Skip to contentSkip to conversation

docs: Convert migration.txt to rst

PATCH Superseded Dr. David Alan Gilbert (git) 3 participants 8 messages Dec 12, 2017 – Dec 14, 2017

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 docs: Convert migration.txt to rst

From: "Dr. David Alan Gilbert" <[email protected]>

Mostly just manual conversion with very minor fixes.

---
 docs/devel/{migration.txt => migration.rst} | 326 +++++++++++++++-------------
 1 file changed, 176 insertions(+), 150 deletions(-)
 rename docs/devel/{migration.txt => migration.rst} (74%)
docs/devel/migration.rst renamed from docs/devel/migration.txt +176 -150
similarity index 74%
rename from docs/devel/migration.txt
rename to docs/devel/migration.rst
index 4030703726..7d64298cd7 100644
@@ -1,4 +1,6 @@
-= Migration =
+=========
+Migration
+=========
 
 QEMU has code to load/save the state of the guest that it is running.
 These are two complementary operations.  Saving the state just does
@@ -26,7 +28,8 @@ the guest to be stopped.  Typically the time that the guest is
 unresponsive during live migration is the low hundred of milliseconds
 (notice that this depends on a lot of things).
 
-=== Types of migration ===
+Types of migration
+==================
 
 Now that we have talked about live migration, there are several ways
 to do migration:
@@ -41,21 +44,25 @@ All these four migration protocols use the same infrastructure to
 save/restore state devices.  This infrastructure is shared with the
 savevm/loadvm functionality.
 
-=== State Live Migration ===
+State Live Migration
+====================
 
 This is used for RAM and block devices.  It is not yet ported to vmstate.
 <Fill more information here>
 
-=== What is the common infrastructure ===
+Common infrastructure
+=====================
 
 QEMU uses a QEMUFile abstraction to be able to do migration.  Any type
 of migration that wants to use QEMU infrastructure has to create a
 QEMUFile with:
 
-QEMUFile *qemu_fopen_ops(void *opaque,
-                         QEMUFilePutBufferFunc *put_buffer,
-                         QEMUFileGetBufferFunc *get_buffer,
-                         QEMUFileCloseFunc *close);
+.. code:: c
+
+  QEMUFile *qemu_fopen_ops(void *opaque,
+                           QEMUFilePutBufferFunc *put_buffer,
+                           QEMUFileGetBufferFunc *get_buffer,
+                           QEMUFileCloseFunc *close);
 
 The functions have the following functionality:
 
@@ -63,19 +70,25 @@ This function writes a chunk of data to a file at the given position.
 The pos argument can be ignored if the file is only used for
 streaming.  The handler should try to write all of the data it can.
 
-typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
+                                      int64_t pos, int size);
 
 Read a chunk of data from a file at the given position.  The pos argument
 can be ignored if the file is only be used for streaming.  The number of
 bytes actually read should be returned.
 
-typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
+                                      int64_t pos, int size);
 
 Close a file and return an error code.
 
-typedef int (QEMUFileCloseFunc)(void *opaque);
+.. code:: c
+
+  typedef int (QEMUFileCloseFunc)(void *opaque);
 
 You can use any internal state that you need using the opaque void *
 pointer that is passed to all functions.
@@ -83,7 +96,8 @@ pointer that is passed to all functions.
 The important functions for us are put_buffer()/get_buffer() that
 allow to write/read a buffer into the QEMUFile.
 
-=== How to save the state of one device ===
+Saving the state of one device
+==============================
 
 The state of a device is saved using intermediate buffers.  There are
 some helper functions to assist this saving.
@@ -97,30 +111,34 @@ associated with a series of fields saved.  The save_state always saves
 the state as the newer version.  But load_state sometimes is able to
 load state from an older version.
 
-=== Legacy way ===
+Legacy way
+----------
 
 This way is going to disappear as soon as all current users are ported to VMSTATE.
 
 Each device has to register two functions, one to save the state and
 another to load the state back.
 
-int register_savevm(DeviceState *dev,
-                    const char *idstr,
-                    int instance_id,
-                    int version_id,
-                    SaveStateHandler *save_state,
-                    LoadStateHandler *load_state,
-                    void *opaque);
+.. code:: c
+
+  int register_savevm(DeviceState *dev,
+                      const char *idstr,
+                      int instance_id,
+                      int version_id,
+                      SaveStateHandler *save_state,
+                      LoadStateHandler *load_state,
+                      void *opaque);
 
-typedef void SaveStateHandler(QEMUFile *f, void *opaque);
-typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
+  typedef void SaveStateHandler(QEMUFile *f, void *opaque);
+  typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
 
 The important functions for the device state format are the save_state
 and load_state.  Notice that load_state receives a version_id
 parameter to know what state format is receiving.  save_state doesn't
 have a version_id parameter because it always uses the latest version.
 
-=== VMState ===
+VMState
+-------
 
 The legacy way of saving/loading state of the device had the problem
 that we have to maintain two functions in sync.  If we did one change
@@ -135,23 +153,27 @@ save/load functions.
 
 An example (from hw/input/pckbd.c)
 
-static const VMStateDescription vmstate_kbd = {
-    .name = "pckbd",
-    .version_id = 3,
-    .minimum_version_id = 3,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT8(write_cmd, KBDState),
-        VMSTATE_UINT8(status, KBDState),
-        VMSTATE_UINT8(mode, KBDState),
-        VMSTATE_UINT8(pending, KBDState),
-        VMSTATE_END_OF_LIST()
-    }
-};
+.. code:: c
+
+  static const VMStateDescription vmstate_kbd = {
+      .name = "pckbd",
+      .version_id = 3,
+      .minimum_version_id = 3,
+      .fields = (VMStateField[]) {
+          VMSTATE_UINT8(write_cmd, KBDState),
+          VMSTATE_UINT8(status, KBDState),
+          VMSTATE_UINT8(mode, KBDState),
+          VMSTATE_UINT8(pending, KBDState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
 
 We are declaring the state with name "pckbd".
 The version_id is 3, and the fields are 4 uint8_t in a KBDState structure.
 We registered this with:
 
+.. code:: c
+
     vmstate_register(NULL, 0, &vmstate_kbd, s);
 
 Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
@@ -159,7 +181,8 @@ Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
 You can search for VMSTATE_* macros for lots of types used in QEMU in
 include/hw/hw.h.
 
-=== More about versions ===
+More about versions
+-------------------
 
 Version numbers are intended for major incompatible changes to the
 migration of a device, and using them breaks backwards-migration
@@ -183,7 +206,8 @@ function is deprecated and will be removed when no more users are left.
 Saving state will always create a section with the 'version_id' value
 and thus can't be loaded by any older QEMU.
 
-===  Massaging functions ===
+Massaging functions
+-------------------
 
 Sometimes, it is not enough to be able to save the state directly
 from one structure, we need to fill the correct values there.  One
@@ -194,20 +218,19 @@ load the state for the cpu that we have just loaded from the QEMUFile.
 
 The functions to do that are inside a vmstate definition, and are called:
 
-- int (*pre_load)(void *opaque);
+- ``int (*pre_load)(void *opaque);``
 
   This function is called before we load the state of one device.
 
-- int (*post_load)(void *opaque, int version_id);
+- ``int (*post_load)(void *opaque, int version_id);``
 
   This function is called after we load the state of one device.
 
-- int (*pre_save)(void *opaque);
+- ``int (*pre_save)(void *opaque);``
 
   This function is called before we save the state of one device.
 
-Example: You can look at hpet.c, that uses the three function to
-         massage the state that is transferred.
+Example: You can look at hpet.c, that uses the three function to massage the state that is transferred.
 
 If you use memory API functions that update memory layout outside
 initialization (i.e., in response to a guest action), this is a strong
@@ -221,7 +244,8 @@ Examples of such memory API functions are:
   - memory_region_set_address()
   - memory_region_set_alias_offset()
 
-=== Subsections ===
+Subsections
+-----------
 
 The use of version_id allows to be able to migrate from older versions
 to newer versions of a device.  But not the other way around.  This
@@ -251,48 +275,50 @@ value that it uses.
 
 Example:
 
-static bool ide_drive_pio_state_needed(void *opaque)
-{
-    IDEState *s = opaque;
-
-    return ((s->status & DRQ_STAT) != 0)
-        || (s->bus->error_status & BM_STATUS_PIO_RETRY);
-}
-
-const VMStateDescription vmstate_ide_drive_pio_state = {
-    .name = "ide_drive/pio_state",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .pre_save = ide_drive_pio_pre_save,
-    .post_load = ide_drive_pio_post_load,
-    .needed = ide_drive_pio_state_needed,
-    .fields = (VMStateField[]) {
-        VMSTATE_INT32(req_nb_sectors, IDEState),
-        VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
-                             vmstate_info_uint8, uint8_t),
-        VMSTATE_INT32(cur_io_buffer_offset, IDEState),
-        VMSTATE_INT32(cur_io_buffer_len, IDEState),
-        VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
-        VMSTATE_INT32(elementary_transfer_size, IDEState),
-        VMSTATE_INT32(packet_transfer_size, IDEState),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
-const VMStateDescription vmstate_ide_drive = {
-    .name = "ide_drive",
-    .version_id = 3,
-    .minimum_version_id = 0,
-    .post_load = ide_drive_post_load,
-    .fields = (VMStateField[]) {
-        .... several fields ....
-        VMSTATE_END_OF_LIST()
-    },
-    .subsections = (const VMStateDescription*[]) {
-        &vmstate_ide_drive_pio_state,
-        NULL
-    }
-};
+.. code:: c
+
+  static bool ide_drive_pio_state_needed(void *opaque)
+  {
+      IDEState *s = opaque;
+
+      return ((s->status & DRQ_STAT) != 0)
+          || (s->bus->error_status & BM_STATUS_PIO_RETRY);
+  }
+
+  const VMStateDescription vmstate_ide_drive_pio_state = {
+      .name = "ide_drive/pio_state",
+      .version_id = 1,
+      .minimum_version_id = 1,
+      .pre_save = ide_drive_pio_pre_save,
+      .post_load = ide_drive_pio_post_load,
+      .needed = ide_drive_pio_state_needed,
+      .fields = (VMStateField[]) {
+          VMSTATE_INT32(req_nb_sectors, IDEState),
+          VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
+                               vmstate_info_uint8, uint8_t),
+          VMSTATE_INT32(cur_io_buffer_offset, IDEState),
+          VMSTATE_INT32(cur_io_buffer_len, IDEState),
+          VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
+          VMSTATE_INT32(elementary_transfer_size, IDEState),
+          VMSTATE_INT32(packet_transfer_size, IDEState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
+
+  const VMStateDescription vmstate_ide_drive = {
+      .name = "ide_drive",
+      .version_id = 3,
+      .minimum_version_id = 0,
+      .post_load = ide_drive_post_load,
+      .fields = (VMStateField[]) {
+          .... several fields ....
+          VMSTATE_END_OF_LIST()
+      },
+      .subsections = (const VMStateDescription*[]) {
+          &vmstate_ide_drive_pio_state,
+          NULL
+      }
+  };
 
 Here we have a subsection for the pio state.  We only need to
 save/send this state when we are in the middle of a pio operation
@@ -305,14 +331,11 @@ to send a subsection allows backwards migration compatibility when
 new subsections are added.
 
 For example;
-   a) Add a new property using DEFINE_PROP_BOOL - e.g. support-foo and
-      default it to true.
-   b) Add an entry to the HW_COMPAT_ for the previous version
-      that sets the property to false.
+   a) Add a new property using ``DEFINE_PROP_BOOL`` - e.g. support-foo and default it to true.
+   b) Add an entry to the ``HW_COMPAT_`` for the previous version that sets the property to false.
    c) Add a static bool  support_foo function that tests the property.
    d) Add a subsection with a .needed set to the support_foo function
-   e) (potentially) Add a pre_load that sets up a default value for 'foo'
-      to be used if the subsection isn't loaded.
+   e) (potentially) Add a pre_load that sets up a default value for 'foo' to be used if the subsection isn't loaded.
 
 Now that subsection will not be generated when using an older
 machine type and the migration stream will be accepted by older
@@ -332,25 +355,28 @@ in most cases.  In general the preference is to tie the subsection to
 the machine type, and allow reliable migrations, unless the behaviour
 from omission of the subsection is really bad.
 
-= Not sending existing elements =
+Not sending existing elements
+-----------------------------
 
 Sometimes members of the VMState are no longer needed;
-  removing them will break migration compatibility
-  making them version dependent and bumping the version will break backwards
-   migration compatibility.
+  - removing them will break migration compatibility
+
+  - making them version dependent and bumping the version will break backwards migration compatibility.
 
 The best way is to:
-  a) Add a new property/compatibility/function in the same way for subsections
-    above.
+  a) Add a new property/compatibility/function in the same way for subsections above.
   b) replace the VMSTATE macro with the _TEST version of the macro, e.g.:
-     VMSTATE_UINT32(foo, barstruct)
+
+   ``VMSTATE_UINT32(foo, barstruct)``
+
    becomes
-     VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)
 
-  Sometime in the future when we no longer care about the ancient
-versions these can be killed off.
+   ``VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)``
 
-= Return path =
+   Sometime in the future when we no longer care about the ancient versions these can be killed off.
+
+Return path
+-----------
 
 In most migration scenarios there is only a single data path that runs
 from the source VM to the destination, typically along a single fd (although
@@ -369,10 +395,11 @@ path.
 
   Destination side
      Forward path - read by main thread
-     Return path  - opened by main thread, written by main thread AND postcopy
-                    thread (protected by rp_mutex)
+     Return path  - opened by main thread, written by main thread AND postcopy thread (protected by rp_mutex)
+
+Postcopy
+========
 
-= Postcopy =
 'Postcopy' migration is a way to deal with migrations that refuse to converge
 (or take too long to converge) its plus side is that there is an upper bound on
 the amount of migration traffic and time it takes, the down side is that during
@@ -386,17 +413,18 @@ a fault that's translated by QEMU into a request to the source QEMU.
 Postcopy can be combined with precopy (i.e. normal migration) so that if precopy
 doesn't finish in a given time the switch is made to postcopy.
 
-=== Enabling postcopy ===
+Enabling postcopy
+-----------------
 
 To enable postcopy, issue this command on the monitor prior to the
 start of migration:
 
-migrate_set_capability postcopy-ram on
+``migrate_set_capability postcopy-ram on``
 
 The normal commands are then used to start a migration, which is still
 started in precopy mode.  Issuing:
 
-migrate_start_postcopy
+``migrate_start_postcopy``
 
 will now cause the transition from precopy to postcopy.
 It can be issued immediately after migration is started or any
@@ -406,7 +434,8 @@ Note: During the postcopy phase, the bandwidth limits set using
 migrate_set_speed is ignored (to avoid delaying requested pages that
 the destination is waiting for).
 
-=== Postcopy device transfer ===
+Postcopy device transfer
+------------------------
 
 Loading of device data may cause the device emulation to access guest RAM
 that may trigger faults that have to be resolved by the source, as such
@@ -416,6 +445,7 @@ before the device load begins to free the stream up.  This is achieved by
 'packaging' the device data into a blob that's read in one go.
 
 Source behaviour
+----------------
 
 Until postcopy is entered the migration stream is identical to normal
 precopy, except for the addition of a 'postcopy advise' command at
@@ -423,11 +453,10 @@ the beginning, to tell the destination that postcopy might happen.
 When postcopy starts the source sends the page discard data and then
 forms the 'package' containing:
 
-   Command: 'postcopy listen'
-   The device state
-      A series of sections, identical to the precopy streams device state stream
-      containing everything except postcopiable devices (i.e. RAM)
-   Command: 'postcopy run'
+   - Command: 'postcopy listen'
+   - The device state
+      A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)
+   - Command: 'postcopy run'
 
 The 'package' is sent as the data part of a Command: 'CMD_PACKAGED', and the
 contents are formatted in the same way as the main migration stream.
@@ -441,44 +470,38 @@ to be sent quickly in the hope that those pages are likely to be used
 by the destination soon.
 
 Destination behaviour
+---------------------
 
 Initially the destination looks the same as precopy, with a single thread
 reading the migration stream; the 'postcopy advise' and 'discard' commands
 are processed to change the way RAM is managed, but don't affect the stream
 processing.
 
-------------------------------------------------------------------------------
-                        1      2   3     4 5                      6   7
-main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
-thread                             |       |
-                                   |     (page request)
-                                   |        \___
-                                   v            \
-listen thread:                     --- page -- page -- page -- page -- page --
-
-                                   a   b        c
-------------------------------------------------------------------------------
-
-On receipt of CMD_PACKAGED (1)
-   All the data associated with the package - the ( ... ) section in the
-diagram - is read into memory, and the main thread recurses into
-qemu_loadvm_state_main to process the contents of the package (2)
-which contains commands (3,6) and devices (4...)
-
-On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
-a new thread (a) is started that takes over servicing the migration stream,
-while the main thread carries on loading the package.   It loads normal
-background page data (b) but if during a device load a fault happens (5) the
-returned page (c) is loaded by the listen thread allowing the main threads
-device load to carry on.
-
-The last thing in the CMD_PACKAGED is a 'RUN' command (6) letting the destination
-CPUs start running.
-At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour
-and is no longer used by migration, while the listen thread carries
-on servicing page data until the end of migration.
-
-=== Postcopy states ===
+::
+
+  ------------------------------------------------------------------------------
+                          1      2   3     4 5                      6   7
+  main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
+  thread                             |       |
+                                     |     (page request)
+                                     |        \___
+                                     v            \
+  listen thread:                     --- page -- page -- page -- page -- page --
+
+                                     a   b        c
+  ------------------------------------------------------------------------------
+
+- On receipt of CMD_PACKAGED (1)
+   All the data associated with the package - the ( ... ) section in the diagram - is read into memory, and the main thread recurses into qemu_loadvm_state_main to process the contents of the package (2) which contains commands (3,6) and devices (4...)
+
+- On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
+   a new thread (a) is started that takes over servicing the migration stream, while the main thread carries on loading the package.   It loads normal background page data (b) but if during a device load a fault happens (5) the returned page (c) is loaded by the listen thread allowing the main threads device load to carry on.
+
+- The last thing in the CMD_PACKAGED is a 'RUN' command (6)
+   letting the destination CPUs start running.  At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour and is no longer used by migration, while the listen thread carries on servicing page data until the end of migration.
+
+Postcopy states
+---------------
 
 Postcopy moves through a series of states (see postcopy_state) from
 ADVISE->DISCARD->LISTEN->RUNNING->END
@@ -516,7 +539,8 @@ ADVISE->DISCARD->LISTEN->RUNNING->END
   End:     The listen thread can now quit, and perform the cleanup of migration
            state, the migration is now complete.
 
-=== Source side page maps ===
+Source side page maps
+---------------------
 
 The source side keeps two bitmaps during postcopy; 'the migration bitmap'
 and 'unsent map'.  The 'migration bitmap' is basically the same as in
@@ -529,6 +553,7 @@ The 'unsent map' is used for the transition to postcopy. It is a bitmap that
 has a bit cleared whenever a page is sent to the destination, however during
 the transition to postcopy mode it is combined with the migration bitmap
 to form a set of pages that:
+
    a) Have been sent but then redirtied (which must be discarded)
    b) Have not yet been sent - which also must be discarded to cause any
       transparent huge pages built during precopy to be broken.
@@ -540,7 +565,8 @@ request for a page that has already been sent is ignored.  Duplicate requests
 such as this can happen as a page is sent at about the same time the
 destination accesses it.
 
-=== Postcopy with hugepages ===
+Postcopy with hugepages
+-----------------------
 
 Postcopy now works with hugetlbfs backed memory:
   a) The linux kernel on the destination must support userfault on hugepages.
Signature · 1 line
2.14.3
Reviewed-by on v1 1/1
#1 lore raw
Quoted 5 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Mostly just manual conversion with very minor fixes.

Some nits below, but r-b is fine with/without changing them. Mostly it's about wrapping lines. I think .rst supports wrapping lines?

Thanks,

Quoted 221 lines
---
 docs/devel/{migration.txt => migration.rst} | 326 +++++++++++++++-------------
 1 file changed, 176 insertions(+), 150 deletions(-)
 rename docs/devel/{migration.txt => migration.rst} (74%)
docs/devel/migration.rst renamed from docs/devel/migration.txt +65 -41
similarity index 74%
rename from docs/devel/migration.txt
rename to docs/devel/migration.rst
index 4030703726..7d64298cd7 100644
@@ -1,4 +1,6 @@
-= Migration =
+=========
+Migration
+=========
 
 QEMU has code to load/save the state of the guest that it is running.
 These are two complementary operations.  Saving the state just does
@@ -26,7 +28,8 @@ the guest to be stopped.  Typically the time that the guest is
 unresponsive during live migration is the low hundred of milliseconds
 (notice that this depends on a lot of things).
 
-=== Types of migration ===
+Types of migration
+==================
 
 Now that we have talked about live migration, there are several ways
 to do migration:
@@ -41,21 +44,25 @@ All these four migration protocols use the same infrastructure to
 save/restore state devices.  This infrastructure is shared with the
 savevm/loadvm functionality.
 
-=== State Live Migration ===
+State Live Migration
+====================
 
 This is used for RAM and block devices.  It is not yet ported to vmstate.
 <Fill more information here>
 
-=== What is the common infrastructure ===
+Common infrastructure
+=====================
 
 QEMU uses a QEMUFile abstraction to be able to do migration.  Any type
 of migration that wants to use QEMU infrastructure has to create a
 QEMUFile with:
 
-QEMUFile *qemu_fopen_ops(void *opaque,
-                         QEMUFilePutBufferFunc *put_buffer,
-                         QEMUFileGetBufferFunc *get_buffer,
-                         QEMUFileCloseFunc *close);
+.. code:: c
+
+  QEMUFile *qemu_fopen_ops(void *opaque,
+                           QEMUFilePutBufferFunc *put_buffer,
+                           QEMUFileGetBufferFunc *get_buffer,
+                           QEMUFileCloseFunc *close);
 
 The functions have the following functionality:
 
@@ -63,19 +70,25 @@ This function writes a chunk of data to a file at the given position.
 The pos argument can be ignored if the file is only used for
 streaming.  The handler should try to write all of the data it can.
 
-typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
+                                      int64_t pos, int size);
 
 Read a chunk of data from a file at the given position.  The pos argument
 can be ignored if the file is only be used for streaming.  The number of
 bytes actually read should be returned.
 
-typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
+                                      int64_t pos, int size);
 
 Close a file and return an error code.
 
-typedef int (QEMUFileCloseFunc)(void *opaque);
+.. code:: c
+
+  typedef int (QEMUFileCloseFunc)(void *opaque);
 
 You can use any internal state that you need using the opaque void *
 pointer that is passed to all functions.
@@ -83,7 +96,8 @@ pointer that is passed to all functions.
 The important functions for us are put_buffer()/get_buffer() that
 allow to write/read a buffer into the QEMUFile.
 
-=== How to save the state of one device ===
+Saving the state of one device
+==============================
 
 The state of a device is saved using intermediate buffers.  There are
 some helper functions to assist this saving.
@@ -97,30 +111,34 @@ associated with a series of fields saved.  The save_state always saves
 the state as the newer version.  But load_state sometimes is able to
 load state from an older version.
 
-=== Legacy way ===
+Legacy way
+----------
 
 This way is going to disappear as soon as all current users are ported to VMSTATE.
 
 Each device has to register two functions, one to save the state and
 another to load the state back.
 
-int register_savevm(DeviceState *dev,
-                    const char *idstr,
-                    int instance_id,
-                    int version_id,
-                    SaveStateHandler *save_state,
-                    LoadStateHandler *load_state,
-                    void *opaque);
+.. code:: c
+
+  int register_savevm(DeviceState *dev,
+                      const char *idstr,
+                      int instance_id,
+                      int version_id,
+                      SaveStateHandler *save_state,
+                      LoadStateHandler *load_state,
+                      void *opaque);
 
-typedef void SaveStateHandler(QEMUFile *f, void *opaque);
-typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
+  typedef void SaveStateHandler(QEMUFile *f, void *opaque);
+  typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
 
 The important functions for the device state format are the save_state
 and load_state.  Notice that load_state receives a version_id
 parameter to know what state format is receiving.  save_state doesn't
 have a version_id parameter because it always uses the latest version.
 
-=== VMState ===
+VMState
+-------
 
 The legacy way of saving/loading state of the device had the problem
 that we have to maintain two functions in sync.  If we did one change
@@ -135,23 +153,27 @@ save/load functions.
 
 An example (from hw/input/pckbd.c)
 
-static const VMStateDescription vmstate_kbd = {
-    .name = "pckbd",
-    .version_id = 3,
-    .minimum_version_id = 3,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT8(write_cmd, KBDState),
-        VMSTATE_UINT8(status, KBDState),
-        VMSTATE_UINT8(mode, KBDState),
-        VMSTATE_UINT8(pending, KBDState),
-        VMSTATE_END_OF_LIST()
-    }
-};
+.. code:: c
+
+  static const VMStateDescription vmstate_kbd = {
+      .name = "pckbd",
+      .version_id = 3,
+      .minimum_version_id = 3,
+      .fields = (VMStateField[]) {
+          VMSTATE_UINT8(write_cmd, KBDState),
+          VMSTATE_UINT8(status, KBDState),
+          VMSTATE_UINT8(mode, KBDState),
+          VMSTATE_UINT8(pending, KBDState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
 
 We are declaring the state with name "pckbd".
 The version_id is 3, and the fields are 4 uint8_t in a KBDState structure.
 We registered this with:
 
+.. code:: c
+
     vmstate_register(NULL, 0, &vmstate_kbd, s);
 
 Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
@@ -159,7 +181,8 @@ Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
 You can search for VMSTATE_* macros for lots of types used in QEMU in
 include/hw/hw.h.
 
-=== More about versions ===
+More about versions
+-------------------
 
 Version numbers are intended for major incompatible changes to the
 migration of a device, and using them breaks backwards-migration
@@ -183,7 +206,8 @@ function is deprecated and will be removed when no more users are left.
 Saving state will always create a section with the 'version_id' value
 and thus can't be loaded by any older QEMU.
 
-===  Massaging functions ===
+Massaging functions
+-------------------
 
 Sometimes, it is not enough to be able to save the state directly
 from one structure, we need to fill the correct values there.  One
@@ -194,20 +218,19 @@ load the state for the cpu that we have just loaded from the QEMUFile.
 
 The functions to do that are inside a vmstate definition, and are called:
 
-- int (*pre_load)(void *opaque);
+- ``int (*pre_load)(void *opaque);``
 
   This function is called before we load the state of one device.
 
-- int (*post_load)(void *opaque, int version_id);
+- ``int (*post_load)(void *opaque, int version_id);``
 

docs/devel/migration.rst load the state for the cpu that we have just loaded from the QEMUFile. · line 233

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -3
   This function is called after we load the state of one device.
 
-- int (*pre_save)(void *opaque);
+- ``int (*pre_save)(void *opaque);``
 
   This function is called before we save the state of one device.
 
-Example: You can look at hpet.c, that uses the three function to
-         massage the state that is transferred.
+Example: You can look at hpet.c, that uses the three function to massage the state that is transferred.

This change seems meaningless, I would prefer keep it wrapped, but I'm fine with it.

Quoted 111 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +46 -44
 If you use memory API functions that update memory layout outside
 initialization (i.e., in response to a guest action), this is a strong
@@ -221,7 +244,8 @@ Examples of such memory API functions are:
   - memory_region_set_address()
   - memory_region_set_alias_offset()
 
-=== Subsections ===
+Subsections
+-----------
 
 The use of version_id allows to be able to migrate from older versions
 to newer versions of a device.  But not the other way around.  This
@@ -251,48 +275,50 @@ value that it uses.
 
 Example:
 
-static bool ide_drive_pio_state_needed(void *opaque)
-{
-    IDEState *s = opaque;
-
-    return ((s->status & DRQ_STAT) != 0)
-        || (s->bus->error_status & BM_STATUS_PIO_RETRY);
-}
-
-const VMStateDescription vmstate_ide_drive_pio_state = {
-    .name = "ide_drive/pio_state",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .pre_save = ide_drive_pio_pre_save,
-    .post_load = ide_drive_pio_post_load,
-    .needed = ide_drive_pio_state_needed,
-    .fields = (VMStateField[]) {
-        VMSTATE_INT32(req_nb_sectors, IDEState),
-        VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
-                             vmstate_info_uint8, uint8_t),
-        VMSTATE_INT32(cur_io_buffer_offset, IDEState),
-        VMSTATE_INT32(cur_io_buffer_len, IDEState),
-        VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
-        VMSTATE_INT32(elementary_transfer_size, IDEState),
-        VMSTATE_INT32(packet_transfer_size, IDEState),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
-const VMStateDescription vmstate_ide_drive = {
-    .name = "ide_drive",
-    .version_id = 3,
-    .minimum_version_id = 0,
-    .post_load = ide_drive_post_load,
-    .fields = (VMStateField[]) {
-        .... several fields ....
-        VMSTATE_END_OF_LIST()
-    },
-    .subsections = (const VMStateDescription*[]) {
-        &vmstate_ide_drive_pio_state,
-        NULL
-    }
-};
+.. code:: c
+
+  static bool ide_drive_pio_state_needed(void *opaque)
+  {
+      IDEState *s = opaque;
+
+      return ((s->status & DRQ_STAT) != 0)
+          || (s->bus->error_status & BM_STATUS_PIO_RETRY);
+  }
+
+  const VMStateDescription vmstate_ide_drive_pio_state = {
+      .name = "ide_drive/pio_state",
+      .version_id = 1,
+      .minimum_version_id = 1,
+      .pre_save = ide_drive_pio_pre_save,
+      .post_load = ide_drive_pio_post_load,
+      .needed = ide_drive_pio_state_needed,
+      .fields = (VMStateField[]) {
+          VMSTATE_INT32(req_nb_sectors, IDEState),
+          VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
+                               vmstate_info_uint8, uint8_t),
+          VMSTATE_INT32(cur_io_buffer_offset, IDEState),
+          VMSTATE_INT32(cur_io_buffer_len, IDEState),
+          VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
+          VMSTATE_INT32(elementary_transfer_size, IDEState),
+          VMSTATE_INT32(packet_transfer_size, IDEState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
+
+  const VMStateDescription vmstate_ide_drive = {
+      .name = "ide_drive",
+      .version_id = 3,
+      .minimum_version_id = 0,
+      .post_load = ide_drive_post_load,
+      .fields = (VMStateField[]) {
+          .... several fields ....
+          VMSTATE_END_OF_LIST()
+      },
+      .subsections = (const VMStateDescription*[]) {
+          &vmstate_ide_drive_pio_state,
+          NULL
+      }
+  };
 
 Here we have a subsection for the pio state.  We only need to
 save/send this state when we are in the middle of a pio operation
@@ -305,14 +331,11 @@ to send a subsection allows backwards migration compatibility when
 new subsections are added.
 
 For example;
-   a) Add a new property using DEFINE_PROP_BOOL - e.g. support-foo and

docs/devel/migration.rst to send a subsection allows backwards migration compatibility when · line 338

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -5
-      default it to true.
-   b) Add an entry to the HW_COMPAT_ for the previous version
-      that sets the property to false.
+   a) Add a new property using ``DEFINE_PROP_BOOL`` - e.g. support-foo and default it to true.
+   b) Add an entry to the ``HW_COMPAT_`` for the previous version that sets the property to false.
    c) Add a static bool  support_foo function that tests the property.
    d) Add a subsection with a .needed set to the support_foo function
-   e) (potentially) Add a pre_load that sets up a default value for 'foo'
-      to be used if the subsection isn't loaded.
+   e) (potentially) Add a pre_load that sets up a default value for 'foo' to be used if the subsection isn't loaded.

(same to these lines)

Quoted 44 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +14 -11
 Now that subsection will not be generated when using an older
 machine type and the migration stream will be accepted by older
@@ -332,25 +355,28 @@ in most cases.  In general the preference is to tie the subsection to
 the machine type, and allow reliable migrations, unless the behaviour
 from omission of the subsection is really bad.
 
-= Not sending existing elements =
+Not sending existing elements
+-----------------------------
 
 Sometimes members of the VMState are no longer needed;
-  removing them will break migration compatibility
-  making them version dependent and bumping the version will break backwards
-   migration compatibility.
+  - removing them will break migration compatibility
+
+  - making them version dependent and bumping the version will break backwards migration compatibility.
 
 The best way is to:
-  a) Add a new property/compatibility/function in the same way for subsections
-    above.
+  a) Add a new property/compatibility/function in the same way for subsections above.
   b) replace the VMSTATE macro with the _TEST version of the macro, e.g.:
-     VMSTATE_UINT32(foo, barstruct)
+
+   ``VMSTATE_UINT32(foo, barstruct)``
+
    becomes
-     VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)
 
-  Sometime in the future when we no longer care about the ancient
-versions these can be killed off.
+   ``VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)``
 
-= Return path =
+   Sometime in the future when we no longer care about the ancient versions these can be killed off.
+
+Return path
+-----------
 
 In most migration scenarios there is only a single data path that runs
 from the source VM to the destination, typically along a single fd (although

docs/devel/migration.rst path. · line 398

Quoted 6 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +1 -2
@@ -369,10 +395,11 @@ path.
 
   Destination side
      Forward path - read by main thread
-     Return path  - opened by main thread, written by main thread AND postcopy
-                    thread (protected by rp_mutex)
+     Return path  - opened by main thread, written by main thread AND postcopy thread (protected by rp_mutex)

(same here)

Quoted 50 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +10 -5
+
+Postcopy
+========
 
-= Postcopy =
 'Postcopy' migration is a way to deal with migrations that refuse to converge
 (or take too long to converge) its plus side is that there is an upper bound on
 the amount of migration traffic and time it takes, the down side is that during
@@ -386,17 +413,18 @@ a fault that's translated by QEMU into a request to the source QEMU.
 Postcopy can be combined with precopy (i.e. normal migration) so that if precopy
 doesn't finish in a given time the switch is made to postcopy.
 
-=== Enabling postcopy ===
+Enabling postcopy
+-----------------
 
 To enable postcopy, issue this command on the monitor prior to the
 start of migration:
 
-migrate_set_capability postcopy-ram on
+``migrate_set_capability postcopy-ram on``
 
 The normal commands are then used to start a migration, which is still
 started in precopy mode.  Issuing:
 
-migrate_start_postcopy
+``migrate_start_postcopy``
 
 will now cause the transition from precopy to postcopy.
 It can be issued immediately after migration is started or any
@@ -406,7 +434,8 @@ Note: During the postcopy phase, the bandwidth limits set using
 migrate_set_speed is ignored (to avoid delaying requested pages that
 the destination is waiting for).
 
-=== Postcopy device transfer ===
+Postcopy device transfer
+------------------------
 
 Loading of device data may cause the device emulation to access guest RAM
 that may trigger faults that have to be resolved by the source, as such
@@ -416,6 +445,7 @@ before the device load begins to free the stream up.  This is achieved by
 'packaging' the device data into a blob that's read in one go.
 
 Source behaviour
+----------------
 
 Until postcopy is entered the migration stream is identical to normal
 precopy, except for the addition of a 'postcopy advise' command at
@@ -423,11 +453,10 @@ the beginning, to tell the destination that postcopy might happen.
 When postcopy starts the source sends the page discard data and then

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 458

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -5
 forms the 'package' containing:
 
-   Command: 'postcopy listen'
-   The device state
-      A series of sections, identical to the precopy streams device state stream
-      containing everything except postcopiable devices (i.e. RAM)
-   Command: 'postcopy run'
+   - Command: 'postcopy listen'
+   - The device state
+      A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

(a super long line, fine too)

Quoted 59 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +14 -32
+   - Command: 'postcopy run'
 
 The 'package' is sent as the data part of a Command: 'CMD_PACKAGED', and the
 contents are formatted in the same way as the main migration stream.
@@ -441,44 +470,38 @@ to be sent quickly in the hope that those pages are likely to be used
 by the destination soon.
 
 Destination behaviour
+---------------------
 
 Initially the destination looks the same as precopy, with a single thread
 reading the migration stream; the 'postcopy advise' and 'discard' commands
 are processed to change the way RAM is managed, but don't affect the stream
 processing.
 
-------------------------------------------------------------------------------
-                        1      2   3     4 5                      6   7
-main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
-thread                             |       |
-                                   |     (page request)
-                                   |        \___
-                                   v            \
-listen thread:                     --- page -- page -- page -- page -- page --
-
-                                   a   b        c
-------------------------------------------------------------------------------
-
-On receipt of CMD_PACKAGED (1)
-   All the data associated with the package - the ( ... ) section in the
-diagram - is read into memory, and the main thread recurses into
-qemu_loadvm_state_main to process the contents of the package (2)
-which contains commands (3,6) and devices (4...)
-
-On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
-a new thread (a) is started that takes over servicing the migration stream,
-while the main thread carries on loading the package.   It loads normal
-background page data (b) but if during a device load a fault happens (5) the
-returned page (c) is loaded by the listen thread allowing the main threads
-device load to carry on.
-
-The last thing in the CMD_PACKAGED is a 'RUN' command (6) letting the destination
-CPUs start running.
-At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour
-and is no longer used by migration, while the listen thread carries
-on servicing page data until the end of migration.
-
-=== Postcopy states ===
+::
+
+  ------------------------------------------------------------------------------
+                          1      2   3     4 5                      6   7
+  main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
+  thread                             |       |
+                                     |     (page request)
+                                     |        \___
+                                     v            \
+  listen thread:                     --- page -- page -- page -- page -- page --
+
+                                     a   b        c

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 501

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +10 -0
+  ------------------------------------------------------------------------------
+
+- On receipt of CMD_PACKAGED (1)
+   All the data associated with the package - the ( ... ) section in the diagram - is read into memory, and the main thread recurses into qemu_loadvm_state_main to process the contents of the package (2) which contains commands (3,6) and devices (4...)
+
+- On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
+   a new thread (a) is started that takes over servicing the migration stream, while the main thread carries on loading the package.   It loads normal background page data (b) but if during a device load a fault happens (5) the returned page (c) is loaded by the listen thread allowing the main threads device load to carry on.
+
+- The last thing in the CMD_PACKAGED is a 'RUN' command (6)
+   letting the destination CPUs start running.  At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour and is no longer used by migration, while the listen thread carries on servicing page data until the end of migration.

(here the lines can be wrapped too?)

Quoted 37 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +8 -2
+
+Postcopy states
+---------------
 
 Postcopy moves through a series of states (see postcopy_state) from
 ADVISE->DISCARD->LISTEN->RUNNING->END
@@ -516,7 +539,8 @@ ADVISE->DISCARD->LISTEN->RUNNING->END
   End:     The listen thread can now quit, and perform the cleanup of migration
            state, the migration is now complete.
 
-=== Source side page maps ===
+Source side page maps
+---------------------
 
 The source side keeps two bitmaps during postcopy; 'the migration bitmap'
 and 'unsent map'.  The 'migration bitmap' is basically the same as in
@@ -529,6 +553,7 @@ The 'unsent map' is used for the transition to postcopy. It is a bitmap that
 has a bit cleared whenever a page is sent to the destination, however during
 the transition to postcopy mode it is combined with the migration bitmap
 to form a set of pages that:
+
    a) Have been sent but then redirtied (which must be discarded)
    b) Have not yet been sent - which also must be discarded to cause any
       transparent huge pages built during precopy to be broken.
@@ -540,7 +565,8 @@ request for a page that has already been sent is ignored.  Duplicate requests
 such as this can happen as a page is sent at about the same time the
 destination accesses it.
 
-=== Postcopy with hugepages ===
+Postcopy with hugepages
+-----------------------
 
 Postcopy now works with hugetlbfs backed memory:
   a) The linux kernel on the destination must support userfault on hugepages.
Signature · 1 line
2.14.3
Signature · 1 line
Peter Xu
#3 lore raw
In reply to Dr. David Alan Gilbert (git)
Quoted 9 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Mostly just manual conversion with very minor fixes.

---
 docs/devel/{migration.txt => migration.rst} | 326 +++++++++++++++-------------
 1 file changed, 176 insertions(+), 150 deletions(-)
 rename docs/devel/{migration.txt => migration.rst} (74%)

Hey Dave,

A good chunk of changes I suggest are pre-existing, so take it with a grain of salt :-). I'm ambivalent on whether to keep the conversion to rST and the other rST syntax changes separate. Maybe we can keep in one change, as this doesn't impact backports in terms of code / functionality.

[...]

Quoted 5 lines

+State Live Migration
+====================

This is used for RAM and block devices. It is not yet ported to vmstate. <Fill more information here>

Not this patch's problem, but do we have the more information that is to be filled in above?

[...]

docs/devel/migration.rst pointer that is passed to all functions. · line 96

Quoted 4 lines

You can use any internal state that you need using the opaque void * pointer that is passed to all functions.

docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
@@ -83,7 +96,8 @@ pointer that is passed to all functions.
 The important functions for us are put_buffer()/get_buffer() that

Might want to highlight the fuctions (as you call them as important) with back ticks (and some spacing): put_buffer() / get_buffer().

docs/devel/migration.rst associated with a series of fields saved. The save_state always saves

Quoted 9 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 allow to write/read a buffer into the QEMUFile.
 
-=== How to save the state of one device ===
+Saving the state of one device
+==============================
 
 The state of a device is saved using intermediate buffers.  There are
 some helper functions to assist this saving.
@@ -97,30 +111,34 @@ associated with a series of fields saved.  The save_state always saves

Here too: s/save_state/``save_state``/

(And other occurrences throughout the doc.)

docs/devel/migration.rst associated with a series of fields saved. The save_state always saves · line 111

Quoted 1 line
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
 the state as the newer version.  But load_state sometimes is able to

s/load_state/``load_state``/

(And other occurrences throughout the doc.)

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 load state from an older version.
 
-=== Legacy way ===
+Legacy way
+----------
 
 This way is going to disappear as soon as all current users are ported to VMSTATE.
 
 Each device has to register two functions, one to save the state and
 another to load the state back.

[...]

Quoted 4 lines

The important functions for the device state format are the save_state and load_state. Notice that load_state receives a version_id parameter to know what state format is receiving. save_state doesn't have a version_id parameter because it always uses the latest version.

In the above paragrah, too (``save_state``, ``load_state``).

Also s/version_id/'version_id'/g

Quoted 4 lines

-=== VMState ===
+VMState
+-------

[...]

Quoted 9 lines

We are declaring the state with name "pckbd".
The version_id is 3, and the fields are 4 uint8_t in a KBDState structure. We registered this with:

+.. code:: c
+

     vmstate_register(NULL, 0, &vmstate_kbd, s);

Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.

Not introduced in this patch, but: s/ids/IDs/

Two points:

- I think the above is a TODO item; so you can use:

.. TODO (dgilbert):: talk about how vmstate <-> qdev interact, and what the instance ids mean.

As that'll keep it only in the source rST; but not in the rendered version.

- Also, other minor thing (again, pre-existing): capitalize all the things like s/tcp/TCP; s/unix/UNIX/ where appropriate.

docs/devel/migration.rst Note: talk about how vmstate <-> qdev interact, and what the instance ids mean. · line 181

Quoted 2 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
@@ -159,7 +181,8 @@ Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
 You can search for VMSTATE_* macros for lots of types used in QEMU in

Maybe: s/VMSTATE_*/``VMSTATE_*``/

docs/devel/migration.rst function is deprecated and will be removed when no more users are left. · line 207

Quoted 11 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 include/hw/hw.h.
 
-=== More about versions ===
+More about versions
+-------------------
 
 Version numbers are intended for major incompatible changes to the
 migration of a device, and using them breaks backwards-migration
@@ -183,7 +206,8 @@ function is deprecated and will be removed when no more users are left.
 Saving state will always create a section with the 'version_id' value
 and thus can't be loaded by any older QEMU.

Also, not in this patch, but keeping up with the theme of highlighting function names:

    s/load_state_old()/``load_state_old()``/

And other fields and strings (you can use a single quote, or a back tick for italics; whatever you choose, might want to keep it consistent):

    s/version_id/'version_id'/
Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
-===  Massaging functions ===
+Massaging functions
+-------------------

[...]

Quoted 5 lines

-=== Subsections ===
+Subsections
+-----------

The use of version_id allows to be able to migrate from older versions

s/version_id/'version_id'/ (or version_id, if you prefer italics)

Again, not in this patch, but noticed in the rendered version locally:

s/post_load()/``post_load()``/

[...]

Quoted 2 lines

Here we have a subsection for the pio state. We only need to save/send this state when we are in the middle of a pio operation

s/ide_drive_pio_state_needed()/``ide_drive_pio_state_needed()``/

Quoted 1 line
(quoted diff) +0 -0
@@ -305,14 +331,11 @@ to send a subsection allows backwards migration compatibility when

[...]

Quoted 4 lines

+Not sending existing elements
+-----------------------------

Sometimes members of the VMState are no longer needed;

Pre-existing: s/;/:/

[...]

Quoted 2 lines

+Return path
+-----------

[...]

Pre-existing, and not this patch's problem. Noticed in my HTML render:
in this section, you might to highlight (double back ticks) the function:

``qemu_file_get_return_path(QEMUFile* fwdpath)``

[...]

Quoted 2 lines

+Postcopy
+========

[...]

Quoted 3 lines

-=== Enabling postcopy ===
+Enabling postcopy
+-----------------

[...]

Not in this patch, but in this section, 'migrate_set_speed' is mentioned; since it's a QMP command, you might want to highlith it:
``migrate_set_speed``

Also, from this section, you might want to use adminitions like:

.. note::
During the postcopy phase, the bandwidth limits set using migrate_set_speed is ignored (to avoid delaying requested pages that the destination is waiting for).

Use this wherever you see fit in the document.

[...]

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 427

Quoted 11 lines

Source behaviour
+----------------

Until postcopy is entered the migration stream is identical to normal precopy, except for the addition of a 'postcopy advise' command at

docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -2
@@ -423,11 +453,10 @@ the beginning, to tell the destination that postcopy might happen.
 When postcopy starts the source sends the page discard data and then
 forms the 'package' containing:
 
-   Command: 'postcopy listen'
-   The device state

In my local Sphinx-based HTML rendering, the above "The device state" ended up being bold somehow.

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 458

Quoted 6 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -3
-      A series of sections, identical to the precopy streams device state stream
-      containing everything except postcopiable devices (i.e. RAM)
-   Command: 'postcopy run'
+   - Command: 'postcopy listen'
+   - The device state
+      A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

Might want to wrap this long line (and other places); as it's causing an extra new line in my local rendering.

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 461

Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +1 -0
+   - Command: 'postcopy run'
 
 The 'package' is sent as the data part of a Command: 'CMD_PACKAGED', and the

s/CMD_PACKAGED/``CMD_PACKAGED``/g ?

[...]

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 501

Quoted 8 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +8 -0
+- On receipt of CMD_PACKAGED (1)
+   All the data associated with the package - the ( ... ) section in the diagram - is read into memory, and the main thread recurses into qemu_loadvm_state_main to process the contents of the package (2) which contains commands (3,6) and devices (4...)
+
+- On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
+   a new thread (a) is started that takes over servicing the migration stream, while the main thread carries on loading the package.   It loads normal background page data (b) but if during a device load a fault happens (5) the returned page (c) is loaded by the listen thread allowing the main threads device load to carry on.
+
+- The last thing in the CMD_PACKAGED is a 'RUN' command (6)
+   letting the destination CPUs start running.  At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour and is no longer used by migration, while the listen thread carries on servicing page data until the end of migration.

Super long lines, needs wrapping. Here too something seem to cause needless "bold" text in my Sphinx-based HTML rendering.

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 504

Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -0
+
+Postcopy states
+---------------

[...]

There's a list in this document. You might want to 'listify' it (using numbers, alphabets, etc), like you did it in other places.

[...]

Quoted 3 lines

-=== Postcopy with hugepages ===
+Postcopy with hugepages
+-----------------------

Here too, noticed after reading the rendered version; you might want to highlight the QEMU command-line parameters & file system paths using the verbatim (``):

-mem-path
/dev/hugepages
-mem-prealloc

[...]

Signature · 1 line
/kashyap
#2 lore raw
In reply to Peter Xu
Quoted 11 lines · * Peter Xu ([email protected]) wrote:
Quoted 5 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Mostly just manual conversion with very minor fixes.

Some nits below, but r-b is fine with/without changing them. Mostly it's about wrapping lines. I think .rst supports wrapping lines?

Thanks, I've fixed them. What I hadn't figured out previously was how to do the wrapping in the middle of lists; you've got to get the indent on the wrapped lines just write for the list not to complain.

Dave

Quoted 581 lines

Thanks,

Quoted 221 lines
---
 docs/devel/{migration.txt => migration.rst} | 326 +++++++++++++++-------------
 1 file changed, 176 insertions(+), 150 deletions(-)
 rename docs/devel/{migration.txt => migration.rst} (74%)
docs/devel/migration.rst renamed from docs/devel/migration.txt +65 -41
similarity index 74%
rename from docs/devel/migration.txt
rename to docs/devel/migration.rst
index 4030703726..7d64298cd7 100644
@@ -1,4 +1,6 @@
-= Migration =
+=========
+Migration
+=========
 
 QEMU has code to load/save the state of the guest that it is running.
 These are two complementary operations.  Saving the state just does
@@ -26,7 +28,8 @@ the guest to be stopped.  Typically the time that the guest is
 unresponsive during live migration is the low hundred of milliseconds
 (notice that this depends on a lot of things).
 
-=== Types of migration ===
+Types of migration
+==================
 
 Now that we have talked about live migration, there are several ways
 to do migration:
@@ -41,21 +44,25 @@ All these four migration protocols use the same infrastructure to
 save/restore state devices.  This infrastructure is shared with the
 savevm/loadvm functionality.
 
-=== State Live Migration ===
+State Live Migration
+====================
 
 This is used for RAM and block devices.  It is not yet ported to vmstate.
 <Fill more information here>
 
-=== What is the common infrastructure ===
+Common infrastructure
+=====================
 
 QEMU uses a QEMUFile abstraction to be able to do migration.  Any type
 of migration that wants to use QEMU infrastructure has to create a
 QEMUFile with:
 
-QEMUFile *qemu_fopen_ops(void *opaque,
-                         QEMUFilePutBufferFunc *put_buffer,
-                         QEMUFileGetBufferFunc *get_buffer,
-                         QEMUFileCloseFunc *close);
+.. code:: c
+
+  QEMUFile *qemu_fopen_ops(void *opaque,
+                           QEMUFilePutBufferFunc *put_buffer,
+                           QEMUFileGetBufferFunc *get_buffer,
+                           QEMUFileCloseFunc *close);
 
 The functions have the following functionality:
 
@@ -63,19 +70,25 @@ This function writes a chunk of data to a file at the given position.
 The pos argument can be ignored if the file is only used for
 streaming.  The handler should try to write all of the data it can.
 
-typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFilePutBufferFunc)(void *opaque, const uint8_t *buf,
+                                      int64_t pos, int size);
 
 Read a chunk of data from a file at the given position.  The pos argument
 can be ignored if the file is only be used for streaming.  The number of
 bytes actually read should be returned.
 
-typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
-                                    int64_t pos, int size);
+.. code:: c
+
+  typedef int (QEMUFileGetBufferFunc)(void *opaque, uint8_t *buf,
+                                      int64_t pos, int size);
 
 Close a file and return an error code.
 
-typedef int (QEMUFileCloseFunc)(void *opaque);
+.. code:: c
+
+  typedef int (QEMUFileCloseFunc)(void *opaque);
 
 You can use any internal state that you need using the opaque void *
 pointer that is passed to all functions.
@@ -83,7 +96,8 @@ pointer that is passed to all functions.
 The important functions for us are put_buffer()/get_buffer() that
 allow to write/read a buffer into the QEMUFile.
 
-=== How to save the state of one device ===
+Saving the state of one device
+==============================
 
 The state of a device is saved using intermediate buffers.  There are
 some helper functions to assist this saving.
@@ -97,30 +111,34 @@ associated with a series of fields saved.  The save_state always saves
 the state as the newer version.  But load_state sometimes is able to
 load state from an older version.
 
-=== Legacy way ===
+Legacy way
+----------
 
 This way is going to disappear as soon as all current users are ported to VMSTATE.
 
 Each device has to register two functions, one to save the state and
 another to load the state back.
 
-int register_savevm(DeviceState *dev,
-                    const char *idstr,
-                    int instance_id,
-                    int version_id,
-                    SaveStateHandler *save_state,
-                    LoadStateHandler *load_state,
-                    void *opaque);
+.. code:: c
+
+  int register_savevm(DeviceState *dev,
+                      const char *idstr,
+                      int instance_id,
+                      int version_id,
+                      SaveStateHandler *save_state,
+                      LoadStateHandler *load_state,
+                      void *opaque);
 
-typedef void SaveStateHandler(QEMUFile *f, void *opaque);
-typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
+  typedef void SaveStateHandler(QEMUFile *f, void *opaque);
+  typedef int LoadStateHandler(QEMUFile *f, void *opaque, int version_id);
 
 The important functions for the device state format are the save_state
 and load_state.  Notice that load_state receives a version_id
 parameter to know what state format is receiving.  save_state doesn't
 have a version_id parameter because it always uses the latest version.
 
-=== VMState ===
+VMState
+-------
 
 The legacy way of saving/loading state of the device had the problem
 that we have to maintain two functions in sync.  If we did one change
@@ -135,23 +153,27 @@ save/load functions.
 
 An example (from hw/input/pckbd.c)
 
-static const VMStateDescription vmstate_kbd = {
-    .name = "pckbd",
-    .version_id = 3,
-    .minimum_version_id = 3,
-    .fields = (VMStateField[]) {
-        VMSTATE_UINT8(write_cmd, KBDState),
-        VMSTATE_UINT8(status, KBDState),
-        VMSTATE_UINT8(mode, KBDState),
-        VMSTATE_UINT8(pending, KBDState),
-        VMSTATE_END_OF_LIST()
-    }
-};
+.. code:: c
+
+  static const VMStateDescription vmstate_kbd = {
+      .name = "pckbd",
+      .version_id = 3,
+      .minimum_version_id = 3,
+      .fields = (VMStateField[]) {
+          VMSTATE_UINT8(write_cmd, KBDState),
+          VMSTATE_UINT8(status, KBDState),
+          VMSTATE_UINT8(mode, KBDState),
+          VMSTATE_UINT8(pending, KBDState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
 
 We are declaring the state with name "pckbd".
 The version_id is 3, and the fields are 4 uint8_t in a KBDState structure.
 We registered this with:
 
+.. code:: c
+
     vmstate_register(NULL, 0, &vmstate_kbd, s);
 
 Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
@@ -159,7 +181,8 @@ Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
 You can search for VMSTATE_* macros for lots of types used in QEMU in
 include/hw/hw.h.
 
-=== More about versions ===
+More about versions
+-------------------
 
 Version numbers are intended for major incompatible changes to the
 migration of a device, and using them breaks backwards-migration
@@ -183,7 +206,8 @@ function is deprecated and will be removed when no more users are left.
 Saving state will always create a section with the 'version_id' value
 and thus can't be loaded by any older QEMU.
 
-===  Massaging functions ===
+Massaging functions
+-------------------
 
 Sometimes, it is not enough to be able to save the state directly
 from one structure, we need to fill the correct values there.  One
@@ -194,20 +218,19 @@ load the state for the cpu that we have just loaded from the QEMUFile.
 
 The functions to do that are inside a vmstate definition, and are called:
 
-- int (*pre_load)(void *opaque);
+- ``int (*pre_load)(void *opaque);``
 
   This function is called before we load the state of one device.
 
-- int (*post_load)(void *opaque, int version_id);
+- ``int (*post_load)(void *opaque, int version_id);``
 

docs/devel/migration.rst load the state for the cpu that we have just loaded from the QEMUFile. · line 233

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -3
   This function is called after we load the state of one device.
 
-- int (*pre_save)(void *opaque);
+- ``int (*pre_save)(void *opaque);``
 
   This function is called before we save the state of one device.
 
-Example: You can look at hpet.c, that uses the three function to
-         massage the state that is transferred.
+Example: You can look at hpet.c, that uses the three function to massage the state that is transferred.

This change seems meaningless, I would prefer keep it wrapped, but I'm fine with it.

Quoted 111 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +46 -44
 If you use memory API functions that update memory layout outside
 initialization (i.e., in response to a guest action), this is a strong
@@ -221,7 +244,8 @@ Examples of such memory API functions are:
   - memory_region_set_address()
   - memory_region_set_alias_offset()
 
-=== Subsections ===
+Subsections
+-----------
 
 The use of version_id allows to be able to migrate from older versions
 to newer versions of a device.  But not the other way around.  This
@@ -251,48 +275,50 @@ value that it uses.
 
 Example:
 
-static bool ide_drive_pio_state_needed(void *opaque)
-{
-    IDEState *s = opaque;
-
-    return ((s->status & DRQ_STAT) != 0)
-        || (s->bus->error_status & BM_STATUS_PIO_RETRY);
-}
-
-const VMStateDescription vmstate_ide_drive_pio_state = {
-    .name = "ide_drive/pio_state",
-    .version_id = 1,
-    .minimum_version_id = 1,
-    .pre_save = ide_drive_pio_pre_save,
-    .post_load = ide_drive_pio_post_load,
-    .needed = ide_drive_pio_state_needed,
-    .fields = (VMStateField[]) {
-        VMSTATE_INT32(req_nb_sectors, IDEState),
-        VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
-                             vmstate_info_uint8, uint8_t),
-        VMSTATE_INT32(cur_io_buffer_offset, IDEState),
-        VMSTATE_INT32(cur_io_buffer_len, IDEState),
-        VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
-        VMSTATE_INT32(elementary_transfer_size, IDEState),
-        VMSTATE_INT32(packet_transfer_size, IDEState),
-        VMSTATE_END_OF_LIST()
-    }
-};
-
-const VMStateDescription vmstate_ide_drive = {
-    .name = "ide_drive",
-    .version_id = 3,
-    .minimum_version_id = 0,
-    .post_load = ide_drive_post_load,
-    .fields = (VMStateField[]) {
-        .... several fields ....
-        VMSTATE_END_OF_LIST()
-    },
-    .subsections = (const VMStateDescription*[]) {
-        &vmstate_ide_drive_pio_state,
-        NULL
-    }
-};
+.. code:: c
+
+  static bool ide_drive_pio_state_needed(void *opaque)
+  {
+      IDEState *s = opaque;
+
+      return ((s->status & DRQ_STAT) != 0)
+          || (s->bus->error_status & BM_STATUS_PIO_RETRY);
+  }
+
+  const VMStateDescription vmstate_ide_drive_pio_state = {
+      .name = "ide_drive/pio_state",
+      .version_id = 1,
+      .minimum_version_id = 1,
+      .pre_save = ide_drive_pio_pre_save,
+      .post_load = ide_drive_pio_post_load,
+      .needed = ide_drive_pio_state_needed,
+      .fields = (VMStateField[]) {
+          VMSTATE_INT32(req_nb_sectors, IDEState),
+          VMSTATE_VARRAY_INT32(io_buffer, IDEState, io_buffer_total_len, 1,
+                               vmstate_info_uint8, uint8_t),
+          VMSTATE_INT32(cur_io_buffer_offset, IDEState),
+          VMSTATE_INT32(cur_io_buffer_len, IDEState),
+          VMSTATE_UINT8(end_transfer_fn_idx, IDEState),
+          VMSTATE_INT32(elementary_transfer_size, IDEState),
+          VMSTATE_INT32(packet_transfer_size, IDEState),
+          VMSTATE_END_OF_LIST()
+      }
+  };
+
+  const VMStateDescription vmstate_ide_drive = {
+      .name = "ide_drive",
+      .version_id = 3,
+      .minimum_version_id = 0,
+      .post_load = ide_drive_post_load,
+      .fields = (VMStateField[]) {
+          .... several fields ....
+          VMSTATE_END_OF_LIST()
+      },
+      .subsections = (const VMStateDescription*[]) {
+          &vmstate_ide_drive_pio_state,
+          NULL
+      }
+  };
 
 Here we have a subsection for the pio state.  We only need to
 save/send this state when we are in the middle of a pio operation
@@ -305,14 +331,11 @@ to send a subsection allows backwards migration compatibility when
 new subsections are added.
 
 For example;
-   a) Add a new property using DEFINE_PROP_BOOL - e.g. support-foo and

docs/devel/migration.rst to send a subsection allows backwards migration compatibility when · line 338

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -5
-      default it to true.
-   b) Add an entry to the HW_COMPAT_ for the previous version
-      that sets the property to false.
+   a) Add a new property using ``DEFINE_PROP_BOOL`` - e.g. support-foo and default it to true.
+   b) Add an entry to the ``HW_COMPAT_`` for the previous version that sets the property to false.
    c) Add a static bool  support_foo function that tests the property.
    d) Add a subsection with a .needed set to the support_foo function
-   e) (potentially) Add a pre_load that sets up a default value for 'foo'
-      to be used if the subsection isn't loaded.
+   e) (potentially) Add a pre_load that sets up a default value for 'foo' to be used if the subsection isn't loaded.

(same to these lines)

Quoted 44 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +14 -11
 Now that subsection will not be generated when using an older
 machine type and the migration stream will be accepted by older
@@ -332,25 +355,28 @@ in most cases.  In general the preference is to tie the subsection to
 the machine type, and allow reliable migrations, unless the behaviour
 from omission of the subsection is really bad.
 
-= Not sending existing elements =
+Not sending existing elements
+-----------------------------
 
 Sometimes members of the VMState are no longer needed;
-  removing them will break migration compatibility
-  making them version dependent and bumping the version will break backwards
-   migration compatibility.
+  - removing them will break migration compatibility
+
+  - making them version dependent and bumping the version will break backwards migration compatibility.
 
 The best way is to:
-  a) Add a new property/compatibility/function in the same way for subsections
-    above.
+  a) Add a new property/compatibility/function in the same way for subsections above.
   b) replace the VMSTATE macro with the _TEST version of the macro, e.g.:
-     VMSTATE_UINT32(foo, barstruct)
+
+   ``VMSTATE_UINT32(foo, barstruct)``
+
    becomes
-     VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)
 
-  Sometime in the future when we no longer care about the ancient
-versions these can be killed off.
+   ``VMSTATE_UINT32_TEST(foo, barstruct, pre_version_baz)``
 
-= Return path =
+   Sometime in the future when we no longer care about the ancient versions these can be killed off.
+
+Return path
+-----------
 
 In most migration scenarios there is only a single data path that runs
 from the source VM to the destination, typically along a single fd (although

docs/devel/migration.rst path. · line 398

Quoted 6 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +1 -2
@@ -369,10 +395,11 @@ path.
 
   Destination side
      Forward path - read by main thread
-     Return path  - opened by main thread, written by main thread AND postcopy
-                    thread (protected by rp_mutex)
+     Return path  - opened by main thread, written by main thread AND postcopy thread (protected by rp_mutex)

(same here)

Quoted 50 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +10 -5
+
+Postcopy
+========
 
-= Postcopy =
 'Postcopy' migration is a way to deal with migrations that refuse to converge
 (or take too long to converge) its plus side is that there is an upper bound on
 the amount of migration traffic and time it takes, the down side is that during
@@ -386,17 +413,18 @@ a fault that's translated by QEMU into a request to the source QEMU.
 Postcopy can be combined with precopy (i.e. normal migration) so that if precopy
 doesn't finish in a given time the switch is made to postcopy.
 
-=== Enabling postcopy ===
+Enabling postcopy
+-----------------
 
 To enable postcopy, issue this command on the monitor prior to the
 start of migration:
 
-migrate_set_capability postcopy-ram on
+``migrate_set_capability postcopy-ram on``
 
 The normal commands are then used to start a migration, which is still
 started in precopy mode.  Issuing:
 
-migrate_start_postcopy
+``migrate_start_postcopy``
 
 will now cause the transition from precopy to postcopy.
 It can be issued immediately after migration is started or any
@@ -406,7 +434,8 @@ Note: During the postcopy phase, the bandwidth limits set using
 migrate_set_speed is ignored (to avoid delaying requested pages that
 the destination is waiting for).
 
-=== Postcopy device transfer ===
+Postcopy device transfer
+------------------------
 
 Loading of device data may cause the device emulation to access guest RAM
 that may trigger faults that have to be resolved by the source, as such
@@ -416,6 +445,7 @@ before the device load begins to free the stream up.  This is achieved by
 'packaging' the device data into a blob that's read in one go.
 
 Source behaviour
+----------------
 
 Until postcopy is entered the migration stream is identical to normal
 precopy, except for the addition of a 'postcopy advise' command at
@@ -423,11 +453,10 @@ the beginning, to tell the destination that postcopy might happen.
 When postcopy starts the source sends the page discard data and then

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 458

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -5
 forms the 'package' containing:
 
-   Command: 'postcopy listen'
-   The device state
-      A series of sections, identical to the precopy streams device state stream
-      containing everything except postcopiable devices (i.e. RAM)
-   Command: 'postcopy run'
+   - Command: 'postcopy listen'
+   - The device state
+      A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

(a super long line, fine too)

Quoted 59 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +14 -32
+   - Command: 'postcopy run'
 
 The 'package' is sent as the data part of a Command: 'CMD_PACKAGED', and the
 contents are formatted in the same way as the main migration stream.
@@ -441,44 +470,38 @@ to be sent quickly in the hope that those pages are likely to be used
 by the destination soon.
 
 Destination behaviour
+---------------------
 
 Initially the destination looks the same as precopy, with a single thread
 reading the migration stream; the 'postcopy advise' and 'discard' commands
 are processed to change the way RAM is managed, but don't affect the stream
 processing.
 
-------------------------------------------------------------------------------
-                        1      2   3     4 5                      6   7
-main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
-thread                             |       |
-                                   |     (page request)
-                                   |        \___
-                                   v            \
-listen thread:                     --- page -- page -- page -- page -- page --
-
-                                   a   b        c
-------------------------------------------------------------------------------
-
-On receipt of CMD_PACKAGED (1)
-   All the data associated with the package - the ( ... ) section in the
-diagram - is read into memory, and the main thread recurses into
-qemu_loadvm_state_main to process the contents of the package (2)
-which contains commands (3,6) and devices (4...)
-
-On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
-a new thread (a) is started that takes over servicing the migration stream,
-while the main thread carries on loading the package.   It loads normal
-background page data (b) but if during a device load a fault happens (5) the
-returned page (c) is loaded by the listen thread allowing the main threads
-device load to carry on.
-
-The last thing in the CMD_PACKAGED is a 'RUN' command (6) letting the destination
-CPUs start running.
-At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour
-and is no longer used by migration, while the listen thread carries
-on servicing page data until the end of migration.
-
-=== Postcopy states ===
+::
+
+  ------------------------------------------------------------------------------
+                          1      2   3     4 5                      6   7
+  main -----DISCARD-CMD_PACKAGED ( LISTEN  DEVICE     DEVICE DEVICE RUN )
+  thread                             |       |
+                                     |     (page request)
+                                     |        \___
+                                     v            \
+  listen thread:                     --- page -- page -- page -- page -- page --
+
+                                     a   b        c

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 501

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +10 -0
+  ------------------------------------------------------------------------------
+
+- On receipt of CMD_PACKAGED (1)
+   All the data associated with the package - the ( ... ) section in the diagram - is read into memory, and the main thread recurses into qemu_loadvm_state_main to process the contents of the package (2) which contains commands (3,6) and devices (4...)
+
+- On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
+   a new thread (a) is started that takes over servicing the migration stream, while the main thread carries on loading the package.   It loads normal background page data (b) but if during a device load a fault happens (5) the returned page (c) is loaded by the listen thread allowing the main threads device load to carry on.
+
+- The last thing in the CMD_PACKAGED is a 'RUN' command (6)
+   letting the destination CPUs start running.  At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour and is no longer used by migration, while the listen thread carries on servicing page data until the end of migration.

(here the lines can be wrapped too?)

Quoted 32 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +8 -2
+
+Postcopy states
+---------------
 
 Postcopy moves through a series of states (see postcopy_state) from
 ADVISE->DISCARD->LISTEN->RUNNING->END
@@ -516,7 +539,8 @@ ADVISE->DISCARD->LISTEN->RUNNING->END
   End:     The listen thread can now quit, and perform the cleanup of migration
            state, the migration is now complete.
 
-=== Source side page maps ===
+Source side page maps
+---------------------
 
 The source side keeps two bitmaps during postcopy; 'the migration bitmap'
 and 'unsent map'.  The 'migration bitmap' is basically the same as in
@@ -529,6 +553,7 @@ The 'unsent map' is used for the transition to postcopy. It is a bitmap that
 has a bit cleared whenever a page is sent to the destination, however during
 the transition to postcopy mode it is combined with the migration bitmap
 to form a set of pages that:
+
    a) Have been sent but then redirtied (which must be discarded)
    b) Have not yet been sent - which also must be discarded to cause any
       transparent huge pages built during precopy to be broken.
@@ -540,7 +565,8 @@ request for a page that has already been sent is ignored.  Duplicate requests
 such as this can happen as a page is sent at about the same time the
 destination accesses it.
 
-=== Postcopy with hugepages ===
+Postcopy with hugepages
+-----------------------
 
Quoted 8 lines
Quoted 5 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
 Postcopy now works with hugetlbfs backed memory:
   a) The linux kernel on the destination must support userfault on hugepages.
Signature · 1 line
2.14.3
Signature · 1 line
Peter Xu

--
Dr. David Alan Gilbert / [email protected] / Manchester, UK

#4 lore raw
In reply to Kashyap Chamarthy
Quoted 11 lines · * Kashyap Chamarthy ([email protected]) wrote:
Quoted 9 lines · Dr. David Alan Gilbert (git)

From: "Dr. David Alan Gilbert" <[email protected]>

Mostly just manual conversion with very minor fixes.

---
 docs/devel/{migration.txt => migration.rst} | 326 +++++++++++++++-------------
 1 file changed, 176 insertions(+), 150 deletions(-)
 rename docs/devel/{migration.txt => migration.rst} (74%)
Quoted 7 lines

Hey Dave,

A good chunk of changes I suggest are pre-existing, so take it with a grain of salt :-). I'm ambivalent on whether to keep the conversion to rST and the other rST syntax changes separate. Maybe we can keep in one change, as this doesn't impact backports in terms of code / functionality.

I'd like to fix the non-rST fixes some other time; there's plenty wrong with the actual contents of this file.
(I need to pull in a fix from Jay Zhou before resending this actually).

Quoted 10 lines

[...]

Quoted 5 lines

+State Live Migration
+====================

This is used for RAM and block devices. It is not yet ported to vmstate. <Fill more information here>

Not this patch's problem, but do we have the more information that is to be filled in above?

Oh yes, plenty!

Quoted 9 lines

[...]

docs/devel/migration.rst pointer that is passed to all functions. · line 96

Quoted 4 lines

You can use any internal state that you need using the opaque void * pointer that is passed to all functions.

docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
@@ -83,7 +96,8 @@ pointer that is passed to all functions.
 The important functions for us are put_buffer()/get_buffer() that

Might want to highlight the fuctions (as you call them as important) with back ticks (and some spacing): put_buffer() / get_buffer().

OK, I've done a bunch of these highlights you suggest; one or two questions below.

Quoted 63 lines

docs/devel/migration.rst associated with a series of fields saved. The save_state always saves

Quoted 9 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 allow to write/read a buffer into the QEMUFile.
 
-=== How to save the state of one device ===
+Saving the state of one device
+==============================
 
 The state of a device is saved using intermediate buffers.  There are
 some helper functions to assist this saving.
@@ -97,30 +111,34 @@ associated with a series of fields saved.  The save_state always saves

Here too: s/save_state/``save_state``/

(And other occurrences throughout the doc.)

docs/devel/migration.rst associated with a series of fields saved. The save_state always saves · line 111

Quoted 1 line
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
 the state as the newer version.  But load_state sometimes is able to

s/load_state/``load_state``/

(And other occurrences throughout the doc.)

Quoted 10 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 load state from an older version.
 
-=== Legacy way ===
+Legacy way
+----------
 
 This way is going to disappear as soon as all current users are ported to VMSTATE.
 
 Each device has to register two functions, one to save the state and
 another to load the state back.

[...]

Quoted 4 lines

The important functions for the device state format are the save_state and load_state. Notice that load_state receives a version_id parameter to know what state format is receiving. save_state doesn't have a version_id parameter because it always uses the latest version.

In the above paragrah, too (``save_state``, ``load_state``).

Also s/version_id/'version_id'/g

Quoted 4 lines

-=== VMState ===
+VMState
+-------

[...]

Quoted 9 lines

We are declaring the state with name "pckbd".
The version_id is 3, and the fields are 4 uint8_t in a KBDState structure. We registered this with:

+.. code:: c
+

     vmstate_register(NULL, 0, &vmstate_kbd, s);

Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.

Not introduced in this patch, but: s/ids/IDs/

Two points:

Quoted 7 lines

- I think the above is a TODO item; so you can use:

.. TODO (dgilbert):: talk about how vmstate <-> qdev interact, and what the instance ids mean.

As that'll keep it only in the source rST; but not in the rendered version.

I'd rather leave the items rendered, otherwise they have even less chance of being fixed.

Quoted 116 lines

- Also, other minor thing (again, pre-existing): capitalize all the things like s/tcp/TCP; s/unix/UNIX/ where appropriate.

docs/devel/migration.rst Note: talk about how vmstate <-> qdev interact, and what the instance ids mean. · line 181

Quoted 2 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
@@ -159,7 +181,8 @@ Note: talk about how vmstate <-> qdev interact, and what the instance ids mean.
 You can search for VMSTATE_* macros for lots of types used in QEMU in

Maybe: s/VMSTATE_*/``VMSTATE_*``/

docs/devel/migration.rst function is deprecated and will be removed when no more users are left. · line 207

Quoted 11 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
 include/hw/hw.h.
 
-=== More about versions ===
+More about versions
+-------------------
 
 Version numbers are intended for major incompatible changes to the
 migration of a device, and using them breaks backwards-migration
@@ -183,7 +206,8 @@ function is deprecated and will be removed when no more users are left.
 Saving state will always create a section with the 'version_id' value
 and thus can't be loaded by any older QEMU.

Also, not in this patch, but keeping up with the theme of highlighting function names:

    s/load_state_old()/``load_state_old()``/

And other fields and strings (you can use a single quote, or a back tick for italics; whatever you choose, might want to keep it consistent):

    s/version_id/'version_id'/
Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +2 -1
-===  Massaging functions ===
+Massaging functions
+-------------------

[...]

Quoted 5 lines

-=== Subsections ===
+Subsections
+-----------

The use of version_id allows to be able to migrate from older versions

s/version_id/'version_id'/ (or version_id, if you prefer italics)

Again, not in this patch, but noticed in the rendered version locally:

s/post_load()/``post_load()``/

[...]

Quoted 2 lines

Here we have a subsection for the pio state. We only need to save/send this state when we are in the middle of a pio operation

s/ide_drive_pio_state_needed()/``ide_drive_pio_state_needed()``/

Quoted 1 line
(quoted diff) +0 -0
@@ -305,14 +331,11 @@ to send a subsection allows backwards migration compatibility when

[...]

Quoted 4 lines

+Not sending existing elements
+-----------------------------

Sometimes members of the VMState are no longer needed;

Pre-existing: s/;/:/

[...]

Quoted 2 lines

+Return path
+-----------

[...]

Pre-existing, and not this patch's problem. Noticed in my HTML render:
in this section, you might to highlight (double back ticks) the function:

``qemu_file_get_return_path(QEMUFile* fwdpath)``

[...]

Quoted 2 lines

+Postcopy
+========

[...]

Quoted 3 lines

-=== Enabling postcopy ===
+Enabling postcopy
+-----------------

[...]

Not in this patch, but in this section, 'migrate_set_speed' is mentioned; since it's a QMP command, you might want to highlith it:
``migrate_set_speed``

Also, from this section, you might want to use adminitions like:

.. note::
During the postcopy phase, the bandwidth limits set using migrate_set_speed is ignored (to avoid delaying requested pages that the destination is waiting for).

Use this wherever you see fit in the document.

[...]

Quoted 9 lines

Source behaviour
+----------------

Until postcopy is entered the migration stream is identical to normal precopy, except for the addition of a 'postcopy advise' command at

docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -0
@@ -423,11 +453,10 @@ the beginning, to tell the destination that postcopy might happen.
 When postcopy starts the source sends the page discard data and then
 forms the 'package' containing:
 
Quoted 5 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 427

Quoted 2 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -2
-   Command: 'postcopy listen'
-   The device state

In my local Sphinx-based HTML rendering, the above "The device state" ended up being bold somehow.

Any idea why? It's fine both in vim and rst2html-3's output.

Quoted 9 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 458

Quoted 6 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -3
-      A series of sections, identical to the precopy streams device state stream
-      containing everything except postcopiable devices (i.e. RAM)
-   Command: 'postcopy run'
+   - Command: 'postcopy listen'
+   - The device state
+      A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

Might want to wrap this long line (and other places); as it's causing an extra new line in my local rendering.

Yes, done (as per Peter's comments); it took me a while to figure out how to wrap them in lists.

Quoted 40 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 461

Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +1 -0
+   - Command: 'postcopy run'
 
 The 'package' is sent as the data part of a Command: 'CMD_PACKAGED', and the

s/CMD_PACKAGED/``CMD_PACKAGED``/g ?

[...]

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 501

Quoted 8 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +8 -0
+- On receipt of CMD_PACKAGED (1)
+   All the data associated with the package - the ( ... ) section in the diagram - is read into memory, and the main thread recurses into qemu_loadvm_state_main to process the contents of the package (2) which contains commands (3,6) and devices (4...)
+
+- On receipt of 'postcopy listen' - 3 -(i.e. the 1st command in the package)
+   a new thread (a) is started that takes over servicing the migration stream, while the main thread carries on loading the package.   It loads normal background page data (b) but if during a device load a fault happens (5) the returned page (c) is loaded by the listen thread allowing the main threads device load to carry on.
+
+- The last thing in the CMD_PACKAGED is a 'RUN' command (6)
+   letting the destination CPUs start running.  At the end of the CMD_PACKAGED (7) the main thread returns to normal running behaviour and is no longer used by migration, while the listen thread carries on servicing page data until the end of migration.

Super long lines, needs wrapping. Here too something seem to cause needless "bold" text in my Sphinx-based HTML rendering.

docs/devel/migration.rst to be sent quickly in the hope that those pages are likely to be used · line 504

Quoted 3 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +3 -0
+
+Postcopy states
+---------------

[...]

There's a list in this document. You might want to 'listify' it (using numbers, alphabets, etc), like you did it in other places.

[...]

Quoted 3 lines

-=== Postcopy with hugepages ===
+Postcopy with hugepages
+-----------------------

Here too, noticed after reading the rendered version; you might want to highlight the QEMU command-line parameters & file system paths using the verbatim (``):

Quoted 8 lines

-mem-path
/dev/hugepages
-mem-prealloc

[...]

Signature · 1 line
/kashyap

Thanks,

Dave

--
Dr. David Alan Gilbert / [email protected] / Manchester, UK

#5 lore raw
Quoted 2 lines · Dr. David Alan Gilbert
Quoted 1 line · * Kashyap Chamarthy ([email protected]) wrote:

On Tue, Dec 12, 2017 at 01:56:00PM +0000, Dr. David Alan Gilbert (git) wrote:

[...]

Quoted 8 lines
Quoted 5 lines

A good chunk of changes I suggest are pre-existing, so take it with a grain of salt :-). I'm ambivalent on whether to keep the conversion to rST and the other rST syntax changes separate. Maybe we can keep in one change, as this doesn't impact backports in terms of code / functionality.

I'd like to fix the non-rST fixes some other time; there's plenty wrong with the actual contents of this file.

That's fine.

Quoted 1 line

(I need to pull in a fix from Jay Zhou before resending this actually).

[...]

Quoted 4 lines
Quoted 2 lines

Not this patch's problem, but do we have the more information that is to be filled in above?

Oh yes, plenty!

Okay, I take it that you'd address them as a future patch.

[...]

Quoted 10 lines
Quoted 7 lines

- I think the above is a TODO item; so you can use:

.. TODO (dgilbert):: talk about how vmstate <-> qdev interact, and what the instance ids mean.

As that'll keep it only in the source rST; but not in the rendered version.

I'd rather leave the items rendered, otherwise they have even less chance of being fixed.

Yeah, I thought about it too; you're correct, a visual reminder works better.

[...]

Quoted 10 lines
Quoted 8 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 427

Quoted 5 lines

When postcopy starts the source sends the page discard data and then forms the 'package' containing:

docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -2
-   Command: 'postcopy listen'
-   The device state

In my local Sphinx-based HTML rendering, the above "The device state" ended up being bold somehow.

Any idea why? It's fine both in vim and rst2html-3's output.

Don't know yet. You can see here in the Sphinx-rendered HTML version:

https://kashyapc.fedorapeople.org/QEMU-Docs/_build/html/docs/migration.html

You can also see in the "Massaging functions" --> "Subsections" --> "For example;" also turned into bold.

Same for the "Postcopy states". And a couple of other places, where text turned into bold.

I'll let you know once I figure it out.

[...]

Signature · 1 line
/kashyap
#6 lore raw
Quoted 1 line · Kashyap Chamarthy

On Wed, Dec 13, 2017 at 08:11:46PM +0000, Dr. David Alan Gilbert wrote:

[...]

Quoted 9 lines
Quoted 7 lines
Quoted 5 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 427

Quoted 2 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -2
-   Command: 'postcopy listen'
-   The device state

In my local Sphinx-based HTML rendering, the above "The device state" ended up being bold somehow.

Any idea why? It's fine both in vim and rst2html-3's output.

Don't know yet. You can see here in the Sphinx-rendered HTML version:

So it's due to the indentation below the line "The device state". You can either put an extra blank line before indenting the next line:

- The device state:

A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

Or, just indent it to the same level:

- The device state: A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

Same applies for other places where bold text came in at unwanted places.

[...]

Signature · 1 line
/kashyap
#7 lore raw
Quoted 15 lines · * Kashyap Chamarthy ([email protected]) wrote:
Quoted 1 line · Kashyap Chamarthy

On Wed, Dec 13, 2017 at 08:11:46PM +0000, Dr. David Alan Gilbert wrote:

[...]

Quoted 9 lines
Quoted 7 lines
Quoted 5 lines

docs/devel/migration.rst the beginning, to tell the destination that postcopy might happen. · line 427

Quoted 2 lines
docs/devel/migration.rst renamed from docs/devel/migration.txt +0 -2
-   Command: 'postcopy listen'
-   The device state

In my local Sphinx-based HTML rendering, the above "The device state" ended up being bold somehow.

Any idea why? It's fine both in vim and rst2html-3's output.

Don't know yet. You can see here in the Sphinx-rendered HTML version:

Quoted 8 lines

So it's due to the indentation below the line "The device state". You can either put an extra blank line before indenting the next line:

- The device state:

A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

I think the problem here is the 'A' - it's picking it up as the first item in a numbered list; if I change that to 'The' it also fixes it. I added the extra blank line, it's OK but doesn't look great.

Quoted 7 lines

Or, just indent it to the same level:

- The device state: A series of sections, identical to the precopy streams device state stream containing everything except postcopiable devices (i.e. RAM)

Same applies for other places where bold text came in at unwanted places.

Dave

Quoted 4 lines

[...]

Signature · 1 line
/kashyap

--
Dr. David Alan Gilbert / [email protected] / Manchester, UK