openkylin
/
qemu
mirror of https://gitee.com/openkylin/qemu.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity
qemu/tests/test-visitor-serialization.c

1188 lines
33 KiB
C
Raw Normal View History

qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
/*
* Unit-tests for visitor-based serialization
*
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
* Copyright (C) 2014-2015 Red Hat, Inc.
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
* Copyright IBM, Corp. 2012
*
* Authors:
* Michael Roth <mdroth@linux.vnet.ibm.com>
*
* This work is licensed under the terms of the GNU GPL, version 2 or later.
* See the COPYING file in the top-level directory.
*/
tests: Clean up includes Clean up includes so that osdep.h is included first and headers which it implies are not included manually. This commit was created with scripts/clean-includes. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Eric Blake <eblake@redhat.com> Tested-by: Eric Blake <eblake@redhat.com>
2016-02-09 02:08:51 +08:00
#include "qemu/osdep.h"
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
#include <glib.h>
#include <float.h>
qapi: move inclusions of qemu-common.h from headers to .c files Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-06 18:22:34 +08:00
#include "qemu-common.h"
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
#include "test-qapi-types.h"
#include "test-qapi-visit.h"
qapi: move include files to include/qobject/ Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2012-12-18 01:19:43 +08:00
#include "qapi/qmp/types.h"
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
#include "qapi/qmp-input-visitor.h"
#include "qapi/qmp-output-visitor.h"
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
#include "qapi/string-input-visitor.h"
#include "qapi/string-output-visitor.h"
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
#include "qapi-types.h"
#include "qapi-visit.h"
#include "qapi/dealloc-visitor.h"
enum PrimitiveTypeKind {
PTYPE_STRING = 0,
PTYPE_BOOLEAN,
PTYPE_NUMBER,
PTYPE_INTEGER,
PTYPE_U8,
PTYPE_U16,
PTYPE_U32,
PTYPE_U64,
PTYPE_S8,
PTYPE_S16,
PTYPE_S32,
PTYPE_S64,
PTYPE_EOL,
};
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
typedef struct PrimitiveType {
union {
const char *string;
bool boolean;
double number;
int64_t integer;
uint8_t u8;
uint16_t u16;
uint32_t u32;
uint64_t u64;
int8_t s8;
int16_t s16;
int32_t s32;
int64_t s64;
intmax_t max;
} value;
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
enum PrimitiveTypeKind type;
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
const char *description;
} PrimitiveType;
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
typedef struct PrimitiveList {
union {
strList *strings;
boolList *booleans;
numberList *numbers;
intList *integers;
int8List *s8_integers;
int16List *s16_integers;
int32List *s32_integers;
int64List *s64_integers;
uint8List *u8_integers;
uint16List *u16_integers;
uint32List *u32_integers;
uint64List *u64_integers;
} value;
enum PrimitiveTypeKind type;
const char *description;
} PrimitiveList;
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
/* test helpers */
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
typedef void (*VisitorFunc)(Visitor *v, void **native, Error **errp);
static void dealloc_helper(void *native_in, VisitorFunc visit, Error **errp)
{
QapiDeallocVisitor *qdv = qapi_dealloc_visitor_new();
visit(qapi_dealloc_get_visitor(qdv), &native_in, errp);
qapi_dealloc_visitor_cleanup(qdv);
}
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
static void visit_primitive_type(Visitor *v, void **native, Error **errp)
{
PrimitiveType *pt = *native;
switch(pt->type) {
case PTYPE_STRING:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_str(v, NULL, (char **)&pt->value.string, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_BOOLEAN:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_bool(v, NULL, &pt->value.boolean, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_NUMBER:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_number(v, NULL, &pt->value.number, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_INTEGER:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int(v, NULL, &pt->value.integer, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_U8:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint8(v, NULL, &pt->value.u8, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_U16:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint16(v, NULL, &pt->value.u16, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_U32:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint32(v, NULL, &pt->value.u32, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_U64:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint64(v, NULL, &pt->value.u64, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_S8:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int8(v, NULL, &pt->value.s8, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_S16:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int16(v, NULL, &pt->value.s16, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_S32:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int32(v, NULL, &pt->value.s32, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_S64:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int64(v, NULL, &pt->value.s64, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
break;
case PTYPE_EOL:
misc: Use g_assert_not_reached for code which is expected to be unreachable The macro g_assert_not_reached is a better self documenting replacement for assert(0) or assert(false). Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2013-07-26 00:21:28 +08:00
g_assert_not_reached();
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
}
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
static void visit_primitive_list(Visitor *v, void **native, Error **errp)
{
PrimitiveList *pl = *native;
switch (pl->type) {
case PTYPE_STRING:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_strList(v, NULL, &pl->value.strings, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_BOOLEAN:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_boolList(v, NULL, &pl->value.booleans, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_NUMBER:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_numberList(v, NULL, &pl->value.numbers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_INTEGER:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_intList(v, NULL, &pl->value.integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_S8:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int8List(v, NULL, &pl->value.s8_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_S16:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int16List(v, NULL, &pl->value.s16_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_S32:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int32List(v, NULL, &pl->value.s32_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_S64:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_int64List(v, NULL, &pl->value.s64_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_U8:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint8List(v, NULL, &pl->value.u8_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_U16:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint16List(v, NULL, &pl->value.u16_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_U32:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint32List(v, NULL, &pl->value.u32_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
case PTYPE_U64:
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_uint64List(v, NULL, &pl->value.u64_integers, errp);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
break;
default:
misc: Use g_assert_not_reached for code which is expected to be unreachable The macro g_assert_not_reached is a better self documenting replacement for assert(0) or assert(false). Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2013-07-26 00:21:28 +08:00
g_assert_not_reached();
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
}
}
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
static TestStruct *struct_create(void)
{
TestStruct *ts = g_malloc0(sizeof(*ts));
ts->integer = -42;
ts->boolean = true;
ts->string = strdup("test string");
return ts;
}
static void struct_compare(TestStruct *ts1, TestStruct *ts2)
{
g_assert(ts1);
g_assert(ts2);
g_assert_cmpint(ts1->integer, ==, ts2->integer);
g_assert(ts1->boolean == ts2->boolean);
g_assert_cmpstr(ts1->string, ==, ts2->string);
}
static void struct_cleanup(TestStruct *ts)
{
g_free(ts->string);
g_free(ts);
}
static void visit_struct(Visitor *v, void **native, Error **errp)
{
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_TestStruct(v, NULL, (TestStruct **)native, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
static UserDefTwo *nested_struct_create(void)
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
{
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
UserDefTwo *udnp = g_malloc0(sizeof(*udnp));
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
udnp->string0 = strdup("test_string0");
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
udnp->dict1 = g_malloc0(sizeof(*udnp->dict1));
udnp->dict1->string1 = strdup("test_string1");
udnp->dict1->dict2 = g_malloc0(sizeof(*udnp->dict1->dict2));
udnp->dict1->dict2->userdef = g_new0(UserDefOne, 1);
qapi: Unbox base members Rather than storing a base class as a pointer to a box, just store the fields of that base class in the same order, so that a child struct can be directly cast to its parent. This gives less malloc overhead, less pointer dereferencing, and even less generated code. Compare to the earlier commit 1e6c1616a "qapi: Generate a nicer struct for flat unions" (although that patch had fewer places to change, as less of qemu was directly using qapi structs for flat unions). It also allows us to turn on automatic type-safe wrappers for upcasting to the base class of a struct. Changes to the generated code look like this in qapi-types.h: | struct SpiceChannel { |- SpiceBasicInfo *base; |+ /* Members inherited from SpiceBasicInfo: */ |+ char *host; |+ char *port; |+ NetworkAddressFamily family; |+ /* Own members: */ | int64_t connection_id; as well as additional upcast functions like qapi_SpiceChannel_base(). Meanwhile, changes to qapi-visit.c look like: | static void visit_type_SpiceChannel_fields(Visitor *v, SpiceChannel **obj, Error **errp) | { | Error *err = NULL; | |- visit_type_implicit_SpiceBasicInfo(v, &(*obj)->base, &err); |+ visit_type_SpiceBasicInfo_fields(v, (SpiceBasicInfo **)obj, &err); | if (err) { (the cast is necessary, since our upcast wrappers only deal with a single pointer, not pointer-to-pointer); plus the wholesale elimination of some now-unused visit_type_implicit_FOO() functions. Without boxing, the corner case of one empty struct having another empty struct as its base type now requires inserting a dummy member (previously, the 'Base *base' member sufficed). And now that we no longer consume a 'base' member in the generated C struct, we can delete the former negative struct-base-clash-base test. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1445898903-12082-11-git-send-email-eblake@redhat.com> [Commit message tweaked slightly] Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-10-27 06:34:49 +08:00
udnp->dict1->dict2->userdef->integer = 42;
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
udnp->dict1->dict2->userdef->string = strdup("test_string");
udnp->dict1->dict2->string = strdup("test_string2");
udnp->dict1->dict3 = g_malloc0(sizeof(*udnp->dict1->dict3));
udnp->dict1->has_dict3 = true;
udnp->dict1->dict3->userdef = g_new0(UserDefOne, 1);
qapi: Unbox base members Rather than storing a base class as a pointer to a box, just store the fields of that base class in the same order, so that a child struct can be directly cast to its parent. This gives less malloc overhead, less pointer dereferencing, and even less generated code. Compare to the earlier commit 1e6c1616a "qapi: Generate a nicer struct for flat unions" (although that patch had fewer places to change, as less of qemu was directly using qapi structs for flat unions). It also allows us to turn on automatic type-safe wrappers for upcasting to the base class of a struct. Changes to the generated code look like this in qapi-types.h: | struct SpiceChannel { |- SpiceBasicInfo *base; |+ /* Members inherited from SpiceBasicInfo: */ |+ char *host; |+ char *port; |+ NetworkAddressFamily family; |+ /* Own members: */ | int64_t connection_id; as well as additional upcast functions like qapi_SpiceChannel_base(). Meanwhile, changes to qapi-visit.c look like: | static void visit_type_SpiceChannel_fields(Visitor *v, SpiceChannel **obj, Error **errp) | { | Error *err = NULL; | |- visit_type_implicit_SpiceBasicInfo(v, &(*obj)->base, &err); |+ visit_type_SpiceBasicInfo_fields(v, (SpiceBasicInfo **)obj, &err); | if (err) { (the cast is necessary, since our upcast wrappers only deal with a single pointer, not pointer-to-pointer); plus the wholesale elimination of some now-unused visit_type_implicit_FOO() functions. Without boxing, the corner case of one empty struct having another empty struct as its base type now requires inserting a dummy member (previously, the 'Base *base' member sufficed). And now that we no longer consume a 'base' member in the generated C struct, we can delete the former negative struct-base-clash-base test. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1445898903-12082-11-git-send-email-eblake@redhat.com> [Commit message tweaked slightly] Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-10-27 06:34:49 +08:00
udnp->dict1->dict3->userdef->integer = 43;
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
udnp->dict1->dict3->userdef->string = strdup("test_string");
udnp->dict1->dict3->string = strdup("test_string3");
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
return udnp;
}
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
static void nested_struct_compare(UserDefTwo *udnp1, UserDefTwo *udnp2)
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
{
g_assert(udnp1);
g_assert(udnp2);
g_assert_cmpstr(udnp1->string0, ==, udnp2->string0);
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
g_assert_cmpstr(udnp1->dict1->string1, ==, udnp2->dict1->string1);
qapi: Unbox base members Rather than storing a base class as a pointer to a box, just store the fields of that base class in the same order, so that a child struct can be directly cast to its parent. This gives less malloc overhead, less pointer dereferencing, and even less generated code. Compare to the earlier commit 1e6c1616a "qapi: Generate a nicer struct for flat unions" (although that patch had fewer places to change, as less of qemu was directly using qapi structs for flat unions). It also allows us to turn on automatic type-safe wrappers for upcasting to the base class of a struct. Changes to the generated code look like this in qapi-types.h: | struct SpiceChannel { |- SpiceBasicInfo *base; |+ /* Members inherited from SpiceBasicInfo: */ |+ char *host; |+ char *port; |+ NetworkAddressFamily family; |+ /* Own members: */ | int64_t connection_id; as well as additional upcast functions like qapi_SpiceChannel_base(). Meanwhile, changes to qapi-visit.c look like: | static void visit_type_SpiceChannel_fields(Visitor *v, SpiceChannel **obj, Error **errp) | { | Error *err = NULL; | |- visit_type_implicit_SpiceBasicInfo(v, &(*obj)->base, &err); |+ visit_type_SpiceBasicInfo_fields(v, (SpiceBasicInfo **)obj, &err); | if (err) { (the cast is necessary, since our upcast wrappers only deal with a single pointer, not pointer-to-pointer); plus the wholesale elimination of some now-unused visit_type_implicit_FOO() functions. Without boxing, the corner case of one empty struct having another empty struct as its base type now requires inserting a dummy member (previously, the 'Base *base' member sufficed). And now that we no longer consume a 'base' member in the generated C struct, we can delete the former negative struct-base-clash-base test. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1445898903-12082-11-git-send-email-eblake@redhat.com> [Commit message tweaked slightly] Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-10-27 06:34:49 +08:00
g_assert_cmpint(udnp1->dict1->dict2->userdef->integer, ==,
udnp2->dict1->dict2->userdef->integer);
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
g_assert_cmpstr(udnp1->dict1->dict2->userdef->string, ==,
udnp2->dict1->dict2->userdef->string);
g_assert_cmpstr(udnp1->dict1->dict2->string, ==,
udnp2->dict1->dict2->string);
g_assert(udnp1->dict1->has_dict3 == udnp2->dict1->has_dict3);
qapi: Unbox base members Rather than storing a base class as a pointer to a box, just store the fields of that base class in the same order, so that a child struct can be directly cast to its parent. This gives less malloc overhead, less pointer dereferencing, and even less generated code. Compare to the earlier commit 1e6c1616a "qapi: Generate a nicer struct for flat unions" (although that patch had fewer places to change, as less of qemu was directly using qapi structs for flat unions). It also allows us to turn on automatic type-safe wrappers for upcasting to the base class of a struct. Changes to the generated code look like this in qapi-types.h: | struct SpiceChannel { |- SpiceBasicInfo *base; |+ /* Members inherited from SpiceBasicInfo: */ |+ char *host; |+ char *port; |+ NetworkAddressFamily family; |+ /* Own members: */ | int64_t connection_id; as well as additional upcast functions like qapi_SpiceChannel_base(). Meanwhile, changes to qapi-visit.c look like: | static void visit_type_SpiceChannel_fields(Visitor *v, SpiceChannel **obj, Error **errp) | { | Error *err = NULL; | |- visit_type_implicit_SpiceBasicInfo(v, &(*obj)->base, &err); |+ visit_type_SpiceBasicInfo_fields(v, (SpiceBasicInfo **)obj, &err); | if (err) { (the cast is necessary, since our upcast wrappers only deal with a single pointer, not pointer-to-pointer); plus the wholesale elimination of some now-unused visit_type_implicit_FOO() functions. Without boxing, the corner case of one empty struct having another empty struct as its base type now requires inserting a dummy member (previously, the 'Base *base' member sufficed). And now that we no longer consume a 'base' member in the generated C struct, we can delete the former negative struct-base-clash-base test. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1445898903-12082-11-git-send-email-eblake@redhat.com> [Commit message tweaked slightly] Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-10-27 06:34:49 +08:00
g_assert_cmpint(udnp1->dict1->dict3->userdef->integer, ==,
udnp2->dict1->dict3->userdef->integer);
qapi: Drop tests for inline nested structs A future patch will be using a 'name':{dictionary} entry in the QAPI schema to specify a default value for an optional argument; but existing use of inline nested structs conflicts with that goal. More precisely, a definition in the QAPI schema associates a name with a set of properties: Example 1: { 'struct': 'Foo', 'data': { MEMBERS... } } associates the global name 'Foo' with properties (meta-type struct) and MEMBERS... Example 2: 'mumble': TYPE within MEMBERS... above associates 'mumble' with properties (type TYPE) and (optional false) within type Foo The syntax of example 1 is extensible; if we need another property, we add another name/value pair to the dictionary (such as 'base':TYPE). The syntax of example 2 is not extensible, because the right hand side can only be a type. We have used name encoding to add a property: "'*mumble': 'int'" associates 'mumble' with (type int) and (optional true). Nice, but doesn't scale. So the solution is to change our existing uses to be syntactic sugar to an extensible form: NAME: TYPE --> NAME: { 'type': TYPE, 'optional': false } *ONAME: TYPE --> ONAME: { 'type': TYPE, 'optional': true } This patch fixes the testsuite to avoid inline nested types, by breaking the nesting into explicit types; it means that the type is now boxed instead of unboxed in C code, but makes no difference on the wire (and if desired, a later patch could change the generator to not do so much boxing in C). When touching code to add new allocations, also convert existing allocations to consistently prefer typesafe g_new0 over g_malloc0 when a type name is involved. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:30 +08:00
g_assert_cmpstr(udnp1->dict1->dict3->userdef->string, ==,
udnp2->dict1->dict3->userdef->string);
g_assert_cmpstr(udnp1->dict1->dict3->string, ==,
udnp2->dict1->dict3->string);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
static void nested_struct_cleanup(UserDefTwo *udnp)
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
{
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
qapi_free_UserDefTwo(udnp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
static void visit_nested_struct(Visitor *v, void **native, Error **errp)
{
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_UserDefTwo(v, NULL, (UserDefTwo **)native, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
static void visit_nested_struct_list(Visitor *v, void **native, Error **errp)
{
qapi: Swap visit_* arguments for consistent 'name' placement JSON uses "name":value, but many of our visitor interfaces were called with visit_type_FOO(v, &value, name, errp). This can be a bit confusing to have to mentally swap the parameter order to match JSON order. It's particularly bad for visit_start_struct(), where the 'name' parameter is smack in the middle of the otherwise-related group of 'obj, kind, size' parameters! It's time to do a global swap of the parameter ordering, so that the 'name' parameter is always immediately after the Visitor argument. Additional reason in favor of the swap: the existing include/qjson.h prefers listing 'name' first in json_prop_*(), and I have plans to unify that file with the qapi visitors; listing 'name' first in qapi will minimize churn to the (admittedly few) qjson.h clients. Later patches will then fix docs, object.h, visitor-impl.h, and those clients to match. Done by first patching scripts/qapi*.py by hand to make generated files do what I want, then by running the following Coccinelle script to affect the rest of the code base: $ spatch --sp-file script `git grep -l '\bvisit_' -- '**/*.[ch]'` I then had to apply some touchups (Coccinelle insisted on TAB indentation in visitor.h, and botched the signature of visit_type_enum() by rewriting 'const char *const strings[]' to the syntactically invalid 'const char*const[] strings'). The movement of parameters is sufficient to provoke compiler errors if any callers were missed. // Part 1: Swap declaration order @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_start_struct -(TV v, TObj OBJ, T1 ARG1, const char *name, T2 ARG2, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type bool, TV, T1; identifier ARG1; @@ bool visit_optional -(TV v, T1 ARG1, const char *name) +(TV v, const char *name, T1 ARG1) { ... } @@ type TV, TErr, TObj, T1; identifier OBJ, ARG1; @@ void visit_get_next_type -(TV v, TObj OBJ, T1 ARG1, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, TErr errp) { ... } @@ type TV, TErr, TObj, T1, T2; identifier OBJ, ARG1, ARG2; @@ void visit_type_enum -(TV v, TObj OBJ, T1 ARG1, T2 ARG2, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, T1 ARG1, T2 ARG2, TErr errp) { ... } @@ type TV, TErr, TObj; identifier OBJ; identifier VISIT_TYPE =~ "^visit_type_"; @@ void VISIT_TYPE -(TV v, TObj OBJ, const char *name, TErr errp) +(TV v, const char *name, TObj OBJ, TErr errp) { ... } // Part 2: swap caller order @@ expression V, NAME, OBJ, ARG1, ARG2, ERR; identifier VISIT_TYPE =~ "^visit_type_"; @@ ( -visit_start_struct(V, OBJ, ARG1, NAME, ARG2, ERR) +visit_start_struct(V, NAME, OBJ, ARG1, ARG2, ERR) | -visit_optional(V, ARG1, NAME) +visit_optional(V, NAME, ARG1) | -visit_get_next_type(V, OBJ, ARG1, NAME, ERR) +visit_get_next_type(V, NAME, OBJ, ARG1, ERR) | -visit_type_enum(V, OBJ, ARG1, ARG2, NAME, ERR) +visit_type_enum(V, NAME, OBJ, ARG1, ARG2, ERR) | -VISIT_TYPE(V, OBJ, NAME, ERR) +VISIT_TYPE(V, NAME, OBJ, ERR) ) Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com> Message-Id: <1454075341-13658-19-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2016-01-29 21:48:54 +08:00
visit_type_UserDefTwoList(v, NULL, (UserDefTwoList **)native, errp);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
/* test cases */
typedef enum VisitorCapabilities {
VCAP_PRIMITIVES = 1,
VCAP_STRUCTURES = 2,
VCAP_LISTS = 4,
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
VCAP_PRIMITIVE_LISTS = 8,
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
} VisitorCapabilities;
typedef struct SerializeOps {
void (*serialize)(void *native_in, void **datap,
VisitorFunc visit, Error **errp);
void (*deserialize)(void **native_out, void *datap,
VisitorFunc visit, Error **errp);
void (*cleanup)(void *datap);
const char *type;
VisitorCapabilities caps;
} SerializeOps;
typedef struct TestArgs {
const SerializeOps *ops;
void *test_data;
} TestArgs;
static void test_primitives(gconstpointer opaque)
{
TestArgs *args = (TestArgs *) opaque;
const SerializeOps *ops = args->ops;
PrimitiveType *pt = args->test_data;
PrimitiveType *pt_copy = g_malloc0(sizeof(*pt_copy));
void *serialize_data;
pt_copy->type = pt->type;
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
ops->serialize(pt, &serialize_data, visit_primitive_type, &error_abort);
ops->deserialize((void **)&pt_copy, serialize_data, visit_primitive_type,
&error_abort);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
g_assert(pt_copy != NULL);
if (pt->type == PTYPE_STRING) {
g_assert_cmpstr(pt->value.string, ==, pt_copy->value.string);
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
g_free((char *)pt_copy->value.string);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
} else if (pt->type == PTYPE_NUMBER) {
qapi: fix visitor serialization tests for numbers/doubles We never actually stored the stringified double values into the strings before we did the comparisons. This left number/double values completely uncovered in test-visitor-serialization tests. Fixing this exposed a bug in our handling of large whole number values in QEMU's JSON parser which is now fixed. Simplify the code while we're at it by dropping the calc_float_string_storage() craziness in favor of GStrings. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:07 +08:00
GString *double_expected = g_string_new("");
GString *double_actual = g_string_new("");
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
/* we serialize with %f for our reference visitors, so rather than fuzzy
* floating math to test "equality", just compare the formatted values
*/
qapi: fix visitor serialization tests for numbers/doubles We never actually stored the stringified double values into the strings before we did the comparisons. This left number/double values completely uncovered in test-visitor-serialization tests. Fixing this exposed a bug in our handling of large whole number values in QEMU's JSON parser which is now fixed. Simplify the code while we're at it by dropping the calc_float_string_storage() craziness in favor of GStrings. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:07 +08:00
g_string_printf(double_expected, "%.6f", pt->value.number);
g_string_printf(double_actual, "%.6f", pt_copy->value.number);
g_assert_cmpstr(double_actual->str, ==, double_expected->str);
g_string_free(double_expected, true);
g_string_free(double_actual, true);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
} else if (pt->type == PTYPE_BOOLEAN) {
g_assert_cmpint(!!pt->value.max, ==, !!pt->value.max);
} else {
g_assert_cmpint(pt->value.max, ==, pt_copy->value.max);
}
ops->cleanup(serialize_data);
g_free(args);
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
g_free(pt_copy);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
static void test_primitive_lists(gconstpointer opaque)
{
TestArgs *args = (TestArgs *) opaque;
const SerializeOps *ops = args->ops;
PrimitiveType *pt = args->test_data;
tests: Add missing 'static' attributes (fix warnings from smatch) Smatch also complains about 0 used for pointers, so replace those by NULL in test-visitor-serialization.c, too. Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2014-07-08 03:03:38 +08:00
PrimitiveList pl = { .value = { NULL } };
PrimitiveList pl_copy = { .value = { NULL } };
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
PrimitiveList *pl_copy_ptr = &pl_copy;
void *serialize_data;
void *cur_head = NULL;
int i;
pl.type = pl_copy.type = pt->type;
/* build up our list of primitive types */
for (i = 0; i < 32; i++) {
switch (pl.type) {
case PTYPE_STRING: {
strList *tmp = g_new0(strList, 1);
tmp->value = g_strdup(pt->value.string);
if (pl.value.strings == NULL) {
pl.value.strings = tmp;
} else {
tmp->next = pl.value.strings;
pl.value.strings = tmp;
}
break;
}
case PTYPE_INTEGER: {
intList *tmp = g_new0(intList, 1);
tmp->value = pt->value.integer;
if (pl.value.integers == NULL) {
pl.value.integers = tmp;
} else {
tmp->next = pl.value.integers;
pl.value.integers = tmp;
}
break;
}
case PTYPE_S8: {
int8List *tmp = g_new0(int8List, 1);
tmp->value = pt->value.s8;
if (pl.value.s8_integers == NULL) {
pl.value.s8_integers = tmp;
} else {
tmp->next = pl.value.s8_integers;
pl.value.s8_integers = tmp;
}
break;
}
case PTYPE_S16: {
int16List *tmp = g_new0(int16List, 1);
tmp->value = pt->value.s16;
if (pl.value.s16_integers == NULL) {
pl.value.s16_integers = tmp;
} else {
tmp->next = pl.value.s16_integers;
pl.value.s16_integers = tmp;
}
break;
}
case PTYPE_S32: {
int32List *tmp = g_new0(int32List, 1);
tmp->value = pt->value.s32;
if (pl.value.s32_integers == NULL) {
pl.value.s32_integers = tmp;
} else {
tmp->next = pl.value.s32_integers;
pl.value.s32_integers = tmp;
}
break;
}
case PTYPE_S64: {
int64List *tmp = g_new0(int64List, 1);
tmp->value = pt->value.s64;
if (pl.value.s64_integers == NULL) {
pl.value.s64_integers = tmp;
} else {
tmp->next = pl.value.s64_integers;
pl.value.s64_integers = tmp;
}
break;
}
case PTYPE_U8: {
uint8List *tmp = g_new0(uint8List, 1);
tmp->value = pt->value.u8;
if (pl.value.u8_integers == NULL) {
pl.value.u8_integers = tmp;
} else {
tmp->next = pl.value.u8_integers;
pl.value.u8_integers = tmp;
}
break;
}
case PTYPE_U16: {
uint16List *tmp = g_new0(uint16List, 1);
tmp->value = pt->value.u16;
if (pl.value.u16_integers == NULL) {
pl.value.u16_integers = tmp;
} else {
tmp->next = pl.value.u16_integers;
pl.value.u16_integers = tmp;
}
break;
}
case PTYPE_U32: {
uint32List *tmp = g_new0(uint32List, 1);
tmp->value = pt->value.u32;
if (pl.value.u32_integers == NULL) {
pl.value.u32_integers = tmp;
} else {
tmp->next = pl.value.u32_integers;
pl.value.u32_integers = tmp;
}
break;
}
case PTYPE_U64: {
uint64List *tmp = g_new0(uint64List, 1);
tmp->value = pt->value.u64;
if (pl.value.u64_integers == NULL) {
pl.value.u64_integers = tmp;
} else {
tmp->next = pl.value.u64_integers;
pl.value.u64_integers = tmp;
}
break;
}
case PTYPE_NUMBER: {
numberList *tmp = g_new0(numberList, 1);
tmp->value = pt->value.number;
if (pl.value.numbers == NULL) {
pl.value.numbers = tmp;
} else {
tmp->next = pl.value.numbers;
pl.value.numbers = tmp;
}
break;
}
case PTYPE_BOOLEAN: {
boolList *tmp = g_new0(boolList, 1);
tmp->value = pt->value.boolean;
if (pl.value.booleans == NULL) {
pl.value.booleans = tmp;
} else {
tmp->next = pl.value.booleans;
pl.value.booleans = tmp;
}
break;
}
default:
misc: Use g_assert_not_reached for code which is expected to be unreachable The macro g_assert_not_reached is a better self documenting replacement for assert(0) or assert(false). Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2013-07-26 00:21:28 +08:00
g_assert_not_reached();
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
}
}
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
ops->serialize((void **)&pl, &serialize_data, visit_primitive_list,
&error_abort);
ops->deserialize((void **)&pl_copy_ptr, serialize_data,
visit_primitive_list, &error_abort);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
i = 0;
/* compare our deserialized list of primitives to the original */
do {
switch (pl_copy.type) {
case PTYPE_STRING: {
strList *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.strings;
}
g_assert_cmpstr(pt->value.string, ==, ptr->value);
break;
}
case PTYPE_INTEGER: {
intList *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.integers;
}
g_assert_cmpint(pt->value.integer, ==, ptr->value);
break;
}
case PTYPE_S8: {
int8List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.s8_integers;
}
g_assert_cmpint(pt->value.s8, ==, ptr->value);
break;
}
case PTYPE_S16: {
int16List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.s16_integers;
}
g_assert_cmpint(pt->value.s16, ==, ptr->value);
break;
}
case PTYPE_S32: {
int32List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.s32_integers;
}
g_assert_cmpint(pt->value.s32, ==, ptr->value);
break;
}
case PTYPE_S64: {
int64List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.s64_integers;
}
g_assert_cmpint(pt->value.s64, ==, ptr->value);
break;
}
case PTYPE_U8: {
uint8List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.u8_integers;
}
g_assert_cmpint(pt->value.u8, ==, ptr->value);
break;
}
case PTYPE_U16: {
uint16List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.u16_integers;
}
g_assert_cmpint(pt->value.u16, ==, ptr->value);
break;
}
case PTYPE_U32: {
uint32List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.u32_integers;
}
g_assert_cmpint(pt->value.u32, ==, ptr->value);
break;
}
case PTYPE_U64: {
uint64List *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.u64_integers;
}
g_assert_cmpint(pt->value.u64, ==, ptr->value);
break;
}
case PTYPE_NUMBER: {
numberList *ptr;
GString *double_expected = g_string_new("");
GString *double_actual = g_string_new("");
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.numbers;
}
/* we serialize with %f for our reference visitors, so rather than
* fuzzy floating math to test "equality", just compare the
* formatted values
*/
g_string_printf(double_expected, "%.6f", pt->value.number);
g_string_printf(double_actual, "%.6f", ptr->value);
g_assert_cmpstr(double_actual->str, ==, double_expected->str);
g_string_free(double_expected, true);
g_string_free(double_actual, true);
break;
}
case PTYPE_BOOLEAN: {
boolList *ptr;
if (cur_head) {
ptr = cur_head;
cur_head = ptr->next;
} else {
cur_head = ptr = pl_copy.value.booleans;
}
g_assert_cmpint(!!pt->value.boolean, ==, !!ptr->value);
break;
}
default:
misc: Use g_assert_not_reached for code which is expected to be unreachable The macro g_assert_not_reached is a better self documenting replacement for assert(0) or assert(false). Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2013-07-26 00:21:28 +08:00
g_assert_not_reached();
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
}
i++;
} while (cur_head);
g_assert_cmpint(i, ==, 33);
ops->cleanup(serialize_data);
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
dealloc_helper(&pl, visit_primitive_list, &error_abort);
dealloc_helper(&pl_copy, visit_primitive_list, &error_abort);
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
g_free(args);
}
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
static void test_struct(gconstpointer opaque)
{
TestArgs *args = (TestArgs *) opaque;
const SerializeOps *ops = args->ops;
TestStruct *ts = struct_create();
TestStruct *ts_copy = NULL;
void *serialize_data;
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
ops->serialize(ts, &serialize_data, visit_struct, &error_abort);
ops->deserialize((void **)&ts_copy, serialize_data, visit_struct,
&error_abort);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
struct_compare(ts, ts_copy);
struct_cleanup(ts);
struct_cleanup(ts_copy);
ops->cleanup(serialize_data);
g_free(args);
}
static void test_nested_struct(gconstpointer opaque)
{
TestArgs *args = (TestArgs *) opaque;
const SerializeOps *ops = args->ops;
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
UserDefTwo *udnp = nested_struct_create();
UserDefTwo *udnp_copy = NULL;
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
void *serialize_data;
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
ops->serialize(udnp, &serialize_data, visit_nested_struct, &error_abort);
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
ops->deserialize((void **)&udnp_copy, serialize_data, visit_nested_struct,
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
&error_abort);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
nested_struct_compare(udnp, udnp_copy);
nested_struct_cleanup(udnp);
nested_struct_cleanup(udnp_copy);
ops->cleanup(serialize_data);
g_free(args);
}
static void test_nested_struct_list(gconstpointer opaque)
{
TestArgs *args = (TestArgs *) opaque;
const SerializeOps *ops = args->ops;
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
UserDefTwoList *listp = NULL, *tmp, *tmp_copy, *listp_copy = NULL;
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
void *serialize_data;
int i = 0;
for (i = 0; i < 8; i++) {
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
tmp = g_new0(UserDefTwoList, 1);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
tmp->value = nested_struct_create();
tmp->next = listp;
listp = tmp;
}
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
ops->serialize(listp, &serialize_data, visit_nested_struct_list,
&error_abort);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
ops->deserialize((void **)&listp_copy, serialize_data,
qapi: Simplify non-error testing in test-qmp-* By using &error_abort, we can avoid a local err variable in situations where we expect success. It also has the nice effect that if the test breaks, the error message from error_abort tends to be nicer than that of g_assert(). This patch has an additional bonus of fixing several call sites that were passing &err to two different functions without checking it in between. In general that is unsafe practice; because if the first function sets an error, the second function could abort() if it tries to set a different error. We got away with it because we were asserting that err was NULL through the entire chain, but switching to &error_abort avoids the questionable practice up front. Signed-off-by: Eric Blake <eblake@redhat.com> Message-Id: <1446791754-23823-7-git-send-email-eblake@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-11-06 14:35:30 +08:00
visit_nested_struct_list, &error_abort);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
tmp = listp;
tmp_copy = listp_copy;
while (listp_copy) {
g_assert(listp);
nested_struct_compare(listp->value, listp_copy->value);
listp = listp->next;
listp_copy = listp_copy->next;
}
qapi: Merge UserDefTwo and UserDefNested in tests In the testsuite, UserDefTwo and UserDefNested were identical structs other than the member names. Reduce code duplication by having just one type, and choose names that also favor reuse. This will also make it easier for a later patch to get rid of inline nested types in QAPI. When touching code related to allocations, convert g_malloc0(sizeof(Type)) to the more typesafe g_new0(Type, 1). Ensure that 'make check-qapi-schema check-unit' still passes. Signed-off-by: Eric Blake <eblake@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2015-05-04 23:05:29 +08:00
qapi_free_UserDefTwoList(tmp);
qapi_free_UserDefTwoList(tmp_copy);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
ops->cleanup(serialize_data);
g_free(args);
}
tests: Add missing 'static' attributes (fix warnings from smatch) Smatch also complains about 0 used for pointers, so replace those by NULL in test-visitor-serialization.c, too. Signed-off-by: Stefan Weil <sw@weilnetz.de> Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2014-07-08 03:03:38 +08:00
static PrimitiveType pt_values[] = {
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
/* string tests */
{
.description = "string_empty",
.type = PTYPE_STRING,
.value.string = "",
},
{
.description = "string_whitespace",
.type = PTYPE_STRING,
.value.string = "a b c\td",
},
{
.description = "string_newlines",
.type = PTYPE_STRING,
.value.string = "a\nb\n",
},
{
.description = "string_commas",
.type = PTYPE_STRING,
.value.string = "a,b, c,d",
},
{
.description = "string_single_quoted",
.type = PTYPE_STRING,
.value.string = "'a b',cd",
},
{
.description = "string_double_quoted",
.type = PTYPE_STRING,
.value.string = "\"a b\",cd",
},
/* boolean tests */
{
.description = "boolean_true1",
.type = PTYPE_BOOLEAN,
.value.boolean = true,
},
{
.description = "boolean_true2",
.type = PTYPE_BOOLEAN,
.value.boolean = 8,
},
{
.description = "boolean_true3",
.type = PTYPE_BOOLEAN,
.value.boolean = -1,
},
{
.description = "boolean_false1",
.type = PTYPE_BOOLEAN,
.value.boolean = false,
},
{
.description = "boolean_false2",
.type = PTYPE_BOOLEAN,
.value.boolean = 0,
},
/* number tests (double) */
/* note: we format these to %.6f before comparing, since that's how
* we serialize them and it doesn't make sense to check precision
* beyond that.
*/
{
.description = "number_sanity1",
.type = PTYPE_NUMBER,
.value.number = -1,
},
{
.description = "number_sanity2",
.type = PTYPE_NUMBER,
.value.number = 3.14159265,
},
{
.description = "number_min",
.type = PTYPE_NUMBER,
.value.number = DBL_MIN,
},
{
.description = "number_max",
.type = PTYPE_NUMBER,
.value.number = DBL_MAX,
},
/* integer tests (int64) */
{
.description = "integer_sanity1",
.type = PTYPE_INTEGER,
.value.integer = -1,
},
{
.description = "integer_sanity2",
.type = PTYPE_INTEGER,
.value.integer = INT64_MAX / 2 + 1,
},
{
.description = "integer_min",
.type = PTYPE_INTEGER,
.value.integer = INT64_MIN,
},
{
.description = "integer_max",
.type = PTYPE_INTEGER,
.value.integer = INT64_MAX,
},
/* uint8 tests */
{
.description = "uint8_sanity1",
.type = PTYPE_U8,
.value.u8 = 1,
},
{
.description = "uint8_sanity2",
.type = PTYPE_U8,
.value.u8 = UINT8_MAX / 2 + 1,
},
{
.description = "uint8_min",
.type = PTYPE_U8,
.value.u8 = 0,
},
{
.description = "uint8_max",
.type = PTYPE_U8,
.value.u8 = UINT8_MAX,
},
/* uint16 tests */
{
.description = "uint16_sanity1",
.type = PTYPE_U16,
.value.u16 = 1,
},
{
.description = "uint16_sanity2",
.type = PTYPE_U16,
.value.u16 = UINT16_MAX / 2 + 1,
},
{
.description = "uint16_min",
.type = PTYPE_U16,
.value.u16 = 0,
},
{
.description = "uint16_max",
.type = PTYPE_U16,
.value.u16 = UINT16_MAX,
},
/* uint32 tests */
{
.description = "uint32_sanity1",
.type = PTYPE_U32,
.value.u32 = 1,
},
{
.description = "uint32_sanity2",
.type = PTYPE_U32,
.value.u32 = UINT32_MAX / 2 + 1,
},
{
.description = "uint32_min",
.type = PTYPE_U32,
.value.u32 = 0,
},
{
.description = "uint32_max",
.type = PTYPE_U32,
.value.u32 = UINT32_MAX,
},
/* uint64 tests */
{
.description = "uint64_sanity1",
.type = PTYPE_U64,
.value.u64 = 1,
},
{
.description = "uint64_sanity2",
.type = PTYPE_U64,
.value.u64 = UINT64_MAX / 2 + 1,
},
{
.description = "uint64_min",
.type = PTYPE_U64,
.value.u64 = 0,
},
{
.description = "uint64_max",
.type = PTYPE_U64,
.value.u64 = UINT64_MAX,
},
/* int8 tests */
{
.description = "int8_sanity1",
.type = PTYPE_S8,
.value.s8 = -1,
},
{
.description = "int8_sanity2",
.type = PTYPE_S8,
.value.s8 = INT8_MAX / 2 + 1,
},
{
.description = "int8_min",
.type = PTYPE_S8,
.value.s8 = INT8_MIN,
},
{
.description = "int8_max",
.type = PTYPE_S8,
.value.s8 = INT8_MAX,
},
/* int16 tests */
{
.description = "int16_sanity1",
.type = PTYPE_S16,
.value.s16 = -1,
},
{
.description = "int16_sanity2",
.type = PTYPE_S16,
.value.s16 = INT16_MAX / 2 + 1,
},
{
.description = "int16_min",
.type = PTYPE_S16,
.value.s16 = INT16_MIN,
},
{
.description = "int16_max",
.type = PTYPE_S16,
.value.s16 = INT16_MAX,
},
/* int32 tests */
{
.description = "int32_sanity1",
.type = PTYPE_S32,
.value.s32 = -1,
},
{
.description = "int32_sanity2",
.type = PTYPE_S32,
.value.s32 = INT32_MAX / 2 + 1,
},
{
.description = "int32_min",
.type = PTYPE_S32,
.value.s32 = INT32_MIN,
},
{
.description = "int32_max",
.type = PTYPE_S32,
.value.s32 = INT32_MAX,
},
/* int64 tests */
{
.description = "int64_sanity1",
.type = PTYPE_S64,
.value.s64 = -1,
},
{
.description = "int64_sanity2",
.type = PTYPE_S64,
.value.s64 = INT64_MAX / 2 + 1,
},
{
.description = "int64_min",
.type = PTYPE_S64,
.value.s64 = INT64_MIN,
},
{
.description = "int64_max",
.type = PTYPE_S64,
.value.s64 = INT64_MAX,
},
{ .type = PTYPE_EOL }
};
/* visitor-specific op implementations */
typedef struct QmpSerializeData {
QmpOutputVisitor *qov;
QmpInputVisitor *qiv;
} QmpSerializeData;
static void qmp_serialize(void *native_in, void **datap,
VisitorFunc visit, Error **errp)
{
QmpSerializeData *d = g_malloc0(sizeof(*d));
d->qov = qmp_output_visitor_new();
visit(qmp_output_get_visitor(d->qov), &native_in, errp);
*datap = d;
}
static void qmp_deserialize(void **native_out, void *datap,
VisitorFunc visit, Error **errp)
{
QmpSerializeData *d = datap;
qapi: fix leak in unit tests qmp_output_get_qobject() increments the qobject's reference count. Since we currently pass this straight into qobject_to_json() so we can feed the data into a QMP input visitor, we never actually free the underlying qobject when qmp_output_visitor_cleanup() is called. This causes leaks on all of the QMP serialization tests. Fix this by holding a pointer to the qobject and decref'ing it before returning from qmp_deserialize(). Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-10 10:20:57 +08:00
QString *output_json;
QObject *obj_orig, *obj;
obj_orig = qmp_output_get_qobject(d->qov);
output_json = qobject_to_json(obj_orig);
obj = qobject_from_json(qstring_get_str(output_json));
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
QDECREF(output_json);
d->qiv = qmp_input_visitor_new(obj);
qapi: fix leak in unit tests qmp_output_get_qobject() increments the qobject's reference count. Since we currently pass this straight into qobject_to_json() so we can feed the data into a QMP input visitor, we never actually free the underlying qobject when qmp_output_visitor_cleanup() is called. This causes leaks on all of the QMP serialization tests. Fix this by holding a pointer to the qobject and decref'ing it before returning from qmp_deserialize(). Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-10 10:20:57 +08:00
qobject_decref(obj_orig);
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
qobject_decref(obj);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
visit(qmp_input_get_visitor(d->qiv), native_out, errp);
}
static void qmp_cleanup(void *datap)
{
QmpSerializeData *d = datap;
qmp_output_visitor_cleanup(d->qov);
qmp_input_visitor_cleanup(d->qiv);
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
g_free(d);
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
typedef struct StringSerializeData {
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
char *string;
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
StringOutputVisitor *sov;
StringInputVisitor *siv;
} StringSerializeData;
static void string_serialize(void *native_in, void **datap,
VisitorFunc visit, Error **errp)
{
StringSerializeData *d = g_malloc0(sizeof(*d));
qapi: Add human mode to StringOutputVisitor This will be used by "info qtree". For numbers it prints both the decimal and hex values. For sizes it rounds to the nearest power of 2^10. For strings, it puts quotes around the string and separates NULL and empty string. Reviewed-by: Igor Mammedov <imammedo@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2014-02-08 18:01:50 +08:00
d->sov = string_output_visitor_new(false);
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
visit(string_output_get_visitor(d->sov), &native_in, errp);
*datap = d;
}
static void string_deserialize(void **native_out, void *datap,
VisitorFunc visit, Error **errp)
{
StringSerializeData *d = datap;
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
d->string = string_output_get_string(d->sov);
d->siv = string_input_visitor_new(d->string);
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
visit(string_input_get_visitor(d->siv), native_out, errp);
}
static void string_cleanup(void *datap)
{
StringSerializeData *d = datap;
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
string_output_visitor_cleanup(d->sov);
string_input_visitor_cleanup(d->siv);
test-visitor-serialization: Fix some memory leaks This patch fixes some of the memory leaks in test-visitor-serialization but not all of them. Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2013-03-28 23:18:40 +08:00
g_free(d->string);
g_free(d);
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
}
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
/* visitor registration, test harness */
/* note: to function interchangeably as a serialization mechanism your
* visitor test implementation should pass the test cases for all visitor
* capabilities: primitives, structures, and lists
*/
static const SerializeOps visitors[] = {
{
.type = "QMP",
.serialize = qmp_serialize,
.deserialize = qmp_deserialize,
.cleanup = qmp_cleanup,
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
.caps = VCAP_PRIMITIVES | VCAP_STRUCTURES | VCAP_LISTS |
VCAP_PRIMITIVE_LISTS
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
},
qapi: Add String visitor coverage to serialization unit tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-23 00:16:31 +08:00
{
.type = "String",
.serialize = string_serialize,
.deserialize = string_deserialize,
.cleanup = string_cleanup,
.caps = VCAP_PRIMITIVES
},
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
{ NULL }
};
static void add_visitor_type(const SerializeOps *ops)
{
char testname_prefix[128];
char testname[128];
TestArgs *args;
int i = 0;
sprintf(testname_prefix, "/visitor/serialization/%s", ops->type);
if (ops->caps & VCAP_PRIMITIVES) {
while (pt_values[i].type != PTYPE_EOL) {
sprintf(testname, "%s/primitives/%s", testname_prefix,
pt_values[i].description);
args = g_malloc0(sizeof(*args));
args->ops = ops;
args->test_data = &pt_values[i];
g_test_add_data_func(testname, args, test_primitives);
i++;
}
}
if (ops->caps & VCAP_STRUCTURES) {
sprintf(testname, "%s/struct", testname_prefix);
args = g_malloc0(sizeof(*args));
args->ops = ops;
args->test_data = NULL;
g_test_add_data_func(testname, args, test_struct);
sprintf(testname, "%s/nested_struct", testname_prefix);
args = g_malloc0(sizeof(*args));
args->ops = ops;
args->test_data = NULL;
g_test_add_data_func(testname, args, test_nested_struct);
}
if (ops->caps & VCAP_LISTS) {
sprintf(testname, "%s/nested_struct_list", testname_prefix);
args = g_malloc0(sizeof(*args));
args->ops = ops;
args->test_data = NULL;
g_test_add_data_func(testname, args, test_nested_struct_list);
}
qapi: add native list coverage for visitor serialization tests Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Amos Kong <akong@redhat.com> Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
2013-05-11 06:46:08 +08:00
if (ops->caps & VCAP_PRIMITIVE_LISTS) {
i = 0;
while (pt_values[i].type != PTYPE_EOL) {
sprintf(testname, "%s/primitive_list/%s", testname_prefix,
pt_values[i].description);
args = g_malloc0(sizeof(*args));
args->ops = ops;
args->test_data = &pt_values[i];
g_test_add_data_func(testname, args, test_primitive_lists);
i++;
}
}
qapi: Unit tests for visitor-based serialization Currently we test our visitors individually, and seperately for input vs. output. This is useful for validating internal representations against the native C types and vice-versa, and other visitor-specific testing, but it doesn't cover the potential use-case of using visitor pairs for serialization/deserialization very well, and makes it hard to easily extend the coverage for different C types / boundary conditions. To cover that we add a set of unit tests that takes a number of native C values, passes them into an output visitor, extracts the values with an input visitor, then compares the result to the original. Plugging in new visitors to the test harness only requires a user to implement the SerializeOps interface and add it to a list. Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Andreas Färber <afaerber@suse.de>
2012-02-22 11:05:07 +08:00
}
int main(int argc, char **argv)
{
int i = 0;
g_test_init(&argc, &argv, NULL);
while (visitors[i].type != NULL) {
add_visitor_type(&visitors[i]);
i++;
}
g_test_run();
return 0;
}