[PATCH v2 0/2] Fix use after free in HPT resizing code and related minor improvements

[PATCH v2 0/2] Fix use after free in HPT resizing code and related minor improvements

[Serhii Popovych]

It is possible to trigger use after free during HPT resize
causing host kernel to crash. More details and analysis of
the problem can be found in change with corresponding subject
(KVM: PPC: Book3S HV: Fix use after free in case of multiple
resize requests).

We need some changes to prepare for the fix, especially
make ->error in HPT resize instance single point for
tracking allocation state.

See individual commit description message to get more
information on changes presented.

v2:
 Serhii Popovych: Tested with current 4.15-rc2 as host kernel on P8
                  with same testcase as for v1: no problems so far.

Serhii Popovych (2):
  KVM: PPC: Book3S HV: Drop prepare_done from struct kvm_resize_hpt and
    cleanups
  KVM: PPC: Book3S HV: Fix use after free in case of multiple resize
    requests

 arch/powerpc/kvm/book3s_64_mmu_hv.c | 84 +++++++++++++++++++++++++------------
 1 file changed, 57 insertions(+), 27 deletions(-)

-- 
1.8.3.1

[PATCH v2 1/2] KVM: PPC: Book3S HV: Drop prepare_done from struct kvm_resize_hpt and cleanups

[Serhii Popovych]

Currently the kvm_resize_hpt structure has two fields relevant to the
state of an ongoing resize: 'prepare_done', which indicates whether
the worker thread has completed or not, and 'error' which indicates
whether it was successful or not.

Since the success/failure isn't known until completion, this is
confusingly redundant.  This patch consolidates the information into
just the 'error' value: -EBUSY indicates the worked is still in
progress, other negative values indicate (completed) failure, 0
indicates successful completion.

As a bonus this reduces size of struct kvm_resize_hpt by
__alignof__(struct kvm_hpt_info) and saves few bytes of code.

While there correct comment in struct kvm_resize_hpt which references
a non-existent semaphore (leftover from an early draft).

Assert with WARN_ON() in case of HPT allocation thread work runs more
than once for resize request or resize_hpt_allocate() returns -EBUSY
that is treated specially.

Change comparison against zero to make checkpatch.pl happy.

Signed-off-by: Serhii Popovych <spopovyc@redhat.com>
[dwg: Changed BUG_ON()s to WARN_ON()s and altered commit message for
 clarity]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 arch/powerpc/kvm/book3s_64_mmu_hv.c | 44 +++++++++++++++++++++++--------------
 1 file changed, 27 insertions(+), 17 deletions(-)

diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index 9660972..31fa710 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -65,11 +65,17 @@ struct kvm_resize_hpt {
 	u32 order;
 
 	/* These fields protected by kvm->lock */
+
+	/* Possible values and their usage:
+	 *  <0     an error occurred during allocation,
+	  * -EBUSY allocation is in the progress,
+	 *  0      allocation made successfuly.
+	 */
 	int error;
-	bool prepare_done;
 
-	/* Private to the work thread, until prepare_done is true,
-	 * then protected by kvm->resize_hpt_sem */
+	/* Private to the work thread, until error != -EBUSY,
+	 * then protected by kvm->lock.
+	 */
 	struct kvm_hpt_info hpt;
 };
 
@@ -1433,15 +1439,23 @@ static void resize_hpt_prepare_work(struct work_struct *work)
 	struct kvm *kvm = resize->kvm;
 	int err;
 
+	if (WARN_ON(resize->error != -EBUSY))
+		return;
+
 	resize_hpt_debug(resize, "resize_hpt_prepare_work(): order = %d\n",
 			 resize->order);
 
 	err = resize_hpt_allocate(resize);
 
+	/* We have strict assumption about -EBUSY
+	 * when preparing for HPT resize.
+	 */
+	if (WARN_ON(err == -EBUSY))
+		err = -EINPROGRESS;
+
 	mutex_lock(&kvm->lock);
 
 	resize->error = err;
-	resize->prepare_done = true;
 
 	mutex_unlock(&kvm->lock);
 }
@@ -1466,14 +1480,12 @@ long kvm_vm_ioctl_resize_hpt_prepare(struct kvm *kvm,
 
 	if (resize) {
 		if (resize->order == shift) {
-			/* Suitable resize in progress */
-			if (resize->prepare_done) {
-				ret = resize->error;
-				if (ret != 0)
-					resize_hpt_release(kvm, resize);
-			} else {
+			/* Suitable resize in progress? */
+			ret = resize->error;
+			if (ret == -EBUSY)
 				ret = 100; /* estimated time in ms */
-			}
+			else if (ret)
+				resize_hpt_release(kvm, resize);
 
 			goto out;
 		}
@@ -1493,6 +1505,8 @@ long kvm_vm_ioctl_resize_hpt_prepare(struct kvm *kvm,
 		ret = -ENOMEM;
 		goto out;
 	}
+
+	resize->error = -EBUSY;
 	resize->order = shift;
 	resize->kvm = kvm;
 	INIT_WORK(&resize->work, resize_hpt_prepare_work);
@@ -1547,16 +1561,12 @@ long kvm_vm_ioctl_resize_hpt_commit(struct kvm *kvm,
 	if (!resize || (resize->order != shift))
 		goto out;
 
-	ret = -EBUSY;
-	if (!resize->prepare_done)
-		goto out;
-
 	ret = resize->error;
-	if (ret != 0)
+	if (ret)
 		goto out;
 
 	ret = resize_hpt_rehash(resize);
-	if (ret != 0)
+	if (ret)
 		goto out;
 
 	resize_hpt_pivot(resize);
-- 
1.8.3.1

[PATCH v2 2/2] KVM: PPC: Book3S HV: Fix use after free in case of multiple resize requests

[Serhii Popovych]

When serving multiple resize requests following could happen:

    CPU0                                    CPU1
    ----                                    ----
    kvm_vm_ioctl_resize_hpt_prepare(1);
      -> schedule_work()
                                            /* system_rq might be busy: delay */
    kvm_vm_ioctl_resize_hpt_prepare(2);
      mutex_lock();
      if (resize) {
         ...
         release_hpt_resize();
      }
      ...                                   resize_hpt_prepare_work()
      -> schedule_work()                    {
      mutex_unlock()                           /* resize->kvm could be wrong */
                                               struct kvm *kvm = resize->kvm;

                                               mutex_lock(&kvm->lock);   <<<< UAF
                                               ...
                                            }

i.e. a second resize request with different order could be started by
kvm_vm_ioctl_resize_hpt_prepare(), causing the previous request to be
free()d when there's still an active worker thread which will try to
access it.  This leads to a use after free in point marked with UAF on
the diagram above.

To prevent this from happening, instead of unconditionally releasing a
pre-existing resize structure from the prepare ioctl(), we check if
the existing structure has an in-progress worker.  We do that by
checking if the resize->error == -EBUSY, which is safe because the
resize->error field is protected by the kvm->lock.  If there is an
active worker, instead of releasing, we mark the structure as stale by
unlinking it from kvm_struct.

In the worker thread we check for a stale structure (with kvm->lock
held), and in that case abort, releasing the stale structure ourself.
We make the check both before and the actual allocation.  Strictly,
only the check afterwards is needed, the check before is an
optimization: if the structure happens to become stale before the
worker thread is dispatched, rather than during the allocation, it
means we can avoid allocating then immediately freeing a potentially
substantial amount of memory.

This fixes following or similar host kernel crash message:

[  635.277361] Unable to handle kernel paging request for data at address 0x00000000
[  635.277438] Faulting instruction address: 0xc00000000052f568
[  635.277446] Oops: Kernel access of bad area, sig: 11 [#1]
[  635.277451] SMP NR_CPUS=2048 NUMA PowerNV
[  635.277470] Modules linked in: xt_CHECKSUM iptable_mangle ipt_MASQUERADE
nf_nat_masquerade_ipv4 iptable_nat nf_nat_ipv4 nf_nat nf_conntrack_ipv4
nf_defrag_ipv4 xt_conntrack nf_conntrack ipt_REJECT nf_reject_ipv4 tun bridge stp llc
ebtable_filter ebtables ip6table_filter ip6_tables iptable_filter nfsv3 nfs_acl nfs
lockd grace fscache kvm_hv kvm rpcrdma sunrpc ib_isert iscsi_target_mod ib_iser libiscsi
scsi_transport_iscsi ib_srpt target_core_mod ext4 ib_srp scsi_transport_srp
ib_ipoib mbcache jbd2 rdma_ucm ib_ucm ib_uverbs ib_umad rdma_cm ib_cm iw_cm ocrdma(T)
ib_core ses enclosure scsi_transport_sas sg shpchp leds_powernv ibmpowernv i2c_opal
i2c_core powernv_rng ipmi_powernv ipmi_devintf ipmi_msghandler ip_tables xfs
libcrc32c sr_mod sd_mod cdrom lpfc nvme_fc(T) nvme_fabrics nvme_core ipr nvmet_fc(T)
tg3 nvmet libata be2net crc_t10dif crct10dif_generic scsi_transport_fc ptp scsi_tgt
pps_core crct10dif_common dm_mirror dm_region_hash dm_log dm_mod
[  635.278687] CPU: 40 PID: 749 Comm: kworker/40:1 Tainted: G
------------ T 3.10.0.bz1510771+ #1
[  635.278782] Workqueue: events resize_hpt_prepare_work [kvm_hv]
[  635.278851] task: c0000007e6840000 ti: c0000007e9180000 task.ti: c0000007e9180000
[  635.278919] NIP: c00000000052f568 LR: c0000000009ea310 CTR: c0000000009ea4f0
[  635.278988] REGS: c0000007e91837f0 TRAP: 0300   Tainted: G
------------ T  (3.10.0.bz1510771+)
[  635.279077] MSR: 9000000100009033 <SF,HV,EE,ME,IR,DR,RI,LE>  CR: 24002022  XER:
00000000
[  635.279248] CFAR: c000000000009368 DAR: 0000000000000000 DSISR: 40000000 SOFTE: 1
GPR00: c0000000009ea310 c0000007e9183a70 c000000001250b00 c0000007e9183b10
GPR04: 0000000000000000 0000000000000000 c0000007e9183650 0000000000000000
GPR08: c0000007ffff7b80 00000000ffffffff 0000000080000028 d00000000d2529a0
GPR12: 0000000000002200 c000000007b56800 c000000000120028 c0000007f135bb40
GPR16: 0000000000000000 c000000005c1e018 c000000005c1e018 0000000000000000
GPR20: 0000000000000001 c0000000011bf778 0000000000000001 fffffffffffffef7
GPR24: 0000000000000000 c000000f1e262e50 0000000000000002 c0000007e9180000
GPR28: c000000f1e262e4c c000000f1e262e50 0000000000000000 c0000007e9183b10
[  635.280149] NIP [c00000000052f568] __list_add+0x38/0x110
[  635.280197] LR [c0000000009ea310] __mutex_lock_slowpath+0xe0/0x2c0
[  635.280253] Call Trace:
[  635.280277] [c0000007e9183af0] [c0000000009ea310] __mutex_lock_slowpath+0xe0/0x2c0
[  635.280356] [c0000007e9183b70] [c0000000009ea554] mutex_lock+0x64/0x70
[  635.280426] [c0000007e9183ba0] [d00000000d24da04]
resize_hpt_prepare_work+0xe4/0x1c0 [kvm_hv]
[  635.280507] [c0000007e9183c40] [c000000000113c0c] process_one_work+0x1dc/0x680
[  635.280587] [c0000007e9183ce0] [c000000000114250] worker_thread+0x1a0/0x520
[  635.280655] [c0000007e9183d80] [c00000000012010c] kthread+0xec/0x100
[  635.280724] [c0000007e9183e30] [c00000000000a4b8] ret_from_kernel_thread+0x5c/0xa4
[  635.280814] Instruction dump:
[  635.280880] 7c0802a6 fba1ffe8 fbc1fff0 7cbd2b78 fbe1fff8 7c9e2378 7c7f1b78
f8010010
[  635.281099] f821ff81 e8a50008 7fa52040 40de00b8 <e8be0000> 7fbd2840 40de008c
7fbff040
[  635.281324] ---[ end trace b628b73449719b9d ]---

Signed-off-by: Serhii Popovych <spopovyc@redhat.com>
[dwg: Replaced BUG_ON()s with WARN_ONs() and reworded commit message
 for clarity]
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
---
 arch/powerpc/kvm/book3s_64_mmu_hv.c | 50 ++++++++++++++++++++++++++-----------
 1 file changed, 35 insertions(+), 15 deletions(-)

diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
index 31fa710..7948b84 100644
--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -1419,16 +1419,20 @@ static void resize_hpt_pivot(struct kvm_resize_hpt *resize)
 
 static void resize_hpt_release(struct kvm *kvm, struct kvm_resize_hpt *resize)
 {
-	BUG_ON(kvm->arch.resize_hpt != resize);
+	if (WARN_ON(!mutex_is_locked(&kvm->lock)))
+		return;
 
 	if (!resize)
 		return;
 
-	if (resize->hpt.virt)
-		kvmppc_free_hpt(&resize->hpt);
+	if (resize->error != -EBUSY) {
+		if (resize->hpt.virt)
+			kvmppc_free_hpt(&resize->hpt);
+		kfree(resize);
+	}
 
-	kvm->arch.resize_hpt = NULL;
-	kfree(resize);
+	if (kvm->arch.resize_hpt == resize)
+		kvm->arch.resize_hpt = NULL;
 }
 
 static void resize_hpt_prepare_work(struct work_struct *work)
@@ -1437,26 +1441,42 @@ static void resize_hpt_prepare_work(struct work_struct *work)
 						     struct kvm_resize_hpt,
 						     work);
 	struct kvm *kvm = resize->kvm;
-	int err;
+	int err = 0;
 
 	if (WARN_ON(resize->error != -EBUSY))
 		return;
 
-	resize_hpt_debug(resize, "resize_hpt_prepare_work(): order = %d\n",
-			 resize->order);
+	mutex_lock(&kvm->lock);
 
-	err = resize_hpt_allocate(resize);
+	/* Request is still current? */
+	if (kvm->arch.resize_hpt == resize) {
+		/* We may request large allocations here:
+		 * do not sleep with kvm->lock held for a while.
+		 */
+		mutex_unlock(&kvm->lock);
 
-	/* We have strict assumption about -EBUSY
-	 * when preparing for HPT resize.
-	 */
-	if (WARN_ON(err == -EBUSY))
-		err = -EINPROGRESS;
+		resize_hpt_debug(resize, "resize_hpt_prepare_work(): order = %d\n",
+				 resize->order);
 
-	mutex_lock(&kvm->lock);
+		err = resize_hpt_allocate(resize);
+
+		/* We have strict assumption about -EBUSY
+		 * when preparing for HPT resize.
+		 */
+		if (WARN_ON(err == -EBUSY))
+			err = -EINPROGRESS;
+
+		mutex_lock(&kvm->lock);
+		/* It is possible that kvm->arch.resize_hpt != resize
+		 * after we grab kvm->lock again.
+		 */
+	}
 
 	resize->error = err;
 
+	if (kvm->arch.resize_hpt != resize)
+		resize_hpt_release(kvm, resize);
+
 	mutex_unlock(&kvm->lock);
 }
 
-- 
1.8.3.1

Re: [PATCH v2 1/2] KVM: PPC: Book3S HV: Drop prepare_done from struct kvm_resize_hpt and cleanups

[Joe Perches]

On Mon, 2017-12-04 at 09:36 -0500, Serhii Popovych wrote:
> Change comparison against zero to make checkpatch.pl happy.

Huh?  Either I'm confused or you are incorrect here.

checkpatch should or does not warn against comparisons
to zero.  There is a test for comparison to NULL.

> diff --git a/arch/powerpc/kvm/book3s_64_mmu_hv.c b/arch/powerpc/kvm/book3s_64_mmu_hv.c
[]
> @@ -1547,16 +1561,12 @@ long kvm_vm_ioctl_resize_hpt_commit(struct kvm *kvm,
>  	if (!resize || (resize->order != shift))
>  		goto out;
>  
> -	ret = -EBUSY;
> -	if (!resize->prepare_done)
> -		goto out;
> -
>  	ret = resize->error;
> -	if (ret != 0)
> +	if (ret)
>  		goto out;
>  
>  	ret = resize_hpt_rehash(resize);
> -	if (ret != 0)
> +	if (ret)
>  		goto out;
>  
>  	resize_hpt_pivot(resize);