Inconsistent use of configASSERT in kernel functions

[PwnVerse]

I was browsing through FreeRTOS's kernel code and wondered why there is inconsistent usage of configASSERTs in the many parts of the kernel code. For instance, in function pvTaskGetThreadLocalStoragePointer, the pxTCB is queried and directly referenced without checking if it exists or not here

void * pvTaskGetThreadLocalStoragePointer( TaskHandle_t xTaskToQuery,
                                               BaseType_t xIndex )
    {
        if( ( xIndex >= 0 ) &&
            ( xIndex < ( BaseType_t ) configNUM_THREAD_LOCAL_STORAGE_POINTERS ) )
        {
            pxTCB = prvGetTCBFromHandle( xTaskToQuery );
            pvReturn = pxTCB->pvThreadLocalStoragePointers[ xIndex ];
        }
        else
        {
            pvReturn = NULL;
        }

        return pvReturn;
    }

But the same pxTCB is guarded by a configASSERT in other functions where it is queried -

char * pcTaskGetName( TaskHandle_t xTaskToQuery )
{
    TCB_t * pxTCB;

    traceENTER_pcTaskGetName( xTaskToQuery );

    /* If null is passed in here then the name of the calling task is being
     * queried. */
    pxTCB = prvGetTCBFromHandle( xTaskToQuery );
    configASSERT( pxTCB );

    traceRETURN_pcTaskGetName( &( pxTCB->pcTaskName[ 0 ] ) );

    return &( pxTCB->pcTaskName[ 0 ] );
}

Is there a reason for using configASSERT in one function and not using it in another? I might be entirely wrong in my judgement. It will be very helpful to know the developer's insight as to why such a choice is made. (PS : I reported this since my fuzzer crashed complaining about NULL dereferences in the pvTaskGetThreadLocalStoragePointer)

[kar-rahul-aws]

Hi @PwnVerse
Thank you for pointing out the issue. I have raised a PR #1177 for adding the assert check for NULL TCB handle . Let me know if that addresses your concern.

[PwnVerse]

Thanks alot for the acknowledgement. Yes, it does resolve my concern.

[ActoryOu]

Closing issue since #1177 has been merged.