From: Steve Sistare <steven.sistare@oracle.com>
To: qemu-devel@nongnu.org
Cc: "Paolo Bonzini" <pbonzini@redhat.com>,
"Stefan Hajnoczi" <stefanha@redhat.com>,
"Marc-André Lureau" <marcandre.lureau@redhat.com>,
"Alex Bennée" <alex.bennee@linaro.org>,
"Dr. David Alan Gilbert" <dgilbert@redhat.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
"Marcel Apfelbaum" <marcel.apfelbaum@gmail.com>,
"Alex Williamson" <alex.williamson@redhat.com>,
"Daniel P. Berrange" <berrange@redhat.com>,
"Juan Quintela" <quintela@redhat.com>,
"Markus Armbruster" <armbru@redhat.com>,
"Eric Blake" <eblake@redhat.com>,
"Jason Zeng" <jason.zeng@linux.intel.com>,
"Zheng Chuan" <zhengchuan@huawei.com>,
"Steve Sistare" <steven.sistare@oracle.com>,
"Mark Kanda" <mark.kanda@oracle.com>,
"Guoyi Tu" <tugy@chinatelecom.cn>,
"Peter Maydell" <peter.maydell@linaro.org>,
"Philippe Mathieu-Daudé" <philippe.mathieu.daude@gmail.com>,
"Igor Mammedov" <imammedo@redhat.com>,
"David Hildenbrand" <david@redhat.com>,
"John Snow" <jsnow@redhat.com>, "Peng Liang" <tcx4c70@gmail.com>
Subject: [PATCH V9 24/46] cpr: ram block blockers
Date: Tue, 26 Jul 2022 09:10:21 -0700 [thread overview]
Message-ID: <1658851843-236870-25-git-send-email-steven.sistare@oracle.com> (raw)
In-Reply-To: <1658851843-236870-1-git-send-email-steven.sistare@oracle.com>
Unlike cpr-reboot mode, cpr-exec mode cannot save volatile ram blocks in the
migration stream file and recreate them later, because the physical memory for
the blocks is pinned and registered for vfio. Add an exec-mode blocker for
volatile ram blocks.
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
include/exec/memory.h | 2 ++
include/exec/ramblock.h | 1 +
softmmu/physmem.c | 45 +++++++++++++++++++++++++++++++++++++++++++++
softmmu/vl.c | 2 ++
4 files changed, 50 insertions(+)
diff --git a/include/exec/memory.h b/include/exec/memory.h
index 9629289..128bf3e 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -3039,6 +3039,8 @@ bool ram_block_discard_is_disabled(void);
*/
bool ram_block_discard_is_required(void);
+void ram_block_add_cpr_blockers(Error **errp);
+
#endif
#endif
diff --git a/include/exec/ramblock.h b/include/exec/ramblock.h
index 6cbedf9..a5cbd9e 100644
--- a/include/exec/ramblock.h
+++ b/include/exec/ramblock.h
@@ -39,6 +39,7 @@ struct RAMBlock {
/* RCU-enabled, writes protected by the ramlist lock */
QLIST_ENTRY(RAMBlock) next;
QLIST_HEAD(, RAMBlockNotifier) ramblock_notifiers;
+ Error *cpr_blocker;
int fd;
size_t page_size;
/* dirty bitmap used during migration */
diff --git a/softmmu/physmem.c b/softmmu/physmem.c
index 29baa0f..ac0ae25 100644
--- a/softmmu/physmem.c
+++ b/softmmu/physmem.c
@@ -67,6 +67,7 @@
#include "qemu/pmem.h"
+#include "migration/blocker.h"
#include "migration/cpr-state.h"
#include "migration/misc.h"
#include "migration/vmstate.h"
@@ -1979,6 +1980,49 @@ static bool memory_region_is_backend(MemoryRegion *mr)
return !!object_dynamic_cast(OBJECT(mr)->parent, TYPE_MEMORY_BACKEND);
}
+/*
+ * Return true if ram contents would be lost during cpr for MIG_MODE_CPR_EXEC.
+ * Return false for ram_device because it is remapped after exec. Do not
+ * exclude rom, even though it is readonly, because the rom file could change
+ * in the new qemu. Return false for non-migratable blocks. They are either
+ * re-created after exec, or are handled specially, or are covered by a
+ * device-level cpr blocker. Return false for an fd, because it is visible and
+ * can be remapped in the new process.
+ */
+static bool ram_is_volatile(RAMBlock *rb)
+{
+ MemoryRegion *mr = rb->mr;
+
+ return mr &&
+ memory_region_is_ram(mr) &&
+ !memory_region_is_ram_device(mr) &&
+ (!qemu_ram_is_shared(rb) || !ramblock_is_named_file(rb)) &&
+ qemu_ram_is_migratable(rb) &&
+ rb->fd < 0;
+}
+
+/*
+ * Add a MIG_MODE_CPR_EXEC blocker for each volatile ram block. This cannot be
+ * performed in ram_block_add because the migratable flag has not been set yet.
+ * No need to examine anonymous (non-backend) blocks, because they are
+ * created using memfd if cpr-exec mode is enabled.
+ */
+void ram_block_add_cpr_blockers(Error **errp)
+{
+ RAMBlock *rb;
+
+ RAMBLOCK_FOREACH(rb) {
+ if (ram_is_volatile(rb) && memory_region_is_backend(rb->mr)) {
+ const char *name = memory_region_name(rb->mr);
+ rb->cpr_blocker = NULL;
+ error_setg(&rb->cpr_blocker,
+ "Memory region %s is volatile. A memory-backend-memfd or"
+ " memory-backend-file with share=on is required.", name);
+ migrate_add_blockers(&rb->cpr_blocker, errp, MIG_MODE_CPR_EXEC, -1);
+ }
+ }
+}
+
static void *qemu_anon_memfd_alloc(RAMBlock *rb, size_t maxlen, Error **errp)
{
size_t len, align;
@@ -2285,6 +2329,7 @@ void qemu_ram_free(RAMBlock *block)
qemu_mutex_lock_ramlist();
cpr_delete_memfd(memory_region_name(block->mr));
+ migrate_del_blocker(&block->cpr_blocker);
QLIST_REMOVE_RCU(block, next);
ram_list.mru_block = NULL;
/* Write list before version */
diff --git a/softmmu/vl.c b/softmmu/vl.c
index 83f3be0..5f6cd8c 100644
--- a/softmmu/vl.c
+++ b/softmmu/vl.c
@@ -28,6 +28,7 @@
#include "qemu/units.h"
#include "exec/cpu-common.h"
#include "exec/page-vary.h"
+#include "exec/memory.h"
#include "hw/qdev-properties.h"
#include "qapi/compat-policy.h"
#include "qapi/error.h"
@@ -2589,6 +2590,7 @@ void qmp_x_exit_preconfig(Error **errp)
qemu_init_board();
qemu_create_cli_devices();
qemu_machine_creation_done();
+ ram_block_add_cpr_blockers(&error_fatal);
if (loadvm) {
load_snapshot(loadvm, NULL, false, NULL, &error_fatal);
--
1.8.3.1
next prev parent reply other threads:[~2022-07-26 16:32 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-07-26 16:09 [PATCH V9 00/46] Live Update Steve Sistare
2022-07-26 16:09 ` [PATCH V9 01/46] migration: fix populate_vfio_info Steve Sistare
2022-07-26 16:09 ` [PATCH V9 02/46] memory: RAM_NAMED_FILE flag Steve Sistare
2022-07-26 16:10 ` [PATCH V9 03/46] migration: file URI Steve Sistare
2022-07-26 16:10 ` [PATCH V9 04/46] migration: mode parameter Steve Sistare
2022-07-26 16:10 ` [PATCH V9 05/46] migration: migrate-enable-mode option Steve Sistare
2022-07-26 16:10 ` [PATCH V9 06/46] migration: simplify blockers Steve Sistare
2022-07-26 16:10 ` [PATCH V9 07/46] migration: per-mode blockers Steve Sistare
2022-07-26 16:10 ` [PATCH V9 08/46] cpr: relax some blockers Steve Sistare
2022-07-26 16:10 ` [PATCH V9 09/46] cpr: reboot mode Steve Sistare
2022-07-26 16:10 ` [PATCH V9 10/46] qdev-properties: strList Steve Sistare
2023-06-08 14:50 ` Steven Sistare
2022-07-26 16:10 ` [PATCH V9 11/46] qapi: strList_from_string Steve Sistare
2022-07-26 16:10 ` [PATCH V9 12/46] qapi: QAPI_LIST_LENGTH Steve Sistare
2022-07-26 16:10 ` [PATCH V9 13/46] qapi: strv_from_strList Steve Sistare
2022-07-26 16:10 ` [PATCH V9 14/46] qapi: strList unit tests Steve Sistare
2022-07-26 16:10 ` [PATCH V9 15/46] migration: cpr-exec-args parameter Steve Sistare
2022-07-26 16:10 ` [PATCH V9 16/46] migration: simplify notifiers Steve Sistare
2022-07-26 16:10 ` [PATCH V9 17/46] migration: check mode in notifiers Steve Sistare
2022-07-26 16:10 ` [PATCH V9 18/46] memory: flat section iterator Steve Sistare
2022-07-26 16:10 ` [PATCH V9 19/46] oslib: qemu_clear_cloexec Steve Sistare
2022-07-26 16:10 ` [PATCH V9 20/46] vl: helper to request re-exec Steve Sistare
2022-07-26 16:10 ` [PATCH V9 21/46] cpr: preserve extra state Steve Sistare
2022-07-26 16:10 ` [PATCH V9 22/46] cpr: exec mode Steve Sistare
2022-07-26 16:10 ` [PATCH V9 23/46] cpr: add exec-mode blockers Steve Sistare
2022-07-26 16:10 ` Steve Sistare [this message]
2022-07-26 16:10 ` [PATCH V9 25/46] cpr: only-cpr-capable Steve Sistare
2022-07-26 16:10 ` [PATCH V9 26/46] cpr: Mismatched GPAs fix Steve Sistare
2022-07-26 16:10 ` [PATCH V9 27/46] hostmem-memfd: cpr support Steve Sistare
2022-07-26 16:10 ` [PATCH V9 28/46] hostmem-epc: " Steve Sistare
2022-07-26 16:10 ` [PATCH V9 29/46] pci: export msix_is_pending Steve Sistare
2022-07-26 16:10 ` [PATCH V9 30/46] vfio-pci: refactor for cpr Steve Sistare
2022-07-26 16:10 ` [PATCH V9 31/46] vfio-pci: cpr part 1 (fd and dma) Steve Sistare
2022-07-26 16:10 ` [PATCH V9 32/46] vfio-pci: cpr part 2 (msi) Steve Sistare
2023-07-05 8:56 ` Kunkun Jiang via
2023-07-10 15:43 ` Steven Sistare
2023-07-13 12:35 ` Kunkun Jiang via
2023-07-13 12:42 ` Marc Zyngier
2022-07-26 16:10 ` [PATCH V9 33/46] vfio-pci: cpr part 3 (intx) Steve Sistare
2022-07-26 16:10 ` [PATCH V9 34/46] vfio-pci: recover from unmap-all-vaddr failure Steve Sistare
2022-07-26 16:10 ` [PATCH V9 35/46] vhost: reset vhost devices for cpr Steve Sistare
2022-07-26 16:10 ` [PATCH V9 36/46] chardev: cpr framework Steve Sistare
2022-07-26 16:10 ` [PATCH V9 37/46] chardev: cpr for simple devices Steve Sistare
2022-07-26 16:10 ` [PATCH V9 38/46] chardev: cpr for pty Steve Sistare
2022-07-26 16:10 ` [PATCH V9 39/46] chardev: cpr for sockets Steve Sistare
2022-07-26 16:10 ` [PATCH V9 40/46] python/machine: QEMUMachine full_args Steve Sistare
2022-07-26 18:00 ` John Snow
2022-07-26 16:10 ` [PATCH V9 41/46] python/machine: QEMUMachine reopen_qmp_connection Steve Sistare
2022-07-26 18:04 ` John Snow
2022-07-27 12:06 ` Steven Sistare
2022-07-26 16:10 ` [PATCH V9 42/46] tests/avocado: add cpr regression test Steve Sistare
2023-12-01 10:44 ` Philippe Mathieu-Daudé
2022-07-26 16:10 ` [PATCH V9 43/46] vl: start on wakeup request Steve Sistare
2022-07-26 16:10 ` [PATCH V9 44/46] migration: fix suspended runstate Steve Sistare
2022-07-26 16:10 ` [PATCH V9 45/46] migration: notifier error reporting Steve Sistare
2022-07-26 16:10 ` [PATCH V9 46/46] vfio: allow cpr-reboot migration if suspended Steve Sistare
2022-12-07 15:48 ` [PATCH V9 00/46] Live Update Steven Sistare
2023-02-07 18:44 ` Steven Sistare
2023-02-07 19:01 ` Steven Sistare
2023-05-30 13:38 ` Philippe Mathieu-Daudé
2023-05-30 13:53 ` Steven Sistare
2023-04-07 17:35 ` Michael Galaxy
2023-04-14 19:20 ` Michael Galaxy
2023-06-06 22:15 ` Michael Galaxy
2023-06-07 15:55 ` Michael Galaxy
2023-06-07 17:37 ` Steven Sistare
2023-06-12 14:59 ` Michael Galaxy
2023-07-10 15:10 ` Steven Sistare
2023-07-13 15:53 ` Michael Galaxy
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=1658851843-236870-25-git-send-email-steven.sistare@oracle.com \
--to=steven.sistare@oracle.com \
--cc=alex.bennee@linaro.org \
--cc=alex.williamson@redhat.com \
--cc=armbru@redhat.com \
--cc=berrange@redhat.com \
--cc=david@redhat.com \
--cc=dgilbert@redhat.com \
--cc=eblake@redhat.com \
--cc=imammedo@redhat.com \
--cc=jason.zeng@linux.intel.com \
--cc=jsnow@redhat.com \
--cc=marcandre.lureau@redhat.com \
--cc=marcel.apfelbaum@gmail.com \
--cc=mark.kanda@oracle.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=philippe.mathieu.daude@gmail.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=stefanha@redhat.com \
--cc=tcx4c70@gmail.com \
--cc=tugy@chinatelecom.cn \
--cc=zhengchuan@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).