`M5-0-8` - `M5-0-9`: Return value of `static_cast` is wrongly considered as a `cvalue`

[nbusser]

Affected rules

• M5-0-8
• M5-0-9

Description

Return value of static_cast seems to be treated as a cvalue interferring with several MISRA rules:

M-0-8

When upcasting variable using static_cast and rightaway using the result in another expression, it triggers a M5-0-9 warning (illustrated in example function false_positive).

It forces the user to create a intermediate variable containing the result of the static_cast, then using this intermediate variable in the expression (illustrated in example function true_negative).

M-0-9

When changing variable's signedness using static_cast and rightaway using the result in another expression, it triggers a M5-0-9 warning (illustrated in example function false_positive).

It forces the user to create a intermediate variable containing the result of the static_cast, then using this intermediate variable in the expression (illustrated in example function true_negative).

Example

M-0-8

void false_positive() { 
    std::vector<std::uint8_t> v{0};

    std::uint32_t u32{0};
    v.at(static_cast<std::size_t>(u32)); // Triggers a M5-0-8 warning
}

void true_negative() {
    std::vector<std::uint8_t> v{0};

    std::size_t st = static_cast<std::size_t>(u32);
    v.at(st); // Does not trigger a M5-0-8 warning
}

M-0-9

void false_positive() { 
  std::vector<std::uint8_t> v{0};

  std::int32_t s32{0};
  v.at(static_cast<std::size_t>(s32)); // Triggers a M5-0-9 warning
}

void true_negative() {
  std::vector<std::uint8_t> v{0};

  std::size_t st = static_cast<std::size_t>(s32);
  v.at(st); // Does not trigger a M5-0-9 warning
}