From 22e1992cf7b034db5325660e98c41ca5afa5f519 Mon Sep 17 00:00:00 2001 From: Gavin Shan Date: Thu, 28 Mar 2024 10:21:47 +1000 Subject: [PATCH 1/5] vhost: Add smp_rmb() in vhost_vq_avail_empty() A smp_rmb() has been missed in vhost_vq_avail_empty(), spotted by Will. Otherwise, it's not ensured the available ring entries pushed by guest can be observed by vhost in time, leading to stale available ring entries fetched by vhost in vhost_get_vq_desc(), as reported by Yihuang Yu on NVidia's grace-hopper (ARM64) platform. /home/gavin/sandbox/qemu.main/build/qemu-system-aarch64 \ -accel kvm -machine virt,gic-version=host -cpu host \ -smp maxcpus=1,cpus=1,sockets=1,clusters=1,cores=1,threads=1 \ -m 4096M,slots=16,maxmem=64G \ -object memory-backend-ram,id=mem0,size=4096M \ : \ -netdev tap,id=vnet0,vhost=true \ -device virtio-net-pci,bus=pcie.8,netdev=vnet0,mac=52:54:00:f1:26:b0 : guest# netperf -H 10.26.1.81 -l 60 -C -c -t UDP_STREAM virtio_net virtio0: output.0:id 100 is not a head! Add the missed smp_rmb() in vhost_vq_avail_empty(). When tx_can_batch() returns true, it means there's still pending tx buffers. Since it might read indices, so it still can bypass the smp_rmb() in vhost_get_vq_desc(). Note that it should be safe until vq->avail_idx is changed by commit 275bf960ac697 ("vhost: better detection of available buffers"). Fixes: 275bf960ac69 ("vhost: better detection of available buffers") Cc: # v4.11+ Reported-by: Yihuang Yu Suggested-by: Will Deacon Signed-off-by: Gavin Shan Acked-by: Jason Wang Message-Id: <20240328002149.1141302-2-gshan@redhat.com> Signed-off-by: Michael S. Tsirkin Reviewed-by: Stefano Garzarella --- drivers/vhost/vhost.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 045f666b4f12..29df65b2ebf2 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -2799,9 +2799,19 @@ bool vhost_vq_avail_empty(struct vhost_dev *dev, struct vhost_virtqueue *vq) r = vhost_get_avail_idx(vq, &avail_idx); if (unlikely(r)) return false; - vq->avail_idx = vhost16_to_cpu(vq, avail_idx); - return vq->avail_idx == vq->last_avail_idx; + vq->avail_idx = vhost16_to_cpu(vq, avail_idx); + if (vq->avail_idx != vq->last_avail_idx) { + /* Since we have updated avail_idx, the following + * call to vhost_get_vq_desc() will read available + * ring entries. Make sure that read happens after + * the avail_idx read. + */ + smp_rmb(); + return false; + } + + return true; } EXPORT_SYMBOL_GPL(vhost_vq_avail_empty); From df9ace7647d4123209395bb9967e998d5758c645 Mon Sep 17 00:00:00 2001 From: Gavin Shan Date: Thu, 28 Mar 2024 10:21:48 +1000 Subject: [PATCH 2/5] vhost: Add smp_rmb() in vhost_enable_notify() A smp_rmb() has been missed in vhost_enable_notify(), inspired by Will. Otherwise, it's not ensured the available ring entries pushed by guest can be observed by vhost in time, leading to stale available ring entries fetched by vhost in vhost_get_vq_desc(), as reported by Yihuang Yu on NVidia's grace-hopper (ARM64) platform. /home/gavin/sandbox/qemu.main/build/qemu-system-aarch64 \ -accel kvm -machine virt,gic-version=host -cpu host \ -smp maxcpus=1,cpus=1,sockets=1,clusters=1,cores=1,threads=1 \ -m 4096M,slots=16,maxmem=64G \ -object memory-backend-ram,id=mem0,size=4096M \ : \ -netdev tap,id=vnet0,vhost=true \ -device virtio-net-pci,bus=pcie.8,netdev=vnet0,mac=52:54:00:f1:26:b0 : guest# netperf -H 10.26.1.81 -l 60 -C -c -t UDP_STREAM virtio_net virtio0: output.0:id 100 is not a head! Add the missed smp_rmb() in vhost_enable_notify(). When it returns true, it means there's still pending tx buffers. Since it might read indices, so it still can bypass the smp_rmb() in vhost_get_vq_desc(). Note that it should be safe until vq->avail_idx is changed by commit d3bb267bbdcb ("vhost: cache avail index in vhost_enable_notify()"). Fixes: d3bb267bbdcb ("vhost: cache avail index in vhost_enable_notify()") Cc: # v5.18+ Reported-by: Yihuang Yu Suggested-by: Will Deacon Signed-off-by: Gavin Shan Acked-by: Jason Wang Message-Id: <20240328002149.1141302-3-gshan@redhat.com> Signed-off-by: Michael S. Tsirkin Reviewed-by: Stefano Garzarella --- drivers/vhost/vhost.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 29df65b2ebf2..32686c79c41d 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -2848,9 +2848,19 @@ bool vhost_enable_notify(struct vhost_dev *dev, struct vhost_virtqueue *vq) &vq->avail->idx, r); return false; } - vq->avail_idx = vhost16_to_cpu(vq, avail_idx); - return vq->avail_idx != vq->last_avail_idx; + vq->avail_idx = vhost16_to_cpu(vq, avail_idx); + if (vq->avail_idx != vq->last_avail_idx) { + /* Since we have updated avail_idx, the following + * call to vhost_get_vq_desc() will read available + * ring entries. Make sure that read happens after + * the avail_idx read. + */ + smp_rmb(); + return true; + } + + return false; } EXPORT_SYMBOL_GPL(vhost_enable_notify); From ffe6176b7f53ca0c99355f13e14a33a40cf49406 Mon Sep 17 00:00:00 2001 From: Krzysztof Kozlowski Date: Sun, 31 Mar 2024 10:43:48 +0200 Subject: [PATCH 3/5] virtio: store owner from modules with register_virtio_driver() Modules registering driver with register_virtio_driver() might forget to set .owner field. i2c-virtio.c for example has it missing. The field is used by some other kernel parts for reference counting (try_module_get()), so it is expected that drivers will set it. Solve the problem by moving this task away from the drivers to the core virtio code, just like we did for platform_driver in commit 9447057eaff8 ("platform_device: use a macro instead of platform_driver_register"). Fixes: 3cfc88380413 ("i2c: virtio: add a virtio i2c frontend driver") Cc: "Jie Deng" Signed-off-by: Krzysztof Kozlowski Message-Id: <20240331-module-owner-virtio-v2-1-98f04bfaf46a@linaro.org> Signed-off-by: Michael S. Tsirkin --- Documentation/driver-api/virtio/writing_virtio_drivers.rst | 1 - drivers/virtio/virtio.c | 6 ++++-- include/linux/virtio.h | 7 +++++-- 3 files changed, 9 insertions(+), 5 deletions(-) diff --git a/Documentation/driver-api/virtio/writing_virtio_drivers.rst b/Documentation/driver-api/virtio/writing_virtio_drivers.rst index e14c58796d25..e5de6f5d061a 100644 --- a/Documentation/driver-api/virtio/writing_virtio_drivers.rst +++ b/Documentation/driver-api/virtio/writing_virtio_drivers.rst @@ -97,7 +97,6 @@ like this:: static struct virtio_driver virtio_dummy_driver = { .driver.name = KBUILD_MODNAME, - .driver.owner = THIS_MODULE, .id_table = id_table, .probe = virtio_dummy_probe, .remove = virtio_dummy_remove, diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c index f173587893cb..9510c551dce8 100644 --- a/drivers/virtio/virtio.c +++ b/drivers/virtio/virtio.c @@ -362,14 +362,16 @@ static const struct bus_type virtio_bus = { .remove = virtio_dev_remove, }; -int register_virtio_driver(struct virtio_driver *driver) +int __register_virtio_driver(struct virtio_driver *driver, struct module *owner) { /* Catch this early. */ BUG_ON(driver->feature_table_size && !driver->feature_table); driver->driver.bus = &virtio_bus; + driver->driver.owner = owner; + return driver_register(&driver->driver); } -EXPORT_SYMBOL_GPL(register_virtio_driver); +EXPORT_SYMBOL_GPL(__register_virtio_driver); void unregister_virtio_driver(struct virtio_driver *driver) { diff --git a/include/linux/virtio.h b/include/linux/virtio.h index b0201747a263..26c4325aa373 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -170,7 +170,7 @@ size_t virtio_max_dma_size(const struct virtio_device *vdev); /** * struct virtio_driver - operations for a virtio I/O driver - * @driver: underlying device driver (populate name and owner). + * @driver: underlying device driver (populate name). * @id_table: the ids serviced by this driver. * @feature_table: an array of feature numbers supported by this driver. * @feature_table_size: number of entries in the feature table array. @@ -208,7 +208,10 @@ static inline struct virtio_driver *drv_to_virtio(struct device_driver *drv) return container_of(drv, struct virtio_driver, driver); } -int register_virtio_driver(struct virtio_driver *drv); +/* use a macro to avoid include chaining to get THIS_MODULE */ +#define register_virtio_driver(drv) \ + __register_virtio_driver(drv, THIS_MODULE) +int __register_virtio_driver(struct virtio_driver *drv, struct module *owner); void unregister_virtio_driver(struct virtio_driver *drv); /* module_virtio_driver() - Helper macro for drivers that don't do From 2855c2a7820bc8198ae937a9a67dbdc3990e9d2c Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Tue, 2 Apr 2024 17:21:43 -0400 Subject: [PATCH 4/5] vhost-vdpa: change ioctl # for VDPA_GET_VRING_SIZE MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit VDPA_GET_VRING_SIZE by mistake uses the already occupied ioctl # 0x80 and we never noticed - it happens to work because the direction and size are different, but confuses tools such as perf which like to look at just the number, and breaks the extra robustness of the ioctl numbering macros. To fix, sort the entries and renumber the ioctl - not too late since it wasn't in any released kernels yet. Cc: Arnaldo Carvalho de Melo Reported-by: Namhyung Kim Fixes: 1496c47065f9 ("vhost-vdpa: uapi to support reporting per vq size") Cc: "Zhu Lingshan" Signed-off-by: Michael S. Tsirkin Message-Id: <41c1c5489688abe5bfef9f7cf15584e3fb872ac5.1712092759.git.mst@redhat.com> Reviewed-by: Eugenio Pérez Reviewed-by: Zhu Lingshan Reviewed-by: Stefano Garzarella Acked-by: Jason Wang --- include/uapi/linux/vhost.h | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git a/include/uapi/linux/vhost.h b/include/uapi/linux/vhost.h index bea697390613..b95dd84eef2d 100644 --- a/include/uapi/linux/vhost.h +++ b/include/uapi/linux/vhost.h @@ -179,12 +179,6 @@ /* Get the config size */ #define VHOST_VDPA_GET_CONFIG_SIZE _IOR(VHOST_VIRTIO, 0x79, __u32) -/* Get the count of all virtqueues */ -#define VHOST_VDPA_GET_VQS_COUNT _IOR(VHOST_VIRTIO, 0x80, __u32) - -/* Get the number of virtqueue groups. */ -#define VHOST_VDPA_GET_GROUP_NUM _IOR(VHOST_VIRTIO, 0x81, __u32) - /* Get the number of address spaces. */ #define VHOST_VDPA_GET_AS_NUM _IOR(VHOST_VIRTIO, 0x7A, unsigned int) @@ -228,10 +222,17 @@ #define VHOST_VDPA_GET_VRING_DESC_GROUP _IOWR(VHOST_VIRTIO, 0x7F, \ struct vhost_vring_state) + +/* Get the count of all virtqueues */ +#define VHOST_VDPA_GET_VQS_COUNT _IOR(VHOST_VIRTIO, 0x80, __u32) + +/* Get the number of virtqueue groups. */ +#define VHOST_VDPA_GET_GROUP_NUM _IOR(VHOST_VIRTIO, 0x81, __u32) + /* Get the queue size of a specific virtqueue. * userspace set the vring index in vhost_vring_state.index * kernel set the queue size in vhost_vring_state.num */ -#define VHOST_VDPA_GET_VRING_SIZE _IOWR(VHOST_VIRTIO, 0x80, \ +#define VHOST_VDPA_GET_VRING_SIZE _IOWR(VHOST_VIRTIO, 0x82, \ struct vhost_vring_state) #endif From 76f408535aab39c33e0a1dcada9fba5631c65595 Mon Sep 17 00:00:00 2001 From: Xianting Tian Date: Mon, 11 Mar 2024 16:21:09 +0800 Subject: [PATCH 5/5] vhost: correct misleading printing information Guest moved avail idx not used idx when we need to print log if '(vq->avail_idx - last_avail_idx) > vq->num', so fix it. Signed-off-by: Xianting Tian Message-Id: <20240311082109.46773-1-xianting.tian@linux.alibaba.com> Signed-off-by: Michael S. Tsirkin --- drivers/vhost/vhost.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 32686c79c41d..8995730ce0bf 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -2515,7 +2515,7 @@ int vhost_get_vq_desc(struct vhost_virtqueue *vq, vq->avail_idx = vhost16_to_cpu(vq, avail_idx); if (unlikely((u16)(vq->avail_idx - last_avail_idx) > vq->num)) { - vq_err(vq, "Guest moved used index from %u to %u", + vq_err(vq, "Guest moved avail index from %u to %u", last_avail_idx, vq->avail_idx); return -EFAULT; }