Re: [bug?] checkout -m doesn't work without a base version

From: Junio C Hamano <gitster@pobox.com>
Date: Dec 14, 2011, 17:54 UTC
Message-ID: <7vhb13qbs6.fsf@alter.siamese.dyndns.org>
In-Reply-To: <4EE8782A.9040507@elegosoft.com>

Michael Schubert <mschub@elegosoft.com> writes:

Show 26 quoted lines

>> ...
>> +	memset(threeway, 0, sizeof(threeway));
>> +	while (pos < active_nr) {
>> +		int stage;
>> +		stage = ce_stage(ce);
>> +		if (!stage || strcmp(path, ce->name))
>> +			break;
>> +		hashcpy(threeway[stage - 1], ce->sha1);
>> +		if (stage == 2)
>> +			mode = create_ce_mode(ce->ce_mode);
>> +		pos++;
>> +		ce = active_cache[pos];
>> +	}
>> +	if (is_null_sha1(threeway[1]) || is_null_sha1(threeway[2]))
>> +		return error(_("path '%s' does not have necessary versions"), path);
>> ...
>> +	ce = make_cache_entry(mode, sha1, path, 2, 0);
>>  	if (!ce)
>>  		die(_("make_cache_entry failed for path '%s'"), path);
>>  	status = checkout_entry(ce, state, NULL);
>
> gcc 4.6.2:
>
> builtin/checkout.c: In function ‘cmd_checkout’:
> builtin/checkout.c:210:5: warning: ‘mode’ may be used uninitialized in this function [-Wuninitialized]
> builtin/checkout.c:160:11: note: ‘mode’ was declared here

Isn't this just your gcc being overly cautious (aka "silly")?

The variable "mode" is assigned to when we see an stage #2 entry in the loop, and we should have updated threeway[1] immediately before doing so. If threeway[1] is not updated, we would have already returned before using the variable in make_cache_entry().