[DTrace-devel] [PATCH 13/17] fbt, syscall: use getline(), not fgets()

[Nick Alcock]

On 1 Sep 2022, Eugene Loh via DTrace-devel verbalised:

> On 8/10/22 18:07, Nick Alcock via DTrace-devel wrote:
>> diff --git a/libdtrace/dt_prov_fbt.c b/libdtrace/dt_prov_fbt.c
>> @@ -77,27 +77,16 @@ static int populate(dtrace_hdl_t *dtp)
>>   -	while (fgets(buf, sizeof(buf), f)) {
>> +	while (getline(&buf, &n, f) >= 0) {
>>   		/*
>>   		 * Here buf is either "funcname\n" or "funcname [modname]\n".
>> +		 * The last line may not have a linefeed.
>>   		 */
>>   		p = strchr(buf, '\n');
>>   		if (p) {
>>   			*p = '\0';
>>   			if (p > buf && *(--p) == ']')
>>   				*p = '\0';
>> -		} else {
>> -			/*
>> -			 * If we didn't see a newline, the line was too long.
>> -			 * Report it, and skip until the end of the line.
>> -			 */
>> -			fprintf(stderr, "%s: Line too long: %s\n",
>> -				PROBE_LIST, buf);
>> -
>> -			do
>> -				fgets(buf, sizeof(buf), f);
>> -			while (strchr(buf, '\n') == NULL);
>> -			continue;
>>   		}
>
> Is there a pre-existing problem here?  What if the last line does not have a linefeed?  Do we then fail to strip off the trailing
> ']'?

Oof. Yeah. However, that can only happen if the kernel truncates the
output, in which case it's pretty unlikely we'll get a trailing ]
anyway. If present, it would be the very last character before the
trailing linefeed, after all. So this is harmless, I think.

-- 
NULL && (void)