tjh.dev / kernel
Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

vfio: Follow a strict lifetime for struct iommu_group

The iommu_group comes from the struct device that a driver has been bound
to and then created a struct vfio_device against. To keep the iommu layer
sane we want to have a simple rule that only an attached driver should be
using the iommu API. Particularly only an attached driver should hold
ownership.

In VFIO's case since it uses the group APIs and it shares between
different drivers it is a bit more complicated, but the principle still
holds.

Solve this by waiting for all users of the vfio_group to stop before
allowing vfio_unregister_group_dev() to complete. This is done with a new
completion to know when the users go away and an additional refcount to
keep track of how many device drivers are sharing the vfio group. The last
driver to be unregistered will clean up the group.

This solves crashes in the S390 iommu driver that come because VFIO ends
up racing releasing ownership (which attaches the default iommu_domain to
the device) with the removal of that same device from the iommu
driver. This is a side case that iommu drivers should not have to cope
with.

iommu driver failed to attach the default/blocking domain
WARNING: CPU: 0 PID: 5082 at drivers/iommu/iommu.c:1961 iommu_detach_group+0x6c/0x80
Modules linked in: macvtap macvlan tap vfio_pci vfio_pci_core irqbypass vfio_virqfd kvm nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reject nft_ct nft_chain_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_set nf_tables nfnetlink mlx5_ib sunrpc ib_uverbs ism smc uvdevice ib_core s390_trng eadm_sch tape_3590 tape tape_class vfio_ccw mdev vfio_iommu_type1 vfio zcrypt_cex4 sch_fq_codel configfs ghash_s390 prng chacha_s390 libchacha aes_s390 mlx5_core des_s390 libdes sha3_512_s390 nvme sha3_256_s390 sha512_s390 sha256_s390 sha1_s390 sha_common nvme_core zfcp scsi_transport_fc pkey zcrypt rng_core autofs4
CPU: 0 PID: 5082 Comm: qemu-system-s39 Tainted: G W 6.0.0-rc3 #5
Hardware name: IBM 3931 A01 782 (LPAR)
Krnl PSW : 0704c00180000000 000000095bb10d28 (iommu_detach_group+0x70/0x80)
R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:0 PM:0 RI:0 EA:3
Krnl GPRS: 0000000000000001 0000000900000027 0000000000000039 000000095c97ffe0
00000000fffeffff 00000009fc290000 00000000af1fda50 00000000af590b58
00000000af1fdaf0 0000000135c7a320 0000000135e52258 0000000135e52200
00000000a29e8000 00000000af590b40 000000095bb10d24 0000038004b13c98
Krnl Code: 000000095bb10d18: c020003d56fc larl %r2,000000095c2bbb10
000000095bb10d1e: c0e50019d901 brasl %r14,000000095be4bf20
#000000095bb10d24: af000000 mc 0,0
>000000095bb10d28: b904002a lgr %r2,%r10
000000095bb10d2c: ebaff0a00004 lmg %r10,%r15,160(%r15)
000000095bb10d32: c0f4001aa867 brcl 15,000000095be65e00
000000095bb10d38: c004002168e0 brcl 0,000000095bf3def8
000000095bb10d3e: eb6ff0480024 stmg %r6,%r15,72(%r15)
Call Trace:
[<000000095bb10d28>] iommu_detach_group+0x70/0x80
([<000000095bb10d24>] iommu_detach_group+0x6c/0x80)
[<000003ff80243b0e>] vfio_iommu_type1_detach_group+0x136/0x6c8 [vfio_iommu_type1]
[<000003ff80137780>] __vfio_group_unset_container+0x58/0x158 [vfio]
[<000003ff80138a16>] vfio_group_fops_unl_ioctl+0x1b6/0x210 [vfio]
pci 0004:00:00.0: Removing from iommu group 4
[<000000095b5b62e8>] __s390x_sys_ioctl+0xc0/0x100
[<000000095be5d3b4>] __do_syscall+0x1d4/0x200
[<000000095be6c072>] system_call+0x82/0xb0
Last Breaking-Event-Address:
[<000000095be4bf80>] __warn_printk+0x60/0x68

It indicates that domain->ops->attach_dev() failed because the driver has
already passed the point of destructing the device.

Fixes: 9ac8545199a1 ("iommu: Fix use-after-free in iommu_release_device")
Reported-by: Matthew Rosato <mjrosato@linux.ibm.com>
Tested-by: Matthew Rosato <mjrosato@linux.ibm.com>
Reviewed-by: Yi Liu <yi.l.liu@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/0-v2-a3c5f4429e2a+55-iommu_group_lifetime_jgg@nvidia.com
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>

authored by

Jason Gunthorpe and committed by
Alex Williamson ca5f21b2 cdc71fe4

+53 -23
+8
drivers/vfio/vfio.h
··· 41 41 struct vfio_group { 42 42 struct device dev; 43 43 struct cdev cdev; 44 + /* 45 + * When drivers is non-zero a driver is attached to the struct device 46 + * that provided the iommu_group and thus the iommu_group is a valid 47 + * pointer. When drivers is 0 the driver is being detached. Once users 48 + * reaches 0 then the iommu_group is invalid. 49 + */ 50 + refcount_t drivers; 44 51 refcount_t users; 52 + struct completion users_comp; 45 53 unsigned int container_users; 46 54 struct iommu_group *iommu_group; 47 55 struct vfio_container *container;
+45 -23
drivers/vfio/vfio_main.c
··· 125 125 xa_unlock(&vfio_device_set_xa); 126 126 } 127 127 128 - static void vfio_group_get(struct vfio_group *group); 129 - 130 128 /* 131 129 * Group objects - create, release, get, put, search 132 130 */ ··· 135 137 136 138 list_for_each_entry(group, &vfio.group_list, vfio_next) { 137 139 if (group->iommu_group == iommu_group) { 138 - vfio_group_get(group); 140 + refcount_inc(&group->drivers); 139 141 return group; 140 142 } 141 143 } ··· 187 189 group->cdev.owner = THIS_MODULE; 188 190 189 191 refcount_set(&group->users, 1); 192 + refcount_set(&group->drivers, 1); 193 + init_completion(&group->users_comp); 190 194 init_rwsem(&group->group_rwsem); 191 195 INIT_LIST_HEAD(&group->device_list); 192 196 mutex_init(&group->device_lock); ··· 247 247 248 248 static void vfio_group_put(struct vfio_group *group) 249 249 { 250 - if (!refcount_dec_and_mutex_lock(&group->users, &vfio.group_lock)) 250 + if (refcount_dec_and_test(&group->users)) 251 + complete(&group->users_comp); 252 + } 253 + 254 + static void vfio_device_remove_group(struct vfio_device *device) 255 + { 256 + struct vfio_group *group = device->group; 257 + 258 + if (group->type == VFIO_NO_IOMMU || group->type == VFIO_EMULATED_IOMMU) 259 + iommu_group_remove_device(device->dev); 260 + 261 + /* Pairs with vfio_create_group() / vfio_group_get_from_iommu() */ 262 + if (!refcount_dec_and_mutex_lock(&group->drivers, &vfio.group_lock)) 251 263 return; 264 + list_del(&group->vfio_next); 265 + 266 + /* 267 + * We could concurrently probe another driver in the group that might 268 + * race vfio_device_remove_group() with vfio_get_group(), so we have to 269 + * ensure that the sysfs is all cleaned up under lock otherwise the 270 + * cdev_device_add() will fail due to the name aready existing. 271 + */ 272 + cdev_device_del(&group->cdev, &group->dev); 273 + mutex_unlock(&vfio.group_lock); 274 + 275 + /* Matches the get from vfio_group_alloc() */ 276 + vfio_group_put(group); 277 + 278 + /* 279 + * Before we allow the last driver in the group to be unplugged the 280 + * group must be sanitized so nothing else is or can reference it. This 281 + * is because the group->iommu_group pointer should only be used so long 282 + * as a device driver is attached to a device in the group. 283 + */ 284 + wait_for_completion(&group->users_comp); 252 285 253 286 /* 254 287 * These data structures all have paired operations that can only be ··· 292 259 WARN_ON(!list_empty(&group->device_list)); 293 260 WARN_ON(group->container || group->container_users); 294 261 WARN_ON(group->notifier.head); 295 - 296 - list_del(&group->vfio_next); 297 - cdev_device_del(&group->cdev, &group->dev); 298 - mutex_unlock(&vfio.group_lock); 262 + group->iommu_group = NULL; 299 263 300 264 put_device(&group->dev); 301 - } 302 - 303 - static void vfio_group_get(struct vfio_group *group) 304 - { 305 - refcount_inc(&group->users); 306 265 } 307 266 308 267 /* ··· 519 494 struct vfio_device *existing_device; 520 495 int ret; 521 496 497 + /* 498 + * In all cases group is the output of one of the group allocation 499 + * functions and we have group->drivers incremented for us. 500 + */ 522 501 if (IS_ERR(group)) 523 502 return PTR_ERR(group); 524 503 ··· 562 533 563 534 return 0; 564 535 err_out: 565 - if (group->type == VFIO_NO_IOMMU || 566 - group->type == VFIO_EMULATED_IOMMU) 567 - iommu_group_remove_device(device->dev); 568 - vfio_group_put(group); 536 + vfio_device_remove_group(device); 569 537 return ret; 570 538 } 571 539 ··· 653 627 /* Balances device_add in register path */ 654 628 device_del(&device->device); 655 629 656 - if (group->type == VFIO_NO_IOMMU || group->type == VFIO_EMULATED_IOMMU) 657 - iommu_group_remove_device(device->dev); 658 - 659 - /* Matches the get in vfio_register_group_dev() */ 660 - vfio_group_put(group); 630 + vfio_device_remove_group(device); 661 631 } 662 632 EXPORT_SYMBOL_GPL(vfio_unregister_group_dev); 663 633 ··· 906 884 907 885 down_write(&group->group_rwsem); 908 886 909 - /* users can be zero if this races with vfio_group_put() */ 887 + /* users can be zero if this races with vfio_device_remove_group() */ 910 888 if (!refcount_inc_not_zero(&group->users)) { 911 889 ret = -ENODEV; 912 890 goto err_unlock;