Commit Graph
340 Commits
Author SHA1 Message Date
Markus Armbruster 4e99f4b12c qapi: Drop simple unions
Simple unions predate flat unions.  Having both complicates the QAPI
schema language and the QAPI generator.  We haven't been using simple
unions in new code for a long time, because they are less flexible and
somewhat awkward on the wire.

The previous commits eliminated simple union from the tree.  Now drop
them from the QAPI schema language entirely, and update mentions of
"flat union" to just "union".

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210917143134.412106-22-armbru@redhat.com>
2021-09-27 08:23:25 +02:00
Markus Armbruster 8ebc3120e1 qapi: Stop enforcing "type name should not end in 'Kind'
I'm about to convert simple unions to flat unions, then drop simple
union support.  The conversion involves making the implict enum types
explicit.  To reduce churn, I'd like to name them exactly like the
implicit types they replace.  However, these names are reserved for
the generator's use.  They won't be once simple unions are gone.  Stop
enforcing this naming rule now rather than then.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-Id: <20210917143134.412106-3-armbru@redhat.com>
2021-09-25 06:55:54 +02:00
Markus Armbruster 62f27589f8 qapi: Fix bogus error for 'if': { 'not': '' }
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210908045428.2689093-6-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
[check_infix()'s type hint fixed]
2021-09-08 15:30:30 +02:00
Markus Armbruster e2ff14a574 qapi: Bury some unused code in class Indentation
.__int__() has never been used.  Drop it.

.decrease() raises ArithmeticError when asked to decrease indentation
level below zero.  Nothing catches it.  It's a programming error.
Dumb down to assert.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210908045428.2689093-4-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-08 15:30:10 +02:00
Markus Armbruster 916fca17c7 qapi: Drop Indentation.__bool__()
Intentation.__bool__() is not worth its keep: it has just one user,
which can just as well check .__str__() instead.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210908045428.2689093-3-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-08 15:30:10 +02:00
Markus Armbruster 7b275cdd69 qapi: Fix a botched type annotation
Mypy is unhappy:

    $ mypy --config-file=scripts/qapi/mypy.ini `git-ls-files scripts/qapi/\*py`
    scripts/qapi/common.py:208: error: Function is missing a return type annotation
    scripts/qapi/common.py:227: error: Returning Any from function declared to return "str"

Messed up in commit ccea6a8637 "qapi: Factor common recursion out of
cgen_ifcond(), docgen_ifcond()".  Tidy up.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210908045428.2689093-2-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-08 15:30:10 +02:00
Markus Armbruster 34f7b25e57 qapi: Tweak error messages for unknown / conflicting 'if' keys
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-13-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster 6dcf03719a qapi: Tweak error messages for missing / conflicting meta-type
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-12-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster 555dd1aaa6 qapi: Use re.fullmatch() where appropriate
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-10-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster d0830ee443 qapi: Use "not COND" instead of "!COND" for generated documentation
Generated documentation uses operators "and", "or", and "!".  Change
the latter to "not".

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-9-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster a7987799d1 qapi: Avoid redundant parens in code generated for conditionals
Commit 6cc2e4817f "qapi: introduce QAPISchemaIfCond.cgen()" caused a
minor regression: redundant parenthesis.  Subsequent commits
eliminated of many of them, but not all.  Get rid of the rest now.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-8-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster ccea6a8637 qapi: Factor common recursion out of cgen_ifcond(), docgen_ifcond()
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-7-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster 82ca72c023 qapi: Fix C code generation for 'if'
When commit 5d83b9a130 "qapi: replace if condition list with dict
{'all': [...]}" made cgen_ifcond() and docgen_ifcond() recursive, it
messed up parenthesises in the former, and got them right in the
latter, as the previous commit demonstrates.

To fix, adopt the latter's working code for the former.  This
generates the correct code from the previous commit's commit message.

Fixes: 5d83b9a130
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-6-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster e46c930cdd qapi: Simplify how QAPISchemaIfCond represents "no condition"
None works fine, there is no need to replace it by {} in .__init__().

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-3-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
2021-09-03 17:09:10 +02:00
Markus Armbruster 1889e57a71 qapi: Simplify QAPISchemaIfCond's interface for generating C
QAPISchemaIfCond.cgen() is only ever used like

    gen_if(ifcond.cgen())

and

    gen_endif(ifcond.cgen())

Simplify to

    ifcond.gen_if()

and

    ifcond.gen_endif()

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210831123809.1107782-2-armbru@redhat.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
[Import statements tidied up with isort]
2021-09-03 17:06:40 +02:00
Marc-André LureauandMarkus Armbruster 8a9f1e1d9c qapi: make 'if' condition strings simple identifiers
Change the 'if' condition strings to be C-agnostic. It will accept
'[A-Z][A-Z0-9_]*' identifiers. This allows to express configuration
conditions in other languages (Rust or Python for ex) or other more
suitable forms.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Tested-by: John Snow <jsnow@redhat.com>
Message-Id: <20210804083105.97531-11-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Rebased with semantic conflict in redefined-event.json]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster 2b7d214536 qapi: add 'not' condition operation
For the sake of completeness, introduce the 'not' condition.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-10-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Long line broken in tests/qapi-schema/qapi-schema-test.json]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster 3ad64edfad qapi: add 'any' condition
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-8-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster 5d83b9a130 qapi: replace if condition list with dict {'all': [...]}
Replace the simple list sugar form with a recursive structure that will
accept other operators in the following commits (all, any or not).

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-7-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Accidental code motion undone.  Degenerate :forms: comment dropped.
Helper _check_if() moved.  Error messages tweaked.  ui.json updated.
Accidental changes to qapi-schema-test.json dropped.]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster d806f89f87 qapidoc: introduce QAPISchemaIfCond.docgen()
Instead of building the condition documentation from a list of string,
use the result generated from QAPISchemaIfCond.docgen().

This changes the generated documentation from:
- COND1, COND2... (where COND1, COND2 are Literal nodes, and ',' is Text)
to:
- COND1 and COND2 (the whole string as a Literal node)

This will allow us to generate more complex conditions in the following
patches, such as "(COND1 and COND2) or COND3".

Adding back the differentiated formatting is left to the wish list.

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-6-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[TODO comment added]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster 6cc2e4817f qapi: introduce QAPISchemaIfCond.cgen()
Instead of building prepocessor conditions from a list of string, use
the result generated from QAPISchemaIfCond.cgen() and hide the
implementation details.

Note: this patch introduces a minor regression, generating a redundant
pair of parenthesis. This is mostly fixed in a later patch in this
series ("qapi: replace if condition list with dict [..]")

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-5-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Commit message tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster 33aa3267ba qapi: add QAPISchemaIfCond.is_present()
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210804083105.97531-4-marcandre.lureau@redhat.com>
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Marc-André LureauandMarkus Armbruster f17539c80d qapi: wrap Sequence[str] in an object
Mechanical change, except for a new assertion in
QAPISchemaEntity.ifcond().

Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210804083105.97531-3-marcandre.lureau@redhat.com>
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Rebased with obvious conflicts, commit message adjusted]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
2021-08-26 13:53:56 +02:00
Markus Armbruster b32abbb2f5 qapi: Fix crash on redefinition with a different condition
QAPISchema._make_implicit_object_type() asserts that when an implicit
object type is used multiple times, @ifcond is the same for all uses.
It will be for legitimate uses, i.e. simple union branch wrapper
types.  A comment explains this.

The assertion fails when a command or event is redefined with a
different condition.  The redefinition is an error, but it's flagged
only later.

Fixing the assertion would complicate matters further.  Not
worthwhile, drop it instead.  We really need to get rid of simple
unions.

Tweak test case redefined-event to cover redefinition with a different
condition.

Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210806120510.2367124-1-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
2021-08-26 13:53:56 +02:00
Markus Armbruster a0c7b99bf7 qapi: Fix crash on missing enum member name
New test case enum-dict-no-name.json crashes:

    $ python3 scripts/qapi-gen.py tests/qapi-schema/enum-dict-no-name.json
    Traceback (most recent call last):
    [...]
      File "/work/armbru/qemu/scripts/qapi/expr.py", line 458, in check_enum
	member_name = member['name']
    KeyError: 'name'

Root cause: we try to retrieve member 'name' before we check for
missing members.  With that fixed, we get the expected error "'data'
member misses key 'name'".

Fixes: 0825f62c84
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20210616072121.626431-1-armbru@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: John Snow <jsnow@redhat.com>
2021-07-15 12:56:41 +02:00