Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Updated references from C11->C23

...

Code Block
bgColor#FFcccc
langc
struct multi_threaded_flags {
  unsigned int flag1 : 2;
  unsigned int flag2 : 2;
};

struct multi_threaded_flags flags;

int thread1(void *arg) {
  flags.flag1 = 1;
  return 0;
}

int thread2(void *arg) {
  flags.flag2 = 2;
  return 0;
}

The C Standard, 3.1417, paragraph 3 [ISO/IEC 9899:20112024], states

NOTE Note 2 to entry: A bit-field and an adjacent non-bit-field member are in separate memory locations. The same applies to two bit-fields, if one is declared inside a nested structure declaration and the other is not, or if the two are separated by a zero-length bit-field declaration, or if they are separated by a non-bit-field member declaration. It is not safe to concurrently update two non-atomic bit-fields in the same structure if all members declared between them are also (non-zerononzero-length) bit-fields, no matter what the sizes of those intervening bit-fields happen to be.

For example, the following instruction sequence is possible:

...

Code Block
bgColor#ccccff
langc
struct multi_threaded_flags {
  unsigned char flag1;
  unsigned char flag2;
};
 
struct multi_threaded_flags flags;
 
int thread1(void *arg) {
  flags.flag1 = 1;
  return 0;
}

int thread2(void *arg) {
  flags.flag2 = 2;
  return 0;
}

Unlike C99, C11 and C23 explicitly defines define a memory location and provides the following note in subclause 3.14.17 paragraph 2 [ISO/IEC 9899:20112024]:

NOTE Note 1 to entry: Two threads of execution can update and access separate memory locations without interfering with each other.

It is almost certain that flag1 and flag2 are stored in the same word. Using a compiler that conforms to C99 or earlier, if both assignments occur on a thread-scheduling interleaving that ends with both stores occurring after one another, it is possible that only one of the flags will be set as intended. The other flag will contain its previous value because both members are represented by the same word, which is the smallest unit the processor can work on. Before the changes were made to the C Standard for C11, there were no guarantees that these flags could be modified concurrently.

...

Rule

Severity

Likelihood

Remediation Cost

Priority

Level

CON32-C

Medium

Probable

Medium

P8

L2

Automated Detection

PRQA QA-C
ToolVersionCheckerDescription
Astrée
Include Page
Astrée_V
Astrée_V

read_data_race

write_data_race

Supported by sound analysis (data race alarm)
Axivion Bauhaus Suite

Include Page
Axivion Bauhaus Suite_V
Axivion Bauhaus Suite_V

CertC-CON32
CodeSonar
Include Page
CodeSonar_V
CodeSonar_V
CONCURRENCY.DATARACE
CONCURRENCY.MAA
Data race
Multiple Accesses of Atomic
Coverity
Include Page
Coverity_V
Coverity_V
MISSING_LOCK

Partially implemented

Cppcheck Premium

Include Page
Cppcheck Premium_V
Cppcheck Premium_V

premium-cert-con32-cPartially implemented
Helix QAC

Include Page
Helix QAC_V
Helix QAC_V

C1774, C1775
Parasoft C/C++test

Include Page
Parasoft_V
Parasoft_V

CERT_C-CON32-a

Use locks to prevent race conditions when modifying bit fields

PC-lint Plus

Include Page
PC-lint Plus_V
PC-lint Plus_V

457

Partially supported: access is detected at the object level (not at the field level)

Polyspace Bug Finder

Include Page
Polyspace Bug Finder_V
Polyspace Bug Finder_V

CERT C: Rule CON32-C

Checks for data race (rule fully covered)

Include Page
PRQA QA-C_vPRQA QA-C_v1774, 1775Enforced by MTA

Related Vulnerabilities

Search for vulnerabilities resulting from the violation of this rule on the CERT website.

Bibliography

[ISO/IEC 9899:20112024]3.1417, "Memory Location"


...