* [PATCH v17 0/2] add debugfs to migration driver @ 2023-10-13 9:04 Longfang Liu 2023-10-13 9:04 ` [PATCH v17 1/2] vfio/migration: Add debugfs to live " Longfang Liu 2023-10-13 9:04 ` [PATCH v17 2/2] Documentation: add debugfs description for vfio Longfang Liu 0 siblings, 2 replies; 8+ messages in thread From: Longfang Liu @ 2023-10-13 9:04 UTC (permalink / raw) To: alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm, liulongfang Add a debugfs function to the migration driver in VFIO to provide a step-by-step debugfs information for the migration driver. Changes v16 -> v17 Add separate VFIO_DEBUGFS Kconfig entries. Changes v15 -> v16 Update the calling order of functions to maintain symmetry Changes v14 -> v15 Update the output status value of live migration. Changes v13 -> v14 Split the patchset and keep the vfio debugfs frame. Changes v12 -> v13 Solve the problem of open and close competition to debugfs. Changes v11 -> v12 Update loading conditions of vfio debugfs. Changes v10 -> v11 Delete the device restore function in debugfs. Changes v9 -> v10 Update the debugfs file of the live migration driver. Changes v8 -> v9 Update the debugfs directory structure of vfio. Changes v7 -> v8 Add support for platform devices. Changes v6 -> v7 Fix some code style issues. Changes v5 -> v6 Control the creation of debugfs through the CONFIG_DEBUG_FS. Changes v4 -> v5 Remove the newly added vfio_migration_ops and use seq_printf to optimize the implementation of debugfs. Changes v3 -> v4 Change the migration_debug_operate interface to debug_root file. Changes v2 -> v3 Extend the debugfs function from hisilicon device to vfio. Changes v1 -> v2 Change the registration method of root_debugfs to register with module initialization. Longfang Liu (2): vfio/migration: Add debugfs to live migration driver Documentation: add debugfs description for vfio Documentation/ABI/testing/debugfs-vfio | 25 +++++++ MAINTAINERS | 1 + drivers/vfio/Kconfig | 10 +++ drivers/vfio/Makefile | 1 + drivers/vfio/debugfs.c | 90 ++++++++++++++++++++++++++ drivers/vfio/vfio.h | 14 ++++ drivers/vfio/vfio_main.c | 14 +++- include/linux/vfio.h | 7 ++ include/uapi/linux/vfio.h | 1 + 9 files changed, 161 insertions(+), 2 deletions(-) create mode 100644 Documentation/ABI/testing/debugfs-vfio create mode 100644 drivers/vfio/debugfs.c -- 2.24.0 ^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-13 9:04 [PATCH v17 0/2] add debugfs to migration driver Longfang Liu @ 2023-10-13 9:04 ` Longfang Liu 2023-10-16 15:17 ` Cédric Le Goater 2023-10-13 9:04 ` [PATCH v17 2/2] Documentation: add debugfs description for vfio Longfang Liu 1 sibling, 1 reply; 8+ messages in thread From: Longfang Liu @ 2023-10-13 9:04 UTC (permalink / raw) To: alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm, liulongfang There are multiple devices, software and operational steps involved in the process of live migration. An error occurred on any node may cause the live migration operation to fail. This complex process makes it very difficult to locate and analyze the cause when the function fails. In order to quickly locate the cause of the problem when the live migration fails, I added a set of debugfs to the vfio live migration driver. +-------------------------------------------+ | | | | | QEMU | | | | | +---+----------------------------+----------+ | ^ | ^ | | | | | | | | v | v | +---------+--+ +---------+--+ |src vfio_dev| |dst vfio_dev| +--+---------+ +--+---------+ | ^ | ^ | | | | v | | | +-----------+----+ +-----------+----+ |src dev debugfs | |dst dev debugfs | +----------------+ +----------------+ The entire debugfs directory will be based on the definition of the CONFIG_DEBUG_FS macro. If this macro is not enabled, the interfaces in vfio.h will be empty definitions, and the creation and initialization of the debugfs directory will not be executed. vfio | +---<dev_name1> | +---migration | +--state | +---<dev_name2> +---migration +--state debugfs will create a public root directory "vfio" file. then create a dev_name() file for each live migration device. First, create a unified state acquisition file of "migration" in this device directory. Then, create a public live migration state lookup file "state". Signed-off-by: Longfang Liu <liulongfang@huawei.com> --- drivers/vfio/Kconfig | 10 +++++ drivers/vfio/Makefile | 1 + drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ drivers/vfio/vfio.h | 14 ++++++ drivers/vfio/vfio_main.c | 14 +++++- include/linux/vfio.h | 7 +++ include/uapi/linux/vfio.h | 1 + 7 files changed, 135 insertions(+), 2 deletions(-) create mode 100644 drivers/vfio/debugfs.c diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig index 6bda6dbb4878..ceae52fd7586 100644 --- a/drivers/vfio/Kconfig +++ b/drivers/vfio/Kconfig @@ -80,6 +80,16 @@ config VFIO_VIRQFD select EVENTFD default n +config VFIO_DEBUGFS + bool "Export VFIO internals in DebugFS" + depends on DEBUG_FS + help + Allows exposure of VFIO device internals. This option enables + the use of debugfs by VFIO drivers as required. The device can + cause the VFIO code create a top-level debug/vfio directory + during initialization, and then populate a subdirectory with + entries as required. + source "drivers/vfio/pci/Kconfig" source "drivers/vfio/platform/Kconfig" source "drivers/vfio/mdev/Kconfig" diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile index c82ea032d352..d43a699d55b1 100644 --- a/drivers/vfio/Makefile +++ b/drivers/vfio/Makefile @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o vfio-$(CONFIG_IOMMUFD) += iommufd.o vfio-$(CONFIG_VFIO_CONTAINER) += container.o vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c new file mode 100644 index 000000000000..ae53d6110f47 --- /dev/null +++ b/drivers/vfio/debugfs.c @@ -0,0 +1,90 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (c) 2023, HiSilicon Ltd. + */ + +#include <linux/device.h> +#include <linux/debugfs.h> +#include <linux/seq_file.h> +#include <linux/vfio.h> +#include "vfio.h" + +static struct dentry *vfio_debugfs_root; + +static int vfio_device_state_read(struct seq_file *seq, void *data) +{ + struct device *vf_dev = seq->private; + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); + enum vfio_device_mig_state state; + int ret; + + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); + + ret = vdev->mig_ops->migration_get_state(vdev, &state); + if (ret) + return -EINVAL; + + switch (state) { + case VFIO_DEVICE_STATE_ERROR: + seq_printf(seq, "%s\n", "ERROR"); + break; + case VFIO_DEVICE_STATE_STOP: + seq_printf(seq, "%s\n", "STOP"); + break; + case VFIO_DEVICE_STATE_RUNNING: + seq_printf(seq, "%s\n", "RUNNING"); + break; + case VFIO_DEVICE_STATE_STOP_COPY: + seq_printf(seq, "%s\n", "STOP_COPY"); + break; + case VFIO_DEVICE_STATE_RESUMING: + seq_printf(seq, "%s\n", "RESUMING"); + break; + case VFIO_DEVICE_STATE_RUNNING_P2P: + seq_printf(seq, "%s\n", "RUNNING_P2P"); + break; + case VFIO_DEVICE_STATE_PRE_COPY: + seq_printf(seq, "%s\n", "PRE_COPY"); + break; + case VFIO_DEVICE_STATE_PRE_COPY_P2P: + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); + break; + default: + seq_printf(seq, "%s\n", "Invalid"); + } + + return 0; +} + +void vfio_device_debugfs_init(struct vfio_device *vdev) +{ + struct device *dev = &vdev->device; + + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); + + if (vdev->mig_ops) { + struct dentry *vfio_dev_migration = NULL; + + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, + vfio_device_state_read); + } +} + +void vfio_device_debugfs_exit(struct vfio_device *vdev) +{ + debugfs_remove_recursive(vdev->debug_root); +} + +void vfio_debugfs_create_root(void) +{ + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); +} + +void vfio_debugfs_remove_root(void) +{ + debugfs_remove_recursive(vfio_debugfs_root); + vfio_debugfs_root = NULL; +} + diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h index 307e3f29b527..bde84ad344e5 100644 --- a/drivers/vfio/vfio.h +++ b/drivers/vfio/vfio.h @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) } #endif +#ifdef CONFIG_VFIO_DEBUGFS +void vfio_debugfs_create_root(void); +void vfio_debugfs_remove_root(void); + +void vfio_device_debugfs_init(struct vfio_device *vdev); +void vfio_device_debugfs_exit(struct vfio_device *vdev); +#else +static inline void vfio_debugfs_create_root(void) { } +static inline void vfio_debugfs_remove_root(void) { } + +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } +#endif /* CONFIG_VFIO_DEBUGFS */ + #endif diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c index e31e1952d7b8..9aec4c22f051 100644 --- a/drivers/vfio/vfio_main.c +++ b/drivers/vfio/vfio_main.c @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, /* Refcounting can't start until the driver calls register */ refcount_set(&device->refcount, 1); - vfio_device_group_register(device); return 0; @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, int vfio_register_group_dev(struct vfio_device *device) { - return __vfio_register_dev(device, VFIO_IOMMU); + int ret; + + ret = __vfio_register_dev(device, VFIO_IOMMU); + if (ret) + return ret; + + vfio_device_debugfs_init(device); + + return 0; } EXPORT_SYMBOL_GPL(vfio_register_group_dev); @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) } } + vfio_device_debugfs_exit(device); /* Balances vfio_device_set_group in register path */ vfio_device_remove_group(device); } @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) if (ret) goto err_alloc_dev_chrdev; + vfio_debugfs_create_root(); pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); return 0; @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) static void __exit vfio_cleanup(void) { + vfio_debugfs_remove_root(); ida_destroy(&vfio.device_ida); vfio_cdev_cleanup(); class_destroy(vfio.device_class); diff --git a/include/linux/vfio.h b/include/linux/vfio.h index 454e9295970c..769d7af86225 100644 --- a/include/linux/vfio.h +++ b/include/linux/vfio.h @@ -69,6 +69,13 @@ struct vfio_device { u8 iommufd_attached:1; #endif u8 cdev_opened:1; +#ifdef CONFIG_DEBUG_FS + /* + * debug_root is a static property of the vfio_device + * which must be set prior to registering the vfio_device. + */ + struct dentry *debug_root; +#endif }; /** diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index 7f5fb010226d..2b68e6cdf190 100644 --- a/include/uapi/linux/vfio.h +++ b/include/uapi/linux/vfio.h @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { VFIO_DEVICE_STATE_RUNNING_P2P = 5, VFIO_DEVICE_STATE_PRE_COPY = 6, VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, + VFIO_DEVICE_STATE_NR, }; /** -- 2.24.0 ^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-13 9:04 ` [PATCH v17 1/2] vfio/migration: Add debugfs to live " Longfang Liu @ 2023-10-16 15:17 ` Cédric Le Goater 2023-10-19 8:03 ` liulongfang 0 siblings, 1 reply; 8+ messages in thread From: Cédric Le Goater @ 2023-10-16 15:17 UTC (permalink / raw) To: Longfang Liu, alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm Hello Longfang, On 10/13/23 11:04, Longfang Liu wrote: > There are multiple devices, software and operational steps involved > in the process of live migration. An error occurred on any node may > cause the live migration operation to fail. > This complex process makes it very difficult to locate and analyze > the cause when the function fails. > > In order to quickly locate the cause of the problem when the > live migration fails, I added a set of debugfs to the vfio > live migration driver. > > +-------------------------------------------+ > | | > | | > | QEMU | > | | > | | > +---+----------------------------+----------+ > | ^ | ^ > | | | | > | | | | > v | v | > +---------+--+ +---------+--+ > |src vfio_dev| |dst vfio_dev| > +--+---------+ +--+---------+ > | ^ | ^ > | | | | > v | | | > +-----------+----+ +-----------+----+ > |src dev debugfs | |dst dev debugfs | > +----------------+ +----------------+ > > The entire debugfs directory will be based on the definition of > the CONFIG_DEBUG_FS macro. If this macro is not enabled, the > interfaces in vfio.h will be empty definitions, and the creation > and initialization of the debugfs directory will not be executed. > > vfio > | > +---<dev_name1> > | +---migration > | +--state > | > +---<dev_name2> > +---migration > +--state > > debugfs will create a public root directory "vfio" file. > then create a dev_name() file for each live migration device. > First, create a unified state acquisition file of "migration" > in this device directory. > Then, create a public live migration state lookup file "state". > > Signed-off-by: Longfang Liu <liulongfang@huawei.com> > --- > drivers/vfio/Kconfig | 10 +++++ > drivers/vfio/Makefile | 1 + > drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ > drivers/vfio/vfio.h | 14 ++++++ > drivers/vfio/vfio_main.c | 14 +++++- > include/linux/vfio.h | 7 +++ > include/uapi/linux/vfio.h | 1 + > 7 files changed, 135 insertions(+), 2 deletions(-) > create mode 100644 drivers/vfio/debugfs.c > > diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig > index 6bda6dbb4878..ceae52fd7586 100644 > --- a/drivers/vfio/Kconfig > +++ b/drivers/vfio/Kconfig > @@ -80,6 +80,16 @@ config VFIO_VIRQFD > select EVENTFD > default n > > +config VFIO_DEBUGFS > + bool "Export VFIO internals in DebugFS" > + depends on DEBUG_FS > + help > + Allows exposure of VFIO device internals. This option enables > + the use of debugfs by VFIO drivers as required. The device can > + cause the VFIO code create a top-level debug/vfio directory > + during initialization, and then populate a subdirectory with > + entries as required. > + > source "drivers/vfio/pci/Kconfig" > source "drivers/vfio/platform/Kconfig" > source "drivers/vfio/mdev/Kconfig" > diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile > index c82ea032d352..d43a699d55b1 100644 > --- a/drivers/vfio/Makefile > +++ b/drivers/vfio/Makefile > @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o > vfio-$(CONFIG_IOMMUFD) += iommufd.o > vfio-$(CONFIG_VFIO_CONTAINER) += container.o > vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o > +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o > > obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o > obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o > diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c > new file mode 100644 > index 000000000000..ae53d6110f47 > --- /dev/null > +++ b/drivers/vfio/debugfs.c > @@ -0,0 +1,90 @@ > +// SPDX-License-Identifier: GPL-2.0-only > +/* > + * Copyright (c) 2023, HiSilicon Ltd. > + */ > + > +#include <linux/device.h> > +#include <linux/debugfs.h> > +#include <linux/seq_file.h> > +#include <linux/vfio.h> > +#include "vfio.h" > + > +static struct dentry *vfio_debugfs_root; > + > +static int vfio_device_state_read(struct seq_file *seq, void *data) > +{ > + struct device *vf_dev = seq->private; > + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); > + enum vfio_device_mig_state state; > + int ret; > + > + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != > + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); > + > + ret = vdev->mig_ops->migration_get_state(vdev, &state); > + if (ret) > + return -EINVAL; > + > + switch (state) { > + case VFIO_DEVICE_STATE_ERROR: > + seq_printf(seq, "%s\n", "ERROR"); > + break; > + case VFIO_DEVICE_STATE_STOP: > + seq_printf(seq, "%s\n", "STOP"); > + break; > + case VFIO_DEVICE_STATE_RUNNING: > + seq_printf(seq, "%s\n", "RUNNING"); > + break; > + case VFIO_DEVICE_STATE_STOP_COPY: > + seq_printf(seq, "%s\n", "STOP_COPY"); > + break; > + case VFIO_DEVICE_STATE_RESUMING: > + seq_printf(seq, "%s\n", "RESUMING"); > + break; > + case VFIO_DEVICE_STATE_RUNNING_P2P: > + seq_printf(seq, "%s\n", "RUNNING_P2P"); > + break; > + case VFIO_DEVICE_STATE_PRE_COPY: > + seq_printf(seq, "%s\n", "PRE_COPY"); > + break; > + case VFIO_DEVICE_STATE_PRE_COPY_P2P: > + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); > + break; > + default: > + seq_printf(seq, "%s\n", "Invalid"); seq_puts() is more appropriate than seq_printf() above. I would suggest to add an array or some helper, that the VFIO drivers could use to debug the migration flow with pr_* primitives. It can be done later. > + } > + > + return 0; > +} > + > +void vfio_device_debugfs_init(struct vfio_device *vdev) > +{ > + struct device *dev = &vdev->device; > + > + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); > + > + if (vdev->mig_ops) { > + struct dentry *vfio_dev_migration = NULL; mig_dir maybe ? It would be easier to understand the nature of the variable IMHO. > + > + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); > + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, > + vfio_device_state_read); > + } > +} > + > +void vfio_device_debugfs_exit(struct vfio_device *vdev) > +{ > + debugfs_remove_recursive(vdev->debug_root); > +} > + > +void vfio_debugfs_create_root(void) > +{ > + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); > +} > + > +void vfio_debugfs_remove_root(void) > +{ > + debugfs_remove_recursive(vfio_debugfs_root); > + vfio_debugfs_root = NULL; > +} > + > diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h > index 307e3f29b527..bde84ad344e5 100644 > --- a/drivers/vfio/vfio.h > +++ b/drivers/vfio/vfio.h > @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) > } > #endif > > +#ifdef CONFIG_VFIO_DEBUGFS > +void vfio_debugfs_create_root(void); > +void vfio_debugfs_remove_root(void); > + > +void vfio_device_debugfs_init(struct vfio_device *vdev); > +void vfio_device_debugfs_exit(struct vfio_device *vdev); > +#else > +static inline void vfio_debugfs_create_root(void) { } > +static inline void vfio_debugfs_remove_root(void) { } > + > +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } > +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } > +#endif /* CONFIG_VFIO_DEBUGFS */ > + > #endif > diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c > index e31e1952d7b8..9aec4c22f051 100644 > --- a/drivers/vfio/vfio_main.c > +++ b/drivers/vfio/vfio_main.c > @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, > > /* Refcounting can't start until the driver calls register */ > refcount_set(&device->refcount, 1); > - superfluous change. > vfio_device_group_register(device); > > return 0; > @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, > > int vfio_register_group_dev(struct vfio_device *device) > { > - return __vfio_register_dev(device, VFIO_IOMMU); > + int ret; > + > + ret = __vfio_register_dev(device, VFIO_IOMMU); > + if (ret) > + return ret; > + > + vfio_device_debugfs_init(device); Can it be called from __vfio_register_dev() instead ? and mdev devices would get debugfs support also. Thanks, C. > + > + return 0; > } > EXPORT_SYMBOL_GPL(vfio_register_group_dev); > > @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) > } > } > > + vfio_device_debugfs_exit(device); > /* Balances vfio_device_set_group in register path */ > vfio_device_remove_group(device); > } > @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) > if (ret) > goto err_alloc_dev_chrdev; > > + vfio_debugfs_create_root(); > pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); > return 0; > > @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) > > static void __exit vfio_cleanup(void) > { > + vfio_debugfs_remove_root(); > ida_destroy(&vfio.device_ida); > vfio_cdev_cleanup(); > class_destroy(vfio.device_class); > diff --git a/include/linux/vfio.h b/include/linux/vfio.h > index 454e9295970c..769d7af86225 100644 > --- a/include/linux/vfio.h > +++ b/include/linux/vfio.h > @@ -69,6 +69,13 @@ struct vfio_device { > u8 iommufd_attached:1; > #endif > u8 cdev_opened:1; > +#ifdef CONFIG_DEBUG_FS > + /* > + * debug_root is a static property of the vfio_device > + * which must be set prior to registering the vfio_device. > + */ > + struct dentry *debug_root; > +#endif > }; > > /** > diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h > index 7f5fb010226d..2b68e6cdf190 100644 > --- a/include/uapi/linux/vfio.h > +++ b/include/uapi/linux/vfio.h > @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { > VFIO_DEVICE_STATE_RUNNING_P2P = 5, > VFIO_DEVICE_STATE_PRE_COPY = 6, > VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, > + VFIO_DEVICE_STATE_NR, > }; > > /** ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-16 15:17 ` Cédric Le Goater @ 2023-10-19 8:03 ` liulongfang 2023-10-19 8:30 ` Cédric Le Goater 0 siblings, 1 reply; 8+ messages in thread From: liulongfang @ 2023-10-19 8:03 UTC (permalink / raw) To: Cédric Le Goater, alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm On 2023/10/16 23:17, Cédric Le Goater wrote: > Hello Longfang, > > On 10/13/23 11:04, Longfang Liu wrote: >> There are multiple devices, software and operational steps involved >> in the process of live migration. An error occurred on any node may >> cause the live migration operation to fail. >> This complex process makes it very difficult to locate and analyze >> the cause when the function fails. >> >> In order to quickly locate the cause of the problem when the >> live migration fails, I added a set of debugfs to the vfio >> live migration driver. >> >> +-------------------------------------------+ >> | | >> | | >> | QEMU | >> | | >> | | >> +---+----------------------------+----------+ >> | ^ | ^ >> | | | | >> | | | | >> v | v | >> +---------+--+ +---------+--+ >> |src vfio_dev| |dst vfio_dev| >> +--+---------+ +--+---------+ >> | ^ | ^ >> | | | | >> v | | | >> +-----------+----+ +-----------+----+ >> |src dev debugfs | |dst dev debugfs | >> +----------------+ +----------------+ >> >> The entire debugfs directory will be based on the definition of >> the CONFIG_DEBUG_FS macro. If this macro is not enabled, the >> interfaces in vfio.h will be empty definitions, and the creation >> and initialization of the debugfs directory will not be executed. >> >> vfio >> | >> +---<dev_name1> >> | +---migration >> | +--state >> | >> +---<dev_name2> >> +---migration >> +--state >> >> debugfs will create a public root directory "vfio" file. >> then create a dev_name() file for each live migration device. >> First, create a unified state acquisition file of "migration" >> in this device directory. >> Then, create a public live migration state lookup file "state". >> >> Signed-off-by: Longfang Liu <liulongfang@huawei.com> >> --- >> drivers/vfio/Kconfig | 10 +++++ >> drivers/vfio/Makefile | 1 + >> drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ >> drivers/vfio/vfio.h | 14 ++++++ >> drivers/vfio/vfio_main.c | 14 +++++- >> include/linux/vfio.h | 7 +++ >> include/uapi/linux/vfio.h | 1 + >> 7 files changed, 135 insertions(+), 2 deletions(-) >> create mode 100644 drivers/vfio/debugfs.c >> >> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig >> index 6bda6dbb4878..ceae52fd7586 100644 >> --- a/drivers/vfio/Kconfig >> +++ b/drivers/vfio/Kconfig >> @@ -80,6 +80,16 @@ config VFIO_VIRQFD >> select EVENTFD >> default n >> +config VFIO_DEBUGFS >> + bool "Export VFIO internals in DebugFS" >> + depends on DEBUG_FS >> + help >> + Allows exposure of VFIO device internals. This option enables >> + the use of debugfs by VFIO drivers as required. The device can >> + cause the VFIO code create a top-level debug/vfio directory >> + during initialization, and then populate a subdirectory with >> + entries as required. >> + >> source "drivers/vfio/pci/Kconfig" >> source "drivers/vfio/platform/Kconfig" >> source "drivers/vfio/mdev/Kconfig" >> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile >> index c82ea032d352..d43a699d55b1 100644 >> --- a/drivers/vfio/Makefile >> +++ b/drivers/vfio/Makefile >> @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o >> vfio-$(CONFIG_IOMMUFD) += iommufd.o >> vfio-$(CONFIG_VFIO_CONTAINER) += container.o >> vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o >> +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o >> obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o >> obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o >> diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c >> new file mode 100644 >> index 000000000000..ae53d6110f47 >> --- /dev/null >> +++ b/drivers/vfio/debugfs.c >> @@ -0,0 +1,90 @@ >> +// SPDX-License-Identifier: GPL-2.0-only >> +/* >> + * Copyright (c) 2023, HiSilicon Ltd. >> + */ >> + >> +#include <linux/device.h> >> +#include <linux/debugfs.h> >> +#include <linux/seq_file.h> >> +#include <linux/vfio.h> >> +#include "vfio.h" >> + >> +static struct dentry *vfio_debugfs_root; >> + >> +static int vfio_device_state_read(struct seq_file *seq, void *data) >> +{ >> + struct device *vf_dev = seq->private; >> + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); >> + enum vfio_device_mig_state state; >> + int ret; >> + >> + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != >> + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); >> + >> + ret = vdev->mig_ops->migration_get_state(vdev, &state); >> + if (ret) >> + return -EINVAL; >> + >> + switch (state) { >> + case VFIO_DEVICE_STATE_ERROR: >> + seq_printf(seq, "%s\n", "ERROR"); >> + break; >> + case VFIO_DEVICE_STATE_STOP: >> + seq_printf(seq, "%s\n", "STOP"); >> + break; >> + case VFIO_DEVICE_STATE_RUNNING: >> + seq_printf(seq, "%s\n", "RUNNING"); >> + break; >> + case VFIO_DEVICE_STATE_STOP_COPY: >> + seq_printf(seq, "%s\n", "STOP_COPY"); >> + break; >> + case VFIO_DEVICE_STATE_RESUMING: >> + seq_printf(seq, "%s\n", "RESUMING"); >> + break; >> + case VFIO_DEVICE_STATE_RUNNING_P2P: >> + seq_printf(seq, "%s\n", "RUNNING_P2P"); >> + break; >> + case VFIO_DEVICE_STATE_PRE_COPY: >> + seq_printf(seq, "%s\n", "PRE_COPY"); >> + break; >> + case VFIO_DEVICE_STATE_PRE_COPY_P2P: >> + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); >> + break; >> + default: >> + seq_printf(seq, "%s\n", "Invalid"); > > seq_puts() is more appropriate than seq_printf() above. > There is no difference between seq_puts() and seq_printf() here, no need to modify it. > I would suggest to add an array or some helper, that the VFIO drivers > could use to debug the migration flow with pr_* primitives. It can be > done later. > If you want to debug this migration process in the VFIO driver, you can refer to vdev->mig_ops->migration_get_state() to read the status. > >> + } >> + >> + return 0; >> +} >> + >> +void vfio_device_debugfs_init(struct vfio_device *vdev) >> +{ >> + struct device *dev = &vdev->device; >> + >> + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); >> + >> + if (vdev->mig_ops) { >> + struct dentry *vfio_dev_migration = NULL; > > mig_dir maybe ? > "vfio_dev_migration " will not affect the readability of the code. > It would be easier to understand the nature of the variable IMHO. > >> + >> + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); >> + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, >> + vfio_device_state_read); >> + } >> +} >> + >> +void vfio_device_debugfs_exit(struct vfio_device *vdev) >> +{ >> + debugfs_remove_recursive(vdev->debug_root); >> +} >> + >> +void vfio_debugfs_create_root(void) >> +{ >> + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); >> +} >> + >> +void vfio_debugfs_remove_root(void) >> +{ >> + debugfs_remove_recursive(vfio_debugfs_root); >> + vfio_debugfs_root = NULL; >> +} >> + >> diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h >> index 307e3f29b527..bde84ad344e5 100644 >> --- a/drivers/vfio/vfio.h >> +++ b/drivers/vfio/vfio.h >> @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) >> } >> #endif >> +#ifdef CONFIG_VFIO_DEBUGFS >> +void vfio_debugfs_create_root(void); >> +void vfio_debugfs_remove_root(void); >> + >> +void vfio_device_debugfs_init(struct vfio_device *vdev); >> +void vfio_device_debugfs_exit(struct vfio_device *vdev); >> +#else >> +static inline void vfio_debugfs_create_root(void) { } >> +static inline void vfio_debugfs_remove_root(void) { } >> + >> +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } >> +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } >> +#endif /* CONFIG_VFIO_DEBUGFS */ >> + >> #endif >> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c >> index e31e1952d7b8..9aec4c22f051 100644 >> --- a/drivers/vfio/vfio_main.c >> +++ b/drivers/vfio/vfio_main.c >> @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, >> /* Refcounting can't start until the driver calls register */ >> refcount_set(&device->refcount, 1); >> - > > superfluous change. > A blank line here is to separate it from the comment above. Makes it easier to be read. >> vfio_device_group_register(device); >> return 0; >> @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, >> int vfio_register_group_dev(struct vfio_device *device) >> { >> - return __vfio_register_dev(device, VFIO_IOMMU); >> + int ret; >> + >> + ret = __vfio_register_dev(device, VFIO_IOMMU); >> + if (ret) >> + return ret; >> + >> + vfio_device_debugfs_init(device); > > Can it be called from __vfio_register_dev() instead ? and mdev devices > would get debugfs support also. > This is for symmetry in function calls. The need for symmetry was mentioned in the previous review. > Thanks, > > C. > Thanks, Longfang. >> + >> + return 0; >> } >> EXPORT_SYMBOL_GPL(vfio_register_group_dev); >> @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) >> } >> } >> + vfio_device_debugfs_exit(device); >> /* Balances vfio_device_set_group in register path */ >> vfio_device_remove_group(device); >> } >> @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) >> if (ret) >> goto err_alloc_dev_chrdev; >> + vfio_debugfs_create_root(); >> pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); >> return 0; >> @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) >> static void __exit vfio_cleanup(void) >> { >> + vfio_debugfs_remove_root(); >> ida_destroy(&vfio.device_ida); >> vfio_cdev_cleanup(); >> class_destroy(vfio.device_class); >> diff --git a/include/linux/vfio.h b/include/linux/vfio.h >> index 454e9295970c..769d7af86225 100644 >> --- a/include/linux/vfio.h >> +++ b/include/linux/vfio.h >> @@ -69,6 +69,13 @@ struct vfio_device { >> u8 iommufd_attached:1; >> #endif >> u8 cdev_opened:1; >> +#ifdef CONFIG_DEBUG_FS >> + /* >> + * debug_root is a static property of the vfio_device >> + * which must be set prior to registering the vfio_device. >> + */ >> + struct dentry *debug_root; >> +#endif >> }; >> /** >> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h >> index 7f5fb010226d..2b68e6cdf190 100644 >> --- a/include/uapi/linux/vfio.h >> +++ b/include/uapi/linux/vfio.h >> @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { >> VFIO_DEVICE_STATE_RUNNING_P2P = 5, >> VFIO_DEVICE_STATE_PRE_COPY = 6, >> VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, >> + VFIO_DEVICE_STATE_NR, >> }; >> /** > > . > ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-19 8:03 ` liulongfang @ 2023-10-19 8:30 ` Cédric Le Goater 2023-10-19 22:37 ` Alex Williamson 2023-10-20 9:38 ` liulongfang 0 siblings, 2 replies; 8+ messages in thread From: Cédric Le Goater @ 2023-10-19 8:30 UTC (permalink / raw) To: liulongfang, alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm On 10/19/23 10:03, liulongfang wrote: > On 2023/10/16 23:17, Cédric Le Goater wrote: >> Hello Longfang, >> >> On 10/13/23 11:04, Longfang Liu wrote: >>> There are multiple devices, software and operational steps involved >>> in the process of live migration. An error occurred on any node may >>> cause the live migration operation to fail. >>> This complex process makes it very difficult to locate and analyze >>> the cause when the function fails. >>> >>> In order to quickly locate the cause of the problem when the >>> live migration fails, I added a set of debugfs to the vfio >>> live migration driver. >>> >>> +-------------------------------------------+ >>> | | >>> | | >>> | QEMU | >>> | | >>> | | >>> +---+----------------------------+----------+ >>> | ^ | ^ >>> | | | | >>> | | | | >>> v | v | >>> +---------+--+ +---------+--+ >>> |src vfio_dev| |dst vfio_dev| >>> +--+---------+ +--+---------+ >>> | ^ | ^ >>> | | | | >>> v | | | >>> +-----------+----+ +-----------+----+ >>> |src dev debugfs | |dst dev debugfs | >>> +----------------+ +----------------+ >>> >>> The entire debugfs directory will be based on the definition of >>> the CONFIG_DEBUG_FS macro. If this macro is not enabled, the >>> interfaces in vfio.h will be empty definitions, and the creation >>> and initialization of the debugfs directory will not be executed. >>> >>> vfio >>> | >>> +---<dev_name1> >>> | +---migration >>> | +--state >>> | >>> +---<dev_name2> >>> +---migration >>> +--state >>> >>> debugfs will create a public root directory "vfio" file. >>> then create a dev_name() file for each live migration device. >>> First, create a unified state acquisition file of "migration" >>> in this device directory. >>> Then, create a public live migration state lookup file "state". >>> >>> Signed-off-by: Longfang Liu <liulongfang@huawei.com> >>> --- >>> drivers/vfio/Kconfig | 10 +++++ >>> drivers/vfio/Makefile | 1 + >>> drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ >>> drivers/vfio/vfio.h | 14 ++++++ >>> drivers/vfio/vfio_main.c | 14 +++++- >>> include/linux/vfio.h | 7 +++ >>> include/uapi/linux/vfio.h | 1 + >>> 7 files changed, 135 insertions(+), 2 deletions(-) >>> create mode 100644 drivers/vfio/debugfs.c >>> >>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig >>> index 6bda6dbb4878..ceae52fd7586 100644 >>> --- a/drivers/vfio/Kconfig >>> +++ b/drivers/vfio/Kconfig >>> @@ -80,6 +80,16 @@ config VFIO_VIRQFD >>> select EVENTFD >>> default n >>> +config VFIO_DEBUGFS >>> + bool "Export VFIO internals in DebugFS" >>> + depends on DEBUG_FS >>> + help >>> + Allows exposure of VFIO device internals. This option enables >>> + the use of debugfs by VFIO drivers as required. The device can >>> + cause the VFIO code create a top-level debug/vfio directory >>> + during initialization, and then populate a subdirectory with >>> + entries as required. >>> + >>> source "drivers/vfio/pci/Kconfig" >>> source "drivers/vfio/platform/Kconfig" >>> source "drivers/vfio/mdev/Kconfig" >>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile >>> index c82ea032d352..d43a699d55b1 100644 >>> --- a/drivers/vfio/Makefile >>> +++ b/drivers/vfio/Makefile >>> @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o >>> vfio-$(CONFIG_IOMMUFD) += iommufd.o >>> vfio-$(CONFIG_VFIO_CONTAINER) += container.o >>> vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o >>> +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o >>> obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o >>> obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o >>> diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c >>> new file mode 100644 >>> index 000000000000..ae53d6110f47 >>> --- /dev/null >>> +++ b/drivers/vfio/debugfs.c >>> @@ -0,0 +1,90 @@ >>> +// SPDX-License-Identifier: GPL-2.0-only >>> +/* >>> + * Copyright (c) 2023, HiSilicon Ltd. >>> + */ >>> + >>> +#include <linux/device.h> >>> +#include <linux/debugfs.h> >>> +#include <linux/seq_file.h> >>> +#include <linux/vfio.h> >>> +#include "vfio.h" >>> + >>> +static struct dentry *vfio_debugfs_root; >>> + >>> +static int vfio_device_state_read(struct seq_file *seq, void *data) >>> +{ >>> + struct device *vf_dev = seq->private; >>> + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); >>> + enum vfio_device_mig_state state; >>> + int ret; >>> + >>> + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != >>> + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); >>> + >>> + ret = vdev->mig_ops->migration_get_state(vdev, &state); >>> + if (ret) >>> + return -EINVAL; >>> + >>> + switch (state) { >>> + case VFIO_DEVICE_STATE_ERROR: >>> + seq_printf(seq, "%s\n", "ERROR"); >>> + break; >>> + case VFIO_DEVICE_STATE_STOP: >>> + seq_printf(seq, "%s\n", "STOP"); >>> + break; >>> + case VFIO_DEVICE_STATE_RUNNING: >>> + seq_printf(seq, "%s\n", "RUNNING"); >>> + break; >>> + case VFIO_DEVICE_STATE_STOP_COPY: >>> + seq_printf(seq, "%s\n", "STOP_COPY"); >>> + break; >>> + case VFIO_DEVICE_STATE_RESUMING: >>> + seq_printf(seq, "%s\n", "RESUMING"); >>> + break; >>> + case VFIO_DEVICE_STATE_RUNNING_P2P: >>> + seq_printf(seq, "%s\n", "RUNNING_P2P"); >>> + break; >>> + case VFIO_DEVICE_STATE_PRE_COPY: >>> + seq_printf(seq, "%s\n", "PRE_COPY"); >>> + break; >>> + case VFIO_DEVICE_STATE_PRE_COPY_P2P: >>> + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); >>> + break; >>> + default: >>> + seq_printf(seq, "%s\n", "Invalid"); >> >> seq_puts() is more appropriate than seq_printf() above. >> > > There is no difference between seq_puts() and seq_printf() here, > no need to modify it. seq_puts is simply preferred for unformatted output. >> I would suggest to add an array or some helper, that the VFIO drivers >> could use to debug the migration flow with pr_* primitives. It can be >> done later. >> > > If you want to debug this migration process in the VFIO driver, > you can refer to vdev->mig_ops->migration_get_state() to read the status. > >> >>> + } >>> + >>> + return 0; >>> +} >>> + >>> +void vfio_device_debugfs_init(struct vfio_device *vdev) >>> +{ >>> + struct device *dev = &vdev->device; >>> + >>> + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); >>> + >>> + if (vdev->mig_ops) { >>> + struct dentry *vfio_dev_migration = NULL; >> >> mig_dir maybe ? >> > > "vfio_dev_migration " will not affect the readability of the code. > >> It would be easier to understand the nature of the variable IMHO. >> >>> + >>> + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); >>> + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, >>> + vfio_device_state_read); >>> + } >>> +} >>> + >>> +void vfio_device_debugfs_exit(struct vfio_device *vdev) >>> +{ >>> + debugfs_remove_recursive(vdev->debug_root); >>> +} >>> + >>> +void vfio_debugfs_create_root(void) >>> +{ >>> + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); >>> +} >>> + >>> +void vfio_debugfs_remove_root(void) >>> +{ >>> + debugfs_remove_recursive(vfio_debugfs_root); >>> + vfio_debugfs_root = NULL; >>> +} >>> + >>> diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h >>> index 307e3f29b527..bde84ad344e5 100644 >>> --- a/drivers/vfio/vfio.h >>> +++ b/drivers/vfio/vfio.h >>> @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) >>> } >>> #endif >>> +#ifdef CONFIG_VFIO_DEBUGFS >>> +void vfio_debugfs_create_root(void); >>> +void vfio_debugfs_remove_root(void); >>> + >>> +void vfio_device_debugfs_init(struct vfio_device *vdev); >>> +void vfio_device_debugfs_exit(struct vfio_device *vdev); >>> +#else >>> +static inline void vfio_debugfs_create_root(void) { } >>> +static inline void vfio_debugfs_remove_root(void) { } >>> + >>> +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } >>> +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } >>> +#endif /* CONFIG_VFIO_DEBUGFS */ >>> + >>> #endif >>> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c >>> index e31e1952d7b8..9aec4c22f051 100644 >>> --- a/drivers/vfio/vfio_main.c >>> +++ b/drivers/vfio/vfio_main.c >>> @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, >>> /* Refcounting can't start until the driver calls register */ >>> refcount_set(&device->refcount, 1); >>> - >> >> superfluous change. >> > > A blank line here is to separate it from the comment above. > Makes it easier to be read. > >>> vfio_device_group_register(device); >>> return 0; >>> @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, >>> int vfio_register_group_dev(struct vfio_device *device) >>> { >>> - return __vfio_register_dev(device, VFIO_IOMMU); >>> + int ret; >>> + >>> + ret = __vfio_register_dev(device, VFIO_IOMMU); >>> + if (ret) >>> + return ret; >>> + >>> + vfio_device_debugfs_init(device); >> >> Can it be called from __vfio_register_dev() instead ? and mdev devices >> would get debugfs support also. >> > > This is for symmetry in function calls. > The need for symmetry was mentioned in the previous review. yes. But this is also exluding the mdev devices which is a large VFIO family. Thanks, C. > >> Thanks, >> >> C. >> > Thanks, > Longfang. > >>> + >>> + return 0; >>> } >>> EXPORT_SYMBOL_GPL(vfio_register_group_dev); >>> @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) >>> } >>> } >>> + vfio_device_debugfs_exit(device); >>> /* Balances vfio_device_set_group in register path */ >>> vfio_device_remove_group(device); >>> } >>> @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) >>> if (ret) >>> goto err_alloc_dev_chrdev; >>> + vfio_debugfs_create_root(); >>> pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); >>> return 0; >>> @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) >>> static void __exit vfio_cleanup(void) >>> { >>> + vfio_debugfs_remove_root(); >>> ida_destroy(&vfio.device_ida); >>> vfio_cdev_cleanup(); >>> class_destroy(vfio.device_class); >>> diff --git a/include/linux/vfio.h b/include/linux/vfio.h >>> index 454e9295970c..769d7af86225 100644 >>> --- a/include/linux/vfio.h >>> +++ b/include/linux/vfio.h >>> @@ -69,6 +69,13 @@ struct vfio_device { >>> u8 iommufd_attached:1; >>> #endif >>> u8 cdev_opened:1; >>> +#ifdef CONFIG_DEBUG_FS >>> + /* >>> + * debug_root is a static property of the vfio_device >>> + * which must be set prior to registering the vfio_device. >>> + */ >>> + struct dentry *debug_root; >>> +#endif >>> }; >>> /** >>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h >>> index 7f5fb010226d..2b68e6cdf190 100644 >>> --- a/include/uapi/linux/vfio.h >>> +++ b/include/uapi/linux/vfio.h >>> @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { >>> VFIO_DEVICE_STATE_RUNNING_P2P = 5, >>> VFIO_DEVICE_STATE_PRE_COPY = 6, >>> VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, >>> + VFIO_DEVICE_STATE_NR, >>> }; >>> /** >> >> . >> > ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-19 8:30 ` Cédric Le Goater @ 2023-10-19 22:37 ` Alex Williamson 2023-10-20 9:38 ` liulongfang 1 sibling, 0 replies; 8+ messages in thread From: Alex Williamson @ 2023-10-19 22:37 UTC (permalink / raw) To: Cédric Le Goater Cc: liulongfang, jgg, shameerali.kolothum.thodi, jonathan.cameron, bcreeley, kvm, linux-kernel, linuxarm On Thu, 19 Oct 2023 10:30:53 +0200 Cédric Le Goater <clegoate@redhat.com> wrote: > On 10/19/23 10:03, liulongfang wrote: > > On 2023/10/16 23:17, Cédric Le Goater wrote: > >> Hello Longfang, > >> > >> On 10/13/23 11:04, Longfang Liu wrote: > >>> There are multiple devices, software and operational steps involved > >>> in the process of live migration. An error occurred on any node may > >>> cause the live migration operation to fail. > >>> This complex process makes it very difficult to locate and analyze > >>> the cause when the function fails. > >>> > >>> In order to quickly locate the cause of the problem when the > >>> live migration fails, I added a set of debugfs to the vfio > >>> live migration driver. > >>> > >>> +-------------------------------------------+ > >>> | | > >>> | | > >>> | QEMU | > >>> | | > >>> | | > >>> +---+----------------------------+----------+ > >>> | ^ | ^ > >>> | | | | > >>> | | | | > >>> v | v | > >>> +---------+--+ +---------+--+ > >>> |src vfio_dev| |dst vfio_dev| > >>> +--+---------+ +--+---------+ > >>> | ^ | ^ > >>> | | | | > >>> v | | | > >>> +-----------+----+ +-----------+----+ > >>> |src dev debugfs | |dst dev debugfs | > >>> +----------------+ +----------------+ > >>> > >>> The entire debugfs directory will be based on the definition of > >>> the CONFIG_DEBUG_FS macro. If this macro is not enabled, the > >>> interfaces in vfio.h will be empty definitions, and the creation > >>> and initialization of the debugfs directory will not be executed. > >>> > >>> vfio > >>> | > >>> +---<dev_name1> > >>> | +---migration > >>> | +--state > >>> | > >>> +---<dev_name2> > >>> +---migration > >>> +--state > >>> > >>> debugfs will create a public root directory "vfio" file. > >>> then create a dev_name() file for each live migration device. > >>> First, create a unified state acquisition file of "migration" > >>> in this device directory. > >>> Then, create a public live migration state lookup file "state". > >>> > >>> Signed-off-by: Longfang Liu <liulongfang@huawei.com> > >>> --- > >>> drivers/vfio/Kconfig | 10 +++++ > >>> drivers/vfio/Makefile | 1 + > >>> drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ > >>> drivers/vfio/vfio.h | 14 ++++++ > >>> drivers/vfio/vfio_main.c | 14 +++++- > >>> include/linux/vfio.h | 7 +++ > >>> include/uapi/linux/vfio.h | 1 + > >>> 7 files changed, 135 insertions(+), 2 deletions(-) > >>> create mode 100644 drivers/vfio/debugfs.c > >>> > >>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig > >>> index 6bda6dbb4878..ceae52fd7586 100644 > >>> --- a/drivers/vfio/Kconfig > >>> +++ b/drivers/vfio/Kconfig > >>> @@ -80,6 +80,16 @@ config VFIO_VIRQFD > >>> select EVENTFD > >>> default n > >>> +config VFIO_DEBUGFS > >>> + bool "Export VFIO internals in DebugFS" > >>> + depends on DEBUG_FS > >>> + help > >>> + Allows exposure of VFIO device internals. This option enables > >>> + the use of debugfs by VFIO drivers as required. The device can > >>> + cause the VFIO code create a top-level debug/vfio directory > >>> + during initialization, and then populate a subdirectory with > >>> + entries as required. > >>> + > >>> source "drivers/vfio/pci/Kconfig" > >>> source "drivers/vfio/platform/Kconfig" > >>> source "drivers/vfio/mdev/Kconfig" > >>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile > >>> index c82ea032d352..d43a699d55b1 100644 > >>> --- a/drivers/vfio/Makefile > >>> +++ b/drivers/vfio/Makefile > >>> @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o > >>> vfio-$(CONFIG_IOMMUFD) += iommufd.o > >>> vfio-$(CONFIG_VFIO_CONTAINER) += container.o > >>> vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o > >>> +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o > >>> obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o > >>> obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o > >>> diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c > >>> new file mode 100644 > >>> index 000000000000..ae53d6110f47 > >>> --- /dev/null > >>> +++ b/drivers/vfio/debugfs.c > >>> @@ -0,0 +1,90 @@ > >>> +// SPDX-License-Identifier: GPL-2.0-only > >>> +/* > >>> + * Copyright (c) 2023, HiSilicon Ltd. > >>> + */ > >>> + > >>> +#include <linux/device.h> > >>> +#include <linux/debugfs.h> > >>> +#include <linux/seq_file.h> > >>> +#include <linux/vfio.h> > >>> +#include "vfio.h" > >>> + > >>> +static struct dentry *vfio_debugfs_root; > >>> + > >>> +static int vfio_device_state_read(struct seq_file *seq, void *data) > >>> +{ > >>> + struct device *vf_dev = seq->private; > >>> + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); > >>> + enum vfio_device_mig_state state; > >>> + int ret; > >>> + > >>> + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != > >>> + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); > >>> + > >>> + ret = vdev->mig_ops->migration_get_state(vdev, &state); > >>> + if (ret) > >>> + return -EINVAL; > >>> + > >>> + switch (state) { > >>> + case VFIO_DEVICE_STATE_ERROR: > >>> + seq_printf(seq, "%s\n", "ERROR"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_STOP: > >>> + seq_printf(seq, "%s\n", "STOP"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_RUNNING: > >>> + seq_printf(seq, "%s\n", "RUNNING"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_STOP_COPY: > >>> + seq_printf(seq, "%s\n", "STOP_COPY"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_RESUMING: > >>> + seq_printf(seq, "%s\n", "RESUMING"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_RUNNING_P2P: > >>> + seq_printf(seq, "%s\n", "RUNNING_P2P"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_PRE_COPY: > >>> + seq_printf(seq, "%s\n", "PRE_COPY"); > >>> + break; > >>> + case VFIO_DEVICE_STATE_PRE_COPY_P2P: > >>> + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); > >>> + break; > >>> + default: > >>> + seq_printf(seq, "%s\n", "Invalid"); > >> > >> seq_puts() is more appropriate than seq_printf() above. > >> > > > > There is no difference between seq_puts() and seq_printf() here, > > no need to modify it. > > seq_puts is simply preferred for unformatted output. Agreed, thanks for pointing this out. I think Documentation/filesystems/seq_file.rst suggests this as well and it's also a bit silly anyway to use %s for a fixed string. > >> I would suggest to add an array or some helper, that the VFIO drivers > >> could use to debug the migration flow with pr_* primitives. It can be > >> done later. > >> > > > > If you want to debug this migration process in the VFIO driver, > > you can refer to vdev->mig_ops->migration_get_state() to read the status. > > > >> > >>> + } > >>> + > >>> + return 0; > >>> +} > >>> + > >>> +void vfio_device_debugfs_init(struct vfio_device *vdev) > >>> +{ > >>> + struct device *dev = &vdev->device; > >>> + > >>> + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); > >>> + > >>> + if (vdev->mig_ops) { > >>> + struct dentry *vfio_dev_migration = NULL; > >> > >> mig_dir maybe ? > >> > > > > "vfio_dev_migration " will not affect the readability of the code. > > > >> It would be easier to understand the nature of the variable IMHO. I don't know that we need to impose a specific style here, but the variable has a very limited scope, so it doesn't really need a super descriptive name. > >>> + > >>> + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); > >>> + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, > >>> + vfio_device_state_read); > >>> + } > >>> +} > >>> + > >>> +void vfio_device_debugfs_exit(struct vfio_device *vdev) > >>> +{ > >>> + debugfs_remove_recursive(vdev->debug_root); > >>> +} > >>> + > >>> +void vfio_debugfs_create_root(void) > >>> +{ > >>> + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); > >>> +} > >>> + > >>> +void vfio_debugfs_remove_root(void) > >>> +{ > >>> + debugfs_remove_recursive(vfio_debugfs_root); > >>> + vfio_debugfs_root = NULL; > >>> +} > >>> + > >>> diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h > >>> index 307e3f29b527..bde84ad344e5 100644 > >>> --- a/drivers/vfio/vfio.h > >>> +++ b/drivers/vfio/vfio.h > >>> @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) > >>> } > >>> #endif > >>> +#ifdef CONFIG_VFIO_DEBUGFS > >>> +void vfio_debugfs_create_root(void); > >>> +void vfio_debugfs_remove_root(void); > >>> + > >>> +void vfio_device_debugfs_init(struct vfio_device *vdev); > >>> +void vfio_device_debugfs_exit(struct vfio_device *vdev); > >>> +#else > >>> +static inline void vfio_debugfs_create_root(void) { } > >>> +static inline void vfio_debugfs_remove_root(void) { } > >>> + > >>> +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } > >>> +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } > >>> +#endif /* CONFIG_VFIO_DEBUGFS */ > >>> + > >>> #endif > >>> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c > >>> index e31e1952d7b8..9aec4c22f051 100644 > >>> --- a/drivers/vfio/vfio_main.c > >>> +++ b/drivers/vfio/vfio_main.c > >>> @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, > >>> /* Refcounting can't start until the driver calls register */ > >>> refcount_set(&device->refcount, 1); > >>> - > >> > >> superfluous change. > >> > > > > A blank line here is to separate it from the comment above. > > Makes it easier to be read. This sounds more like a justification for keeping the blank line than for removing it. The original intent was to provide some logical separation. Regardless, introducing unrelated formatting changes while implementing something else is generally considered poor form. > >>> vfio_device_group_register(device); > >>> return 0; > >>> @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, > >>> int vfio_register_group_dev(struct vfio_device *device) > >>> { > >>> - return __vfio_register_dev(device, VFIO_IOMMU); > >>> + int ret; > >>> + > >>> + ret = __vfio_register_dev(device, VFIO_IOMMU); > >>> + if (ret) > >>> + return ret; > >>> + > >>> + vfio_device_debugfs_init(device); > >> > >> Can it be called from __vfio_register_dev() instead ? and mdev devices > >> would get debugfs support also. > >> > > > > This is for symmetry in function calls. > > The need for symmetry was mentioned in the previous review. > > yes. But this is also exluding the mdev devices which is a large VFIO family. +1 It needs to support mdev as well. In fact the previous comments related to symmetry suggested it be placed at the end of __vfio_register_dev(), moving it to only cover the VFIO_IOMMU path is a bug. Thanks, Alex > >>> + > >>> + return 0; > >>> } > >>> EXPORT_SYMBOL_GPL(vfio_register_group_dev); > >>> @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) > >>> } > >>> } > >>> + vfio_device_debugfs_exit(device); > >>> /* Balances vfio_device_set_group in register path */ > >>> vfio_device_remove_group(device); > >>> } > >>> @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) > >>> if (ret) > >>> goto err_alloc_dev_chrdev; > >>> + vfio_debugfs_create_root(); > >>> pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); > >>> return 0; > >>> @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) > >>> static void __exit vfio_cleanup(void) > >>> { > >>> + vfio_debugfs_remove_root(); > >>> ida_destroy(&vfio.device_ida); > >>> vfio_cdev_cleanup(); > >>> class_destroy(vfio.device_class); > >>> diff --git a/include/linux/vfio.h b/include/linux/vfio.h > >>> index 454e9295970c..769d7af86225 100644 > >>> --- a/include/linux/vfio.h > >>> +++ b/include/linux/vfio.h > >>> @@ -69,6 +69,13 @@ struct vfio_device { > >>> u8 iommufd_attached:1; > >>> #endif > >>> u8 cdev_opened:1; > >>> +#ifdef CONFIG_DEBUG_FS > >>> + /* > >>> + * debug_root is a static property of the vfio_device > >>> + * which must be set prior to registering the vfio_device. > >>> + */ > >>> + struct dentry *debug_root; > >>> +#endif > >>> }; > >>> /** > >>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h > >>> index 7f5fb010226d..2b68e6cdf190 100644 > >>> --- a/include/uapi/linux/vfio.h > >>> +++ b/include/uapi/linux/vfio.h > >>> @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { > >>> VFIO_DEVICE_STATE_RUNNING_P2P = 5, > >>> VFIO_DEVICE_STATE_PRE_COPY = 6, > >>> VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, > >>> + VFIO_DEVICE_STATE_NR, > >>> }; > >>> /** > >> > >> . > >> > > > ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH v17 1/2] vfio/migration: Add debugfs to live migration driver 2023-10-19 8:30 ` Cédric Le Goater 2023-10-19 22:37 ` Alex Williamson @ 2023-10-20 9:38 ` liulongfang 1 sibling, 0 replies; 8+ messages in thread From: liulongfang @ 2023-10-20 9:38 UTC (permalink / raw) To: Cédric Le Goater, alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm On 2023/10/19 16:30, Cédric Le Goater wrote: > On 10/19/23 10:03, liulongfang wrote: >> On 2023/10/16 23:17, Cédric Le Goater wrote: >>> Hello Longfang, >>> >>> On 10/13/23 11:04, Longfang Liu wrote: >>>> There are multiple devices, software and operational steps involved >>>> in the process of live migration. An error occurred on any node may >>>> cause the live migration operation to fail. >>>> This complex process makes it very difficult to locate and analyze >>>> the cause when the function fails. >>>> >>>> In order to quickly locate the cause of the problem when the >>>> live migration fails, I added a set of debugfs to the vfio >>>> live migration driver. >>>> >>>> +-------------------------------------------+ >>>> | | >>>> | | >>>> | QEMU | >>>> | | >>>> | | >>>> +---+----------------------------+----------+ >>>> | ^ | ^ >>>> | | | | >>>> | | | | >>>> v | v | >>>> +---------+--+ +---------+--+ >>>> |src vfio_dev| |dst vfio_dev| >>>> +--+---------+ +--+---------+ >>>> | ^ | ^ >>>> | | | | >>>> v | | | >>>> +-----------+----+ +-----------+----+ >>>> |src dev debugfs | |dst dev debugfs | >>>> +----------------+ +----------------+ >>>> >>>> The entire debugfs directory will be based on the definition of >>>> the CONFIG_DEBUG_FS macro. If this macro is not enabled, the >>>> interfaces in vfio.h will be empty definitions, and the creation >>>> and initialization of the debugfs directory will not be executed. >>>> >>>> vfio >>>> | >>>> +---<dev_name1> >>>> | +---migration >>>> | +--state >>>> | >>>> +---<dev_name2> >>>> +---migration >>>> +--state >>>> >>>> debugfs will create a public root directory "vfio" file. >>>> then create a dev_name() file for each live migration device. >>>> First, create a unified state acquisition file of "migration" >>>> in this device directory. >>>> Then, create a public live migration state lookup file "state". >>>> >>>> Signed-off-by: Longfang Liu <liulongfang@huawei.com> >>>> --- >>>> drivers/vfio/Kconfig | 10 +++++ >>>> drivers/vfio/Makefile | 1 + >>>> drivers/vfio/debugfs.c | 90 +++++++++++++++++++++++++++++++++++++++ >>>> drivers/vfio/vfio.h | 14 ++++++ >>>> drivers/vfio/vfio_main.c | 14 +++++- >>>> include/linux/vfio.h | 7 +++ >>>> include/uapi/linux/vfio.h | 1 + >>>> 7 files changed, 135 insertions(+), 2 deletions(-) >>>> create mode 100644 drivers/vfio/debugfs.c >>>> >>>> diff --git a/drivers/vfio/Kconfig b/drivers/vfio/Kconfig >>>> index 6bda6dbb4878..ceae52fd7586 100644 >>>> --- a/drivers/vfio/Kconfig >>>> +++ b/drivers/vfio/Kconfig >>>> @@ -80,6 +80,16 @@ config VFIO_VIRQFD >>>> select EVENTFD >>>> default n >>>> +config VFIO_DEBUGFS >>>> + bool "Export VFIO internals in DebugFS" >>>> + depends on DEBUG_FS >>>> + help >>>> + Allows exposure of VFIO device internals. This option enables >>>> + the use of debugfs by VFIO drivers as required. The device can >>>> + cause the VFIO code create a top-level debug/vfio directory >>>> + during initialization, and then populate a subdirectory with >>>> + entries as required. >>>> + >>>> source "drivers/vfio/pci/Kconfig" >>>> source "drivers/vfio/platform/Kconfig" >>>> source "drivers/vfio/mdev/Kconfig" >>>> diff --git a/drivers/vfio/Makefile b/drivers/vfio/Makefile >>>> index c82ea032d352..d43a699d55b1 100644 >>>> --- a/drivers/vfio/Makefile >>>> +++ b/drivers/vfio/Makefile >>>> @@ -8,6 +8,7 @@ vfio-$(CONFIG_VFIO_GROUP) += group.o >>>> vfio-$(CONFIG_IOMMUFD) += iommufd.o >>>> vfio-$(CONFIG_VFIO_CONTAINER) += container.o >>>> vfio-$(CONFIG_VFIO_VIRQFD) += virqfd.o >>>> +vfio-$(CONFIG_VFIO_DEBUGFS) += debugfs.o >>>> obj-$(CONFIG_VFIO_IOMMU_TYPE1) += vfio_iommu_type1.o >>>> obj-$(CONFIG_VFIO_IOMMU_SPAPR_TCE) += vfio_iommu_spapr_tce.o >>>> diff --git a/drivers/vfio/debugfs.c b/drivers/vfio/debugfs.c >>>> new file mode 100644 >>>> index 000000000000..ae53d6110f47 >>>> --- /dev/null >>>> +++ b/drivers/vfio/debugfs.c >>>> @@ -0,0 +1,90 @@ >>>> +// SPDX-License-Identifier: GPL-2.0-only >>>> +/* >>>> + * Copyright (c) 2023, HiSilicon Ltd. >>>> + */ >>>> + >>>> +#include <linux/device.h> >>>> +#include <linux/debugfs.h> >>>> +#include <linux/seq_file.h> >>>> +#include <linux/vfio.h> >>>> +#include "vfio.h" >>>> + >>>> +static struct dentry *vfio_debugfs_root; >>>> + >>>> +static int vfio_device_state_read(struct seq_file *seq, void *data) >>>> +{ >>>> + struct device *vf_dev = seq->private; >>>> + struct vfio_device *vdev = container_of(vf_dev, struct vfio_device, device); >>>> + enum vfio_device_mig_state state; >>>> + int ret; >>>> + >>>> + BUILD_BUG_ON(VFIO_DEVICE_STATE_NR != >>>> + VFIO_DEVICE_STATE_PRE_COPY_P2P + 1); >>>> + >>>> + ret = vdev->mig_ops->migration_get_state(vdev, &state); >>>> + if (ret) >>>> + return -EINVAL; >>>> + >>>> + switch (state) { >>>> + case VFIO_DEVICE_STATE_ERROR: >>>> + seq_printf(seq, "%s\n", "ERROR"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_STOP: >>>> + seq_printf(seq, "%s\n", "STOP"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_RUNNING: >>>> + seq_printf(seq, "%s\n", "RUNNING"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_STOP_COPY: >>>> + seq_printf(seq, "%s\n", "STOP_COPY"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_RESUMING: >>>> + seq_printf(seq, "%s\n", "RESUMING"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_RUNNING_P2P: >>>> + seq_printf(seq, "%s\n", "RUNNING_P2P"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_PRE_COPY: >>>> + seq_printf(seq, "%s\n", "PRE_COPY"); >>>> + break; >>>> + case VFIO_DEVICE_STATE_PRE_COPY_P2P: >>>> + seq_printf(seq, "%s\n", "PRE_COPY_P2P"); >>>> + break; >>>> + default: >>>> + seq_printf(seq, "%s\n", "Invalid"); >>> >>> seq_puts() is more appropriate than seq_printf() above. >>> >> >> There is no difference between seq_puts() and seq_printf() here, >> no need to modify it. > > seq_puts is simply preferred for unformatted output. > I think there is nothing wrong with using "seq_printf". >>> I would suggest to add an array or some helper, that the VFIO drivers >>> could use to debug the migration flow with pr_* primitives. It can be >>> done later. >>> >> >> If you want to debug this migration process in the VFIO driver, >> you can refer to vdev->mig_ops->migration_get_state() to read the status. >> >>> >>>> + } >>>> + >>>> + return 0; >>>> +} >>>> + >>>> +void vfio_device_debugfs_init(struct vfio_device *vdev) >>>> +{ >>>> + struct device *dev = &vdev->device; >>>> + >>>> + vdev->debug_root = debugfs_create_dir(dev_name(vdev->dev), vfio_debugfs_root); >>>> + >>>> + if (vdev->mig_ops) { >>>> + struct dentry *vfio_dev_migration = NULL; >>> >>> mig_dir maybe ? >>> >> >> "vfio_dev_migration " will not affect the readability of the code. >> >>> It would be easier to understand the nature of the variable IMHO. >>> >>>> + >>>> + vfio_dev_migration = debugfs_create_dir("migration", vdev->debug_root); >>>> + debugfs_create_devm_seqfile(dev, "state", vfio_dev_migration, >>>> + vfio_device_state_read); >>>> + } >>>> +} >>>> + >>>> +void vfio_device_debugfs_exit(struct vfio_device *vdev) >>>> +{ >>>> + debugfs_remove_recursive(vdev->debug_root); >>>> +} >>>> + >>>> +void vfio_debugfs_create_root(void) >>>> +{ >>>> + vfio_debugfs_root = debugfs_create_dir("vfio", NULL); >>>> +} >>>> + >>>> +void vfio_debugfs_remove_root(void) >>>> +{ >>>> + debugfs_remove_recursive(vfio_debugfs_root); >>>> + vfio_debugfs_root = NULL; >>>> +} >>>> + >>>> diff --git a/drivers/vfio/vfio.h b/drivers/vfio/vfio.h >>>> index 307e3f29b527..bde84ad344e5 100644 >>>> --- a/drivers/vfio/vfio.h >>>> +++ b/drivers/vfio/vfio.h >>>> @@ -448,4 +448,18 @@ static inline void vfio_device_put_kvm(struct vfio_device *device) >>>> } >>>> #endif >>>> +#ifdef CONFIG_VFIO_DEBUGFS >>>> +void vfio_debugfs_create_root(void); >>>> +void vfio_debugfs_remove_root(void); >>>> + >>>> +void vfio_device_debugfs_init(struct vfio_device *vdev); >>>> +void vfio_device_debugfs_exit(struct vfio_device *vdev); >>>> +#else >>>> +static inline void vfio_debugfs_create_root(void) { } >>>> +static inline void vfio_debugfs_remove_root(void) { } >>>> + >>>> +static inline void vfio_device_debugfs_init(struct vfio_device *vdev) { } >>>> +static inline void vfio_device_debugfs_exit(struct vfio_device *vdev) { } >>>> +#endif /* CONFIG_VFIO_DEBUGFS */ >>>> + >>>> #endif >>>> diff --git a/drivers/vfio/vfio_main.c b/drivers/vfio/vfio_main.c >>>> index e31e1952d7b8..9aec4c22f051 100644 >>>> --- a/drivers/vfio/vfio_main.c >>>> +++ b/drivers/vfio/vfio_main.c >>>> @@ -309,7 +309,6 @@ static int __vfio_register_dev(struct vfio_device *device, >>>> /* Refcounting can't start until the driver calls register */ >>>> refcount_set(&device->refcount, 1); >>>> - >>> >>> superfluous change. >>> >> >> A blank line here is to separate it from the comment above. >> Makes it easier to be read. >> >>>> vfio_device_group_register(device); >>>> return 0; >>>> @@ -320,7 +319,15 @@ static int __vfio_register_dev(struct vfio_device *device, >>>> int vfio_register_group_dev(struct vfio_device *device) >>>> { >>>> - return __vfio_register_dev(device, VFIO_IOMMU); >>>> + int ret; >>>> + >>>> + ret = __vfio_register_dev(device, VFIO_IOMMU); >>>> + if (ret) >>>> + return ret; >>>> + >>>> + vfio_device_debugfs_init(device); >>> >>> Can it be called from __vfio_register_dev() instead ? and mdev devices >>> would get debugfs support also. >>> >> >> This is for symmetry in function calls. >> The need for symmetry was mentioned in the previous review. > > yes. But this is also exluding the mdev devices which is a large VFIO family. > Here it is symmetrically encoded like this. Are there any problems when using it in mdev? Thanks, Longfang. > Thanks, > > C. > >> >>> Thanks, >>> >>> C. >>> >> Thanks, >> Longfang. >> >>>> + >>>> + return 0; >>>> } >>>> EXPORT_SYMBOL_GPL(vfio_register_group_dev); >>>> @@ -378,6 +385,7 @@ void vfio_unregister_group_dev(struct vfio_device *device) >>>> } >>>> } >>>> + vfio_device_debugfs_exit(device); >>>> /* Balances vfio_device_set_group in register path */ >>>> vfio_device_remove_group(device); >>>> } >>>> @@ -1676,6 +1684,7 @@ static int __init vfio_init(void) >>>> if (ret) >>>> goto err_alloc_dev_chrdev; >>>> + vfio_debugfs_create_root(); >>>> pr_info(DRIVER_DESC " version: " DRIVER_VERSION "\n"); >>>> return 0; >>>> @@ -1691,6 +1700,7 @@ static int __init vfio_init(void) >>>> static void __exit vfio_cleanup(void) >>>> { >>>> + vfio_debugfs_remove_root(); >>>> ida_destroy(&vfio.device_ida); >>>> vfio_cdev_cleanup(); >>>> class_destroy(vfio.device_class); >>>> diff --git a/include/linux/vfio.h b/include/linux/vfio.h >>>> index 454e9295970c..769d7af86225 100644 >>>> --- a/include/linux/vfio.h >>>> +++ b/include/linux/vfio.h >>>> @@ -69,6 +69,13 @@ struct vfio_device { >>>> u8 iommufd_attached:1; >>>> #endif >>>> u8 cdev_opened:1; >>>> +#ifdef CONFIG_DEBUG_FS >>>> + /* >>>> + * debug_root is a static property of the vfio_device >>>> + * which must be set prior to registering the vfio_device. >>>> + */ >>>> + struct dentry *debug_root; >>>> +#endif >>>> }; >>>> /** >>>> diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h >>>> index 7f5fb010226d..2b68e6cdf190 100644 >>>> --- a/include/uapi/linux/vfio.h >>>> +++ b/include/uapi/linux/vfio.h >>>> @@ -1219,6 +1219,7 @@ enum vfio_device_mig_state { >>>> VFIO_DEVICE_STATE_RUNNING_P2P = 5, >>>> VFIO_DEVICE_STATE_PRE_COPY = 6, >>>> VFIO_DEVICE_STATE_PRE_COPY_P2P = 7, >>>> + VFIO_DEVICE_STATE_NR, >>>> }; >>>> /** >>> >>> . >>> >> > > . > ^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH v17 2/2] Documentation: add debugfs description for vfio 2023-10-13 9:04 [PATCH v17 0/2] add debugfs to migration driver Longfang Liu 2023-10-13 9:04 ` [PATCH v17 1/2] vfio/migration: Add debugfs to live " Longfang Liu @ 2023-10-13 9:04 ` Longfang Liu 1 sibling, 0 replies; 8+ messages in thread From: Longfang Liu @ 2023-10-13 9:04 UTC (permalink / raw) To: alex.williamson, jgg, shameerali.kolothum.thodi, jonathan.cameron Cc: bcreeley, kvm, linux-kernel, linuxarm, liulongfang 1.Add an debugfs document description file to help users understand how to use the accelerator live migration driver's debugfs. 2.Update the file paths that need to be maintained in MAINTAINERS Signed-off-by: Longfang Liu <liulongfang@huawei.com> --- Documentation/ABI/testing/debugfs-vfio | 25 +++++++++++++++++++++++++ MAINTAINERS | 1 + 2 files changed, 26 insertions(+) create mode 100644 Documentation/ABI/testing/debugfs-vfio diff --git a/Documentation/ABI/testing/debugfs-vfio b/Documentation/ABI/testing/debugfs-vfio new file mode 100644 index 000000000000..445e9f58f924 --- /dev/null +++ b/Documentation/ABI/testing/debugfs-vfio @@ -0,0 +1,25 @@ +What: /sys/kernel/debug/vfio +Date: Oct 2023 +KernelVersion: 6.7 +Contact: Longfang Liu <liulongfang@huawei.com> +Description: This debugfs file directory is used for debugging + of vfio devices, it's a common directory for all vfio devices. + Vfio core will create a device subdirectory under this + directory. + +What: /sys/kernel/debug/vfio/<device>/migration +Date: Oct 2023 +KernelVersion: 6.7 +Contact: Longfang Liu <liulongfang@huawei.com> +Description: This debugfs file directory is used for debugging + of vfio devices that support live migration. + The debugfs of each vfio device that supports live migration + could be created under this directory. + +What: /sys/kernel/debug/vfio/<device>/migration/state +Date: Oct 2023 +KernelVersion: 6.7 +Contact: Longfang Liu <liulongfang@huawei.com> +Description: Read the live migration status of the vfio device. + The contents of the state file reflects the migration state + relative to those defined in the vfio_device_mig_state enum diff --git a/MAINTAINERS b/MAINTAINERS index b19995690904..a6be3b4219c7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -22591,6 +22591,7 @@ L: kvm@vger.kernel.org S: Maintained T: git https://github.com/awilliam/linux-vfio.git F: Documentation/ABI/testing/sysfs-devices-vfio-dev +F: Documentation/ABI/testing/debugfs-vfio F: Documentation/driver-api/vfio.rst F: drivers/vfio/ F: include/linux/vfio.h -- 2.24.0 ^ permalink raw reply related [flat|nested] 8+ messages in thread
end of thread, other threads:[~2023-10-20 9:38 UTC | newest] Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2023-10-13 9:04 [PATCH v17 0/2] add debugfs to migration driver Longfang Liu 2023-10-13 9:04 ` [PATCH v17 1/2] vfio/migration: Add debugfs to live " Longfang Liu 2023-10-16 15:17 ` Cédric Le Goater 2023-10-19 8:03 ` liulongfang 2023-10-19 8:30 ` Cédric Le Goater 2023-10-19 22:37 ` Alex Williamson 2023-10-20 9:38 ` liulongfang 2023-10-13 9:04 ` [PATCH v17 2/2] Documentation: add debugfs description for vfio Longfang Liu
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).