From: Igor Mammedov <imammedo@redhat.com>
To: qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org, mst@redhat.com, dgilbert@redhat.com,
qemu-stable@nongnu.org, amit.shah@redhat.com,
pbonzini@redhat.com, lersek@redhat.com
Subject: [Qemu-devel] [PATCH for-2.1 v2 1/2] migration: load smaller RAMBlock to a bigger one if permitted
Date: Mon, 28 Jul 2014 08:03:24 +0000 [thread overview]
Message-ID: <1406534605-18911-2-git-send-email-imammedo@redhat.com> (raw)
In-Reply-To: <1406534605-18911-1-git-send-email-imammedo@redhat.com>
Add API to mark memory region as extend-able on migration,
to allow migration code to load smaller RAMBlock into
a bigger one on destination QEMU instance.
This will allow to fix broken migration from QEMU 1.7/2.0 to
QEMU 2.1 due to ACPI tables size changes across 1.7/2.0/2.1
versions by marking ACPI tables ROM blob as extend-able.
So that smaller tables from previous version could be always
migrated to a bigger rom blob on new version.
Credits-for-idea: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
---
v2:
fixed patch as suggested by Laszlo
---
arch_init.c | 22 +++++++++++++++++-----
exec.c | 8 ++++++++
include/exec/memory.h | 11 +++++++++++
include/exec/ram_addr.h | 3 +++
memory.c | 5 +++++
5 files changed, 44 insertions(+), 5 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 8ddaf35..2c0c238 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -1071,11 +1071,23 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
QTAILQ_FOREACH(block, &ram_list.blocks, next) {
if (!strncmp(id, block->idstr, sizeof(id))) {
- if (block->length != length) {
- error_report("Length mismatch: %s: " RAM_ADDR_FMT
- " in != " RAM_ADDR_FMT, id, length,
- block->length);
- ret = -EINVAL;
+ if (block->flags & RAM_EXTENDABLE_ON_MIGRATE) {
+ if (block->length < length) {
+ error_report("Length too big: %s: " RAM_ADDR_FMT
+ " in > " RAM_ADDR_FMT, id, length,
+ block->length);
+ ret = -EINVAL;
+ } else {
+ memset(block->host, 0, block->length);
+ }
+ } else {
+ if (block->length != length) {
+ error_report("Length mismatch: %s: "
+ RAM_ADDR_FMT " in != "
+ RAM_ADDR_FMT,
+ id, length, block->length);
+ ret = -EINVAL;
+ }
}
break;
}
diff --git a/exec.c b/exec.c
index 765bd94..02536f8e 100644
--- a/exec.c
+++ b/exec.c
@@ -1214,6 +1214,14 @@ void qemu_ram_unset_idstr(ram_addr_t addr)
}
}
+void qemu_ram_set_extendable_on_migration(ram_addr_t addr)
+{
+ RAMBlock *block = find_ram_block(addr);
+
+ assert(block != NULL);
+ block->flags |= RAM_EXTENDABLE_ON_MIGRATE;
+}
+
static int memory_try_enable_merging(void *addr, size_t len)
{
if (!qemu_opt_get_bool(qemu_get_machine_opts(), "mem-merge", true)) {
diff --git a/include/exec/memory.h b/include/exec/memory.h
index e2c8e3e..f96ddbb 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -894,6 +894,17 @@ bool memory_region_present(MemoryRegion *container, hwaddr addr);
bool memory_region_is_mapped(MemoryRegion *mr);
/**
+ * memory_region_permit_extendable_migration: marks #MemoryRegion
+ * as extendable on migration, allowing the migration code to load
+ * source memory block of smaller size than destination memory block
+ * at migration time
+ *
+ * @mr: a #MemoryRegion whose #RAMBlock should be marked as
+ * extendable on migration
+ */
+void memory_region_permit_extendable_migration(MemoryRegion *mr);
+
+/**
* memory_region_find: translate an address/size relative to a
* MemoryRegion into a #MemoryRegionSection.
*
diff --git a/include/exec/ram_addr.h b/include/exec/ram_addr.h
index 6593be1..7a6b782 100644
--- a/include/exec/ram_addr.h
+++ b/include/exec/ram_addr.h
@@ -34,6 +34,9 @@ void *qemu_get_ram_ptr(ram_addr_t addr);
void qemu_ram_free(ram_addr_t addr);
void qemu_ram_free_from_ptr(ram_addr_t addr);
+#define RAM_EXTENDABLE_ON_MIGRATE (1U << 31)
+void qemu_ram_set_extendable_on_migration(ram_addr_t addr);
+
static inline bool cpu_physical_memory_get_dirty(ram_addr_t start,
ram_addr_t length,
unsigned client)
diff --git a/memory.c b/memory.c
index 64d7176..744c746 100644
--- a/memory.c
+++ b/memory.c
@@ -1791,6 +1791,11 @@ bool memory_region_is_mapped(MemoryRegion *mr)
return mr->container ? true : false;
}
+void memory_region_permit_extendable_migration(MemoryRegion *mr)
+{
+ qemu_ram_set_extendable_on_migration(mr->ram_addr);
+}
+
MemoryRegionSection memory_region_find(MemoryRegion *mr,
hwaddr addr, uint64_t size)
{
--
1.8.3.1
next prev parent reply other threads:[~2014-07-28 8:04 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-28 8:03 [Qemu-devel] [PATCH for-2.1 v2 0/2] Fix migration failure due to ACPI tables size changes Igor Mammedov
2014-07-28 8:03 ` Igor Mammedov [this message]
2014-07-28 8:19 ` [Qemu-devel] [PATCH for-2.1 v2 1/2] migration: load smaller RAMBlock to a bigger one if permitted Laszlo Ersek
2014-07-28 8:03 ` [Qemu-devel] [PATCH for-2.1 v2 2/2] acpi: mark ACPI tables ROM blob as extend-able on migration Igor Mammedov
2014-07-28 8:44 ` Igor Mammedov
2014-07-28 9:02 ` [Qemu-devel] [PATCH for-2.1 v3 " Igor Mammedov
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=1406534605-18911-2-git-send-email-imammedo@redhat.com \
--to=imammedo@redhat.com \
--cc=amit.shah@redhat.com \
--cc=dgilbert@redhat.com \
--cc=lersek@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=qemu-stable@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).