`goto` refactor broke engine initialization, blank sem_name is not an error

[skliper]

Checklist (Please check before submitting)

• I reviewed the Contributing Guide.
• I performed a cursory search to see if the bug report is relevant, not redundant, nor in conflict with other tickets.

Describe the bug
Semaphore name being blank is not an error condition, it should simply skip the semaphore get by name call. The recent refactor changed this behavior which causes the engine to crash.

Error introduced here:

CF/fsw/src/cf_cfdp.c

Lines 1046 to 1084 in eba87df

	else if (CF_AppData.config_table->chan[i].sem_name[0])
	{
	ret = OS_CountSemGetIdByName(&CF_AppData.engine.channels[i].sem_id,
	CF_AppData.config_table->chan[i].sem_name);
	if (ret != OS_SUCCESS)
	{
	CFE_EVS_SendEvent(CF_EID_ERR_INIT_SEM, CFE_EVS_EventType_ERROR,
	"CF: failed to get sem id for name %s, error=%ld",
	CF_AppData.config_table->chan[i].sem_name, (long)ret);
	break;
	}
	else
	{
	for (j = 0; j < CF_NUM_TRANSACTIONS_PER_CHANNEL; ++j, ++t)
	{
	int k;
	t->chan_num = i;
	CF_FreeTransaction(t);
	for (k = 0; k < CF_Direction_NUM; ++k, ++c)
	{
	CF_Assert((chunk_mem_offset + CF_DIR_MAX_CHUNKS[k][i]) <= CF_NUM_CHUNKS_ALL_CHANNELS);
	CF_ChunkListInit(&c->chunks, CF_DIR_MAX_CHUNKS[k][i],
	&CF_AppData.engine.chunk_mem[chunk_mem_offset]);
	chunk_mem_offset += CF_DIR_MAX_CHUNKS[k][i];
	CF_CList_InitNode(&c->cl_node);
	CF_CList_InsertBack(&CF_AppData.engine.channels[i].cs[k], &c->cl_node);
	}
	}
	for (j = 0; j < CF_NUM_HISTORIES_PER_CHANNEL; ++j)
	{
	CF_History_t *h = &CF_AppData.engine.histories[(i * CF_NUM_HISTORIES_PER_CHANNEL) + j];
	CF_CList_InitNode(&h->cl_node);
	CF_CList_InsertBack_Ex(&CF_AppData.engine.channels[i], CF_QueueIdx_HIST_FREE, &h->cl_node);
	}
	}
	}

Old code:

CF/fsw/src/cf_cfdp.c

Lines 1029 to 1064 in bff67f8

	if (CF_AppData.config_table->chan[i].sem_name[0])
	{
	ret = OS_CountSemGetIdByName(&CF_AppData.engine.channels[i].sem_id,
	CF_AppData.config_table->chan[i].sem_name);
	if (ret != OS_SUCCESS)
	{
	CFE_EVS_SendEvent(CF_EID_ERR_INIT_SEM, CFE_EVS_EventType_ERROR,
	"CF: failed to get sem id for name %s, error=%ld",
	CF_AppData.config_table->chan[i].sem_name, (long)ret);
	goto err_out;
	}
	}
	for (j = 0; j < CF_NUM_TRANSACTIONS_PER_CHANNEL; ++j, ++t)
	{
	int k;
	t->chan_num = i;
	CF_FreeTransaction(t);
	for (k = 0; k < CF_Direction_NUM; ++k, ++c)
	{
	CF_Assert((chunk_mem_offset + CF_DIR_MAX_CHUNKS[k][i]) <= CF_NUM_CHUNKS_ALL_CHANNELS);
	CF_ChunkListInit(&c->chunks, CF_DIR_MAX_CHUNKS[k][i], &CF_AppData.engine.chunk_mem[chunk_mem_offset]);
	chunk_mem_offset += CF_DIR_MAX_CHUNKS[k][i];
	CF_CList_InitNode(&c->cl_node);
	CF_CList_InsertBack(&CF_AppData.engine.channels[i].cs[k], &c->cl_node);
	}
	}
	for (j = 0; j < CF_NUM_HISTORIES_PER_CHANNEL; ++j)
	{
	CF_History_t *h = &CF_AppData.engine.histories[(i * CF_NUM_HISTORIES_PER_CHANNEL) + j];
	CF_CList_InitNode(&h->cl_node);
	CF_CList_InsertBack_Ex(&CF_AppData.engine.channels[i], CF_QueueIdx_HIST_FREE, &h->cl_node);
	}

To Reproduce
Initialize engine without a semaphore name defined

Expected behavior
Still initialize the transactions if the semaphore name is blank

System observed on:
Observed from CTF testing (on Linux I assume)

Additional context
None

Reporter Info
Jacob Hageman - NASA/GSFC

[skliper]

I think you can flatten much of that logic since the break makes the additional nesting unnecessary... pseudocode pattern:

ret = function
if (ret != SUCCESS)
{
   send event:
   break;
}

ret = function
...

still wrap getting the sem id such that it's only performed if it's not empty.