qemu/hw/virtio/virtio-balloon.c

466 lines
14 KiB
C
Raw Normal View History

/*
* Virtio Balloon Device
*
* Copyright IBM, Corp. 2008
* Copyright (C) 2011 Red Hat, Inc.
* Copyright (C) 2011 Amit Shah <amit.shah@redhat.com>
*
* Authors:
* Anthony Liguori <aliguori@us.ibm.com>
*
* This work is licensed under the terms of the GNU GPL, version 2. See
* the COPYING file in the top-level directory.
*
*/
#include "qemu/iov.h"
#include "qemu/timer.h"
#include "qemu-common.h"
#include "hw/virtio/virtio.h"
#include "hw/i386/pc.h"
#include "cpu.h"
#include "sysemu/balloon.h"
#include "hw/virtio/virtio-balloon.h"
#include "sysemu/kvm.h"
#include "exec/address-spaces.h"
#include "qapi/visitor.h"
#include "qapi-event.h"
#include "trace.h"
#if defined(__linux__)
#include <sys/mman.h>
#endif
#include "hw/virtio/virtio-bus.h"
#include "hw/virtio/virtio-access.h"
static void balloon_page(void *addr, int deflate)
{
#if defined(__linux__)
if (!qemu_balloon_is_inhibited() && (!kvm_enabled() ||
kvm_has_sync_mmu())) {
Introduce qemu_madvise() vl.c has a Sun-specific hack to supply a prototype for madvise(), but the call site has apparently moved to arch_init.c. Haiku doesn't implement madvise() in favor of posix_madvise(). OpenBSD and Solaris 10 don't implement posix_madvise() but madvise(). MinGW implements neither. Check for madvise() and posix_madvise() in configure and supply qemu_madvise() as wrapper. Prefer madvise() over posix_madvise() due to flag availability. Convert all callers to use qemu_madvise() and QEMU_MADV_*. Note that on Solaris the warning is fixed by moving the madvise() prototype, not by qemu_madvise() itself. It helps with porting though, and it simplifies most call sites. v7 -> v8: * Some versions of MinGW have no sys/mman.h header. Reported by Blue Swirl. v6 -> v7: * Adopt madvise() rather than posix_madvise() semantics for returning errors. * Use EINVAL in place of ENOTSUP. v5 -> v6: * Replace two leftover instances of POSIX_MADV_NORMAL with QEMU_MADV_INVALID. Spotted by Blue Swirl. v4 -> v5: * Introduce QEMU_MADV_INVALID, suggested by Alexander Graf. Note that this relies on -1 not being a valid advice value. v3 -> v4: * Eliminate #ifdefs at qemu_advise() call sites. Requested by Blue Swirl. This will currently break the check in kvm-all.c by calling madvise() with a supported flag, which will not fail. Ideas/patches welcome. v2 -> v3: * Reuse the *_MADV_* defines for QEMU_MADV_*. Suggested by Alexander Graf. * Add configure check for madvise(), too. Add defines to Makefile, not QEMU_CFLAGS. Convert all callers, untested. Suggested by Blue Swirl. * Keep Solaris' madvise() prototype around. Pointed out by Alexander Graf. * Display configure check results. v1 -> v2: * Don't rely on posix_madvise() availability, add qemu_madvise(). Suggested by Blue Swirl. Signed-off-by: Andreas Färber <afaerber@opensolaris.org> Cc: Blue Swirl <blauwirbel@gmail.com> Cc: Alexander Graf <agraf@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Signed-off-by: Blue Swirl <blauwirbel@gmail.com>
2010-09-25 15:26:05 +04:00
qemu_madvise(addr, TARGET_PAGE_SIZE,
deflate ? QEMU_MADV_WILLNEED : QEMU_MADV_DONTNEED);
}
#endif
}
static const char *balloon_stat_names[] = {
[VIRTIO_BALLOON_S_SWAP_IN] = "stat-swap-in",
[VIRTIO_BALLOON_S_SWAP_OUT] = "stat-swap-out",
[VIRTIO_BALLOON_S_MAJFLT] = "stat-major-faults",
[VIRTIO_BALLOON_S_MINFLT] = "stat-minor-faults",
[VIRTIO_BALLOON_S_MEMFREE] = "stat-free-memory",
[VIRTIO_BALLOON_S_MEMTOT] = "stat-total-memory",
[VIRTIO_BALLOON_S_NR] = NULL
};
/*
* reset_stats - Mark all items in the stats array as unset
*
* This function needs to be called at device initialization and before
* updating to a set of newly-generated stats. This will ensure that no
* stale values stick around in case the guest reports a subset of the supported
* statistics.
*/
static inline void reset_stats(VirtIOBalloon *dev)
{
int i;
for (i = 0; i < VIRTIO_BALLOON_S_NR; dev->stats[i++] = -1);
}
static bool balloon_stats_supported(const VirtIOBalloon *s)
{
VirtIODevice *vdev = VIRTIO_DEVICE(s);
return virtio_vdev_has_feature(vdev, VIRTIO_BALLOON_F_STATS_VQ);
}
static bool balloon_stats_enabled(const VirtIOBalloon *s)
{
return s->stats_poll_interval > 0;
}
static void balloon_stats_destroy_timer(VirtIOBalloon *s)
{
if (balloon_stats_enabled(s)) {
timer_del(s->stats_timer);
timer_free(s->stats_timer);
s->stats_timer = NULL;
s->stats_poll_interval = 0;
}
}
static void balloon_stats_change_timer(VirtIOBalloon *s, int64_t secs)
{
timer_mod(s->stats_timer, qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL) + secs * 1000);
}
static void balloon_stats_poll_cb(void *opaque)
{
VirtIOBalloon *s = opaque;
VirtIODevice *vdev = VIRTIO_DEVICE(s);
if (!balloon_stats_supported(s)) {
/* re-schedule */
balloon_stats_change_timer(s, s->stats_poll_interval);
return;
}
virtqueue_push(s->svq, &s->stats_vq_elem, s->stats_vq_offset);
virtio_notify(vdev, s->svq);
}
static void balloon_stats_get_all(Object *obj, struct Visitor *v,
void *opaque, const char *name, Error **errp)
{
Error *err = NULL;
VirtIOBalloon *s = opaque;
int i;
visit_start_struct(v, NULL, "guest-stats", name, 0, &err);
if (err) {
goto out;
}
visit_type_int(v, &s->stats_last_update, "last-update", &err);
qapi: Replace uncommon use of the error API by the common one We commonly use the error API like this: err = NULL; foo(..., &err); if (err) { goto out; } bar(..., &err); Every error source is checked separately. The second function is only called when the first one succeeds. Both functions are free to pass their argument to error_set(). Because error_set() asserts no error has been set, this effectively means they must not be called with an error set. The qapi-generated code uses the error API differently: // *errp was initialized to NULL somewhere up the call chain frob(..., errp); gnat(..., errp); Errors accumulate in *errp: first error wins, subsequent errors get dropped. To make this work, the second function does nothing when called with an error set. Requires non-null errp, or else the second function can't see the first one fail. This usage has also bled into visitor tests, and two device model object property getters rtc_get_date() and balloon_stats_get_all(). With the "accumulate" technique, you need fewer error checks in callers, and buy that with an error check in every callee. Can be nice. However, mixing the two techniques is confusing. You can't use the "accumulate" technique with functions designed for the "check separately" technique. You can use the "check separately" technique with functions designed for the "accumulate" technique, but then error_set() can't catch you setting an error more than once. Standardize on the "check separately" technique for now, because it's overwhelmingly prevalent. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2014-05-07 11:53:54 +04:00
if (err) {
goto out_end;
}
visit_start_struct(v, NULL, NULL, "stats", 0, &err);
if (err) {
goto out_end;
}
qapi: Replace uncommon use of the error API by the common one We commonly use the error API like this: err = NULL; foo(..., &err); if (err) { goto out; } bar(..., &err); Every error source is checked separately. The second function is only called when the first one succeeds. Both functions are free to pass their argument to error_set(). Because error_set() asserts no error has been set, this effectively means they must not be called with an error set. The qapi-generated code uses the error API differently: // *errp was initialized to NULL somewhere up the call chain frob(..., errp); gnat(..., errp); Errors accumulate in *errp: first error wins, subsequent errors get dropped. To make this work, the second function does nothing when called with an error set. Requires non-null errp, or else the second function can't see the first one fail. This usage has also bled into visitor tests, and two device model object property getters rtc_get_date() and balloon_stats_get_all(). With the "accumulate" technique, you need fewer error checks in callers, and buy that with an error check in every callee. Can be nice. However, mixing the two techniques is confusing. You can't use the "accumulate" technique with functions designed for the "check separately" technique. You can use the "check separately" technique with functions designed for the "accumulate" technique, but then error_set() can't catch you setting an error more than once. Standardize on the "check separately" technique for now, because it's overwhelmingly prevalent. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2014-05-07 11:53:54 +04:00
for (i = 0; !err && i < VIRTIO_BALLOON_S_NR; i++) {
visit_type_int64(v, (int64_t *) &s->stats[i], balloon_stat_names[i],
&err);
}
qapi: Replace uncommon use of the error API by the common one We commonly use the error API like this: err = NULL; foo(..., &err); if (err) { goto out; } bar(..., &err); Every error source is checked separately. The second function is only called when the first one succeeds. Both functions are free to pass their argument to error_set(). Because error_set() asserts no error has been set, this effectively means they must not be called with an error set. The qapi-generated code uses the error API differently: // *errp was initialized to NULL somewhere up the call chain frob(..., errp); gnat(..., errp); Errors accumulate in *errp: first error wins, subsequent errors get dropped. To make this work, the second function does nothing when called with an error set. Requires non-null errp, or else the second function can't see the first one fail. This usage has also bled into visitor tests, and two device model object property getters rtc_get_date() and balloon_stats_get_all(). With the "accumulate" technique, you need fewer error checks in callers, and buy that with an error check in every callee. Can be nice. However, mixing the two techniques is confusing. You can't use the "accumulate" technique with functions designed for the "check separately" technique. You can use the "check separately" technique with functions designed for the "accumulate" technique, but then error_set() can't catch you setting an error more than once. Standardize on the "check separately" technique for now, because it's overwhelmingly prevalent. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2014-05-07 11:53:54 +04:00
error_propagate(errp, err);
err = NULL;
visit_end_struct(v, &err);
out_end:
qapi: Replace uncommon use of the error API by the common one We commonly use the error API like this: err = NULL; foo(..., &err); if (err) { goto out; } bar(..., &err); Every error source is checked separately. The second function is only called when the first one succeeds. Both functions are free to pass their argument to error_set(). Because error_set() asserts no error has been set, this effectively means they must not be called with an error set. The qapi-generated code uses the error API differently: // *errp was initialized to NULL somewhere up the call chain frob(..., errp); gnat(..., errp); Errors accumulate in *errp: first error wins, subsequent errors get dropped. To make this work, the second function does nothing when called with an error set. Requires non-null errp, or else the second function can't see the first one fail. This usage has also bled into visitor tests, and two device model object property getters rtc_get_date() and balloon_stats_get_all(). With the "accumulate" technique, you need fewer error checks in callers, and buy that with an error check in every callee. Can be nice. However, mixing the two techniques is confusing. You can't use the "accumulate" technique with functions designed for the "check separately" technique. You can use the "check separately" technique with functions designed for the "accumulate" technique, but then error_set() can't catch you setting an error more than once. Standardize on the "check separately" technique for now, because it's overwhelmingly prevalent. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2014-05-07 11:53:54 +04:00
error_propagate(errp, err);
err = NULL;
visit_end_struct(v, &err);
out:
error_propagate(errp, err);
}
static void balloon_stats_get_poll_interval(Object *obj, struct Visitor *v,
void *opaque, const char *name,
Error **errp)
{
VirtIOBalloon *s = opaque;
visit_type_int(v, &s->stats_poll_interval, name, errp);
}
static void balloon_stats_set_poll_interval(Object *obj, struct Visitor *v,
void *opaque, const char *name,
Error **errp)
{
VirtIOBalloon *s = opaque;
Error *local_err = NULL;
int64_t value;
visit_type_int(v, &value, name, &local_err);
if (local_err) {
error_propagate(errp, local_err);
return;
}
if (value < 0) {
error_setg(errp, "timer value must be greater than zero");
return;
}
if (value > UINT32_MAX) {
error_setg(errp, "timer value is too big");
return;
}
if (value == s->stats_poll_interval) {
return;
}
if (value == 0) {
/* timer=0 disables the timer */
balloon_stats_destroy_timer(s);
return;
}
if (balloon_stats_enabled(s)) {
/* timer interval change */
s->stats_poll_interval = value;
balloon_stats_change_timer(s, value);
return;
}
/* create a new timer */
g_assert(s->stats_timer == NULL);
s->stats_timer = timer_new_ms(QEMU_CLOCK_VIRTUAL, balloon_stats_poll_cb, s);
s->stats_poll_interval = value;
balloon_stats_change_timer(s, 0);
}
static void virtio_balloon_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
VirtIOBalloon *s = VIRTIO_BALLOON(vdev);
VirtQueueElement elem;
MemoryRegionSection section;
while (virtqueue_pop(vq, &elem)) {
size_t offset = 0;
uint32_t pfn;
while (iov_to_buf(elem.out_sg, elem.out_num, offset, &pfn, 4) == 4) {
ram_addr_t pa;
ram_addr_t addr;
int p = virtio_ldl_p(vdev, &pfn);
pa = (ram_addr_t) p << VIRTIO_BALLOON_PFN_SHIFT;
offset += 4;
/* FIXME: remove get_system_memory(), but how? */
section = memory_region_find(get_system_memory(), pa, 1);
if (!int128_nz(section.size) || !memory_region_is_ram(section.mr))
continue;
trace_virtio_balloon_handle_output(memory_region_name(section.mr),
pa);
/* Using memory_region_get_ram_ptr is bending the rules a bit, but
should be OK because we only want a single page. */
addr = section.offset_within_region;
balloon_page(memory_region_get_ram_ptr(section.mr) + addr,
!!(vq == s->dvq));
memory_region_unref(section.mr);
}
virtqueue_push(vq, &elem, offset);
virtio_notify(vdev, vq);
}
}
static void virtio_balloon_receive_stats(VirtIODevice *vdev, VirtQueue *vq)
{
VirtIOBalloon *s = VIRTIO_BALLOON(vdev);
VirtQueueElement *elem = &s->stats_vq_elem;
VirtIOBalloonStat stat;
size_t offset = 0;
qemu_timeval tv;
if (!virtqueue_pop(vq, elem)) {
goto out;
}
/* Initialize the stats to get rid of any stale values. This is only
* needed to handle the case where a guest supports fewer stats than it
* used to (ie. it has booted into an old kernel).
*/
reset_stats(s);
while (iov_to_buf(elem->out_sg, elem->out_num, offset, &stat, sizeof(stat))
== sizeof(stat)) {
uint16_t tag = virtio_tswap16(vdev, stat.tag);
uint64_t val = virtio_tswap64(vdev, stat.val);
offset += sizeof(stat);
if (tag < VIRTIO_BALLOON_S_NR)
s->stats[tag] = val;
}
s->stats_vq_offset = offset;
if (qemu_gettimeofday(&tv) < 0) {
fprintf(stderr, "warning: %s: failed to get time of day\n", __func__);
goto out;
}
s->stats_last_update = tv.tv_sec;
out:
if (balloon_stats_enabled(s)) {
balloon_stats_change_timer(s, s->stats_poll_interval);
}
}
static void virtio_balloon_get_config(VirtIODevice *vdev, uint8_t *config_data)
{
VirtIOBalloon *dev = VIRTIO_BALLOON(vdev);
struct virtio_balloon_config config;
config.num_pages = cpu_to_le32(dev->num_pages);
config.actual = cpu_to_le32(dev->actual);
trace_virtio_balloon_get_config(config.num_pages, config.actual);
memcpy(config_data, &config, sizeof(struct virtio_balloon_config));
}
static void virtio_balloon_set_config(VirtIODevice *vdev,
const uint8_t *config_data)
{
VirtIOBalloon *dev = VIRTIO_BALLOON(vdev);
struct virtio_balloon_config config;
uint32_t oldactual = dev->actual;
ram_addr_t vm_ram_size = get_current_ram_size();
memcpy(&config, config_data, sizeof(struct virtio_balloon_config));
dev->actual = le32_to_cpu(config.actual);
if (dev->actual != oldactual) {
qapi_event_send_balloon_change(vm_ram_size -
((ram_addr_t) dev->actual << VIRTIO_BALLOON_PFN_SHIFT),
&error_abort);
}
trace_virtio_balloon_set_config(dev->actual, oldactual);
}
static uint64_t virtio_balloon_get_features(VirtIODevice *vdev, uint64_t f,
Error **errp)
{
balloon: add a feature bit to let Guest OS deflate balloon on oom Excessive virtio_balloon inflation can cause invocation of OOM-killer, when Linux is under severe memory pressure. Various mechanisms are responsible for correct virtio_balloon memory management. Nevertheless it is often the case that these control tools does not have enough time to react on fast changing memory load. As a result OS runs out of memory and invokes OOM-killer. The balancing of memory by use of the virtio balloon should not cause the termination of processes while there are pages in the balloon. Now there is no way for virtio balloon driver to free memory at the last moment before some process get killed by OOM-killer. This does not provide a security breach as balloon itself is running inside Guest OS and is working in the cooperation with the host. Thus some improvements from Guest side should be considered as normal. To solve the problem, introduce a virtio_balloon callback which is expected to be called from the oom notifier call chain in out_of_memory() function. If virtio balloon could release some memory, it will make the system return and retry the allocation that forced the out of memory killer to run. This behavior should be enabled if and only if appropriate feature bit is set on the device. It is off by default. This functionality was recently merged into vanilla Linux. commit 5a10b7dbf904bfe01bb9fcc6298f7df09eed77d5 Author: Raushaniya Maksudova <rmaksudova@parallels.com> Date: Mon Nov 10 09:36:29 2014 +1030 This patch adds respective control bits into QEMU. It introduces deflate-on-oom option for balloon device which does the trick. Signed-off-by: Denis V. Lunev <den@openvz.org> CC: Raushaniya Maksudova <rmaksudova@parallels.com> CC: Anthony Liguori <aliguori@amazon.com> CC: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: James Bottomley <JBottomley@Odin.com> Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
2015-06-15 13:52:52 +03:00
VirtIOBalloon *dev = VIRTIO_BALLOON(vdev);
f |= dev->host_features;
virtio_add_feature(&f, VIRTIO_BALLOON_F_STATS_VQ);
return f;
}
static void virtio_balloon_stat(void *opaque, BalloonInfo *info)
{
VirtIOBalloon *dev = opaque;
info->actual = get_current_ram_size() - ((uint64_t) dev->actual <<
VIRTIO_BALLOON_PFN_SHIFT);
}
balloon: Separate out stat and balloon handling Passing on '0' as ballooning target to indicate retrieval of stats is bad API. It also makes 'balloon 0' in the monitor cause a segfault. Have two different functions handle the different functionality instead. Detailed explanation from Markus's review: 1. do_info_balloon() is an info_async() method. It receives a callback with argument, to be called exactly once (callback frees the argument). It passes the callback via qemu_balloon_status() and indirectly through qemu_balloon_event to virtio_balloon_to_target(). virtio_balloon_to_target() executes its balloon stats half. It stores the callback in the device state. If it can't send a stats request, it resets stats and calls the callback right away. Else, it sends a stats request. The device model runs the callback when it receives the answer. Works. 2. do_balloon() is a cmd_async() method. It receives a callback with argument, to be called when the command completes. do_balloon() calls it right before it succeeds. Odd, but should work. Nevertheless, it passes the callback on via qemu_ballon() and indirectly through qemu_balloon_event to virtio_balloon_to_target(). a. If the argument is non-zero, virtio_balloon_to_target() executes its balloon half, which doesn't use the callback in any way. Odd, but works. b. If the argument is zero, virtio_balloon_to_target() executes its balloon stats half, just like in 1. It either calls the callback right away, or arranges for it to be called later. Thus, the callback runs twice: use after free and double free. Test case: start with -S -device virtio-balloon, execute "balloon 0" in human monitor. Runs the callback first from virtio_balloon_to_target(), then again from do_balloon(). Reported-by: Mike Cao <bcao@redhat.com> Signed-off-by: Amit Shah <amit.shah@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com>
2011-07-20 12:00:56 +04:00
static void virtio_balloon_to_target(void *opaque, ram_addr_t target)
{
VirtIOBalloon *dev = VIRTIO_BALLOON(opaque);
VirtIODevice *vdev = VIRTIO_DEVICE(dev);
ram_addr_t vm_ram_size = get_current_ram_size();
if (target > vm_ram_size) {
target = vm_ram_size;
}
if (target) {
dev->num_pages = (vm_ram_size - target) >> VIRTIO_BALLOON_PFN_SHIFT;
virtio_notify_config(vdev);
}
trace_virtio_balloon_to_target(target, dev->num_pages);
}
static void virtio_balloon_save(QEMUFile *f, void *opaque)
{
virtio_save(VIRTIO_DEVICE(opaque), f);
}
static void virtio_balloon_save_device(VirtIODevice *vdev, QEMUFile *f)
{
VirtIOBalloon *s = VIRTIO_BALLOON(vdev);
qemu_put_be32(f, s->num_pages);
qemu_put_be32(f, s->actual);
}
static int virtio_balloon_load(QEMUFile *f, void *opaque, int version_id)
{
if (version_id != 1)
return -EINVAL;
return virtio_load(VIRTIO_DEVICE(opaque), f, version_id);
}
static int virtio_balloon_load_device(VirtIODevice *vdev, QEMUFile *f,
int version_id)
{
VirtIOBalloon *s = VIRTIO_BALLOON(vdev);
s->num_pages = qemu_get_be32(f);
s->actual = qemu_get_be32(f);
return 0;
}
static void virtio_balloon_device_realize(DeviceState *dev, Error **errp)
{
VirtIODevice *vdev = VIRTIO_DEVICE(dev);
VirtIOBalloon *s = VIRTIO_BALLOON(dev);
int ret;
virtio_init(vdev, "virtio-balloon", VIRTIO_ID_BALLOON,
sizeof(struct virtio_balloon_config));
ret = qemu_add_balloon_handler(virtio_balloon_to_target,
virtio_balloon_stat, s);
if (ret < 0) {
2015-03-31 20:00:26 +03:00
error_setg(errp, "Only one balloon device is supported");
virtio_cleanup(vdev);
return;
}
s->ivq = virtio_add_queue(vdev, 128, virtio_balloon_handle_output);
s->dvq = virtio_add_queue(vdev, 128, virtio_balloon_handle_output);
s->svq = virtio_add_queue(vdev, 128, virtio_balloon_receive_stats);
reset_stats(s);
register_savevm(dev, "virtio-balloon", -1, 1,
virtio_balloon_save, virtio_balloon_load, s);
}
static void virtio_balloon_device_unrealize(DeviceState *dev, Error **errp)
{
VirtIODevice *vdev = VIRTIO_DEVICE(dev);
VirtIOBalloon *s = VIRTIO_BALLOON(dev);
balloon_stats_destroy_timer(s);
qemu_remove_balloon_handler(s);
unregister_savevm(dev, "virtio-balloon", s);
virtio_cleanup(vdev);
}
static void virtio_balloon_instance_init(Object *obj)
{
VirtIOBalloon *s = VIRTIO_BALLOON(obj);
object_property_add(obj, "guest-stats", "guest statistics",
balloon_stats_get_all, NULL, NULL, s, NULL);
object_property_add(obj, "guest-stats-polling-interval", "int",
balloon_stats_get_poll_interval,
balloon_stats_set_poll_interval,
NULL, s, NULL);
}
static Property virtio_balloon_properties[] = {
balloon: add a feature bit to let Guest OS deflate balloon on oom Excessive virtio_balloon inflation can cause invocation of OOM-killer, when Linux is under severe memory pressure. Various mechanisms are responsible for correct virtio_balloon memory management. Nevertheless it is often the case that these control tools does not have enough time to react on fast changing memory load. As a result OS runs out of memory and invokes OOM-killer. The balancing of memory by use of the virtio balloon should not cause the termination of processes while there are pages in the balloon. Now there is no way for virtio balloon driver to free memory at the last moment before some process get killed by OOM-killer. This does not provide a security breach as balloon itself is running inside Guest OS and is working in the cooperation with the host. Thus some improvements from Guest side should be considered as normal. To solve the problem, introduce a virtio_balloon callback which is expected to be called from the oom notifier call chain in out_of_memory() function. If virtio balloon could release some memory, it will make the system return and retry the allocation that forced the out of memory killer to run. This behavior should be enabled if and only if appropriate feature bit is set on the device. It is off by default. This functionality was recently merged into vanilla Linux. commit 5a10b7dbf904bfe01bb9fcc6298f7df09eed77d5 Author: Raushaniya Maksudova <rmaksudova@parallels.com> Date: Mon Nov 10 09:36:29 2014 +1030 This patch adds respective control bits into QEMU. It introduces deflate-on-oom option for balloon device which does the trick. Signed-off-by: Denis V. Lunev <den@openvz.org> CC: Raushaniya Maksudova <rmaksudova@parallels.com> CC: Anthony Liguori <aliguori@amazon.com> CC: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: James Bottomley <JBottomley@Odin.com> Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com>
2015-06-15 13:52:52 +03:00
DEFINE_PROP_BIT("deflate-on-oom", VirtIOBalloon, host_features,
VIRTIO_BALLOON_F_DEFLATE_ON_OOM, false),
DEFINE_PROP_END_OF_LIST(),
};
static void virtio_balloon_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
VirtioDeviceClass *vdc = VIRTIO_DEVICE_CLASS(klass);
dc->props = virtio_balloon_properties;
set_bit(DEVICE_CATEGORY_MISC, dc->categories);
vdc->realize = virtio_balloon_device_realize;
vdc->unrealize = virtio_balloon_device_unrealize;
vdc->get_config = virtio_balloon_get_config;
vdc->set_config = virtio_balloon_set_config;
vdc->get_features = virtio_balloon_get_features;
vdc->save = virtio_balloon_save_device;
vdc->load = virtio_balloon_load_device;
}
static const TypeInfo virtio_balloon_info = {
.name = TYPE_VIRTIO_BALLOON,
.parent = TYPE_VIRTIO_DEVICE,
.instance_size = sizeof(VirtIOBalloon),
.instance_init = virtio_balloon_instance_init,
.class_init = virtio_balloon_class_init,
};
static void virtio_register_types(void)
{
type_register_static(&virtio_balloon_info);
}
type_init(virtio_register_types)