From: Shiraz Saleem <shiraz.saleem@intel.com>
To: stable@vger.kernel.org
Cc: jgg@nvidia.com, "Saleem, Shiraz" <shiraz.saleem@intel.com>,
Di Zhu <zhudi21@huawei.com>
Subject: [PATCH 4.14-stable] RDMA/i40iw: Address an mmap handler exploit in i40iw
Date: Wed, 2 Dec 2020 11:20:09 -0600 [thread overview]
Message-ID: <20201202172012.801-1-shiraz.saleem@intel.com> (raw)
From: "Saleem, Shiraz" <shiraz.saleem@intel.com>
backport of commit 2ed381439e89fa6d1a0839ef45ccd45d99d8e915 upstream.
i40iw_mmap manipulates the vma->vm_pgoff to differentiate a push page mmap
vs a doorbell mmap, and uses it to compute the pfn in remap_pfn_range
without any validation. This is vulnerable to an mmap exploit as described
in: https://lore.kernel.org/r/20201119093523.7588-1-zhudi21@huawei.com
The push feature is disabled in the driver currently and therefore no push
mmaps are issued from user-space. The feature does not work as expected in
the x722 product.
Remove the push module parameter and all VMA attribute manipulations for
this feature in i40iw_mmap. Update i40iw_mmap to only allow DB user
mmapings at offset = 0. Check vm_pgoff for zero and if the mmaps are bound
to a single page.
Fixes: d37498417947 ("i40iw: add files for iwarp interface")
Link: https://lore.kernel.org/r/20201125005616.1800-2-shiraz.saleem@intel.com
Reported-by: Di Zhu <zhudi21@huawei.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
---
drivers/infiniband/hw/i40iw/i40iw_main.c | 5 -----
drivers/infiniband/hw/i40iw/i40iw_verbs.c | 36 ++++++-------------------------
2 files changed, 7 insertions(+), 34 deletions(-)
diff --git a/drivers/infiniband/hw/i40iw/i40iw_main.c b/drivers/infiniband/hw/i40iw/i40iw_main.c
index 27590ae..f41ac28 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_main.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_main.c
@@ -54,10 +54,6 @@
#define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \
__stringify(DRV_VERSION_MINOR) "." __stringify(DRV_VERSION_BUILD)
-static int push_mode;
-module_param(push_mode, int, 0644);
-MODULE_PARM_DESC(push_mode, "Low latency mode: 0=disabled (default), 1=enabled)");
-
static int debug;
module_param(debug, int, 0644);
MODULE_PARM_DESC(debug, "debug flags: 0=disabled (default), 0x7fffffff=all");
@@ -1564,7 +1560,6 @@ static enum i40iw_status_code i40iw_setup_init_state(struct i40iw_handler *hdl,
if (status)
goto exit;
iwdev->obj_next = iwdev->obj_mem;
- iwdev->push_mode = push_mode;
init_waitqueue_head(&iwdev->vchnl_waitq);
init_waitqueue_head(&dev->vf_reqs);
diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 57bfe48..cc943b4 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -199,38 +199,16 @@ static int i40iw_dealloc_ucontext(struct ib_ucontext *context)
*/
static int i40iw_mmap(struct ib_ucontext *context, struct vm_area_struct *vma)
{
- struct i40iw_ucontext *ucontext;
- u64 db_addr_offset;
- u64 push_offset;
-
- ucontext = to_ucontext(context);
- if (ucontext->iwdev->sc_dev.is_pf) {
- db_addr_offset = I40IW_DB_ADDR_OFFSET;
- push_offset = I40IW_PUSH_OFFSET;
- if (vma->vm_pgoff)
- vma->vm_pgoff += I40IW_PF_FIRST_PUSH_PAGE_INDEX - 1;
- } else {
- db_addr_offset = I40IW_VF_DB_ADDR_OFFSET;
- push_offset = I40IW_VF_PUSH_OFFSET;
- if (vma->vm_pgoff)
- vma->vm_pgoff += I40IW_VF_FIRST_PUSH_PAGE_INDEX - 1;
- }
+ struct i40iw_ucontext *ucontext = to_ucontext(context);
+ u64 dbaddr;
- vma->vm_pgoff += db_addr_offset >> PAGE_SHIFT;
+ if (vma->vm_pgoff || vma->vm_end - vma->vm_start != PAGE_SIZE)
+ return -EINVAL;
- if (vma->vm_pgoff == (db_addr_offset >> PAGE_SHIFT)) {
- vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
- vma->vm_private_data = ucontext;
- } else {
- if ((vma->vm_pgoff - (push_offset >> PAGE_SHIFT)) % 2)
- vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
- else
- vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
- }
+ dbaddr = I40IW_DB_ADDR_OFFSET + pci_resource_start(ucontext->iwdev->ldev->pcidev, 0);
- if (io_remap_pfn_range(vma, vma->vm_start,
- vma->vm_pgoff + (pci_resource_start(ucontext->iwdev->ldev->pcidev, 0) >> PAGE_SHIFT),
- PAGE_SIZE, vma->vm_page_prot))
+ if (io_remap_pfn_range(vma, vma->vm_start, dbaddr >> PAGE_SHIFT, PAGE_SIZE,
+ pgprot_noncached(vma->vm_page_prot)))
return -EAGAIN;
return 0;
--
1.8.3.1
next reply other threads:[~2020-12-02 17:21 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-02 17:20 Shiraz Saleem [this message]
2020-12-02 17:20 ` [PATCH 4.19-stable] RDMA/i40iw: Address an mmap handler exploit in i40iw Shiraz Saleem
2020-12-02 17:20 ` [PATCH 4.9-stable] " Shiraz Saleem
2020-12-02 17:20 ` [PATCH 5.4-stable] " Shiraz Saleem
2020-12-06 9:53 ` [PATCH 4.14-stable] " Greg KH
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=20201202172012.801-1-shiraz.saleem@intel.com \
--to=shiraz.saleem@intel.com \
--cc=jgg@nvidia.com \
--cc=stable@vger.kernel.org \
--cc=zhudi21@huawei.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;
as well as URLs for NNTP newsgroup(s).