[PULL 00/14] QAPI patches for 2026-06-16

[PULL 00/14] QAPI patches for 2026-06-16

[Markus Armbruster]

The following changes since commit 2f28d34ea0aead9830478cd1d3d0dd9d9191d82e:

  Merge tag 'pull-tcg-20260612' of https://gitlab.com/rth7680/qemu into staging (2026-06-13 14:02:34 -0400)

are available in the Git repository at:

  https://repo.or.cz/qemu/armbru.git tags/pull-qapi-2026-06-16

for you to fetch changes up to ebb49d4bb6bcc4f97401fabf3310937d44a2226b:

  qapi: add doc comment "Intro" section parsing (2026-06-16 15:36:08 +0200)

----------------------------------------------------------------
QAPI patches for 2026-06-16

----------------------------------------------------------------
Filip Hejsek (1):
      qapi: drop "must exist" from ID descriptions for consistency

John Snow (13):
      python: temporarily restrict max mypy version
      tests/qapi: generate output in source order
      qapi/parser: remove unused QAPIDoc subsection members
      qapi/parser: add has_features property
      qapi/parser: make remaining subsection members "private"
      qapi/parser: fix comment phrasing
      qapi: new doc comment "Intro" section
      qapi/parser: move _insert_near_kind() method
      qapi/parser: add mea culpa comment for ensure_returns
      qapi: adjust doc comment stub member insertion algorithm
      qapi: remove implicit doc comment Plain section
      qapi/qapidoc: add rendering for INTRO sections
      qapi: add doc comment "Intro" section parsing

 docs/devel/qapi-code-gen.rst   |  16 ++--
 docs/sphinx/qapidoc.py         |  61 +++++++++------
 qapi/char.json                 |   6 +-
 python/setup.cfg               |   3 +-
 scripts/qapi/parser.py         | 167 ++++++++++++++++++++++++-----------------
 tests/qapi-schema/doc-good.out |  86 +++++++++++----------
 tests/qapi-schema/test-qapi.py |  14 ++--
 7 files changed, 200 insertions(+), 153 deletions(-)

-- 
2.54.0

[PULL 01/14] qapi: drop "must exist" from ID descriptions for consistency

[Markus Armbruster]

From: Filip Hejsek <filip.hejsek@gmail.com>

Make chardev ID param descriptions more consistent with ID descriptions
elsewhere.

Signed-off-by: Filip Hejsek <filip.hejsek@gmail.com>
Message-ID: <20260609120221.461303-1-filip.hejsek@gmail.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 qapi/char.json | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/qapi/char.json b/qapi/char.json
index aa5ee9ffcd..abaccefa97 100644
--- a/qapi/char.json
+++ b/qapi/char.json
@@ -821,7 +821,7 @@
 #
 # Change a character device backend
 #
-# @id: the chardev's ID, must exist
+# @id: the chardev's ID
 #
 # @backend: new backend type and parameters
 #
@@ -862,7 +862,7 @@
 #
 # Remove a character device backend
 #
-# @id: the chardev's ID, must exist and not be in use
+# @id: the chardev's ID, must not be in use
 #
 # Since: 1.4
 #
@@ -879,7 +879,7 @@
 #
 # Send a break to a character device
 #
-# @id: the chardev's ID, must exist
+# @id: the chardev's ID
 #
 # Since: 2.10
 #
-- 
2.54.0

[PULL 02/14] python: temporarily restrict max mypy version

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

The newest versions of mypy do not support targeting Python 3.9, which
we still support. I want to address that soon, but in the meantime
it's nice if the tests pass.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-2-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 python/setup.cfg | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/python/setup.cfg b/python/setup.cfg
index 8097593397..ad4121a58d 100644
--- a/python/setup.cfg
+++ b/python/setup.cfg
@@ -40,7 +40,8 @@ devel =
     flake8 >= 5.0.4
     fusepy >= 2.0.4
     isort >= 5.6.0
-    mypy >= 1.4.0
+    # Max ver is temporary until we drop 3.9
+    mypy >= 1.4.0, <2.0.0
     pylint >= 2.17.3
     pylint != 3.2.4; python_version<"3.9"
     pytest >= 6.0.2
-- 
2.54.0

[PULL 03/14] tests/qapi: generate output in source order

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Rewrite the test doc generator to produce output in source order instead
of arbitrarily by section name.

This patch removes our last use of the "body" field, which has an
effect on how the sections of each test documention block are
printed. We now print the name of the section followed by the section
text for all sections except Members and Features, which are printed
as "Member=%s" or "Feature=%s" followed by the section text,
respectively.

This patch is motivated by a desire to move the QAPIDoc API away from
named fields for specific sections in a bid to force all users to simply
iterate through all_sections in order, instead - and to remove the named
subsections.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-3-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 tests/qapi-schema/doc-good.out | 82 +++++++++++++++++-----------------
 tests/qapi-schema/test-qapi.py | 14 +++---
 2 files changed, 48 insertions(+), 48 deletions(-)

diff --git a/tests/qapi-schema/doc-good.out b/tests/qapi-schema/doc-good.out
index 04a5507264..b9829e2f84 100644
--- a/tests/qapi-schema/doc-good.out
+++ b/tests/qapi-schema/doc-good.out
@@ -54,15 +54,15 @@ event EVT_BOXED Object
     boxed=True
     feature feat3
 doc freeform
-    body=
+    Plain
 *******
 Section
 *******
 doc freeform
-    body=
+    Plain
 Just text, no heading.
 doc freeform
-    body=
+    Plain
 Subsection
 ==========
 
@@ -106,84 +106,84 @@ Examples:
 - *verbatim*
 - {braces}
 doc symbol=Enum
-    body=
+    Plain
 
-    arg=one
+    Member=one
 The _one_ {and only}, description on the same line
-    arg=two
+    Member=two
 
-    feature=enum-feat
+    Feature=enum-feat
 Also _one_ {and only}
-    feature=enum-member-feat
+    Feature=enum-member-feat
 a member feature
-    section=Plain
+    Plain
 @two is undocumented
 doc symbol=Base
-    body=
+    Plain
 
-    arg=base1
+    Member=base1
  description starts on a new line,
  minimally indented
 doc symbol=Variant1
-    body=
+    Plain
 A paragraph
 
 Another paragraph
 
 @var1 is undocumented
-    arg=var1
+    Member=var1
 
-    feature=variant1-feat
+    Feature=variant1-feat
 a feature
-    feature=member-feat
+    Feature=member-feat
 a member feature
 doc symbol=Variant2
-    body=
+    Plain
 
 doc symbol=Object
-    body=
+    Plain
 
-    feature=union-feat1
+    Feature=union-feat1
 a feature
 doc symbol=Alternate
-    body=
+    Plain
 
-    arg=i
+    Member=i
 description starts on the same line
     remainder indented the same
     @b is undocumented
-    arg=b
+    Member=b
 
-    feature=alt-feat
+    Feature=alt-feat
 a feature
 doc freeform
-    body=
+    Plain
 Another subsection
 ==================
 doc symbol=cmd
-    body=
+    Plain
 
-    arg=arg1
+    Member=arg1
     description starts on a new line,
     indented
-    arg=arg2
+    Member=arg2
 description starts on the same line
     remainder indented differently
-    arg=arg3
+    Member=arg3
 
-    feature=cmd-feat1
+    Feature=cmd-feat1
 a feature
-    feature=cmd-feat2
+    Feature=cmd-feat2
 another feature
-    section=Plain
+    Plain
 .. note:: @arg3 is undocumented
-    section=Returns
+    Returns
 @Object
-    section=Errors
+    Errors
 some
-    section=Todo
+    Todo
 frobnicate
-    section=Plain
+    Plain
 .. admonition:: Notes
 
  - Lorem ipsum dolor sit amet
@@ -207,23 +207,23 @@ Examples::
 
 Note::
     Ceci n'est pas une note
-    section=Since
+    Since
 2.10
 doc symbol=cmd-boxed
-    body=
+    Plain
 If you're bored enough to read this, go see a video of boxed cats
-    feature=cmd-feat1
+    Feature=cmd-feat1
 a feature
-    feature=cmd-feat2
+    Feature=cmd-feat2
 another feature
-    section=Plain
+    Plain
 .. qmp-example::
 
    -> "this example"
 
    <- ... has no title ...
 doc symbol=EVT_BOXED
-    body=
+    Plain
 
-    feature=feat3
+    Feature=feat3
 a feature
diff --git a/tests/qapi-schema/test-qapi.py b/tests/qapi-schema/test-qapi.py
index cf7fb8a6df..5beb96e894 100755
--- a/tests/qapi-schema/test-qapi.py
+++ b/tests/qapi-schema/test-qapi.py
@@ -19,6 +19,7 @@
 from io import StringIO
 
 from qapi.error import QAPIError
+from qapi.parser import QAPIDoc
 from qapi.schema import QAPISchema, QAPISchemaVisitor
 
 
@@ -116,13 +117,12 @@ def test_frontend(fname):
             print('doc symbol=%s' % doc.symbol)
         else:
             print('doc freeform')
-        print('    body=\n%s' % doc.body.text)
-        for arg, section in doc.args.items():
-            print('    arg=%s\n%s' % (arg, section.text))
-        for feat, section in doc.features.items():
-            print('    feature=%s\n%s' % (feat, section.text))
-        for section in doc.sections:
-            print('    section=%s\n%s' % (section.kind, section.text))
+        for section in doc.all_sections:
+            if isinstance(section, QAPIDoc.ArgSection):
+                print('    %s=%s' % (section.kind, section.name))
+            else:
+                print('    %s' % section.kind)
+            print(section.text)
 
 
 def open_test_result(dir_name, file_name, update):
-- 
2.54.0

[PULL 04/14] qapi/parser: remove unused QAPIDoc subsection members

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

All users of the QAPIDoc object should be iterating over all_sections
and not grabbing arbitrary subsections, if possible. Remove the 'body'
and 'sections' members, as they are no longer used.

Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20260611042332.482979-4-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index c3cf33904e..b33edbba74 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -732,8 +732,6 @@ def __init__(self, info: QAPISourceInfo, symbol: Optional[str] = None):
         self.all_sections: List[QAPIDoc.Section] = [
             QAPIDoc.Section(info, QAPIDoc.Kind.PLAIN)
         ]
-        # the body section
-        self.body: Optional[QAPIDoc.Section] = self.all_sections[0]
         # dicts mapping parameter/feature names to their description
         self.args: Dict[str, QAPIDoc.ArgSection] = {}
         self.features: Dict[str, QAPIDoc.ArgSection] = {}
@@ -742,8 +740,6 @@ def __init__(self, info: QAPISourceInfo, symbol: Optional[str] = None):
         self.errors: Optional[QAPIDoc.Section] = None
         # "Since" section
         self.since: Optional[QAPIDoc.Section] = None
-        # sections other than .body, .args, .features
-        self.sections: List[QAPIDoc.Section] = []
 
     def end(self) -> None:
         for section in self.all_sections:
@@ -766,7 +762,6 @@ def ensure_untagged_section(self, info: QAPISourceInfo) -> None:
 
         # start new section
         section = self.Section(info, kind)
-        self.sections.append(section)
         self.all_sections.append(section)
 
     def new_tagged_section(
@@ -790,7 +785,6 @@ def new_tagged_section(
                 raise QAPISemError(
                     info, "duplicated '%s' section" % kind)
             self.since = section
-        self.sections.append(section)
         self.all_sections.append(section)
 
     def _new_description(
-- 
2.54.0

[PULL 05/14] qapi/parser: add has_features property

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Begin preparing to remove public access to the .features member by
introducing a semantic "has features" property instead.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-5-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index b33edbba74..dcc58170bd 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -555,7 +555,7 @@ def get_doc(self) -> 'QAPIDoc':
                     break
                 # Non-blank line, first of a section
                 if line == 'Features:':
-                    if doc.features:
+                    if doc.has_features:
                         raise QAPIParseError(
                             self, "duplicated 'Features:' line")
                     self.accept(False)
@@ -570,7 +570,7 @@ def get_doc(self) -> 'QAPIDoc':
                         if text:
                             doc.append_line(text)
                         line = self.get_doc_indented(doc)
-                    if not doc.features:
+                    if not doc.has_features:
                         raise QAPIParseError(
                             self, 'feature descriptions expected')
                     no_more_args = True
@@ -741,6 +741,10 @@ def __init__(self, info: QAPISourceInfo, symbol: Optional[str] = None):
         # "Since" section
         self.since: Optional[QAPIDoc.Section] = None
 
+    @property
+    def has_features(self) -> bool:
+        return bool(self.features)
+
     def end(self) -> None:
         for section in self.all_sections:
             section.text = section.text.strip('\n')
-- 
2.54.0

[PULL 06/14] qapi/parser: make remaining subsection members "private"

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

These fields are used to provide error checking and internal logistics
and should not be used by a user of the library to directly access
documentation sections, so make them private.

The "since" field alone is left public, as the qapidoc generator does
use this field to pull that section out of the regular flow of the
document.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-6-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 48 +++++++++++++++++++++---------------------
 1 file changed, 24 insertions(+), 24 deletions(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index dcc58170bd..da4756a742 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -733,17 +733,17 @@ def __init__(self, info: QAPISourceInfo, symbol: Optional[str] = None):
             QAPIDoc.Section(info, QAPIDoc.Kind.PLAIN)
         ]
         # dicts mapping parameter/feature names to their description
-        self.args: Dict[str, QAPIDoc.ArgSection] = {}
-        self.features: Dict[str, QAPIDoc.ArgSection] = {}
+        self._args: Dict[str, QAPIDoc.ArgSection] = {}
+        self._features: Dict[str, QAPIDoc.ArgSection] = {}
         # a command's "Returns" and "Errors" section
-        self.returns: Optional[QAPIDoc.Section] = None
-        self.errors: Optional[QAPIDoc.Section] = None
+        self._returns: Optional[QAPIDoc.Section] = None
+        self._errors: Optional[QAPIDoc.Section] = None
         # "Since" section
         self.since: Optional[QAPIDoc.Section] = None
 
     @property
     def has_features(self) -> bool:
-        return bool(self.features)
+        return bool(self._features)
 
     def end(self) -> None:
         for section in self.all_sections:
@@ -775,15 +775,15 @@ def new_tagged_section(
     ) -> None:
         section = self.Section(info, kind)
         if kind == QAPIDoc.Kind.RETURNS:
-            if self.returns:
+            if self._returns:
                 raise QAPISemError(
                     info, "duplicated '%s' section" % kind)
-            self.returns = section
+            self._returns = section
         elif kind == QAPIDoc.Kind.ERRORS:
-            if self.errors:
+            if self._errors:
                 raise QAPISemError(
                     info, "duplicated '%s' section" % kind)
-            self.errors = section
+            self._errors = section
         elif kind == QAPIDoc.Kind.SINCE:
             if self.since:
                 raise QAPISemError(
@@ -807,16 +807,16 @@ def _new_description(
         desc[name] = section
 
     def new_argument(self, info: QAPISourceInfo, name: str) -> None:
-        self._new_description(info, name, QAPIDoc.Kind.MEMBER, self.args)
+        self._new_description(info, name, QAPIDoc.Kind.MEMBER, self._args)
 
     def new_feature(self, info: QAPISourceInfo, name: str) -> None:
-        self._new_description(info, name, QAPIDoc.Kind.FEATURE, self.features)
+        self._new_description(info, name, QAPIDoc.Kind.FEATURE, self._features)
 
     def append_line(self, line: str) -> None:
         self.all_sections[-1].append_line(line)
 
     def connect_member(self, member: 'QAPISchemaMember') -> None:
-        if member.name not in self.args:
+        if member.name not in self._args:
             assert member.info
             if self.symbol not in member.info.pragma.documentation_exceptions:
                 raise QAPISemError(member.info,
@@ -827,7 +827,7 @@ def connect_member(self, member: 'QAPISchemaMember') -> None:
 
             section = QAPIDoc.ArgSection(
                 self.info, QAPIDoc.Kind.MEMBER, member.name)
-            self.args[member.name] = section
+            self._args[member.name] = section
 
             # Determine where to insert stub doc - it should go at the
             # end of the members section(s), if any. Note that index 0
@@ -839,14 +839,14 @@ def connect_member(self, member: 'QAPISchemaMember') -> None:
                     index += 1
             self.all_sections.insert(index, section)
 
-        self.args[member.name].connect(member)
+        self._args[member.name].connect(member)
 
     def connect_feature(self, feature: 'QAPISchemaFeature') -> None:
-        if feature.name not in self.features:
+        if feature.name not in self._features:
             raise QAPISemError(feature.info,
                                "feature '%s' lacks documentation"
                                % feature.name)
-        self.features[feature.name].connect(feature)
+        self._features[feature.name].connect(feature)
 
     def ensure_returns(self, info: QAPISourceInfo) -> None:
 
@@ -887,18 +887,18 @@ def _insert_near_kind(
 
     def check_expr(self, expr: QAPIExpression) -> None:
         if 'command' in expr:
-            if self.returns and 'returns' not in expr:
+            if self._returns and 'returns' not in expr:
                 raise QAPISemError(
-                    self.returns.info,
+                    self._returns.info,
                     "'Returns' section, but command doesn't return anything")
         else:
-            if self.returns:
+            if self._returns:
                 raise QAPISemError(
-                    self.returns.info,
+                    self._returns.info,
                     "'Returns' section is only valid for commands")
-            if self.errors:
+            if self._errors:
                 raise QAPISemError(
-                    self.errors.info,
+                    self._errors.info,
                     "'Errors' section is only valid for commands")
 
     def check(self) -> None:
@@ -918,5 +918,5 @@ def check_args_section(
                         "do" if len(bogus) > 1 else "does"
                     ))
 
-        check_args_section(self.args, 'member')
-        check_args_section(self.features, 'feature')
+        check_args_section(self._args, 'member')
+        check_args_section(self._features, 'feature')
-- 
2.54.0

[PULL 07/14] qapi/parser: fix comment phrasing

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Plaintext sections can be one or more paragraphs, the original comment
was a mistake.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-7-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index da4756a742..d8068bb1bf 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -631,7 +631,7 @@ def get_doc(self) -> 'QAPIDoc':
                     line = self.get_doc_indented(doc)
                     no_more_args = True
                 else:
-                    # plain paragraph
+                    # plain paragraph(s)
                     doc.ensure_untagged_section(self.info)
                     doc.append_line(line)
                     line = self.get_doc_paragraph(doc)
-- 
2.54.0

[PULL 08/14] qapi: new doc comment "Intro" section

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

This patch adds an "Intro" doc section, intended to eventually replace
the "Plain" doc section alongside a forthcoming "Details" section.

For now, this section is not actually instantiated or used, but
subsequent patches will slowly convert the leading introductory sections
of QAPIDoc documentation to use this new section.

A main motivation of this series of changes is to more explicitly
delineate the "Introductory" documentation for each QAPI definition for
the sake of the inliner. When inlining members, examples, and details
from multiple QAPIDoc sections, we will want to omit the "Introductory"
text from inlined definitions while keeping notes, caution boxes,
examples, and so on. This new section facilitates that.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-8-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 docs/sphinx/qapidoc.py | 2 +-
 scripts/qapi/parser.py | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/docs/sphinx/qapidoc.py b/docs/sphinx/qapidoc.py
index c2f09bac16..1f7c15b707 100644
--- a/docs/sphinx/qapidoc.py
+++ b/docs/sphinx/qapidoc.py
@@ -368,7 +368,7 @@ def visit_sections(self, ent: QAPISchemaDefinition) -> None:
         for i, section in enumerate(sections):
             section.text = self.reformat_arobase(section.text)
 
-            if section.kind == QAPIDoc.Kind.PLAIN:
+            if section.kind.name in ("PLAIN", "INTRO"):
                 self.visit_paragraph(section)
             elif section.kind == QAPIDoc.Kind.MEMBER:
                 assert isinstance(section, QAPIDoc.ArgSection)
diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index d8068bb1bf..97e7dacb0f 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -681,6 +681,7 @@ class Kind(enum.Enum):
         ERRORS = 4
         SINCE = 5
         TODO = 6
+        INTRO = 7
 
         @staticmethod
         def from_string(kind: str) -> 'QAPIDoc.Kind':
@@ -748,7 +749,7 @@ def has_features(self) -> bool:
     def end(self) -> None:
         for section in self.all_sections:
             section.text = section.text.strip('\n')
-            if section.kind != QAPIDoc.Kind.PLAIN and section.text == '':
+            if not (section.kind.name in ("INTRO", "PLAIN") or section.text):
                 raise QAPISemError(
                     section.info, "text required after '%s:'" % section.kind)
 
-- 
2.54.0

[PULL 09/14] qapi/parser: move _insert_near_kind() method

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Move this function out from underneath `ensure_returns` and make it
available for general purpose use as an object method instead. This is
purely a scope-level patch with no functional changes.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-9-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 33 +++++++++++++++++----------------
 1 file changed, 17 insertions(+), 16 deletions(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index 97e7dacb0f..261f8ba9f8 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -816,6 +816,22 @@ def new_feature(self, info: QAPISourceInfo, name: str) -> None:
     def append_line(self, line: str) -> None:
         self.all_sections[-1].append_line(line)
 
+    def _insert_near_kind(
+        self,
+        kind: 'QAPIDoc.Kind',
+        new_sect: 'QAPIDoc.Section',
+        after: bool = False,
+    ) -> bool:
+        """Insert or append a new doc section at a specific point."""
+        for idx, sect in enumerate(reversed(self.all_sections)):
+            if sect.kind == kind:
+                pos = len(self.all_sections) - idx - 1
+                if after:
+                    pos += 1
+                self.all_sections.insert(pos, new_sect)
+                return True
+        return False
+
     def connect_member(self, member: 'QAPISchemaMember') -> None:
         if member.name not in self._args:
             assert member.info
@@ -850,28 +866,13 @@ def connect_feature(self, feature: 'QAPISchemaFeature') -> None:
         self._features[feature.name].connect(feature)
 
     def ensure_returns(self, info: QAPISourceInfo) -> None:
-
-        def _insert_near_kind(
-            kind: QAPIDoc.Kind,
-            new_sect: QAPIDoc.Section,
-            after: bool = False,
-        ) -> bool:
-            for idx, sect in enumerate(reversed(self.all_sections)):
-                if sect.kind == kind:
-                    pos = len(self.all_sections) - idx - 1
-                    if after:
-                        pos += 1
-                    self.all_sections.insert(pos, new_sect)
-                    return True
-            return False
-
         if any(s.kind == QAPIDoc.Kind.RETURNS for s in self.all_sections):
             return
 
         # Stub "Returns" section for undocumented returns value
         stub = QAPIDoc.Section(info, QAPIDoc.Kind.RETURNS)
 
-        if any(_insert_near_kind(kind, stub, after) for kind, after in (
+        if any(self._insert_near_kind(kind, stub, after) for kind, after in (
                 # 1. If arguments, right after those.
                 (QAPIDoc.Kind.MEMBER, True),
                 # 2. Elif errors, right *before* those.
-- 
2.54.0

[PULL 10/14] qapi/parser: add mea culpa comment for ensure_returns

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

These algorithms are quite a mess currently, but they are temporary
until we add the inliner which will address these issues more
holistically. For now, add the "mea culpa".

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-10-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index 261f8ba9f8..b0cead38b1 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -866,6 +866,13 @@ def connect_feature(self, feature: 'QAPISchemaFeature') -> None:
         self._features[feature.name].connect(feature)
 
     def ensure_returns(self, info: QAPISourceInfo) -> None:
+        # This is more complicated than it ought to be.  The doc
+        # parser should already know where a generated RETURNS section
+        # should go.  It currently doesn't, mostly because it accepts
+        # tagged sections in any order.
+        #
+        # TODO: Tighten doc syntax and simplify.
+
         if any(s.kind == QAPIDoc.Kind.RETURNS for s in self.all_sections):
             return
 
-- 
2.54.0

[PULL 11/14] qapi: adjust doc comment stub member insertion algorithm

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

A forthcoming patch removes the implicit PLAIN section that always
starts a QAPIDoc section list. Further future changes begin converting
"PLAIN" sections to "INTRO" sections.

This will affect the code that inserts "Not documented" descriptions
for undocumented members ("stub sections") and the dummy section that
marks the spot for "The members of ..." references.

Adjust the algorithm to cope with not only the finished state, but
temporary intermediate states while the series is merged.

This algorithm can handle zero-or-more PLAIN *or* INTRO sections at
the beginning of a QAPIDoc object, in contrast to the previous
algorithm which assumed and relied upon there being always one PLAIN
section at the beginning of every QAPIDoc section list.

In other words: (PLAIN | INTRO)* <EverythingElse>

This does not impact what the parser itself will actually produce. As
of this patch, the parser will still always generate QAPIDoc section
lists that start with precisely one PLAIN section (whether or not it
is empty), followed by the remaining sections. Those remaining
sections may or may not include additional PLAIN sections, but never
two such sections contiguously as the parser will always treat that
layout as one PLAIN section consisting of multiple paragraph(s).

In other other words: This insertion algorithm is more lenient than
the parser, but this is on purpose for flexibility mid-stream as we
convert QAPI to using explicit introductory sections. The allowed
order of sections will eventually become strictly enforced in the
parser, which will in turn allow dramatic simplifications to the
insertion algorithm. This only exists as transitory code until we are
able to enforce that order.

Fear not: the intermediate ReST output before and after this patch
are byte identical, so failing all else, we at least know it doesn't
make anything worse.

Lastly, because we have three places in the code that need to insert
stub/dummy sections, we take the opportunity to consolidate this code
to handle all three cases with one function. This winds up
necessitating the qapidoc.py generator actually modify the section
list to insert a "dummy" member that acts as a placeholder for "The
members of ..." text. While it looks like a code smell to modify the
caller's argument, it is ultimately safe because the QAPI Schema
object is re-parsed and re-constructed in memory for each individual
process that needs to operate on it. In other words, the Sphinx
document generator already does have "its own copy" of the section
lists, so it is "safe" to modify here without regards to other
consumers of the QAPIDoc objects. It only *looks* like it smells
bad. Ultimately, this code will also be removed once the inliner is
merged, so it is only a temporary aesthetic issue regardless.

That's my story and I'm sticking to it.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-11-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 docs/sphinx/qapidoc.py | 42 ++++++++++++++++--------------
 scripts/qapi/parser.py | 58 +++++++++++++++++++++++++++++-------------
 2 files changed, 64 insertions(+), 36 deletions(-)

diff --git a/docs/sphinx/qapidoc.py b/docs/sphinx/qapidoc.py
index 1f7c15b707..16ad15fe94 100644
--- a/docs/sphinx/qapidoc.py
+++ b/docs/sphinx/qapidoc.py
@@ -349,30 +349,38 @@ def _get_target(
                 )
 
     def visit_sections(self, ent: QAPISchemaDefinition) -> None:
+        # Generate a placeholder right after the member section(s) which
+        # may be used to generate documentation for "The members of..."
+        # pointers in the rendered document.
+        #
+        # This is a temporary hack until the inliner is merged. Note
+        # that although we modify the caller's section list, the
+        # Sphinx document generator has its own copy of the parsed
+        # schema in memory, so this action does not interfere with
+        # other users of the QAPISchema or QAPIDoc objects outside of
+        # the document generator. Fishy, but not harmful.
+        if ent.doc:
+            ent.doc.append_member_stub(
+                QAPIDoc.ArgSection(
+                    ent.doc.info, QAPIDoc.Kind.MEMBER, "q_dummy"
+                )
+            )
+
         sections = ent.doc.all_sections if ent.doc else []
 
-        # Determine the index location at which we should generate
-        # documentation for "The members of ..." pointers. This should
-        # go at the end of the members section(s) if any. Note that
-        # index 0 is assumed to be a plain intro section, even if it is
-        # empty; and that a members section if present will always
-        # immediately follow the opening PLAIN section.
-        gen_index = 1
-        if len(sections) > 1:
-            while sections[gen_index].kind == QAPIDoc.Kind.MEMBER:
-                gen_index += 1
-                if gen_index >= len(sections):
-                    break
-
         # Add sections in source order:
-        for i, section in enumerate(sections):
+        for section in sections:
             section.text = self.reformat_arobase(section.text)
 
             if section.kind.name in ("PLAIN", "INTRO"):
                 self.visit_paragraph(section)
             elif section.kind == QAPIDoc.Kind.MEMBER:
                 assert isinstance(section, QAPIDoc.ArgSection)
-                self.visit_member(section)
+                if section.name == "q_dummy":
+                    # Generate "The members of ..." entries if necessary
+                    self._insert_member_pointer(ent)
+                else:
+                    self.visit_member(section)
             elif section.kind == QAPIDoc.Kind.FEATURE:
                 assert isinstance(section, QAPIDoc.ArgSection)
                 self.visit_feature(section)
@@ -386,10 +394,6 @@ def visit_sections(self, ent: QAPISchemaDefinition) -> None:
             else:
                 assert False
 
-            # Generate "The members of ..." entries if necessary:
-            if i == gen_index - 1:
-                self._insert_member_pointer(ent)
-
         self.ensure_blank_line()
 
     # Transmogrification core methods
diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index b0cead38b1..09720a2c27 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -832,6 +832,42 @@ def _insert_near_kind(
                 return True
         return False
 
+    def _insert_after_intro(
+        self,
+        section: 'QAPIDoc.Section',
+    ) -> None:
+        """
+        Insert a section immediately after the intro section.
+
+        While we convert PLAIN sections to INTRO sections, all
+        contiguous INTRO/PLAIN sections at the start of a QAPIDoc
+        section list are treated as "the intro".
+
+        Once INTRO conversion is complete, this helper will no longer be
+        needed and ``_insert_near_kind(QAPIDoc.Kind.INTRO, ...)`` will
+        be sufficient.
+        """
+        index = 0
+        for index, ref_section in enumerate(self.all_sections):
+            if ref_section.kind.name in ("PLAIN", "INTRO"):
+                continue
+            break
+        else:
+            index += 1
+
+        self.all_sections.insert(index, section)
+
+    def append_member_stub(self, stub: 'QAPIDoc.Section') -> None:
+
+        """
+        Append a stub section after any Member sections.
+        """
+        if self._insert_near_kind(QAPIDoc.Kind.MEMBER, stub, True):
+            return
+
+        # No MEMBER sections present. Insert after INTRO/PLAIN sections.
+        self._insert_after_intro(stub)
+
     def connect_member(self, member: 'QAPISchemaMember') -> None:
         if member.name not in self._args:
             assert member.info
@@ -841,20 +877,10 @@ def connect_member(self, member: 'QAPISchemaMember') -> None:
                                    % (member.role, member.name))
             # Insert stub documentation section for missing member docs.
             # TODO: drop when undocumented members are outlawed
-
-            section = QAPIDoc.ArgSection(
+            stub_section = QAPIDoc.ArgSection(
                 self.info, QAPIDoc.Kind.MEMBER, member.name)
-            self._args[member.name] = section
-
-            # Determine where to insert stub doc - it should go at the
-            # end of the members section(s), if any. Note that index 0
-            # is assumed to be an untagged intro section, even if it is
-            # empty.
-            index = 1
-            if len(self.all_sections) > 1:
-                while self.all_sections[index].kind == QAPIDoc.Kind.MEMBER:
-                    index += 1
-            self.all_sections.insert(index, section)
+            self._args[member.name] = stub_section
+            self.append_member_stub(stub_section)
 
         self._args[member.name].connect(member)
 
@@ -889,10 +915,8 @@ def ensure_returns(self, info: QAPISourceInfo) -> None:
         )):
             return
 
-        # Otherwise, it should go right after the intro. The intro
-        # is always the first section and is always present (even
-        # when empty), so we can insert directly at index=1 blindly.
-        self.all_sections.insert(1, stub)
+        # Otherwise, it should go right after the intro.
+        self._insert_after_intro(stub)
 
     def check_expr(self, expr: QAPIExpression) -> None:
         if 'command' in expr:
-- 
2.54.0

[PULL 12/14] qapi: remove implicit doc comment Plain section

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Prior to this patch, we always create an empty Plain section. Removing
this allows us to gradually phase out the "Plain" section in favor of
"Intro" and "Details" sections while keeping "Plain" around for the
interim churn during the series - meaning that we don't actually know
at __init__ time which type of section we'll have first.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-12-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 scripts/qapi/parser.py         |  4 +---
 tests/qapi-schema/doc-good.out | 14 --------------
 2 files changed, 1 insertion(+), 17 deletions(-)

diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index 09720a2c27..b7a7b9465a 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -730,9 +730,7 @@ def __init__(self, info: QAPISourceInfo, symbol: Optional[str] = None):
         # definition doc's symbol, None for free-form doc
         self.symbol: Optional[str] = symbol
         # the sections in textual order
-        self.all_sections: List[QAPIDoc.Section] = [
-            QAPIDoc.Section(info, QAPIDoc.Kind.PLAIN)
-        ]
+        self.all_sections: List[QAPIDoc.Section] = []
         # dicts mapping parameter/feature names to their description
         self._args: Dict[str, QAPIDoc.ArgSection] = {}
         self._features: Dict[str, QAPIDoc.ArgSection] = {}
diff --git a/tests/qapi-schema/doc-good.out b/tests/qapi-schema/doc-good.out
index b9829e2f84..16f4422177 100644
--- a/tests/qapi-schema/doc-good.out
+++ b/tests/qapi-schema/doc-good.out
@@ -106,8 +106,6 @@ Examples:
 - *verbatim*
 - {braces}
 doc symbol=Enum
-    Plain
-
     Member=one
 The _one_ {and only}, description on the same line
     Member=two
@@ -119,8 +117,6 @@ a member feature
     Plain
 @two is undocumented
 doc symbol=Base
-    Plain
-
     Member=base1
  description starts on a new line,
  minimally indented
@@ -138,16 +134,10 @@ a feature
     Feature=member-feat
 a member feature
 doc symbol=Variant2
-    Plain
-
 doc symbol=Object
-    Plain
-
     Feature=union-feat1
 a feature
 doc symbol=Alternate
-    Plain
-
     Member=i
 description starts on the same line
     remainder indented the same
@@ -161,8 +151,6 @@ doc freeform
 Another subsection
 ==================
 doc symbol=cmd
-    Plain
-
     Member=arg1
     description starts on a new line,
     indented
@@ -223,7 +211,5 @@ another feature
 
    <- ... has no title ...
 doc symbol=EVT_BOXED
-    Plain
-
     Feature=feat3
 a feature
-- 
2.54.0

[PULL 13/14] qapi/qapidoc: add rendering for INTRO sections

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Amend the qapidoc generator to handle and render INTRO sections.

The only real difference here from other sections is that we need to
dedent the text so it renders correctly. Members and Features are also
indented, but do not require a dedent() because they are always used
in tandem with an rST construct that forms the start of a new indented
block; there is coincidental harmony.

Plaintext sections, however, do not start their own block and thus
need to be dedented to prevent accidentally rendering them as a
blockquote or a syntax error.

This dedent transformation on the text does not reflow the text, so
source line information remains accurate, and the "blame" chain of
custody for sphinx rST parsing error messages continues to be correct
even through this transformation.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-13-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 docs/sphinx/qapidoc.py | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/docs/sphinx/qapidoc.py b/docs/sphinx/qapidoc.py
index 16ad15fe94..54a32e45f7 100644
--- a/docs/sphinx/qapidoc.py
+++ b/docs/sphinx/qapidoc.py
@@ -35,6 +35,7 @@
 from pathlib import Path
 import re
 import sys
+import textwrap
 from typing import TYPE_CHECKING
 
 from docutils import nodes
@@ -150,8 +151,15 @@ def add_lines(
         self,
         content: str,
         info: QAPISourceInfo,
+        dedent: bool = False,
     ) -> None:
         lines = content.splitlines(True)
+
+        if dedent:
+            lines = textwrap.dedent(content).splitlines(True)
+        else:
+            lines = content.splitlines(True)
+
         for i, line in enumerate(lines):
             self.add_line_raw(line, info.fname, info.line + i)
 
@@ -223,13 +231,16 @@ def reformat_arobase(text: str) -> str:
 
     # Transmogrification helpers
 
-    def visit_paragraph(self, section: QAPIDoc.Section) -> None:
+    def visit_plaintext(self, section: QAPIDoc.Section) -> None:
         # Squelch empty paragraphs.
         if not section.text:
             return
 
+        # Intro sections, which are indented in QAPI source, need to
+        # be dedented to avoid accidental block quotes in ReST syntax.
+        dedent = bool(section.kind == QAPIDoc.Kind.INTRO)
         self.ensure_blank_line()
-        self.add_lines(section.text, section.info)
+        self.add_lines(section.text, section.info, dedent)
         self.ensure_blank_line()
 
     def visit_member(self, section: QAPIDoc.ArgSection) -> None:
@@ -373,7 +384,7 @@ def visit_sections(self, ent: QAPISchemaDefinition) -> None:
             section.text = self.reformat_arobase(section.text)
 
             if section.kind.name in ("PLAIN", "INTRO"):
-                self.visit_paragraph(section)
+                self.visit_plaintext(section)
             elif section.kind == QAPIDoc.Kind.MEMBER:
                 assert isinstance(section, QAPIDoc.ArgSection)
                 if section.name == "q_dummy":
-- 
2.54.0

[PULL 14/14] qapi: add doc comment "Intro" section parsing

[Markus Armbruster]

From: John Snow <jsnow@redhat.com>

Add parsing for explicit Intro section syntax.

A side effect of this patch is that we will (currently) always create
an empty Intro section, similar to how we used to have an empty Plain
section. The tests are adjusted accordingly, rendered document output
does not change at all.

Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20260611042332.482979-14-jsnow@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
---
 docs/devel/qapi-code-gen.rst   | 16 +++++++---------
 scripts/qapi/parser.py         |  4 ++--
 tests/qapi-schema/doc-good.out | 18 ++++++++++++++++++
 3 files changed, 27 insertions(+), 11 deletions(-)

diff --git a/docs/devel/qapi-code-gen.rst b/docs/devel/qapi-code-gen.rst
index 3a632b4a64..b1cc5b5f0d 100644
--- a/docs/devel/qapi-code-gen.rst
+++ b/docs/devel/qapi-code-gen.rst
@@ -984,11 +984,11 @@ definition it documents.
 When documentation is required (see pragma_ 'doc-required'), every
 definition must have documentation.
 
-Definition documentation starts with a line naming the definition,
-followed by an optional overview, a description of each argument (for
-commands and events), member (for structs and unions), branch (for
-alternates), or value (for enums), a description of each feature (if
-any), and finally optional tagged sections.
+Definition documentation starts with a description naming the
+definition with an optional indented overview, a description of each
+argument (for commands and events), member (for structs and unions),
+branch (for alternates), or value (for enums), a description of each
+feature (if any), and finally optional tagged sections.
 
 Descriptions start with '\@name:'.  The description text must be
 indented like this::
@@ -1093,8 +1093,7 @@ Examples of complete definition documentation::
 
  ##
  # @BlockStats:
- #
- # Statistics of a virtual block device or a block backing device.
+ #     Statistics of a virtual block device or a block backing device.
  #
  # @device: If the stats are for a virtual block device, the name
  #     corresponding to the virtual block device.
@@ -1111,8 +1110,7 @@ Examples of complete definition documentation::
 
  ##
  # @query-blockstats:
- #
- # Query the @BlockStats for all virtual block devices.
+ #     Query the @BlockStats for all virtual block devices.
  #
  # @query-nodes: If true, the command will query all the block nodes
  #     ... explain, explain ...
diff --git a/scripts/qapi/parser.py b/scripts/qapi/parser.py
index b7a7b9465a..9e14c2f792 100644
--- a/scripts/qapi/parser.py
+++ b/scripts/qapi/parser.py
@@ -542,8 +542,8 @@ def get_doc(self) -> 'QAPIDoc':
             if not symbol:
                 raise QAPIParseError(self, "name required after '@'")
             doc = QAPIDoc(info, symbol)
-            self.accept(False)
-            line = self.get_doc_line()
+            doc.all_sections.append(QAPIDoc.Section(info, QAPIDoc.Kind.INTRO))
+            line = self.get_doc_indented(doc)
             no_more_args = False
 
             while line is not None:
diff --git a/tests/qapi-schema/doc-good.out b/tests/qapi-schema/doc-good.out
index 16f4422177..371dd25ffc 100644
--- a/tests/qapi-schema/doc-good.out
+++ b/tests/qapi-schema/doc-good.out
@@ -106,6 +106,8 @@ Examples:
 - *verbatim*
 - {braces}
 doc symbol=Enum
+    Intro
+
     Member=one
 The _one_ {and only}, description on the same line
     Member=two
@@ -117,10 +119,14 @@ a member feature
     Plain
 @two is undocumented
 doc symbol=Base
+    Intro
+
     Member=base1
  description starts on a new line,
  minimally indented
 doc symbol=Variant1
+    Intro
+
     Plain
 A paragraph
 
@@ -134,10 +140,16 @@ a feature
     Feature=member-feat
 a member feature
 doc symbol=Variant2
+    Intro
+
 doc symbol=Object
+    Intro
+
     Feature=union-feat1
 a feature
 doc symbol=Alternate
+    Intro
+
     Member=i
 description starts on the same line
     remainder indented the same
@@ -151,6 +163,8 @@ doc freeform
 Another subsection
 ==================
 doc symbol=cmd
+    Intro
+
     Member=arg1
     description starts on a new line,
     indented
@@ -198,6 +212,8 @@ Note::
     Since
 2.10
 doc symbol=cmd-boxed
+    Intro
+
     Plain
 If you're bored enough to read this, go see a video of boxed cats
     Feature=cmd-feat1
@@ -211,5 +227,7 @@ another feature
 
    <- ... has no title ...
 doc symbol=EVT_BOXED
+    Intro
+
     Feature=feat3
 a feature
-- 
2.54.0

Re: [PULL 00/14] QAPI patches for 2026-06-16

[Stefan Hajnoczi]

[-- Attachment #1: Type: text/plain, Size: 116 bytes --]

Applied, thanks.

Please update the changelog at https://wiki.qemu.org/ChangeLog/11.1 for any user-visible changes.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]