maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/flang/test/Semantics/resolve37.f90

36 lines
1.1 KiB
Fortran
Raw Normal View History

[flang] Semantic checks for C702 C702 (R701) A colon shall not be used as a type-param-value except in the declaration of an entity that has the POINTER or ALLOCATABLE attribute. I added code to the visitor for a TypeDeclarationStmt to check for the 'LEN' type parameter for strings and to loop over the type parameters for derived types. I also ran into a few situations where previous tests had erroneously used a colon for type parameters without either the POINTER or ALLOCATABLE attribute and fixed them up. Original-commit: flang-compiler/f18@a1a95bfcd1811d697f1386fe57af664b8a16ffb5 Reviewed-on: https://github.com/flang-compiler/f18/pull/973
2020-02-07 04:26:51 +08:00
! C701 The type-param-value for a kind type parameter shall be a constant
! expression. This constraint looks like a mistake in the standard.
[flang] Rework when symbol table expressions are evaluated It's not good enough to evaluate expressions in the symbol table after name resolution has completed. This is because we need the values of constant expressions for types, for example, we need to evaluate `k` in `integer(k) :: x` to know the type of `x`. So, eliminate `LazyExpr` and call `EvaluateExpr()` on expressions that we need in the symbol table. The latter evaluates and folds an expression in the current context. This is now possible because symbols are added to `parser::Name` as soon as possible rather than in a pass after name resolution. Along with `LazyExpr` we can eliminate the whole `ResolveSymbolExprs` pass that used to resolve them. In resolve-names.cc, many `Pre` functions are changed to `Post` so that names are resolved before doing the associated processing. For example, with intrinsic type specs, names in the kind expression must be resolved before attempting to evaluate that expression. In `GetSymbolType()` in type.cc, handle both `ObjectEntityDetails` and `EntityDetails` by using `Symbol::GetType()`. Add explicit declarations in label01.F90 because we can't handle implicitly typed array bounds yet. Original-commit: flang-compiler/f18@d67716640baeedca150205f9f3845ceb5d2f6402 Reviewed-on: https://github.com/flang-compiler/f18/pull/238
2018-12-06 22:59:37 +08:00
integer, parameter :: k = 8
real, parameter :: l = 8.0
integer :: n = 2
[flang] update error message texts in test Original-commit: flang-compiler/f18@42c0d5d950aba45b5ed0463345d5508631648b03 Reviewed-on: https://github.com/flang-compiler/f18/pull/250 Tree-same-pre-rewrite: false
2018-12-29 08:01:46 +08:00
!ERROR: Must be a constant value
[flang] Rework when symbol table expressions are evaluated It's not good enough to evaluate expressions in the symbol table after name resolution has completed. This is because we need the values of constant expressions for types, for example, we need to evaluate `k` in `integer(k) :: x` to know the type of `x`. So, eliminate `LazyExpr` and call `EvaluateExpr()` on expressions that we need in the symbol table. The latter evaluates and folds an expression in the current context. This is now possible because symbols are added to `parser::Name` as soon as possible rather than in a pass after name resolution. Along with `LazyExpr` we can eliminate the whole `ResolveSymbolExprs` pass that used to resolve them. In resolve-names.cc, many `Pre` functions are changed to `Post` so that names are resolved before doing the associated processing. For example, with intrinsic type specs, names in the kind expression must be resolved before attempting to evaluate that expression. In `GetSymbolType()` in type.cc, handle both `ObjectEntityDetails` and `EntityDetails` by using `Symbol::GetType()`. Add explicit declarations in label01.F90 because we can't handle implicitly typed array bounds yet. Original-commit: flang-compiler/f18@d67716640baeedca150205f9f3845ceb5d2f6402 Reviewed-on: https://github.com/flang-compiler/f18/pull/238
2018-12-06 22:59:37 +08:00
parameter(m=n)
integer(k) :: x
[flang] Reactivate tree-driven constraint checking on expressions. Original-commit: flang-compiler/f18@85c4a6aab601cf2a1647ec27b29f9c6decbe1ed6 Reviewed-on: https://github.com/flang-compiler/f18/pull/406 Tree-same-pre-rewrite: false
2019-04-12 08:22:16 +08:00
!ERROR: Must have INTEGER type, but is REAL(4)
[flang] Rework when symbol table expressions are evaluated It's not good enough to evaluate expressions in the symbol table after name resolution has completed. This is because we need the values of constant expressions for types, for example, we need to evaluate `k` in `integer(k) :: x` to know the type of `x`. So, eliminate `LazyExpr` and call `EvaluateExpr()` on expressions that we need in the symbol table. The latter evaluates and folds an expression in the current context. This is now possible because symbols are added to `parser::Name` as soon as possible rather than in a pass after name resolution. Along with `LazyExpr` we can eliminate the whole `ResolveSymbolExprs` pass that used to resolve them. In resolve-names.cc, many `Pre` functions are changed to `Post` so that names are resolved before doing the associated processing. For example, with intrinsic type specs, names in the kind expression must be resolved before attempting to evaluate that expression. In `GetSymbolType()` in type.cc, handle both `ObjectEntityDetails` and `EntityDetails` by using `Symbol::GetType()`. Add explicit declarations in label01.F90 because we can't handle implicitly typed array bounds yet. Original-commit: flang-compiler/f18@d67716640baeedca150205f9f3845ceb5d2f6402 Reviewed-on: https://github.com/flang-compiler/f18/pull/238
2018-12-06 22:59:37 +08:00
integer(l) :: y
[flang] update error message texts in test Original-commit: flang-compiler/f18@42c0d5d950aba45b5ed0463345d5508631648b03 Reviewed-on: https://github.com/flang-compiler/f18/pull/250 Tree-same-pre-rewrite: false
2018-12-29 08:01:46 +08:00
!ERROR: Must be a constant value
[flang] Rework when symbol table expressions are evaluated It's not good enough to evaluate expressions in the symbol table after name resolution has completed. This is because we need the values of constant expressions for types, for example, we need to evaluate `k` in `integer(k) :: x` to know the type of `x`. So, eliminate `LazyExpr` and call `EvaluateExpr()` on expressions that we need in the symbol table. The latter evaluates and folds an expression in the current context. This is now possible because symbols are added to `parser::Name` as soon as possible rather than in a pass after name resolution. Along with `LazyExpr` we can eliminate the whole `ResolveSymbolExprs` pass that used to resolve them. In resolve-names.cc, many `Pre` functions are changed to `Post` so that names are resolved before doing the associated processing. For example, with intrinsic type specs, names in the kind expression must be resolved before attempting to evaluate that expression. In `GetSymbolType()` in type.cc, handle both `ObjectEntityDetails` and `EntityDetails` by using `Symbol::GetType()`. Add explicit declarations in label01.F90 because we can't handle implicitly typed array bounds yet. Original-commit: flang-compiler/f18@d67716640baeedca150205f9f3845ceb5d2f6402 Reviewed-on: https://github.com/flang-compiler/f18/pull/238
2018-12-06 22:59:37 +08:00
integer(n) :: z
type t(k)
integer, kind :: k
end type
[flang] basic skeleton of assignment analyzer remove needless template<> on some function overloads dodge bogus compiler warning from gcc 8.1.0 only stricter typing of expressions in symbols adjust modfile12.f90 expected test results add Unwrap, massage folding a bit Use Unwrap to simplify folding Move KindSelector analysis into expression semantics fix crash checkpoint updates to TypeParamInquiry support of %KIND type parameter inquiry equality testing for expressions checkpoint during PDT implementation reformat checkpoint derived type instantiation checkpoint resolve merge debugging failed tests fix failing resolve37.f90 test all existing tests pass clean up all build warnings fix bug update copyright dates fix copyright dates address review comments review comment merge with master after peeling off changes bugfixing new feature fix warning from old g++s tweaks after merging with latest head more bugfixing making modfile17.f90 test work Make kinds into expressions in symbol table types big refactor for deferring kinds in intrinsic types modfile17.f90 test passes clean up TODOs Simplify types as stored in scopes Test KIND parameter default init expressions, debug them Update copyright dates address comments remove dead line address comments Original-commit: flang-compiler/f18@1f43d0a04813dc02909e9952ee479d13f630df4c Reviewed-on: https://github.com/flang-compiler/f18/pull/260 Tree-same-pre-rewrite: false
2018-12-05 02:55:32 +08:00
!ERROR: Type parameter 'k' lacks a value and has no default
type(t( &
[flang] Reactivate tree-driven constraint checking on expressions. Original-commit: flang-compiler/f18@85c4a6aab601cf2a1647ec27b29f9c6decbe1ed6 Reviewed-on: https://github.com/flang-compiler/f18/pull/406 Tree-same-pre-rewrite: false
2019-04-12 08:22:16 +08:00
!ERROR: Must have INTEGER type, but is LOGICAL(4)
[flang] basic skeleton of assignment analyzer remove needless template<> on some function overloads dodge bogus compiler warning from gcc 8.1.0 only stricter typing of expressions in symbols adjust modfile12.f90 expected test results add Unwrap, massage folding a bit Use Unwrap to simplify folding Move KindSelector analysis into expression semantics fix crash checkpoint updates to TypeParamInquiry support of %KIND type parameter inquiry equality testing for expressions checkpoint during PDT implementation reformat checkpoint derived type instantiation checkpoint resolve merge debugging failed tests fix failing resolve37.f90 test all existing tests pass clean up all build warnings fix bug update copyright dates fix copyright dates address review comments review comment merge with master after peeling off changes bugfixing new feature fix warning from old g++s tweaks after merging with latest head more bugfixing making modfile17.f90 test work Make kinds into expressions in symbol table types big refactor for deferring kinds in intrinsic types modfile17.f90 test passes clean up TODOs Simplify types as stored in scopes Test KIND parameter default init expressions, debug them Update copyright dates address comments remove dead line address comments Original-commit: flang-compiler/f18@1f43d0a04813dc02909e9952ee479d13f630df4c Reviewed-on: https://github.com/flang-compiler/f18/pull/260 Tree-same-pre-rewrite: false
2018-12-05 02:55:32 +08:00
.true.)) :: w
[flang] Reactivate tree-driven constraint checking on expressions. Original-commit: flang-compiler/f18@85c4a6aab601cf2a1647ec27b29f9c6decbe1ed6 Reviewed-on: https://github.com/flang-compiler/f18/pull/406 Tree-same-pre-rewrite: false
2019-04-12 08:22:16 +08:00
!ERROR: Must have INTEGER type, but is REAL(4)
[flang] done with merge Original-commit: flang-compiler/f18@9c63eb7a1a87bcd075938b4ec93a45ef5bd4c1bf Reviewed-on: https://github.com/flang-compiler/f18/pull/260 Tree-same-pre-rewrite: false
2019-01-18 08:14:36 +08:00
real :: u(l*2)
[flang] Reactivate tree-driven constraint checking on expressions. Original-commit: flang-compiler/f18@85c4a6aab601cf2a1647ec27b29f9c6decbe1ed6 Reviewed-on: https://github.com/flang-compiler/f18/pull/406 Tree-same-pre-rewrite: false
2019-04-12 08:22:16 +08:00
!ERROR: Must have INTEGER type, but is REAL(4)
[flang] Add support for character type specs Recognize the various ways of specifying character lengths. Define CharacterTypeSpec with length and kind and store them in the current scope, as is done with DerivedTypeSpec (which can also have length parameters). Note: IntrinsicTypeSpec is no longer used for characters, so it should have a different name. Similarly, in DeclTypeSpec::Category, Intrinsic does not include Character. Original-commit: flang-compiler/f18@5f847851933b27dd30bd8165d16a7c390e787b9a Reviewed-on: https://github.com/flang-compiler/f18/pull/247 Tree-same-pre-rewrite: false
2018-12-15 06:04:15 +08:00
character(len=l) :: v
[flang] add test for parameter infinite loop Original-commit: flang-compiler/f18@552568c39764972b4e21ebeb6e9f4337eee2eb38 Reviewed-on: https://github.com/flang-compiler/f18/pull/469
2019-05-23 17:11:33 +08:00
!ERROR: Initialization expression for PARAMETER 'o' (o) cannot be computed as a constant value
real, parameter :: o = o
[flang] Add semantic error test for integer constant containing division by zero Original-commit: flang-compiler/f18@650cfac5397236373e3f8c4ca7a1e15e6f33e0a4 Reviewed-on: https://github.com/flang-compiler/f18/pull/468
2019-05-24 23:22:11 +08:00
!ERROR: Must be a constant value
integer, parameter :: p = 0/0
!ERROR: Must be a constant value
integer, parameter :: q = 1+2*(1/0)
!ERROR: Must be a constant value
integer(kind=2/0) r
[flang] Fix errors in tests and predefined modules There were many places in tests and predefined modules that had incorrect code that we weren't detecting until now. Most of the problems were deferred-shape arrays that should have been implied-shape. For example: `real, parameter :: a(:) = [1.0, 2.0]` should have `(*)` rather than `(:)`. There were also a few places with deferred-shape arrays that were not allocatable/pointer or explicit-shape ones that were. Original-commit: flang-compiler/f18@0a959ce1d81dd2e54b72886db699b12c7ad59d88 Reviewed-on: https://github.com/flang-compiler/f18/pull/630
2019-08-07 05:14:33 +08:00
integer, parameter :: sok(*)=[1,2]/[1,2]
[flang] Add semantic error test for integer constant containing division by zero Original-commit: flang-compiler/f18@650cfac5397236373e3f8c4ca7a1e15e6f33e0a4 Reviewed-on: https://github.com/flang-compiler/f18/pull/468
2019-05-24 23:22:11 +08:00
!ERROR: Must be a constant value
[flang] Fix errors in tests and predefined modules There were many places in tests and predefined modules that had incorrect code that we weren't detecting until now. Most of the problems were deferred-shape arrays that should have been implied-shape. For example: `real, parameter :: a(:) = [1.0, 2.0]` should have `(*)` rather than `(:)`. There were also a few places with deferred-shape arrays that were not allocatable/pointer or explicit-shape ones that were. Original-commit: flang-compiler/f18@0a959ce1d81dd2e54b72886db699b12c7ad59d88 Reviewed-on: https://github.com/flang-compiler/f18/pull/630
2019-08-07 05:14:33 +08:00
integer, parameter :: snok(*)=[1,2]/[1,0]
[flang] Rework when symbol table expressions are evaluated It's not good enough to evaluate expressions in the symbol table after name resolution has completed. This is because we need the values of constant expressions for types, for example, we need to evaluate `k` in `integer(k) :: x` to know the type of `x`. So, eliminate `LazyExpr` and call `EvaluateExpr()` on expressions that we need in the symbol table. The latter evaluates and folds an expression in the current context. This is now possible because symbols are added to `parser::Name` as soon as possible rather than in a pass after name resolution. Along with `LazyExpr` we can eliminate the whole `ResolveSymbolExprs` pass that used to resolve them. In resolve-names.cc, many `Pre` functions are changed to `Post` so that names are resolved before doing the associated processing. For example, with intrinsic type specs, names in the kind expression must be resolved before attempting to evaluate that expression. In `GetSymbolType()` in type.cc, handle both `ObjectEntityDetails` and `EntityDetails` by using `Symbol::GetType()`. Add explicit declarations in label01.F90 because we can't handle implicitly typed array bounds yet. Original-commit: flang-compiler/f18@d67716640baeedca150205f9f3845ceb5d2f6402 Reviewed-on: https://github.com/flang-compiler/f18/pull/238
2018-12-06 22:59:37 +08:00
end