stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org,
	Liu Ping Fan <pingfank@linux.vnet.ibm.com>,
	Paul Mackerras <paulus@samba.org>, Alexander Graf <agraf@suse.de>
Subject: [PATCH 3.4 04/43] powerpc: kvm: fix rare but potential deadlock scene
Date: Mon,  6 Jan 2014 14:39:21 -0800	[thread overview]
Message-ID: <20140106223942.384564191@linuxfoundation.org> (raw)
In-Reply-To: <20140106223942.259651490@linuxfoundation.org>

3.4-stable review patch.  If anyone has any objections, please let me know.

------------------

From: pingfan liu <qemulist@gmail.com>

commit 91648ec09c1ef69c4d840ab6dab391bfb452d554 upstream.

Since kvmppc_hv_find_lock_hpte() is called from both virtmode and
realmode, so it can trigger the deadlock.

Suppose the following scene:

Two physical cpuM, cpuN, two VM instances A, B, each VM has a group of
vcpus.

If on cpuM, vcpu_A_1 holds bitlock X (HPTE_V_HVLOCK), then is switched
out, and on cpuN, vcpu_A_2 try to lock X in realmode, then cpuN will be
caught in realmode for a long time.

What makes things even worse if the following happens,
  On cpuM, bitlockX is hold, on cpuN, Y is hold.
  vcpu_B_2 try to lock Y on cpuM in realmode
  vcpu_A_2 try to lock X on cpuN in realmode

Oops! deadlock happens

Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
Reviewed-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 arch/powerpc/kvm/book3s_64_mmu_hv.c |    6 +++++-
 arch/powerpc/kvm/book3s_hv_rm_mmu.c |    4 ++++
 2 files changed, 9 insertions(+), 1 deletion(-)

--- a/arch/powerpc/kvm/book3s_64_mmu_hv.c
+++ b/arch/powerpc/kvm/book3s_64_mmu_hv.c
@@ -393,11 +393,14 @@ static int kvmppc_mmu_book3s_64_hv_xlate
 		slb_v = vcpu->kvm->arch.vrma_slb_v;
 	}
 
+	preempt_disable();
 	/* Find the HPTE in the hash table */
 	index = kvmppc_hv_find_lock_hpte(kvm, eaddr, slb_v,
 					 HPTE_V_VALID | HPTE_V_ABSENT);
-	if (index < 0)
+	if (index < 0) {
+		preempt_enable();
 		return -ENOENT;
+	}
 	hptep = (unsigned long *)(kvm->arch.hpt_virt + (index << 4));
 	v = hptep[0] & ~HPTE_V_HVLOCK;
 	gr = kvm->arch.revmap[index].guest_rpte;
@@ -405,6 +408,7 @@ static int kvmppc_mmu_book3s_64_hv_xlate
 	/* Unlock the HPTE */
 	asm volatile("lwsync" : : : "memory");
 	hptep[0] = v;
+	preempt_enable();
 
 	gpte->eaddr = eaddr;
 	gpte->vpage = ((v & HPTE_V_AVPN) << 4) | ((eaddr >> 12) & 0xfff);
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -649,6 +649,10 @@ static int slb_base_page_shift[4] = {
 	20,	/* 1M, unsupported */
 };
 
+/* When called from virtmode, this func should be protected by
+ * preempt_disable(), otherwise, the holding of HPTE_V_HVLOCK
+ * can trigger deadlock issue.
+ */
 long kvmppc_hv_find_lock_hpte(struct kvm *kvm, gva_t eaddr, unsigned long slb_v,
 			      unsigned long valid)
 {



  parent reply	other threads:[~2014-01-06 22:39 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-06 22:39 [PATCH 3.4 00/43] 3.4.76-stable review Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 01/43] USB: serial: fix race in generic write Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 02/43] ceph: cleanup aborted requests when re-sending requests Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 03/43] ceph: wake up safe waiters when unregistering request Greg Kroah-Hartman
2014-01-06 22:39 ` Greg Kroah-Hartman [this message]
2014-01-06 22:39 ` [PATCH 3.4 05/43] TTY: pmac_zilog, check existence of ports in pmz_console_init() Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 06/43] ASoC: wm8904: fix DSP mode B configuration Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 07/43] ALSA: Add SNDRV_PCM_STATE_PAUSED case in wait_for_avail function Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 09/43] selinux: fix broken peer recv check Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 10/43] selinux: selinux_setprocattr()->ptrace_parent() needs rcu_read_lock() Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 11/43] ftrace: Initialize the ftrace profiler for each possible cpu Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 12/43] intel_idle: initial IVB support Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 13/43] intel_idle: enable IVB Xeon support Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 14/43] ext4: fix use-after-free in ext4_mb_new_blocks Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 16/43] sched/rt: Fix rqs cpupri leak while enqueue/dequeue child RT entities Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 18/43] net_dma: mark broken Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 19/43] drm/radeon: fix asic gfx values for scrapper asics Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 20/43] drm/radeon: 0x9649 is SUMO2 not SUMO Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 21/43] ceph: Avoid data inconsistency due to d-cache aliasing in readpage() Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 22/43] tg3: Expand 4g_overflow_test workaround to skb fragments of any size Greg Kroah-Hartman
2014-01-06 22:58   ` Eric Dumazet
2014-01-06 23:05     ` Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 23/43] dm9601: fix reception of full size ethernet frames on dm9620/dm9621a Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 24/43] dm9601: work around tx fifo sync issue on dm962x Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 25/43] ath9k: Fix interrupt handling for the AR9002 family Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 26/43] ath9k_htc: properly set MAC address and BSSID mask Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 27/43] powerpc: Fix bad stack check in exception entry Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 28/43] powerpc: Align p_end Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 29/43] cpupower: Fix segfault due to incorrect getopt_long arugments Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 30/43] libata: add ATA_HORKAGE_BROKEN_FPDMA_AA quirk for Seagate Momentus SpinPoint M8 Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 31/43] radiotap: fix bitmap-end-finding buffer overrun Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 32/43] rtlwifi: pci: Fix oops on driver unload Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 33/43] mm/hugetlb: check for pte NULL pointer in __page_check_address() Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 34/43] Input: allocate absinfo data when setting ABS capability Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 35/43] GFS2: dont hold s_umount over blkdev_put Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 36/43] GFS2: Fix incorrect invalidation for DIO/buffered I/O Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 37/43] jbd2: dont BUG but return ENOSPC if a handle runs out of space Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 38/43] gpio: msm: Fix irq mask/unmask by writing bits instead of numbers Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 39/43] sched: Avoid throttle_cfs_rq() racing with period_timer stopping Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 40/43] sh: always link in helper functions extracted from libgcc Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 41/43] selinux: look for IPsec labels on both inbound and outbound packets Greg Kroah-Hartman
2014-01-06 22:39 ` [PATCH 3.4 42/43] selinux: process labeled IPsec TCP SYN-ACK packets properly in selinux_ip_postroute() Greg Kroah-Hartman
2014-01-06 22:40 ` [PATCH 3.4 43/43] hwmon: (w83l768ng) Fix fan speed control range Greg Kroah-Hartman
2014-01-07  5:01 ` [PATCH 3.4 00/43] 3.4.76-stable review Guenter Roeck
2014-01-07 15:22   ` Greg Kroah-Hartman
2014-01-07 19:09 ` Shuah Khan

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=20140106223942.384564191@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=agraf@suse.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=paulus@samba.org \
    --cc=pingfank@linux.vnet.ibm.com \
    --cc=stable@vger.kernel.org \
    /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).