From: Jason Gunthorpe <jgg@nvidia.com>
To: iommu@lists.linux.dev
Cc: Kevin Tian <kevin.tian@intel.com>,
Nicolin Chen <nicolinc@nvidia.com>,
patches@lists.linux.dev, Yi Liu <yi.l.liu@intel.com>
Subject: [PATCH] iommufd: Require drivers to supply the cache_invalidate_user ops
Date: Fri, 28 Jun 2024 13:11:11 -0300 [thread overview]
Message-ID: <0-v1-e153859bd707+61-iommufd_check_ops_jgg@nvidia.com> (raw)
If drivers don't do this then iommufd will oops invalidation ioctls with
something like:
Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
Mem abort info:
ESR = 0x0000000086000004
EC = 0x21: IABT (current EL), IL = 32 bits
SET = 0, FnV = 0
EA = 0, S1PTW = 0
FSC = 0x04: level 0 translation fault
user pgtable: 4k pages, 48-bit VAs, pgdp=0000000101059000
[0000000000000000] pgd=0000000000000000, p4d=0000000000000000
Internal error: Oops: 0000000086000004 [#1] PREEMPT SMP
Modules linked in:
CPU: 2 PID: 371 Comm: qemu-system-aar Not tainted 6.8.0-rc7-gde77230ac23a #9
Hardware name: linux,dummy-virt (DT)
pstate: 81400809 (Nzcv daif +PAN -UAO -TCO +DIT -SSBS BTYPE=-c)
pc : 0x0
lr : iommufd_hwpt_invalidate+0xa4/0x204
sp : ffff800080f3bcc0
x29: ffff800080f3bcf0 x28: ffff0000c369b300 x27: 0000000000000000
x26: 0000000000000000 x25: 0000000000000000 x24: 0000000000000000
x23: 0000000000000000 x22: 00000000c1e334a0 x21: ffff0000c1e334a0
x20: ffff800080f3bd38 x19: ffff800080f3bd58 x18: 0000000000000000
x17: 0000000000000000 x16: 0000000000000000 x15: 0000ffff8240d6d8
x14: 0000000000000000 x13: 0000000000000000 x12: 0000000000000000
x11: 0000000000000000 x10: 0000000000000000 x9 : 0000000000000000
x8 : 0000001000000002 x7 : 0000fffeac1ec950 x6 : 0000000000000000
x5 : ffff800080f3bd78 x4 : 0000000000000003 x3 : 0000000000000002
x2 : 0000000000000000 x1 : ffff800080f3bcc8 x0 : ffff0000c6034d80
Call trace:
0x0
iommufd_fops_ioctl+0x154/0x274
__arm64_sys_ioctl+0xac/0xf0
invoke_syscall+0x48/0x110
el0_svc_common.constprop.0+0x40/0xe0
do_el0_svc+0x1c/0x28
el0_svc+0x34/0xb4
el0t_64_sync_handler+0x120/0x12c
el0t_64_sync+0x190/0x194
All existing drivers implement this op for nesting, this is mostly a
bisection aid.
Fixes: 8c6eabae3807 ("iommufd: Add IOMMU_HWPT_INVALIDATE")
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
drivers/iommu/iommufd/hw_pagetable.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c
index 33d142f8057d70..a9f1fe44c4c0bd 100644
--- a/drivers/iommu/iommufd/hw_pagetable.c
+++ b/drivers/iommu/iommufd/hw_pagetable.c
@@ -236,7 +236,8 @@ iommufd_hwpt_nested_alloc(struct iommufd_ctx *ictx,
}
hwpt->domain->owner = ops;
- if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED)) {
+ if (WARN_ON_ONCE(hwpt->domain->type != IOMMU_DOMAIN_NESTED ||
+ !hwpt->domain->ops->cache_invalidate_user)) {
rc = -EINVAL;
goto out_abort;
}
base-commit: f2661062f16b2de5d7b6a5c42a9a5c96326b8454
--
2.45.2
next reply other threads:[~2024-06-28 16:11 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-28 16:11 Jason Gunthorpe [this message]
2024-06-28 21:27 ` [PATCH] iommufd: Require drivers to supply the cache_invalidate_user ops Nicolin Chen
2024-06-30 6:23 ` Yi Liu
2024-07-01 9:12 ` Tian, Kevin
2024-07-10 13:51 ` Jason Gunthorpe
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=0-v1-e153859bd707+61-iommufd_check_ops_jgg@nvidia.com \
--to=jgg@nvidia.com \
--cc=iommu@lists.linux.dev \
--cc=kevin.tian@intel.com \
--cc=nicolinc@nvidia.com \
--cc=patches@lists.linux.dev \
--cc=yi.l.liu@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox