This patch series enhances the existing virtio-balloon with the following
new features:
1) fast ballooning: transfer ballooned pages between the guest and host in
chunks using sgs, instead of one array each time; and
2) free page block reporting: a new virtqueue to report guest free pages
to the host.
The second feature can be used to accelerate live migration of VMs. Here
are some details:
Live migration needs to transfer the VM's memory from the source machine
to the destination round by round. For the 1st round, all the VM's memory
is transferred. From the 2nd round, only the pieces of memory that were
written by the guest (after the 1st round) are transferred. One method
that is popularly used by the hypervisor to track which part of memory is
written is to write-protect all the guest memory.
The second feature enables the optimization of the 1st round memory
transfer - the hypervisor can skip the transfer of guest free pages in the
1st round. It is not concerned that the memory pages are used after they
are given to the hypervisor as a hint of the free pages, because they will
be tracked by the hypervisor and transferred in the next round if they are
used and written.
ChangeLog:
v20 (RESEND)->v21:
1) patch 1:
- xb_zero: fix the bug related to bitmap_clear;
- testing/radix-tree/Makefile: removed the "$CC.." line;
- added a test case for xbitmap_check_zero_bits;
2) patch 2: virtio-balloon
- changed the usage of the xb_ related APIs
v19->v20:
1) patch 1: xbitmap
- add __rcu to "void **slot";
- remove the exceptional path.
2) patch 3: xbitmap
- DeveloperNotes: add an item to comment that the current bit range
related APIs operating on extremely large ranges (e.g.
[0, ULONG_MAX)) will take too long time. This can be optimized in
the future.
- remove the exceptional path;
- remove xb_preload_and_set();
- reimplement xb_clear_bit_range to make its usage close to
bitmap_clear;
- rename xb_find_next_set_bit to xb_find_set, and re-implement it
in a style close to find_next_bit;
- rename xb_find_next_zero_bit to xb_find_clear, and re-implement
it in a stytle close to find_next_zero_bit;
- separate the implementation of xb_find_set and xb_find_clear for
the convenience of future updates.
3) patch 4: virtio-balloon
- xb_set_page: change the way to call xb_ related APIs
v18->v19:
1) patch 3:
- xb_clear_bit_range and xb_find_next_bit will deal with range [start,
end), where end is changed to be exclusive of the range.
- add overflow checks at the end of xb_clear_bit_range and
xb_find_next_bit
- add overflow related test cases
2) patch 4:
- change back to the previous add_one_sg methond, which is based on the
scatterlist struct
- tell_host_sgs: use "uint64_t len" to avoid overflow
- batch_balloon_page_sg: a simpler function to implement the batching of
sgs
3) patch 6: batch_free_page_sg: batch sgs using the previous scatterlist struct
4) patch 7: add a config field, poison_val, to tell the host about the poison
value
v17->v18:
1) patch 1-2: new to solve some tools related compilation issues
2) patch 3: revert to the original xbitmap implementation from Matthew
Wilcox with some minor changes (e.g. comments added to the exported
functions)
3) patch 4: summarize the changes we want to make to patch 3
4) patch 5: add the developer notes as a reminder for users to avoid
concurrent accesses to the ida bitmap
5) patch 6: a new vring API to allow users to directly pass in a physical
address to a vring desc
6) patch 7: ballooning time is reduced from ~490ms to ~440ms with the new
implementation
- use the new API from patch 6 to send balloon pages
- xb_preload with "GFP_NOWAIT | __GFP_NOWARN" flag;
- handle the case when xb_set_page() fails to avoid memory leak;
- put xb_set_page() under the balloon lock
7) patch 9: simper implementation
- start free page reporting by sending a new cmd id from the host
- guest acks the start or stop via adding a cmd id to the free page vq
- use vb->report_free_page, instead of vb->report_free_page_stop
- use WRITE_ONCE/READ_ONCE to access vb->report_free_page
- use the new API from patch 6 to send free pages to avoid the
unnecessary use of kaddr.
8) patch 10: new patch to solve the page posioning issue reported by
Michael S. Tsirkin
v16->v17:
1) patch 1: please check the commit log there;
2) patch 3: included Michael S. Tsirkin patch to fix the potential
deadlock issue;
3) patch 4: use BUG_ON if virtqueue_add_ returns error, which is
expected never to happen;
4) patch 4: add leak_balloon_sg_oom, which is used in the oom case when
VIRTIO_BALLOON_F_SG is in use;
5) patch 6: use config registers, instead of a vq, as the command channel
between the host and guest;
6) patch 6: add the command sequence id support.
v15->v16:
1) mm: stop reporting the free pfn range if the callback returns false;
2) mm: move some implementaion of walk_free_mem_block into a function to
make the code layout looks better;
3) xbitmap: added some optimizations suggested by Matthew, please refer to
the ChangLog in the xbitmap patch for details.
4) xbitmap: added a test suite
5) virtio-balloon: bail out with a warning when virtqueue_add_inbuf returns
an error
6) virtio-balloon: some small code re-arrangement, e.g. detachinf used buf
from the vq before adding a new buf
v14->v15:
1) mm: make the report callback return a bool value - returning 1 to stop
walking through the free page list.
2) virtio-balloon: batching sgs of balloon pages till the vq is full
3) virtio-balloon: create a new workqueue, rather than using the default
system_wq, to queue the free page reporting work item.
4) virtio-balloon: add a ctrl_vq to be a central control plane which will
handle all the future control related commands between the host and guest.
Add free page report as the first feature controlled under ctrl_vq, and
the free_page_vq is a data plane vq dedicated to the transmission of free
page blocks.
v13->v14:
1) xbitmap: move the code from lib/radix-tree.c to lib/xbitmap.c.
2) xbitmap: consolidate the implementation of xb_bit_set/clear/test into
one xb_bit_ops.
3) xbitmap: add documents for the exported APIs.
4) mm: rewrite the function to walk through free page blocks.
5) virtio-balloon: when reporting a free page blcok to the device, if the
vq is full (less likey to happen in practice), just skip reporting this
block, instead of busywaiting till an entry gets released.
6) virtio-balloon: fail the probe function if adding the signal buf in
init_vqs fails.
v12->v13:
1) mm: use a callback function to handle the the free page blocks from the
report function. This avoids exposing the zone internal to a kernel
module.
2) virtio-balloon: send balloon pages or a free page block using a single
sg each time. This has the benefits of simpler implementation with no new
APIs.
3) virtio-balloon: the free_page_vq is used to report free pages only (no
multiple usages interleaving)
4) virtio-balloon: Balloon pages and free page blocks are sent via input
sgs, and the completion signal to the host is sent via an output sg.
v11->v12:
1) xbitmap: use the xbitmap from Matthew Wilcox to record ballooned pages.
2) virtio-ring: enable the driver to build up a desc chain using vring
desc.
3) virtio-ring: Add locking to the existing START_USE() and END_USE()
macro to lock/unlock the vq when a vq operation starts/ends.
4) virtio-ring: add virtqueue_kick_sync() and virtqueue_kick_async()
5) virtio-balloon: describe chunks of ballooned pages and free pages
blocks directly using one or more chains of desc from the vq.
v10->v11:
1) virtio_balloon: use vring_desc to describe a chunk;
2) virtio_ring: support to add an indirect desc table to virtqueue;
3) virtio_balloon: use cmdq to report guest memory statistics.
v9->v10:
1) mm: put report_unused_page_block() under CONFIG_VIRTIO_BALLOON;
2) virtio-balloon: add virtballoon_validate();
3) virtio-balloon: msg format change;
4) virtio-balloon: move miscq handling to a task on system_freezable_wq;
5) virtio-balloon: code cleanup.
v8->v9:
1) Split the two new features, VIRTIO_BALLOON_F_BALLOON_CHUNKS and
VIRTIO_BALLOON_F_MISC_VQ, which were mixed together in the previous
implementation;
2) Simpler function to get the free page block.
v7->v8:
1) Use only one chunk format, instead of two.
2) re-write the virtio-balloon implementation patch.
3) commit changes
4) patch re-org
Matthew Wilcox (1):
xbitmap: Introduce xbitmap
Wei Wang (4):
virtio-balloon: VIRTIO_BALLOON_F_SG
mm: support reporting free page blocks
virtio-balloon: VIRTIO_BALLOON_F_FREE_PAGE_VQ
virtio-balloon: don't report free pages when page poisoning is enabled
drivers/virtio/virtio_balloon.c | 443 ++++++++++++++++++++++++++----
include/linux/mm.h | 6 +
include/linux/xbitmap.h | 48 ++++
include/uapi/linux/virtio_balloon.h | 7 +
lib/Makefile | 2 +-
lib/radix-tree.c | 38 ++-
lib/xbitmap.c | 444 +++++++++++++++++++++++++++++++
mm/page_alloc.c | 91 +++++++
tools/include/linux/bitmap.h | 34 +++
tools/include/linux/kernel.h | 2 +
tools/testing/radix-tree/Makefile | 17 +-
tools/testing/radix-tree/linux/kernel.h | 2 -
tools/testing/radix-tree/linux/xbitmap.h | 1 +
tools/testing/radix-tree/main.c | 4 +
tools/testing/radix-tree/test.h | 1 +
15 files changed, 1080 insertions(+), 60 deletions(-)
create mode 100644 include/linux/xbitmap.h
create mode 100644 lib/xbitmap.c
create mode 100644 tools/testing/radix-tree/linux/xbitmap.h
--
2.7.4
From: Matthew Wilcox <mawilcox at microsoft.com>
The eXtensible Bitmap is a sparse bitmap representation which is
efficient for set bits which tend to cluster. It supports up to
'unsigned long' worth of bits.
Signed-off-by: Matthew Wilcox <mawilcox at microsoft.com>
Signed-off-by: Wei Wang <wei.w.wang at intel.com>
Cc: Andrew Morton <akpm at linux-foundation.org>
Cc: Michal Hocko <mhocko at kernel.org>
Cc: Michael S. Tsirkin <mst at redhat.com>
Cc: Tetsuo Handa <penguin-kernel at I-love.SAKURA.ne.jp>
---
include/linux/xbitmap.h | 48 ++++
lib/Makefile | 2 +-
lib/radix-tree.c | 38 ++-
lib/xbitmap.c | 444 +++++++++++++++++++++++++++++++
tools/include/linux/bitmap.h | 34 +++
tools/include/linux/kernel.h | 2 +
tools/testing/radix-tree/Makefile | 17 +-
tools/testing/radix-tree/linux/kernel.h | 2 -
tools/testing/radix-tree/linux/xbitmap.h | 1 +
tools/testing/radix-tree/main.c | 4 +
tools/testing/radix-tree/test.h | 1 +
11 files changed, 583 insertions(+), 10 deletions(-)
create mode 100644 include/linux/xbitmap.h
create mode 100644 lib/xbitmap.c
create mode 100644 tools/testing/radix-tree/linux/xbitmap.h
diff --git a/include/linux/xbitmap.h b/include/linux/xbitmap.h
new file mode 100644
index 0000000..c008309
--- /dev/null
+++ b/include/linux/xbitmap.h
@@ -0,0 +1,48 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * eXtensible Bitmaps
+ * Copyright (c) 2017 Microsoft Corporation
+ * Author: Matthew Wilcox <mawilcox at microsoft.com>
+ *
+ * eXtensible Bitmaps provide an unlimited-size sparse bitmap facility.
+ * All bits are initially zero.
+ *
+ * Locking is to be provided by the user. No xb_ function is safe to
+ * call concurrently with any other xb_ function.
+ */
+
+#include <linux/idr.h>
+
+struct xb {
+ struct radix_tree_root xbrt;
+};
+
+#define XB_INIT { \
+ .xbrt = RADIX_TREE_INIT(IDR_RT_MARKER | GFP_NOWAIT), \
+}
+#define DEFINE_XB(name) struct xb name = XB_INIT
+
+static inline void xb_init(struct xb *xb)
+{
+ INIT_RADIX_TREE(&xb->xbrt, IDR_RT_MARKER | GFP_NOWAIT);
+}
+
+int xb_set_bit(struct xb *xb, unsigned long bit);
+bool xb_test_bit(const struct xb *xb, unsigned long bit);
+void xb_clear_bit(struct xb *xb, unsigned long bit);
+void xb_zero(struct xb *xb, unsigned long min, unsigned long max);
+void xb_fill(struct xb *xb, unsigned long min, unsigned long max);
+bool xb_find_set(const struct xb *xb, unsigned long max, unsigned long *bit);
+bool xb_find_zero(const struct xb *xb, unsigned long max, unsigned long *bit);
+
+static inline bool xb_empty(const struct xb *xb)
+{
+ return radix_tree_empty(&xb->xbrt);
+}
+
+int __must_check xb_preload(gfp_t);
+
+static inline void xb_preload_end(void)
+{
+ preempt_enable();
+}
diff --git a/lib/Makefile b/lib/Makefile
index d11c48e..08a8183 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -19,7 +19,7 @@ KCOV_INSTRUMENT_dynamic_debug.o := n
lib-y := ctype.o string.o vsprintf.o cmdline.o \
rbtree.o radix-tree.o dump_stack.o timerqueue.o\
- idr.o int_sqrt.o extable.o \
+ idr.o xbitmap.o int_sqrt.o extable.o \
sha1.o chacha20.o irq_regs.o argv_split.o \
flex_proportions.o ratelimit.o show_mem.o \
is_single_threaded.o plist.o decompress.o kobject_uevent.o \
diff --git a/lib/radix-tree.c b/lib/radix-tree.c
index c8d5556..d2bd8fe 100644
--- a/lib/radix-tree.c
+++ b/lib/radix-tree.c
@@ -37,7 +37,7 @@
#include <linux/rcupdate.h>
#include <linux/slab.h>
#include <linux/string.h>
-
+#include <linux/xbitmap.h>
/* Number of nodes in fully populated tree of given height */
static unsigned long height_to_maxnodes[RADIX_TREE_MAX_PATH + 1] __read_mostly;
@@ -77,6 +77,11 @@ static struct kmem_cache *radix_tree_node_cachep;
RADIX_TREE_MAP_SHIFT))
#define IDA_PRELOAD_SIZE (IDA_MAX_PATH * 2 - 1)
+#define XB_INDEX_BITS (BITS_PER_LONG - ilog2(IDA_BITMAP_BITS))
+#define XB_MAX_PATH (DIV_ROUND_UP(XB_INDEX_BITS, \
+ RADIX_TREE_MAP_SHIFT))
+#define XB_PRELOAD_SIZE (XB_MAX_PATH * 2 - 1)
+
/*
* Per-cpu pool of preloaded nodes
*/
@@ -1781,7 +1786,7 @@ void __rcu **radix_tree_next_chunk(const struct
radix_tree_root *root,
child = rcu_dereference_raw(node->slots[offset]);
}
- if (!child)
+ if (!child && !is_idr(root))
goto restart;
if (child == RADIX_TREE_RETRY)
break;
@@ -2135,6 +2140,35 @@ int ida_pre_get(struct ida *ida, gfp_t gfp)
}
EXPORT_SYMBOL(ida_pre_get);
+/**
+ * xb_preload - preload for xb_set_bit()
+ * @gfp_mask: allocation mask to use for preloading
+ *
+ * Preallocate memory to use for the next call to xb_set_bit(). On success,
+ * return zero, with preemption disabled. On error, return -ENOMEM with
+ * preemption not disabled.
+ */
+int xb_preload(gfp_t gfp)
+{
+ if (!this_cpu_read(ida_bitmap)) {
+ struct ida_bitmap *bitmap = kmalloc(sizeof(*bitmap), gfp);
+
+ if (!bitmap)
+ return -ENOMEM;
+ /*
+ * The per-CPU variable is updated with preemption enabled.
+ * If the calling task is unlucky to be scheduled to another
+ * CPU which has no ida_bitmap allocation, it will be detected
+ * when setting a bit (i.e. xb_set_bit()).
+ */
+ bitmap = this_cpu_cmpxchg(ida_bitmap, NULL, bitmap);
+ kfree(bitmap);
+ }
+
+ return __radix_tree_preload(gfp, XB_PRELOAD_SIZE);
+}
+EXPORT_SYMBOL(xb_preload);
+
void __rcu **idr_get_free_cmn(struct radix_tree_root *root,
struct radix_tree_iter *iter, gfp_t gfp,
unsigned long max)
diff --git a/lib/xbitmap.c b/lib/xbitmap.c
new file mode 100644
index 0000000..62b2211
--- /dev/null
+++ b/lib/xbitmap.c
@@ -0,0 +1,444 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * XBitmap implementation
+ * Copyright (c) 2017 Microsoft Corporation
+ * Author: Matthew Wilcox <mawilcox at microsoft.com>
+ */
+
+#include <linux/bitmap.h>
+#include <linux/export.h>
+#include <linux/slab.h>
+#include <linux/xbitmap.h>
+
+/**
+ * xb_set_bit() - Set a bit in the XBitmap.
+ * @xb: The XBitmap.
+ * @bit: Index of the bit to set.
+ *
+ * This function is used to set a bit in the xbitmap.
+ *
+ * Return: 0 on success. -ENOMEM if memory could not be allocated.
+ */
+int xb_set_bit(struct xb *xb, unsigned long bit)
+{
+ unsigned long index = bit / IDA_BITMAP_BITS;
+ struct radix_tree_root *root = &xb->xbrt;
+ struct radix_tree_iter iter;
+ void __rcu **slot;
+ struct ida_bitmap *bitmap;
+
+ bit %= IDA_BITMAP_BITS;
+ radix_tree_iter_init(&iter, index);
+ slot = idr_get_free_cmn(root, &iter, GFP_NOWAIT | __GFP_NOWARN, index);
+ if (IS_ERR(slot)) {
+ if (slot == ERR_PTR(-ENOSPC))
+ return 0; /* Already set */
+ return -ENOMEM;
+ }
+ bitmap = rcu_dereference_raw(*slot);
+ if (!bitmap) {
+ bitmap = this_cpu_xchg(ida_bitmap, NULL);
+ if (!bitmap)
+ return -ENOMEM;
+ memset(bitmap, 0, sizeof(*bitmap));
+ radix_tree_iter_replace(root, &iter, slot, bitmap);
+ }
+
+ __set_bit(bit, bitmap->bitmap);
+ if (bitmap_full(bitmap->bitmap, IDA_BITMAP_BITS))
+ radix_tree_iter_tag_clear(root, &iter, IDR_FREE);
+ return 0;
+}
+EXPORT_SYMBOL(xb_set_bit);
+
+/**
+ * xb_clear_bit() - Clear a bit in the XBitmap.
+ * @xb: The XBitmap.
+ * @bit: Index of the bit to clear.
+ *
+ * This function is used to clear a bit in the xbitmap.
+ */
+void xb_clear_bit(struct xb *xb, unsigned long bit)
+{
+ unsigned long index = bit / IDA_BITMAP_BITS;
+ struct radix_tree_root *root = &xb->xbrt;
+ struct radix_tree_iter iter;
+ void __rcu **slot;
+ struct ida_bitmap *bitmap;
+
+ bit %= IDA_BITMAP_BITS;
+ slot = radix_tree_iter_lookup(root, &iter, index);
+ if (!slot)
+ return;
+ bitmap = radix_tree_deref_slot(slot);
+ if (!bitmap)
+ return;
+
+ radix_tree_iter_tag_set(root, &iter, IDR_FREE);
+ __clear_bit(bit, bitmap->bitmap);
+ if (bitmap_empty(bitmap->bitmap, IDA_BITMAP_BITS)) {
+ kfree(bitmap);
+ radix_tree_iter_delete(root, &iter, slot);
+ }
+}
+EXPORT_SYMBOL(xb_clear_bit);
+
+/**
+ * xb_zero() - Clear a range of bits in the XBitmap.
+ * @xb: The XBitmap.
+ * @min: The first bit to clear.
+ * @max: The last bit to clear.
+ *
+ * This function is used to clear a range of bits in the xbitmap.
+ */
+void xb_zero(struct xb *xb, unsigned long min, unsigned long max)
+{
+ struct radix_tree_root *root = &xb->xbrt;
+ struct radix_tree_iter iter;
+ void __rcu **slot;
+ struct ida_bitmap *bitmap;
+ unsigned long index = min / IDA_BITMAP_BITS;
+ unsigned long first = min % IDA_BITMAP_BITS;
+ unsigned long maxindex = max / IDA_BITMAP_BITS;
+
+ radix_tree_for_each_slot(slot, root, &iter, index) {
+ unsigned long nbits = IDA_BITMAP_BITS;
+
+ if (index > maxindex)
+ break;
+ bitmap = radix_tree_deref_slot(slot);
+ if (!bitmap)
+ continue;
+ radix_tree_iter_tag_set(root, &iter, IDR_FREE);
+
+ if (!first && iter.index < maxindex)
+ goto delete;
+ if (iter.index == maxindex)
+ nbits = max % IDA_BITMAP_BITS + 1;
+ bitmap_clear(bitmap->bitmap, first, nbits - first);
+ first = 0;
+ if (bitmap_empty(bitmap->bitmap, IDA_BITMAP_BITS))
+ goto delete;
+ continue;
+delete:
+ kfree(bitmap);
+ radix_tree_iter_delete(root, &iter, slot);
+ }
+}
+EXPORT_SYMBOL(xb_zero);
+
+/**
+ * xb_test_bit() - Test a bit in the xbitmap.
+ * @xb: The XBitmap.
+ * @bit: Index of the bit to test.
+ *
+ * This function is used to test a bit in the xbitmap.
+ *
+ * Return: %true if the bit is set.
+ */
+bool xb_test_bit(const struct xb *xb, unsigned long bit)
+{
+ unsigned long index = bit / IDA_BITMAP_BITS;
+ struct ida_bitmap *bitmap = radix_tree_lookup(&xb->xbrt, index);
+
+ bit %= IDA_BITMAP_BITS;
+
+ if (!bitmap)
+ return false;
+ return test_bit(bit, bitmap->bitmap);
+}
+EXPORT_SYMBOL(xb_test_bit);
+
+/**
+ * xb_find_set() - Find the next set bit in a range of bits.
+ * @xb: The XBitmap.
+ * @max: The maximum position to search.
+ * @bit: The first bit to examine, and on exit, the found bit.
+ *
+ * On entry, @bit points to the index of the first bit to search. On exit,
+ * if this function returns %true, @bit will be updated to the index of the
+ * first found bit. It will not be updated if this function returns %false.
+ *
+ * Return: %true if a set bit was found.
+ */
+bool xb_find_set(const struct xb *xb, unsigned long max, unsigned long *bit)
+{
+ struct radix_tree_iter iter;
+ void __rcu **slot;
+ struct ida_bitmap *bitmap;
+ unsigned long index = *bit / IDA_BITMAP_BITS;
+ unsigned int first = *bit % IDA_BITMAP_BITS;
+ unsigned long maxindex = max / IDA_BITMAP_BITS;
+
+ radix_tree_for_each_slot(slot, &xb->xbrt, &iter, index) {
+ if (iter.index > maxindex)
+ break;
+ bitmap = radix_tree_deref_slot(slot);
+ if (bitmap) {
+ unsigned int nbits = IDA_BITMAP_BITS;
+
+ if (iter.index == maxindex)
+ nbits = max % IDA_BITMAP_BITS + 1;
+ first = find_next_bit(bitmap->bitmap, nbits, first);
+ if (first != nbits) {
+ *bit = first + iter.index * IDA_BITMAP_BITS;
+ return true;
+ }
+ }
+ first = 0;
+ }
+
+ return false;
+}
+EXPORT_SYMBOL(xb_find_set);
+
+/**
+ * xb_find_zero() - Find the next zero bit in a range of bits
+ * @xb: The XBitmap.
+ * @max: The maximum index to search.
+ * @bit: Pointer to an index.
+ *
+ * On entry, @bit points to the index of the first bit to search. On exit,
+ * if this function returns %true, @bit will be updated to the index of the
+ * first found bit. It will not be updated if this function returns %false.
+ *
+ * Return: %true if a clear bit was found.
+ */
+bool xb_find_zero(const struct xb *xb, unsigned long max, unsigned long *bit)
+{
+ void __rcu **slot;
+ struct radix_tree_iter iter;
+ struct ida_bitmap *bitmap;
+ unsigned long index = *bit / IDA_BITMAP_BITS;
+ unsigned long first = *bit % IDA_BITMAP_BITS;
+ unsigned long maxindex = max / IDA_BITMAP_BITS;
+
+ radix_tree_for_each_tagged(slot, &xb->xbrt, &iter, index, IDR_FREE)
{
+ unsigned int nbits = IDA_BITMAP_BITS;
+
+ if (iter.index > maxindex)
+ return false;
+ bitmap = radix_tree_deref_slot(slot);
+ if (!bitmap)
+ break;
+ if (iter.index == maxindex)
+ nbits = max % IDA_BITMAP_BITS + 1;
+ first = find_next_zero_bit(bitmap->bitmap, nbits, first);
+ if (first != nbits)
+ break;
+ first = 0;
+ }
+
+ *bit = first + iter.index * IDA_BITMAP_BITS;
+ return true;
+}
+EXPORT_SYMBOL(xb_find_zero);
+
+#ifndef __KERNEL__
+
+static DEFINE_XB(xb1);
+
+static void xbitmap_check_bit(unsigned long bit)
+{
+ unsigned long nbit = 0;
+
+ xb_preload(GFP_KERNEL);
+ assert(!xb_test_bit(&xb1, bit));
+ assert(xb_set_bit(&xb1, bit) == 0);
+ assert(xb_test_bit(&xb1, bit));
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == bit);
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == bit);
+ nbit++;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == false);
+ assert(nbit == bit + 1);
+ xb_clear_bit(&xb1, bit);
+ assert(xb_empty(&xb1));
+ xb_clear_bit(&xb1, bit);
+ assert(xb_empty(&xb1));
+ nbit = 0;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == false);
+ assert(nbit == 0);
+ xb_preload_end();
+}
+
+/*
+ * In the following tests, preload is called once when all the bits to set
+ * locate in the same ida bitmap. Otherwise, it is recommended to call
+ * preload for each xb_set_bit.
+ */
+static void xbitmap_check_bit_range(void)
+{
+ unsigned long nbit = 0;
+
+ /* Regular test1: node = NULL */
+ xb_preload(GFP_KERNEL);
+ xb_set_bit(&xb1, 700);
+ xb_preload_end();
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == 700);
+ nbit++;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == false);
+ assert(nbit == 701);
+ xb_zero(&xb1, 0, 1023);
+
+ /*
+ * Regular test2
+ * set bit 2000, 2001, 2040
+ * Next 1 in [0, 2048] --> 2000
+ * Next 1 in [2000, 2002] --> 2000
+ * Next 1 in [2002, 2040] --> 2040
+ * Next 1 in [2002, 2039] --> none
+ * Next 0 in [2000, 2048] --> 2002
+ * Next 0 in [2048, 2060] --> 2048
+ */
+ xb_preload(GFP_KERNEL);
+ assert(!xb_set_bit(&xb1, 2000));
+ assert(!xb_set_bit(&xb1, 2001));
+ assert(!xb_set_bit(&xb1, 2040));
+ nbit = 0;
+ assert(xb_find_set(&xb1, 2048, &nbit) == true);
+ assert(nbit == 2000);
+ assert(xb_find_set(&xb1, 2002, &nbit) == true);
+ assert(nbit == 2000);
+ nbit = 2002;
+ assert(xb_find_set(&xb1, 2040, &nbit) == true);
+ assert(nbit == 2040);
+ nbit = 2002;
+ assert(xb_find_set(&xb1, 2039, &nbit) == false);
+ assert(nbit == 2002);
+ nbit = 2000;
+ assert(xb_find_zero(&xb1, 2048, &nbit) == true);
+ assert(nbit == 2002);
+ nbit = 2048;
+ assert(xb_find_zero(&xb1, 2060, &nbit) == true);
+ assert(nbit == 2048);
+ xb_zero(&xb1, 0, 2048);
+ nbit = 0;
+ assert(xb_find_set(&xb1, 2048, &nbit) == false);
+ assert(nbit == 0);
+ xb_preload_end();
+
+ /*
+ * Overflow tests:
+ * Set bit 1 and ULONG_MAX - 4
+ * Next 1 in [0, ULONG_MAX] --> 1
+ * Next 1 in [1, ULONG_MAX] --> 1
+ * Next 1 in [2, ULONG_MAX] --> ULONG_MAX - 4
+ * Next 1 in [ULONG_MAX - 3, 2] --> none
+ * Next 0 in [ULONG_MAX - 4, ULONG_MAX] --> ULONG_MAX - 3
+ * Zero [ULONG_MAX - 4, ULONG_MAX]
+ * Next 1 in [ULONG_MAX - 10, ULONG_MAX] --> none
+ * Next 1 in [ULONG_MAX - 1, 2] --> none
+ * Zero [0, 1]
+ * Next 1 in [0, 2] --> none
+ */
+ xb_preload(GFP_KERNEL);
+ assert(!xb_set_bit(&xb1, 1));
+ xb_preload_end();
+ xb_preload(GFP_KERNEL);
+ assert(!xb_set_bit(&xb1, ULONG_MAX - 4));
+ nbit = 0;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == 1);
+ nbit = 1;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == 1);
+ nbit = 2;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == ULONG_MAX - 4);
+ nbit++;
+ assert(xb_find_set(&xb1, 2, &nbit) == false);
+ assert(nbit == ULONG_MAX - 3);
+ nbit--;
+ assert(xb_find_zero(&xb1, ULONG_MAX, &nbit) == true);
+ assert(nbit == ULONG_MAX - 3);
+ xb_zero(&xb1, ULONG_MAX - 4, ULONG_MAX);
+ nbit = ULONG_MAX - 10;
+ assert(xb_find_set(&xb1, ULONG_MAX, &nbit) == false);
+ assert(nbit == ULONG_MAX - 10);
+ nbit = ULONG_MAX - 1;
+ assert(xb_find_set(&xb1, 2, &nbit) == false);
+ xb_zero(&xb1, 0, 1);
+ nbit = 0;
+ assert(xb_find_set(&xb1, 2, &nbit) == false);
+ assert(nbit == 0);
+ xb_preload_end();
+ assert(xb_empty(&xb1));
+}
+
+static void xbitmap_check_zero_bits(void)
+{
+ assert(xb_empty(&xb1));
+
+ /* Zero an empty xbitmap should work though no real work to do */
+ xb_zero(&xb1, 0, ULONG_MAX);
+ assert(xb_empty(&xb1));
+
+ xb_preload(GFP_KERNEL);
+ assert(xb_set_bit(&xb1, 0) == 0);
+ xb_preload_end();
+
+ /* Overflow test */
+ xb_zero(&xb1, ULONG_MAX - 10, ULONG_MAX);
+ assert(xb_test_bit(&xb1, 0));
+
+ xb_preload(GFP_KERNEL);
+ assert(xb_set_bit(&xb1, ULONG_MAX) == 0);
+ xb_preload_end();
+
+ xb_zero(&xb1, 0, ULONG_MAX);
+ assert(xb_empty(&xb1));
+}
+
+/* Check that setting an already-full bitmap works */
+static void xbitmap_check_set(unsigned long base)
+{
+ unsigned long i;
+
+ assert(xb_empty(&xb1));
+
+ for (i = 0; i < 64 * 1024; i++) {
+ xb_preload(GFP_KERNEL);
+ assert(xb_set_bit(&xb1, base + i) == 0);
+ xb_preload_end();
+ }
+ for (i = 0; i < 64 * 1024; i++)
+ assert(xb_set_bit(&xb1, base + i) == 0);
+
+ for (i = 0; i < 64 * 1024; i++)
+ xb_clear_bit(&xb1, base + i);
+
+ assert(xb_empty(&xb1));
+}
+
+static void xbitmap_checks(void)
+{
+ xb_init(&xb1);
+ xbitmap_check_bit(0);
+ xbitmap_check_bit(30);
+ xbitmap_check_bit(31);
+ xbitmap_check_bit(62);
+ xbitmap_check_bit(63);
+ xbitmap_check_bit(64);
+ xbitmap_check_bit(700);
+ xbitmap_check_bit(1023);
+ xbitmap_check_bit(1024);
+ xbitmap_check_bit(1025);
+ xbitmap_check_bit((1UL << 63) | (1UL << 24));
+ xbitmap_check_bit((1UL << 63) | (1UL << 24) | 70);
+
+ xbitmap_check_bit_range();
+ xbitmap_check_zero_bits();
+ xbitmap_check_set(0);
+ xbitmap_check_set(1024);
+ xbitmap_check_set(1024 * 64);
+}
+
+int __weak main(void)
+{
+ radix_tree_init();
+ xbitmap_checks();
+}
+#endif
diff --git a/tools/include/linux/bitmap.h b/tools/include/linux/bitmap.h
index ca16027..8d0bc1b 100644
--- a/tools/include/linux/bitmap.h
+++ b/tools/include/linux/bitmap.h
@@ -37,6 +37,40 @@ static inline void bitmap_zero(unsigned long *dst, int nbits)
}
}
+static inline void __bitmap_clear(unsigned long *map, unsigned int start,
+ int len)
+{
+ unsigned long *p = map + BIT_WORD(start);
+ const unsigned int size = start + len;
+ int bits_to_clear = BITS_PER_LONG - (start % BITS_PER_LONG);
+ unsigned long mask_to_clear = BITMAP_FIRST_WORD_MASK(start);
+
+ while (len - bits_to_clear >= 0) {
+ *p &= ~mask_to_clear;
+ len -= bits_to_clear;
+ bits_to_clear = BITS_PER_LONG;
+ mask_to_clear = ~0UL;
+ p++;
+ }
+ if (len) {
+ mask_to_clear &= BITMAP_LAST_WORD_MASK(size);
+ *p &= ~mask_to_clear;
+ }
+}
+
+static inline __always_inline void bitmap_clear(unsigned long *map,
+ unsigned int start,
+ unsigned int nbits)
+{
+ if (__builtin_constant_p(nbits) && nbits == 1)
+ __clear_bit(start, map);
+ else if (__builtin_constant_p(start & 7) && IS_ALIGNED(start, 8)
&&
+ __builtin_constant_p(nbits & 7) && IS_ALIGNED(nbits, 8))
+ memset((char *)map + start / 8, 0, nbits / 8);
+ else
+ __bitmap_clear(map, start, nbits);
+}
+
static inline void bitmap_fill(unsigned long *dst, unsigned int nbits)
{
unsigned int nlongs = BITS_TO_LONGS(nbits);
diff --git a/tools/include/linux/kernel.h b/tools/include/linux/kernel.h
index 0ad8844..3c992ae 100644
--- a/tools/include/linux/kernel.h
+++ b/tools/include/linux/kernel.h
@@ -13,6 +13,8 @@
#define UINT_MAX (~0U)
#endif
+#define IS_ALIGNED(x, a) (((x) & ((typeof(x))(a) - 1)) == 0)
+
#define DIV_ROUND_UP(n,d) (((n) + (d) - 1) / (d))
#define PERF_ALIGN(x, a) __PERF_ALIGN_MASK(x, (typeof(x))(a)-1)
diff --git a/tools/testing/radix-tree/Makefile
b/tools/testing/radix-tree/Makefile
index fa7ee36..788e526 100644
--- a/tools/testing/radix-tree/Makefile
+++ b/tools/testing/radix-tree/Makefile
@@ -1,12 +1,13 @@
# SPDX-License-Identifier: GPL-2.0
CFLAGS += -I. -I../../include -g -O2 -Wall -D_LGPL_SOURCE -fsanitize=address
-LDFLAGS += -fsanitize=address
-LDLIBS+= -lpthread -lurcu
-TARGETS = main idr-test multiorder
+LDFLAGS += -fsanitize=address $(LDLIBS)
+LDLIBS := -lpthread -lurcu
+TARGETS = main idr-test multiorder xbitmap
CORE_OFILES := radix-tree.o idr.o linux.o test.o find_bit.o
OFILES = main.o $(CORE_OFILES) regression1.o regression2.o regression3.o \
- tag_check.o multiorder.o idr-test.o iteration_check.o benchmark.o
+ tag_check.o multiorder.o idr-test.o iteration_check.o benchmark.o \
+ xbitmap.o
ifndef SHIFT
SHIFT=3
@@ -25,8 +26,10 @@ idr-test: idr-test.o $(CORE_OFILES)
multiorder: multiorder.o $(CORE_OFILES)
+xbitmap: xbitmap.o $(CORE_OFILES)
+
clean:
- $(RM) $(TARGETS) *.o radix-tree.c idr.c generated/map-shift.h
+ $(RM) $(TARGETS) *.o radix-tree.c idr.c xbitmap.c generated/map-shift.h
vpath %.c ../../lib
@@ -34,6 +37,7 @@ $(OFILES): Makefile *.h */*.h generated/map-shift.h \
../../include/linux/*.h \
../../include/asm/*.h \
../../../include/linux/radix-tree.h \
+ ../../../include/linux/xbitmap.h \
../../../include/linux/idr.h
radix-tree.c: ../../../lib/radix-tree.c
@@ -42,6 +46,9 @@ radix-tree.c: ../../../lib/radix-tree.c
idr.c: ../../../lib/idr.c
sed -e 's/^static //' -e 's/__always_inline //' -e
's/inline //' < $< > $@
+xbitmap.c: ../../../lib/xbitmap.c
+ sed -e 's/^static //' -e 's/__always_inline //' -e
's/inline //' < $< > $@
+
.PHONY: mapshift
mapshift:
diff --git a/tools/testing/radix-tree/linux/kernel.h
b/tools/testing/radix-tree/linux/kernel.h
index c3bc3f3..426f32f 100644
--- a/tools/testing/radix-tree/linux/kernel.h
+++ b/tools/testing/radix-tree/linux/kernel.h
@@ -17,6 +17,4 @@
#define pr_debug printk
#define pr_cont printk
-#define ARRAY_SIZE(arr) (sizeof(arr) / sizeof((arr)[0]))
-
#endif /* _KERNEL_H */
diff --git a/tools/testing/radix-tree/linux/xbitmap.h
b/tools/testing/radix-tree/linux/xbitmap.h
new file mode 100644
index 0000000..61de214
--- /dev/null
+++ b/tools/testing/radix-tree/linux/xbitmap.h
@@ -0,0 +1 @@
+#include "../../../../include/linux/xbitmap.h"
diff --git a/tools/testing/radix-tree/main.c b/tools/testing/radix-tree/main.c
index 257f3f8..d112363 100644
--- a/tools/testing/radix-tree/main.c
+++ b/tools/testing/radix-tree/main.c
@@ -326,6 +326,10 @@ static void single_thread_tests(bool long_run)
rcu_barrier();
printv(2, "after idr_checks: %d allocated, preempt %d\n",
nr_allocated, preempt_count);
+ xbitmap_checks();
+ rcu_barrier();
+ printv(2, "after xbitmap_checks: %d allocated, preempt %d\n",
+ nr_allocated, preempt_count);
big_gang_check(long_run);
rcu_barrier();
printv(2, "after big_gang_check: %d allocated, preempt %d\n",
diff --git a/tools/testing/radix-tree/test.h b/tools/testing/radix-tree/test.h
index d9c031d..8175d6b 100644
--- a/tools/testing/radix-tree/test.h
+++ b/tools/testing/radix-tree/test.h
@@ -38,6 +38,7 @@ void benchmark(void);
void idr_checks(void);
void ida_checks(void);
void ida_thread_tests(void);
+void xbitmap_checks(void);
struct item *
item_tag_set(struct radix_tree_root *root, unsigned long index, int tag);
--
2.7.4
Add a new feature, VIRTIO_BALLOON_F_SG, which enables the transfer of
balloon (i.e. inflated/deflated) pages using scatter-gather lists to the
host.
The implementation of the previous virtio-balloon is not very efficient,
because the balloon pages are transferred to the host by one array each
time. Here is the breakdown of the time in percentage spent on each step
of the balloon inflating process (inflating 7GB of an 8GB idle guest).
1) allocating pages (6.5%)
2) sending PFNs to host (68.3%)
3) address translation (6.1%)
4) madvise (19%)
It takes about 4126ms for the inflating process to complete. The above
profiling shows that the bottlenecks are stage 2) and stage 4).
This patch optimizes step 2) by transferring pages to host in sgs. An sg
describes a chunk of guest physically continuous pages. With this
mechanism, step 4) can also be optimized by doing address translation and
madvise() in chunks rather than page by page.
With this new feature, the above ballooning process takes ~460ms resulting
in an improvement of ~89%.
TODO:
- optimize stage 1) by allocating/freeing a chunk of pages instead of a
single page each time.
- sort the internal balloon page queue.
- enable OOM to free inflated pages maintained in the local temporary
list.
Signed-off-by: Wei Wang <wei.w.wang at intel.com>
Signed-off-by: Liang Li <liang.z.li at intel.com>
Suggested-by: Michael S. Tsirkin <mst at redhat.com>
Cc: Tetsuo Handa <penguin-kernel at I-love.SAKURA.ne.jp>
---
drivers/virtio/virtio_balloon.c | 233 +++++++++++++++++++++++++++++++++---
include/uapi/linux/virtio_balloon.h | 1 +
2 files changed, 216 insertions(+), 18 deletions(-)
diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
index a1fb52c..10876ea 100644
--- a/drivers/virtio/virtio_balloon.c
+++ b/drivers/virtio/virtio_balloon.c
@@ -32,6 +32,8 @@
#include <linux/mm.h>
#include <linux/mount.h>
#include <linux/magic.h>
+#include <linux/xbitmap.h>
+#include <asm/page.h>
/*
* Balloon device works in 4K page units. So each page is pointed to by
@@ -79,6 +81,9 @@ struct virtio_balloon {
/* Synchronize access/update to this struct virtio_balloon elements */
struct mutex balloon_lock;
+ /* The xbitmap used to record balloon pages */
+ struct xb page_xb;
+
/* The array of pfns we tell the Host about. */
unsigned int num_pfns;
__virtio32 pfns[VIRTIO_BALLOON_ARRAY_PFNS_MAX];
@@ -141,15 +146,128 @@ static void set_page_pfns(struct virtio_balloon *vb,
page_to_balloon_pfn(page) + i);
}
+static void kick_and_wait(struct virtqueue *vq, wait_queue_head_t wq_head)
+{
+ unsigned int len;
+
+ virtqueue_kick(vq);
+ wait_event(wq_head, virtqueue_get_buf(vq, &len));
+}
+
+static void add_one_sg(struct virtqueue *vq, unsigned long pfn, uint32_t len)
+{
+ struct scatterlist sg;
+ unsigned int unused;
+ int err;
+
+ sg_init_table(&sg, 1);
+ sg_set_page(&sg, pfn_to_page(pfn), len, 0);
+
+ /* Detach all the used buffers from the vq */
+ while (virtqueue_get_buf(vq, &unused))
+ ;
+
+ err = virtqueue_add_inbuf(vq, &sg, 1, vq, GFP_KERNEL);
+ /*
+ * This is expected to never fail: there is always at least 1 entry
+ * available on the vq, because when the vq is full the worker thread
+ * that adds the sg will be put into sleep until at least 1 entry is
+ * available to use.
+ */
+ BUG_ON(err);
+}
+
+static void batch_balloon_page_sg(struct virtio_balloon *vb,
+ struct virtqueue *vq,
+ unsigned long pfn,
+ uint32_t len)
+{
+ add_one_sg(vq, pfn, len);
+
+ /* Batch till the vq is full */
+ if (!vq->num_free)
+ kick_and_wait(vq, vb->acked);
+}
+
+/*
+ * Send balloon pages in sgs to host. The balloon pages are recorded in the
+ * page xbitmap. Each bit in the bitmap corresponds to a page of PAGE_SIZE.
+ * The page xbitmap is searched for continuous "1" bits, which
correspond
+ * to continuous pages, to chunk into sgs.
+ *
+ * @page_xb_start and @page_xb_end form the range of bits in the xbitmap that
+ * need to be searched.
+ */
+static void tell_host_sgs(struct virtio_balloon *vb,
+ struct virtqueue *vq,
+ unsigned long page_xb_start,
+ unsigned long page_xb_end)
+{
+ unsigned long pfn_start, pfn_end;
+ uint32_t max_len = round_down(UINT_MAX, PAGE_SIZE);
+ uint64_t len;
+
+ pfn_start = page_xb_start;
+ while (pfn_start < page_xb_end) {
+ if (!xb_find_set(&vb->page_xb, page_xb_end, &pfn_start))
+ break;
+ pfn_end = pfn_start + 1;
+ if (!xb_find_zero(&vb->page_xb, page_xb_end, &pfn_end))
+ pfn_end = page_xb_end + 1;
+ len = (pfn_end - pfn_start) << PAGE_SHIFT;
+ while (len > max_len) {
+ batch_balloon_page_sg(vb, vq, pfn_start, max_len);
+ pfn_start += max_len >> PAGE_SHIFT;
+ len -= max_len;
+ }
+ batch_balloon_page_sg(vb, vq, pfn_start, (uint32_t)len);
+ pfn_start = pfn_end + 1;
+ }
+
+ /*
+ * The last few sgs may not reach the batch size, but need a kick to
+ * notify the device to handle them.
+ */
+ if (vq->num_free != virtqueue_get_vring_size(vq))
+ kick_and_wait(vq, vb->acked);
+
+ xb_zero(&vb->page_xb, page_xb_start, page_xb_end);
+}
+
+static inline int xb_set_page(struct virtio_balloon *vb,
+ struct page *page,
+ unsigned long *pfn_min,
+ unsigned long *pfn_max)
+{
+ unsigned long pfn = page_to_pfn(page);
+ int ret;
+
+ *pfn_min = min(pfn, *pfn_min);
+ *pfn_max = max(pfn, *pfn_max);
+
+ do {
+ if (xb_preload(GFP_NOWAIT | __GFP_NOWARN) < 0)
+ return -ENOMEM;
+
+ ret = xb_set_bit(&vb->page_xb, pfn);
+ xb_preload_end();
+ } while (unlikely(ret == -EAGAIN));
+
+ return ret;
+}
+
static unsigned fill_balloon(struct virtio_balloon *vb, size_t num)
{
unsigned num_allocated_pages;
unsigned num_pfns;
struct page *page;
LIST_HEAD(pages);
+ bool use_sg = virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_SG);
+ unsigned long pfn_max = 0, pfn_min = ULONG_MAX;
/* We can only do one array worth at a time. */
- num = min(num, ARRAY_SIZE(vb->pfns));
+ if (!use_sg)
+ num = min(num, ARRAY_SIZE(vb->pfns));
for (num_pfns = 0; num_pfns < num;
num_pfns += VIRTIO_BALLOON_PAGES_PER_PAGE) {
@@ -173,8 +291,15 @@ static unsigned fill_balloon(struct virtio_balloon *vb,
size_t num)
while ((page = balloon_page_pop(&pages))) {
balloon_page_enqueue(&vb->vb_dev_info, page);
+ if (use_sg) {
+ if (xb_set_page(vb, page, &pfn_min, &pfn_max) < 0) {
+ __free_page(page);
+ continue;
+ }
+ } else {
+ set_page_pfns(vb, vb->pfns + vb->num_pfns, page);
+ }
- set_page_pfns(vb, vb->pfns + vb->num_pfns, page);
vb->num_pages += VIRTIO_BALLOON_PAGES_PER_PAGE;
if (!virtio_has_feature(vb->vdev,
VIRTIO_BALLOON_F_DEFLATE_ON_OOM))
@@ -184,8 +309,12 @@ static unsigned fill_balloon(struct virtio_balloon *vb,
size_t num)
num_allocated_pages = vb->num_pfns;
/* Did we get any? */
- if (vb->num_pfns != 0)
- tell_host(vb, vb->inflate_vq);
+ if (vb->num_pfns) {
+ if (use_sg)
+ tell_host_sgs(vb, vb->inflate_vq, pfn_min, pfn_max);
+ else
+ tell_host(vb, vb->inflate_vq);
+ }
mutex_unlock(&vb->balloon_lock);
return num_allocated_pages;
@@ -211,9 +340,12 @@ static unsigned leak_balloon(struct virtio_balloon *vb,
size_t num)
struct page *page;
struct balloon_dev_info *vb_dev_info = &vb->vb_dev_info;
LIST_HEAD(pages);
+ bool use_sg = virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_SG);
+ unsigned long pfn_max = 0, pfn_min = ULONG_MAX;
- /* We can only do one array worth at a time. */
- num = min(num, ARRAY_SIZE(vb->pfns));
+ /* Traditionally, we can only do one array worth at a time. */
+ if (!use_sg)
+ num = min(num, ARRAY_SIZE(vb->pfns));
mutex_lock(&vb->balloon_lock);
/* We can't release more pages than taken */
@@ -223,7 +355,14 @@ static unsigned leak_balloon(struct virtio_balloon *vb,
size_t num)
page = balloon_page_dequeue(vb_dev_info);
if (!page)
break;
- set_page_pfns(vb, vb->pfns + vb->num_pfns, page);
+ if (use_sg) {
+ if (xb_set_page(vb, page, &pfn_min, &pfn_max) < 0) {
+ balloon_page_enqueue(&vb->vb_dev_info, page);
+ break;
+ }
+ } else {
+ set_page_pfns(vb, vb->pfns + vb->num_pfns, page);
+ }
list_add(&page->lru, &pages);
vb->num_pages -= VIRTIO_BALLOON_PAGES_PER_PAGE;
}
@@ -234,13 +373,55 @@ static unsigned leak_balloon(struct virtio_balloon *vb,
size_t num)
* virtio_has_feature(vdev, VIRTIO_BALLOON_F_MUST_TELL_HOST);
* is true, we *have* to do it in this order
*/
- if (vb->num_pfns != 0)
- tell_host(vb, vb->deflate_vq);
+ if (vb->num_pfns) {
+ if (use_sg)
+ tell_host_sgs(vb, vb->deflate_vq, pfn_min, pfn_max);
+ else
+ tell_host(vb, vb->deflate_vq);
+ }
release_pages_balloon(vb, &pages);
mutex_unlock(&vb->balloon_lock);
return num_freed_pages;
}
+/*
+ * The regular leak_balloon() with VIRTIO_BALLOON_F_SG needs memory allocation
+ * for xbitmap, which is not suitable for the oom case. This function does not
+ * use xbitmap to chunk pages, so it can be used by oom notifier to deflate
+ * pages when VIRTIO_BALLOON_F_SG is negotiated.
+ */
+static unsigned int leak_balloon_sg_oom(struct virtio_balloon *vb)
+{
+ unsigned int n;
+ struct page *page;
+ struct balloon_dev_info *vb_dev_info = &vb->vb_dev_info;
+ struct virtqueue *vq = vb->deflate_vq;
+ LIST_HEAD(pages);
+
+ mutex_lock(&vb->balloon_lock);
+ for (n = 0; n < oom_pages; n++) {
+ page = balloon_page_dequeue(vb_dev_info);
+ if (!page)
+ break;
+
+ list_add(&page->lru, &pages);
+ vb->num_pages -= VIRTIO_BALLOON_PAGES_PER_PAGE;
+ batch_balloon_page_sg(vb, vb->deflate_vq, page_to_pfn(page),
+ PAGE_SIZE);
+ release_pages_balloon(vb, &pages);
+ }
+
+ /*
+ * The last few sgs may not reach the batch size, but need a kick to
+ * notify the device to handle them.
+ */
+ if (vq->num_free != virtqueue_get_vring_size(vq))
+ kick_and_wait(vq, vb->acked);
+ mutex_unlock(&vb->balloon_lock);
+
+ return n;
+}
+
static inline void update_stat(struct virtio_balloon *vb, int idx,
u16 tag, u64 val)
{
@@ -380,7 +561,10 @@ static int virtballoon_oom_notify(struct notifier_block
*self,
return NOTIFY_OK;
freed = parm;
- num_freed_pages = leak_balloon(vb, oom_pages);
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_SG))
+ num_freed_pages = leak_balloon_sg_oom(vb);
+ else
+ num_freed_pages = leak_balloon(vb, oom_pages);
update_balloon_size(vb);
*freed += num_freed_pages;
@@ -477,6 +661,7 @@ static int virtballoon_migratepage(struct balloon_dev_info
*vb_dev_info,
{
struct virtio_balloon *vb = container_of(vb_dev_info,
struct virtio_balloon, vb_dev_info);
+ bool use_sg = virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_SG);
unsigned long flags;
/*
@@ -498,16 +683,24 @@ static int virtballoon_migratepage(struct balloon_dev_info
*vb_dev_info,
vb_dev_info->isolated_pages--;
__count_vm_event(BALLOON_MIGRATE);
spin_unlock_irqrestore(&vb_dev_info->pages_lock, flags);
- vb->num_pfns = VIRTIO_BALLOON_PAGES_PER_PAGE;
- set_page_pfns(vb, vb->pfns, newpage);
- tell_host(vb, vb->inflate_vq);
-
+ if (use_sg) {
+ add_one_sg(vb->inflate_vq, page_to_pfn(newpage), PAGE_SIZE);
+ kick_and_wait(vb->inflate_vq, vb->acked);
+ } else {
+ vb->num_pfns = VIRTIO_BALLOON_PAGES_PER_PAGE;
+ set_page_pfns(vb, vb->pfns, newpage);
+ tell_host(vb, vb->inflate_vq);
+ }
/* balloon's page migration 2nd step -- deflate "page" */
balloon_page_delete(page);
- vb->num_pfns = VIRTIO_BALLOON_PAGES_PER_PAGE;
- set_page_pfns(vb, vb->pfns, page);
- tell_host(vb, vb->deflate_vq);
-
+ if (use_sg) {
+ add_one_sg(vb->deflate_vq, page_to_pfn(page), PAGE_SIZE);
+ kick_and_wait(vb->deflate_vq, vb->acked);
+ } else {
+ vb->num_pfns = VIRTIO_BALLOON_PAGES_PER_PAGE;
+ set_page_pfns(vb, vb->pfns, page);
+ tell_host(vb, vb->deflate_vq);
+ }
mutex_unlock(&vb->balloon_lock);
put_page(page); /* balloon reference */
@@ -566,6 +759,9 @@ static int virtballoon_probe(struct virtio_device *vdev)
if (err)
goto out_free_vb;
+ if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_SG))
+ xb_init(&vb->page_xb);
+
vb->nb.notifier_call = virtballoon_oom_notify;
vb->nb.priority = VIRTBALLOON_OOM_NOTIFY_PRIORITY;
err = register_oom_notifier(&vb->nb);
@@ -682,6 +878,7 @@ static unsigned int features[] = {
VIRTIO_BALLOON_F_MUST_TELL_HOST,
VIRTIO_BALLOON_F_STATS_VQ,
VIRTIO_BALLOON_F_DEFLATE_ON_OOM,
+ VIRTIO_BALLOON_F_SG,
};
static struct virtio_driver virtio_balloon_driver = {
diff --git a/include/uapi/linux/virtio_balloon.h
b/include/uapi/linux/virtio_balloon.h
index 343d7dd..37780a7 100644
--- a/include/uapi/linux/virtio_balloon.h
+++ b/include/uapi/linux/virtio_balloon.h
@@ -34,6 +34,7 @@
#define VIRTIO_BALLOON_F_MUST_TELL_HOST 0 /* Tell before reclaiming pages */
#define VIRTIO_BALLOON_F_STATS_VQ 1 /* Memory Stats virtqueue */
#define VIRTIO_BALLOON_F_DEFLATE_ON_OOM 2 /* Deflate balloon on OOM */
+#define VIRTIO_BALLOON_F_SG 3 /* Use sg instead of PFN lists */
/* Size of a PFN in the balloon interface. */
#define VIRTIO_BALLOON_PFN_SHIFT 12
--
2.7.4
This patch adds support to walk through the free page blocks in the
system and report them via a callback function. Some page blocks may
leave the free list after zone->lock is released, so it is the caller's
responsibility to either detect or prevent the use of such pages.
One use example of this patch is to accelerate live migration by skipping
the transfer of free pages reported from the guest. A popular method used
by the hypervisor to track which part of memory is written during live
migration is to write-protect all the guest memory. So, those pages that
are reported as free pages but are written after the report function
returns will be captured by the hypervisor, and they will be added to the
next round of memory transfer.
Signed-off-by: Wei Wang <wei.w.wang at intel.com>
Signed-off-by: Liang Li <liang.z.li at intel.com>
Cc: Michal Hocko <mhocko at kernel.org>
Cc: Michael S. Tsirkin <mst at redhat.com>
Acked-by: Michal Hocko <mhocko at kernel.org>
---
include/linux/mm.h | 6 ++++
mm/page_alloc.c | 91 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 97 insertions(+)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index ea818ff..b3077dd 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -1938,6 +1938,12 @@ extern void free_area_init_node(int nid, unsigned long *
zones_size,
unsigned long zone_start_pfn, unsigned long *zholes_size);
extern void free_initmem(void);
+extern void walk_free_mem_block(void *opaque,
+ int min_order,
+ bool (*report_pfn_range)(void *opaque,
+ unsigned long pfn,
+ unsigned long num));
+
/*
* Free reserved pages within range [PAGE_ALIGN(start), end & PAGE_MASK)
* into the buddy system. The freed pages will be poisoned with pattern
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 76c9688..705de22 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -4899,6 +4899,97 @@ void show_free_areas(unsigned int filter, nodemask_t
*nodemask)
show_swap_cache_info();
}
+/*
+ * Walk through a free page list and report the found pfn range via the
+ * callback.
+ *
+ * Return false if the callback requests to stop reporting. Otherwise,
+ * return true.
+ */
+static bool walk_free_page_list(void *opaque,
+ struct zone *zone,
+ int order,
+ enum migratetype mt,
+ bool (*report_pfn_range)(void *,
+ unsigned long,
+ unsigned long))
+{
+ struct page *page;
+ struct list_head *list;
+ unsigned long pfn, flags;
+ bool ret;
+
+ spin_lock_irqsave(&zone->lock, flags);
+ list = &zone->free_area[order].free_list[mt];
+ list_for_each_entry(page, list, lru) {
+ pfn = page_to_pfn(page);
+ ret = report_pfn_range(opaque, pfn, 1 << order);
+ if (!ret)
+ break;
+ }
+ spin_unlock_irqrestore(&zone->lock, flags);
+
+ return ret;
+}
+
+/**
+ * walk_free_mem_block - Walk through the free page blocks in the system
+ * @opaque: the context passed from the caller
+ * @min_order: the minimum order of free lists to check
+ * @report_pfn_range: the callback to report the pfn range of the free pages
+ *
+ * If the callback returns false, stop iterating the list of free page blocks.
+ * Otherwise, continue to report.
+ *
+ * Please note that there are no locking guarantees for the callback and
+ * that the reported pfn range might be freed or disappear after the
+ * callback returns so the caller has to be very careful how it is used.
+ *
+ * The callback itself must not sleep or perform any operations which would
+ * require any memory allocations directly (not even GFP_NOWAIT/GFP_ATOMIC)
+ * or via any lock dependency. It is generally advisable to implement
+ * the callback as simple as possible and defer any heavy lifting to a
+ * different context.
+ *
+ * There is no guarantee that each free range will be reported only once
+ * during one walk_free_mem_block invocation.
+ *
+ * pfn_to_page on the given range is strongly discouraged and if there is
+ * an absolute need for that make sure to contact MM people to discuss
+ * potential problems.
+ *
+ * The function itself might sleep so it cannot be called from atomic
+ * contexts.
+ *
+ * In general low orders tend to be very volatile and so it makes more
+ * sense to query larger ones first for various optimizations which like
+ * ballooning etc... This will reduce the overhead as well.
+ */
+void walk_free_mem_block(void *opaque,
+ int min_order,
+ bool (*report_pfn_range)(void *opaque,
+ unsigned long pfn,
+ unsigned long num))
+{
+ struct zone *zone;
+ int order;
+ enum migratetype mt;
+ bool ret;
+
+ for_each_populated_zone(zone) {
+ for (order = MAX_ORDER - 1; order >= min_order; order--) {
+ for (mt = 0; mt < MIGRATE_TYPES; mt++) {
+ ret = walk_free_page_list(opaque, zone,
+ order, mt,
+ report_pfn_range);
+ if (!ret)
+ return;
+ }
+ }
+ }
+}
+EXPORT_SYMBOL_GPL(walk_free_mem_block);
+
static void zoneref_set_zone(struct zone *zone, struct zoneref *zoneref)
{
zoneref->zone = zone;
--
2.7.4
Wei Wang
2018-Jan-09 11:11 UTC
[PATCH v21 4/5] virtio-balloon: VIRTIO_BALLOON_F_FREE_PAGE_VQ
Negotiation of the VIRTIO_BALLOON_F_FREE_PAGE_VQ feature indicates the
support of reporting hints of guest free pages to host via virtio-balloon.
Host requests the guest to report free pages by sending a new cmd
id to the guest via the free_page_report_cmd_id configuration register.
When the guest starts to report, the first element added to the free page
vq is the cmd id given by host. When the guest finishes the reporting
of all the free pages, VIRTIO_BALLOON_FREE_PAGE_REPORT_STOP_ID is added
to the vq to tell host that the reporting is done. Host may also requests
the guest to stop the reporting in advance by sending the stop cmd id to
the guest via the configuration register.
Signed-off-by: Wei Wang <wei.w.wang at intel.com>
Signed-off-by: Liang Li <liang.z.li at intel.com>
Cc: Michael S. Tsirkin <mst at redhat.com>
Cc: Michal Hocko <mhocko at kernel.org>
---
drivers/virtio/virtio_balloon.c | 202 ++++++++++++++++++++++++++++++------
include/uapi/linux/virtio_balloon.h | 4 +
2 files changed, 174 insertions(+), 32 deletions(-)
diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
index 10876ea..f7e8830 100644
--- a/drivers/virtio/virtio_balloon.c
+++ b/drivers/virtio/virtio_balloon.c
@@ -55,7 +55,12 @@ static struct vfsmount *balloon_mnt;
struct virtio_balloon {
struct virtio_device *vdev;
- struct virtqueue *inflate_vq, *deflate_vq, *stats_vq;
+ struct virtqueue *inflate_vq, *deflate_vq, *stats_vq, *free_page_vq;
+
+ /* Balloon's own wq for cpu-intensive work items */
+ struct workqueue_struct *balloon_wq;
+ /* The free page reporting work item submitted to the balloon wq */
+ struct work_struct report_free_page_work;
/* The balloon servicing is delegated to a freezable workqueue. */
struct work_struct update_balloon_stats_work;
@@ -65,6 +70,13 @@ struct virtio_balloon {
spinlock_t stop_update_lock;
bool stop_update;
+ /* Start to report free pages */
+ bool report_free_page;
+ /* Stores the cmd id given by host to start the free page reporting */
+ uint32_t start_cmd_id;
+ /* Stores STOP_ID as a sign to tell host that the reporting is done */
+ uint32_t stop_cmd_id;
+
/* Waiting for host to ack the pages we released. */
wait_queue_head_t acked;
@@ -189,6 +201,28 @@ static void batch_balloon_page_sg(struct virtio_balloon
*vb,
kick_and_wait(vq, vb->acked);
}
+static void batch_free_page_sg(struct virtqueue *vq,
+ unsigned long pfn,
+ uint32_t len)
+{
+ add_one_sg(vq, pfn, len);
+
+ /* Batch till the vq is full */
+ if (!vq->num_free)
+ virtqueue_kick(vq);
+}
+
+static void send_cmd_id(struct virtio_balloon *vb, void *addr)
+{
+ struct scatterlist sg;
+ int err;
+
+ sg_init_one(&sg, addr, sizeof(uint32_t));
+ err = virtqueue_add_outbuf(vb->free_page_vq, &sg, 1, vb, GFP_KERNEL);
+ BUG_ON(err);
+ virtqueue_kick(vb->free_page_vq);
+}
+
/*
* Send balloon pages in sgs to host. The balloon pages are recorded in the
* page xbitmap. Each bit in the bitmap corresponds to a page of PAGE_SIZE.
@@ -497,17 +531,6 @@ static void stats_handle_request(struct virtio_balloon *vb)
virtqueue_kick(vq);
}
-static void virtballoon_changed(struct virtio_device *vdev)
-{
- struct virtio_balloon *vb = vdev->priv;
- unsigned long flags;
-
- spin_lock_irqsave(&vb->stop_update_lock, flags);
- if (!vb->stop_update)
- queue_work(system_freezable_wq, &vb->update_balloon_size_work);
- spin_unlock_irqrestore(&vb->stop_update_lock, flags);
-}
-
static inline s64 towards_target(struct virtio_balloon *vb)
{
s64 target;
@@ -524,6 +547,36 @@ static inline s64 towards_target(struct virtio_balloon *vb)
return target - vb->num_pages;
}
+static void virtballoon_changed(struct virtio_device *vdev)
+{
+ struct virtio_balloon *vb = vdev->priv;
+ unsigned long flags;
+ __u32 cmd_id;
+ s64 diff = towards_target(vb);
+
+ if (diff) {
+ spin_lock_irqsave(&vb->stop_update_lock, flags);
+ if (!vb->stop_update)
+ queue_work(system_freezable_wq,
+ &vb->update_balloon_size_work);
+ spin_unlock_irqrestore(&vb->stop_update_lock, flags);
+ }
+
+ virtio_cread(vb->vdev, struct virtio_balloon_config,
+ free_page_report_cmd_id, &cmd_id);
+ if (cmd_id == VIRTIO_BALLOON_FREE_PAGE_REPORT_STOP_ID) {
+ WRITE_ONCE(vb->report_free_page, false);
+ } else if (cmd_id != vb->start_cmd_id) {
+ /*
+ * Host requests to start the reporting by sending a new cmd
+ * id.
+ */
+ WRITE_ONCE(vb->report_free_page, true);
+ vb->start_cmd_id = cmd_id;
+ queue_work(vb->balloon_wq, &vb->report_free_page_work);
+ }
+}
+
static void update_balloon_size(struct virtio_balloon *vb)
{
u32 actual = vb->num_pages;
@@ -601,40 +654,116 @@ static void update_balloon_size_func(struct work_struct
*work)
static int init_vqs(struct virtio_balloon *vb)
{
- struct virtqueue *vqs[3];
- vq_callback_t *callbacks[] = { balloon_ack, balloon_ack, stats_request };
- static const char * const names[] = { "inflate",
"deflate", "stats" };
- int err, nvqs;
+ struct virtqueue **vqs;
+ vq_callback_t **callbacks;
+ const char **names;
+ struct scatterlist sg;
+ int i, nvqs, err = -ENOMEM;
+
+ /* Inflateq and deflateq are used unconditionally */
+ nvqs = 2;
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ))
+ nvqs++;
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_VQ))
+ nvqs++;
+
+ /* Allocate space for find_vqs parameters */
+ vqs = kcalloc(nvqs, sizeof(*vqs), GFP_KERNEL);
+ if (!vqs)
+ goto err_vq;
+ callbacks = kmalloc_array(nvqs, sizeof(*callbacks), GFP_KERNEL);
+ if (!callbacks)
+ goto err_callback;
+ names = kmalloc_array(nvqs, sizeof(*names), GFP_KERNEL);
+ if (!names)
+ goto err_names;
+
+ callbacks[0] = balloon_ack;
+ names[0] = "inflate";
+ callbacks[1] = balloon_ack;
+ names[1] = "deflate";
+
+ i = 2;
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ)) {
+ callbacks[i] = stats_request;
+ names[i] = "stats";
+ i++;
+ }
- /*
- * We expect two virtqueues: inflate and deflate, and
- * optionally stat.
- */
- nvqs = virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ) ? 3 : 2;
- err = virtio_find_vqs(vb->vdev, nvqs, vqs, callbacks, names, NULL);
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_VQ)) {
+ callbacks[i] = NULL;
+ names[i] = "free_page_vq";
+ }
+
+ err = vb->vdev->config->find_vqs(vb->vdev, nvqs, vqs, callbacks,
names,
+ NULL, NULL);
if (err)
- return err;
+ goto err_find;
vb->inflate_vq = vqs[0];
vb->deflate_vq = vqs[1];
+ i = 2;
if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ)) {
- struct scatterlist sg;
- unsigned int num_stats;
- vb->stats_vq = vqs[2];
-
+ vb->stats_vq = vqs[i++];
/*
* Prime this virtqueue with one buffer so the hypervisor can
* use it to signal us later (it can't be broken yet!).
*/
- num_stats = update_balloon_stats(vb);
-
- sg_init_one(&sg, vb->stats, sizeof(vb->stats[0]) * num_stats);
+ sg_init_one(&sg, vb->stats, sizeof(vb->stats));
if (virtqueue_add_outbuf(vb->stats_vq, &sg, 1, vb, GFP_KERNEL)
- < 0)
- BUG();
+ < 0) {
+ dev_warn(&vb->vdev->dev, "%s: add stat_vq failed\n",
+ __func__);
+ goto err_find;
+ }
virtqueue_kick(vb->stats_vq);
}
+
+ if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_VQ))
+ vb->free_page_vq = vqs[i];
+
+ kfree(names);
+ kfree(callbacks);
+ kfree(vqs);
return 0;
+
+err_find:
+ kfree(names);
+err_names:
+ kfree(callbacks);
+err_callback:
+ kfree(vqs);
+err_vq:
+ return err;
+}
+
+static bool virtio_balloon_send_free_pages(void *opaque, unsigned long pfn,
+ unsigned long nr_pages)
+{
+ struct virtio_balloon *vb = (struct virtio_balloon *)opaque;
+ uint32_t len = nr_pages << PAGE_SHIFT;
+
+ if (!READ_ONCE(vb->report_free_page))
+ return false;
+
+ batch_free_page_sg(vb->free_page_vq, pfn, len);
+
+ return true;
+}
+
+static void report_free_page(struct work_struct *work)
+{
+ struct virtio_balloon *vb;
+
+ vb = container_of(work, struct virtio_balloon, report_free_page_work);
+ /* Start by sending the obtained cmd id to the host with an outbuf */
+ send_cmd_id(vb, &vb->start_cmd_id);
+ walk_free_mem_block(vb, 0, &virtio_balloon_send_free_pages);
+ /*
+ * End by sending the stop id to the host with an outbuf. Use the
+ * non-batching mode here to trigger a kick after adding the stop id.
+ */
+ send_cmd_id(vb, &vb->stop_cmd_id);
}
#ifdef CONFIG_BALLOON_COMPACTION
@@ -762,6 +891,13 @@ static int virtballoon_probe(struct virtio_device *vdev)
if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_SG))
xb_init(&vb->page_xb);
+ if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_FREE_PAGE_VQ)) {
+ vb->balloon_wq = alloc_workqueue("balloon-wq",
+ WQ_FREEZABLE | WQ_CPU_INTENSIVE, 0);
+ INIT_WORK(&vb->report_free_page_work, report_free_page);
+ vb->stop_cmd_id = VIRTIO_BALLOON_FREE_PAGE_REPORT_STOP_ID;
+ }
+
vb->nb.notifier_call = virtballoon_oom_notify;
vb->nb.priority = VIRTBALLOON_OOM_NOTIFY_PRIORITY;
err = register_oom_notifier(&vb->nb);
@@ -826,6 +962,7 @@ static void virtballoon_remove(struct virtio_device *vdev)
spin_unlock_irq(&vb->stop_update_lock);
cancel_work_sync(&vb->update_balloon_size_work);
cancel_work_sync(&vb->update_balloon_stats_work);
+ cancel_work_sync(&vb->report_free_page_work);
remove_common(vb);
#ifdef CONFIG_BALLOON_COMPACTION
@@ -879,6 +1016,7 @@ static unsigned int features[] = {
VIRTIO_BALLOON_F_STATS_VQ,
VIRTIO_BALLOON_F_DEFLATE_ON_OOM,
VIRTIO_BALLOON_F_SG,
+ VIRTIO_BALLOON_F_FREE_PAGE_VQ,
};
static struct virtio_driver virtio_balloon_driver = {
diff --git a/include/uapi/linux/virtio_balloon.h
b/include/uapi/linux/virtio_balloon.h
index 37780a7..58f1274 100644
--- a/include/uapi/linux/virtio_balloon.h
+++ b/include/uapi/linux/virtio_balloon.h
@@ -35,15 +35,19 @@
#define VIRTIO_BALLOON_F_STATS_VQ 1 /* Memory Stats virtqueue */
#define VIRTIO_BALLOON_F_DEFLATE_ON_OOM 2 /* Deflate balloon on OOM */
#define VIRTIO_BALLOON_F_SG 3 /* Use sg instead of PFN lists */
+#define VIRTIO_BALLOON_F_FREE_PAGE_VQ 4 /* VQ to report free pages */
/* Size of a PFN in the balloon interface. */
#define VIRTIO_BALLOON_PFN_SHIFT 12
+#define VIRTIO_BALLOON_FREE_PAGE_REPORT_STOP_ID 0
struct virtio_balloon_config {
/* Number of pages host wants Guest to give up. */
__u32 num_pages;
/* Number of pages we've actually got in balloon. */
__u32 actual;
+ /* Free page report command id, readonly by guest */
+ __u32 free_page_report_cmd_id;
};
#define VIRTIO_BALLOON_S_SWAP_IN 0 /* Amount of memory swapped in */
--
2.7.4
Wei Wang
2018-Jan-09 11:11 UTC
[PATCH v21 5/5] virtio-balloon: don't report free pages when page poisoning is enabled
The guest free pages should not be discarded by the live migration thread
when page poisoning is enabled with PAGE_POISONING_NO_SANITY=n, because
skipping the transfer of such poisoned free pages will trigger false
positive when new pages are allocated and checked on the destination.
This patch adds a config field, poison_val. Guest writes to the config
field to tell the host about the poisoning value. The value will be 0 in
the following cases:
1) PAGE_POISONING_NO_SANITY is enabled;
2) page poisoning is disabled; or
3) PAGE_POISONING_ZERO is enabled.
Signed-off-by: Wei Wang <wei.w.wang at intel.com>
Suggested-by: Michael S. Tsirkin <mst at redhat.com>
Cc: Michal Hocko <mhocko at suse.com>
---
drivers/virtio/virtio_balloon.c | 8 ++++++++
include/uapi/linux/virtio_balloon.h | 2 ++
2 files changed, 10 insertions(+)
diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
index f7e8830..7429894 100644
--- a/drivers/virtio/virtio_balloon.c
+++ b/drivers/virtio/virtio_balloon.c
@@ -859,6 +859,7 @@ static struct file_system_type balloon_fs = {
static int virtballoon_probe(struct virtio_device *vdev)
{
struct virtio_balloon *vb;
+ __u32 poison_val;
int err;
if (!vdev->config->get) {
@@ -896,6 +897,13 @@ static int virtballoon_probe(struct virtio_device *vdev)
WQ_FREEZABLE | WQ_CPU_INTENSIVE, 0);
INIT_WORK(&vb->report_free_page_work, report_free_page);
vb->stop_cmd_id = VIRTIO_BALLOON_FREE_PAGE_REPORT_STOP_ID;
+ if (IS_ENABLED(CONFIG_PAGE_POISONING_NO_SANITY) ||
+ !page_poisoning_enabled())
+ poison_val = 0;
+ else
+ poison_val = PAGE_POISON;
+ virtio_cwrite(vb->vdev, struct virtio_balloon_config,
+ poison_val, &poison_val);
}
vb->nb.notifier_call = virtballoon_oom_notify;
diff --git a/include/uapi/linux/virtio_balloon.h
b/include/uapi/linux/virtio_balloon.h
index 58f1274..f270e9e 100644
--- a/include/uapi/linux/virtio_balloon.h
+++ b/include/uapi/linux/virtio_balloon.h
@@ -48,6 +48,8 @@ struct virtio_balloon_config {
__u32 actual;
/* Free page report command id, readonly by guest */
__u32 free_page_report_cmd_id;
+ /* Stores PAGE_POISON if page poisoning with sanity check is in use */
+ __u32 poison_val;
};
#define VIRTIO_BALLOON_S_SWAP_IN 0 /* Amount of memory swapped in */
--
2.7.4
On Tue, Jan 9, 2018 at 1:10 PM, Wei Wang <wei.w.wang at intel.com> wrote:> From: Matthew Wilcox <mawilcox at microsoft.com> > > The eXtensible Bitmap is a sparse bitmap representation which is > efficient for set bits which tend to cluster. It supports up to > 'unsigned long' worth of bits.> lib/xbitmap.c | 444 +++++++++++++++++++++++++++++++Please, split tests to a separate module. -- With Best Regards, Andy Shevchenko