qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Max Reitz <mreitz@redhat.com>
To: qemu-block@nongnu.org
Cc: Kevin Wolf <kwolf@redhat.com>,
	Peter Maydell <peter.maydell@linaro.org>,
	qemu-devel@nongnu.org, Max Reitz <mreitz@redhat.com>
Subject: [PULL 39/69] util: introduce SharedResource
Date: Mon, 28 Oct 2019 13:14:31 +0100	[thread overview]
Message-ID: <20191028121501.15279-40-mreitz@redhat.com> (raw)
In-Reply-To: <20191028121501.15279-1-mreitz@redhat.com>

From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Introduce an API for some shared splittable resource, like memory.
It's going to be used by backup. Backup uses both read/write io and
copy_range. copy_range may consume memory implictly, so the new API is
abstract: it doesn't allocate any real memory but only hands out
tickets.

The idea is that we have some total amount of something and callers
should wait in coroutine queue if there is not enough of the resource
at the moment.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Reviewed-by: Max Reitz <mreitz@redhat.com>
Message-id: 20191022111805.3432-5-vsementsov@virtuozzo.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
---
 util/Makefile.objs                |  1 +
 include/qemu/co-shared-resource.h | 71 +++++++++++++++++++++++++++++
 util/qemu-co-shared-resource.c    | 76 +++++++++++++++++++++++++++++++
 3 files changed, 148 insertions(+)
 create mode 100644 include/qemu/co-shared-resource.h
 create mode 100644 util/qemu-co-shared-resource.c

diff --git a/util/Makefile.objs b/util/Makefile.objs
index 41bf59d127..df124af1c5 100644
--- a/util/Makefile.objs
+++ b/util/Makefile.objs
@@ -37,6 +37,7 @@ util-obj-y += rcu.o
 util-obj-$(CONFIG_MEMBARRIER) += sys_membarrier.o
 util-obj-y += qemu-coroutine.o qemu-coroutine-lock.o qemu-coroutine-io.o
 util-obj-y += qemu-coroutine-sleep.o
+util-obj-y += qemu-co-shared-resource.o
 util-obj-y += coroutine-$(CONFIG_COROUTINE_BACKEND).o
 util-obj-y += buffer.o
 util-obj-y += timed-average.o
diff --git a/include/qemu/co-shared-resource.h b/include/qemu/co-shared-resource.h
new file mode 100644
index 0000000000..4e4503004c
--- /dev/null
+++ b/include/qemu/co-shared-resource.h
@@ -0,0 +1,71 @@
+/*
+ * Helper functionality for distributing a fixed total amount of
+ * an abstract resource among multiple coroutines.
+ *
+ * Copyright (c) 2019 Virtuozzo International GmbH
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#ifndef QEMU_CO_SHARED_RESOURCE_H
+#define QEMU_CO_SHARED_RESOURCE_H
+
+
+typedef struct SharedResource SharedResource;
+
+/*
+ * Create SharedResource structure
+ *
+ * @total: total amount of some resource to be shared between clients
+ *
+ * Note: this API is not thread-safe.
+ */
+SharedResource *shres_create(uint64_t total);
+
+/*
+ * Release SharedResource structure
+ *
+ * This function may only be called once everything allocated by all
+ * clients has been deallocated.
+ */
+void shres_destroy(SharedResource *s);
+
+/*
+ * Try to allocate an amount of @n.  Return true on success, and false
+ * if there is too little left of the collective resource to fulfill
+ * the request.
+ */
+bool co_try_get_from_shres(SharedResource *s, uint64_t n);
+
+/*
+ * Allocate an amount of @n, and, if necessary, yield until
+ * that becomes possible.
+ */
+void coroutine_fn co_get_from_shres(SharedResource *s, uint64_t n);
+
+/*
+ * Deallocate an amount of @n.  The total amount allocated by a caller
+ * does not need to be deallocated/released with a single call, but may
+ * be split over several calls.  For example, get(4), get(3), and then
+ * put(5), put(2).
+ */
+void coroutine_fn co_put_to_shres(SharedResource *s, uint64_t n);
+
+
+#endif /* QEMU_CO_SHARED_RESOURCE_H */
diff --git a/util/qemu-co-shared-resource.c b/util/qemu-co-shared-resource.c
new file mode 100644
index 0000000000..1c83cd9d29
--- /dev/null
+++ b/util/qemu-co-shared-resource.c
@@ -0,0 +1,76 @@
+/*
+ * Helper functionality for distributing a fixed total amount of
+ * an abstract resource among multiple coroutines.
+ *
+ * Copyright (c) 2019 Virtuozzo International GmbH
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/coroutine.h"
+#include "qemu/co-shared-resource.h"
+
+struct SharedResource {
+    uint64_t total;
+    uint64_t available;
+
+    CoQueue queue;
+};
+
+SharedResource *shres_create(uint64_t total)
+{
+    SharedResource *s = g_new0(SharedResource, 1);
+
+    s->total = s->available = total;
+    qemu_co_queue_init(&s->queue);
+
+    return s;
+}
+
+void shres_destroy(SharedResource *s)
+{
+    assert(s->available == s->total);
+    g_free(s);
+}
+
+bool co_try_get_from_shres(SharedResource *s, uint64_t n)
+{
+    if (s->available >= n) {
+        s->available -= n;
+        return true;
+    }
+
+    return false;
+}
+
+void coroutine_fn co_get_from_shres(SharedResource *s, uint64_t n)
+{
+    assert(n <= s->total);
+    while (!co_try_get_from_shres(s, n)) {
+        qemu_co_queue_wait(&s->queue, NULL);
+    }
+}
+
+void coroutine_fn co_put_to_shres(SharedResource *s, uint64_t n)
+{
+    assert(s->total - s->available >= n);
+    s->available += n;
+    qemu_co_queue_restart_all(&s->queue);
+}
-- 
2.21.0



  parent reply	other threads:[~2019-10-28 12:56 UTC|newest]

Thread overview: 71+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-28 12:13 [PULL 00/69] Block patches for softfreeze Max Reitz
2019-10-28 12:13 ` [PULL 01/69] iotests: Prefer null-co over null-aio Max Reitz
2019-10-28 12:13 ` [PULL 02/69] iotests: Allow skipping test cases Max Reitz
2019-10-28 12:13 ` [PULL 03/69] iotests: Use case_skip() in skip_if_unsupported() Max Reitz
2019-10-28 12:13 ` [PULL 04/69] iotests: Let skip_if_unsupported accept a function Max Reitz
2019-10-28 12:13 ` [PULL 05/69] iotests: Test driver whitelisting in 093 Max Reitz
2019-10-28 12:13 ` [PULL 06/69] iotests: Test driver whitelisting in 136 Max Reitz
2019-10-28 12:13 ` [PULL 07/69] iotests: Cache supported_formats() Max Reitz
2019-10-28 12:14 ` [PULL 08/69] hbitmap: handle set/reset with zero length Max Reitz
2019-10-28 12:14 ` [PULL 09/69] block/mirror: simplify do_sync_target_write Max Reitz
2019-10-28 12:14 ` [PULL 10/69] block/block-backend: add blk_co_pwritev_part Max Reitz
2019-10-28 12:14 ` [PULL 11/69] block/mirror: support unaligned write in active mirror Max Reitz
2019-10-28 12:14 ` [PULL 12/69] Revert "mirror: Only mirror granularity-aligned chunks" Max Reitz
2019-10-28 12:14 ` [PULL 13/69] iotests: Introduce $SOCK_DIR Max Reitz
2019-10-28 12:14 ` [PULL 14/69] iotests.py: Store socket files in $SOCK_DIR Max Reitz
2019-10-28 12:14 ` [PULL 15/69] iotests.py: Add @base_dir to FilePaths etc Max Reitz
2019-10-28 12:14 ` [PULL 16/69] iotests: Filter $SOCK_DIR Max Reitz
2019-10-28 12:14 ` [PULL 17/69] iotests: Let common.nbd create socket in $SOCK_DIR Max Reitz
2019-10-28 12:14 ` [PULL 18/69] iotests/083: Create " Max Reitz
2019-10-28 12:14 ` [PULL 19/69] iotests/140: " Max Reitz
2019-10-28 12:14 ` [PULL 20/69] iotests/143: " Max Reitz
2019-10-28 12:14 ` [PULL 21/69] iotests/147: " Max Reitz
2019-10-28 12:14 ` [PULL 22/69] iotests/181: " Max Reitz
2019-10-28 12:14 ` [PULL 23/69] iotests/182: " Max Reitz
2019-10-28 12:14 ` [PULL 24/69] iotests/183: " Max Reitz
2019-10-28 12:14 ` [PULL 25/69] iotests/192: " Max Reitz
2019-10-28 12:14 ` [PULL 26/69] iotests/194: Create sockets " Max Reitz
2019-10-28 12:14 ` [PULL 27/69] iotests/201: Create socket " Max Reitz
2019-10-28 12:14 ` [PULL 28/69] iotests/205: " Max Reitz
2019-10-28 12:14 ` [PULL 29/69] iotests/208: " Max Reitz
2019-10-28 12:14 ` [PULL 30/69] iotests/209: " Max Reitz
2019-10-28 12:14 ` [PULL 31/69] iotests/222: " Max Reitz
2019-10-28 12:14 ` [PULL 32/69] iotests/223: " Max Reitz
2019-10-28 12:14 ` [PULL 33/69] iotests/240: " Max Reitz
2019-10-28 12:14 ` [PULL 34/69] iotests/267: " Max Reitz
2019-10-28 12:14 ` [PULL 35/69] iotests: Drop TEST_DIR filter from _filter_nbd Max Reitz
2019-10-28 12:14 ` [PULL 36/69] block/block-copy: allocate buffer in block_copy_with_bounce_buffer Max Reitz
2019-10-28 12:14 ` [PULL 37/69] block/block-copy: limit copy_range_size to 16 MiB Max Reitz
2019-10-28 12:14 ` [PULL 38/69] block/block-copy: refactor copying Max Reitz
2019-10-28 12:14 ` Max Reitz [this message]
2019-10-28 12:14 ` [PULL 40/69] block/block-copy: add memory limit Max Reitz
2019-10-28 12:14 ` [PULL 41/69] block/block-copy: increase buffered copy request Max Reitz
2019-10-28 12:14 ` [PULL 42/69] block/nvme: add support for write zeros Max Reitz
2019-10-28 12:14 ` [PULL 43/69] block/nvme: add support for discard Max Reitz
2019-10-28 12:14 ` [PULL 44/69] mirror: Do not dereference invalid pointers Max Reitz
2019-10-28 12:14 ` [PULL 45/69] include: Move endof() up from hw/virtio/virtio.h Max Reitz
2019-10-28 12:14 ` [PULL 46/69] qcow2: Use endof() Max Reitz
2019-10-28 12:14 ` [PULL 47/69] qcow2: Add Error ** to qcow2_read_snapshots() Max Reitz
2019-10-28 12:14 ` [PULL 48/69] qcow2: Keep unknown extra snapshot data Max Reitz
2019-10-28 12:14 ` [PULL 49/69] qcow2: Make qcow2_write_snapshots() public Max Reitz
2019-10-28 12:14 ` [PULL 50/69] qcow2: Put qcow2_upgrade() into its own function Max Reitz
2019-10-28 12:14 ` [PULL 51/69] qcow2: Write v3-compliant snapshot list on upgrade Max Reitz
2019-10-28 12:14 ` [PULL 52/69] qcow2: Separate qcow2_check_read_snapshot_table() Max Reitz
2019-10-28 12:14 ` [PULL 53/69] qcow2: Add qcow2_check_fix_snapshot_table() Max Reitz
2019-10-28 12:14 ` [PULL 54/69] qcow2: Fix broken snapshot table entries Max Reitz
2019-10-28 12:14 ` [PULL 55/69] qcow2: Keep track of the snapshot table length Max Reitz
2019-10-28 12:14 ` [PULL 56/69] qcow2: Fix overly long snapshot tables Max Reitz
2019-10-28 12:14 ` [PULL 57/69] qcow2: Repair snapshot table with too many entries Max Reitz
2019-10-28 12:14 ` [PULL 58/69] qcow2: Fix v3 snapshot table entry compliancy Max Reitz
2019-10-28 12:14 ` [PULL 59/69] iotests: Add peek_file* functions Max Reitz
2019-10-28 12:14 ` [PULL 60/69] iotests: Test qcow2's snapshot table handling Max Reitz
2019-10-28 12:14 ` [PULL 61/69] block: Handle filter truncation like native impl Max Reitz
2019-10-28 12:14 ` [PULL 62/69] block/cor: Drop cor_co_truncate() Max Reitz
2019-10-28 12:14 ` [PULL 63/69] block: Do not truncate file node when formatting Max Reitz
2019-10-28 12:14 ` [PULL 64/69] block: Add @exact parameter to bdrv_co_truncate() Max Reitz
2019-10-28 12:14 ` [PULL 65/69] block: Evaluate @exact in protocol drivers Max Reitz
2019-10-28 12:14 ` [PULL 66/69] block: Let format drivers pass @exact Max Reitz
2019-10-28 12:14 ` [PULL 67/69] block: Pass truncate exact=true where reasonable Max Reitz
2019-10-28 12:15 ` [PULL 68/69] Revert "qemu-img: Check post-truncation size" Max Reitz
2019-10-28 12:15 ` [PULL 69/69] qemu-iotests: restrict 264 to qcow2 only Max Reitz
2019-10-28 21:13 ` [PULL 00/69] Block patches for softfreeze Peter Maydell

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=20191028121501.15279-40-mreitz@redhat.com \
    --to=mreitz@redhat.com \
    --cc=kwolf@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-block@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).