrupt. + */ + vbasedev->ops->vfio_eoi(vbasedev); +} + +uint64_t vfio_region_read(void *opaque, + hwaddr addr, unsigned size) +{ + VFIORegion *region = opaque; + VFIODevice *vbasedev = region->vbasedev; + union { + uint8_t byte; + uint16_t word; + uint32_t dword; + uint64_t qword; + } buf; + uint64_t data = 0; + + if (pread(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { + error_report("%s(%s:region%d+0x%"HWADDR_PRIx", %d) failed: %m", + __func__, vbasedev->name, region->nr, + addr, size); + return (uint64_t)-1; + } + switch (size) { + case 1: + data = buf.byte; + break; + case 2: + data = le16_to_cpu(buf.word); + break; + case 4: + data = le32_to_cpu(buf.dword); + break; + case 8: + data = le64_to_cpu(buf.qword); + break; + default: + hw_error("vfio: unsupported read size, %u bytes", size); + break; + } + + trace_vfio_region_read(vbasedev->name, region->nr, addr, size, data); + + /* Same as write above */ + vbasedev->ops->vfio_eoi(vbasedev); + + return data; +} + +static const MemoryRegionOps vfio_region_ops = { + .read = vfio_region_read, + .write = vfio_region_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 1, + .max_access_size = 8, + }, + .impl = { + .min_access_size = 1, + .max_access_size = 8, + }, +}; + +static int vfio_setup_region_sparse_mmaps(VFIORegion *region, + struct vfio_region_info *info) +{ + struct vfio_info_cap_header *hdr; + struct vfio_region_info_cap_sparse_mmap *sparse; + int i, j; + + hdr = vfio_get_region_info_cap(info, VFIO_REGION_INFO_CAP_SPARSE_MMAP); + if (!hdr) { + return -ENODEV; + } + + sparse = container_of(hdr, struct vfio_region_info_cap_sparse_mmap, header); + + trace_vfio_region_sparse_mmap_header(region->vbasedev->name, + region->nr, sparse->nr_areas); + + region->mmaps = g_new0(VFIOMmap, sparse->nr_areas); + + for (i = 0, j = 0; i < sparse->nr_areas; i++) { + if (sparse->areas[i].size) { + trace_vfio_region_sparse_mmap_entry(i, sparse->areas[i].offset, + sparse->areas[i].offset + + sparse->areas[i].size - 1); + region->mmaps[j].offset = sparse->areas[i].offset; + region->mmaps[j].size = sparse->areas[i].size; + j++; + } + } + + region->nr_mmaps = j; + region->mmaps = g_realloc(region->mmaps, j * sizeof(VFIOMmap)); + + return 0; +} + +int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, + int index, const char *name) +{ + g_autofree struct vfio_region_info *info = NULL; + int ret; + + ret = vfio_get_region_info(vbasedev, index, &info); + if (ret) { + return ret; + } + + region->vbasedev = vbasedev; + region->flags = info->flags; + region->size = info->size; + region->fd_offset = info->offset; + region->nr = index; + + if (region->size) { + region->mem = g_new0(MemoryRegion, 1); + memory_region_init_io(region->mem, obj, &vfio_region_ops, + region, name, region->size); + + if (!vbasedev->no_mmap && + region->flags & VFIO_REGION_INFO_FLAG_MMAP) { + + ret = vfio_setup_region_sparse_mmaps(region, info); + + if (ret) { + region->nr_mmaps = 1; + region->mmaps = g_new0(VFIOMmap, region->nr_mmaps); + region->mmaps[0].offset = 0; + region->mmaps[0].size = region->size; + } + } + } + + trace_vfio_region_setup(vbasedev->name, index, name, + region->flags, region->fd_offset, region->size); + return 0; +} + +static void vfio_subregion_unmap(VFIORegion *region, int index) +{ + trace_vfio_region_unmap(memory_region_name(&region->mmaps[index].mem), + region->mmaps[index].offset, + region->mmaps[index].offset + + region->mmaps[index].size - 1); + memory_region_del_subregion(region->mem, &region->mmaps[index].mem); + munmap(region->mmaps[index].mmap, region->mmaps[index].size); + object_unparent(OBJECT(&region->mmaps[index].mem)); + region->mmaps[index].mmap = NULL; +} + +int vfio_region_mmap(VFIORegion *region) +{ + int i, ret, prot = 0; + char *name; + + if (!region->mem) { + return 0; + } + + prot |= region->flags & VFIO_REGION_INFO_FLAG_READ ? PROT_READ : 0; + prot |= region->flags & VFIO_REGION_INFO_FLAG_WRITE ? PROT_WRITE : 0; + + for (i = 0; i < region->nr_mmaps; i++) { + size_t align = MIN(1ULL << ctz64(region->mmaps[i].size), 1 * GiB); + void *map_base, *map_align; + + /* + * Align the mmap for more efficient mapping in the kernel. Ideally + * we'd know the PMD and PUD mapping sizes to use as discrete alignment + * intervals, but we don't. As of Linux v6.12, the largest PUD size + * supporting huge pfnmap is 1GiB (ARCH_SUPPORTS_PUD_PFNMAP is only set + * on x86_64). Align by power-of-two size, capped at 1GiB. + * + * NB. qemu_memalign() and friends actually allocate memory, whereas + * the region size here can exceed host memory, therefore we manually + * create an oversized anonymous mapping and clean it up for alignment. + */ + map_base = mmap(0, region->mmaps[i].size + align, PROT_NONE, + MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); + if (map_base == MAP_FAILED) { + ret = -errno; + goto no_mmap; + } + + map_align = (void *)ROUND_UP((uintptr_t)map_base, (uintptr_t)align); + munmap(map_base, map_align - map_base); + munmap(map_align + region->mmaps[i].size, + align - (map_align - map_base)); + + region->mmaps[i].mmap = mmap(map_align, region->mmaps[i].size, prot, + MAP_SHARED | MAP_FIXED, + region->vbasedev->fd, + region->fd_offset + + region->mmaps[i].offset); + if (region->mmaps[i].mmap == MAP_FAILED) { + ret = -errno; + goto no_mmap; + } + + name = g_strdup_printf("%s mmaps[%d]", + memory_region_name(region->mem), i); + memory_region_init_ram_device_ptr(&region->mmaps[i].mem, + memory_region_owner(region->mem), + name, region->mmaps[i].size, + region->mmaps[i].mmap); + g_free(name); + memory_region_add_subregion(region->mem, region->mmaps[i].offset, + &region->mmaps[i].mem); + + trace_vfio_region_mmap(memory_region_name(&region->mmaps[i].mem), + region->mmaps[i].offset, + region->mmaps[i].offset + + region->mmaps[i].size - 1); + } + + return 0; + +no_mmap: + trace_vfio_region_mmap_fault(memory_region_name(region->mem), i, + region->fd_offset + region->mmaps[i].offset, + region->fd_offset + region->mmaps[i].offset + + region->mmaps[i].size - 1, ret); + + region->mmaps[i].mmap = NULL; + + for (i--; i >= 0; i--) { + vfio_subregion_unmap(region, i); + } + + return ret; +} + +void vfio_region_unmap(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + vfio_subregion_unmap(region, i); + } + } +} + +void vfio_region_exit(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + memory_region_del_subregion(region->mem, &region->mmaps[i].mem); + } + } + + trace_vfio_region_exit(region->vbasedev->name, region->nr); +} + +void vfio_region_finalize(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + munmap(region->mmaps[i].mmap, region->mmaps[i].size); + object_unparent(OBJECT(&region->mmaps[i].mem)); + } + } + + object_unparent(OBJECT(region->mem)); + + g_free(region->mem); + g_free(region->mmaps); + + trace_vfio_region_finalize(region->vbasedev->name, region->nr); + + region->mem = NULL; + region->mmaps = NULL; + region->nr_mmaps = 0; + region->size = 0; + region->flags = 0; + region->nr = 0; +} + +void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + memory_region_set_enabled(&region->mmaps[i].mem, enabled); + } + } + + trace_vfio_region_mmaps_set_enabled(memory_region_name(region->mem), + enabled); +} diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'migration.c', 'migration-multifd.c', 'cpr.c', + 'region.c', )) system_ss.add(when: ['CONFIG_VFIO', 'CONFIG_IOMMUFD'], if_true: files( 'iommufd.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_igd_host_bridge_enabled(const char *name) "%s" vfio_pci_igd_lpc_bridge_enabled(const char *name) "%s" # common.c -vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" -vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 vfio_iommu_map_notify(const char *op, uint64_t iova_start, uint64_t iova_end) "iommu %s @ 0x%"PRIx64" - 0x%"PRIx64 vfio_listener_region_skip(const char *name, uint64_t start, uint64_t end) "SKIPPING %s 0x%"PRIx64" - 0x%"PRIx64 vfio_spapr_group_attach(int groupfd, int tablefd) "Attached groupfd %d to liobn fd %d" @@ -XXX,XX +XXX,XX @@ vfio_disconnect_container(int fd) "close container->fd=%d" vfio_put_group(int fd) "close group->fd=%d" vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" vfio_put_base_device(int fd) "close vdev->fd=%d" +vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_legacy_dma_unmap_overflow_workaround(void) "" +vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 +vfio_reset_handler(void) "" + +# region.c +vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" +vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 vfio_region_setup(const char *dev, int index, const char *name, unsigned long flags, unsigned long offset, unsigned long size) "Device %s, region %d \"%s\", flags: 0x%lx, offset: 0x%lx, size: 0x%lx" vfio_region_mmap_fault(const char *name, int index, unsigned long offset, unsigned long size, int fault) "Region %s mmaps[%d], [0x%lx - 0x%lx], fault: %d" vfio_region_mmap(const char *name, unsigned long offset, unsigned long end) "Region %s [0x%lx - 0x%lx]" @@ -XXX,XX +XXX,XX @@ vfio_region_mmaps_set_enabled(const char *name, bool enabled) "Region %s mmaps e vfio_region_unmap(const char *name, unsigned long offset, unsigned long end) "Region %s unmap [0x%lx - 0x%lx]" vfio_region_sparse_mmap_header(const char *name, int index, int nr_areas) "Device %s region %d: %d sparse mmap entries" vfio_region_sparse_mmap_entry(int i, unsigned long start, unsigned long end) "sparse entry %d [0x%lx - 0x%lx]" -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -vfio_legacy_dma_unmap_overflow_workaround(void) "" -vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 -vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 -vfio_reset_handler(void) "" # platform.c vfio_platform_realize(char *name, char *compat) "vfio device %s, compat = %s" -- 2.48.1 Gather all VFIOcontainer related declarations into "container.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". These declarations are made available externally for PPC and s390x. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 19 ----------------- include/hw/vfio/vfio-container.h | 35 ++++++++++++++++++++++++++++++++ hw/ppc/spapr_pci_vfio.c | 1 + hw/s390x/s390-pci-vfio.c | 2 +- hw/vfio/container.c | 1 + hw/vfio/spapr.c | 1 + 6 files changed, 39 insertions(+), 20 deletions(-) create mode 100644 include/hw/vfio/vfio-container.h diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ enum { VFIO_DEVICE_TYPE_CCW = 2, VFIO_DEVICE_TYPE_AP = 3, }; -struct VFIOGroup; - -typedef struct VFIOContainer { - VFIOContainerBase bcontainer; - int fd; /* /dev/vfio/vfio, empowered by the attached groups */ - unsigned iommu_type; - QLIST_HEAD(, VFIOGroup) group_list; -} VFIOContainer; - -OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); typedef struct VFIODeviceOps VFIODeviceOps; typedef struct VFIOMigration VFIOMigration; @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; -typedef struct VFIOGroup { - int fd; - int groupid; - VFIOContainer *container; - QLIST_HEAD(, VFIODevice) device_list; - QLIST_ENTRY(VFIOGroup) next; - QLIST_ENTRY(VFIOGroup) container_next; - bool ram_block_discard_allowed; -} VFIOGroup; #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ diff --git a/include/hw/vfio/vfio-container.h b/include/hw/vfio/vfio-container.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-container.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO container + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_CONTAINER_H +#define HW_VFIO_CONTAINER_H + +#include "hw/vfio/vfio-container-base.h" + +typedef struct VFIOContainer VFIOContainer; + +typedef struct VFIOGroup { + int fd; + int groupid; + VFIOContainer *container; + QLIST_HEAD(, VFIODevice) device_list; + QLIST_ENTRY(VFIOGroup) next; + QLIST_ENTRY(VFIOGroup) container_next; + bool ram_block_discard_allowed; +} VFIOGroup; + +typedef struct VFIOContainer { + VFIOContainerBase bcontainer; + int fd; /* /dev/vfio/vfio, empowered by the attached groups */ + unsigned iommu_type; + QLIST_HEAD(, VFIOGroup) group_list; +} VFIOContainer; + +OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); + +#endif /* HW_VFIO_CONTAINER_H */ diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/pci/msix.h" #include "hw/pci/pci_device.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" #include "qemu/error-report.h" #include CONFIG_DEVICES /* CONFIG_VFIO_PCI */ diff --git a/hw/s390x/s390-pci-vfio.c b/hw/s390x/s390-pci-vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/s390-pci-vfio.c +++ b/hw/s390x/s390-pci-vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/s390x/s390-pci-clp.h" #include "hw/s390x/s390-pci-vfio.h" #include "hw/vfio/pci.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" /* * Get the current DMA available count from vfio. Returns true if vfio is diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "qapi/error.h" #include "pci.h" +#include "hw/vfio/vfio-container.h" VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "exec/address-spaces.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" #include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" -- 2.48.1 vfio_group_list is only used in file "container.c". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/container.c | 3 ++- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp); bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); -typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; -extern VFIOGroupList vfio_group_list; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" -VFIOGroupList vfio_group_list = +typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; +static VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); static int vfio_ram_block_discard_disable(VFIOContainer *container, bool state) -- 2.48.1 VFIOAddressSpace is a common object used by VFIOContainerBase which is declared in "hw/vfio/vfio-container-base.h". Move the VFIOAddressSpace related services into "container-base.c". While at it, rename : vfio_get_address_space -> vfio_address_space_get vfio_put_address_space -> vfio_address_space_put to better reflect the namespace these routines belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 6 --- include/hw/vfio/vfio-container-base.h | 5 ++ hw/ppc/spapr_pci_vfio.c | 5 +- hw/vfio/common.c | 66 ------------------------- hw/vfio/container-base.c | 69 +++++++++++++++++++++++++++ hw/vfio/container.c | 6 +-- hw/vfio/iommufd.c | 6 +-- hw/vfio/trace-events | 4 +- 8 files changed, 85 insertions(+), 82 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" -VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); -void vfio_put_address_space(VFIOAddressSpace *space); -void vfio_address_space_insert(VFIOAddressSpace *space, - VFIOContainerBase *bcontainer); - void vfio_disable_irqindex(VFIODevice *vbasedev, int index); void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); -void vfio_reset_handler(void *opaque); struct vfio_device_info *vfio_get_device_info(int fd); bool vfio_device_is_mdev(VFIODevice *vbasedev); bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIORamDiscardListener { QLIST_ENTRY(VFIORamDiscardListener) next; } VFIORamDiscardListener; +VFIOAddressSpace *vfio_address_space_get(AddressSpace *as); +void vfio_address_space_put(VFIOAddressSpace *space); +void vfio_address_space_insert(VFIOAddressSpace *space, + VFIOContainerBase *bcontainer); + int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly); diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/pci-host/spapr.h" #include "hw/pci/msix.h" #include "hw/pci/pci_device.h" -#include "hw/vfio/vfio-common.h" #include "hw/vfio/vfio-container.h" #include "qemu/error-report.h" #include CONFIG_DEVICES /* CONFIG_VFIO_PCI */ @@ -XXX,XX +XXX,XX @@ static int vfio_eeh_container_op(VFIOContainer *container, uint32_t op) static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) { - VFIOAddressSpace *space = vfio_get_address_space(as); + VFIOAddressSpace *space = vfio_address_space_get(as); VFIOContainerBase *bcontainer = NULL; if (QLIST_EMPTY(&space->containers)) { @@ -XXX,XX +XXX,XX @@ static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) } out: - vfio_put_address_space(space); + vfio_address_space_put(space); return container_of(bcontainer, VFIOContainer, bcontainer); } diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "qemu/main-loop.h" #include "qemu/range.h" #include "system/kvm.h" -#include "system/reset.h" #include "system/runstate.h" #include "trace.h" #include "qapi/error.h" @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); -static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = - QLIST_HEAD_INITIALIZER(vfio_address_spaces); #ifdef CONFIG_KVM /* @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_sync = vfio_listener_log_sync, }; -void vfio_reset_handler(void *opaque) -{ - VFIODevice *vbasedev; - - trace_vfio_reset_handler(); - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->dev->realized) { - vbasedev->ops->vfio_compute_needs_reset(vbasedev); - } - } - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->dev->realized && vbasedev->needs_reset) { - vbasedev->ops->vfio_hot_reset_multi(vbasedev); - } - } -} - int vfio_kvm_device_add_fd(int fd, Error **errp) { #ifdef CONFIG_KVM @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp) return 0; } -VFIOAddressSpace *vfio_get_address_space(AddressSpace *as) -{ - VFIOAddressSpace *space; - - QLIST_FOREACH(space, &vfio_address_spaces, list) { - if (space->as == as) { - return space; - } - } - - /* No suitable VFIOAddressSpace, create a new one */ - space = g_malloc0(sizeof(*space)); - space->as = as; - QLIST_INIT(&space->containers); - - if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_register_reset(vfio_reset_handler, NULL); - } - - QLIST_INSERT_HEAD(&vfio_address_spaces, space, list); - - return space; -} - -void vfio_put_address_space(VFIOAddressSpace *space) -{ - if (!QLIST_EMPTY(&space->containers)) { - return; - } - - QLIST_REMOVE(space, list); - g_free(space); - - if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_unregister_reset(vfio_reset_handler, NULL); - } -} - -void vfio_address_space_insert(VFIOAddressSpace *space, - VFIOContainerBase *bcontainer) -{ - QLIST_INSERT_HEAD(&space->containers, bcontainer, next); - bcontainer->space = space; -} - struct vfio_device_info *vfio_get_device_info(int fd) { struct vfio_device_info *info; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "qapi/error.h" #include "qemu/error-report.h" +#include "system/reset.h" #include "hw/vfio/vfio-container-base.h" +#include "hw/vfio/vfio-common.h" /* for vfio_device_list */ +#include "trace.h" + +static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = + QLIST_HEAD_INITIALIZER(vfio_address_spaces); + +static void vfio_reset_handler(void *opaque) +{ + VFIODevice *vbasedev; + + trace_vfio_reset_handler(); + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->dev->realized) { + vbasedev->ops->vfio_compute_needs_reset(vbasedev); + } + } + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->dev->realized && vbasedev->needs_reset) { + vbasedev->ops->vfio_hot_reset_multi(vbasedev); + } + } +} + +VFIOAddressSpace *vfio_address_space_get(AddressSpace *as) +{ + VFIOAddressSpace *space; + + QLIST_FOREACH(space, &vfio_address_spaces, list) { + if (space->as == as) { + return space; + } + } + + /* No suitable VFIOAddressSpace, create a new one */ + space = g_malloc0(sizeof(*space)); + space->as = as; + QLIST_INIT(&space->containers); + + if (QLIST_EMPTY(&vfio_address_spaces)) { + qemu_register_reset(vfio_reset_handler, NULL); + } + + QLIST_INSERT_HEAD(&vfio_address_spaces, space, list); + + return space; +} + +void vfio_address_space_put(VFIOAddressSpace *space) +{ + if (!QLIST_EMPTY(&space->containers)) { + return; + } + + QLIST_REMOVE(space, list); + g_free(space); + + if (QLIST_EMPTY(&vfio_address_spaces)) { + qemu_unregister_reset(vfio_reset_handler, NULL); + } +} + +void vfio_address_space_insert(VFIOAddressSpace *space, + VFIOContainerBase *bcontainer) +{ + QLIST_INSERT_HEAD(&space->containers, bcontainer, next); + bcontainer->space = space; +} int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, VFIOAddressSpace *space; VFIOIOMMUClass *vioc; - space = vfio_get_address_space(as); + space = vfio_address_space_get(as); /* * VFIO is currently incompatible with discarding of RAM insofar as the @@ -XXX,XX +XXX,XX @@ close_fd_exit: close(fd); put_space_exit: - vfio_put_address_space(space); + vfio_address_space_put(space); return false; } @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) close(container->fd); object_unref(container); - vfio_put_address_space(space); + vfio_address_space_put(space); } } diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ static bool iommufd_cdev_attach(const char *name, VFIODevice *vbasedev, goto err_connect_bind; } - space = vfio_get_address_space(as); + space = vfio_address_space_get(as); /* * The HostIOMMUDevice data from legacy backend is static and doesn't need @@ -XXX,XX +XXX,XX @@ err_discard_disable: err_attach_container: iommufd_cdev_container_destroy(container); err_alloc_ioas: - vfio_put_address_space(space); + vfio_address_space_put(space); iommufd_cdev_unbind_and_disconnect(vbasedev); err_connect_bind: close(vbasedev->fd); @@ -XXX,XX +XXX,XX @@ static void iommufd_cdev_detach(VFIODevice *vbasedev) vfio_cpr_unregister_container(bcontainer); iommufd_cdev_detach_container(vbasedev, container); iommufd_cdev_container_destroy(container); - vfio_put_address_space(space); + vfio_address_space_put(space); iommufd_cdev_unbind_and_disconnect(vbasedev); close(vbasedev->fd); diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 -vfio_reset_handler(void) "" # region.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" @@ -XXX,XX +XXX,XX @@ iommufd_cdev_fail_attach_existing_container(const char *msg) " %s" iommufd_cdev_alloc_ioas(int iommufd, int ioas_id) " [iommufd=%d] new IOMMUFD container with ioasid=%d" iommufd_cdev_device_info(char *name, int devfd, int num_irqs, int num_regions, int flags) " %s (%d) num_irqs=%d num_regions=%d flags=%d" iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int dev_id) "\t%04x:%02x:%02x.%x devid %d" + +# container-base.c +vfio_reset_handler(void) "" -- 2.48.1 These definitions don't have any use outside of their respective submodules. There is no need to expose them externally. Keep them private. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 5 ----- hw/vfio/container.c | 2 ++ hw/vfio/iommufd.c | 3 +++ 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; - -#define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" -#define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ - TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" - void vfio_disable_irqindex(VFIODevice *vbasedev, int index); void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" +#define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" + typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; static VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "migration.h" #include "iommufd.h" +#define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ + TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" + static int iommufd_cdev_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) { -- 2.48.1 Gather all helper routine declarations into "helpers.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/helpers.h | 26 ++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 7 ------- hw/s390x/s390-pci-vfio.c | 1 + hw/vfio/common.c | 1 + hw/vfio/container.c | 1 + hw/vfio/helpers.c | 1 + hw/vfio/pci.c | 1 + hw/vfio/region.c | 1 + 8 files changed, 32 insertions(+), 7 deletions(-) create mode 100644 hw/vfio/helpers.h diff --git a/hw/vfio/helpers.h b/hw/vfio/helpers.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/helpers.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO helpers + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_HELPERS_H +#define HW_VFIO_HELPERS_H + +#ifdef CONFIG_LINUX +#include <linux/vfio.h> + +struct vfio_info_cap_header * +vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); +struct vfio_info_cap_header * +vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); +struct vfio_info_cap_header * +vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); + +#endif + +int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); + +#endif /* HW_VFIO_HELPERS_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_get_region_info(VFIODevice *vbasedev, int index, int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); -struct vfio_info_cap_header * -vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, unsigned int *avail); -struct vfio_info_cap_header * -vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); -struct vfio_info_cap_header * -vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); #endif -int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); bool vfio_devices_all_dirty_tracking_started( const VFIOContainerBase *bcontainer); bool diff --git a/hw/s390x/s390-pci-vfio.c b/hw/s390x/s390-pci-vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/s390-pci-vfio.c +++ b/hw/s390x/s390-pci-vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/s390x/s390-pci-vfio.h" #include "hw/vfio/pci.h" #include "hw/vfio/vfio-container.h" +#include "hw/vfio/helpers.h" /* * Get the current DMA available count from vfio. Returns true if vfio is diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "system/tcg.h" #include "system/tpm.h" #include "migration.h" +#include "helpers.h" VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "pci.h" #include "hw/vfio/vfio-container.h" +#include "helpers.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "qemu/units.h" #include "monitor/monitor.h" +#include "helpers.h" /* * Common VFIO interrupt disable diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "system/iommufd.h" #include "migration.h" #include "hw/vfio/vfio-region.h" +#include "helpers.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "qemu/units.h" #include "monitor/monitor.h" +#include "helpers.h" /* * IO Port/MMIO - Beware of the endians, VFIO is always little endian -- 2.48.1 vfio_get_info_dma_avail() is a low level routine similar to the other routines extracting capabilities from 'struct vfio_iommu_type1_info'. It belongs to file "helpers.c". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/helpers.h | 5 ++++- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/container.c | 31 ------------------------------- hw/vfio/helpers.c | 31 +++++++++++++++++++++++++++++++ 4 files changed, 35 insertions(+), 34 deletions(-) diff --git a/hw/vfio/helpers.h b/hw/vfio/helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.h +++ b/hw/vfio/helpers.h @@ -XXX,XX +XXX,XX @@ struct vfio_info_cap_header * vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); - +struct vfio_info_cap_header * +vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id); +bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, + unsigned int *avail); #endif int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_get_region_info(VFIODevice *vbasedev, int index, int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); -bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, - unsigned int *avail); #endif bool vfio_devices_all_dirty_tracking_started( diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_query_dirty_bitmap(const VFIOContainerBase *bcontainer, return ret; } -static struct vfio_info_cap_header * -vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) -{ - if (!(info->flags & VFIO_IOMMU_INFO_CAPS)) { - return NULL; - } - - return vfio_get_cap((void *)info, info->cap_offset, id); -} - -bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, - unsigned int *avail) -{ - struct vfio_info_cap_header *hdr; - struct vfio_iommu_type1_info_dma_avail *cap; - - /* If the capability cannot be found, assume no DMA limiting */ - hdr = vfio_get_iommu_type1_info_cap(info, - VFIO_IOMMU_TYPE1_INFO_DMA_AVAIL); - if (!hdr) { - return false; - } - - if (avail != NULL) { - cap = (void *) hdr; - *avail = cap->avail; - } - - return true; -} - static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, VFIOContainerBase *bcontainer) { diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ retry: return 0; } +struct vfio_info_cap_header * +vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) +{ + if (!(info->flags & VFIO_IOMMU_INFO_CAPS)) { + return NULL; + } + + return vfio_get_cap((void *)info, info->cap_offset, id); +} + +bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, + unsigned int *avail) +{ + struct vfio_info_cap_header *hdr; + struct vfio_iommu_type1_info_dma_avail *cap; + + /* If the capability cannot be found, assume no DMA limiting */ + hdr = vfio_get_iommu_type1_info_cap(info, + VFIO_IOMMU_TYPE1_INFO_DMA_AVAIL); + if (!hdr) { + return false; + } + + if (avail != NULL) { + cap = (void *) hdr; + *avail = cap->avail; + } + + return true; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { -- 2.48.1 vfio_kvm_device_add/del_fd() are low level routines. Move them with the other helpers. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/helpers.h | 3 ++ include/hw/vfio/vfio-common.h | 3 -- hw/vfio/common.c | 58 ---------------------------------- hw/vfio/helpers.c | 59 +++++++++++++++++++++++++++++++++++ hw/vfio/iommufd.c | 1 + hw/vfio/meson.build | 2 +- 6 files changed, 64 insertions(+), 62 deletions(-) diff --git a/hw/vfio/helpers.h b/hw/vfio/helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.h +++ b/hw/vfio/helpers.h @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); +int vfio_kvm_device_add_fd(int fd, Error **errp); +int vfio_kvm_device_del_fd(int fd, Error **errp); + #endif /* HW_VFIO_HELPERS_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, void vfio_detach_device(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); -int vfio_kvm_device_add_fd(int fd, Error **errp); -int vfio_kvm_device_del_fd(int fd, Error **errp); - bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_sync = vfio_listener_log_sync, }; -int vfio_kvm_device_add_fd(int fd, Error **errp) -{ -#ifdef CONFIG_KVM - struct kvm_device_attr attr = { - .group = KVM_DEV_VFIO_FILE, - .attr = KVM_DEV_VFIO_FILE_ADD, - .addr = (uint64_t)(unsigned long)&fd, - }; - - if (!kvm_enabled()) { - return 0; - } - - if (vfio_kvm_device_fd < 0) { - struct kvm_create_device cd = { - .type = KVM_DEV_TYPE_VFIO, - }; - - if (kvm_vm_ioctl(kvm_state, KVM_CREATE_DEVICE, &cd)) { - error_setg_errno(errp, errno, "Failed to create KVM VFIO device"); - return -errno; - } - - vfio_kvm_device_fd = cd.fd; - } - - if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { - error_setg_errno(errp, errno, "Failed to add fd %d to KVM VFIO device", - fd); - return -errno; - } -#endif - return 0; -} - -int vfio_kvm_device_del_fd(int fd, Error **errp) -{ -#ifdef CONFIG_KVM - struct kvm_device_attr attr = { - .group = KVM_DEV_VFIO_FILE, - .attr = KVM_DEV_VFIO_FILE_DEL, - .addr = (uint64_t)(unsigned long)&fd, - }; - - if (vfio_kvm_device_fd < 0) { - error_setg(errp, "KVM VFIO device isn't created yet"); - return -EINVAL; - } - - if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { - error_setg_errno(errp, errno, - "Failed to remove fd %d from KVM VFIO device", fd); - return -errno; - } -#endif - return 0; -} - struct vfio_device_info *vfio_get_device_info(int fd) { struct vfio_device_info *info; diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> +#include "system/kvm.h" #include "hw/vfio/vfio-common.h" #include "hw/vfio/pci.h" #include "hw/hw.h" @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, return true; } +int vfio_kvm_device_add_fd(int fd, Error **errp) +{ +#ifdef CONFIG_KVM + struct kvm_device_attr attr = { + .group = KVM_DEV_VFIO_FILE, + .attr = KVM_DEV_VFIO_FILE_ADD, + .addr = (uint64_t)(unsigned long)&fd, + }; + + if (!kvm_enabled()) { + return 0; + } + + if (vfio_kvm_device_fd < 0) { + struct kvm_create_device cd = { + .type = KVM_DEV_TYPE_VFIO, + }; + + if (kvm_vm_ioctl(kvm_state, KVM_CREATE_DEVICE, &cd)) { + error_setg_errno(errp, errno, "Failed to create KVM VFIO device"); + return -errno; + } + + vfio_kvm_device_fd = cd.fd; + } + + if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { + error_setg_errno(errp, errno, "Failed to add fd %d to KVM VFIO device", + fd); + return -errno; + } +#endif + return 0; +} + +int vfio_kvm_device_del_fd(int fd, Error **errp) +{ +#ifdef CONFIG_KVM + struct kvm_device_attr attr = { + .group = KVM_DEV_VFIO_FILE, + .attr = KVM_DEV_VFIO_FILE_DEL, + .addr = (uint64_t)(unsigned long)&fd, + }; + + if (vfio_kvm_device_fd < 0) { + error_setg(errp, "KVM VFIO device isn't created yet"); + return -EINVAL; + } + + if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { + error_setg_errno(errp, errno, + "Failed to remove fd %d from KVM VFIO device", fd); + return -errno; + } +#endif + return 0; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "migration.h" #include "iommufd.h" +#include "helpers.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ vfio_ss = ss.source_set() vfio_ss.add(files( 'common.c', 'container.c', + 'helpers.c', )) vfio_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr.c')) vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files( @@ -XXX,XX +XXX,XX @@ specific_ss.add_all(when: 'CONFIG_VFIO', if_true: vfio_ss) system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( - 'helpers.c', 'container-base.c', 'migration.c', 'migration-multifd.c', -- 2.48.1 vfio_get_device_info() is a low level routine. Move it with the other helpers. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/helpers.h | 1 + include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 24 ------------------------ hw/vfio/helpers.c | 24 ++++++++++++++++++++++++ 4 files changed, 25 insertions(+), 25 deletions(-) diff --git a/hw/vfio/helpers.h b/hw/vfio/helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.h +++ b/hw/vfio/helpers.h @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, #endif int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); +struct vfio_device_info *vfio_get_device_info(int fd); int vfio_kvm_device_add_fd(int fd, Error **errp); int vfio_kvm_device_del_fd(int fd, Error **errp); diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); -struct vfio_device_info *vfio_get_device_info(int fd); bool vfio_device_is_mdev(VFIODevice *vbasedev); bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp); bool vfio_attach_device(char *name, VFIODevice *vbasedev, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_sync = vfio_listener_log_sync, }; -struct vfio_device_info *vfio_get_device_info(int fd) -{ - struct vfio_device_info *info; - uint32_t argsz = sizeof(*info); - - info = g_malloc0(argsz); - -retry: - info->argsz = argsz; - - if (ioctl(fd, VFIO_DEVICE_GET_INFO, info)) { - g_free(info); - return NULL; - } - - if (info->argsz > argsz) { - argsz = info->argsz; - info = g_realloc(info, argsz); - goto retry; - } - - return info; -} - bool vfio_attach_device(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp) return 0; } +struct vfio_device_info *vfio_get_device_info(int fd) +{ + struct vfio_device_info *info; + uint32_t argsz = sizeof(*info); + + info = g_malloc0(argsz); + +retry: + info->argsz = argsz; + + if (ioctl(fd, VFIO_DEVICE_GET_INFO, info)) { + g_free(info); + return NULL; + } + + if (info->argsz > argsz) { + argsz = info->argsz; + info = g_realloc(info, argsz); + goto retry; + } + + return info; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { -- 2.48.1 Move all VFIODevice related routines of helpers.c into a new "device.c" file. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/device.c | 331 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/helpers.c | 303 --------------------------------------- hw/vfio/meson.build | 1 + hw/vfio/trace-events | 4 +- 4 files changed, 335 insertions(+), 304 deletions(-) create mode 100644 hw/vfio/device.c diff --git a/hw/vfio/device.c b/hw/vfio/device.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ +/* + * low level and IOMMU backend agnostic helpers used by VFIO devices, + * related to regions, interrupts, capabilities + * + * Copyright Red Hat, Inc. 2012 + * + * Authors: + * Alex Williamson <alex.williamson@redhat.com> + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + * + * Based on qemu-kvm device-assignment: + * Adapted for KVM by Qumranet. + * Copyright (c) 2007, Neocleus, Alex Novik (alex@neocleus.com) + * Copyright (c) 2007, Neocleus, Guy Zana (guy@neocleus.com) + * Copyright (C) 2008, Qumranet, Amit Shah (amit.shah@qumranet.com) + * Copyright (C) 2008, Red Hat, Amit Shah (amit.shah@redhat.com) + * Copyright (C) 2008, IBM, Muli Ben-Yehuda (muli@il.ibm.com) + */ + +#include "qemu/osdep.h" +#include <sys/ioctl.h> + +#include "hw/vfio/vfio-common.h" +#include "hw/vfio/pci.h" +#include "hw/hw.h" +#include "trace.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/units.h" +#include "monitor/monitor.h" +#include "helpers.h" + +/* + * Common VFIO interrupt disable + */ +void vfio_disable_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_TRIGGER, + .index = index, + .start = 0, + .count = 0, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_UNMASK, + .index = index, + .start = 0, + .count = 1, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_MASK, + .index = index, + .start = 0, + .count = 1, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +static inline const char *action_to_str(int action) +{ + switch (action) { + case VFIO_IRQ_SET_ACTION_MASK: + return "MASK"; + case VFIO_IRQ_SET_ACTION_UNMASK: + return "UNMASK"; + case VFIO_IRQ_SET_ACTION_TRIGGER: + return "TRIGGER"; + default: + return "UNKNOWN ACTION"; + } +} + +static const char *index_to_str(VFIODevice *vbasedev, int index) +{ + if (vbasedev->type != VFIO_DEVICE_TYPE_PCI) { + return NULL; + } + + switch (index) { + case VFIO_PCI_INTX_IRQ_INDEX: + return "INTX"; + case VFIO_PCI_MSI_IRQ_INDEX: + return "MSI"; + case VFIO_PCI_MSIX_IRQ_INDEX: + return "MSIX"; + case VFIO_PCI_ERR_IRQ_INDEX: + return "ERR"; + case VFIO_PCI_REQ_IRQ_INDEX: + return "REQ"; + default: + return NULL; + } +} + +bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, + int action, int fd, Error **errp) +{ + ERRP_GUARD(); + g_autofree struct vfio_irq_set *irq_set = NULL; + int argsz; + const char *name; + int32_t *pfd; + + argsz = sizeof(*irq_set) + sizeof(*pfd); + + irq_set = g_malloc0(argsz); + irq_set->argsz = argsz; + irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | action; + irq_set->index = index; + irq_set->start = subindex; + irq_set->count = 1; + pfd = (int32_t *)&irq_set->data; + *pfd = fd; + + if (!ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set)) { + return true; + } + + error_setg_errno(errp, errno, "VFIO_DEVICE_SET_IRQS failure"); + + name = index_to_str(vbasedev, index); + if (name) { + error_prepend(errp, "%s-%d: ", name, subindex); + } else { + error_prepend(errp, "index %d-%d: ", index, subindex); + } + error_prepend(errp, + "Failed to %s %s eventfd signaling for interrupt ", + fd < 0 ? "tear down" : "set up", action_to_str(action)); + return false; +} + +int vfio_get_region_info(VFIODevice *vbasedev, int index, + struct vfio_region_info **info) +{ + size_t argsz = sizeof(struct vfio_region_info); + + *info = g_malloc0(argsz); + + (*info)->index = index; +retry: + (*info)->argsz = argsz; + + if (ioctl(vbasedev->fd, VFIO_DEVICE_GET_REGION_INFO, *info)) { + g_free(*info); + *info = NULL; + return -errno; + } + + if ((*info)->argsz > argsz) { + argsz = (*info)->argsz; + *info = g_realloc(*info, argsz); + + goto retry; + } + + return 0; +} + +int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, + uint32_t subtype, struct vfio_region_info **info) +{ + int i; + + for (i = 0; i < vbasedev->num_regions; i++) { + struct vfio_info_cap_header *hdr; + struct vfio_region_info_cap_type *cap_type; + + if (vfio_get_region_info(vbasedev, i, info)) { + continue; + } + + hdr = vfio_get_region_info_cap(*info, VFIO_REGION_INFO_CAP_TYPE); + if (!hdr) { + g_free(*info); + continue; + } + + cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); + + trace_vfio_get_dev_region(vbasedev->name, i, + cap_type->type, cap_type->subtype); + + if (cap_type->type == type && cap_type->subtype == subtype) { + return 0; + } + + g_free(*info); + } + + *info = NULL; + return -ENODEV; +} + +bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) +{ + g_autofree struct vfio_region_info *info = NULL; + bool ret = false; + + if (!vfio_get_region_info(vbasedev, region, &info)) { + if (vfio_get_region_info_cap(info, cap_type)) { + ret = true; + } + } + + return ret; +} + +bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp) +{ + ERRP_GUARD(); + struct stat st; + + if (vbasedev->fd < 0) { + if (stat(vbasedev->sysfsdev, &st) < 0) { + error_setg_errno(errp, errno, "no such host device"); + error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); + return false; + } + /* User may specify a name, e.g: VFIO platform device */ + if (!vbasedev->name) { + vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); + } + } else { + if (!vbasedev->iommufd) { + error_setg(errp, "Use FD passing only with iommufd backend"); + return false; + } + /* + * Give a name with fd so any function printing out vbasedev->name + * will not break. + */ + if (!vbasedev->name) { + vbasedev->name = g_strdup_printf("VFIO_FD%d", vbasedev->fd); + } + } + + return true; +} + +void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp) +{ + ERRP_GUARD(); + int fd = monitor_fd_param(monitor_cur(), str, errp); + + if (fd < 0) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vbasedev->fd = fd; +} + +void vfio_device_init(VFIODevice *vbasedev, int type, VFIODeviceOps *ops, + DeviceState *dev, bool ram_discard) +{ + vbasedev->type = type; + vbasedev->ops = ops; + vbasedev->dev = dev; + vbasedev->fd = -1; + + vbasedev->ram_block_discard_allowed = ram_discard; +} + +int vfio_device_get_aw_bits(VFIODevice *vdev) +{ + /* + * iova_ranges is a sorted list. For old kernels that support + * VFIO but not support query of iova ranges, iova_ranges is NULL, + * in this case HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX(64) is returned. + */ + GList *l = g_list_last(vdev->bcontainer->iova_ranges); + + if (l) { + Range *range = l->data; + return range_get_last_bit(range) + 1; + } + + return HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX; +} + +bool vfio_device_is_mdev(VFIODevice *vbasedev) +{ + g_autofree char *subsys = NULL; + g_autofree char *tmp = NULL; + + if (!vbasedev->sysfsdev) { + return false; + } + + tmp = g_strdup_printf("%s/subsystem", vbasedev->sysfsdev); + subsys = realpath(tmp, NULL); + return subsys && (strcmp(subsys, "/sys/bus/mdev") == 0); +} + +bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp) +{ + HostIOMMUDevice *hiod = vbasedev->hiod; + + if (!hiod) { + return true; + } + + return HOST_IOMMU_DEVICE_GET_CLASS(hiod)->realize(hiod, vbasedev, errp); +} + +VFIODevice *vfio_get_vfio_device(Object *obj) +{ + if (object_dynamic_cast(obj, TYPE_VFIO_PCI)) { + return &VFIO_PCI(obj)->vbasedev; + } else { + return NULL; + } +} diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "system/kvm.h" #include "hw/vfio/vfio-common.h" -#include "hw/vfio/pci.h" #include "hw/hw.h" -#include "trace.h" #include "qapi/error.h" -#include "qemu/error-report.h" -#include "qemu/units.h" -#include "monitor/monitor.h" #include "helpers.h" -/* - * Common VFIO interrupt disable - */ -void vfio_disable_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_TRIGGER, - .index = index, - .start = 0, - .count = 0, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_UNMASK, - .index = index, - .start = 0, - .count = 1, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_MASK, - .index = index, - .start = 0, - .count = 1, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -static inline const char *action_to_str(int action) -{ - switch (action) { - case VFIO_IRQ_SET_ACTION_MASK: - return "MASK"; - case VFIO_IRQ_SET_ACTION_UNMASK: - return "UNMASK"; - case VFIO_IRQ_SET_ACTION_TRIGGER: - return "TRIGGER"; - default: - return "UNKNOWN ACTION"; - } -} - -static const char *index_to_str(VFIODevice *vbasedev, int index) -{ - if (vbasedev->type != VFIO_DEVICE_TYPE_PCI) { - return NULL; - } - - switch (index) { - case VFIO_PCI_INTX_IRQ_INDEX: - return "INTX"; - case VFIO_PCI_MSI_IRQ_INDEX: - return "MSI"; - case VFIO_PCI_MSIX_IRQ_INDEX: - return "MSIX"; - case VFIO_PCI_ERR_IRQ_INDEX: - return "ERR"; - case VFIO_PCI_REQ_IRQ_INDEX: - return "REQ"; - default: - return NULL; - } -} - -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, - int action, int fd, Error **errp) -{ - ERRP_GUARD(); - g_autofree struct vfio_irq_set *irq_set = NULL; - int argsz; - const char *name; - int32_t *pfd; - - argsz = sizeof(*irq_set) + sizeof(*pfd); - - irq_set = g_malloc0(argsz); - irq_set->argsz = argsz; - irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | action; - irq_set->index = index; - irq_set->start = subindex; - irq_set->count = 1; - pfd = (int32_t *)&irq_set->data; - *pfd = fd; - - if (!ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set)) { - return true; - } - - error_setg_errno(errp, errno, "VFIO_DEVICE_SET_IRQS failure"); - - name = index_to_str(vbasedev, index); - if (name) { - error_prepend(errp, "%s-%d: ", name, subindex); - } else { - error_prepend(errp, "index %d-%d: ", index, subindex); - } - error_prepend(errp, - "Failed to %s %s eventfd signaling for interrupt ", - fd < 0 ? "tear down" : "set up", action_to_str(action)); - return false; -} - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size) { vbmap->pages = REAL_HOST_PAGE_ALIGN(size) / qemu_real_host_page_size(); @@ -XXX,XX +XXX,XX @@ vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id) return vfio_get_cap((void *)info, info->cap_offset, id); } -int vfio_get_region_info(VFIODevice *vbasedev, int index, - struct vfio_region_info **info) -{ - size_t argsz = sizeof(struct vfio_region_info); - - *info = g_malloc0(argsz); - - (*info)->index = index; -retry: - (*info)->argsz = argsz; - - if (ioctl(vbasedev->fd, VFIO_DEVICE_GET_REGION_INFO, *info)) { - g_free(*info); - *info = NULL; - return -errno; - } - - if ((*info)->argsz > argsz) { - argsz = (*info)->argsz; - *info = g_realloc(*info, argsz); - - goto retry; - } - - return 0; -} - struct vfio_info_cap_header * vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) { @@ -XXX,XX +XXX,XX @@ retry: return info; } - -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, - uint32_t subtype, struct vfio_region_info **info) -{ - int i; - - for (i = 0; i < vbasedev->num_regions; i++) { - struct vfio_info_cap_header *hdr; - struct vfio_region_info_cap_type *cap_type; - - if (vfio_get_region_info(vbasedev, i, info)) { - continue; - } - - hdr = vfio_get_region_info_cap(*info, VFIO_REGION_INFO_CAP_TYPE); - if (!hdr) { - g_free(*info); - continue; - } - - cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); - - trace_vfio_get_dev_region(vbasedev->name, i, - cap_type->type, cap_type->subtype); - - if (cap_type->type == type && cap_type->subtype == subtype) { - return 0; - } - - g_free(*info); - } - - *info = NULL; - return -ENODEV; -} - -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) -{ - g_autofree struct vfio_region_info *info = NULL; - bool ret = false; - - if (!vfio_get_region_info(vbasedev, region, &info)) { - if (vfio_get_region_info_cap(info, cap_type)) { - ret = true; - } - } - - return ret; -} - -bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp) -{ - ERRP_GUARD(); - struct stat st; - - if (vbasedev->fd < 0) { - if (stat(vbasedev->sysfsdev, &st) < 0) { - error_setg_errno(errp, errno, "no such host device"); - error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); - return false; - } - /* User may specify a name, e.g: VFIO platform device */ - if (!vbasedev->name) { - vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); - } - } else { - if (!vbasedev->iommufd) { - error_setg(errp, "Use FD passing only with iommufd backend"); - return false; - } - /* - * Give a name with fd so any function printing out vbasedev->name - * will not break. - */ - if (!vbasedev->name) { - vbasedev->name = g_strdup_printf("VFIO_FD%d", vbasedev->fd); - } - } - - return true; -} - -void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp) -{ - ERRP_GUARD(); - int fd = monitor_fd_param(monitor_cur(), str, errp); - - if (fd < 0) { - error_prepend(errp, "Could not parse remote object fd %s:", str); - return; - } - vbasedev->fd = fd; -} - -void vfio_device_init(VFIODevice *vbasedev, int type, VFIODeviceOps *ops, - DeviceState *dev, bool ram_discard) -{ - vbasedev->type = type; - vbasedev->ops = ops; - vbasedev->dev = dev; - vbasedev->fd = -1; - - vbasedev->ram_block_discard_allowed = ram_discard; -} - -int vfio_device_get_aw_bits(VFIODevice *vdev) -{ - /* - * iova_ranges is a sorted list. For old kernels that support - * VFIO but not support query of iova ranges, iova_ranges is NULL, - * in this case HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX(64) is returned. - */ - GList *l = g_list_last(vdev->bcontainer->iova_ranges); - - if (l) { - Range *range = l->data; - return range_get_last_bit(range) + 1; - } - - return HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX; -} - -bool vfio_device_is_mdev(VFIODevice *vbasedev) -{ - g_autofree char *subsys = NULL; - g_autofree char *tmp = NULL; - - if (!vbasedev->sysfsdev) { - return false; - } - - tmp = g_strdup_printf("%s/subsystem", vbasedev->sysfsdev); - subsys = realpath(tmp, NULL); - return subsys && (strcmp(subsys, "/sys/bus/mdev") == 0); -} - -bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp) -{ - HostIOMMUDevice *hiod = vbasedev->hiod; - - if (!hiod) { - return true; - } - - return HOST_IOMMU_DEVICE_GET_CLASS(hiod)->realize(hiod, vbasedev, errp); -} - -VFIODevice *vfio_get_vfio_device(Object *obj) -{ - if (object_dynamic_cast(obj, TYPE_VFIO_PCI)) { - return &VFIO_PCI(obj)->vbasedev; - } else { - return NULL; - } -} diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'container-base.c', + 'device.c', 'migration.c', 'migration-multifd.c', 'cpr.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_disconnect_container(int fd) "close container->fd=%d" vfio_put_group(int fd) "close group->fd=%d" vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" vfio_put_base_device(int fd) "close vdev->fd=%d" -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 @@ -XXX,XX +XXX,XX @@ iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int functi # container-base.c vfio_reset_handler(void) "" + +# device.c +vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -- 2.48.1 Gather all CPR related declarations into "cpr.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Order file list in meson.build while at it. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/cpr.h | 15 +++++++++++++++ include/hw/vfio/vfio-common.h | 3 --- hw/vfio/container.c | 1 + hw/vfio/cpr.c | 1 + hw/vfio/iommufd.c | 1 + hw/vfio/meson.build | 2 +- 6 files changed, 19 insertions(+), 4 deletions(-) create mode 100644 hw/vfio/cpr.h diff --git a/hw/vfio/cpr.h b/hw/vfio/cpr.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/cpr.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO display + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_CPR_H +#define HW_VFIO_CPR_H + +bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); +void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); + +#endif /* HW_VFIO_CPR_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, void vfio_detach_device(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); -bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); -void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); - typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" #include "helpers.h" +#include "cpr.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/cpr.c +++ b/hw/vfio/cpr.c @@ -XXX,XX +XXX,XX @@ #include "migration/misc.h" #include "qapi/error.h" #include "system/runstate.h" +#include "cpr.h" static int vfio_cpr_reboot_notifier(NotifierWithReturn *notifier, MigrationEvent *e, Error **errp) diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "migration.h" #include "iommufd.h" #include "helpers.h" +#include "cpr.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'container-base.c', + 'cpr.c', 'device.c', 'migration.c', 'migration-multifd.c', - 'cpr.c', 'region.c', )) system_ss.add(when: ['CONFIG_VFIO', 'CONFIG_IOMMUFD'], if_true: files( -- 2.48.1 The vfio_kvm_device_add/del_fd() routines opening the VFIO pseudo device are defined in "helpers.c". Move 'vfio_kvm_device_fd' definition there and its declaration into "helpers.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/helpers.h | 2 ++ include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 11 ----------- hw/vfio/helpers.c | 11 +++++++++++ hw/vfio/spapr.c | 2 +- 5 files changed, 14 insertions(+), 13 deletions(-) diff --git a/hw/vfio/helpers.h b/hw/vfio/helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.h +++ b/hw/vfio/helpers.h @@ -XXX,XX +XXX,XX @@ #ifdef CONFIG_LINUX #include <linux/vfio.h> +extern int vfio_kvm_device_fd; + struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); struct vfio_info_cap_header * diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; -extern int vfio_kvm_device_fd; #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); -#ifdef CONFIG_KVM -/* - * We have a single VFIO pseudo device per KVM VM. Once created it lives - * for the life of the VM. Closing the file descriptor only drops our - * reference to it and the device's reference to kvm. Therefore once - * initialized, this file descriptor is only released on QEMU exit and - * we'll re-use it should another vfio device be attached before then. - */ -int vfio_kvm_device_fd = -1; -#endif - /* * Device state interfaces */ diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, return true; } +#ifdef CONFIG_KVM +/* + * We have a single VFIO pseudo device per KVM VM. Once created it lives + * for the life of the VM. Closing the file descriptor only drops our + * reference to it and the device's reference to kvm. Therefore once + * initialized, this file descriptor is only released on QEMU exit and + * we'll re-use it should another vfio device be attached before then. + */ +int vfio_kvm_device_fd = -1; +#endif + int vfio_kvm_device_add_fd(int fd, Error **errp) { #ifdef CONFIG_KVM diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "system/hostmem.h" #include "exec/address-spaces.h" -#include "hw/vfio/vfio-common.h" #include "hw/vfio/vfio-container.h" #include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" #include "qapi/error.h" #include "trace.h" +#include "helpers.h" typedef struct VFIOHostDMAWindow { hwaddr min_iova; -- 2.48.1 'vfio_device_list' is VFIODevice related. Move its definitions into "device.c". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 3 --- hw/vfio/device.c | 3 +++ 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "migration.h" #include "helpers.h" -VFIODeviceList vfio_device_list = - QLIST_HEAD_INITIALIZER(vfio_device_list); - /* * Device state interfaces */ diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ #include "monitor/monitor.h" #include "helpers.h" +VFIODeviceList vfio_device_list = + QLIST_HEAD_INITIALIZER(vfio_device_list); + /* * Common VFIO interrupt disable */ -- 2.48.1 These routines are VFIODevice related. Move their definitions into "device.c". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 37 ------------------------------------- hw/vfio/device.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+), 37 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_global_stop = vfio_listener_log_global_stop, .log_sync = vfio_listener_log_sync, }; - -bool vfio_attach_device(char *name, VFIODevice *vbasedev, - AddressSpace *as, Error **errp) -{ - const VFIOIOMMUClass *ops = - VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_LEGACY)); - HostIOMMUDevice *hiod = NULL; - - if (vbasedev->iommufd) { - ops = VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_IOMMUFD)); - } - - assert(ops); - - - if (!vbasedev->mdev) { - hiod = HOST_IOMMU_DEVICE(object_new(ops->hiod_typename)); - vbasedev->hiod = hiod; - } - - if (!ops->attach_device(name, vbasedev, as, errp)) { - object_unref(hiod); - vbasedev->hiod = NULL; - return false; - } - - return true; -} - -void vfio_detach_device(VFIODevice *vbasedev) -{ - if (!vbasedev->bcontainer) { - return; - } - object_unref(vbasedev->hiod); - VFIO_IOMMU_GET_CLASS(vbasedev->bcontainer)->detach_device(vbasedev); -} diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj) return NULL; } } + +bool vfio_attach_device(char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp) +{ + const VFIOIOMMUClass *ops = + VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_LEGACY)); + HostIOMMUDevice *hiod = NULL; + + if (vbasedev->iommufd) { + ops = VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_IOMMUFD)); + } + + assert(ops); + + + if (!vbasedev->mdev) { + hiod = HOST_IOMMU_DEVICE(object_new(ops->hiod_typename)); + vbasedev->hiod = hiod; + } + + if (!ops->attach_device(name, vbasedev, as, errp)) { + object_unref(hiod); + vbasedev->hiod = NULL; + return false; + } + + return true; +} + +void vfio_detach_device(VFIODevice *vbasedev) +{ + if (!vbasedev->bcontainer) { + return; + } + object_unref(vbasedev->hiod); + VFIO_IOMMU_GET_CLASS(vbasedev->bcontainer)->detach_device(vbasedev); +} -- 2.48.1 File "common.c" has been emptied of most of its definitions by the previous changes and the only definitions left are related to dirty tracking. Rename it to "dirty-tracking.c" and introduce its associated "dirty-tracking.h" header file for the declarations. Cleanup a little the includes while at it. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/dirty-tracking.h | 22 ++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 10 ---------- hw/vfio/container.c | 1 + hw/vfio/{common.c => dirty-tracking.c} | 5 +---- hw/vfio/iommufd.c | 1 + hw/vfio/meson.build | 2 +- hw/vfio/trace-events | 2 +- 7 files changed, 27 insertions(+), 16 deletions(-) create mode 100644 hw/vfio/dirty-tracking.h rename hw/vfio/{common.c => dirty-tracking.c} (99%) diff --git a/hw/vfio/dirty-tracking.h b/hw/vfio/dirty-tracking.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/dirty-tracking.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO dirty page tracking routines + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_DIRTY_TRACKING_H +#define HW_VFIO_DIRTY_TRACKING_H + +extern const MemoryListener vfio_memory_listener; + +bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); +bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); +int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, hwaddr iova, hwaddr size, + Error **errp); +int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, + uint64_t size, ram_addr_t ram_addr, Error **errp); + +#endif /* HW_VFIO_DIRTY_TRACKING_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; -extern const MemoryListener vfio_memory_listener; #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); #endif -bool vfio_devices_all_dirty_tracking_started( - const VFIOContainerBase *bcontainer); -bool -vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp); -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, - uint64_t size, ram_addr_t ram_addr, Error **errp); - /* Returns 0 on success, or a negative errno. */ bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp); void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "hw/vfio/vfio-container.h" #include "helpers.h" #include "cpr.h" +#include "dirty-tracking.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/common.c b/hw/vfio/dirty-tracking.c similarity index 99% rename from hw/vfio/common.c rename to hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> -#ifdef CONFIG_KVM -#include <linux/kvm.h> -#endif #include <linux/vfio.h> #include "hw/vfio/vfio-common.h" #include "hw/vfio/pci.h" -#include "exec/address-spaces.h" #include "exec/memory.h" #include "exec/ram_addr.h" #include "exec/target_page.h" @@ -XXX,XX +XXX,XX @@ #include "system/tpm.h" #include "migration.h" #include "helpers.h" +#include "dirty-tracking.h" /* * Device state interfaces diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "iommufd.h" #include "helpers.h" #include "cpr.h" +#include "dirty-tracking.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ vfio_ss = ss.source_set() vfio_ss.add(files( - 'common.c', + 'dirty-tracking.c', 'container.c', 'helpers.c', )) diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_igd_bdsm_enabled(const char *name, int size) "%s %dMB" vfio_pci_igd_host_bridge_enabled(const char *name) "%s" vfio_pci_igd_lpc_bridge_enabled(const char *name) "%s" -# common.c +# dirty-tracking.c vfio_iommu_map_notify(const char *op, uint64_t iova_start, uint64_t iova_end) "iommu %s @ 0x%"PRIx64" - 0x%"PRIx64 vfio_listener_region_skip(const char *name, uint64_t start, uint64_t end) "SKIPPING %s 0x%"PRIx64" - 0x%"PRIx64 vfio_spapr_group_attach(int groupfd, int tablefd) "Attached groupfd %d to liobn fd %d" -- 2.48.1 This routine is related to VFIO migration. It belongs to "migration.c". While at it, rename it to better reflect the namespace it belongs to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/migration.h | 1 + hw/vfio/dirty-tracking.c | 19 +++++-------------- hw/vfio/migration.c | 7 +++++++ 3 files changed, 13 insertions(+), 14 deletions(-) diff --git a/hw/vfio/migration.h b/hw/vfio/migration.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.h +++ b/hw/vfio/migration.h @@ -XXX,XX +XXX,XX @@ int vfio_migration_set_state(VFIODevice *vbasedev, enum vfio_device_mig_state recover_state, Error **errp); #endif +void vfio_migration_set_error(int ret); #endif /* HW_VFIO_MIGRATION_H */ diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ #include "system/runstate.h" #include "trace.h" #include "qapi/error.h" -#include "migration/misc.h" -#include "migration/qemu-file.h" #include "system/tcg.h" #include "system/tpm.h" #include "migration.h" @@ -XXX,XX +XXX,XX @@ * Device state interfaces */ -static void vfio_set_migration_error(int ret) -{ - if (migration_is_running()) { - migration_file_set_error(ret, NULL); - } -} - static bool vfio_devices_all_device_dirty_tracking_started( const VFIOContainerBase *bcontainer) { @@ -XXX,XX +XXX,XX @@ static void vfio_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) if (iotlb->target_as != &address_space_memory) { error_report("Wrong target AS \"%s\", only system memory is allowed", iotlb->target_as->name ? iotlb->target_as->name : "none"); - vfio_set_migration_error(-EINVAL); + vfio_migration_set_error(-EINVAL); return; } @@ -XXX,XX +XXX,XX @@ static void vfio_iommu_map_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) "0x%"HWADDR_PRIx") = %d (%s)", bcontainer, iova, iotlb->addr_mask + 1, ret, strerror(-ret)); - vfio_set_migration_error(ret); + vfio_migration_set_error(ret); } } out: @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_global_stop(MemoryListener *listener) error_prepend(&local_err, "vfio: Could not stop dirty page tracking - "); error_report_err(local_err); - vfio_set_migration_error(ret); + vfio_migration_set_error(ret); } } @@ -XXX,XX +XXX,XX @@ out_unlock: out: if (ret) { - vfio_set_migration_error(ret); + vfio_migration_set_error(ret); } } @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_sync(MemoryListener *listener, ret = vfio_sync_dirty_bitmap(bcontainer, section, &local_err); if (ret) { error_report_err(local_err); - vfio_set_migration_error(ret); + vfio_migration_set_error(ret); } } } diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ bool vfio_device_state_is_precopy(VFIODevice *vbasedev) return migration->device_state == VFIO_DEVICE_STATE_PRE_COPY || migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; } + +void vfio_migration_set_error(int ret) +{ + if (migration_is_running()) { + migration_file_set_error(ret, NULL); + } +} -- 2.48.1 "hw/vfio/vfio-common.h" has been emptied of most of its declarations by the previous changes and the only declarations left are related to VFIODevice. Rename it to "hw/vfio/vfio-device.h" and make the necessary adjustments. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/migration-multifd.h | 2 +- hw/vfio/pci.h | 2 +- include/hw/s390x/vfio-ccw.h | 2 +- include/hw/vfio/{vfio-common.h => vfio-device.h} | 0 include/hw/vfio/vfio-platform.h | 2 +- backends/iommufd.c | 2 +- hw/vfio/ap.c | 2 +- hw/vfio/ccw.c | 2 +- hw/vfio/container-base.c | 2 +- hw/vfio/container.c | 2 +- hw/vfio/cpr.c | 2 +- hw/vfio/device.c | 2 +- hw/vfio/dirty-tracking.c | 2 +- hw/vfio/helpers.c | 2 +- hw/vfio/iommufd.c | 2 +- hw/vfio/migration-multifd.c | 2 +- hw/vfio/migration.c | 2 +- hw/vfio/region.c | 4 ++-- 18 files changed, 18 insertions(+), 18 deletions(-) rename include/hw/vfio/{vfio-common.h => vfio-device.h} (100%) diff --git a/hw/vfio/migration-multifd.h b/hw/vfio/migration-multifd.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.h +++ b/hw/vfio/migration-multifd.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_MIGRATION_MULTIFD_H #define HW_VFIO_MIGRATION_MULTIFD_H -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" bool vfio_multifd_setup(VFIODevice *vbasedev, bool alloc_multifd, Error **errp); void vfio_multifd_cleanup(VFIODevice *vbasedev); diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "hw/pci/pci_device.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/event_notifier.h" #include "qemu/queue.h" #include "qemu/timer.h" diff --git a/include/hw/s390x/vfio-ccw.h b/include/hw/s390x/vfio-ccw.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/s390x/vfio-ccw.h +++ b/include/hw/s390x/vfio-ccw.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_CCW_H #define HW_VFIO_CCW_H -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/ccw-device.h" #include "qom/object.h" diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-device.h similarity index 100% rename from include/hw/vfio/vfio-common.h rename to include/hw/vfio/vfio-device.h diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-platform.h +++ b/include/hw/vfio/vfio-platform.h @@ -XXX,XX +XXX,XX @@ #define HW_VFIO_VFIO_PLATFORM_H #include "hw/sysbus.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/event_notifier.h" #include "qemu/queue.h" #include "qom/object.h" diff --git a/backends/iommufd.c b/backends/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/backends/iommufd.c +++ b/backends/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "monitor/monitor.h" #include "trace.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include <sys/ioctl.h> #include <linux/iommufd.h> diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -XXX,XX +XXX,XX @@ #include <linux/vfio.h> #include <sys/ioctl.h> #include "qapi/error.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "system/iommufd.h" #include "hw/s390x/ap-device.h" #include "qemu/error-report.h" diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "qapi/error.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "system/iommufd.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/vfio-ccw.h" diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "system/reset.h" #include "hw/vfio/vfio-container-base.h" -#include "hw/vfio/vfio-common.h" /* for vfio_device_list */ +#include "hw/vfio/vfio-device.h" /* for vfio_device_list */ #include "trace.h" static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include <linux/vfio.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "exec/address-spaces.h" #include "exec/memory.h" #include "exec/ram_addr.h" diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/cpr.c +++ b/hw/vfio/cpr.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "migration/misc.h" #include "qapi/error.h" #include "system/runstate.h" diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/pci.h" #include "hw/hw.h" #include "trace.h" diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include <linux/vfio.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/pci.h" #include "exec/memory.h" #include "exec/ram_addr.h" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "system/kvm.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/hw.h" #include "qapi/error.h" #include "helpers.h" diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include <linux/vfio.h> #include <linux/iommufd.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/error-report.h" #include "trace.h" #include "qapi/error.h" diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "qapi/error.h" diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "system/runstate.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "migration/savevm.h" diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> -#include "hw/vfio/vfio-common.h" -#include "hw/vfio/pci.h" +#include "hw/vfio/vfio-region.h" +#include "hw/vfio/vfio-device.h" #include "hw/hw.h" #include "trace.h" #include "qapi/error.h" -- 2.48.1 Rename these routines : vfio_disable_irqindex -> vfio_device_irq_disable vfio_unmask_single_irqindex -> vfio_device_irq_unmask vfio_mask_single_irqindex -> vfio_device_irq_mask vfio_set_irq_signaling -> vfio_device_irq_set_signaling vfio_attach_device -> vfio_device_attach vfio_detach_device -> vfio_device_detach vfio_get_region_info -> vfio_device_get_region_info vfio_get_dev_region_info -> vfio_device_get_region_info_type vfio_has_region_cap -> vfio_device_has_region_cap to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-device.h | 18 ++++++------- hw/vfio/ap.c | 8 +++--- hw/vfio/ccw.c | 18 ++++++------- hw/vfio/container.c | 6 ++--- hw/vfio/device.c | 24 +++++++++--------- hw/vfio/display.c | 2 +- hw/vfio/igd.c | 8 +++--- hw/vfio/pci.c | 48 +++++++++++++++++------------------ hw/vfio/platform.c | 10 ++++---- hw/vfio/region.c | 2 +- hw/vfio/trace-events | 6 ++--- 11 files changed, 75 insertions(+), 75 deletions(-) diff --git a/include/hw/vfio/vfio-device.h b/include/hw/vfio/vfio-device.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-device.h +++ b/include/hw/vfio/vfio-device.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; -void vfio_disable_irqindex(VFIODevice *vbasedev, int index); -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, +void vfio_device_irq_disable(VFIODevice *vbasedev, int index); +void vfio_device_irq_unmask(VFIODevice *vbasedev, int index); +void vfio_device_irq_mask(VFIODevice *vbasedev, int index); +bool vfio_device_irq_set_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); bool vfio_device_is_mdev(VFIODevice *vbasedev); bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp); -bool vfio_attach_device(char *name, VFIODevice *vbasedev, +bool vfio_device_attach(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp); -void vfio_detach_device(VFIODevice *vbasedev); +void vfio_device_detach(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; #ifdef CONFIG_LINUX -int vfio_get_region_info(VFIODevice *vbasedev, int index, +int vfio_device_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info); -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, +int vfio_device_get_region_info_type(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); +bool vfio_device_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); #endif /* Returns 0 on success, or a negative errno. */ diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -XXX,XX +XXX,XX @@ static bool vfio_ap_register_irq_notifier(VFIOAPDevice *vapdev, fd = event_notifier_get_fd(notifier); qemu_set_fd_handler(fd, fd_read, NULL, vapdev); - if (!vfio_set_irq_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, + if (!vfio_device_irq_set_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vapdev); event_notifier_cleanup(notifier); @@ -XXX,XX +XXX,XX @@ static void vfio_ap_unregister_irq_notifier(VFIOAPDevice *vapdev, return; } - if (!vfio_set_irq_signaling(&vapdev->vdev, irq, 0, + if (!vfio_device_irq_set_signaling(&vapdev->vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { warn_reportf_err(err, VFIO_MSG_PREFIX, vapdev->vdev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_ap_realize(DeviceState *dev, Error **errp) return; } - if (!vfio_attach_device(vbasedev->name, vbasedev, + if (!vfio_device_attach(vbasedev->name, vbasedev, &address_space_memory, errp)) { goto error; } @@ -XXX,XX +XXX,XX @@ static void vfio_ap_unrealize(DeviceState *dev) VFIOAPDevice *vapdev = VFIO_AP_DEVICE(dev); vfio_ap_unregister_irq_notifier(vapdev, VFIO_AP_REQ_IRQ_INDEX); - vfio_detach_device(&vapdev->vdev); + vfio_device_detach(&vapdev->vdev); g_free(vapdev->vdev.name); } diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_register_irq_notifier(VFIOCCWDevice *vcdev, fd = event_notifier_get_fd(notifier); qemu_set_fd_handler(fd, fd_read, NULL, vcdev); - if (!vfio_set_irq_signaling(vdev, irq, 0, + if (!vfio_device_irq_set_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vcdev); event_notifier_cleanup(notifier); @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_unregister_irq_notifier(VFIOCCWDevice *vcdev, return; } - if (!vfio_set_irq_signaling(&vcdev->vdev, irq, 0, + if (!vfio_device_irq_set_signaling(&vcdev->vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { warn_reportf_err(err, VFIO_MSG_PREFIX, vcdev->vdev.name); } @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) return false; } - ret = vfio_get_region_info(vdev, VFIO_CCW_CONFIG_REGION_INDEX, &info); + ret = vfio_device_get_region_info(vdev, VFIO_CCW_CONFIG_REGION_INDEX, &info); if (ret) { error_setg_errno(errp, -ret, "vfio: Error getting config info"); return false; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); /* check for the optional async command region */ - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_ASYNC_CMD, &info); if (!ret) { vcdev->async_cmd_region_size = info->size; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); } - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_SCHIB, &info); if (!ret) { vcdev->schib_region_size = info->size; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); } - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_CRW, &info); if (!ret) { @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_realize(DeviceState *dev, Error **errp) goto out_unrealize; } - if (!vfio_attach_device(cdev->mdevid, vbasedev, + if (!vfio_device_attach(cdev->mdevid, vbasedev, &address_space_memory, errp)) { goto out_attach_dev_err; } @@ -XXX,XX +XXX,XX @@ out_irq_notifier_err: out_io_notifier_err: vfio_ccw_put_region(vcdev); out_region_err: - vfio_detach_device(vbasedev); + vfio_device_detach(vbasedev); out_attach_dev_err: g_free(vbasedev->name); out_unrealize: @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_unrealize(DeviceState *dev) vfio_ccw_unregister_irq_notifier(vcdev, VFIO_CCW_CRW_IRQ_INDEX); vfio_ccw_unregister_irq_notifier(vcdev, VFIO_CCW_IO_IRQ_INDEX); vfio_ccw_put_region(vcdev); - vfio_detach_device(&vcdev->vdev); + vfio_device_detach(&vcdev->vdev); g_free(vcdev->vdev.name); if (cdc->unrealize) { diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) } /* - * vfio_attach_device: attach a device to a security context + * vfio_device_attach: attach a device to a security context * @name and @vbasedev->name are likely to be different depending * on the type of the device, hence the need for passing @name */ @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, return false; } - trace_vfio_attach_device(vbasedev->name, groupid); + trace_vfio_device_attach(vbasedev->name, groupid); if (!vfio_device_hiod_realize(vbasedev, errp)) { return false; @@ -XXX,XX +XXX,XX @@ static void vfio_legacy_detach_device(VFIODevice *vbasedev) QLIST_REMOVE(vbasedev, global_next); QLIST_REMOVE(vbasedev, container_next); vbasedev->bcontainer = NULL; - trace_vfio_detach_device(vbasedev->name, group->groupid); + trace_vfio_device_detach(vbasedev->name, group->groupid); vfio_put_base_device(vbasedev); vfio_put_group(group); } diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = /* * Common VFIO interrupt disable */ -void vfio_disable_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_disable(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ void vfio_disable_irqindex(VFIODevice *vbasedev, int index) ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); } -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_unmask(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); } -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_mask(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ static const char *index_to_str(VFIODevice *vbasedev, int index) } } -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, +bool vfio_device_irq_set_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp) { ERRP_GUARD(); @@ -XXX,XX +XXX,XX @@ bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, return false; } -int vfio_get_region_info(VFIODevice *vbasedev, int index, +int vfio_device_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info) { size_t argsz = sizeof(struct vfio_region_info); @@ -XXX,XX +XXX,XX @@ retry: return 0; } -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, +int vfio_device_get_region_info_type(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { int i; @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, struct vfio_info_cap_header *hdr; struct vfio_region_info_cap_type *cap_type; - if (vfio_get_region_info(vbasedev, i, info)) { + if (vfio_device_get_region_info(vbasedev, i, info)) { continue; } @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); - trace_vfio_get_dev_region(vbasedev->name, i, + trace_vfio_device_get_region_info_type(vbasedev->name, i, cap_type->type, cap_type->subtype); if (cap_type->type == type && cap_type->subtype == subtype) { @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, return -ENODEV; } -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) +bool vfio_device_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) { g_autofree struct vfio_region_info *info = NULL; bool ret = false; - if (!vfio_get_region_info(vbasedev, region, &info)) { + if (!vfio_device_get_region_info(vbasedev, region, &info)) { if (vfio_get_region_info_cap(info, cap_type)) { ret = true; } @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj) } } -bool vfio_attach_device(char *name, VFIODevice *vbasedev, +bool vfio_device_attach(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { const VFIOIOMMUClass *ops = @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, return true; } -void vfio_detach_device(VFIODevice *vbasedev) +void vfio_device_detach(VFIODevice *vbasedev) { if (!vbasedev->bcontainer) { return; diff --git a/hw/vfio/display.c b/hw/vfio/display.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/display.c +++ b/hw/vfio/display.c @@ -XXX,XX +XXX,XX @@ static bool vfio_display_edid_init(VFIOPCIDevice *vdev, Error **errp) int fd = vdev->vbasedev.fd; int ret; - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_GFX, VFIO_REGION_SUBTYPE_GFX_EDID, &dpy->edid_info); diff --git a/hw/vfio/igd.c b/hw/vfio/igd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/igd.c +++ b/hw/vfio/igd.c @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_opregion(VFIOPCIDevice *vdev, Error **errp) return false; } - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_OPREGION, &opregion); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_lpc_bridge(VFIOPCIDevice *vdev, Error **errp) * Check whether we have all the vfio device specific regions to * support LPC quirk (added in Linux v4.6). */ - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_LPC_CFG, &lpc); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_lpc_bridge(VFIOPCIDevice *vdev, Error **errp) return false; } - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_HOST_CFG, &host); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_config_quirk(VFIOPCIDevice *vdev, Error **errp) * there's no ROM, there's no point in setting up this quirk. * NB. We only seem to get BIOS ROMs, so UEFI VM would need CSM support. */ - ret = vfio_get_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info(&vdev->vbasedev, VFIO_PCI_ROM_REGION_INDEX, &rom); if ((ret || !rom->size) && !vdev->pdev.romfile) { error_setg(&err, "Device has no ROM"); diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ static void vfio_intx_eoi(VFIODevice *vbasedev) vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); - vfio_unmask_single_irqindex(vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(vbasedev, VFIO_PCI_INTX_IRQ_INDEX); } static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) /* Get to a known interrupt state */ qemu_set_fd_handler(irq_fd, NULL, NULL, vdev); - vfio_mask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_mask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) goto fail_irqfd; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_UNMASK, event_notifier_get_fd(&vdev->intx.unmask), errp)) { @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) } /* Let'em rip */ - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.kvm_accel = true; @@ -XXX,XX +XXX,XX @@ fail_irqfd: event_notifier_cleanup(&vdev->intx.unmask); fail: qemu_set_fd_handler(irq_fd, vfio_intx_interrupt, NULL, vdev); - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); return false; #else return true; @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable_kvm(VFIOPCIDevice *vdev) * Get to a known state, hardware masked, QEMU ready to accept new * interrupts, QEMU IRQ de-asserted. */ - vfio_mask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_mask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable_kvm(VFIOPCIDevice *vdev) vdev->intx.kvm_accel = false; /* If we've missed an event, let it re-fire through QEMU */ - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); trace_vfio_intx_disable_kvm(vdev->vbasedev.name); #endif @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable(VFIOPCIDevice *vdev, Error **errp) fd = event_notifier_get_fd(&vdev->intx.interrupt); qemu_set_fd_handler(fd, vfio_intx_interrupt, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vdev); event_notifier_cleanup(&vdev->intx.interrupt); @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable(VFIOPCIDevice *vdev) timer_del(vdev->intx.mmap_timer); vfio_intx_disable_kvm(vdev); - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); vfio_mmap_set_enabled(vdev, true); @@ -XXX,XX +XXX,XX @@ static int vfio_msix_vector_do_use(PCIDevice *pdev, unsigned int nr, if (!vdev->defer_kvm_irq_routing) { if (vdev->msix->noresize && resizing) { - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); ret = vfio_enable_vectors(vdev, true); if (ret) { error_report("vfio: failed to enable vectors, %d", ret); @@ -XXX,XX +XXX,XX @@ static int vfio_msix_vector_do_use(PCIDevice *pdev, unsigned int nr, fd = event_notifier_get_fd(&vector->interrupt); } - if (!vfio_set_irq_signaling(&vdev->vbasedev, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, nr, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { @@ -XXX,XX +XXX,XX @@ static void vfio_msix_vector_release(PCIDevice *pdev, unsigned int nr) int32_t fd = event_notifier_get_fd(&vector->interrupt); Error *err = NULL; - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, nr, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); @@ -XXX,XX +XXX,XX @@ static void vfio_msix_disable(VFIOPCIDevice *vdev) * Always clear MSI-X IRQ index. A PF device could have enabled * MSI-X with no vectors. See vfio_msix_enable(). */ - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); vfio_msi_disable_common(vdev); if (!vfio_intx_enable(vdev, &err)) { @@ -XXX,XX +XXX,XX @@ static void vfio_msi_disable(VFIOPCIDevice *vdev) { Error *err = NULL; - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSI_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSI_IRQ_INDEX); vfio_msi_disable_common(vdev); vfio_intx_enable(vdev, &err); if (err) { @@ -XXX,XX +XXX,XX @@ static void vfio_pci_load_rom(VFIOPCIDevice *vdev) off_t off = 0; ssize_t bytes; - if (vfio_get_region_info(&vdev->vbasedev, + if (vfio_device_get_region_info(&vdev->vbasedev, VFIO_PCI_ROM_REGION_INDEX, &reg_info)) { error_report("vfio: Error getting ROM info: %m"); return; @@ -XXX,XX +XXX,XX @@ static void vfio_pci_fixup_msix_region(VFIOPCIDevice *vdev) * If the host driver allows mapping of a MSIX data, we are going to * do map the entire BAR and emulate MSIX table on top of that. */ - if (vfio_has_region_cap(&vdev->vbasedev, region->nr, + if (vfio_device_has_region_cap(&vdev->vbasedev, region->nr, VFIO_REGION_INFO_CAP_MSIX_MAPPABLE)) { return; } @@ -XXX,XX +XXX,XX @@ bool vfio_populate_vga(VFIOPCIDevice *vdev, Error **errp) g_autofree struct vfio_region_info *reg_info = NULL; int ret; - ret = vfio_get_region_info(vbasedev, VFIO_PCI_VGA_REGION_INDEX, &reg_info); + ret = vfio_device_get_region_info(vbasedev, VFIO_PCI_VGA_REGION_INDEX, &reg_info); if (ret) { error_setg_errno(errp, -ret, "failed getting region info for VGA region index %d", @@ -XXX,XX +XXX,XX @@ static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp) QLIST_INIT(&vdev->bars[i].quirks); } - ret = vfio_get_region_info(vbasedev, + ret = vfio_device_get_region_info(vbasedev, VFIO_PCI_CONFIG_REGION_INDEX, &reg_info); if (ret) { error_setg_errno(errp, -ret, "failed to get config info"); @@ -XXX,XX +XXX,XX @@ static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp) static void vfio_pci_put_device(VFIOPCIDevice *vdev) { - vfio_detach_device(&vdev->vbasedev); + vfio_device_detach(&vdev->vbasedev); g_free(vdev->vbasedev.name); g_free(vdev->msix); @@ -XXX,XX +XXX,XX @@ static void vfio_register_err_notifier(VFIOPCIDevice *vdev) fd = event_notifier_get_fd(&vdev->err_notifier); qemu_set_fd_handler(fd, vfio_err_notifier_handler, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); qemu_set_fd_handler(fd, NULL, NULL, vdev); @@ -XXX,XX +XXX,XX @@ static void vfio_unregister_err_notifier(VFIOPCIDevice *vdev) return; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_register_req_notifier(VFIOPCIDevice *vdev) fd = event_notifier_get_fd(&vdev->req_notifier); qemu_set_fd_handler(fd, vfio_req_notifier_handler, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); qemu_set_fd_handler(fd, NULL, NULL, vdev); @@ -XXX,XX +XXX,XX @@ static void vfio_unregister_req_notifier(VFIOPCIDevice *vdev) return; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_realize(PCIDevice *pdev, Error **errp) name = g_strdup(vbasedev->name); } - if (!vfio_attach_device(name, vbasedev, + if (!vfio_device_attach(name, vbasedev, pci_device_iommu_address_space(pdev), errp)) { goto error; } diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -XXX,XX +XXX,XX @@ static int vfio_set_trigger_eventfd(VFIOINTp *intp, qemu_set_fd_handler(fd, (IOHandler *)handler, NULL, intp); - if (!vfio_set_irq_signaling(vbasedev, intp->pin, 0, + if (!vfio_device_irq_set_signaling(vbasedev, intp->pin, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vbasedev->name); qemu_set_fd_handler(fd, NULL, NULL, NULL); @@ -XXX,XX +XXX,XX @@ static void vfio_platform_eoi(VFIODevice *vbasedev) if (vfio_irq_is_automasked(intp)) { /* unmasks the physical level-sensitive IRQ */ - vfio_unmask_single_irqindex(vbasedev, intp->pin); + vfio_device_irq_unmask(vbasedev, intp->pin); } /* a single IRQ can be active at a time */ @@ -XXX,XX +XXX,XX @@ static int vfio_set_resample_eventfd(VFIOINTp *intp) Error *err = NULL; qemu_set_fd_handler(fd, NULL, NULL, NULL); - if (!vfio_set_irq_signaling(vbasedev, intp->pin, 0, + if (!vfio_device_irq_set_signaling(vbasedev, intp->pin, 0, VFIO_IRQ_SET_ACTION_UNMASK, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vbasedev->name); return -EINVAL; @@ -XXX,XX +XXX,XX @@ static bool vfio_base_device_init(VFIODevice *vbasedev, Error **errp) return false; } - if (!vfio_attach_device(vbasedev->name, vbasedev, + if (!vfio_device_attach(vbasedev->name, vbasedev, &address_space_memory, errp)) { return false; } @@ -XXX,XX +XXX,XX @@ static bool vfio_base_device_init(VFIODevice *vbasedev, Error **errp) return true; } - vfio_detach_device(vbasedev); + vfio_device_detach(vbasedev); return false; } diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, g_autofree struct vfio_region_info *info = NULL; int ret; - ret = vfio_get_region_info(vbasedev, index, &info); + ret = vfio_device_get_region_info(vbasedev, index, &info); if (ret) { return ret; } diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int vfio_pci_hot_reset_result(const char *name, const char *result) "%s hot reset: %s" vfio_populate_device_config(const char *name, unsigned long size, unsigned long offset, unsigned long flags) "Device '%s' config: size: 0x%lx, offset: 0x%lx, flags: 0x%lx" vfio_populate_device_get_irq_info_failure(const char *errstr) "VFIO_DEVICE_GET_IRQ_INFO failure: %s" -vfio_attach_device(const char *name, int group_id) " (%s) group %d" -vfio_detach_device(const char *name, int group_id) " (%s) group %d" vfio_mdev(const char *name, bool is_mdev) " (%s) is_mdev %d" vfio_add_ext_cap_dropped(const char *name, uint16_t cap, uint16_t offset) "%s 0x%x@0x%x" vfio_pci_reset(const char *name) " (%s)" @@ -XXX,XX +XXX,XX @@ iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int functi vfio_reset_handler(void) "" # device.c -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_device_get_region_info_type(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_device_attach(const char *name, int group_id) " (%s) group %d" +vfio_device_detach(const char *name, int group_id) " (%s) group %d" -- 2.48.1 vfio_devices_query_dirty_bitmap() is only used in "dirty-tracking.c". Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/dirty-tracking.h | 3 --- hw/vfio/dirty-tracking.c | 2 +- 2 files changed, 1 insertion(+), 4 deletions(-) diff --git a/hw/vfio/dirty-tracking.h b/hw/vfio/dirty-tracking.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.h +++ b/hw/vfio/dirty-tracking.h @@ -XXX,XX +XXX,XX @@ extern const MemoryListener vfio_memory_listener; bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, - Error **errp); int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, +static int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) { VFIODevice *vbasedev; -- 2.48.1 Rename these routines : vfio_put_group -> vfio_group_put vfio_get_group -> vfio_group_get vfio_kvm_device_del_group -> vfio_group_del_kvm_device vfio_kvm_device_add_group -> vfio_group_add_kvm_device vfio_get_device -> vfio_device_get vfio_put_base_device -> vfio_device_put vfio_device_groupid -> vfio_device_get_groupid vfio_connect_container -> vfio_container_connect vfio_disconnect_container -> vfio_container_disconnect to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/container.c | 52 ++++++++++++++++++++++---------------------- hw/vfio/trace-events | 10 +++++---- 2 files changed, 32 insertions(+), 30 deletions(-) diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, return true; } -static void vfio_kvm_device_add_group(VFIOGroup *group) +static void vfio_group_add_kvm_device(VFIOGroup *group) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void vfio_kvm_device_add_group(VFIOGroup *group) } } -static void vfio_kvm_device_del_group(VFIOGroup *group) +static void vfio_group_del_kvm_device(VFIOGroup *group) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_setup(VFIOContainerBase *bcontainer, Error **errp) return true; } -static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, +static bool vfio_container_connect(VFIOGroup *group, AddressSpace *as, Error **errp) { VFIOContainer *container; @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, } group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); - vfio_kvm_device_add_group(group); + vfio_group_add_kvm_device(group); return true; } } @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, goto enable_discards_exit; } - vfio_kvm_device_add_group(group); + vfio_group_add_kvm_device(group); vfio_address_space_insert(space, bcontainer); @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, return true; listener_release_exit: QLIST_REMOVE(group, container_next); - vfio_kvm_device_del_group(group); + vfio_group_del_kvm_device(group); memory_listener_unregister(&bcontainer->listener); if (vioc->release) { vioc->release(bcontainer); @@ -XXX,XX +XXX,XX @@ put_space_exit: return false; } -static void vfio_disconnect_container(VFIOGroup *group) +static void vfio_container_disconnect(VFIOGroup *group) { VFIOContainer *container = group->container; VFIOContainerBase *bcontainer = &container->bcontainer; @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) if (QLIST_EMPTY(&container->group_list)) { VFIOAddressSpace *space = bcontainer->space; - trace_vfio_disconnect_container(container->fd); + trace_vfio_container_disconnect(container->fd); vfio_cpr_unregister_container(bcontainer); close(container->fd); object_unref(container); @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) } } -static VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp) +static VFIOGroup *vfio_group_get(int groupid, AddressSpace *as, Error **errp) { ERRP_GUARD(); VFIOGroup *group; @@ -XXX,XX +XXX,XX @@ static VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp) group->groupid = groupid; QLIST_INIT(&group->device_list); - if (!vfio_connect_container(group, as, errp)) { + if (!vfio_container_connect(group, as, errp)) { error_prepend(errp, "failed to setup container for group %d: ", groupid); goto close_fd_exit; @@ -XXX,XX +XXX,XX @@ free_group_exit: return NULL; } -static void vfio_put_group(VFIOGroup *group) +static void vfio_group_put(VFIOGroup *group) { if (!group || !QLIST_EMPTY(&group->device_list)) { return; @@ -XXX,XX +XXX,XX @@ static void vfio_put_group(VFIOGroup *group) if (!group->ram_block_discard_allowed) { vfio_ram_block_discard_disable(group->container, false); } - vfio_kvm_device_del_group(group); - vfio_disconnect_container(group); + vfio_group_del_kvm_device(group); + vfio_container_disconnect(group); QLIST_REMOVE(group, next); - trace_vfio_put_group(group->fd); + trace_vfio_group_put(group->fd); close(group->fd); g_free(group); } -static bool vfio_get_device(VFIOGroup *group, const char *name, +static bool vfio_device_get(VFIOGroup *group, const char *name, VFIODevice *vbasedev, Error **errp) { g_autofree struct vfio_device_info *info = NULL; @@ -XXX,XX +XXX,XX @@ static bool vfio_get_device(VFIOGroup *group, const char *name, vbasedev->num_regions = info->num_regions; vbasedev->flags = info->flags; - trace_vfio_get_device(name, info->flags, info->num_regions, info->num_irqs); + trace_vfio_device_get(name, info->flags, info->num_regions, info->num_irqs); vbasedev->reset_works = !!(info->flags & VFIO_DEVICE_FLAGS_RESET); return true; } -static void vfio_put_base_device(VFIODevice *vbasedev) +static void vfio_device_put(VFIODevice *vbasedev) { if (!vbasedev->group) { return; } QLIST_REMOVE(vbasedev, next); vbasedev->group = NULL; - trace_vfio_put_base_device(vbasedev->fd); + trace_vfio_device_put(vbasedev->fd); close(vbasedev->fd); } -static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) +static int vfio_device_get_groupid(VFIODevice *vbasedev, Error **errp) { char *tmp, group_path[PATH_MAX]; g_autofree char *group_name = NULL; @@ -XXX,XX +XXX,XX @@ static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { - int groupid = vfio_device_groupid(vbasedev, errp); + int groupid = vfio_device_get_groupid(vbasedev, errp); VFIODevice *vbasedev_iter; VFIOGroup *group; VFIOContainerBase *bcontainer; @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, return false; } - group = vfio_get_group(groupid, as, errp); + group = vfio_group_get(groupid, as, errp); if (!group) { return false; } @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, QLIST_FOREACH(vbasedev_iter, &group->device_list, next) { if (strcmp(vbasedev_iter->name, vbasedev->name) == 0) { error_setg(errp, "device is already attached"); - vfio_put_group(group); + vfio_group_put(group); return false; } } - if (!vfio_get_device(group, name, vbasedev, errp)) { - vfio_put_group(group); + if (!vfio_device_get(group, name, vbasedev, errp)) { + vfio_group_put(group); return false; } @@ -XXX,XX +XXX,XX @@ static void vfio_legacy_detach_device(VFIODevice *vbasedev) QLIST_REMOVE(vbasedev, container_next); vbasedev->bcontainer = NULL; trace_vfio_device_detach(vbasedev->name, group->groupid); - vfio_put_base_device(vbasedev); - vfio_put_group(group); + vfio_device_put(vbasedev); + vfio_group_put(group); } static int vfio_legacy_pci_hot_reset(VFIODevice *vbasedev, bool single) diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_listener_region_add_no_dma_map(const char *name, uint64_t iova, uint64_t si vfio_listener_region_del(uint64_t start, uint64_t end) "region_del 0x%"PRIx64" - 0x%"PRIx64 vfio_device_dirty_tracking_update(uint64_t start, uint64_t end, uint64_t min, uint64_t max) "section 0x%"PRIx64" - 0x%"PRIx64" -> update [0x%"PRIx64" - 0x%"PRIx64"]" vfio_device_dirty_tracking_start(int nr_ranges, uint64_t min32, uint64_t max32, uint64_t min64, uint64_t max64, uint64_t minpci, uint64_t maxpci) "nr_ranges %d 32:[0x%"PRIx64" - 0x%"PRIx64"], 64:[0x%"PRIx64" - 0x%"PRIx64"], pci64:[0x%"PRIx64" - 0x%"PRIx64"]" -vfio_disconnect_container(int fd) "close container->fd=%d" -vfio_put_group(int fd) "close group->fd=%d" -vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" -vfio_put_base_device(int fd) "close vdev->fd=%d" vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 @@ -XXX,XX +XXX,XX @@ iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int functi # container-base.c vfio_reset_handler(void) "" +# container.c +vfio_container_disconnect(int fd) "close container->fd=%d" +vfio_group_put(int fd) "close group->fd=%d" +vfio_device_get(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" +vfio_device_put(int fd) "close vdev->fd=%d" + # device.c vfio_device_get_region_info_type(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" vfio_device_attach(const char *name, int group_id) " (%s) group %d" -- 2.48.1 Rename these routines : vfio_devices_all_device_dirty_tracking_started -> vfio_dirty_tracking_devices_is_started_all vfio_devices_all_dirty_tracking_started -> vfio_dirty_tracking_devices_is_started vfio_devices_all_device_dirty_tracking -> vfio_dirty_tracking_devices_is_supported vfio_devices_dma_logging_start -> vfio_dirty_tracking_devices_dma_logging_start vfio_devices_dma_logging_stop -> vfio_dirty_tracking_devices_dma_logging_stop vfio_devices_query_dirty_bitmap -> vfio_dirty_tracking_devices_query_dirty_bitmap vfio_get_dirty_bitmap -> vfio_dirty_tracking_query_dirty_bitmap to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/dirty-tracking.h | 6 +++--- hw/vfio/container.c | 6 +++--- hw/vfio/dirty-tracking.c | 44 ++++++++++++++++++++-------------------- hw/vfio/trace-events | 2 +- 4 files changed, 29 insertions(+), 29 deletions(-) diff --git a/hw/vfio/dirty-tracking.h b/hw/vfio/dirty-tracking.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.h +++ b/hw/vfio/dirty-tracking.h @@ -XXX,XX +XXX,XX @@ extern const MemoryListener vfio_memory_listener; -bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); -bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, +bool vfio_dirty_tracking_devices_is_started(const VFIOContainerBase *bcontainer); +bool vfio_dirty_tracking_devices_is_supported(const VFIOContainerBase *bcontainer); +int vfio_dirty_tracking_query_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); #endif /* HW_VFIO_DIRTY_TRACKING_H */ diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, int ret; Error *local_err = NULL; - if (iotlb && vfio_devices_all_dirty_tracking_started(bcontainer)) { - if (!vfio_devices_all_device_dirty_tracking(bcontainer) && + if (iotlb && vfio_dirty_tracking_devices_is_started(bcontainer)) { + if (!vfio_dirty_tracking_devices_is_supported(bcontainer) && bcontainer->dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); } @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, } if (need_dirty_sync) { - ret = vfio_get_dirty_bitmap(bcontainer, iova, size, + ret = vfio_dirty_tracking_query_dirty_bitmap(bcontainer, iova, size, iotlb->translated_addr, &local_err); if (ret) { error_report_err(local_err); diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ * Device state interfaces */ -static bool vfio_devices_all_device_dirty_tracking_started( +static bool vfio_dirty_tracking_devices_is_started_all( const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ static bool vfio_devices_all_device_dirty_tracking_started( return true; } -bool vfio_devices_all_dirty_tracking_started( - const VFIOContainerBase *bcontainer) +bool vfio_dirty_tracking_devices_is_started(const VFIOContainerBase *bcontainer) { - return vfio_devices_all_device_dirty_tracking_started(bcontainer) || + return vfio_dirty_tracking_devices_is_started_all(bcontainer) || bcontainer->dirty_pages_started; } @@ -XXX,XX +XXX,XX @@ static bool vfio_log_sync_needed(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; - if (!vfio_devices_all_dirty_tracking_started(bcontainer)) { + if (!vfio_dirty_tracking_devices_is_started(bcontainer)) { return false; } @@ -XXX,XX +XXX,XX @@ static bool vfio_log_sync_needed(const VFIOContainerBase *bcontainer) return true; } -bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer) +bool vfio_dirty_tracking_devices_is_supported(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ static void vfio_dirty_tracking_init(VFIOContainerBase *bcontainer, memory_listener_unregister(&dirty.listener); } -static void vfio_devices_dma_logging_stop(VFIOContainerBase *bcontainer) +static void vfio_dirty_tracking_devices_dma_logging_stop(VFIOContainerBase *bcontainer) { uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature), sizeof(uint64_t))] = {}; @@ -XXX,XX +XXX,XX @@ static void vfio_device_feature_dma_logging_start_destroy( g_free(feature); } -static bool vfio_devices_dma_logging_start(VFIOContainerBase *bcontainer, +static bool vfio_dirty_tracking_devices_dma_logging_start(VFIOContainerBase *bcontainer, Error **errp) { struct vfio_device_feature *feature; @@ -XXX,XX +XXX,XX @@ static bool vfio_devices_dma_logging_start(VFIOContainerBase *bcontainer, out: if (ret) { - vfio_devices_dma_logging_stop(bcontainer); + vfio_dirty_tracking_devices_dma_logging_stop(bcontainer); } vfio_device_feature_dma_logging_start_destroy(feature); @@ -XXX,XX +XXX,XX @@ static bool vfio_listener_log_global_start(MemoryListener *listener, listener); bool ret; - if (vfio_devices_all_device_dirty_tracking(bcontainer)) { - ret = vfio_devices_dma_logging_start(bcontainer, errp); + if (vfio_dirty_tracking_devices_is_supported(bcontainer)) { + ret = vfio_dirty_tracking_devices_dma_logging_start(bcontainer, errp); } else { ret = vfio_container_set_dirty_page_tracking(bcontainer, true, errp) == 0; } @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_global_stop(MemoryListener *listener) Error *local_err = NULL; int ret = 0; - if (vfio_devices_all_device_dirty_tracking(bcontainer)) { - vfio_devices_dma_logging_stop(bcontainer); + if (vfio_dirty_tracking_devices_is_supported(bcontainer)) { + vfio_dirty_tracking_devices_dma_logging_stop(bcontainer); } else { ret = vfio_container_set_dirty_page_tracking(bcontainer, false, &local_err); @@ -XXX,XX +XXX,XX @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } -static int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, +static int vfio_dirty_tracking_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ static int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, return 0; } -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, +int vfio_dirty_tracking_query_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp) { bool all_device_dirty_tracking = - vfio_devices_all_device_dirty_tracking(bcontainer); + vfio_dirty_tracking_devices_is_supported(bcontainer); uint64_t dirty_pages; VFIOBitmap vbmap; int ret; @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, } if (all_device_dirty_tracking) { - ret = vfio_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, - errp); + ret = vfio_dirty_tracking_devices_query_dirty_bitmap(bcontainer, &vbmap, + iova, size, errp); } else { ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size, errp); @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, dirty_pages = cpu_physical_memory_set_dirty_lebitmap(vbmap.bitmap, ram_addr, vbmap.pages); - trace_vfio_get_dirty_bitmap(iova, size, vbmap.size, ram_addr, dirty_pages); + trace_vfio_dirty_tracking_query_dirty_bitmap(iova, size, vbmap.size, ram_addr, + dirty_pages); out: g_free(vbmap.bitmap); @@ -XXX,XX +XXX,XX @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) goto out_unlock; } - ret = vfio_get_dirty_bitmap(bcontainer, iova, iotlb->addr_mask + 1, + ret = vfio_dirty_tracking_query_dirty_bitmap(bcontainer, iova, iotlb->addr_mask + 1, translated_addr, &local_err); if (ret) { error_prepend(&local_err, @@ -XXX,XX +XXX,XX @@ static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, * Sync the whole mapped region (spanning multiple individual mappings) * in one go. */ - ret = vfio_get_dirty_bitmap(vrdl->bcontainer, iova, size, ram_addr, + ret = vfio_dirty_tracking_query_dirty_bitmap(vrdl->bcontainer, iova, size, ram_addr, &local_err); if (ret) { error_report_err(local_err); @@ -XXX,XX +XXX,XX @@ static int vfio_sync_dirty_bitmap(VFIOContainerBase *bcontainer, ram_addr = memory_region_get_ram_addr(section->mr) + section->offset_within_region; - return vfio_get_dirty_bitmap(bcontainer, + return vfio_dirty_tracking_query_dirty_bitmap(bcontainer, REAL_HOST_PAGE_ALIGN(section->offset_within_address_space), int128_get64(section->size), ram_addr, errp); } diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_listener_region_del(uint64_t start, uint64_t end) "region_del 0x%"PRIx64" - vfio_device_dirty_tracking_update(uint64_t start, uint64_t end, uint64_t min, uint64_t max) "section 0x%"PRIx64" - 0x%"PRIx64" -> update [0x%"PRIx64" - 0x%"PRIx64"]" vfio_device_dirty_tracking_start(int nr_ranges, uint64_t min32, uint64_t max32, uint64_t min64, uint64_t max64, uint64_t minpci, uint64_t maxpci) "nr_ranges %d 32:[0x%"PRIx64" - 0x%"PRIx64"], 64:[0x%"PRIx64" - 0x%"PRIx64"], pci64:[0x%"PRIx64" - 0x%"PRIx64"]" vfio_legacy_dma_unmap_overflow_workaround(void) "" -vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +vfio_dirty_tracking_query_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 # region.c -- 2.48.1 This hides the MemoryListener implementation and makes the code common to both IOMMU backends, legacy and IOMMUFD. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/dirty-tracking.h | 4 ++-- hw/vfio/container.c | 11 +++-------- hw/vfio/dirty-tracking.c | 21 ++++++++++++++++++++- hw/vfio/iommufd.c | 9 ++------- 4 files changed, 27 insertions(+), 18 deletions(-) diff --git a/hw/vfio/dirty-tracking.h b/hw/vfio/dirty-tracking.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.h +++ b/hw/vfio/dirty-tracking.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_DIRTY_TRACKING_H #define HW_VFIO_DIRTY_TRACKING_H -extern const MemoryListener vfio_memory_listener; - bool vfio_dirty_tracking_devices_is_started(const VFIOContainerBase *bcontainer); bool vfio_dirty_tracking_devices_is_supported(const VFIOContainerBase *bcontainer); int vfio_dirty_tracking_query_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); +bool vfio_dirty_tracking_register(VFIOContainerBase *bcontainer, Error **errp); +void vfio_dirty_tracking_unregister(VFIOContainerBase *bcontainer); #endif /* HW_VFIO_DIRTY_TRACKING_H */ diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_container_connect(VFIOGroup *group, AddressSpace *as, group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); - bcontainer->listener = vfio_memory_listener; - memory_listener_register(&bcontainer->listener, bcontainer->space->as); - - if (bcontainer->error) { - error_propagate_prepend(errp, bcontainer->error, - "memory listener initialization failed: "); + if (!vfio_dirty_tracking_register(bcontainer, errp)) { goto listener_release_exit; } @@ -XXX,XX +XXX,XX @@ static bool vfio_container_connect(VFIOGroup *group, AddressSpace *as, listener_release_exit: QLIST_REMOVE(group, container_next); vfio_group_del_kvm_device(group); - memory_listener_unregister(&bcontainer->listener); + vfio_dirty_tracking_unregister(bcontainer); if (vioc->release) { vioc->release(bcontainer); } @@ -XXX,XX +XXX,XX @@ static void vfio_container_disconnect(VFIOGroup *group) * group. */ if (QLIST_EMPTY(&container->group_list)) { - memory_listener_unregister(&bcontainer->listener); + vfio_dirty_tracking_unregister(bcontainer); if (vioc->release) { vioc->release(bcontainer); } diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_sync(MemoryListener *listener, } } -const MemoryListener vfio_memory_listener = { +static const MemoryListener vfio_memory_listener = { .name = "vfio", .region_add = vfio_listener_region_add, .region_del = vfio_listener_region_del, @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_global_stop = vfio_listener_log_global_stop, .log_sync = vfio_listener_log_sync, }; + +bool vfio_dirty_tracking_register(VFIOContainerBase *bcontainer, Error **errp) +{ + bcontainer->listener = vfio_memory_listener; + memory_listener_register(&bcontainer->listener, bcontainer->space->as); + + if (bcontainer->error) { + error_propagate_prepend(errp, bcontainer->error, + "memory listener initialization failed: "); + return false; + } + + return true; +} + +void vfio_dirty_tracking_unregister(VFIOContainerBase *bcontainer) +{ + memory_listener_unregister(&bcontainer->listener); +} diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ static void iommufd_cdev_container_destroy(VFIOIOMMUFDContainer *container) if (!QLIST_EMPTY(&bcontainer->device_list)) { return; } - memory_listener_unregister(&bcontainer->listener); + vfio_dirty_tracking_unregister(bcontainer); iommufd_backend_free_id(container->be, container->ioas_id); object_unref(container); } @@ -XXX,XX +XXX,XX @@ static bool iommufd_cdev_attach(const char *name, VFIODevice *vbasedev, bcontainer->pgsizes = qemu_real_host_page_size(); } - bcontainer->listener = vfio_memory_listener; - memory_listener_register(&bcontainer->listener, bcontainer->space->as); - - if (bcontainer->error) { - error_propagate_prepend(errp, bcontainer->error, - "memory listener initialization failed: "); + if (!vfio_dirty_tracking_register(bcontainer, errp)) { goto err_listener_register; } -- 2.48.1 Rename some routines to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/dirty-tracking.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/hw/vfio/dirty-tracking.c b/hw/vfio/dirty-tracking.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/dirty-tracking.c +++ b/hw/vfio/dirty-tracking.c @@ -XXX,XX +XXX,XX @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, return 0; } -static void vfio_register_ram_discard_listener(VFIOContainerBase *bcontainer, +static void vfio_ram_discard_register_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); @@ -XXX,XX +XXX,XX @@ static void vfio_register_ram_discard_listener(VFIOContainerBase *bcontainer, } } -static void vfio_unregister_ram_discard_listener(VFIOContainerBase *bcontainer, +static void vfio_ram_discard_unregister_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); @@ -XXX,XX +XXX,XX @@ static void vfio_listener_region_add(MemoryListener *listener, * about changes. */ if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_register_ram_discard_listener(bcontainer, section); + vfio_ram_discard_register_listener(bcontainer, section); return; } @@ -XXX,XX +XXX,XX @@ static void vfio_listener_region_del(MemoryListener *listener, pgmask = (1ULL << ctz64(bcontainer->pgsizes)) - 1; try_unmap = !((iova & pgmask) || (int128_get64(llsize) & pgmask)); } else if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_unregister_ram_discard_listener(bcontainer, section); + vfio_ram_discard_unregister_listener(bcontainer, section); /* Unregistering will trigger an unmap. */ try_unmap = false; } @@ -XXX,XX +XXX,XX @@ out: } } -static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, +static int vfio_ram_discard_query_dirty_bitmap(MemoryRegionSection *section, void *opaque) { const hwaddr size = int128_get64(section->size); @@ -XXX,XX +XXX,XX @@ vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainerBase *bcontainer, * which correspond to populated parts. Replay all populated parts. */ return ram_discard_manager_replay_populated(rdm, section, - vfio_ram_discard_get_dirty_bitmap, + vfio_ram_discard_query_dirty_bitmap, &vrdl); } -- 2.48.1 Hello, Several large extensions were merged in VFIO recently: migration support with dirty tracking, support for different host IOMMU backend devices, multifd support, etc. This adds up to the previous extensions: vfio-platform, AP, CCW. The result is that VFIO is now a subsystem of over +16,000 lines of code : QEMU 2.0 : 3988 total ... QEMU 10.0 : 16607 total Organization is weak, naming inconsistent, the vfio-common.h header file and common.c are quite messy. It's time to address the technical debt before adding new features. This proposal reorganizes some of the VFIO files to isolate features, introduces new files and renames services to better reflect the namespace they belong to. This is code reshuffling and there are no intentional functional changes. If more could be done, please propose ! Timing seems right. I have taken care to preserve all existing copyright notices in the file headers. I have added one on behalf of my current employer for newly created files. However, original authors may wish to include their own notices as well. If so, please respond to the patch, and I will update the patch in the next spin or before applying. Here is a (short) list requiring your attention : * include/hw/vfio/vfio-migration.h * hw/vfio/vfio-migration-internal.h Kirti Wankhede and Avihai Horon, NVIDIA * hw/vfio/vfio-iommufd.h Joao Martins and Yi Liu, Oracle and Intel * include/hw/vfio/vfio-region.h * hw/vfio/region.c Eric Auger, may be we could reduce the list ? * hw/vfio/vfio-cpr.h Steve Sistare, Oracle I think the remaining new files are correctly covered but I am human, so please review and let me know. I hope we can merge this when the QEMU 10.1 cycle starts and then address the two large series waiting : live update and vfio-user. What next in terms of cleanups : - container.c is quite messy - the pci* files need some love too (add vfio- prefix ?) - improve overall documentation, structs and routines documentation too would be great - isolate all the low level routines (kvm ioctls) into helpers.c to improve build ? - improve build to reduce the number of files built per target - continue the never ending quest of adding 'Error **' parameters, Look for migration_file_set_error, in MemoryListener handlers : vfio_listener_region_add vfio_listener_log_global_stop vfio_listener_log_sync and in callback routines for IOMMU notifiers : vfio_iommu_map_notify vfio_iommu_map_dirty_notify memory_region_iommu_replay() would be a start. - remove vfio-platform (start of QEMU 10.2 cycle ~ September 2025) Thanks, C. Changes in v2: - Dropped vfio_migration_set_error() https://lore.kernel.org/qemu-devel/20250324123315.637827-1-clg@redhat.com/ - Dropped R-b trailers on patches which were modified too much (context changes are ok) - Improved commit logs of patches adding new files - Fixed top comment in header files - Used a 'vfio_migration_' prefix instead of 'vfio_mig_' - Made vfio_migration_add_bytes_transferred() internal - Added extra patch for vfio_device_state_is_running/precopy() - Moved vfio_reset_handler() in device.c - Moved "dirty tracking" related services into container-base.c and improved naming - Introduced listener.* files instead of dirty-tracking.* - Introduced vfio_listener_un/register() routines Cédric Le Goater (37): vfio: Move vfio_mig_active() into migration.c vfio: Rename vfio_reset_bytes_transferred() vfio: Introduce a new header file for external migration services vfio: Make vfio_un/block_multiple_devices_migration() static vfio: Make vfio_viommu_preset() static vfio: Introduce a new header file for internal migration services vfio: Move vfio_device_state_is_running/precopy() into migration.c vfio: Introduce a new header file for VFIOdisplay declarations vfio: Move VFIOHostDMAWindow definition into spapr.c vfio: Introduce a new header file for VFIOIOMMUFD declarations vfio: Introduce new files for VFIORegion definitions and declarations vfio: Introduce a new header file for VFIOcontainer declarations vfio: Make vfio_group_list static vfio: Move VFIOAddressSpace helpers into container-base.c vfio: Move Host IOMMU type declarations into their respective files vfio: Introduce a new header file for helper services vfio: Move vfio_get_info_dma_avail() into helpers.c vfio: Move vfio_kvm_device_add/del_fd() to helpers.c vfio: Move vfio_get_device_info() to helpers.c vfio: Introduce a new file for VFIODevice definitions vfio: Introduce new files for CPR definitions and declarations vfio: Move vfio_kvm_device_fd() into helpers.c vfio: Move vfio_device_list into device.c vfio: Move vfio_de/attach_device() into device.c vfio: Move vfio_reset_handler() into device.c vfio: Move dirty tracking related services into container-base.c vfio: Make vfio_devices_query_dirty_bitmap() static vfio: Make vfio_container_query_dirty_bitmap() static vfio: Rename vfio_devices_all_dirty_tracking_started() vfio: Rename vfio_devices_all_device_dirty_tracking() vfio: Rename vfio_get_dirty_bitmap() vfio: Introduce new files for VFIO MemoryListener vfio: Rename RAM discard related services vfio: Introduce vfio_listener_un/register() routines vfio: Rename vfio-common.h to vfio-device.h vfio: Rename VFIODevice related services vfio: Rename VFIOContainer related services hw/vfio/migration-multifd.h | 2 +- hw/vfio/pci.h | 3 +- hw/vfio/vfio-cpr.h | 15 + hw/vfio/vfio-display.h | 42 ++ hw/vfio/vfio-helpers.h | 35 ++ hw/vfio/vfio-iommufd.h | 34 ++ hw/vfio/vfio-listener.h | 15 + hw/vfio/vfio-migration-internal.h | 74 +++ include/hw/s390x/vfio-ccw.h | 2 +- include/hw/vfio/vfio-common.h | 346 ------------- include/hw/vfio/vfio-container-base.h | 12 +- include/hw/vfio/vfio-container.h | 36 ++ include/hw/vfio/vfio-device.h | 149 ++++++ include/hw/vfio/vfio-migration.h | 16 + include/hw/vfio/vfio-platform.h | 4 +- include/hw/vfio/vfio-region.h | 47 ++ backends/iommufd.c | 2 +- hw/core/sysbus-fdt.c | 1 + hw/ppc/spapr_pci_vfio.c | 6 +- hw/s390x/s390-pci-vfio.c | 3 +- hw/vfio/ap.c | 10 +- hw/vfio/ccw.c | 20 +- hw/vfio/container-base.c | 192 ++++++- hw/vfio/container.c | 123 ++--- hw/vfio/cpr.c | 3 +- hw/vfio/device.c | 405 +++++++++++++++ hw/vfio/display.c | 4 +- hw/vfio/helpers.c | 702 +++----------------------- hw/vfio/igd.c | 8 +- hw/vfio/iommufd.c | 24 +- hw/vfio/{common.c => listener.c} | 455 +---------------- hw/vfio/migration-multifd.c | 8 +- hw/vfio/migration.c | 111 +++- hw/vfio/pci-quirks.c | 1 + hw/vfio/pci.c | 51 +- hw/vfio/platform.c | 11 +- hw/vfio/region.c | 395 +++++++++++++++ hw/vfio/spapr.c | 10 +- migration/target.c | 8 +- hw/vfio/meson.build | 10 +- hw/vfio/trace-events | 36 +- 41 files changed, 1854 insertions(+), 1577 deletions(-) create mode 100644 hw/vfio/vfio-cpr.h create mode 100644 hw/vfio/vfio-display.h create mode 100644 hw/vfio/vfio-helpers.h create mode 100644 hw/vfio/vfio-iommufd.h create mode 100644 hw/vfio/vfio-listener.h create mode 100644 hw/vfio/vfio-migration-internal.h delete mode 100644 include/hw/vfio/vfio-common.h create mode 100644 include/hw/vfio/vfio-container.h create mode 100644 include/hw/vfio/vfio-device.h create mode 100644 include/hw/vfio/vfio-migration.h create mode 100644 include/hw/vfio/vfio-region.h create mode 100644 hw/vfio/device.c rename hw/vfio/{common.c => listener.c} (76%) create mode 100644 hw/vfio/region.c -- 2.49.0 vfio_mig_active() is part of the VFIO migration API. Move the definitions where VFIO migration is implemented. Reviewed-by: Avihai Horon <avihaih@nvidia.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-2-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 16 ---------------- hw/vfio/migration.c | 16 ++++++++++++++++ 2 files changed, 16 insertions(+), 16 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; * Device state interfaces */ -bool vfio_mig_active(void) -{ - VFIODevice *vbasedev; - - if (QLIST_EMPTY(&vfio_device_list)) { - return false; - } - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->migration_blocker) { - return false; - } - } - return true; -} - static Error *multiple_devices_migration_blocker; /* diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ void vfio_mig_add_bytes_transferred(unsigned long val) qatomic_add(&bytes_transferred, val); } +bool vfio_mig_active(void) +{ + VFIODevice *vbasedev; + + if (QLIST_EMPTY(&vfio_device_list)) { + return false; + } + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->migration_blocker) { + return false; + } + } + return true; +} + /* * Return true when either migration initialized or blocker registered. * Currently only return false when adding blocker fails which will -- 2.49.0 Enforce a 'vfio_mig_' prefix for the VFIO migration API to better reflect the namespace these routines belong to. Reviewed-by: Avihai Horon <avihaih@nvidia.com> Reviewed-by: John Levon <john.levon@nutanix.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-3-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 +- hw/vfio/migration.c | 2 +- migration/target.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); int64_t vfio_mig_bytes_transferred(void); -void vfio_reset_bytes_transferred(void); +void vfio_mig_reset_bytes_transferred(void); void vfio_mig_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ int64_t vfio_mig_bytes_transferred(void) return MIN(qatomic_read(&bytes_transferred), INT64_MAX); } -void vfio_reset_bytes_transferred(void) +void vfio_mig_reset_bytes_transferred(void) { qatomic_set(&bytes_transferred, 0); } diff --git a/migration/target.c b/migration/target.c index XXXXXXX..XXXXXXX 100644 --- a/migration/target.c +++ b/migration/target.c @@ -XXX,XX +XXX,XX @@ void migration_populate_vfio_info(MigrationInfo *info) void migration_reset_vfio_bytes_transferred(void) { - vfio_reset_bytes_transferred(); + vfio_mig_reset_bytes_transferred(); } #else void migration_populate_vfio_info(MigrationInfo *info) -- 2.49.0 The migration core subsystem makes use of the VFIO migration API to collect statistics on the number of bytes transferred. These services are declared in "hw/vfio/vfio-common.h" which also contains VFIO internal declarations. Move the migration declarations into a new header file "hw/vfio/vfio-migration.h" to reduce the exposure of VFIO internals. While at it, use a 'vfio_migration_' prefix for these services. To be noted, vfio_migration_add_bytes_transferred() is a VFIO migration internal service which we will moved in the subsequent patches. Cc: Kirti Wankhede <kwankhede@nvidia.com> Cc: Avihai Horon <avihaih@nvidia.com> Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 5 +---- include/hw/vfio/vfio-migration.h | 16 ++++++++++++++++ hw/vfio/migration-multifd.c | 5 +++-- hw/vfio/migration.c | 11 ++++++----- migration/target.c | 8 ++++---- 5 files changed, 30 insertions(+), 15 deletions(-) create mode 100644 include/hw/vfio/vfio-migration.h diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_mig_active(void); int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); -int64_t vfio_mig_bytes_transferred(void); -void vfio_mig_reset_bytes_transferred(void); -void vfio_mig_add_bytes_transferred(unsigned long val); +void vfio_migration_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/include/hw/vfio/vfio-migration.h b/include/hw/vfio/vfio-migration.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-migration.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO migration interface + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_MIGRATION_H +#define HW_VFIO_VFIO_MIGRATION_H + +bool vfio_migration_active(void); +int64_t vfio_migration_bytes_transferred(void); +void vfio_migration_reset_bytes_transferred(void); + +#endif /* HW_VFIO_VFIO_MIGRATION_H */ diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "qapi/error.h" #include "qemu/bswap.h" @@ -XXX,XX +XXX,XX @@ vfio_save_complete_precopy_thread_config_state(VFIODevice *vbasedev, return false; } - vfio_mig_add_bytes_transferred(packet_len); + vfio_migration_add_bytes_transferred(packet_len); return true; } @@ -XXX,XX +XXX,XX @@ vfio_multifd_save_complete_precopy_thread(SaveLiveCompletePrecopyThreadData *d, goto thread_exit; } - vfio_mig_add_bytes_transferred(packet_size); + vfio_migration_add_bytes_transferred(packet_size); } if (!vfio_save_complete_precopy_thread_config_state(vbasedev, diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include "system/runstate.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "migration/savevm.h" #include "migration/vmstate.h" @@ -XXX,XX +XXX,XX @@ static ssize_t vfio_save_block(QEMUFile *f, VFIOMigration *migration) qemu_put_be64(f, VFIO_MIG_FLAG_DEV_DATA_STATE); qemu_put_be64(f, data_size); qemu_put_buffer(f, migration->data_buffer, data_size); - vfio_mig_add_bytes_transferred(data_size); + vfio_migration_add_bytes_transferred(data_size); trace_vfio_save_block(migration->vbasedev->name, data_size); @@ -XXX,XX +XXX,XX @@ static int vfio_block_migration(VFIODevice *vbasedev, Error *err, Error **errp) /* ---------------------------------------------------------------------- */ -int64_t vfio_mig_bytes_transferred(void) +int64_t vfio_migration_bytes_transferred(void) { return MIN(qatomic_read(&bytes_transferred), INT64_MAX); } -void vfio_mig_reset_bytes_transferred(void) +void vfio_migration_reset_bytes_transferred(void) { qatomic_set(&bytes_transferred, 0); } -void vfio_mig_add_bytes_transferred(unsigned long val) +void vfio_migration_add_bytes_transferred(unsigned long val) { qatomic_add(&bytes_transferred, val); } -bool vfio_mig_active(void) +bool vfio_migration_active(void) { VFIODevice *vbasedev; diff --git a/migration/target.c b/migration/target.c index XXXXXXX..XXXXXXX 100644 --- a/migration/target.c +++ b/migration/target.c @@ -XXX,XX +XXX,XX @@ #include CONFIG_DEVICES #ifdef CONFIG_VFIO -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-migration.h" #endif #ifdef CONFIG_VFIO void migration_populate_vfio_info(MigrationInfo *info) { - if (vfio_mig_active()) { + if (vfio_migration_active()) { info->vfio = g_malloc0(sizeof(*info->vfio)); - info->vfio->transferred = vfio_mig_bytes_transferred(); + info->vfio->transferred = vfio_migration_bytes_transferred(); } } void migration_reset_vfio_bytes_transferred(void) { - vfio_mig_reset_bytes_transferred(); + vfio_migration_reset_bytes_transferred(); } #else void migration_populate_vfio_info(MigrationInfo *info) -- 2.49.0 Both of these routines are only used in file "migration.c". Move them there. Reviewed-by: Joao Martins <joao.m.martins@oracle.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-5-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/common.c | 62 ----------------------------------- hw/vfio/migration.c | 62 +++++++++++++++++++++++++++++++++++ 3 files changed, 62 insertions(+), 64 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp); -void vfio_unblock_multiple_devices_migration(void); bool vfio_viommu_preset(VFIODevice *vbasedev); void vfio_migration_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "qapi/error.h" #include "migration/misc.h" -#include "migration/blocker.h" #include "migration/qemu-file.h" #include "system/tcg.h" #include "system/tpm.h" @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; * Device state interfaces */ -static Error *multiple_devices_migration_blocker; - -/* - * Multiple devices migration is allowed only if all devices support P2P - * migration. Single device migration is allowed regardless of P2P migration - * support. - */ -static bool vfio_multiple_devices_migration_is_supported(void) -{ - VFIODevice *vbasedev; - unsigned int device_num = 0; - bool all_support_p2p = true; - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->migration) { - device_num++; - - if (!(vbasedev->migration->mig_flags & VFIO_MIGRATION_P2P)) { - all_support_p2p = false; - } - } - } - - return all_support_p2p || device_num <= 1; -} - -int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp) -{ - int ret; - - if (vfio_multiple_devices_migration_is_supported()) { - return 0; - } - - if (vbasedev->enable_migration == ON_OFF_AUTO_ON) { - error_setg(errp, "Multiple VFIO devices migration is supported only if " - "all of them support P2P migration"); - return -EINVAL; - } - - if (multiple_devices_migration_blocker) { - return 0; - } - - error_setg(&multiple_devices_migration_blocker, - "Multiple VFIO devices migration is supported only if all of " - "them support P2P migration"); - ret = migrate_add_blocker_normal(&multiple_devices_migration_blocker, errp); - - return ret; -} - -void vfio_unblock_multiple_devices_migration(void) -{ - if (!multiple_devices_migration_blocker || - !vfio_multiple_devices_migration_is_supported()) { - return; - } - - migrate_del_blocker(&multiple_devices_migration_blocker); -} bool vfio_viommu_preset(VFIODevice *vbasedev) { diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ static int vfio_migration_init(VFIODevice *vbasedev) return 0; } +static Error *multiple_devices_migration_blocker; + +/* + * Multiple devices migration is allowed only if all devices support P2P + * migration. Single device migration is allowed regardless of P2P migration + * support. + */ +static bool vfio_multiple_devices_migration_is_supported(void) +{ + VFIODevice *vbasedev; + unsigned int device_num = 0; + bool all_support_p2p = true; + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->migration) { + device_num++; + + if (!(vbasedev->migration->mig_flags & VFIO_MIGRATION_P2P)) { + all_support_p2p = false; + } + } + } + + return all_support_p2p || device_num <= 1; +} + +static int vfio_block_multiple_devices_migration(VFIODevice *vbasedev, Error **errp) +{ + int ret; + + if (vfio_multiple_devices_migration_is_supported()) { + return 0; + } + + if (vbasedev->enable_migration == ON_OFF_AUTO_ON) { + error_setg(errp, "Multiple VFIO devices migration is supported only if " + "all of them support P2P migration"); + return -EINVAL; + } + + if (multiple_devices_migration_blocker) { + return 0; + } + + error_setg(&multiple_devices_migration_blocker, + "Multiple VFIO devices migration is supported only if all of " + "them support P2P migration"); + ret = migrate_add_blocker_normal(&multiple_devices_migration_blocker, errp); + + return ret; +} + +static void vfio_unblock_multiple_devices_migration(void) +{ + if (!multiple_devices_migration_blocker || + !vfio_multiple_devices_migration_is_supported()) { + return; + } + + migrate_del_blocker(&multiple_devices_migration_blocker); +} + static void vfio_migration_deinit(VFIODevice *vbasedev) { VFIOMigration *migration = vbasedev->migration; -- 2.49.0 This routine is only used in file "migration.c". Move it there. Reviewed-by: John Levon <john.levon@nutanix.com> Reviewed-by: Joao Martins <joao.m.martins@oracle.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-6-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 5 ----- hw/vfio/migration.c | 5 +++++ 3 files changed, 5 insertions(+), 6 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_viommu_preset(VFIODevice *vbasedev); void vfio_migration_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; */ -bool vfio_viommu_preset(VFIODevice *vbasedev) -{ - return vbasedev->bcontainer->space->as != &address_space_memory; -} - bool vfio_device_state_is_running(VFIODevice *vbasedev) { VFIOMigration *migration = vbasedev->migration; diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ bool vfio_migration_active(void) return true; } +static bool vfio_viommu_preset(VFIODevice *vbasedev) +{ + return vbasedev->bcontainer->space->as != &address_space_memory; +} + /* * Return true when either migration initialized or blocker registered. * Currently only return false when adding blocker fails which will -- 2.49.0 Gather all VFIO migration related declarations into "vfio-migration-internal.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Cc: Kirti Wankhede <kwankhede@nvidia.com> Cc: Avihai Horon <avihaih@nvidia.com> Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-migration-internal.h | 72 +++++++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 52 +--------------------- hw/vfio/common.c | 1 + hw/vfio/migration-multifd.c | 1 + hw/vfio/migration.c | 1 + hw/vfio/pci.c | 1 + 6 files changed, 77 insertions(+), 51 deletions(-) create mode 100644 hw/vfio/vfio-migration-internal.h diff --git a/hw/vfio/vfio-migration-internal.h b/hw/vfio/vfio-migration-internal.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-migration-internal.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO migration + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_MIGRATION_INTERNAL_H +#define HW_VFIO_VFIO_MIGRATION_INTERNAL_H + +#ifdef CONFIG_LINUX +#include <linux/vfio.h> +#endif + +#include "qemu/typedefs.h" +#include "qemu/notify.h" + +/* + * Flags to be used as unique delimiters for VFIO devices in the migration + * stream. These flags are composed as: + * 0xffffffff => MSB 32-bit all 1s + * 0xef10 => Magic ID, represents emulated (virtual) function IO + * 0x0000 => 16-bits reserved for flags + * + * The beginning of state information is marked by _DEV_CONFIG_STATE, + * _DEV_SETUP_STATE, or _DEV_DATA_STATE, respectively. The end of a + * certain state information is marked by _END_OF_STATE. + */ +#define VFIO_MIG_FLAG_END_OF_STATE (0xffffffffef100001ULL) +#define VFIO_MIG_FLAG_DEV_CONFIG_STATE (0xffffffffef100002ULL) +#define VFIO_MIG_FLAG_DEV_SETUP_STATE (0xffffffffef100003ULL) +#define VFIO_MIG_FLAG_DEV_DATA_STATE (0xffffffffef100004ULL) +#define VFIO_MIG_FLAG_DEV_INIT_DATA_SENT (0xffffffffef100005ULL) + +typedef struct VFIODevice VFIODevice; +typedef struct VFIOMultifd VFIOMultifd; + +typedef struct VFIOMigration { + struct VFIODevice *vbasedev; + VMChangeStateEntry *vm_state; + NotifierWithReturn migration_state; + uint32_t device_state; + int data_fd; + void *data_buffer; + size_t data_buffer_size; + uint64_t mig_flags; + uint64_t precopy_init_size; + uint64_t precopy_dirty_size; + bool multifd_transfer; + VFIOMultifd *multifd; + bool initial_data_sent; + + bool event_save_iterate_started; + bool event_precopy_empty_hit; +} VFIOMigration; + +bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); +void vfio_migration_exit(VFIODevice *vbasedev); +int vfio_save_device_config_state(QEMUFile *f, void *opaque, Error **errp); +int vfio_load_device_config_state(QEMUFile *f, void *opaque); + +#ifdef CONFIG_LINUX +int vfio_migration_set_state(VFIODevice *vbasedev, + enum vfio_device_mig_state new_state, + enum vfio_device_mig_state recover_state, + Error **errp); +#endif + +void vfio_migration_add_bytes_transferred(unsigned long val); + +#endif /* HW_VFIO_VFIO_MIGRATION_INTERNAL_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "qemu/queue.h" -#include "qemu/notify.h" #include "ui/console.h" #include "hw/display/ramfb.h" #ifdef CONFIG_LINUX @@ -XXX,XX +XXX,XX @@ #define VFIO_MSG_PREFIX "vfio %s: " -/* - * Flags to be used as unique delimiters for VFIO devices in the migration - * stream. These flags are composed as: - * 0xffffffff => MSB 32-bit all 1s - * 0xef10 => Magic ID, represents emulated (virtual) function IO - * 0x0000 => 16-bits reserved for flags - * - * The beginning of state information is marked by _DEV_CONFIG_STATE, - * _DEV_SETUP_STATE, or _DEV_DATA_STATE, respectively. The end of a - * certain state information is marked by _END_OF_STATE. - */ -#define VFIO_MIG_FLAG_END_OF_STATE (0xffffffffef100001ULL) -#define VFIO_MIG_FLAG_DEV_CONFIG_STATE (0xffffffffef100002ULL) -#define VFIO_MIG_FLAG_DEV_SETUP_STATE (0xffffffffef100003ULL) -#define VFIO_MIG_FLAG_DEV_DATA_STATE (0xffffffffef100004ULL) -#define VFIO_MIG_FLAG_DEV_INIT_DATA_SENT (0xffffffffef100005ULL) - enum { VFIO_DEVICE_TYPE_PCI = 0, VFIO_DEVICE_TYPE_PLATFORM = 1, @@ -XXX,XX +XXX,XX @@ typedef struct VFIORegion { uint8_t nr; /* cache the region number for debug */ } VFIORegion; -typedef struct VFIOMultifd VFIOMultifd; - -typedef struct VFIOMigration { - struct VFIODevice *vbasedev; - VMChangeStateEntry *vm_state; - NotifierWithReturn migration_state; - uint32_t device_state; - int data_fd; - void *data_buffer; - size_t data_buffer_size; - uint64_t mig_flags; - uint64_t precopy_init_size; - uint64_t precopy_dirty_size; - bool multifd_transfer; - VFIOMultifd *multifd; - bool initial_data_sent; - - bool event_save_iterate_started; - bool event_precopy_empty_hit; -} VFIOMigration; - struct VFIOGroup; typedef struct VFIOContainer { @@ -XXX,XX +XXX,XX @@ typedef struct VFIOIOMMUFDContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); typedef struct VFIODeviceOps VFIODeviceOps; +typedef struct VFIOMigration VFIOMigration; typedef struct VFIODevice { QLIST_ENTRY(VFIODevice) next; @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -void vfio_migration_add_bytes_transferred(unsigned long val); bool vfio_device_state_is_running(VFIODevice *vbasedev); bool vfio_device_state_is_precopy(VFIODevice *vbasedev); -int vfio_save_device_config_state(QEMUFile *f, void *opaque, Error **errp); -int vfio_load_device_config_state(QEMUFile *f, void *opaque); - #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info); @@ -XXX,XX +XXX,XX @@ struct vfio_info_cap_header * vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); - -int vfio_migration_set_state(VFIODevice *vbasedev, - enum vfio_device_mig_state new_state, - enum vfio_device_mig_state recover_state, - Error **errp); #endif -bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); -void vfio_migration_exit(VFIODevice *vbasedev); - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); bool vfio_devices_all_dirty_tracking_started( const VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "migration/qemu-file.h" #include "system/tcg.h" #include "system/tpm.h" +#include "vfio-migration-internal.h" VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ #include "io/channel-buffer.h" #include "migration/qemu-file.h" #include "migration-multifd.h" +#include "vfio-migration-internal.h" #include "trace.h" #define VFIO_DEVICE_STATE_CONFIG_STATE (1) diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "trace.h" #include "hw/hw.h" +#include "vfio-migration-internal.h" /* * This is an arbitrary size based on migration of mlx5 devices, where typically diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "migration/blocker.h" #include "migration/qemu-file.h" #include "system/iommufd.h" +#include "vfio-migration-internal.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" -- 2.49.0 These routines are migration related. Move their declaration and implementation under the migration files. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-migration-internal.h | 2 ++ include/hw/vfio/vfio-common.h | 3 --- hw/vfio/common.c | 16 ---------------- hw/vfio/migration.c | 16 ++++++++++++++++ 4 files changed, 18 insertions(+), 19 deletions(-) diff --git a/hw/vfio/vfio-migration-internal.h b/hw/vfio/vfio-migration-internal.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-migration-internal.h +++ b/hw/vfio/vfio-migration-internal.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOMigration { bool vfio_migration_realize(VFIODevice *vbasedev, Error **errp); void vfio_migration_exit(VFIODevice *vbasedev); +bool vfio_device_state_is_running(VFIODevice *vbasedev); +bool vfio_device_state_is_precopy(VFIODevice *vbasedev); int vfio_save_device_config_state(QEMUFile *f, void *opaque, Error **errp); int vfio_load_device_config_state(QEMUFile *f, void *opaque); diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; -bool vfio_device_state_is_running(VFIODevice *vbasedev); -bool vfio_device_state_is_precopy(VFIODevice *vbasedev); - #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_fd = -1; */ -bool vfio_device_state_is_running(VFIODevice *vbasedev) -{ - VFIOMigration *migration = vbasedev->migration; - - return migration->device_state == VFIO_DEVICE_STATE_RUNNING || - migration->device_state == VFIO_DEVICE_STATE_RUNNING_P2P; -} - -bool vfio_device_state_is_precopy(VFIODevice *vbasedev) -{ - VFIOMigration *migration = vbasedev->migration; - - return migration->device_state == VFIO_DEVICE_STATE_PRE_COPY || - migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; -} - static bool vfio_devices_all_device_dirty_tracking_started( const VFIOContainerBase *bcontainer) { diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ void vfio_migration_exit(VFIODevice *vbasedev) migrate_del_blocker(&vbasedev->migration_blocker); } + +bool vfio_device_state_is_running(VFIODevice *vbasedev) +{ + VFIOMigration *migration = vbasedev->migration; + + return migration->device_state == VFIO_DEVICE_STATE_RUNNING || + migration->device_state == VFIO_DEVICE_STATE_RUNNING_P2P; +} + +bool vfio_device_state_is_precopy(VFIODevice *vbasedev) +{ + VFIOMigration *migration = vbasedev->migration; + + return migration->device_state == VFIO_DEVICE_STATE_PRE_COPY || + migration->device_state == VFIO_DEVICE_STATE_PRE_COPY_P2P; +} -- 2.49.0 Gather all VFIOdisplay related declarations into "vfio-display.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-8-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/pci.h | 1 + hw/vfio/vfio-display.h | 41 +++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 28 ------------------------ hw/vfio/display.c | 2 +- 4 files changed, 43 insertions(+), 29 deletions(-) create mode 100644 hw/vfio/vfio-display.h diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -XXX,XX +XXX,XX @@ #include "qemu/timer.h" #include "qom/object.h" #include "system/kvm.h" +#include "vfio-display.h" #define PCI_ANY_ID (~0) diff --git a/hw/vfio/vfio-display.h b/hw/vfio/vfio-display.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-display.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO display + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_DISPLAY_H +#define HW_VFIO_VFIO_DISPLAY_H + +#include "ui/console.h" +#include "hw/display/ramfb.h" + +typedef struct VFIODMABuf { + QemuDmaBuf *buf; + uint32_t pos_x, pos_y, pos_updates; + uint32_t hot_x, hot_y, hot_updates; + int dmabuf_id; + QTAILQ_ENTRY(VFIODMABuf) next; +} VFIODMABuf; + +typedef struct VFIODisplay { + QemuConsole *con; + RAMFBState *ramfb; + struct vfio_region_info *edid_info; + struct vfio_region_gfx_edid *edid_regs; + uint8_t *edid_blob; + QEMUTimer *edid_link_timer; + struct { + VFIORegion buffer; + DisplaySurface *surface; + } region; + struct { + QTAILQ_HEAD(, VFIODMABuf) bufs; + VFIODMABuf *primary; + VFIODMABuf *cursor; + } dmabuf; +} VFIODisplay; + +#endif /* HW_VFIO_VFIO_DISPLAY_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "qemu/queue.h" -#include "ui/console.h" -#include "hw/display/ramfb.h" #ifdef CONFIG_LINUX #include <linux/vfio.h> #endif @@ -XXX,XX +XXX,XX @@ typedef struct VFIOGroup { #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" -typedef struct VFIODMABuf { - QemuDmaBuf *buf; - uint32_t pos_x, pos_y, pos_updates; - uint32_t hot_x, hot_y, hot_updates; - int dmabuf_id; - QTAILQ_ENTRY(VFIODMABuf) next; -} VFIODMABuf; - -typedef struct VFIODisplay { - QemuConsole *con; - RAMFBState *ramfb; - struct vfio_region_info *edid_info; - struct vfio_region_gfx_edid *edid_regs; - uint8_t *edid_blob; - QEMUTimer *edid_link_timer; - struct { - VFIORegion buffer; - DisplaySurface *surface; - } region; - struct { - QTAILQ_HEAD(, VFIODMABuf) bufs; - VFIODMABuf *primary; - VFIODMABuf *cursor; - } dmabuf; -} VFIODisplay; - VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); void vfio_put_address_space(VFIOAddressSpace *space); void vfio_address_space_insert(VFIOAddressSpace *space, diff --git a/hw/vfio/display.c b/hw/vfio/display.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/display.c +++ b/hw/vfio/display.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "hw/display/edid.h" -#include "ui/console.h" #include "qapi/error.h" #include "pci.h" +#include "vfio-display.h" #include "trace.h" #ifndef DRM_PLANE_TYPE_PRIMARY -- 2.49.0 VFIOHostDMAWindow is only used in file "spapr.c". Move it there. Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-9-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 7 ------- hw/vfio/spapr.c | 7 +++++++ 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); -typedef struct VFIOHostDMAWindow { - hwaddr min_iova; - hwaddr max_iova; - uint64_t iova_pgsizes; - QLIST_ENTRY(VFIOHostDMAWindow) hostwin_next; -} VFIOHostDMAWindow; - typedef struct IOMMUFDBackend IOMMUFDBackend; typedef struct VFIOIOASHwpt { diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "trace.h" +typedef struct VFIOHostDMAWindow { + hwaddr min_iova; + hwaddr max_iova; + uint64_t iova_pgsizes; + QLIST_ENTRY(VFIOHostDMAWindow) hostwin_next; +} VFIOHostDMAWindow; + typedef struct VFIOSpaprContainer { VFIOContainer container; MemoryListener prereg_listener; -- 2.49.0 Gather all VFIOIOMMUFD related declarations introduced by commits 5ee3dc7af785 ("vfio/iommufd: Implement the iommufd backend") and 5b1e96e65403 ("vfio/iommufd: Introduce auto domain creation") into "vfio-iommufd.h". This to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Cc: Joao Martins <joao.m.martins@oracle.com> Cc: Yi Liu <yi.l.liu@intel.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-10-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-iommufd.h | 34 ++++++++++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 21 +++------------------ hw/vfio/iommufd.c | 1 + 3 files changed, 38 insertions(+), 18 deletions(-) create mode 100644 hw/vfio/vfio-iommufd.h diff --git a/hw/vfio/vfio-iommufd.h b/hw/vfio/vfio-iommufd.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-iommufd.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO iommufd + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_IOMMUFD_H +#define HW_VFIO_VFIO_IOMMUFD_H + +#include "hw/vfio/vfio-container-base.h" + +typedef struct VFIODevice VFIODevice; + +typedef struct VFIOIOASHwpt { + uint32_t hwpt_id; + uint32_t hwpt_flags; + QLIST_HEAD(, VFIODevice) device_list; + QLIST_ENTRY(VFIOIOASHwpt) next; +} VFIOIOASHwpt; + +typedef struct IOMMUFDBackend IOMMUFDBackend; + +typedef struct VFIOIOMMUFDContainer { + VFIOContainerBase bcontainer; + IOMMUFDBackend *be; + uint32_t ioas_id; + QLIST_HEAD(, VFIOIOASHwpt) hwpt_list; +} VFIOIOMMUFDContainer; + +OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); + +#endif /* HW_VFIO_VFIO_IOMMUFD_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOContainer { OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); -typedef struct IOMMUFDBackend IOMMUFDBackend; - -typedef struct VFIOIOASHwpt { - uint32_t hwpt_id; - uint32_t hwpt_flags; - QLIST_HEAD(, VFIODevice) device_list; - QLIST_ENTRY(VFIOIOASHwpt) next; -} VFIOIOASHwpt; - -typedef struct VFIOIOMMUFDContainer { - VFIOContainerBase bcontainer; - IOMMUFDBackend *be; - uint32_t ioas_id; - QLIST_HEAD(, VFIOIOASHwpt) hwpt_list; -} VFIOIOMMUFDContainer; - -OBJECT_DECLARE_SIMPLE_TYPE(VFIOIOMMUFDContainer, VFIO_IOMMU_IOMMUFD); - typedef struct VFIODeviceOps VFIODeviceOps; typedef struct VFIOMigration VFIOMigration; +typedef struct IOMMUFDBackend IOMMUFDBackend; +typedef struct VFIOIOASHwpt VFIOIOASHwpt; + typedef struct VFIODevice { QLIST_ENTRY(VFIODevice) next; QLIST_ENTRY(VFIODevice) container_next; diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/cutils.h" #include "qemu/chardev_open.h" #include "pci.h" +#include "vfio-iommufd.h" static int iommufd_cdev_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) -- 2.49.0 Gather all VFIORegion related declarations and definitions into their own files to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". They were introduced for 'vfio-platform' support in commits db0da029a185 ("vfio: Generalize region support") and a664477db8da ("hw/vfio/pci: Introduce VFIORegion"). To be noted that the 'vfio-platform' devices have been deprecated and will be removed in QEMU 10.2. Until then, make the declarations available externally for 'sysbus-fdt.c'. Cc: Eric Auger <eric.auger@redhat.com> Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-display.h | 1 + include/hw/vfio/vfio-common.h | 32 +-- include/hw/vfio/vfio-platform.h | 2 + include/hw/vfio/vfio-region.h | 47 ++++ hw/core/sysbus-fdt.c | 1 + hw/vfio/helpers.c | 363 ----------------------------- hw/vfio/pci-quirks.c | 1 + hw/vfio/pci.c | 1 + hw/vfio/platform.c | 1 + hw/vfio/region.c | 394 ++++++++++++++++++++++++++++++++ hw/vfio/meson.build | 1 + hw/vfio/trace-events | 16 +- 12 files changed, 459 insertions(+), 401 deletions(-) create mode 100644 include/hw/vfio/vfio-region.h create mode 100644 hw/vfio/region.c diff --git a/hw/vfio/vfio-display.h b/hw/vfio/vfio-display.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-display.h +++ b/hw/vfio/vfio-display.h @@ -XXX,XX +XXX,XX @@ #include "ui/console.h" #include "hw/display/ramfb.h" +#include "hw/vfio/vfio-region.h" typedef struct VFIODMABuf { QemuDmaBuf *buf; diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ enum { VFIO_DEVICE_TYPE_CCW = 2, VFIO_DEVICE_TYPE_AP = 3, }; - -typedef struct VFIOMmap { - MemoryRegion mem; - void *mmap; - off_t offset; - size_t size; -} VFIOMmap; - -typedef struct VFIORegion { - struct VFIODevice *vbasedev; - off_t fd_offset; /* offset of region within device fd */ - MemoryRegion *mem; /* slow, read/write access */ - size_t size; - uint32_t flags; /* VFIO region flags (rd/wr/mmap) */ - uint32_t nr_mmaps; - VFIOMmap *mmaps; - uint8_t nr; /* cache the region number for debug */ -} VFIORegion; - struct VFIOGroup; typedef struct VFIOContainer { @@ -XXX,XX +XXX,XX @@ void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); -void vfio_region_write(void *opaque, hwaddr addr, - uint64_t data, unsigned size); -uint64_t vfio_region_read(void *opaque, - hwaddr addr, unsigned size); -int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, - int index, const char *name); -int vfio_region_mmap(VFIORegion *region); -void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled); -void vfio_region_unmap(VFIORegion *region); -void vfio_region_exit(VFIORegion *region); -void vfio_region_finalize(VFIORegion *region); + void vfio_reset_handler(void *opaque); struct vfio_device_info *vfio_get_device_info(int fd); bool vfio_device_is_mdev(VFIODevice *vbasedev); @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp); bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); -extern const MemoryRegionOps vfio_region_ops; typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIOGroupList vfio_group_list; diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-platform.h +++ b/include/hw/vfio/vfio-platform.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIOINTp { /* function type for user side eventfd handler */ typedef void (*eventfd_user_side_handler_t)(VFIOINTp *intp); +typedef struct VFIORegion VFIORegion; + struct VFIOPlatformDevice { SysBusDevice sbdev; VFIODevice vbasedev; /* not a QOM object */ diff --git a/include/hw/vfio/vfio-region.h b/include/hw/vfio/vfio-region.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-region.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO region + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_REGION_H +#define HW_VFIO_REGION_H + +#include "exec/memory.h" + +typedef struct VFIOMmap { + MemoryRegion mem; + void *mmap; + off_t offset; + size_t size; +} VFIOMmap; + +typedef struct VFIODevice VFIODevice; + +typedef struct VFIORegion { + struct VFIODevice *vbasedev; + off_t fd_offset; /* offset of region within device fd */ + MemoryRegion *mem; /* slow, read/write access */ + size_t size; + uint32_t flags; /* VFIO region flags (rd/wr/mmap) */ + uint32_t nr_mmaps; + VFIOMmap *mmaps; + uint8_t nr; /* cache the region number for debug */ +} VFIORegion; + + +void vfio_region_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size); +uint64_t vfio_region_read(void *opaque, + hwaddr addr, unsigned size); +int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, + int index, const char *name); +int vfio_region_mmap(VFIORegion *region); +void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled); +void vfio_region_unmap(VFIORegion *region); +void vfio_region_exit(VFIORegion *region); +void vfio_region_finalize(VFIORegion *region); + +#endif /* HW_VFIO_REGION_H */ diff --git a/hw/core/sysbus-fdt.c b/hw/core/sysbus-fdt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/sysbus-fdt.c +++ b/hw/core/sysbus-fdt.c @@ -XXX,XX +XXX,XX @@ #include "hw/vfio/vfio-platform.h" #include "hw/vfio/vfio-calxeda-xgmac.h" #include "hw/vfio/vfio-amd-xgbe.h" +#include "hw/vfio/vfio-region.h" #include "hw/display/ramfb.h" #include "hw/uefi/var-service-api.h" #include "hw/arm/fdt.h" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, return false; } -/* - * IO Port/MMIO - Beware of the endians, VFIO is always little endian - */ -void vfio_region_write(void *opaque, hwaddr addr, - uint64_t data, unsigned size) -{ - VFIORegion *region = opaque; - VFIODevice *vbasedev = region->vbasedev; - union { - uint8_t byte; - uint16_t word; - uint32_t dword; - uint64_t qword; - } buf; - - switch (size) { - case 1: - buf.byte = data; - break; - case 2: - buf.word = cpu_to_le16(data); - break; - case 4: - buf.dword = cpu_to_le32(data); - break; - case 8: - buf.qword = cpu_to_le64(data); - break; - default: - hw_error("vfio: unsupported write size, %u bytes", size); - break; - } - - if (pwrite(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { - error_report("%s(%s:region%d+0x%"HWADDR_PRIx", 0x%"PRIx64 - ",%d) failed: %m", - __func__, vbasedev->name, region->nr, - addr, data, size); - } - - trace_vfio_region_write(vbasedev->name, region->nr, addr, data, size); - - /* - * A read or write to a BAR always signals an INTx EOI. This will - * do nothing if not pending (including not in INTx mode). We assume - * that a BAR access is in response to an interrupt and that BAR - * accesses will service the interrupt. Unfortunately, we don't know - * which access will service the interrupt, so we're potentially - * getting quite a few host interrupts per guest interrupt. - */ - vbasedev->ops->vfio_eoi(vbasedev); -} - -uint64_t vfio_region_read(void *opaque, - hwaddr addr, unsigned size) -{ - VFIORegion *region = opaque; - VFIODevice *vbasedev = region->vbasedev; - union { - uint8_t byte; - uint16_t word; - uint32_t dword; - uint64_t qword; - } buf; - uint64_t data = 0; - - if (pread(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { - error_report("%s(%s:region%d+0x%"HWADDR_PRIx", %d) failed: %m", - __func__, vbasedev->name, region->nr, - addr, size); - return (uint64_t)-1; - } - switch (size) { - case 1: - data = buf.byte; - break; - case 2: - data = le16_to_cpu(buf.word); - break; - case 4: - data = le32_to_cpu(buf.dword); - break; - case 8: - data = le64_to_cpu(buf.qword); - break; - default: - hw_error("vfio: unsupported read size, %u bytes", size); - break; - } - - trace_vfio_region_read(vbasedev->name, region->nr, addr, size, data); - - /* Same as write above */ - vbasedev->ops->vfio_eoi(vbasedev); - - return data; -} - -const MemoryRegionOps vfio_region_ops = { - .read = vfio_region_read, - .write = vfio_region_write, - .endianness = DEVICE_LITTLE_ENDIAN, - .valid = { - .min_access_size = 1, - .max_access_size = 8, - }, - .impl = { - .min_access_size = 1, - .max_access_size = 8, - }, -}; - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size) { vbmap->pages = REAL_HOST_PAGE_ALIGN(size) / qemu_real_host_page_size(); @@ -XXX,XX +XXX,XX @@ vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id) return vfio_get_cap((void *)info, info->cap_offset, id); } -static int vfio_setup_region_sparse_mmaps(VFIORegion *region, - struct vfio_region_info *info) -{ - struct vfio_info_cap_header *hdr; - struct vfio_region_info_cap_sparse_mmap *sparse; - int i, j; - - hdr = vfio_get_region_info_cap(info, VFIO_REGION_INFO_CAP_SPARSE_MMAP); - if (!hdr) { - return -ENODEV; - } - - sparse = container_of(hdr, struct vfio_region_info_cap_sparse_mmap, header); - - trace_vfio_region_sparse_mmap_header(region->vbasedev->name, - region->nr, sparse->nr_areas); - - region->mmaps = g_new0(VFIOMmap, sparse->nr_areas); - - for (i = 0, j = 0; i < sparse->nr_areas; i++) { - if (sparse->areas[i].size) { - trace_vfio_region_sparse_mmap_entry(i, sparse->areas[i].offset, - sparse->areas[i].offset + - sparse->areas[i].size - 1); - region->mmaps[j].offset = sparse->areas[i].offset; - region->mmaps[j].size = sparse->areas[i].size; - j++; - } - } - - region->nr_mmaps = j; - region->mmaps = g_realloc(region->mmaps, j * sizeof(VFIOMmap)); - - return 0; -} - -int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, - int index, const char *name) -{ - g_autofree struct vfio_region_info *info = NULL; - int ret; - - ret = vfio_get_region_info(vbasedev, index, &info); - if (ret) { - return ret; - } - - region->vbasedev = vbasedev; - region->flags = info->flags; - region->size = info->size; - region->fd_offset = info->offset; - region->nr = index; - - if (region->size) { - region->mem = g_new0(MemoryRegion, 1); - memory_region_init_io(region->mem, obj, &vfio_region_ops, - region, name, region->size); - - if (!vbasedev->no_mmap && - region->flags & VFIO_REGION_INFO_FLAG_MMAP) { - - ret = vfio_setup_region_sparse_mmaps(region, info); - - if (ret) { - region->nr_mmaps = 1; - region->mmaps = g_new0(VFIOMmap, region->nr_mmaps); - region->mmaps[0].offset = 0; - region->mmaps[0].size = region->size; - } - } - } - - trace_vfio_region_setup(vbasedev->name, index, name, - region->flags, region->fd_offset, region->size); - return 0; -} - -static void vfio_subregion_unmap(VFIORegion *region, int index) -{ - trace_vfio_region_unmap(memory_region_name(&region->mmaps[index].mem), - region->mmaps[index].offset, - region->mmaps[index].offset + - region->mmaps[index].size - 1); - memory_region_del_subregion(region->mem, &region->mmaps[index].mem); - munmap(region->mmaps[index].mmap, region->mmaps[index].size); - object_unparent(OBJECT(&region->mmaps[index].mem)); - region->mmaps[index].mmap = NULL; -} - -int vfio_region_mmap(VFIORegion *region) -{ - int i, ret, prot = 0; - char *name; - - if (!region->mem) { - return 0; - } - - prot |= region->flags & VFIO_REGION_INFO_FLAG_READ ? PROT_READ : 0; - prot |= region->flags & VFIO_REGION_INFO_FLAG_WRITE ? PROT_WRITE : 0; - - for (i = 0; i < region->nr_mmaps; i++) { - size_t align = MIN(1ULL << ctz64(region->mmaps[i].size), 1 * GiB); - void *map_base, *map_align; - - /* - * Align the mmap for more efficient mapping in the kernel. Ideally - * we'd know the PMD and PUD mapping sizes to use as discrete alignment - * intervals, but we don't. As of Linux v6.12, the largest PUD size - * supporting huge pfnmap is 1GiB (ARCH_SUPPORTS_PUD_PFNMAP is only set - * on x86_64). Align by power-of-two size, capped at 1GiB. - * - * NB. qemu_memalign() and friends actually allocate memory, whereas - * the region size here can exceed host memory, therefore we manually - * create an oversized anonymous mapping and clean it up for alignment. - */ - map_base = mmap(0, region->mmaps[i].size + align, PROT_NONE, - MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); - if (map_base == MAP_FAILED) { - ret = -errno; - goto no_mmap; - } - - map_align = (void *)ROUND_UP((uintptr_t)map_base, (uintptr_t)align); - munmap(map_base, map_align - map_base); - munmap(map_align + region->mmaps[i].size, - align - (map_align - map_base)); - - region->mmaps[i].mmap = mmap(map_align, region->mmaps[i].size, prot, - MAP_SHARED | MAP_FIXED, - region->vbasedev->fd, - region->fd_offset + - region->mmaps[i].offset); - if (region->mmaps[i].mmap == MAP_FAILED) { - ret = -errno; - goto no_mmap; - } - - name = g_strdup_printf("%s mmaps[%d]", - memory_region_name(region->mem), i); - memory_region_init_ram_device_ptr(&region->mmaps[i].mem, - memory_region_owner(region->mem), - name, region->mmaps[i].size, - region->mmaps[i].mmap); - g_free(name); - memory_region_add_subregion(region->mem, region->mmaps[i].offset, - &region->mmaps[i].mem); - - trace_vfio_region_mmap(memory_region_name(&region->mmaps[i].mem), - region->mmaps[i].offset, - region->mmaps[i].offset + - region->mmaps[i].size - 1); - } - - return 0; - -no_mmap: - trace_vfio_region_mmap_fault(memory_region_name(region->mem), i, - region->fd_offset + region->mmaps[i].offset, - region->fd_offset + region->mmaps[i].offset + - region->mmaps[i].size - 1, ret); - - region->mmaps[i].mmap = NULL; - - for (i--; i >= 0; i--) { - vfio_subregion_unmap(region, i); - } - - return ret; -} - -void vfio_region_unmap(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - vfio_subregion_unmap(region, i); - } - } -} - -void vfio_region_exit(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - memory_region_del_subregion(region->mem, &region->mmaps[i].mem); - } - } - - trace_vfio_region_exit(region->vbasedev->name, region->nr); -} - -void vfio_region_finalize(VFIORegion *region) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - munmap(region->mmaps[i].mmap, region->mmaps[i].size); - object_unparent(OBJECT(&region->mmaps[i].mem)); - } - } - - object_unparent(OBJECT(region->mem)); - - g_free(region->mem); - g_free(region->mmaps); - - trace_vfio_region_finalize(region->vbasedev->name, region->nr); - - region->mem = NULL; - region->mmaps = NULL; - region->nr_mmaps = 0; - region->size = 0; - region->flags = 0; - region->nr = 0; -} - -void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled) -{ - int i; - - if (!region->mem) { - return; - } - - for (i = 0; i < region->nr_mmaps; i++) { - if (region->mmaps[i].mmap) { - memory_region_set_enabled(&region->mmaps[i].mem, enabled); - } - } - - trace_vfio_region_mmaps_set_enabled(memory_region_name(region->mem), - enabled); -} - int vfio_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info) { diff --git a/hw/vfio/pci-quirks.c b/hw/vfio/pci-quirks.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci-quirks.c +++ b/hw/vfio/pci-quirks.c @@ -XXX,XX +XXX,XX @@ #include "hw/qdev-properties.h" #include "pci.h" #include "pci-quirks.h" +#include "hw/vfio/vfio-region.h" #include "trace.h" /* diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "migration/qemu-file.h" #include "system/iommufd.h" #include "vfio-migration-internal.h" +#include "hw/vfio/vfio-region.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -XXX,XX +XXX,XX @@ #include "hw/platform-bus.h" #include "hw/qdev-properties.h" #include "system/kvm.h" +#include "hw/vfio/vfio-region.h" /* * Functions used whatever the injection method diff --git a/hw/vfio/region.c b/hw/vfio/region.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ +/* + * VFIO regions + * + * Copyright Red Hat, Inc. 2012 + * + * Authors: + * Alex Williamson <alex.williamson@redhat.com> + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + * + * Based on qemu-kvm device-assignment: + * Adapted for KVM by Qumranet. + * Copyright (c) 2007, Neocleus, Alex Novik (alex@neocleus.com) + * Copyright (c) 2007, Neocleus, Guy Zana (guy@neocleus.com) + * Copyright (C) 2008, Qumranet, Amit Shah (amit.shah@qumranet.com) + * Copyright (C) 2008, Red Hat, Amit Shah (amit.shah@redhat.com) + * Copyright (C) 2008, IBM, Muli Ben-Yehuda (muli@il.ibm.com) + */ + +#include "qemu/osdep.h" +#include <sys/ioctl.h> + +#include "hw/vfio/vfio-common.h" +#include "hw/vfio/pci.h" +#include "hw/hw.h" +#include "trace.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/units.h" +#include "monitor/monitor.h" + +/* + * IO Port/MMIO - Beware of the endians, VFIO is always little endian + */ +void vfio_region_write(void *opaque, hwaddr addr, + uint64_t data, unsigned size) +{ + VFIORegion *region = opaque; + VFIODevice *vbasedev = region->vbasedev; + union { + uint8_t byte; + uint16_t word; + uint32_t dword; + uint64_t qword; + } buf; + + switch (size) { + case 1: + buf.byte = data; + break; + case 2: + buf.word = cpu_to_le16(data); + break; + case 4: + buf.dword = cpu_to_le32(data); + break; + case 8: + buf.qword = cpu_to_le64(data); + break; + default: + hw_error("vfio: unsupported write size, %u bytes", size); + break; + } + + if (pwrite(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { + error_report("%s(%s:region%d+0x%"HWADDR_PRIx", 0x%"PRIx64 + ",%d) failed: %m", + __func__, vbasedev->name, region->nr, + addr, data, size); + } + + trace_vfio_region_write(vbasedev->name, region->nr, addr, data, size); + + /* + * A read or write to a BAR always signals an INTx EOI. This will + * do nothing if not pending (including not in INTx mode). We assume + * that a BAR access is in response to an interrupt and that BAR + * accesses will service the interrupt. Unfortunately, we don't know + * which access will service the interrupt, so we're potentially + * getting quite a few host interrupts per guest interrupt. + */ + vbasedev->ops->vfio_eoi(vbasedev); +} + +uint64_t vfio_region_read(void *opaque, + hwaddr addr, unsigned size) +{ + VFIORegion *region = opaque; + VFIODevice *vbasedev = region->vbasedev; + union { + uint8_t byte; + uint16_t word; + uint32_t dword; + uint64_t qword; + } buf; + uint64_t data = 0; + + if (pread(vbasedev->fd, &buf, size, region->fd_offset + addr) != size) { + error_report("%s(%s:region%d+0x%"HWADDR_PRIx", %d) failed: %m", + __func__, vbasedev->name, region->nr, + addr, size); + return (uint64_t)-1; + } + switch (size) { + case 1: + data = buf.byte; + break; + case 2: + data = le16_to_cpu(buf.word); + break; + case 4: + data = le32_to_cpu(buf.dword); + break; + case 8: + data = le64_to_cpu(buf.qword); + break; + default: + hw_error("vfio: unsupported read size, %u bytes", size); + break; + } + + trace_vfio_region_read(vbasedev->name, region->nr, addr, size, data); + + /* Same as write above */ + vbasedev->ops->vfio_eoi(vbasedev); + + return data; +} + +static const MemoryRegionOps vfio_region_ops = { + .read = vfio_region_read, + .write = vfio_region_write, + .endianness = DEVICE_LITTLE_ENDIAN, + .valid = { + .min_access_size = 1, + .max_access_size = 8, + }, + .impl = { + .min_access_size = 1, + .max_access_size = 8, + }, +}; + +static int vfio_setup_region_sparse_mmaps(VFIORegion *region, + struct vfio_region_info *info) +{ + struct vfio_info_cap_header *hdr; + struct vfio_region_info_cap_sparse_mmap *sparse; + int i, j; + + hdr = vfio_get_region_info_cap(info, VFIO_REGION_INFO_CAP_SPARSE_MMAP); + if (!hdr) { + return -ENODEV; + } + + sparse = container_of(hdr, struct vfio_region_info_cap_sparse_mmap, header); + + trace_vfio_region_sparse_mmap_header(region->vbasedev->name, + region->nr, sparse->nr_areas); + + region->mmaps = g_new0(VFIOMmap, sparse->nr_areas); + + for (i = 0, j = 0; i < sparse->nr_areas; i++) { + if (sparse->areas[i].size) { + trace_vfio_region_sparse_mmap_entry(i, sparse->areas[i].offset, + sparse->areas[i].offset + + sparse->areas[i].size - 1); + region->mmaps[j].offset = sparse->areas[i].offset; + region->mmaps[j].size = sparse->areas[i].size; + j++; + } + } + + region->nr_mmaps = j; + region->mmaps = g_realloc(region->mmaps, j * sizeof(VFIOMmap)); + + return 0; +} + +int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, + int index, const char *name) +{ + g_autofree struct vfio_region_info *info = NULL; + int ret; + + ret = vfio_get_region_info(vbasedev, index, &info); + if (ret) { + return ret; + } + + region->vbasedev = vbasedev; + region->flags = info->flags; + region->size = info->size; + region->fd_offset = info->offset; + region->nr = index; + + if (region->size) { + region->mem = g_new0(MemoryRegion, 1); + memory_region_init_io(region->mem, obj, &vfio_region_ops, + region, name, region->size); + + if (!vbasedev->no_mmap && + region->flags & VFIO_REGION_INFO_FLAG_MMAP) { + + ret = vfio_setup_region_sparse_mmaps(region, info); + + if (ret) { + region->nr_mmaps = 1; + region->mmaps = g_new0(VFIOMmap, region->nr_mmaps); + region->mmaps[0].offset = 0; + region->mmaps[0].size = region->size; + } + } + } + + trace_vfio_region_setup(vbasedev->name, index, name, + region->flags, region->fd_offset, region->size); + return 0; +} + +static void vfio_subregion_unmap(VFIORegion *region, int index) +{ + trace_vfio_region_unmap(memory_region_name(&region->mmaps[index].mem), + region->mmaps[index].offset, + region->mmaps[index].offset + + region->mmaps[index].size - 1); + memory_region_del_subregion(region->mem, &region->mmaps[index].mem); + munmap(region->mmaps[index].mmap, region->mmaps[index].size); + object_unparent(OBJECT(&region->mmaps[index].mem)); + region->mmaps[index].mmap = NULL; +} + +int vfio_region_mmap(VFIORegion *region) +{ + int i, ret, prot = 0; + char *name; + + if (!region->mem) { + return 0; + } + + prot |= region->flags & VFIO_REGION_INFO_FLAG_READ ? PROT_READ : 0; + prot |= region->flags & VFIO_REGION_INFO_FLAG_WRITE ? PROT_WRITE : 0; + + for (i = 0; i < region->nr_mmaps; i++) { + size_t align = MIN(1ULL << ctz64(region->mmaps[i].size), 1 * GiB); + void *map_base, *map_align; + + /* + * Align the mmap for more efficient mapping in the kernel. Ideally + * we'd know the PMD and PUD mapping sizes to use as discrete alignment + * intervals, but we don't. As of Linux v6.12, the largest PUD size + * supporting huge pfnmap is 1GiB (ARCH_SUPPORTS_PUD_PFNMAP is only set + * on x86_64). Align by power-of-two size, capped at 1GiB. + * + * NB. qemu_memalign() and friends actually allocate memory, whereas + * the region size here can exceed host memory, therefore we manually + * create an oversized anonymous mapping and clean it up for alignment. + */ + map_base = mmap(0, region->mmaps[i].size + align, PROT_NONE, + MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); + if (map_base == MAP_FAILED) { + ret = -errno; + goto no_mmap; + } + + map_align = (void *)ROUND_UP((uintptr_t)map_base, (uintptr_t)align); + munmap(map_base, map_align - map_base); + munmap(map_align + region->mmaps[i].size, + align - (map_align - map_base)); + + region->mmaps[i].mmap = mmap(map_align, region->mmaps[i].size, prot, + MAP_SHARED | MAP_FIXED, + region->vbasedev->fd, + region->fd_offset + + region->mmaps[i].offset); + if (region->mmaps[i].mmap == MAP_FAILED) { + ret = -errno; + goto no_mmap; + } + + name = g_strdup_printf("%s mmaps[%d]", + memory_region_name(region->mem), i); + memory_region_init_ram_device_ptr(&region->mmaps[i].mem, + memory_region_owner(region->mem), + name, region->mmaps[i].size, + region->mmaps[i].mmap); + g_free(name); + memory_region_add_subregion(region->mem, region->mmaps[i].offset, + &region->mmaps[i].mem); + + trace_vfio_region_mmap(memory_region_name(&region->mmaps[i].mem), + region->mmaps[i].offset, + region->mmaps[i].offset + + region->mmaps[i].size - 1); + } + + return 0; + +no_mmap: + trace_vfio_region_mmap_fault(memory_region_name(region->mem), i, + region->fd_offset + region->mmaps[i].offset, + region->fd_offset + region->mmaps[i].offset + + region->mmaps[i].size - 1, ret); + + region->mmaps[i].mmap = NULL; + + for (i--; i >= 0; i--) { + vfio_subregion_unmap(region, i); + } + + return ret; +} + +void vfio_region_unmap(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + vfio_subregion_unmap(region, i); + } + } +} + +void vfio_region_exit(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + memory_region_del_subregion(region->mem, &region->mmaps[i].mem); + } + } + + trace_vfio_region_exit(region->vbasedev->name, region->nr); +} + +void vfio_region_finalize(VFIORegion *region) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + munmap(region->mmaps[i].mmap, region->mmaps[i].size); + object_unparent(OBJECT(&region->mmaps[i].mem)); + } + } + + object_unparent(OBJECT(region->mem)); + + g_free(region->mem); + g_free(region->mmaps); + + trace_vfio_region_finalize(region->vbasedev->name, region->nr); + + region->mem = NULL; + region->mmaps = NULL; + region->nr_mmaps = 0; + region->size = 0; + region->flags = 0; + region->nr = 0; +} + +void vfio_region_mmaps_set_enabled(VFIORegion *region, bool enabled) +{ + int i; + + if (!region->mem) { + return; + } + + for (i = 0; i < region->nr_mmaps; i++) { + if (region->mmaps[i].mmap) { + memory_region_set_enabled(&region->mmaps[i].mem, enabled); + } + } + + trace_vfio_region_mmaps_set_enabled(memory_region_name(region->mem), + enabled); +} diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'migration.c', 'migration-multifd.c', 'cpr.c', + 'region.c', )) system_ss.add(when: ['CONFIG_VFIO', 'CONFIG_IOMMUFD'], if_true: files( 'iommufd.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_igd_host_bridge_enabled(const char *name) "%s" vfio_pci_igd_lpc_bridge_enabled(const char *name) "%s" # common.c -vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" -vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 vfio_iommu_map_notify(const char *op, uint64_t iova_start, uint64_t iova_end) "iommu %s @ 0x%"PRIx64" - 0x%"PRIx64 vfio_listener_region_skip(const char *name, uint64_t start, uint64_t end) "SKIPPING %s 0x%"PRIx64" - 0x%"PRIx64 vfio_spapr_group_attach(int groupfd, int tablefd) "Attached groupfd %d to liobn fd %d" @@ -XXX,XX +XXX,XX @@ vfio_disconnect_container(int fd) "close container->fd=%d" vfio_put_group(int fd) "close group->fd=%d" vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" vfio_put_base_device(int fd) "close vdev->fd=%d" +vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_legacy_dma_unmap_overflow_workaround(void) "" +vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 +vfio_reset_handler(void) "" + +# region.c +vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" +vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 vfio_region_setup(const char *dev, int index, const char *name, unsigned long flags, unsigned long offset, unsigned long size) "Device %s, region %d \"%s\", flags: 0x%lx, offset: 0x%lx, size: 0x%lx" vfio_region_mmap_fault(const char *name, int index, unsigned long offset, unsigned long size, int fault) "Region %s mmaps[%d], [0x%lx - 0x%lx], fault: %d" vfio_region_mmap(const char *name, unsigned long offset, unsigned long end) "Region %s [0x%lx - 0x%lx]" @@ -XXX,XX +XXX,XX @@ vfio_region_mmaps_set_enabled(const char *name, bool enabled) "Region %s mmaps e vfio_region_unmap(const char *name, unsigned long offset, unsigned long end) "Region %s unmap [0x%lx - 0x%lx]" vfio_region_sparse_mmap_header(const char *name, int index, int nr_areas) "Device %s region %d: %d sparse mmap entries" vfio_region_sparse_mmap_entry(int i, unsigned long start, unsigned long end) "sparse entry %d [0x%lx - 0x%lx]" -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -vfio_legacy_dma_unmap_overflow_workaround(void) "" -vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 -vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 -vfio_reset_handler(void) "" # platform.c vfio_platform_realize(char *name, char *compat) "vfio device %s, compat = %s" -- 2.49.0 Gather all VFIOcontainer related declarations into "hw/vfio/vfio-container.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". These declarations were initially introduced in commit 65501a745dba ("vfio: vfio-pci device assignment driver"). They are made available externally for PPC and s390x. Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-12-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 19 ----------------- include/hw/vfio/vfio-container.h | 36 ++++++++++++++++++++++++++++++++ hw/ppc/spapr_pci_vfio.c | 1 + hw/s390x/s390-pci-vfio.c | 2 +- hw/vfio/container.c | 1 + hw/vfio/spapr.c | 1 + 6 files changed, 40 insertions(+), 20 deletions(-) create mode 100644 include/hw/vfio/vfio-container.h diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ enum { VFIO_DEVICE_TYPE_CCW = 2, VFIO_DEVICE_TYPE_AP = 3, }; -struct VFIOGroup; - -typedef struct VFIOContainer { - VFIOContainerBase bcontainer; - int fd; /* /dev/vfio/vfio, empowered by the attached groups */ - unsigned iommu_type; - QLIST_HEAD(, VFIOGroup) group_list; -} VFIOContainer; - -OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); typedef struct VFIODeviceOps VFIODeviceOps; typedef struct VFIOMigration VFIOMigration; @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; -typedef struct VFIOGroup { - int fd; - int groupid; - VFIOContainer *container; - QLIST_HEAD(, VFIODevice) device_list; - QLIST_ENTRY(VFIOGroup) next; - QLIST_ENTRY(VFIOGroup) container_next; - bool ram_block_discard_allowed; -} VFIOGroup; #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ diff --git a/include/hw/vfio/vfio-container.h b/include/hw/vfio/vfio-container.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/vfio/vfio-container.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO container + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_CONTAINER_H +#define HW_VFIO_CONTAINER_H + +#include "hw/vfio/vfio-container-base.h" + +typedef struct VFIOContainer VFIOContainer; +typedef struct VFIODevice VFIODevice; + +typedef struct VFIOGroup { + int fd; + int groupid; + VFIOContainer *container; + QLIST_HEAD(, VFIODevice) device_list; + QLIST_ENTRY(VFIOGroup) next; + QLIST_ENTRY(VFIOGroup) container_next; + bool ram_block_discard_allowed; +} VFIOGroup; + +typedef struct VFIOContainer { + VFIOContainerBase bcontainer; + int fd; /* /dev/vfio/vfio, empowered by the attached groups */ + unsigned iommu_type; + QLIST_HEAD(, VFIOGroup) group_list; +} VFIOContainer; + +OBJECT_DECLARE_SIMPLE_TYPE(VFIOContainer, VFIO_IOMMU_LEGACY); + +#endif /* HW_VFIO_CONTAINER_H */ diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/pci/msix.h" #include "hw/pci/pci_device.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" #include "qemu/error-report.h" #include CONFIG_DEVICES /* CONFIG_VFIO_PCI */ diff --git a/hw/s390x/s390-pci-vfio.c b/hw/s390x/s390-pci-vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/s390-pci-vfio.c +++ b/hw/s390x/s390-pci-vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/s390x/s390-pci-clp.h" #include "hw/s390x/s390-pci-vfio.h" #include "hw/vfio/pci.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" /* * Get the current DMA available count from vfio. Returns true if vfio is diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "trace.h" #include "qapi/error.h" #include "pci.h" +#include "hw/vfio/vfio-container.h" VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "exec/address-spaces.h" #include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-container.h" #include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" -- 2.49.0 vfio_group_list is only used in file "container.c". Reviewed-by: John Levon <john.levon@nutanix.com> Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-13-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/container.c | 3 ++- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp); bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); -typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; -extern VFIOGroupList vfio_group_list; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; extern int vfio_kvm_device_fd; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" -VFIOGroupList vfio_group_list = +typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; +static VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); static int vfio_ram_block_discard_disable(VFIOContainer *container, bool state) -- 2.49.0 VFIOAddressSpace is a common object used by VFIOContainerBase which is declared in "hw/vfio/vfio-container-base.h". Move the VFIOAddressSpace related services into "container-base.c". While at it, rename : vfio_get_address_space -> vfio_address_space_get vfio_put_address_space -> vfio_address_space_put to better reflect the namespace these routines belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 5 --- include/hw/vfio/vfio-container-base.h | 6 ++++ hw/ppc/spapr_pci_vfio.c | 5 ++- hw/vfio/common.c | 47 ------------------------- hw/vfio/container-base.c | 50 +++++++++++++++++++++++++++ hw/vfio/container.c | 6 ++-- hw/vfio/iommufd.c | 6 ++-- 7 files changed, 64 insertions(+), 61 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" -VFIOAddressSpace *vfio_get_address_space(AddressSpace *as); -void vfio_put_address_space(VFIOAddressSpace *space); -void vfio_address_space_insert(VFIOAddressSpace *space, - VFIOContainerBase *bcontainer); - void vfio_disable_irqindex(VFIODevice *vbasedev, int index); void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ typedef struct VFIORamDiscardListener { QLIST_ENTRY(VFIORamDiscardListener) next; } VFIORamDiscardListener; +VFIOAddressSpace *vfio_address_space_get(AddressSpace *as); +void vfio_address_space_put(VFIOAddressSpace *space); +void vfio_address_space_insert(VFIOAddressSpace *space, + VFIOContainerBase *bcontainer); + int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly); @@ -XXX,XX +XXX,XX @@ struct VFIOIOMMUClass { MemoryRegionSection *section); void (*release)(VFIOContainerBase *bcontainer); }; + #endif /* HW_VFIO_VFIO_CONTAINER_BASE_H */ diff --git a/hw/ppc/spapr_pci_vfio.c b/hw/ppc/spapr_pci_vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ppc/spapr_pci_vfio.c +++ b/hw/ppc/spapr_pci_vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/pci-host/spapr.h" #include "hw/pci/msix.h" #include "hw/pci/pci_device.h" -#include "hw/vfio/vfio-common.h" #include "hw/vfio/vfio-container.h" #include "qemu/error-report.h" #include CONFIG_DEVICES /* CONFIG_VFIO_PCI */ @@ -XXX,XX +XXX,XX @@ static int vfio_eeh_container_op(VFIOContainer *container, uint32_t op) static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) { - VFIOAddressSpace *space = vfio_get_address_space(as); + VFIOAddressSpace *space = vfio_address_space_get(as); VFIOContainerBase *bcontainer = NULL; if (QLIST_EMPTY(&space->containers)) { @@ -XXX,XX +XXX,XX @@ static VFIOContainer *vfio_eeh_as_container(AddressSpace *as) } out: - vfio_put_address_space(space); + vfio_address_space_put(space); return container_of(bcontainer, VFIOContainer, bcontainer); } diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); -static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = - QLIST_HEAD_INITIALIZER(vfio_address_spaces); #ifdef CONFIG_KVM /* @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp) return 0; } -VFIOAddressSpace *vfio_get_address_space(AddressSpace *as) -{ - VFIOAddressSpace *space; - - QLIST_FOREACH(space, &vfio_address_spaces, list) { - if (space->as == as) { - return space; - } - } - - /* No suitable VFIOAddressSpace, create a new one */ - space = g_malloc0(sizeof(*space)); - space->as = as; - QLIST_INIT(&space->containers); - - if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_register_reset(vfio_reset_handler, NULL); - } - - QLIST_INSERT_HEAD(&vfio_address_spaces, space, list); - - return space; -} - -void vfio_put_address_space(VFIOAddressSpace *space) -{ - if (!QLIST_EMPTY(&space->containers)) { - return; - } - - QLIST_REMOVE(space, list); - g_free(space); - - if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_unregister_reset(vfio_reset_handler, NULL); - } -} - -void vfio_address_space_insert(VFIOAddressSpace *space, - VFIOContainerBase *bcontainer) -{ - QLIST_INSERT_HEAD(&space->containers, bcontainer, next); - bcontainer->space = space; -} - struct vfio_device_info *vfio_get_device_info(int fd) { struct vfio_device_info *info; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "qemu/error-report.h" #include "hw/vfio/vfio-container-base.h" +#include "hw/vfio/vfio-common.h" /* vfio_reset_handler */ +#include "system/reset.h" + +static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = + QLIST_HEAD_INITIALIZER(vfio_address_spaces); + +VFIOAddressSpace *vfio_address_space_get(AddressSpace *as) +{ + VFIOAddressSpace *space; + + QLIST_FOREACH(space, &vfio_address_spaces, list) { + if (space->as == as) { + return space; + } + } + + /* No suitable VFIOAddressSpace, create a new one */ + space = g_malloc0(sizeof(*space)); + space->as = as; + QLIST_INIT(&space->containers); + + if (QLIST_EMPTY(&vfio_address_spaces)) { + qemu_register_reset(vfio_reset_handler, NULL); + } + + QLIST_INSERT_HEAD(&vfio_address_spaces, space, list); + + return space; +} + +void vfio_address_space_put(VFIOAddressSpace *space) +{ + if (!QLIST_EMPTY(&space->containers)) { + return; + } + + QLIST_REMOVE(space, list); + g_free(space); + + if (QLIST_EMPTY(&vfio_address_spaces)) { + qemu_unregister_reset(vfio_reset_handler, NULL); + } +} + +void vfio_address_space_insert(VFIOAddressSpace *space, + VFIOContainerBase *bcontainer) +{ + QLIST_INSERT_HEAD(&space->containers, bcontainer, next); + bcontainer->space = space; +} int vfio_container_dma_map(VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, VFIOAddressSpace *space; VFIOIOMMUClass *vioc; - space = vfio_get_address_space(as); + space = vfio_address_space_get(as); /* * VFIO is currently incompatible with discarding of RAM insofar as the @@ -XXX,XX +XXX,XX @@ close_fd_exit: close(fd); put_space_exit: - vfio_put_address_space(space); + vfio_address_space_put(space); return false; } @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) close(container->fd); object_unref(container); - vfio_put_address_space(space); + vfio_address_space_put(space); } } diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ static bool iommufd_cdev_attach(const char *name, VFIODevice *vbasedev, goto err_connect_bind; } - space = vfio_get_address_space(as); + space = vfio_address_space_get(as); /* * The HostIOMMUDevice data from legacy backend is static and doesn't need @@ -XXX,XX +XXX,XX @@ err_discard_disable: err_attach_container: iommufd_cdev_container_destroy(container); err_alloc_ioas: - vfio_put_address_space(space); + vfio_address_space_put(space); iommufd_cdev_unbind_and_disconnect(vbasedev); err_connect_bind: close(vbasedev->fd); @@ -XXX,XX +XXX,XX @@ static void iommufd_cdev_detach(VFIODevice *vbasedev) vfio_cpr_unregister_container(bcontainer); iommufd_cdev_detach_container(vbasedev, container); iommufd_cdev_container_destroy(container); - vfio_put_address_space(space); + vfio_address_space_put(space); iommufd_cdev_unbind_and_disconnect(vbasedev); close(vbasedev->fd); -- 2.49.0 These definitions don't have any use outside of their respective submodules. There is no need to expose them externally. Keep them private. Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-15-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 5 ----- hw/vfio/container.c | 2 ++ hw/vfio/iommufd.c | 3 +++ 3 files changed, 5 insertions(+), 5 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; - -#define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" -#define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ - TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" - void vfio_disable_irqindex(VFIODevice *vbasedev, int index); void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" +#define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" + typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) VFIOGroupList; static VFIOGroupList vfio_group_list = QLIST_HEAD_INITIALIZER(vfio_group_list); diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "vfio-iommufd.h" +#define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ + TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" + static int iommufd_cdev_map(const VFIOContainerBase *bcontainer, hwaddr iova, ram_addr_t size, void *vaddr, bool readonly) { -- 2.49.0 Gather all helper routine declarations into "vfio-helpers.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-16-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-helpers.h | 26 ++++++++++++++++++++++++++ include/hw/vfio/vfio-common.h | 7 ------- hw/s390x/s390-pci-vfio.c | 1 + hw/vfio/common.c | 1 + hw/vfio/container.c | 1 + hw/vfio/helpers.c | 1 + hw/vfio/pci.c | 1 + hw/vfio/region.c | 1 + 8 files changed, 32 insertions(+), 7 deletions(-) create mode 100644 hw/vfio/vfio-helpers.h diff --git a/hw/vfio/vfio-helpers.h b/hw/vfio/vfio-helpers.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-helpers.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO helpers + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_HELPERS_H +#define HW_VFIO_VFIO_HELPERS_H + +#ifdef CONFIG_LINUX +#include <linux/vfio.h> + +struct vfio_info_cap_header * +vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); +struct vfio_info_cap_header * +vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); +struct vfio_info_cap_header * +vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); + +#endif + +int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); + +#endif /* HW_VFIO_VFIO_HELPERS_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_get_region_info(VFIODevice *vbasedev, int index, int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); -struct vfio_info_cap_header * -vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, unsigned int *avail); -struct vfio_info_cap_header * -vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); -struct vfio_info_cap_header * -vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); #endif -int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); bool vfio_devices_all_dirty_tracking_started( const VFIOContainerBase *bcontainer); bool diff --git a/hw/s390x/s390-pci-vfio.c b/hw/s390x/s390-pci-vfio.c index XXXXXXX..XXXXXXX 100644 --- a/hw/s390x/s390-pci-vfio.c +++ b/hw/s390x/s390-pci-vfio.c @@ -XXX,XX +XXX,XX @@ #include "hw/s390x/s390-pci-vfio.h" #include "hw/vfio/pci.h" #include "hw/vfio/vfio-container.h" +#include "hw/vfio/vfio-helpers.h" /* * Get the current DMA available count from vfio. Returns true if vfio is diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "system/tcg.h" #include "system/tpm.h" #include "vfio-migration-internal.h" +#include "vfio-helpers.h" VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "pci.h" #include "hw/vfio/vfio-container.h" +#include "vfio-helpers.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "qemu/units.h" #include "monitor/monitor.h" +#include "vfio-helpers.h" /* * Common VFIO interrupt disable diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ #include "system/iommufd.h" #include "vfio-migration-internal.h" #include "hw/vfio/vfio-region.h" +#include "vfio-helpers.h" #define TYPE_VFIO_PCI_NOHOTPLUG "vfio-pci-nohotplug" diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "qemu/units.h" #include "monitor/monitor.h" +#include "vfio-helpers.h" /* * IO Port/MMIO - Beware of the endians, VFIO is always little endian -- 2.49.0 vfio_get_info_dma_avail() is a low level routine similar to the other routines extracting capabilities from 'struct vfio_iommu_type1_info'. It belongs to file "helpers.c". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-17-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-helpers.h | 5 ++++- include/hw/vfio/vfio-common.h | 2 -- hw/vfio/container.c | 31 ------------------------------- hw/vfio/helpers.c | 31 +++++++++++++++++++++++++++++++ 4 files changed, 35 insertions(+), 34 deletions(-) diff --git a/hw/vfio/vfio-helpers.h b/hw/vfio/vfio-helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-helpers.h +++ b/hw/vfio/vfio-helpers.h @@ -XXX,XX +XXX,XX @@ struct vfio_info_cap_header * vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id); struct vfio_info_cap_header * vfio_get_region_info_cap(struct vfio_region_info *info, uint16_t id); - +struct vfio_info_cap_header * +vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id); +bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, + unsigned int *avail); #endif int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_get_region_info(VFIODevice *vbasedev, int index, int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); -bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, - unsigned int *avail); #endif bool vfio_devices_all_dirty_tracking_started( diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_query_dirty_bitmap(const VFIOContainerBase *bcontainer, return ret; } -static struct vfio_info_cap_header * -vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) -{ - if (!(info->flags & VFIO_IOMMU_INFO_CAPS)) { - return NULL; - } - - return vfio_get_cap((void *)info, info->cap_offset, id); -} - -bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, - unsigned int *avail) -{ - struct vfio_info_cap_header *hdr; - struct vfio_iommu_type1_info_dma_avail *cap; - - /* If the capability cannot be found, assume no DMA limiting */ - hdr = vfio_get_iommu_type1_info_cap(info, - VFIO_IOMMU_TYPE1_INFO_DMA_AVAIL); - if (!hdr) { - return false; - } - - if (avail != NULL) { - cap = (void *) hdr; - *avail = cap->avail; - } - - return true; -} - static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, VFIOContainerBase *bcontainer) { diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ retry: return 0; } +struct vfio_info_cap_header * +vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) +{ + if (!(info->flags & VFIO_IOMMU_INFO_CAPS)) { + return NULL; + } + + return vfio_get_cap((void *)info, info->cap_offset, id); +} + +bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, + unsigned int *avail) +{ + struct vfio_info_cap_header *hdr; + struct vfio_iommu_type1_info_dma_avail *cap; + + /* If the capability cannot be found, assume no DMA limiting */ + hdr = vfio_get_iommu_type1_info_cap(info, + VFIO_IOMMU_TYPE1_INFO_DMA_AVAIL); + if (!hdr) { + return false; + } + + if (avail != NULL) { + cap = (void *) hdr; + *avail = cap->avail; + } + + return true; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { -- 2.49.0 vfio_kvm_device_add/del_fd() are low level routines. Move them with the other helpers. Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-18-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-helpers.h | 3 ++ include/hw/vfio/vfio-common.h | 3 -- hw/vfio/common.c | 58 ---------------------------------- hw/vfio/helpers.c | 59 +++++++++++++++++++++++++++++++++++ hw/vfio/iommufd.c | 1 + hw/vfio/meson.build | 2 +- 6 files changed, 64 insertions(+), 62 deletions(-) diff --git a/hw/vfio/vfio-helpers.h b/hw/vfio/vfio-helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-helpers.h +++ b/hw/vfio/vfio-helpers.h @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); +int vfio_kvm_device_add_fd(int fd, Error **errp); +int vfio_kvm_device_del_fd(int fd, Error **errp); + #endif /* HW_VFIO_VFIO_HELPERS_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, void vfio_detach_device(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); -int vfio_kvm_device_add_fd(int fd, Error **errp); -int vfio_kvm_device_del_fd(int fd, Error **errp); - bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ void vfio_reset_handler(void *opaque) } } -int vfio_kvm_device_add_fd(int fd, Error **errp) -{ -#ifdef CONFIG_KVM - struct kvm_device_attr attr = { - .group = KVM_DEV_VFIO_FILE, - .attr = KVM_DEV_VFIO_FILE_ADD, - .addr = (uint64_t)(unsigned long)&fd, - }; - - if (!kvm_enabled()) { - return 0; - } - - if (vfio_kvm_device_fd < 0) { - struct kvm_create_device cd = { - .type = KVM_DEV_TYPE_VFIO, - }; - - if (kvm_vm_ioctl(kvm_state, KVM_CREATE_DEVICE, &cd)) { - error_setg_errno(errp, errno, "Failed to create KVM VFIO device"); - return -errno; - } - - vfio_kvm_device_fd = cd.fd; - } - - if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { - error_setg_errno(errp, errno, "Failed to add fd %d to KVM VFIO device", - fd); - return -errno; - } -#endif - return 0; -} - -int vfio_kvm_device_del_fd(int fd, Error **errp) -{ -#ifdef CONFIG_KVM - struct kvm_device_attr attr = { - .group = KVM_DEV_VFIO_FILE, - .attr = KVM_DEV_VFIO_FILE_DEL, - .addr = (uint64_t)(unsigned long)&fd, - }; - - if (vfio_kvm_device_fd < 0) { - error_setg(errp, "KVM VFIO device isn't created yet"); - return -EINVAL; - } - - if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { - error_setg_errno(errp, errno, - "Failed to remove fd %d from KVM VFIO device", fd); - return -errno; - } -#endif - return 0; -} - struct vfio_device_info *vfio_get_device_info(int fd) { struct vfio_device_info *info; diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> +#include "system/kvm.h" #include "hw/vfio/vfio-common.h" #include "hw/vfio/pci.h" #include "hw/hw.h" @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, return true; } +int vfio_kvm_device_add_fd(int fd, Error **errp) +{ +#ifdef CONFIG_KVM + struct kvm_device_attr attr = { + .group = KVM_DEV_VFIO_FILE, + .attr = KVM_DEV_VFIO_FILE_ADD, + .addr = (uint64_t)(unsigned long)&fd, + }; + + if (!kvm_enabled()) { + return 0; + } + + if (vfio_kvm_device_fd < 0) { + struct kvm_create_device cd = { + .type = KVM_DEV_TYPE_VFIO, + }; + + if (kvm_vm_ioctl(kvm_state, KVM_CREATE_DEVICE, &cd)) { + error_setg_errno(errp, errno, "Failed to create KVM VFIO device"); + return -errno; + } + + vfio_kvm_device_fd = cd.fd; + } + + if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { + error_setg_errno(errp, errno, "Failed to add fd %d to KVM VFIO device", + fd); + return -errno; + } +#endif + return 0; +} + +int vfio_kvm_device_del_fd(int fd, Error **errp) +{ +#ifdef CONFIG_KVM + struct kvm_device_attr attr = { + .group = KVM_DEV_VFIO_FILE, + .attr = KVM_DEV_VFIO_FILE_DEL, + .addr = (uint64_t)(unsigned long)&fd, + }; + + if (vfio_kvm_device_fd < 0) { + error_setg(errp, "KVM VFIO device isn't created yet"); + return -EINVAL; + } + + if (ioctl(vfio_kvm_device_fd, KVM_SET_DEVICE_ATTR, &attr)) { + error_setg_errno(errp, errno, + "Failed to remove fd %d from KVM VFIO device", fd); + return -errno; + } +#endif + return 0; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/chardev_open.h" #include "pci.h" #include "vfio-iommufd.h" +#include "vfio-helpers.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ vfio_ss = ss.source_set() vfio_ss.add(files( 'common.c', 'container.c', + 'helpers.c', )) vfio_ss.add(when: 'CONFIG_PSERIES', if_true: files('spapr.c')) vfio_ss.add(when: 'CONFIG_VFIO_PCI', if_true: files( @@ -XXX,XX +XXX,XX @@ specific_ss.add_all(when: 'CONFIG_VFIO', if_true: vfio_ss) system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( - 'helpers.c', 'container-base.c', 'migration.c', 'migration-multifd.c', -- 2.49.0 vfio_get_device_info() is a low level routine. Move it with the other helpers. Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-19-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-helpers.h | 1 + include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 24 ------------------------ hw/vfio/helpers.c | 24 ++++++++++++++++++++++++ 4 files changed, 25 insertions(+), 25 deletions(-) diff --git a/hw/vfio/vfio-helpers.h b/hw/vfio/vfio-helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-helpers.h +++ b/hw/vfio/vfio-helpers.h @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, #endif int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size); +struct vfio_device_info *vfio_get_device_info(int fd); int vfio_kvm_device_add_fd(int fd, Error **errp); int vfio_kvm_device_del_fd(int fd, Error **errp); diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); void vfio_reset_handler(void *opaque); -struct vfio_device_info *vfio_get_device_info(int fd); bool vfio_device_is_mdev(VFIODevice *vbasedev); bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp); bool vfio_attach_device(char *name, VFIODevice *vbasedev, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ void vfio_reset_handler(void *opaque) } } -struct vfio_device_info *vfio_get_device_info(int fd) -{ - struct vfio_device_info *info; - uint32_t argsz = sizeof(*info); - - info = g_malloc0(argsz); - -retry: - info->argsz = argsz; - - if (ioctl(fd, VFIO_DEVICE_GET_INFO, info)) { - g_free(info); - return NULL; - } - - if (info->argsz > argsz) { - argsz = info->argsz; - info = g_realloc(info, argsz); - goto retry; - } - - return info; -} - bool vfio_attach_device(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ int vfio_kvm_device_del_fd(int fd, Error **errp) return 0; } +struct vfio_device_info *vfio_get_device_info(int fd) +{ + struct vfio_device_info *info; + uint32_t argsz = sizeof(*info); + + info = g_malloc0(argsz); + +retry: + info->argsz = argsz; + + if (ioctl(fd, VFIO_DEVICE_GET_INFO, info)) { + g_free(info); + return NULL; + } + + if (info->argsz > argsz) { + argsz = info->argsz; + info = g_realloc(info, argsz); + goto retry; + } + + return info; +} + int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { -- 2.49.0 Move all VFIODevice related routines of "helpers.c" into a new "device.c" file. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/device.c | 330 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/helpers.c | 303 --------------------------------------- hw/vfio/meson.build | 1 + hw/vfio/trace-events | 4 +- 4 files changed, 334 insertions(+), 304 deletions(-) create mode 100644 hw/vfio/device.c diff --git a/hw/vfio/device.c b/hw/vfio/device.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ +/* + * VFIO device + * + * Copyright Red Hat, Inc. 2012 + * + * Authors: + * Alex Williamson <alex.williamson@redhat.com> + * + * This work is licensed under the terms of the GNU GPL, version 2. See + * the COPYING file in the top-level directory. + * + * Based on qemu-kvm device-assignment: + * Adapted for KVM by Qumranet. + * Copyright (c) 2007, Neocleus, Alex Novik (alex@neocleus.com) + * Copyright (c) 2007, Neocleus, Guy Zana (guy@neocleus.com) + * Copyright (C) 2008, Qumranet, Amit Shah (amit.shah@qumranet.com) + * Copyright (C) 2008, Red Hat, Amit Shah (amit.shah@redhat.com) + * Copyright (C) 2008, IBM, Muli Ben-Yehuda (muli@il.ibm.com) + */ + +#include "qemu/osdep.h" +#include <sys/ioctl.h> + +#include "hw/vfio/vfio-common.h" +#include "hw/vfio/pci.h" +#include "hw/hw.h" +#include "trace.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/units.h" +#include "monitor/monitor.h" +#include "vfio-helpers.h" + +/* + * Common VFIO interrupt disable + */ +void vfio_disable_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_TRIGGER, + .index = index, + .start = 0, + .count = 0, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_UNMASK, + .index = index, + .start = 0, + .count = 1, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) +{ + struct vfio_irq_set irq_set = { + .argsz = sizeof(irq_set), + .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_MASK, + .index = index, + .start = 0, + .count = 1, + }; + + ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); +} + +static inline const char *action_to_str(int action) +{ + switch (action) { + case VFIO_IRQ_SET_ACTION_MASK: + return "MASK"; + case VFIO_IRQ_SET_ACTION_UNMASK: + return "UNMASK"; + case VFIO_IRQ_SET_ACTION_TRIGGER: + return "TRIGGER"; + default: + return "UNKNOWN ACTION"; + } +} + +static const char *index_to_str(VFIODevice *vbasedev, int index) +{ + if (vbasedev->type != VFIO_DEVICE_TYPE_PCI) { + return NULL; + } + + switch (index) { + case VFIO_PCI_INTX_IRQ_INDEX: + return "INTX"; + case VFIO_PCI_MSI_IRQ_INDEX: + return "MSI"; + case VFIO_PCI_MSIX_IRQ_INDEX: + return "MSIX"; + case VFIO_PCI_ERR_IRQ_INDEX: + return "ERR"; + case VFIO_PCI_REQ_IRQ_INDEX: + return "REQ"; + default: + return NULL; + } +} + +bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, + int action, int fd, Error **errp) +{ + ERRP_GUARD(); + g_autofree struct vfio_irq_set *irq_set = NULL; + int argsz; + const char *name; + int32_t *pfd; + + argsz = sizeof(*irq_set) + sizeof(*pfd); + + irq_set = g_malloc0(argsz); + irq_set->argsz = argsz; + irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | action; + irq_set->index = index; + irq_set->start = subindex; + irq_set->count = 1; + pfd = (int32_t *)&irq_set->data; + *pfd = fd; + + if (!ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set)) { + return true; + } + + error_setg_errno(errp, errno, "VFIO_DEVICE_SET_IRQS failure"); + + name = index_to_str(vbasedev, index); + if (name) { + error_prepend(errp, "%s-%d: ", name, subindex); + } else { + error_prepend(errp, "index %d-%d: ", index, subindex); + } + error_prepend(errp, + "Failed to %s %s eventfd signaling for interrupt ", + fd < 0 ? "tear down" : "set up", action_to_str(action)); + return false; +} + +int vfio_get_region_info(VFIODevice *vbasedev, int index, + struct vfio_region_info **info) +{ + size_t argsz = sizeof(struct vfio_region_info); + + *info = g_malloc0(argsz); + + (*info)->index = index; +retry: + (*info)->argsz = argsz; + + if (ioctl(vbasedev->fd, VFIO_DEVICE_GET_REGION_INFO, *info)) { + g_free(*info); + *info = NULL; + return -errno; + } + + if ((*info)->argsz > argsz) { + argsz = (*info)->argsz; + *info = g_realloc(*info, argsz); + + goto retry; + } + + return 0; +} + +int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, + uint32_t subtype, struct vfio_region_info **info) +{ + int i; + + for (i = 0; i < vbasedev->num_regions; i++) { + struct vfio_info_cap_header *hdr; + struct vfio_region_info_cap_type *cap_type; + + if (vfio_get_region_info(vbasedev, i, info)) { + continue; + } + + hdr = vfio_get_region_info_cap(*info, VFIO_REGION_INFO_CAP_TYPE); + if (!hdr) { + g_free(*info); + continue; + } + + cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); + + trace_vfio_get_dev_region(vbasedev->name, i, + cap_type->type, cap_type->subtype); + + if (cap_type->type == type && cap_type->subtype == subtype) { + return 0; + } + + g_free(*info); + } + + *info = NULL; + return -ENODEV; +} + +bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) +{ + g_autofree struct vfio_region_info *info = NULL; + bool ret = false; + + if (!vfio_get_region_info(vbasedev, region, &info)) { + if (vfio_get_region_info_cap(info, cap_type)) { + ret = true; + } + } + + return ret; +} + +bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp) +{ + ERRP_GUARD(); + struct stat st; + + if (vbasedev->fd < 0) { + if (stat(vbasedev->sysfsdev, &st) < 0) { + error_setg_errno(errp, errno, "no such host device"); + error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); + return false; + } + /* User may specify a name, e.g: VFIO platform device */ + if (!vbasedev->name) { + vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); + } + } else { + if (!vbasedev->iommufd) { + error_setg(errp, "Use FD passing only with iommufd backend"); + return false; + } + /* + * Give a name with fd so any function printing out vbasedev->name + * will not break. + */ + if (!vbasedev->name) { + vbasedev->name = g_strdup_printf("VFIO_FD%d", vbasedev->fd); + } + } + + return true; +} + +void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp) +{ + ERRP_GUARD(); + int fd = monitor_fd_param(monitor_cur(), str, errp); + + if (fd < 0) { + error_prepend(errp, "Could not parse remote object fd %s:", str); + return; + } + vbasedev->fd = fd; +} + +void vfio_device_init(VFIODevice *vbasedev, int type, VFIODeviceOps *ops, + DeviceState *dev, bool ram_discard) +{ + vbasedev->type = type; + vbasedev->ops = ops; + vbasedev->dev = dev; + vbasedev->fd = -1; + + vbasedev->ram_block_discard_allowed = ram_discard; +} + +int vfio_device_get_aw_bits(VFIODevice *vdev) +{ + /* + * iova_ranges is a sorted list. For old kernels that support + * VFIO but not support query of iova ranges, iova_ranges is NULL, + * in this case HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX(64) is returned. + */ + GList *l = g_list_last(vdev->bcontainer->iova_ranges); + + if (l) { + Range *range = l->data; + return range_get_last_bit(range) + 1; + } + + return HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX; +} + +bool vfio_device_is_mdev(VFIODevice *vbasedev) +{ + g_autofree char *subsys = NULL; + g_autofree char *tmp = NULL; + + if (!vbasedev->sysfsdev) { + return false; + } + + tmp = g_strdup_printf("%s/subsystem", vbasedev->sysfsdev); + subsys = realpath(tmp, NULL); + return subsys && (strcmp(subsys, "/sys/bus/mdev") == 0); +} + +bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp) +{ + HostIOMMUDevice *hiod = vbasedev->hiod; + + if (!hiod) { + return true; + } + + return HOST_IOMMU_DEVICE_GET_CLASS(hiod)->realize(hiod, vbasedev, errp); +} + +VFIODevice *vfio_get_vfio_device(Object *obj) +{ + if (object_dynamic_cast(obj, TYPE_VFIO_PCI)) { + return &VFIO_PCI(obj)->vbasedev; + } else { + return NULL; + } +} diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include "system/kvm.h" #include "hw/vfio/vfio-common.h" -#include "hw/vfio/pci.h" #include "hw/hw.h" -#include "trace.h" #include "qapi/error.h" -#include "qemu/error-report.h" -#include "qemu/units.h" -#include "monitor/monitor.h" #include "vfio-helpers.h" -/* - * Common VFIO interrupt disable - */ -void vfio_disable_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_TRIGGER, - .index = index, - .start = 0, - .count = 0, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_UNMASK, - .index = index, - .start = 0, - .count = 1, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) -{ - struct vfio_irq_set irq_set = { - .argsz = sizeof(irq_set), - .flags = VFIO_IRQ_SET_DATA_NONE | VFIO_IRQ_SET_ACTION_MASK, - .index = index, - .start = 0, - .count = 1, - }; - - ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); -} - -static inline const char *action_to_str(int action) -{ - switch (action) { - case VFIO_IRQ_SET_ACTION_MASK: - return "MASK"; - case VFIO_IRQ_SET_ACTION_UNMASK: - return "UNMASK"; - case VFIO_IRQ_SET_ACTION_TRIGGER: - return "TRIGGER"; - default: - return "UNKNOWN ACTION"; - } -} - -static const char *index_to_str(VFIODevice *vbasedev, int index) -{ - if (vbasedev->type != VFIO_DEVICE_TYPE_PCI) { - return NULL; - } - - switch (index) { - case VFIO_PCI_INTX_IRQ_INDEX: - return "INTX"; - case VFIO_PCI_MSI_IRQ_INDEX: - return "MSI"; - case VFIO_PCI_MSIX_IRQ_INDEX: - return "MSIX"; - case VFIO_PCI_ERR_IRQ_INDEX: - return "ERR"; - case VFIO_PCI_REQ_IRQ_INDEX: - return "REQ"; - default: - return NULL; - } -} - -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, - int action, int fd, Error **errp) -{ - ERRP_GUARD(); - g_autofree struct vfio_irq_set *irq_set = NULL; - int argsz; - const char *name; - int32_t *pfd; - - argsz = sizeof(*irq_set) + sizeof(*pfd); - - irq_set = g_malloc0(argsz); - irq_set->argsz = argsz; - irq_set->flags = VFIO_IRQ_SET_DATA_EVENTFD | action; - irq_set->index = index; - irq_set->start = subindex; - irq_set->count = 1; - pfd = (int32_t *)&irq_set->data; - *pfd = fd; - - if (!ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, irq_set)) { - return true; - } - - error_setg_errno(errp, errno, "VFIO_DEVICE_SET_IRQS failure"); - - name = index_to_str(vbasedev, index); - if (name) { - error_prepend(errp, "%s-%d: ", name, subindex); - } else { - error_prepend(errp, "index %d-%d: ", index, subindex); - } - error_prepend(errp, - "Failed to %s %s eventfd signaling for interrupt ", - fd < 0 ? "tear down" : "set up", action_to_str(action)); - return false; -} - int vfio_bitmap_alloc(VFIOBitmap *vbmap, hwaddr size) { vbmap->pages = REAL_HOST_PAGE_ALIGN(size) / qemu_real_host_page_size(); @@ -XXX,XX +XXX,XX @@ vfio_get_device_info_cap(struct vfio_device_info *info, uint16_t id) return vfio_get_cap((void *)info, info->cap_offset, id); } -int vfio_get_region_info(VFIODevice *vbasedev, int index, - struct vfio_region_info **info) -{ - size_t argsz = sizeof(struct vfio_region_info); - - *info = g_malloc0(argsz); - - (*info)->index = index; -retry: - (*info)->argsz = argsz; - - if (ioctl(vbasedev->fd, VFIO_DEVICE_GET_REGION_INFO, *info)) { - g_free(*info); - *info = NULL; - return -errno; - } - - if ((*info)->argsz > argsz) { - argsz = (*info)->argsz; - *info = g_realloc(*info, argsz); - - goto retry; - } - - return 0; -} - struct vfio_info_cap_header * vfio_get_iommu_type1_info_cap(struct vfio_iommu_type1_info *info, uint16_t id) { @@ -XXX,XX +XXX,XX @@ retry: return info; } - -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, - uint32_t subtype, struct vfio_region_info **info) -{ - int i; - - for (i = 0; i < vbasedev->num_regions; i++) { - struct vfio_info_cap_header *hdr; - struct vfio_region_info_cap_type *cap_type; - - if (vfio_get_region_info(vbasedev, i, info)) { - continue; - } - - hdr = vfio_get_region_info_cap(*info, VFIO_REGION_INFO_CAP_TYPE); - if (!hdr) { - g_free(*info); - continue; - } - - cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); - - trace_vfio_get_dev_region(vbasedev->name, i, - cap_type->type, cap_type->subtype); - - if (cap_type->type == type && cap_type->subtype == subtype) { - return 0; - } - - g_free(*info); - } - - *info = NULL; - return -ENODEV; -} - -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) -{ - g_autofree struct vfio_region_info *info = NULL; - bool ret = false; - - if (!vfio_get_region_info(vbasedev, region, &info)) { - if (vfio_get_region_info_cap(info, cap_type)) { - ret = true; - } - } - - return ret; -} - -bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp) -{ - ERRP_GUARD(); - struct stat st; - - if (vbasedev->fd < 0) { - if (stat(vbasedev->sysfsdev, &st) < 0) { - error_setg_errno(errp, errno, "no such host device"); - error_prepend(errp, VFIO_MSG_PREFIX, vbasedev->sysfsdev); - return false; - } - /* User may specify a name, e.g: VFIO platform device */ - if (!vbasedev->name) { - vbasedev->name = g_path_get_basename(vbasedev->sysfsdev); - } - } else { - if (!vbasedev->iommufd) { - error_setg(errp, "Use FD passing only with iommufd backend"); - return false; - } - /* - * Give a name with fd so any function printing out vbasedev->name - * will not break. - */ - if (!vbasedev->name) { - vbasedev->name = g_strdup_printf("VFIO_FD%d", vbasedev->fd); - } - } - - return true; -} - -void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp) -{ - ERRP_GUARD(); - int fd = monitor_fd_param(monitor_cur(), str, errp); - - if (fd < 0) { - error_prepend(errp, "Could not parse remote object fd %s:", str); - return; - } - vbasedev->fd = fd; -} - -void vfio_device_init(VFIODevice *vbasedev, int type, VFIODeviceOps *ops, - DeviceState *dev, bool ram_discard) -{ - vbasedev->type = type; - vbasedev->ops = ops; - vbasedev->dev = dev; - vbasedev->fd = -1; - - vbasedev->ram_block_discard_allowed = ram_discard; -} - -int vfio_device_get_aw_bits(VFIODevice *vdev) -{ - /* - * iova_ranges is a sorted list. For old kernels that support - * VFIO but not support query of iova ranges, iova_ranges is NULL, - * in this case HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX(64) is returned. - */ - GList *l = g_list_last(vdev->bcontainer->iova_ranges); - - if (l) { - Range *range = l->data; - return range_get_last_bit(range) + 1; - } - - return HOST_IOMMU_DEVICE_CAP_AW_BITS_MAX; -} - -bool vfio_device_is_mdev(VFIODevice *vbasedev) -{ - g_autofree char *subsys = NULL; - g_autofree char *tmp = NULL; - - if (!vbasedev->sysfsdev) { - return false; - } - - tmp = g_strdup_printf("%s/subsystem", vbasedev->sysfsdev); - subsys = realpath(tmp, NULL); - return subsys && (strcmp(subsys, "/sys/bus/mdev") == 0); -} - -bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp) -{ - HostIOMMUDevice *hiod = vbasedev->hiod; - - if (!hiod) { - return true; - } - - return HOST_IOMMU_DEVICE_GET_CLASS(hiod)->realize(hiod, vbasedev, errp); -} - -VFIODevice *vfio_get_vfio_device(Object *obj) -{ - if (object_dynamic_cast(obj, TYPE_VFIO_PCI)) { - return &VFIO_PCI(obj)->vbasedev; - } else { - return NULL; - } -} diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'container-base.c', + 'device.c', 'migration.c', 'migration-multifd.c', 'cpr.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_disconnect_container(int fd) "close container->fd=%d" vfio_put_group(int fd) "close group->fd=%d" vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" vfio_put_base_device(int fd) "close vdev->fd=%d" -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 @@ -XXX,XX +XXX,XX @@ iommufd_cdev_fail_attach_existing_container(const char *msg) " %s" iommufd_cdev_alloc_ioas(int iommufd, int ioas_id) " [iommufd=%d] new IOMMUFD container with ioasid=%d" iommufd_cdev_device_info(char *name, int devfd, int num_irqs, int num_regions, int flags) " %s (%d) num_irqs=%d num_regions=%d flags=%d" iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int dev_id) "\t%04x:%02x:%02x.%x devid %d" + +# device.c +vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -- 2.49.0 Gather all CPR related declarations into "vfio-cpr.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". These were introduced in commit d9fa4223b30a ("vfio: register container for cpr"). Order file list in meson.build while at it. Cc: Steve Sistare <steven.sistare@oracle.com> Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-cpr.h | 15 +++++++++++++++ include/hw/vfio/vfio-common.h | 3 --- hw/vfio/container.c | 1 + hw/vfio/cpr.c | 1 + hw/vfio/iommufd.c | 1 + hw/vfio/meson.build | 2 +- 6 files changed, 19 insertions(+), 4 deletions(-) create mode 100644 hw/vfio/vfio-cpr.h diff --git a/hw/vfio/vfio-cpr.h b/hw/vfio/vfio-cpr.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-cpr.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO CPR + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_CPR_H +#define HW_VFIO_CPR_H + +bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); +void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); + +#endif /* HW_VFIO_CPR_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, void vfio_detach_device(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); -bool vfio_cpr_register_container(VFIOContainerBase *bcontainer, Error **errp); -void vfio_cpr_unregister_container(VFIOContainerBase *bcontainer); - typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "hw/vfio/vfio-container.h" #include "vfio-helpers.h" +#include "vfio-cpr.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/cpr.c +++ b/hw/vfio/cpr.c @@ -XXX,XX +XXX,XX @@ #include "migration/misc.h" #include "qapi/error.h" #include "system/runstate.h" +#include "vfio-cpr.h" static int vfio_cpr_reboot_notifier(NotifierWithReturn *notifier, MigrationEvent *e, Error **errp) diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "pci.h" #include "vfio-iommufd.h" #include "vfio-helpers.h" +#include "vfio-cpr.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( 'container-base.c', + 'cpr.c', 'device.c', 'migration.c', 'migration-multifd.c', - 'cpr.c', 'region.c', )) system_ss.add(when: ['CONFIG_VFIO', 'CONFIG_IOMMUFD'], if_true: files( -- 2.49.0 The vfio_kvm_device_add/del_fd() routines opening the VFIO pseudo device are defined in "helpers.c". Move 'vfio_kvm_device_fd' definition there and its declaration into "vfio-helpers.h" to reduce exposure of VFIO internals in "hw/vfio/vfio-common.h". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-22-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-helpers.h | 2 ++ include/hw/vfio/vfio-common.h | 1 - hw/vfio/common.c | 11 ----------- hw/vfio/helpers.c | 11 +++++++++++ hw/vfio/spapr.c | 2 +- 5 files changed, 14 insertions(+), 13 deletions(-) diff --git a/hw/vfio/vfio-helpers.h b/hw/vfio/vfio-helpers.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-helpers.h +++ b/hw/vfio/vfio-helpers.h @@ -XXX,XX +XXX,XX @@ #ifdef CONFIG_LINUX #include <linux/vfio.h> +extern int vfio_kvm_device_fd; + struct vfio_info_cap_header * vfio_get_cap(void *ptr, uint32_t cap_offset, uint16_t id); struct vfio_info_cap_header * diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; extern const MemoryListener vfio_memory_listener; -extern int vfio_kvm_device_fd; #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); -#ifdef CONFIG_KVM -/* - * We have a single VFIO pseudo device per KVM VM. Once created it lives - * for the life of the VM. Closing the file descriptor only drops our - * reference to it and the device's reference to kvm. Therefore once - * initialized, this file descriptor is only released on QEMU exit and - * we'll re-use it should another vfio device be attached before then. - */ -int vfio_kvm_device_fd = -1; -#endif - /* * Device state interfaces */ diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ bool vfio_get_info_dma_avail(struct vfio_iommu_type1_info *info, return true; } +#ifdef CONFIG_KVM +/* + * We have a single VFIO pseudo device per KVM VM. Once created it lives + * for the life of the VM. Closing the file descriptor only drops our + * reference to it and the device's reference to kvm. Therefore once + * initialized, this file descriptor is only released on QEMU exit and + * we'll re-use it should another vfio device be attached before then. + */ +int vfio_kvm_device_fd = -1; +#endif + int vfio_kvm_device_add_fd(int fd, Error **errp) { #ifdef CONFIG_KVM diff --git a/hw/vfio/spapr.c b/hw/vfio/spapr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/spapr.c +++ b/hw/vfio/spapr.c @@ -XXX,XX +XXX,XX @@ #include "system/hostmem.h" #include "exec/address-spaces.h" -#include "hw/vfio/vfio-common.h" #include "hw/vfio/vfio-container.h" #include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" #include "qapi/error.h" #include "trace.h" +#include "vfio-helpers.h" typedef struct VFIOHostDMAWindow { hwaddr min_iova; -- 2.49.0 'vfio_device_list' is VFIODevice related. Move its definitions into "device.c". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-23-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 3 --- hw/vfio/device.c | 3 +++ 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ #include "vfio-migration-internal.h" #include "vfio-helpers.h" -VFIODeviceList vfio_device_list = - QLIST_HEAD_INITIALIZER(vfio_device_list); - /* * Device state interfaces */ diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ #include "monitor/monitor.h" #include "vfio-helpers.h" +VFIODeviceList vfio_device_list = + QLIST_HEAD_INITIALIZER(vfio_device_list); + /* * Common VFIO interrupt disable */ -- 2.49.0 These routines are VFIODevice related. Move their definitions into "device.c". Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-24-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 37 ------------------------------------- hw/vfio/device.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 37 insertions(+), 37 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ void vfio_reset_handler(void *opaque) } } } - -bool vfio_attach_device(char *name, VFIODevice *vbasedev, - AddressSpace *as, Error **errp) -{ - const VFIOIOMMUClass *ops = - VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_LEGACY)); - HostIOMMUDevice *hiod = NULL; - - if (vbasedev->iommufd) { - ops = VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_IOMMUFD)); - } - - assert(ops); - - - if (!vbasedev->mdev) { - hiod = HOST_IOMMU_DEVICE(object_new(ops->hiod_typename)); - vbasedev->hiod = hiod; - } - - if (!ops->attach_device(name, vbasedev, as, errp)) { - object_unref(hiod); - vbasedev->hiod = NULL; - return false; - } - - return true; -} - -void vfio_detach_device(VFIODevice *vbasedev) -{ - if (!vbasedev->bcontainer) { - return; - } - object_unref(vbasedev->hiod); - VFIO_IOMMU_GET_CLASS(vbasedev->bcontainer)->detach_device(vbasedev); -} diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj) return NULL; } } + +bool vfio_attach_device(char *name, VFIODevice *vbasedev, + AddressSpace *as, Error **errp) +{ + const VFIOIOMMUClass *ops = + VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_LEGACY)); + HostIOMMUDevice *hiod = NULL; + + if (vbasedev->iommufd) { + ops = VFIO_IOMMU_CLASS(object_class_by_name(TYPE_VFIO_IOMMU_IOMMUFD)); + } + + assert(ops); + + + if (!vbasedev->mdev) { + hiod = HOST_IOMMU_DEVICE(object_new(ops->hiod_typename)); + vbasedev->hiod = hiod; + } + + if (!ops->attach_device(name, vbasedev, as, errp)) { + object_unref(hiod); + vbasedev->hiod = NULL; + return false; + } + + return true; +} + +void vfio_detach_device(VFIODevice *vbasedev) +{ + if (!vbasedev->bcontainer) { + return; + } + object_unref(vbasedev->hiod); + VFIO_IOMMU_GET_CLASS(vbasedev->bcontainer)->detach_device(vbasedev); +} -- 2.49.0 Pass-through devices of a VM are not necessarily in the same group and all groups/address_spaces need to be scanned when the machine is reset. Commit f16f39c3fc97 ("Implement PCI hot reset") introduced a VM reset handler for this purpose. Move it under device.c Also reintroduce the comment which explained the context and was lost along the way. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/common.c | 18 ------------------ hw/vfio/device.c | 35 +++++++++++++++++++++++++++++++++++ hw/vfio/trace-events | 2 +- 3 files changed, 36 insertions(+), 19 deletions(-) diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_global_stop = vfio_listener_log_global_stop, .log_sync = vfio_listener_log_sync, }; - -void vfio_reset_handler(void *opaque) -{ - VFIODevice *vbasedev; - - trace_vfio_reset_handler(); - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->dev->realized) { - vbasedev->ops->vfio_compute_needs_reset(vbasedev); - } - } - - QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { - if (vbasedev->dev->realized && vbasedev->needs_reset) { - vbasedev->ops->vfio_hot_reset_multi(vbasedev); - } - } -} diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = QLIST_HEAD_INITIALIZER(vfio_device_list); +/* + * We want to differentiate hot reset of multiple in-use devices vs + * hot reset of a single in-use device. VFIO_DEVICE_RESET will already + * handle the case of doing hot resets when there is only a single + * device per bus. The in-use here refers to how many VFIODevices are + * affected. A hot reset that affects multiple devices, but only a + * single in-use device, means that we can call it from our bus + * ->reset() callback since the extent is effectively a single + * device. This allows us to make use of it in the hotplug path. When + * there are multiple in-use devices, we can only trigger the hot + * reset during a system reset and thus from our reset handler. We + * separate _one vs _multi here so that we don't overlap and do a + * double reset on the system reset path where both our reset handler + * and ->reset() callback are used. Calling _one() will only do a hot + * reset for the one in-use devices case, calling _multi() will do + * nothing if a _one() would have been sufficient. + */ +void vfio_reset_handler(void *opaque) +{ + VFIODevice *vbasedev; + + trace_vfio_reset_handler(); + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->dev->realized) { + vbasedev->ops->vfio_compute_needs_reset(vbasedev); + } + } + + QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { + if (vbasedev->dev->realized && vbasedev->needs_reset) { + vbasedev->ops->vfio_hot_reset_multi(vbasedev); + } + } +} + /* * Common VFIO interrupt disable */ diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_put_base_device(int fd) "close vdev->fd=%d" vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 -vfio_reset_handler(void) "" # region.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" @@ -XXX,XX +XXX,XX @@ iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int functi # device.c vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_reset_handler(void) "" -- 2.49.0 Routines of common.c : vfio_devices_all_dirty_tracking_started vfio_devices_all_device_dirty_tracking vfio_devices_query_dirty_bitmap vfio_get_dirty_bitmap are all related to dirty page tracking directly at the container level or at the container device level. Naming is a bit confusing. We will propose new names in the following changes. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-common.h | 9 -- include/hw/vfio/vfio-container-base.h | 7 ++ hw/vfio/common.c | 130 ------------------------ hw/vfio/container-base.c | 138 ++++++++++++++++++++++++++ hw/vfio/meson.build | 2 +- hw/vfio/trace-events | 4 +- 6 files changed, 149 insertions(+), 141 deletions(-) diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); #endif -bool vfio_devices_all_dirty_tracking_started( - const VFIOContainerBase *bcontainer); -bool -vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp); -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, - uint64_t size, ram_addr_t ram_addr, Error **errp); - /* Returns 0 on success, or a negative errno. */ bool vfio_device_get_name(VFIODevice *vbasedev, Error **errp); void vfio_device_set_fd(VFIODevice *vbasedev, const char *str, Error **errp); diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start, Error **errp); int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp); +bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); +bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); +int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, hwaddr iova, hwaddr size, + Error **errp); +int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, + uint64_t size, ram_addr_t ram_addr, Error **errp); GList *vfio_container_get_iova_ranges(const VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ */ -static bool vfio_devices_all_device_dirty_tracking_started( - const VFIOContainerBase *bcontainer) -{ - VFIODevice *vbasedev; - - QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { - if (!vbasedev->dirty_tracking) { - return false; - } - } - - return true; -} - -bool vfio_devices_all_dirty_tracking_started( - const VFIOContainerBase *bcontainer) -{ - return vfio_devices_all_device_dirty_tracking_started(bcontainer) || - bcontainer->dirty_pages_started; -} - static bool vfio_log_sync_needed(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ static bool vfio_log_sync_needed(const VFIOContainerBase *bcontainer) return true; } -bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer) -{ - VFIODevice *vbasedev; - - QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { - if (vbasedev->device_dirty_page_tracking == ON_OFF_AUTO_OFF) { - return false; - } - if (!vbasedev->dirty_pages_supported) { - return false; - } - } - - return true; -} - static bool vfio_listener_skipped_section(MemoryRegionSection *section) { return (!memory_region_is_ram(section->mr) && @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_global_stop(MemoryListener *listener) } } -static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, - hwaddr size, void *bitmap) -{ - uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature) + - sizeof(struct vfio_device_feature_dma_logging_report), - sizeof(uint64_t))] = {}; - struct vfio_device_feature *feature = (struct vfio_device_feature *)buf; - struct vfio_device_feature_dma_logging_report *report = - (struct vfio_device_feature_dma_logging_report *)feature->data; - - report->iova = iova; - report->length = size; - report->page_size = qemu_real_host_page_size(); - report->bitmap = (uintptr_t)bitmap; - - feature->argsz = sizeof(buf); - feature->flags = VFIO_DEVICE_FEATURE_GET | - VFIO_DEVICE_FEATURE_DMA_LOGGING_REPORT; - - if (ioctl(vbasedev->fd, VFIO_DEVICE_FEATURE, feature)) { - return -errno; - } - - return 0; -} - -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) -{ - VFIODevice *vbasedev; - int ret; - - QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { - ret = vfio_device_dma_logging_report(vbasedev, iova, size, - vbmap->bitmap); - if (ret) { - error_setg_errno(errp, -ret, - "%s: Failed to get DMA logging report, iova: " - "0x%" HWADDR_PRIx ", size: 0x%" HWADDR_PRIx, - vbasedev->name, iova, size); - - return ret; - } - } - - return 0; -} - -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, - uint64_t size, ram_addr_t ram_addr, Error **errp) -{ - bool all_device_dirty_tracking = - vfio_devices_all_device_dirty_tracking(bcontainer); - uint64_t dirty_pages; - VFIOBitmap vbmap; - int ret; - - if (!bcontainer->dirty_pages_supported && !all_device_dirty_tracking) { - cpu_physical_memory_set_dirty_range(ram_addr, size, - tcg_enabled() ? DIRTY_CLIENTS_ALL : - DIRTY_CLIENTS_NOCODE); - return 0; - } - - ret = vfio_bitmap_alloc(&vbmap, size); - if (ret) { - error_setg_errno(errp, -ret, - "Failed to allocate dirty tracking bitmap"); - return ret; - } - - if (all_device_dirty_tracking) { - ret = vfio_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, - errp); - } else { - ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size, - errp); - } - - if (ret) { - goto out; - } - - dirty_pages = cpu_physical_memory_set_dirty_lebitmap(vbmap.bitmap, ram_addr, - vbmap.pages); - - trace_vfio_get_dirty_bitmap(iova, size, vbmap.size, ram_addr, dirty_pages); -out: - g_free(vbmap.bitmap); - - return ret; -} - typedef struct { IOMMUNotifier n; VFIOGuestIOMMU *giommu; diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ * SPDX-License-Identifier: GPL-2.0-or-later */ +#include <sys/ioctl.h> +#include <linux/vfio.h> + +#include "system/tcg.h" #include "qemu/osdep.h" +#include "exec/ram_addr.h" #include "qapi/error.h" #include "qemu/error-report.h" #include "hw/vfio/vfio-container-base.h" #include "hw/vfio/vfio-common.h" /* vfio_reset_handler */ #include "system/reset.h" +#include "vfio-helpers.h" + +#include "trace.h" static QLIST_HEAD(, VFIOAddressSpace) vfio_address_spaces = QLIST_HEAD_INITIALIZER(vfio_address_spaces); @@ -XXX,XX +XXX,XX @@ int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, errp); } +static bool vfio_devices_all_device_dirty_tracking_started( + const VFIOContainerBase *bcontainer) +{ + VFIODevice *vbasedev; + + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { + if (!vbasedev->dirty_tracking) { + return false; + } + } + + return true; +} + +bool vfio_devices_all_dirty_tracking_started( + const VFIOContainerBase *bcontainer) +{ + return vfio_devices_all_device_dirty_tracking_started(bcontainer) || + bcontainer->dirty_pages_started; +} + +bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer) +{ + VFIODevice *vbasedev; + + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { + if (vbasedev->device_dirty_page_tracking == ON_OFF_AUTO_OFF) { + return false; + } + if (!vbasedev->dirty_pages_supported) { + return false; + } + } + + return true; +} + +static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, + hwaddr size, void *bitmap) +{ + uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature) + + sizeof(struct vfio_device_feature_dma_logging_report), + sizeof(uint64_t))] = {}; + struct vfio_device_feature *feature = (struct vfio_device_feature *)buf; + struct vfio_device_feature_dma_logging_report *report = + (struct vfio_device_feature_dma_logging_report *)feature->data; + + report->iova = iova; + report->length = size; + report->page_size = qemu_real_host_page_size(); + report->bitmap = (uintptr_t)bitmap; + + feature->argsz = sizeof(buf); + feature->flags = VFIO_DEVICE_FEATURE_GET | + VFIO_DEVICE_FEATURE_DMA_LOGGING_REPORT; + + if (ioctl(vbasedev->fd, VFIO_DEVICE_FEATURE, feature)) { + return -errno; + } + + return 0; +} + +int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) +{ + VFIODevice *vbasedev; + int ret; + + QLIST_FOREACH(vbasedev, &bcontainer->device_list, container_next) { + ret = vfio_device_dma_logging_report(vbasedev, iova, size, + vbmap->bitmap); + if (ret) { + error_setg_errno(errp, -ret, + "%s: Failed to get DMA logging report, iova: " + "0x%" HWADDR_PRIx ", size: 0x%" HWADDR_PRIx, + vbasedev->name, iova, size); + + return ret; + } + } + + return 0; +} + +int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, + uint64_t size, ram_addr_t ram_addr, Error **errp) +{ + bool all_device_dirty_tracking = + vfio_devices_all_device_dirty_tracking(bcontainer); + uint64_t dirty_pages; + VFIOBitmap vbmap; + int ret; + + if (!bcontainer->dirty_pages_supported && !all_device_dirty_tracking) { + cpu_physical_memory_set_dirty_range(ram_addr, size, + tcg_enabled() ? DIRTY_CLIENTS_ALL : + DIRTY_CLIENTS_NOCODE); + return 0; + } + + ret = vfio_bitmap_alloc(&vbmap, size); + if (ret) { + error_setg_errno(errp, -ret, + "Failed to allocate dirty tracking bitmap"); + return ret; + } + + if (all_device_dirty_tracking) { + ret = vfio_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, + errp); + } else { + ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size, + errp); + } + + if (ret) { + goto out; + } + + dirty_pages = cpu_physical_memory_set_dirty_lebitmap(vbmap.bitmap, ram_addr, + vbmap.pages); + + trace_vfio_get_dirty_bitmap(iova, size, vbmap.size, ram_addr, dirty_pages); +out: + g_free(vbmap.bitmap); + + return ret; +} + static gpointer copy_iova_range(gconstpointer src, gpointer data) { Range *source = (Range *)src; diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ vfio_ss = ss.source_set() vfio_ss.add(files( 'common.c', + 'container-base.c', 'container.c', 'helpers.c', )) @@ -XXX,XX +XXX,XX @@ specific_ss.add_all(when: 'CONFIG_VFIO', if_true: vfio_ss) system_ss.add(when: 'CONFIG_VFIO_XGMAC', if_true: files('calxeda-xgmac.c')) system_ss.add(when: 'CONFIG_VFIO_AMD_XGBE', if_true: files('amd-xgbe.c')) system_ss.add(when: 'CONFIG_VFIO', if_true: files( - 'container-base.c', 'cpr.c', 'device.c', 'migration.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_put_group(int fd) "close group->fd=%d" vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" vfio_put_base_device(int fd) "close vdev->fd=%d" vfio_legacy_dma_unmap_overflow_workaround(void) "" -vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 +# container-base.c +vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 + # region.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 -- 2.49.0 vfio_devices_query_dirty_bitmap() is only used in "container-base.c". Also, rename to vfio_container_devices_query_dirty_bitmap() to reflect with the prefix 'vfio_container_devices_' that it simply loops over the container's device list. This rename also frees the vfio_devices_query_dirty_bitmap() name which we will reuse in the next changes. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-container-base.h | 3 --- hw/vfio/container-base.c | 6 +++--- 2 files changed, 3 insertions(+), 6 deletions(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp); bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, - Error **errp); int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } -int vfio_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, +static int vfio_container_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, } if (all_device_dirty_tracking) { - ret = vfio_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, - errp); + ret = vfio_container_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, + errp); } else { ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size, errp); -- 2.49.0 vfio_container_query_dirty_bitmap() is only used in "container-base.c". Also, rename to vfio_container_vioc_query_dirty_bitmap() to reflect it is using the VFIO IOMMU backend device ->query_dirty_bitmap() handler. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-container-base.h | 2 -- hw/vfio/container-base.c | 24 ++++++++++++------------ 2 files changed, 12 insertions(+), 14 deletions(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ void vfio_container_del_section_window(VFIOContainerBase *bcontainer, MemoryRegionSection *section); int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start, Error **errp); -int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp); bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, return ret; } -int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, - VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) -{ - VFIOIOMMUClass *vioc = VFIO_IOMMU_GET_CLASS(bcontainer); - - g_assert(vioc->query_dirty_bitmap); - return vioc->query_dirty_bitmap(bcontainer, vbmap, iova, size, - errp); -} - static bool vfio_devices_all_device_dirty_tracking_started( const VFIOContainerBase *bcontainer) { @@ -XXX,XX +XXX,XX @@ static int vfio_device_dma_logging_report(VFIODevice *vbasedev, hwaddr iova, return 0; } +static int vfio_container_vioc_query_dirty_bitmap(const VFIOContainerBase *bcontainer, + VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) +{ + VFIOIOMMUClass *vioc = VFIO_IOMMU_GET_CLASS(bcontainer); + + g_assert(vioc->query_dirty_bitmap); + return vioc->query_dirty_bitmap(bcontainer, vbmap, iova, size, + errp); +} + static int vfio_container_devices_query_dirty_bitmap(const VFIOContainerBase *bcontainer, VFIOBitmap *vbmap, hwaddr iova, hwaddr size, Error **errp) { @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, ret = vfio_container_devices_query_dirty_bitmap(bcontainer, &vbmap, iova, size, errp); } else { - ret = vfio_container_query_dirty_bitmap(bcontainer, &vbmap, iova, size, - errp); + ret = vfio_container_vioc_query_dirty_bitmap(bcontainer, &vbmap, iova, size, + errp); } if (ret) { -- 2.49.0 Also rename vfio_devices_all_device_dirty_tracking_started() while at it and use the prefix 'vfio_container_devices_' for routines simply looping over the container's device list. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-container-base.h | 3 ++- hw/vfio/common.c | 2 +- hw/vfio/container-base.c | 6 +++--- hw/vfio/container.c | 2 +- 4 files changed, 7 insertions(+), 6 deletions(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ void vfio_container_del_section_window(VFIOContainerBase *bcontainer, MemoryRegionSection *section); int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start, Error **errp); -bool vfio_devices_all_dirty_tracking_started(const VFIOContainerBase *bcontainer); +bool vfio_container_dirty_tracking_is_started( + const VFIOContainerBase *bcontainer); bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ static bool vfio_log_sync_needed(const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; - if (!vfio_devices_all_dirty_tracking_started(bcontainer)) { + if (!vfio_container_dirty_tracking_is_started(bcontainer)) { return false; } diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, return ret; } -static bool vfio_devices_all_device_dirty_tracking_started( +static bool vfio_container_devices_dirty_tracking_is_started( const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ static bool vfio_devices_all_device_dirty_tracking_started( return true; } -bool vfio_devices_all_dirty_tracking_started( +bool vfio_container_dirty_tracking_is_started( const VFIOContainerBase *bcontainer) { - return vfio_devices_all_device_dirty_tracking_started(bcontainer) || + return vfio_container_devices_dirty_tracking_is_started(bcontainer) || bcontainer->dirty_pages_started; } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, int ret; Error *local_err = NULL; - if (iotlb && vfio_devices_all_dirty_tracking_started(bcontainer)) { + if (iotlb && vfio_container_dirty_tracking_is_started(bcontainer)) { if (!vfio_devices_all_device_dirty_tracking(bcontainer) && bcontainer->dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); -- 2.49.0 Use the prefix 'vfio_container_devices_' to reflect the routine simply loops over the container's device list. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-container-base.h | 3 ++- hw/vfio/common.c | 4 ++-- hw/vfio/container-base.c | 5 +++-- hw/vfio/container.c | 2 +- 4 files changed, 8 insertions(+), 6 deletions(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ int vfio_container_set_dirty_page_tracking(VFIOContainerBase *bcontainer, bool start, Error **errp); bool vfio_container_dirty_tracking_is_started( const VFIOContainerBase *bcontainer); -bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer); +bool vfio_container_devices_dirty_tracking_is_supported( + const VFIOContainerBase *bcontainer); int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ static bool vfio_listener_log_global_start(MemoryListener *listener, listener); bool ret; - if (vfio_devices_all_device_dirty_tracking(bcontainer)) { + if (vfio_container_devices_dirty_tracking_is_supported(bcontainer)) { ret = vfio_devices_dma_logging_start(bcontainer, errp); } else { ret = vfio_container_set_dirty_page_tracking(bcontainer, true, errp) == 0; @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_global_stop(MemoryListener *listener) Error *local_err = NULL; int ret = 0; - if (vfio_devices_all_device_dirty_tracking(bcontainer)) { + if (vfio_container_devices_dirty_tracking_is_supported(bcontainer)) { vfio_devices_dma_logging_stop(bcontainer); } else { ret = vfio_container_set_dirty_page_tracking(bcontainer, false, diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ bool vfio_container_dirty_tracking_is_started( bcontainer->dirty_pages_started; } -bool vfio_devices_all_device_dirty_tracking(const VFIOContainerBase *bcontainer) +bool vfio_container_devices_dirty_tracking_is_supported( + const VFIOContainerBase *bcontainer) { VFIODevice *vbasedev; @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp) { bool all_device_dirty_tracking = - vfio_devices_all_device_dirty_tracking(bcontainer); + vfio_container_devices_dirty_tracking_is_supported(bcontainer); uint64_t dirty_pages; VFIOBitmap vbmap; int ret; diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, Error *local_err = NULL; if (iotlb && vfio_container_dirty_tracking_is_started(bcontainer)) { - if (!vfio_devices_all_device_dirty_tracking(bcontainer) && + if (!vfio_container_devices_dirty_tracking_is_supported(bcontainer) && bcontainer->dirty_pages_supported) { return vfio_dma_unmap_bitmap(container, iova, size, iotlb); } -- 2.49.0 vfio_container_query_dirty_bitmap() is consistent the VFIO container routine naming scheme and is now free to use. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-container-base.h | 4 ++-- hw/vfio/common.c | 6 +++--- hw/vfio/container-base.c | 5 +++-- hw/vfio/container.c | 2 +- hw/vfio/trace-events | 2 +- 5 files changed, 10 insertions(+), 9 deletions(-) diff --git a/include/hw/vfio/vfio-container-base.h b/include/hw/vfio/vfio-container-base.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-container-base.h +++ b/include/hw/vfio/vfio-container-base.h @@ -XXX,XX +XXX,XX @@ bool vfio_container_dirty_tracking_is_started( const VFIOContainerBase *bcontainer); bool vfio_container_devices_dirty_tracking_is_supported( const VFIOContainerBase *bcontainer); -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, - uint64_t size, ram_addr_t ram_addr, Error **errp); +int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, + uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp); GList *vfio_container_get_iova_ranges(const VFIOContainerBase *bcontainer); diff --git a/hw/vfio/common.c b/hw/vfio/common.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/common.c +++ b/hw/vfio/common.c @@ -XXX,XX +XXX,XX @@ static void vfio_iommu_map_dirty_notify(IOMMUNotifier *n, IOMMUTLBEntry *iotlb) goto out_unlock; } - ret = vfio_get_dirty_bitmap(bcontainer, iova, iotlb->addr_mask + 1, + ret = vfio_container_query_dirty_bitmap(bcontainer, iova, iotlb->addr_mask + 1, translated_addr, &local_err); if (ret) { error_prepend(&local_err, @@ -XXX,XX +XXX,XX @@ static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, * Sync the whole mapped region (spanning multiple individual mappings) * in one go. */ - ret = vfio_get_dirty_bitmap(vrdl->bcontainer, iova, size, ram_addr, + ret = vfio_container_query_dirty_bitmap(vrdl->bcontainer, iova, size, ram_addr, &local_err); if (ret) { error_report_err(local_err); @@ -XXX,XX +XXX,XX @@ static int vfio_sync_dirty_bitmap(VFIOContainerBase *bcontainer, ram_addr = memory_region_get_ram_addr(section->mr) + section->offset_within_region; - return vfio_get_dirty_bitmap(bcontainer, + return vfio_container_query_dirty_bitmap(bcontainer, REAL_HOST_PAGE_ALIGN(section->offset_within_address_space), int128_get64(section->size), ram_addr, errp); } diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ static int vfio_container_devices_query_dirty_bitmap(const VFIOContainerBase *bc return 0; } -int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, +int vfio_container_query_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, uint64_t size, ram_addr_t ram_addr, Error **errp) { bool all_device_dirty_tracking = @@ -XXX,XX +XXX,XX @@ int vfio_get_dirty_bitmap(const VFIOContainerBase *bcontainer, uint64_t iova, dirty_pages = cpu_physical_memory_set_dirty_lebitmap(vbmap.bitmap, ram_addr, vbmap.pages); - trace_vfio_get_dirty_bitmap(iova, size, vbmap.size, ram_addr, dirty_pages); + trace_vfio_container_query_dirty_bitmap(iova, size, vbmap.size, ram_addr, + dirty_pages); out: g_free(vbmap.bitmap); diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_legacy_dma_unmap(const VFIOContainerBase *bcontainer, } if (need_dirty_sync) { - ret = vfio_get_dirty_bitmap(bcontainer, iova, size, + ret = vfio_container_query_dirty_bitmap(bcontainer, iova, size, iotlb->translated_addr, &local_err); if (ret) { error_report_err(local_err); diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 # container-base.c -vfio_get_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +vfio_container_query_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 # region.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" -- 2.49.0 File "common.c" has been emptied of most of its definitions by the previous changes and the only definitions left are related to the VFIO MemoryListener handlers. Rename it to "listener.c" and introduce its associated "vfio-listener.h" header file for the declarations. Cleanup a little the includes while at it. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-listener.h | 14 ++++++++++++++ include/hw/vfio/vfio-common.h | 1 - hw/vfio/container.c | 1 + hw/vfio/iommufd.c | 1 + hw/vfio/{common.c => listener.c} | 0 hw/vfio/meson.build | 2 +- hw/vfio/trace-events | 2 +- 7 files changed, 18 insertions(+), 3 deletions(-) create mode 100644 hw/vfio/vfio-listener.h rename hw/vfio/{common.c => listener.c} (100%) diff --git a/hw/vfio/vfio-listener.h b/hw/vfio/vfio-listener.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/vfio/vfio-listener.h @@ -XXX,XX +XXX,XX @@ +/* + * VFIO MemoryListener services + * + * Copyright Red Hat, Inc. 2025 + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef HW_VFIO_VFIO_LISTENER_H +#define HW_VFIO_VFIO_LISTENER_H + +extern const MemoryListener vfio_memory_listener; + +#endif /* HW_VFIO_VFIO_LISTENER_H */ diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-common.h +++ b/include/hw/vfio/vfio-common.h @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; -extern const MemoryListener vfio_memory_listener; #ifdef CONFIG_LINUX int vfio_get_region_info(VFIODevice *vbasedev, int index, diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include "hw/vfio/vfio-container.h" #include "vfio-helpers.h" #include "vfio-cpr.h" +#include "vfio-listener.h" #define TYPE_HOST_IOMMU_DEVICE_LEGACY_VFIO TYPE_HOST_IOMMU_DEVICE "-legacy-vfio" diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "vfio-iommufd.h" #include "vfio-helpers.h" #include "vfio-cpr.h" +#include "vfio-listener.h" #define TYPE_HOST_IOMMU_DEVICE_IOMMUFD_VFIO \ TYPE_HOST_IOMMU_DEVICE_IOMMUFD "-vfio" diff --git a/hw/vfio/common.c b/hw/vfio/listener.c similarity index 100% rename from hw/vfio/common.c rename to hw/vfio/listener.c diff --git a/hw/vfio/meson.build b/hw/vfio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/meson.build +++ b/hw/vfio/meson.build @@ -XXX,XX +XXX,XX @@ vfio_ss = ss.source_set() vfio_ss.add(files( - 'common.c', + 'listener.c', 'container-base.c', 'container.c', 'helpers.c', diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_igd_bdsm_enabled(const char *name, int size) "%s %dMB" vfio_pci_igd_host_bridge_enabled(const char *name) "%s" vfio_pci_igd_lpc_bridge_enabled(const char *name) "%s" -# common.c +# listener.c vfio_iommu_map_notify(const char *op, uint64_t iova_start, uint64_t iova_end) "iommu %s @ 0x%"PRIx64" - 0x%"PRIx64 vfio_listener_region_skip(const char *name, uint64_t start, uint64_t end) "SKIPPING %s 0x%"PRIx64" - 0x%"PRIx64 vfio_spapr_group_attach(int groupfd, int tablefd) "Attached groupfd %d to liobn fd %d" -- 2.49.0 Rename some routines to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/listener.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/hw/vfio/listener.c b/hw/vfio/listener.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/listener.c +++ b/hw/vfio/listener.c @@ -XXX,XX +XXX,XX @@ static int vfio_ram_discard_notify_populate(RamDiscardListener *rdl, return 0; } -static void vfio_register_ram_discard_listener(VFIOContainerBase *bcontainer, +static void vfio_ram_discard_register_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); @@ -XXX,XX +XXX,XX @@ static void vfio_register_ram_discard_listener(VFIOContainerBase *bcontainer, } } -static void vfio_unregister_ram_discard_listener(VFIOContainerBase *bcontainer, +static void vfio_ram_discard_unregister_listener(VFIOContainerBase *bcontainer, MemoryRegionSection *section) { RamDiscardManager *rdm = memory_region_get_ram_discard_manager(section->mr); @@ -XXX,XX +XXX,XX @@ static void vfio_listener_region_add(MemoryListener *listener, * about changes. */ if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_register_ram_discard_listener(bcontainer, section); + vfio_ram_discard_register_listener(bcontainer, section); return; } @@ -XXX,XX +XXX,XX @@ static void vfio_listener_region_del(MemoryListener *listener, pgmask = (1ULL << ctz64(bcontainer->pgsizes)) - 1; try_unmap = !((iova & pgmask) || (int128_get64(llsize) & pgmask)); } else if (memory_region_has_ram_discard_manager(section->mr)) { - vfio_unregister_ram_discard_listener(bcontainer, section); + vfio_ram_discard_unregister_listener(bcontainer, section); /* Unregistering will trigger an unmap. */ try_unmap = false; } @@ -XXX,XX +XXX,XX @@ out: } } -static int vfio_ram_discard_get_dirty_bitmap(MemoryRegionSection *section, +static int vfio_ram_discard_query_dirty_bitmap(MemoryRegionSection *section, void *opaque) { const hwaddr size = int128_get64(section->size); @@ -XXX,XX +XXX,XX @@ vfio_sync_ram_discard_listener_dirty_bitmap(VFIOContainerBase *bcontainer, * which correspond to populated parts. Replay all populated parts. */ return ram_discard_manager_replay_populated(rdm, section, - vfio_ram_discard_get_dirty_bitmap, + vfio_ram_discard_query_dirty_bitmap, &vrdl); } -- 2.49.0 This hides the MemoryListener implementation and makes the code common to both IOMMU backends, legacy and IOMMUFD. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/vfio-listener.h | 3 ++- hw/vfio/container.c | 11 +++-------- hw/vfio/iommufd.c | 9 ++------- hw/vfio/listener.c | 22 +++++++++++++++++++++- 4 files changed, 28 insertions(+), 17 deletions(-) diff --git a/hw/vfio/vfio-listener.h b/hw/vfio/vfio-listener.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/vfio-listener.h +++ b/hw/vfio/vfio-listener.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_VFIO_LISTENER_H #define HW_VFIO_VFIO_LISTENER_H -extern const MemoryListener vfio_memory_listener; +bool vfio_listener_register(VFIOContainerBase *bcontainer, Error **errp); +void vfio_listener_unregister(VFIOContainerBase *bcontainer); #endif /* HW_VFIO_VFIO_LISTENER_H */ diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); - bcontainer->listener = vfio_memory_listener; - memory_listener_register(&bcontainer->listener, bcontainer->space->as); - - if (bcontainer->error) { - error_propagate_prepend(errp, bcontainer->error, - "memory listener initialization failed: "); + if (!vfio_listener_register(bcontainer, errp)) { goto listener_release_exit; } @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, listener_release_exit: QLIST_REMOVE(group, container_next); vfio_kvm_device_del_group(group); - memory_listener_unregister(&bcontainer->listener); + vfio_listener_unregister(bcontainer); if (vioc->release) { vioc->release(bcontainer); } @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) * group. */ if (QLIST_EMPTY(&container->group_list)) { - memory_listener_unregister(&bcontainer->listener); + vfio_listener_unregister(bcontainer); if (vioc->release) { vioc->release(bcontainer); } diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ static void iommufd_cdev_container_destroy(VFIOIOMMUFDContainer *container) if (!QLIST_EMPTY(&bcontainer->device_list)) { return; } - memory_listener_unregister(&bcontainer->listener); + vfio_listener_unregister(bcontainer); iommufd_backend_free_id(container->be, container->ioas_id); object_unref(container); } @@ -XXX,XX +XXX,XX @@ static bool iommufd_cdev_attach(const char *name, VFIODevice *vbasedev, bcontainer->pgsizes = qemu_real_host_page_size(); } - bcontainer->listener = vfio_memory_listener; - memory_listener_register(&bcontainer->listener, bcontainer->space->as); - - if (bcontainer->error) { - error_propagate_prepend(errp, bcontainer->error, - "memory listener initialization failed: "); + if (!vfio_listener_register(bcontainer, errp)) { goto err_listener_register; } diff --git a/hw/vfio/listener.c b/hw/vfio/listener.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/listener.c +++ b/hw/vfio/listener.c @@ -XXX,XX +XXX,XX @@ #include "system/tpm.h" #include "vfio-migration-internal.h" #include "vfio-helpers.h" +#include "vfio-listener.h" /* * Device state interfaces @@ -XXX,XX +XXX,XX @@ static void vfio_listener_log_sync(MemoryListener *listener, } } -const MemoryListener vfio_memory_listener = { +static const MemoryListener vfio_memory_listener = { .name = "vfio", .region_add = vfio_listener_region_add, .region_del = vfio_listener_region_del, @@ -XXX,XX +XXX,XX @@ const MemoryListener vfio_memory_listener = { .log_global_stop = vfio_listener_log_global_stop, .log_sync = vfio_listener_log_sync, }; + +bool vfio_listener_register(VFIOContainerBase *bcontainer, Error **errp) +{ + bcontainer->listener = vfio_memory_listener; + memory_listener_register(&bcontainer->listener, bcontainer->space->as); + + if (bcontainer->error) { + error_propagate_prepend(errp, bcontainer->error, + "memory listener initialization failed: "); + return false; + } + + return true; +} + +void vfio_listener_unregister(VFIOContainerBase *bcontainer) +{ + memory_listener_unregister(&bcontainer->listener); +} -- 2.49.0 "hw/vfio/vfio-common.h" has been emptied of most of its declarations by the previous changes and the only declarations left are related to VFIODevice. Rename it to "hw/vfio/vfio-device.h" and make the necessary adjustments. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/migration-multifd.h | 2 +- hw/vfio/pci.h | 2 +- include/hw/s390x/vfio-ccw.h | 2 +- include/hw/vfio/{vfio-common.h => vfio-device.h} | 0 include/hw/vfio/vfio-platform.h | 2 +- backends/iommufd.c | 2 +- hw/vfio/ap.c | 2 +- hw/vfio/ccw.c | 2 +- hw/vfio/container-base.c | 2 +- hw/vfio/container.c | 2 +- hw/vfio/cpr.c | 2 +- hw/vfio/device.c | 2 +- hw/vfio/helpers.c | 2 +- hw/vfio/iommufd.c | 2 +- hw/vfio/listener.c | 2 +- hw/vfio/migration-multifd.c | 2 +- hw/vfio/migration.c | 2 +- hw/vfio/region.c | 4 ++-- 18 files changed, 18 insertions(+), 18 deletions(-) rename include/hw/vfio/{vfio-common.h => vfio-device.h} (100%) diff --git a/hw/vfio/migration-multifd.h b/hw/vfio/migration-multifd.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.h +++ b/hw/vfio/migration-multifd.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_MIGRATION_MULTIFD_H #define HW_VFIO_MIGRATION_MULTIFD_H -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" bool vfio_multifd_setup(VFIODevice *vbasedev, bool alloc_multifd, Error **errp); void vfio_multifd_cleanup(VFIODevice *vbasedev); diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.h +++ b/hw/vfio/pci.h @@ -XXX,XX +XXX,XX @@ #include "exec/memory.h" #include "hw/pci/pci_device.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/event_notifier.h" #include "qemu/queue.h" #include "qemu/timer.h" diff --git a/include/hw/s390x/vfio-ccw.h b/include/hw/s390x/vfio-ccw.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/s390x/vfio-ccw.h +++ b/include/hw/s390x/vfio-ccw.h @@ -XXX,XX +XXX,XX @@ #ifndef HW_VFIO_CCW_H #define HW_VFIO_CCW_H -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/ccw-device.h" #include "qom/object.h" diff --git a/include/hw/vfio/vfio-common.h b/include/hw/vfio/vfio-device.h similarity index 100% rename from include/hw/vfio/vfio-common.h rename to include/hw/vfio/vfio-device.h diff --git a/include/hw/vfio/vfio-platform.h b/include/hw/vfio/vfio-platform.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-platform.h +++ b/include/hw/vfio/vfio-platform.h @@ -XXX,XX +XXX,XX @@ #define HW_VFIO_VFIO_PLATFORM_H #include "hw/sysbus.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/event_notifier.h" #include "qemu/queue.h" #include "qom/object.h" diff --git a/backends/iommufd.c b/backends/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/backends/iommufd.c +++ b/backends/iommufd.c @@ -XXX,XX +XXX,XX @@ #include "qemu/error-report.h" #include "monitor/monitor.h" #include "trace.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include <sys/ioctl.h> #include <linux/iommufd.h> diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -XXX,XX +XXX,XX @@ #include <linux/vfio.h> #include <sys/ioctl.h> #include "qapi/error.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "system/iommufd.h" #include "hw/s390x/ap-device.h" #include "qemu/error-report.h" diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "qapi/error.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "system/iommufd.h" #include "hw/s390x/s390-ccw.h" #include "hw/s390x/vfio-ccw.h" diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "qemu/error-report.h" #include "hw/vfio/vfio-container-base.h" -#include "hw/vfio/vfio-common.h" /* vfio_reset_handler */ +#include "hw/vfio/vfio-device.h" /* vfio_reset_handler */ #include "system/reset.h" #include "vfio-helpers.h" diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include <linux/vfio.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "exec/address-spaces.h" #include "exec/memory.h" #include "exec/ram_addr.h" diff --git a/hw/vfio/cpr.c b/hw/vfio/cpr.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/cpr.c +++ b/hw/vfio/cpr.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "migration/misc.h" #include "qapi/error.h" #include "system/runstate.h" diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/pci.h" #include "hw/hw.h" #include "trace.h" diff --git a/hw/vfio/helpers.c b/hw/vfio/helpers.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/helpers.c +++ b/hw/vfio/helpers.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "system/kvm.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/hw.h" #include "qapi/error.h" #include "vfio-helpers.h" diff --git a/hw/vfio/iommufd.c b/hw/vfio/iommufd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/iommufd.c +++ b/hw/vfio/iommufd.c @@ -XXX,XX +XXX,XX @@ #include <linux/vfio.h> #include <linux/iommufd.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "qemu/error-report.h" #include "trace.h" #include "qapi/error.h" diff --git a/hw/vfio/listener.c b/hw/vfio/listener.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/listener.c +++ b/hw/vfio/listener.c @@ -XXX,XX +XXX,XX @@ #endif #include <linux/vfio.h> -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/pci.h" #include "exec/address-spaces.h" #include "exec/memory.h" diff --git a/hw/vfio/migration-multifd.c b/hw/vfio/migration-multifd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration-multifd.c +++ b/hw/vfio/migration-multifd.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "qapi/error.h" diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/migration.c +++ b/hw/vfio/migration.c @@ -XXX,XX +XXX,XX @@ #include <sys/ioctl.h> #include "system/runstate.h" -#include "hw/vfio/vfio-common.h" +#include "hw/vfio/vfio-device.h" #include "hw/vfio/vfio-migration.h" #include "migration/misc.h" #include "migration/savevm.h" diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include <sys/ioctl.h> -#include "hw/vfio/vfio-common.h" -#include "hw/vfio/pci.h" +#include "hw/vfio/vfio-region.h" +#include "hw/vfio/vfio-device.h" #include "hw/hw.h" #include "trace.h" #include "qapi/error.h" -- 2.49.0 Rename these routines : vfio_disable_irqindex -> vfio_device_irq_disable vfio_unmask_single_irqindex -> vfio_device_irq_unmask vfio_mask_single_irqindex -> vfio_device_irq_mask vfio_set_irq_signaling -> vfio_device_irq_set_signaling vfio_attach_device -> vfio_device_attach vfio_detach_device -> vfio_device_detach vfio_get_region_info -> vfio_device_get_region_info vfio_get_dev_region_info -> vfio_device_get_region_info_type vfio_has_region_cap -> vfio_device_has_region_cap vfio_reset_handler -> vfio_device_reset_hander to better reflect the namespace they belong to. Signed-off-by: Cédric Le Goater <clg@redhat.com> --- include/hw/vfio/vfio-device.h | 20 +++++++-------- hw/vfio/ap.c | 8 +++--- hw/vfio/ccw.c | 18 ++++++------- hw/vfio/container-base.c | 6 ++--- hw/vfio/container.c | 6 ++--- hw/vfio/device.c | 28 ++++++++++---------- hw/vfio/display.c | 2 +- hw/vfio/igd.c | 8 +++--- hw/vfio/pci.c | 48 +++++++++++++++++------------------ hw/vfio/platform.c | 10 ++++---- hw/vfio/region.c | 2 +- hw/vfio/trace-events | 8 +++--- 12 files changed, 82 insertions(+), 82 deletions(-) diff --git a/include/hw/vfio/vfio-device.h b/include/hw/vfio/vfio-device.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/vfio/vfio-device.h +++ b/include/hw/vfio/vfio-device.h @@ -XXX,XX +XXX,XX @@ struct VFIODeviceOps { int (*vfio_load_config)(VFIODevice *vdev, QEMUFile *f); }; -void vfio_disable_irqindex(VFIODevice *vbasedev, int index); -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index); -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index); -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, +void vfio_device_irq_disable(VFIODevice *vbasedev, int index); +void vfio_device_irq_unmask(VFIODevice *vbasedev, int index); +void vfio_device_irq_mask(VFIODevice *vbasedev, int index); +bool vfio_device_irq_set_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp); -void vfio_reset_handler(void *opaque); +void vfio_device_reset_handler(void *opaque); bool vfio_device_is_mdev(VFIODevice *vbasedev); bool vfio_device_hiod_realize(VFIODevice *vbasedev, Error **errp); -bool vfio_attach_device(char *name, VFIODevice *vbasedev, +bool vfio_device_attach(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp); -void vfio_detach_device(VFIODevice *vbasedev); +void vfio_device_detach(VFIODevice *vbasedev); VFIODevice *vfio_get_vfio_device(Object *obj); typedef QLIST_HEAD(VFIODeviceList, VFIODevice) VFIODeviceList; extern VFIODeviceList vfio_device_list; #ifdef CONFIG_LINUX -int vfio_get_region_info(VFIODevice *vbasedev, int index, +int vfio_device_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info); -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, +int vfio_device_get_region_info_type(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info); -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); +bool vfio_device_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type); #endif /* Returns 0 on success, or a negative errno. */ diff --git a/hw/vfio/ap.c b/hw/vfio/ap.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ap.c +++ b/hw/vfio/ap.c @@ -XXX,XX +XXX,XX @@ static bool vfio_ap_register_irq_notifier(VFIOAPDevice *vapdev, fd = event_notifier_get_fd(notifier); qemu_set_fd_handler(fd, fd_read, NULL, vapdev); - if (!vfio_set_irq_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, + if (!vfio_device_irq_set_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vapdev); event_notifier_cleanup(notifier); @@ -XXX,XX +XXX,XX @@ static void vfio_ap_unregister_irq_notifier(VFIOAPDevice *vapdev, return; } - if (!vfio_set_irq_signaling(&vapdev->vdev, irq, 0, + if (!vfio_device_irq_set_signaling(&vapdev->vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { warn_reportf_err(err, VFIO_MSG_PREFIX, vapdev->vdev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_ap_realize(DeviceState *dev, Error **errp) return; } - if (!vfio_attach_device(vbasedev->name, vbasedev, + if (!vfio_device_attach(vbasedev->name, vbasedev, &address_space_memory, errp)) { goto error; } @@ -XXX,XX +XXX,XX @@ static void vfio_ap_unrealize(DeviceState *dev) VFIOAPDevice *vapdev = VFIO_AP_DEVICE(dev); vfio_ap_unregister_irq_notifier(vapdev, VFIO_AP_REQ_IRQ_INDEX); - vfio_detach_device(&vapdev->vdev); + vfio_device_detach(&vapdev->vdev); g_free(vapdev->vdev.name); } diff --git a/hw/vfio/ccw.c b/hw/vfio/ccw.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/ccw.c +++ b/hw/vfio/ccw.c @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_register_irq_notifier(VFIOCCWDevice *vcdev, fd = event_notifier_get_fd(notifier); qemu_set_fd_handler(fd, fd_read, NULL, vcdev); - if (!vfio_set_irq_signaling(vdev, irq, 0, + if (!vfio_device_irq_set_signaling(vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vcdev); event_notifier_cleanup(notifier); @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_unregister_irq_notifier(VFIOCCWDevice *vcdev, return; } - if (!vfio_set_irq_signaling(&vcdev->vdev, irq, 0, + if (!vfio_device_irq_set_signaling(&vcdev->vdev, irq, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { warn_reportf_err(err, VFIO_MSG_PREFIX, vcdev->vdev.name); } @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) return false; } - ret = vfio_get_region_info(vdev, VFIO_CCW_CONFIG_REGION_INDEX, &info); + ret = vfio_device_get_region_info(vdev, VFIO_CCW_CONFIG_REGION_INDEX, &info); if (ret) { error_setg_errno(errp, -ret, "vfio: Error getting config info"); return false; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); /* check for the optional async command region */ - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_ASYNC_CMD, &info); if (!ret) { vcdev->async_cmd_region_size = info->size; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); } - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_SCHIB, &info); if (!ret) { vcdev->schib_region_size = info->size; @@ -XXX,XX +XXX,XX @@ static bool vfio_ccw_get_region(VFIOCCWDevice *vcdev, Error **errp) g_free(info); } - ret = vfio_get_dev_region_info(vdev, VFIO_REGION_TYPE_CCW, + ret = vfio_device_get_region_info_type(vdev, VFIO_REGION_TYPE_CCW, VFIO_REGION_SUBTYPE_CCW_CRW, &info); if (!ret) { @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_realize(DeviceState *dev, Error **errp) goto out_unrealize; } - if (!vfio_attach_device(cdev->mdevid, vbasedev, + if (!vfio_device_attach(cdev->mdevid, vbasedev, &address_space_memory, errp)) { goto out_attach_dev_err; } @@ -XXX,XX +XXX,XX @@ out_irq_notifier_err: out_io_notifier_err: vfio_ccw_put_region(vcdev); out_region_err: - vfio_detach_device(vbasedev); + vfio_device_detach(vbasedev); out_attach_dev_err: g_free(vbasedev->name); out_unrealize: @@ -XXX,XX +XXX,XX @@ static void vfio_ccw_unrealize(DeviceState *dev) vfio_ccw_unregister_irq_notifier(vcdev, VFIO_CCW_CRW_IRQ_INDEX); vfio_ccw_unregister_irq_notifier(vcdev, VFIO_CCW_IO_IRQ_INDEX); vfio_ccw_put_region(vcdev); - vfio_detach_device(&vcdev->vdev); + vfio_device_detach(&vcdev->vdev); g_free(vcdev->vdev.name); if (cdc->unrealize) { diff --git a/hw/vfio/container-base.c b/hw/vfio/container-base.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container-base.c +++ b/hw/vfio/container-base.c @@ -XXX,XX +XXX,XX @@ #include "qapi/error.h" #include "qemu/error-report.h" #include "hw/vfio/vfio-container-base.h" -#include "hw/vfio/vfio-device.h" /* vfio_reset_handler */ +#include "hw/vfio/vfio-device.h" /* vfio_device_reset_handler */ #include "system/reset.h" #include "vfio-helpers.h" @@ -XXX,XX +XXX,XX @@ VFIOAddressSpace *vfio_address_space_get(AddressSpace *as) QLIST_INIT(&space->containers); if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_register_reset(vfio_reset_handler, NULL); + qemu_register_reset(vfio_device_reset_handler, NULL); } QLIST_INSERT_HEAD(&vfio_address_spaces, space, list); @@ -XXX,XX +XXX,XX @@ void vfio_address_space_put(VFIOAddressSpace *space) g_free(space); if (QLIST_EMPTY(&vfio_address_spaces)) { - qemu_unregister_reset(vfio_reset_handler, NULL); + qemu_unregister_reset(vfio_device_reset_handler, NULL); } } diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) } /* - * vfio_attach_device: attach a device to a security context + * vfio_device_attach: attach a device to a security context * @name and @vbasedev->name are likely to be different depending * on the type of the device, hence the need for passing @name */ @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, return false; } - trace_vfio_attach_device(vbasedev->name, groupid); + trace_vfio_device_attach(vbasedev->name, groupid); if (!vfio_device_hiod_realize(vbasedev, errp)) { return false; @@ -XXX,XX +XXX,XX @@ static void vfio_legacy_detach_device(VFIODevice *vbasedev) QLIST_REMOVE(vbasedev, global_next); QLIST_REMOVE(vbasedev, container_next); vbasedev->bcontainer = NULL; - trace_vfio_detach_device(vbasedev->name, group->groupid); + trace_vfio_device_detach(vbasedev->name, group->groupid); vfio_put_base_device(vbasedev); vfio_put_group(group); } diff --git a/hw/vfio/device.c b/hw/vfio/device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/device.c +++ b/hw/vfio/device.c @@ -XXX,XX +XXX,XX @@ VFIODeviceList vfio_device_list = * reset for the one in-use devices case, calling _multi() will do * nothing if a _one() would have been sufficient. */ -void vfio_reset_handler(void *opaque) +void vfio_device_reset_handler(void *opaque) { VFIODevice *vbasedev; - trace_vfio_reset_handler(); + trace_vfio_device_reset_handler(); QLIST_FOREACH(vbasedev, &vfio_device_list, global_next) { if (vbasedev->dev->realized) { vbasedev->ops->vfio_compute_needs_reset(vbasedev); @@ -XXX,XX +XXX,XX @@ void vfio_reset_handler(void *opaque) /* * Common VFIO interrupt disable */ -void vfio_disable_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_disable(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ void vfio_disable_irqindex(VFIODevice *vbasedev, int index) ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); } -void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_unmask(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ void vfio_unmask_single_irqindex(VFIODevice *vbasedev, int index) ioctl(vbasedev->fd, VFIO_DEVICE_SET_IRQS, &irq_set); } -void vfio_mask_single_irqindex(VFIODevice *vbasedev, int index) +void vfio_device_irq_mask(VFIODevice *vbasedev, int index) { struct vfio_irq_set irq_set = { .argsz = sizeof(irq_set), @@ -XXX,XX +XXX,XX @@ static const char *index_to_str(VFIODevice *vbasedev, int index) } } -bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, +bool vfio_device_irq_set_signaling(VFIODevice *vbasedev, int index, int subindex, int action, int fd, Error **errp) { ERRP_GUARD(); @@ -XXX,XX +XXX,XX @@ bool vfio_set_irq_signaling(VFIODevice *vbasedev, int index, int subindex, return false; } -int vfio_get_region_info(VFIODevice *vbasedev, int index, +int vfio_device_get_region_info(VFIODevice *vbasedev, int index, struct vfio_region_info **info) { size_t argsz = sizeof(struct vfio_region_info); @@ -XXX,XX +XXX,XX @@ retry: return 0; } -int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, +int vfio_device_get_region_info_type(VFIODevice *vbasedev, uint32_t type, uint32_t subtype, struct vfio_region_info **info) { int i; @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, struct vfio_info_cap_header *hdr; struct vfio_region_info_cap_type *cap_type; - if (vfio_get_region_info(vbasedev, i, info)) { + if (vfio_device_get_region_info(vbasedev, i, info)) { continue; } @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, cap_type = container_of(hdr, struct vfio_region_info_cap_type, header); - trace_vfio_get_dev_region(vbasedev->name, i, + trace_vfio_device_get_region_info_type(vbasedev->name, i, cap_type->type, cap_type->subtype); if (cap_type->type == type && cap_type->subtype == subtype) { @@ -XXX,XX +XXX,XX @@ int vfio_get_dev_region_info(VFIODevice *vbasedev, uint32_t type, return -ENODEV; } -bool vfio_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) +bool vfio_device_has_region_cap(VFIODevice *vbasedev, int region, uint16_t cap_type) { g_autofree struct vfio_region_info *info = NULL; bool ret = false; - if (!vfio_get_region_info(vbasedev, region, &info)) { + if (!vfio_device_get_region_info(vbasedev, region, &info)) { if (vfio_get_region_info_cap(info, cap_type)) { ret = true; } @@ -XXX,XX +XXX,XX @@ VFIODevice *vfio_get_vfio_device(Object *obj) } } -bool vfio_attach_device(char *name, VFIODevice *vbasedev, +bool vfio_device_attach(char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { const VFIOIOMMUClass *ops = @@ -XXX,XX +XXX,XX @@ bool vfio_attach_device(char *name, VFIODevice *vbasedev, return true; } -void vfio_detach_device(VFIODevice *vbasedev) +void vfio_device_detach(VFIODevice *vbasedev) { if (!vbasedev->bcontainer) { return; diff --git a/hw/vfio/display.c b/hw/vfio/display.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/display.c +++ b/hw/vfio/display.c @@ -XXX,XX +XXX,XX @@ static bool vfio_display_edid_init(VFIOPCIDevice *vdev, Error **errp) int fd = vdev->vbasedev.fd; int ret; - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_GFX, VFIO_REGION_SUBTYPE_GFX_EDID, &dpy->edid_info); diff --git a/hw/vfio/igd.c b/hw/vfio/igd.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/igd.c +++ b/hw/vfio/igd.c @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_opregion(VFIOPCIDevice *vdev, Error **errp) return false; } - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_OPREGION, &opregion); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_lpc_bridge(VFIOPCIDevice *vdev, Error **errp) * Check whether we have all the vfio device specific regions to * support LPC quirk (added in Linux v4.6). */ - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_LPC_CFG, &lpc); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_setup_lpc_bridge(VFIOPCIDevice *vdev, Error **errp) return false; } - ret = vfio_get_dev_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info_type(&vdev->vbasedev, VFIO_REGION_TYPE_PCI_VENDOR_TYPE | PCI_VENDOR_ID_INTEL, VFIO_REGION_SUBTYPE_INTEL_IGD_HOST_CFG, &host); if (ret) { @@ -XXX,XX +XXX,XX @@ static bool vfio_pci_igd_config_quirk(VFIOPCIDevice *vdev, Error **errp) * there's no ROM, there's no point in setting up this quirk. * NB. We only seem to get BIOS ROMs, so UEFI VM would need CSM support. */ - ret = vfio_get_region_info(&vdev->vbasedev, + ret = vfio_device_get_region_info(&vdev->vbasedev, VFIO_PCI_ROM_REGION_INDEX, &rom); if ((ret || !rom->size) && !vdev->pdev.romfile) { error_setg(&err, "Device has no ROM"); diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/pci.c +++ b/hw/vfio/pci.c @@ -XXX,XX +XXX,XX @@ static void vfio_intx_eoi(VFIODevice *vbasedev) vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); - vfio_unmask_single_irqindex(vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(vbasedev, VFIO_PCI_INTX_IRQ_INDEX); } static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) /* Get to a known interrupt state */ qemu_set_fd_handler(irq_fd, NULL, NULL, vdev); - vfio_mask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_mask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) goto fail_irqfd; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_UNMASK, event_notifier_get_fd(&vdev->intx.unmask), errp)) { @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable_kvm(VFIOPCIDevice *vdev, Error **errp) } /* Let'em rip */ - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.kvm_accel = true; @@ -XXX,XX +XXX,XX @@ fail_irqfd: event_notifier_cleanup(&vdev->intx.unmask); fail: qemu_set_fd_handler(irq_fd, vfio_intx_interrupt, NULL, vdev); - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); return false; #else return true; @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable_kvm(VFIOPCIDevice *vdev) * Get to a known state, hardware masked, QEMU ready to accept new * interrupts, QEMU IRQ de-asserted. */ - vfio_mask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_mask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable_kvm(VFIOPCIDevice *vdev) vdev->intx.kvm_accel = false; /* If we've missed an event, let it re-fire through QEMU */ - vfio_unmask_single_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_unmask(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); trace_vfio_intx_disable_kvm(vdev->vbasedev.name); #endif @@ -XXX,XX +XXX,XX @@ static bool vfio_intx_enable(VFIOPCIDevice *vdev, Error **errp) fd = event_notifier_get_fd(&vdev->intx.interrupt); qemu_set_fd_handler(fd, vfio_intx_interrupt, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, errp)) { qemu_set_fd_handler(fd, NULL, NULL, vdev); event_notifier_cleanup(&vdev->intx.interrupt); @@ -XXX,XX +XXX,XX @@ static void vfio_intx_disable(VFIOPCIDevice *vdev) timer_del(vdev->intx.mmap_timer); vfio_intx_disable_kvm(vdev); - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_INTX_IRQ_INDEX); vdev->intx.pending = false; pci_irq_deassert(&vdev->pdev); vfio_mmap_set_enabled(vdev, true); @@ -XXX,XX +XXX,XX @@ static int vfio_msix_vector_do_use(PCIDevice *pdev, unsigned int nr, if (!vdev->defer_kvm_irq_routing) { if (vdev->msix->noresize && resizing) { - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); ret = vfio_enable_vectors(vdev, true); if (ret) { error_report("vfio: failed to enable vectors, %d", ret); @@ -XXX,XX +XXX,XX @@ static int vfio_msix_vector_do_use(PCIDevice *pdev, unsigned int nr, fd = event_notifier_get_fd(&vector->interrupt); } - if (!vfio_set_irq_signaling(&vdev->vbasedev, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, nr, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { @@ -XXX,XX +XXX,XX @@ static void vfio_msix_vector_release(PCIDevice *pdev, unsigned int nr) int32_t fd = event_notifier_get_fd(&vector->interrupt); Error *err = NULL; - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX, nr, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); @@ -XXX,XX +XXX,XX @@ static void vfio_msix_disable(VFIOPCIDevice *vdev) * Always clear MSI-X IRQ index. A PF device could have enabled * MSI-X with no vectors. See vfio_msix_enable(). */ - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSIX_IRQ_INDEX); vfio_msi_disable_common(vdev); if (!vfio_intx_enable(vdev, &err)) { @@ -XXX,XX +XXX,XX @@ static void vfio_msi_disable(VFIOPCIDevice *vdev) { Error *err = NULL; - vfio_disable_irqindex(&vdev->vbasedev, VFIO_PCI_MSI_IRQ_INDEX); + vfio_device_irq_disable(&vdev->vbasedev, VFIO_PCI_MSI_IRQ_INDEX); vfio_msi_disable_common(vdev); vfio_intx_enable(vdev, &err); if (err) { @@ -XXX,XX +XXX,XX @@ static void vfio_pci_load_rom(VFIOPCIDevice *vdev) off_t off = 0; ssize_t bytes; - if (vfio_get_region_info(&vdev->vbasedev, + if (vfio_device_get_region_info(&vdev->vbasedev, VFIO_PCI_ROM_REGION_INDEX, &reg_info)) { error_report("vfio: Error getting ROM info: %m"); return; @@ -XXX,XX +XXX,XX @@ static void vfio_pci_fixup_msix_region(VFIOPCIDevice *vdev) * If the host driver allows mapping of a MSIX data, we are going to * do map the entire BAR and emulate MSIX table on top of that. */ - if (vfio_has_region_cap(&vdev->vbasedev, region->nr, + if (vfio_device_has_region_cap(&vdev->vbasedev, region->nr, VFIO_REGION_INFO_CAP_MSIX_MAPPABLE)) { return; } @@ -XXX,XX +XXX,XX @@ bool vfio_populate_vga(VFIOPCIDevice *vdev, Error **errp) g_autofree struct vfio_region_info *reg_info = NULL; int ret; - ret = vfio_get_region_info(vbasedev, VFIO_PCI_VGA_REGION_INDEX, &reg_info); + ret = vfio_device_get_region_info(vbasedev, VFIO_PCI_VGA_REGION_INDEX, &reg_info); if (ret) { error_setg_errno(errp, -ret, "failed getting region info for VGA region index %d", @@ -XXX,XX +XXX,XX @@ static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp) QLIST_INIT(&vdev->bars[i].quirks); } - ret = vfio_get_region_info(vbasedev, + ret = vfio_device_get_region_info(vbasedev, VFIO_PCI_CONFIG_REGION_INDEX, &reg_info); if (ret) { error_setg_errno(errp, -ret, "failed to get config info"); @@ -XXX,XX +XXX,XX @@ static bool vfio_populate_device(VFIOPCIDevice *vdev, Error **errp) static void vfio_pci_put_device(VFIOPCIDevice *vdev) { - vfio_detach_device(&vdev->vbasedev); + vfio_device_detach(&vdev->vbasedev); g_free(vdev->vbasedev.name); g_free(vdev->msix); @@ -XXX,XX +XXX,XX @@ static void vfio_register_err_notifier(VFIOPCIDevice *vdev) fd = event_notifier_get_fd(&vdev->err_notifier); qemu_set_fd_handler(fd, vfio_err_notifier_handler, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); qemu_set_fd_handler(fd, NULL, NULL, vdev); @@ -XXX,XX +XXX,XX @@ static void vfio_unregister_err_notifier(VFIOPCIDevice *vdev) return; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_ERR_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_register_req_notifier(VFIOPCIDevice *vdev) fd = event_notifier_get_fd(&vdev->req_notifier); qemu_set_fd_handler(fd, vfio_req_notifier_handler, NULL, vdev); - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); qemu_set_fd_handler(fd, NULL, NULL, vdev); @@ -XXX,XX +XXX,XX @@ static void vfio_unregister_req_notifier(VFIOPCIDevice *vdev) return; } - if (!vfio_set_irq_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, + if (!vfio_device_irq_set_signaling(&vdev->vbasedev, VFIO_PCI_REQ_IRQ_INDEX, 0, VFIO_IRQ_SET_ACTION_TRIGGER, -1, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vdev->vbasedev.name); } @@ -XXX,XX +XXX,XX @@ static void vfio_realize(PCIDevice *pdev, Error **errp) name = g_strdup(vbasedev->name); } - if (!vfio_attach_device(name, vbasedev, + if (!vfio_device_attach(name, vbasedev, pci_device_iommu_address_space(pdev), errp)) { goto error; } diff --git a/hw/vfio/platform.c b/hw/vfio/platform.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/platform.c +++ b/hw/vfio/platform.c @@ -XXX,XX +XXX,XX @@ static int vfio_set_trigger_eventfd(VFIOINTp *intp, qemu_set_fd_handler(fd, (IOHandler *)handler, NULL, intp); - if (!vfio_set_irq_signaling(vbasedev, intp->pin, 0, + if (!vfio_device_irq_set_signaling(vbasedev, intp->pin, 0, VFIO_IRQ_SET_ACTION_TRIGGER, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vbasedev->name); qemu_set_fd_handler(fd, NULL, NULL, NULL); @@ -XXX,XX +XXX,XX @@ static void vfio_platform_eoi(VFIODevice *vbasedev) if (vfio_irq_is_automasked(intp)) { /* unmasks the physical level-sensitive IRQ */ - vfio_unmask_single_irqindex(vbasedev, intp->pin); + vfio_device_irq_unmask(vbasedev, intp->pin); } /* a single IRQ can be active at a time */ @@ -XXX,XX +XXX,XX @@ static int vfio_set_resample_eventfd(VFIOINTp *intp) Error *err = NULL; qemu_set_fd_handler(fd, NULL, NULL, NULL); - if (!vfio_set_irq_signaling(vbasedev, intp->pin, 0, + if (!vfio_device_irq_set_signaling(vbasedev, intp->pin, 0, VFIO_IRQ_SET_ACTION_UNMASK, fd, &err)) { error_reportf_err(err, VFIO_MSG_PREFIX, vbasedev->name); return -EINVAL; @@ -XXX,XX +XXX,XX @@ static bool vfio_base_device_init(VFIODevice *vbasedev, Error **errp) return false; } - if (!vfio_attach_device(vbasedev->name, vbasedev, + if (!vfio_device_attach(vbasedev->name, vbasedev, &address_space_memory, errp)) { return false; } @@ -XXX,XX +XXX,XX @@ static bool vfio_base_device_init(VFIODevice *vbasedev, Error **errp) return true; } - vfio_detach_device(vbasedev); + vfio_device_detach(vbasedev); return false; } diff --git a/hw/vfio/region.c b/hw/vfio/region.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/region.c +++ b/hw/vfio/region.c @@ -XXX,XX +XXX,XX @@ int vfio_region_setup(Object *obj, VFIODevice *vbasedev, VFIORegion *region, g_autofree struct vfio_region_info *info = NULL; int ret; - ret = vfio_get_region_info(vbasedev, index, &info); + ret = vfio_device_get_region_info(vbasedev, index, &info); if (ret) { return ret; } diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int vfio_pci_hot_reset_result(const char *name, const char *result) "%s hot reset: %s" vfio_populate_device_config(const char *name, unsigned long size, unsigned long offset, unsigned long flags) "Device '%s' config: size: 0x%lx, offset: 0x%lx, flags: 0x%lx" vfio_populate_device_get_irq_info_failure(const char *errstr) "VFIO_DEVICE_GET_IRQ_INFO failure: %s" -vfio_attach_device(const char *name, int group_id) " (%s) group %d" -vfio_detach_device(const char *name, int group_id) " (%s) group %d" vfio_mdev(const char *name, bool is_mdev) " (%s) is_mdev %d" vfio_add_ext_cap_dropped(const char *name, uint16_t cap, uint16_t offset) "%s 0x%x@0x%x" vfio_pci_reset(const char *name) " (%s)" @@ -XXX,XX +XXX,XX @@ iommufd_cdev_device_info(char *name, int devfd, int num_irqs, int num_regions, i iommufd_cdev_pci_hot_reset_dep_devices(int domain, int bus, int slot, int function, int dev_id) "\t%04x:%02x:%02x.%x devid %d" # device.c -vfio_get_dev_region(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" -vfio_reset_handler(void) "" +vfio_device_get_region_info_type(const char *name, int index, uint32_t type, uint32_t subtype) "%s index %d, %08x/%08x" +vfio_device_reset_handler(void) "" +vfio_device_attach(const char *name, int group_id) " (%s) group %d" +vfio_device_detach(const char *name, int group_id) " (%s) group %d" -- 2.49.0 Rename these routines : vfio_put_group -> vfio_group_put vfio_get_group -> vfio_group_get vfio_kvm_device_del_group -> vfio_group_del_kvm_device vfio_kvm_device_add_group -> vfio_group_add_kvm_device vfio_get_device -> vfio_device_get vfio_put_base_device -> vfio_device_put vfio_device_groupid -> vfio_device_get_groupid vfio_connect_container -> vfio_container_connect vfio_disconnect_container -> vfio_container_disconnect to better reflect the namespace they belong to. Reviewed-by: John Levon <john.levon@nutanix.com> Link: https://lore.kernel.org/qemu-devel/20250318095415.670319-30-clg@redhat.com Signed-off-by: Cédric Le Goater <clg@redhat.com> --- hw/vfio/container.c | 52 ++++++++++++++++++++++---------------------- hw/vfio/trace-events | 12 +++++----- 2 files changed, 33 insertions(+), 31 deletions(-) diff --git a/hw/vfio/container.c b/hw/vfio/container.c index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/container.c +++ b/hw/vfio/container.c @@ -XXX,XX +XXX,XX @@ static bool vfio_get_info_iova_range(struct vfio_iommu_type1_info *info, return true; } -static void vfio_kvm_device_add_group(VFIOGroup *group) +static void vfio_group_add_kvm_device(VFIOGroup *group) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static void vfio_kvm_device_add_group(VFIOGroup *group) } } -static void vfio_kvm_device_del_group(VFIOGroup *group) +static void vfio_group_del_kvm_device(VFIOGroup *group) { Error *err = NULL; @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_setup(VFIOContainerBase *bcontainer, Error **errp) return true; } -static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, +static bool vfio_container_connect(VFIOGroup *group, AddressSpace *as, Error **errp) { VFIOContainer *container; @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, } group->container = container; QLIST_INSERT_HEAD(&container->group_list, group, container_next); - vfio_kvm_device_add_group(group); + vfio_group_add_kvm_device(group); return true; } } @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, goto enable_discards_exit; } - vfio_kvm_device_add_group(group); + vfio_group_add_kvm_device(group); vfio_address_space_insert(space, bcontainer); @@ -XXX,XX +XXX,XX @@ static bool vfio_connect_container(VFIOGroup *group, AddressSpace *as, return true; listener_release_exit: QLIST_REMOVE(group, container_next); - vfio_kvm_device_del_group(group); + vfio_group_del_kvm_device(group); vfio_listener_unregister(bcontainer); if (vioc->release) { vioc->release(bcontainer); @@ -XXX,XX +XXX,XX @@ put_space_exit: return false; } -static void vfio_disconnect_container(VFIOGroup *group) +static void vfio_container_disconnect(VFIOGroup *group) { VFIOContainer *container = group->container; VFIOContainerBase *bcontainer = &container->bcontainer; @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) if (QLIST_EMPTY(&container->group_list)) { VFIOAddressSpace *space = bcontainer->space; - trace_vfio_disconnect_container(container->fd); + trace_vfio_container_disconnect(container->fd); vfio_cpr_unregister_container(bcontainer); close(container->fd); object_unref(container); @@ -XXX,XX +XXX,XX @@ static void vfio_disconnect_container(VFIOGroup *group) } } -static VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp) +static VFIOGroup *vfio_group_get(int groupid, AddressSpace *as, Error **errp) { ERRP_GUARD(); VFIOGroup *group; @@ -XXX,XX +XXX,XX @@ static VFIOGroup *vfio_get_group(int groupid, AddressSpace *as, Error **errp) group->groupid = groupid; QLIST_INIT(&group->device_list); - if (!vfio_connect_container(group, as, errp)) { + if (!vfio_container_connect(group, as, errp)) { error_prepend(errp, "failed to setup container for group %d: ", groupid); goto close_fd_exit; @@ -XXX,XX +XXX,XX @@ free_group_exit: return NULL; } -static void vfio_put_group(VFIOGroup *group) +static void vfio_group_put(VFIOGroup *group) { if (!group || !QLIST_EMPTY(&group->device_list)) { return; @@ -XXX,XX +XXX,XX @@ static void vfio_put_group(VFIOGroup *group) if (!group->ram_block_discard_allowed) { vfio_ram_block_discard_disable(group->container, false); } - vfio_kvm_device_del_group(group); - vfio_disconnect_container(group); + vfio_group_del_kvm_device(group); + vfio_container_disconnect(group); QLIST_REMOVE(group, next); - trace_vfio_put_group(group->fd); + trace_vfio_group_put(group->fd); close(group->fd); g_free(group); } -static bool vfio_get_device(VFIOGroup *group, const char *name, +static bool vfio_device_get(VFIOGroup *group, const char *name, VFIODevice *vbasedev, Error **errp) { g_autofree struct vfio_device_info *info = NULL; @@ -XXX,XX +XXX,XX @@ static bool vfio_get_device(VFIOGroup *group, const char *name, vbasedev->num_regions = info->num_regions; vbasedev->flags = info->flags; - trace_vfio_get_device(name, info->flags, info->num_regions, info->num_irqs); + trace_vfio_device_get(name, info->flags, info->num_regions, info->num_irqs); vbasedev->reset_works = !!(info->flags & VFIO_DEVICE_FLAGS_RESET); return true; } -static void vfio_put_base_device(VFIODevice *vbasedev) +static void vfio_device_put(VFIODevice *vbasedev) { if (!vbasedev->group) { return; } QLIST_REMOVE(vbasedev, next); vbasedev->group = NULL; - trace_vfio_put_base_device(vbasedev->fd); + trace_vfio_device_put(vbasedev->fd); close(vbasedev->fd); } -static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) +static int vfio_device_get_groupid(VFIODevice *vbasedev, Error **errp) { char *tmp, group_path[PATH_MAX]; g_autofree char *group_name = NULL; @@ -XXX,XX +XXX,XX @@ static int vfio_device_groupid(VFIODevice *vbasedev, Error **errp) static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, AddressSpace *as, Error **errp) { - int groupid = vfio_device_groupid(vbasedev, errp); + int groupid = vfio_device_get_groupid(vbasedev, errp); VFIODevice *vbasedev_iter; VFIOGroup *group; VFIOContainerBase *bcontainer; @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, return false; } - group = vfio_get_group(groupid, as, errp); + group = vfio_group_get(groupid, as, errp); if (!group) { return false; } @@ -XXX,XX +XXX,XX @@ static bool vfio_legacy_attach_device(const char *name, VFIODevice *vbasedev, QLIST_FOREACH(vbasedev_iter, &group->device_list, next) { if (strcmp(vbasedev_iter->name, vbasedev->name) == 0) { error_setg(errp, "device is already attached"); - vfio_put_group(group); + vfio_group_put(group); return false; } } - if (!vfio_get_device(group, name, vbasedev, errp)) { - vfio_put_group(group); + if (!vfio_device_get(group, name, vbasedev, errp)) { + vfio_group_put(group); return false; } @@ -XXX,XX +XXX,XX @@ static void vfio_legacy_detach_device(VFIODevice *vbasedev) QLIST_REMOVE(vbasedev, container_next); vbasedev->bcontainer = NULL; trace_vfio_device_detach(vbasedev->name, group->groupid); - vfio_put_base_device(vbasedev); - vfio_put_group(group); + vfio_device_put(vbasedev); + vfio_group_put(group); } static int vfio_legacy_pci_hot_reset(VFIODevice *vbasedev, bool single) diff --git a/hw/vfio/trace-events b/hw/vfio/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/vfio/trace-events +++ b/hw/vfio/trace-events @@ -XXX,XX +XXX,XX @@ vfio_listener_region_add_no_dma_map(const char *name, uint64_t iova, uint64_t si vfio_listener_region_del(uint64_t start, uint64_t end) "region_del 0x%"PRIx64" - 0x%"PRIx64 vfio_device_dirty_tracking_update(uint64_t start, uint64_t end, uint64_t min, uint64_t max) "section 0x%"PRIx64" - 0x%"PRIx64" -> update [0x%"PRIx64" - 0x%"PRIx64"]" vfio_device_dirty_tracking_start(int nr_ranges, uint64_t min32, uint64_t max32, uint64_t min64, uint64_t max64, uint64_t minpci, uint64_t maxpci) "nr_ranges %d 32:[0x%"PRIx64" - 0x%"PRIx64"], 64:[0x%"PRIx64" - 0x%"PRIx64"], pci64:[0x%"PRIx64" - 0x%"PRIx64"]" -vfio_disconnect_container(int fd) "close container->fd=%d" -vfio_put_group(int fd) "close group->fd=%d" -vfio_get_device(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" -vfio_put_base_device(int fd) "close vdev->fd=%d" -vfio_legacy_dma_unmap_overflow_workaround(void) "" vfio_iommu_map_dirty_notify(uint64_t iova_start, uint64_t iova_end) "iommu dirty @ 0x%"PRIx64" - 0x%"PRIx64 # container-base.c vfio_container_query_dirty_bitmap(uint64_t iova, uint64_t size, uint64_t bitmap_size, uint64_t start, uint64_t dirty_pages) "iova=0x%"PRIx64" size= 0x%"PRIx64" bitmap_size=0x%"PRIx64" start=0x%"PRIx64" dirty_pages=%"PRIu64 +# container.c +vfio_container_disconnect(int fd) "close container->fd=%d" +vfio_group_put(int fd) "close group->fd=%d" +vfio_device_get(const char * name, unsigned int flags, unsigned int num_regions, unsigned int num_irqs) "Device %s flags: %u, regions: %u, irqs: %u" +vfio_device_put(int fd) "close vdev->fd=%d" +vfio_legacy_dma_unmap_overflow_workaround(void) "" + # region.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64 -- 2.49.0