From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Will Deacon <will.deacon@arm.com>,
Wei Huang <wei@redhat.com>, Marc Zyngier <marc.zyngier@arm.com>
Subject: [PATCH 4.8 05/67] KVM: arm64: Fix the issues when guest PMCCFILTR is configured
Date: Thu, 24 Nov 2016 16:26:59 +0100 [thread overview]
Message-ID: <20161124145457.487432214@linuxfoundation.org> (raw)
In-Reply-To: <20161124145457.061710350@linuxfoundation.org>
4.8-stable review patch. If anyone has any objections, please let me know.
------------------
From: Wei Huang <wei@redhat.com>
commit b112c84a6ff035271d41d548c10215f18443d6a6 upstream.
KVM calls kvm_pmu_set_counter_event_type() when PMCCFILTR is configured.
But this function can't deals with PMCCFILTR correctly because the evtCount
bits of PMCCFILTR, which is reserved 0, conflits with the SW_INCR event
type of other PMXEVTYPER<n> registers. To fix it, when eventsel == 0, this
function shouldn't return immediately; instead it needs to check further
if select_idx is ARMV8_PMU_CYCLE_IDX.
Another issue is that KVM shouldn't copy the eventsel bits of PMCCFILTER
blindly to attr.config. Instead it ought to convert the request to the
"cpu cycle" event type (i.e. 0x11).
To support this patch and to prevent duplicated definitions, a limited
set of ARMv8 perf event types were relocated from perf_event.c to
asm/perf_event.h.
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Wei Huang <wei@redhat.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
arch/arm64/include/asm/perf_event.h | 10 +++++++++-
arch/arm64/kernel/perf_event.c | 10 +---------
virt/kvm/arm/pmu.c | 8 +++++---
3 files changed, 15 insertions(+), 13 deletions(-)
--- a/arch/arm64/include/asm/perf_event.h
+++ b/arch/arm64/include/asm/perf_event.h
@@ -46,7 +46,15 @@
#define ARMV8_PMU_EVTYPE_MASK 0xc800ffff /* Mask for writable bits */
#define ARMV8_PMU_EVTYPE_EVENT 0xffff /* Mask for EVENT bits */
-#define ARMV8_PMU_EVTYPE_EVENT_SW_INCR 0 /* Software increment event */
+/*
+ * PMUv3 event types: required events
+ */
+#define ARMV8_PMUV3_PERFCTR_SW_INCR 0x00
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL 0x03
+#define ARMV8_PMUV3_PERFCTR_L1D_CACHE 0x04
+#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED 0x10
+#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES 0x11
+#define ARMV8_PMUV3_PERFCTR_BR_PRED 0x12
/*
* Event filters for PMUv3
--- a/arch/arm64/kernel/perf_event.c
+++ b/arch/arm64/kernel/perf_event.c
@@ -30,17 +30,9 @@
/*
* ARMv8 PMUv3 Performance Events handling code.
- * Common event types.
+ * Common event types (some are defined in asm/perf_event.h).
*/
-/* Required events. */
-#define ARMV8_PMUV3_PERFCTR_SW_INCR 0x00
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE_REFILL 0x03
-#define ARMV8_PMUV3_PERFCTR_L1D_CACHE 0x04
-#define ARMV8_PMUV3_PERFCTR_BR_MIS_PRED 0x10
-#define ARMV8_PMUV3_PERFCTR_CPU_CYCLES 0x11
-#define ARMV8_PMUV3_PERFCTR_BR_PRED 0x12
-
/* At least one of the following is required. */
#define ARMV8_PMUV3_PERFCTR_INST_RETIRED 0x08
#define ARMV8_PMUV3_PERFCTR_INST_SPEC 0x1B
--- a/virt/kvm/arm/pmu.c
+++ b/virt/kvm/arm/pmu.c
@@ -305,7 +305,7 @@ void kvm_pmu_software_increment(struct k
continue;
type = vcpu_sys_reg(vcpu, PMEVTYPER0_EL0 + i)
& ARMV8_PMU_EVTYPE_EVENT;
- if ((type == ARMV8_PMU_EVTYPE_EVENT_SW_INCR)
+ if ((type == ARMV8_PMUV3_PERFCTR_SW_INCR)
&& (enable & BIT(i))) {
reg = vcpu_sys_reg(vcpu, PMEVCNTR0_EL0 + i) + 1;
reg = lower_32_bits(reg);
@@ -379,7 +379,8 @@ void kvm_pmu_set_counter_event_type(stru
eventsel = data & ARMV8_PMU_EVTYPE_EVENT;
/* Software increment event does't need to be backed by a perf event */
- if (eventsel == ARMV8_PMU_EVTYPE_EVENT_SW_INCR)
+ if (eventsel == ARMV8_PMUV3_PERFCTR_SW_INCR &&
+ select_idx != ARMV8_PMU_CYCLE_IDX)
return;
memset(&attr, 0, sizeof(struct perf_event_attr));
@@ -391,7 +392,8 @@ void kvm_pmu_set_counter_event_type(stru
attr.exclude_kernel = data & ARMV8_PMU_EXCLUDE_EL1 ? 1 : 0;
attr.exclude_hv = 1; /* Don't count EL2 events */
attr.exclude_host = 1; /* Don't count host events */
- attr.config = eventsel;
+ attr.config = (select_idx == ARMV8_PMU_CYCLE_IDX) ?
+ ARMV8_PMUV3_PERFCTR_CPU_CYCLES : eventsel;
counter = kvm_pmu_get_counter_value(vcpu, select_idx);
/* The initial sample period (overflow count) of an event. */
next prev parent reply other threads:[~2016-11-24 15:29 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-24 15:26 [PATCH 4.8 00/67] 4.8.11-stable review Greg Kroah-Hartman
2016-11-24 15:26 ` [PATCH 4.8 01/67] x86/cpu/AMD: Fix cpu_llc_id for AMD Fam17h systems Greg Kroah-Hartman
2016-11-24 15:26 ` [PATCH 4.8 04/67] arm64: KVM: pmu: Fix AArch32 cycle counter access Greg Kroah-Hartman
2016-11-24 15:26 ` Greg Kroah-Hartman [this message]
2016-11-24 15:27 ` [PATCH 4.8 06/67] ftrace: Ignore FTRACE_FL_DISABLED while walking dyn_ftrace records Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 07/67] ftrace: Add more checks for FTRACE_FL_DISABLED in processing ip records Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 08/67] genirq: Use irq type from irqdata instead of irqdesc Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 09/67] fuse: fix fuse_write_end() if zero bytes were copied Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 10/67] IB/rdmavt: rdmavt can handle non aligned page maps Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 11/67] IB/hfi1: Fix rnr_timer addition Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 12/67] mfd: intel-lpss: Do not put device in reset state on suspend Greg Kroah-Hartman
2016-11-24 15:30 ` Shaikh, Azhar
2016-11-24 15:27 ` [PATCH 4.8 13/67] mfd: stmpe: Fix RESET regression on STMPE2401 Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 14/67] can: bcm: fix warning in bcm_connect/proc_register Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 15/67] gpio: do not double-check direction on sleeping chips Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 16/67] ALSA: usb-audio: Fix use-after-free of usb_device at disconnect Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 17/67] ALSA: hda - add a new condition to check if it is thinkpad Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 18/67] ALSA: hda - Fix mic regression by ASRock mobo fixup Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 19/67] i2c: mux: fix up dependencies Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 20/67] i2c: i2c-mux-pca954x: fix deselect enabling for device-tree Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 21/67] Disable the __builtin_return_address() warning globally after all Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 23/67] scripts/has-stack-protector: add -fno-PIE Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 24/67] x86/kexec: " Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 25/67] kbuild: Steal gccs pie from the very beginning Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 26/67] ext4: sanity check the block and cluster size at mount time Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 27/67] ARM: dts: imx53-qsb: Fix regulator constraints Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 29/67] powerpc/64: Fix setting of AIL in hypervisor mode Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 33/67] virtio-net: drop legacy features in virtio 1 mode Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 34/67] clk: mmp: pxa910: fix return value check in pxa910_clk_init() Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 35/67] clk: mmp: pxa168: fix return value check in pxa168_clk_init() Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 36/67] clk: mmp: mmp2: fix return value check in mmp2_clk_init() Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 37/67] clk: imx: fix integer overflow in AV PLL round rate Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 38/67] rtc: omap: Fix selecting external osc Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 39/67] iwlwifi: pcie: fix SPLC structure parsing Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 40/67] iwlwifi: pcie: mark command queue lock with separate lockdep class Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 41/67] iwlwifi: mvm: fix netdetect starting/stopping for unified images Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 42/67] iwlwifi: mvm: fix d3_test with unified D0/D3 images Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 43/67] iwlwifi: mvm: wake the wait queue when the RX sync counter is zero Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 44/67] mfd: core: Fix device reference leak in mfd_clone_cell Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 45/67] sunrpc: svc_age_temp_xprts_now should not call setsockopt non-tcp transports Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 46/67] uwb: fix device reference leaks Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 47/67] PM / sleep: fix device reference leak in test_suspend Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 48/67] PM / sleep: dont suspend parent when async child suspend_{noirq, late} fails Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 49/67] perf hists: Fix column length on --hierarchy Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 50/67] IB/rxe: Update qp state for user query Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 51/67] IB/rxe: Fix kernel panic in UDP tunnel with GRO and RX checksum Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 52/67] IB/rxe: Fix handling of erroneous WR Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 53/67] IB/rxe: Clear queue buffer when modifying QP to reset Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 54/67] IB/mlx4: Check gid_index return value Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 55/67] IB/mlx4: Fix create CQ error flow Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 56/67] IB/mlx5: Validate requested RQT size Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 57/67] IB/mlx5: Use cache line size to select CQE stride Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 58/67] IB/mlx5: Fix memory leak in query device Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 59/67] IB/mlx5: Fix fatal error dispatching Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 60/67] IB/mlx5: Fix NULL pointer dereference on debug print Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 61/67] IB/core: Avoid unsigned int overflow in sg_alloc_table Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 62/67] IB/hfi1: Remove incorrect IS_ERR check Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 63/67] IB/uverbs: Fix leak of XRC target QPs Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 64/67] IB/cm: Mark stale CM ids whenever the mad agent was unregistered Greg Kroah-Hartman
2016-11-24 15:27 ` [PATCH 4.8 65/67] netfilter: nft_dynset: fix element timeout for HZ != 1000 Greg Kroah-Hartman
2016-11-24 15:28 ` [PATCH 4.8 66/67] gpio: pca953x: Move memcpy into mutex lock for set multiple Greg Kroah-Hartman
2016-11-24 15:28 ` [PATCH 4.8 67/67] gpio: pca953x: Fix corruption of other gpios in set_multiple Greg Kroah-Hartman
2016-11-25 0:48 ` [PATCH 4.8 00/67] 4.8.11-stable review Guenter Roeck
2016-11-25 9:43 ` Greg Kroah-Hartman
[not found] ` <5837ca28.52301c0a.d82bc.07dd@mx.google.com>
2016-11-25 9:46 ` Greg Kroah-Hartman
[not found] ` <m2mvgjtr7m.fsf@baylibre.com>
2016-11-28 19:22 ` Javier Martinez Canillas
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=20161124145457.487432214@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=marc.zyngier@arm.com \
--cc=stable@vger.kernel.org \
--cc=wei@redhat.com \
--cc=will.deacon@arm.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).