Clarify the advice about reporting errors in test hooks
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
This commit is contained in:
parent
76e9c64c3e
commit
00d0ad4036
1 changed files with 10 additions and 3 deletions
|
@ -265,9 +265,16 @@ void test_hooks_init( void );
|
|||
|
||||
/** Check if any test hooks detected a problem.
|
||||
*
|
||||
* \note When implementing a test hook, make sure to print a message
|
||||
* to standard error either at the time the problem is detected
|
||||
* or during the execution of this function.
|
||||
* If a problem was detected, it's ok for the calling program to keep going,
|
||||
* but it should ultimately exit with an error status.
|
||||
*
|
||||
* \note When implementing a test hook that detects errors on its own
|
||||
* (as opposed to e.g. leaving the error for a memory sanitizer to
|
||||
* report), make sure to print a message to standard error either at
|
||||
* the time the problem is detected or during the execution of this
|
||||
* function. This function does not indicate what problem was detected,
|
||||
* so printing a message is the only way to provide feedback in the
|
||||
* logs of the calling program.
|
||||
*
|
||||
* \return Nonzero if a problem was detected.
|
||||
* \c 0 if no problem was detected.
|
||||
|
|
Loading…
Reference in a new issue