Comments (6)
daveedvdv
commented on July 20, 2024
1
On Oct 3, 2018, at 5:24 PM, Richard Smith ***@***.***> wrote:
Testcase:
struct alignas(2 * sizeof(unsigned)) Base {
unsigned x;
~Base() = default;
};
struct Der : Base {
unsigned y;
};
If Base is POD for the purpose of layout, then sizeof(Der) == 4 * sizeof(unsigned). Otherwise, sizeof(Der) == 2 * sizeof(unsigned). All verrsions of GCC disagree with all versions of Clang on this question -- GCC believes that Base is POD for the purpose of layout, and Clang believes that it is not.
The ABI doesn't say who is right. It says that we must use the C++03 definition of POD to answer the question, which says: "A POD-struct is an aggregate class that has no non-static data members of type non-POD-struct, non-POD-union (or array of such types) or reference, and has no user-defined copy assignment operator and no user-defined destructor."
But "user-defined" is meaningless in C++11 onwards. It looks like Clang interprets it as meaning "user-declared", which makes Base non-POD, and GCC interprets it as meaning "user-provided", which makes Base POD.
So what's the rule? Is it "user-declared" or "user-provided”?
I think it should be user-provided. Ideally, adding “<special-member> = default;” where the resulting member is trivial shouldn’t change the ABI.
Daveed
… A similar situation occurs if we default the default constructor instead of the destructor. Again, GCC believes that Base is POD and Clang believes that it is not. In that case, the C++03 rules are applicable: Base is not an aggregate because in C++03, "An aggregate is an array or a class with no user-declared constructors [...]" (and for what it's worth, this rule has been changed many times since then, but in C++20, we're back to "no user-declared constructors").
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub <#66>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA8ZT-5a4kX21UPt-huhq6ItO1fHcw6hks5uhSsQgaJpZM4XG6pM>.
from cxx-abi.
zygoloid
commented on July 20, 2024
OK, that makes sense to me. From a specification point of view, I think at this point we should no longer rely on the C++03 definition and should instead directly describe what we mean:
"""
A type is considered a POD for the purposes of layout if (ignoring cv-qualifications) it is a scalar type, an array of whose element type is considered POD for the purposes of layout, or a class type that has:
- no user-provided or inherited constructors,
- no user-provided copy or move assignment operators,
- no user-provided destructors,
- no private or protected non-static data members,
- no non-static reference members,
- no bit-fields whose declared width is wider than the declared type of the bit-field,
- no potentially-overlapping non-static data members,
- no base classes,
- no virtual functions,
and for which all non-static data members are of types considered POD for the purposes of layout.
This is intended to match the definition of POD in C++03, suitably extended to later language standards, and to identify a subset of types for which the base C ABI for the platform can be expected to compute a layout.
"""
Emphasized portions of the above are suggested changes from the current rule.
from cxx-abi.
jyknight
commented on July 20, 2024
Isn't saying "no inherited constructors" unnecessary, as there's a "no base classes" rule anyways?
And just to confirm: deleted constructors/destructors/move-assignment operators also won't make a type non-pod-for-layout, correct? Including if they're implicitly deleted by the declaration of a move assignment operator.
from cxx-abi.
jwakely
commented on July 20, 2024
@jyknight special members that are implicitly defined as deleted are not "user-provided".
[dcl.fct.def.default] p5
A function is user-provided if it is user-declared and not explicitly defaulted or deleted on its first declaration.
from cxx-abi.
jyknight
commented on July 20, 2024
GCC doesn't currently check for the existence of user-provided move-assignment operators, since C++03 didn't mention those. It checks only for user-provided constructor, copy assignment operator, and destructor.
Should the list above should be modified to remove the mention of move assignment operators?
from cxx-abi.
dwblaikie
commented on July 20, 2024
This has come up again in Clang here: https://reviews.llvm.org/D119051
from cxx-abi.
Related Issues (20)
- "Deducing this" mangling HOT 14
- Should std::rethrow_exception be covered by the EH ABI? HOT 2
- Emergency EH buffer is overspecified HOT 6
- Where is the most recent ABI document? HOT 1
- Add `[[trivial_abi]]` attribute
- Lambda POD for the purposes of layout? HOT 2
- Mangling the name of an externally visible lambda in a static data member of a class HOT 1
- Proposal: Include an optional specification for mangling names that reference anonymous symbols HOT 4
- Is it possible to form a pointer-to-data-member with offset -1 using explicit derived-to-base conversions without UB? HOT 3
- unnecessary `E`s after <expression> and mangling collisions between <expression> and <number> HOT 1
- need mangling for lambdas appearing in unevaluated operands within a class body HOT 3
- What does "forbidding the use of function templates" mean? HOT 2
- [C++20] [Modules] Do we need the concept of `key function` for class defined in module purview? HOT 25
- Missing HTML encoding in 2.3.1 Data Member Pointers HOT 2
- Proposal: document or somehow notice __cxa_init_primary_exception HOT 3
- Mangling for C++ pack indexing HOT 1
- Function and function pointer types with vendor calling conventions HOT 2
- Ambiguity in mangling grammar around type qualifiers HOT 8
- Question about section 2.9.4 HOT 4
- Questions About Non-POD Types Data Layout HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cxx-abi.