[PATCH V1 0/4] skip memory init during CPR

[PATCH V1 0/4] skip memory init during CPR

[Steve Sistare]

Fix bugs where the realize method re-initializes some memory regions during
CPR.  See the individual commit messages for details.

Steve Sistare (4):
  migration: cpr_is_incoming
  pflash: fix cpr
  hw/loader: fix roms during cpr
  hw/qxl: fix cpr

 hw/block/block.c        |  5 +++++
 hw/core/loader.c        |  5 ++++-
 hw/display/qxl.c        | 27 ++++++++++++++++++++++++---
 include/migration/cpr.h |  1 +
 migration/cpr.c         |  5 +++++
 5 files changed, 39 insertions(+), 4 deletions(-)

-- 
1.8.3.1

[PATCH V1 1/4] migration: cpr_is_incoming

[Steve Sistare]

Define the cpr_is_incoming helper, to be used in several cpr fix patches.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
 include/migration/cpr.h | 1 +
 migration/cpr.c         | 5 +++++
 2 files changed, 6 insertions(+)

diff --git a/include/migration/cpr.h b/include/migration/cpr.h
index 3a6deb7..7561fc7 100644
--- a/include/migration/cpr.h
+++ b/include/migration/cpr.h
@@ -21,6 +21,7 @@ int cpr_find_fd(const char *name, int id);
 
 MigMode cpr_get_incoming_mode(void);
 void cpr_set_incoming_mode(MigMode mode);
+bool cpr_is_incoming(void);
 
 int cpr_state_save(MigrationChannel *channel, Error **errp);
 int cpr_state_load(MigrationChannel *channel, Error **errp);
diff --git a/migration/cpr.c b/migration/cpr.c
index 584b0b9..0e935ab 100644
--- a/migration/cpr.c
+++ b/migration/cpr.c
@@ -128,6 +128,11 @@ void cpr_set_incoming_mode(MigMode mode)
     incoming_mode = mode;
 }
 
+bool cpr_is_incoming(void)
+{
+    return incoming_mode != MIG_MODE_NONE;
+}
+
 int cpr_state_save(MigrationChannel *channel, Error **errp)
 {
     int ret;
-- 
1.8.3.1

Re: [PATCH V1 1/4] migration: cpr_is_incoming

[Peter Xu]

On Fri, Mar 07, 2025 at 12:55:51PM -0800, Steve Sistare wrote:
> Define the cpr_is_incoming helper, to be used in several cpr fix patches.
> 
> Signed-off-by: Steve Sistare <steven.sistare@oracle.com>

Reviewed-by: Peter Xu <peterx@redhat.com>

-- 
Peter Xu

[PATCH V1 2/4] pflash: fix cpr

[Steve Sistare]

During normal migration, new QEMU creates and initializes memory regions,
then loads the preserved contents of the region from vmstate.

During CPR, memory regions are preserved in place, then the realize
method initializes the regions contents, losing the old contents.  To
fix, skip the re-init during CPR.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
 hw/block/block.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/hw/block/block.c b/hw/block/block.c
index 1d405e0..2e10611 100644
--- a/hw/block/block.c
+++ b/hw/block/block.c
@@ -12,6 +12,7 @@
 #include "system/blockdev.h"
 #include "system/block-backend.h"
 #include "hw/block/block.h"
+#include "migration/cpr.h"
 #include "qapi/error.h"
 #include "qapi/qapi-types-block.h"
 
@@ -66,6 +67,10 @@ bool blk_check_size_and_read_all(BlockBackend *blk, DeviceState *dev,
     int ret;
     g_autofree char *dev_id = NULL;
 
+    if (cpr_is_incoming()) {
+        return true;
+    }
+
     blk_len = blk_getlength(blk);
     if (blk_len < 0) {
         error_setg_errno(errp, -blk_len,
-- 
1.8.3.1

[PATCH V1 3/4] hw/loader: fix roms during cpr

[Steve Sistare]

During normal migration, new QEMU creates and initializes memory regions,
then loads the preserved contents of the region from vmstate.

During CPR, memory regions are preserved in place, then the realize
method initializes the regions contents, losing the old contents.  To
fix, skip the re-init during CPR.

Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
 hw/core/loader.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/hw/core/loader.c b/hw/core/loader.c
index fd25c5e..3c3a9a0 100644
--- a/hw/core/loader.c
+++ b/hw/core/loader.c
@@ -51,6 +51,7 @@
 #include "trace.h"
 #include "hw/hw.h"
 #include "disas/disas.h"
+#include "migration/cpr.h"
 #include "migration/vmstate.h"
 #include "monitor/monitor.h"
 #include "system/reset.h"
@@ -1029,7 +1030,9 @@ static void *rom_set_mr(Rom *rom, Object *owner, const char *name, bool ro)
     vmstate_register_ram_global(rom->mr);
 
     data = memory_region_get_ram_ptr(rom->mr);
-    memcpy(data, rom->data, rom->datasize);
+    if (!cpr_is_incoming()) {
+        memcpy(data, rom->data, rom->datasize);
+    }
 
     return data;
 }
-- 
1.8.3.1

[PATCH V1 4/4] hw/qxl: fix cpr

[Steve Sistare]

During normal migration, new QEMU creates and initializes memory regions,
then loads the preserved contents of the region from vmstate.

During CPR, memory regions are preserved in place, then the realize
method initializes the regions contents, losing the old contents.  To
fix, skip writes to the qxl memory regions during CPR load.

Reported-by: andrey.drobyshev@virtuozzo.com
Tested-by: andrey.drobyshev@virtuozzo.com
Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
---
 hw/display/qxl.c | 27 ++++++++++++++++++++++++---
 1 file changed, 24 insertions(+), 3 deletions(-)

diff --git a/hw/display/qxl.c b/hw/display/qxl.c
index 2efdc77..da14da5 100644
--- a/hw/display/qxl.c
+++ b/hw/display/qxl.c
@@ -30,6 +30,7 @@
 #include "qemu/module.h"
 #include "hw/qdev-properties.h"
 #include "system/runstate.h"
+#include "migration/cpr.h"
 #include "migration/vmstate.h"
 #include "trace.h"
 
@@ -333,6 +334,10 @@ static void init_qxl_rom(PCIQXLDevice *d)
     uint32_t fb;
     int i, n;
 
+    if (cpr_is_incoming()) {
+        goto skip_init;
+    }
+
     memset(rom, 0, d->rom_size);
 
     rom->magic         = cpu_to_le32(QXL_ROM_MAGIC);
@@ -390,6 +395,7 @@ static void init_qxl_rom(PCIQXLDevice *d)
             sizeof(rom->client_monitors_config));
     }
 
+skip_init:
     d->shadow_rom = *rom;
     d->rom        = rom;
     d->modes      = modes;
@@ -403,6 +409,9 @@ static void init_qxl_ram(PCIQXLDevice *d)
 
     buf = d->vga.vram_ptr;
     d->ram = (QXLRam *)(buf + le32_to_cpu(d->shadow_rom.ram_header_offset));
+    if (cpr_is_incoming()) {
+        return;
+    }
     d->ram->magic       = cpu_to_le32(QXL_RAM_MAGIC);
     d->ram->int_pending = cpu_to_le32(0);
     d->ram->int_mask    = cpu_to_le32(0);
@@ -539,6 +548,10 @@ static void interface_set_compression_level(QXLInstance *sin, int level)
 
     trace_qxl_interface_set_compression_level(qxl->id, level);
     qxl->shadow_rom.compression_level = cpu_to_le32(level);
+    if (cpr_is_incoming()) {
+        assert(qxl->rom->compression_level == cpu_to_le32(level));
+        return;
+    }
     qxl->rom->compression_level = cpu_to_le32(level);
     qxl_rom_set_dirty(qxl);
 }
@@ -997,7 +1010,8 @@ static void interface_set_client_capabilities(QXLInstance *sin,
     }
 
     if (runstate_check(RUN_STATE_INMIGRATE) ||
-        runstate_check(RUN_STATE_POSTMIGRATE)) {
+        runstate_check(RUN_STATE_POSTMIGRATE) ||
+        cpr_is_incoming()) {
         return;
     }
 
@@ -1200,6 +1214,10 @@ static void qxl_reset_state(PCIQXLDevice *d)
 {
     QXLRom *rom = d->rom;
 
+    if (cpr_is_incoming()) {
+        return;
+    }
+
     qxl_check_state(d);
     d->shadow_rom.update_id = cpu_to_le32(0);
     *rom = d->shadow_rom;
@@ -1370,8 +1388,11 @@ static int qxl_add_memslot(PCIQXLDevice *d, uint32_t slot_id, uint64_t delta,
     memslot.virt_start = virt_start + (guest_start - pci_start);
     memslot.virt_end   = virt_start + (guest_end   - pci_start);
     memslot.addr_delta = memslot.virt_start - delta;
-    memslot.generation = d->rom->slot_generation = 0;
-    qxl_rom_set_dirty(d);
+    if (!cpr_is_incoming()) {
+        d->rom->slot_generation = 0;
+        qxl_rom_set_dirty(d);
+    }
+    memslot.generation = d->rom->slot_generation;
 
     qemu_spice_add_memslot(&d->ssd, &memslot, async);
     d->guest_slots[slot_id].mr = mr;
-- 
1.8.3.1

Re: [PATCH V1 0/4] skip memory init during CPR

[Fabiano Rosas]

Steve Sistare <steven.sistare@oracle.com> writes:

> Fix bugs where the realize method re-initializes some memory regions during
> CPR.  See the individual commit messages for details.
>
> Steve Sistare (4):
>   migration: cpr_is_incoming
>   pflash: fix cpr
>   hw/loader: fix roms during cpr
>   hw/qxl: fix cpr
>
>  hw/block/block.c        |  5 +++++
>  hw/core/loader.c        |  5 ++++-
>  hw/display/qxl.c        | 27 ++++++++++++++++++++++++---
>  include/migration/cpr.h |  1 +
>  migration/cpr.c         |  5 +++++
>  5 files changed, 39 insertions(+), 4 deletions(-)

Series:

Reviewed-by: Fabiano Rosas <farosas@suse.de>

Re: [PATCH V1 0/4] skip memory init during CPR

[Fabiano Rosas]

Steve Sistare <steven.sistare@oracle.com> writes:

> Fix bugs where the realize method re-initializes some memory regions during
> CPR.  See the individual commit messages for details.
>
> Steve Sistare (4):
>   migration: cpr_is_incoming
>   pflash: fix cpr
>   hw/loader: fix roms during cpr
>   hw/qxl: fix cpr
>
>  hw/block/block.c        |  5 +++++
>  hw/core/loader.c        |  5 ++++-
>  hw/display/qxl.c        | 27 ++++++++++++++++++++++++---
>  include/migration/cpr.h |  1 +
>  migration/cpr.c         |  5 +++++
>  5 files changed, 39 insertions(+), 4 deletions(-)

I'm queing this for the freeze to ensure people using CPR with qxl can
get a complete feature in 10.0. If anyone has concerns, please speak up.