linux-kernel - Re: [PATCH 5/9] perf util: Handle failure case in trace

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 20 Mar 2013 10:12:41 +0900
From:	Namhyung Kim <namhyung@...nel.org>
To:	Steven Rostedt <rostedt@...dmis.org>
Cc:	Arnaldo Carvalho de Melo <acme@...stprotocols.net>,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Paul Mackerras <paulus@...ba.org>,
	Ingo Molnar <mingo@...nel.org>,
	Namhyung Kim <namhyung.kim@....com>,
	LKML <linux-kernel@...r.kernel.org>,
	Frederic Weisbecker <fweisbec@...il.com>
Subject: Re: [PATCH 5/9] perf util: Handle failure case in trace_report()

On Tue, 19 Mar 2013 10:47:53 -0400, Steven Rostedt wrote:
> On Tue, 2013-03-19 at 17:53 +0900, Namhyung Kim wrote:
>
>> @@ -3100,7 +3105,7 @@ int perf_event__process_tracing_data(union perf_event *event,
>>  		}
>>  	}
>>  
>> -	if (size_read + padding != size) {
>> +	if (size_read + padding != size || session->pevent == NULL) {
>>  		pr_err("%s: tracing data size mismatch", __func__);
>
> This is a strange error to give when pevent is NULL. Could have been a
> malloc failure.

Hmm.. right.  I was just too lazy. ;)

Will break NULL check and print "failed to process tracing data" as we
set it to NULL if something broken.

>
>
>>  		return -1;
>>  	}
>> diff --git a/tools/perf/util/trace-event-read.c b/tools/perf/util/trace-event-read.c
>> index 7cb24635adf2..129362b97ca5 100644
>> --- a/tools/perf/util/trace-event-read.c
>> +++ b/tools/perf/util/trace-event-read.c
>> @@ -293,7 +293,10 @@ ssize_t trace_report(int fd, struct pevent **ppevent, bool __repipe)
>>  	int show_version = 0;
>>  	int show_funcs = 0;
>>  	int show_printk = 0;
>> -	ssize_t size;
>> +	ssize_t size = -1;
>> +	struct pevent *pevent;
>> +
>> +	*ppevent = NULL;
>>  
>>  	calc_data_size = 1;
>>  	repipe = __repipe;
>> @@ -317,34 +320,38 @@ ssize_t trace_report(int fd, struct pevent **ppevent, bool __repipe)
>>  	file_bigendian = buf[0];
>>  	host_bigendian = bigendian();
>>  
>> -	*ppevent = read_trace_init(file_bigendian, host_bigendian);
>> -	if (*ppevent == NULL)
>> -		die("read_trace_init failed");
>> +	pevent = read_trace_init(file_bigendian, host_bigendian);
>> +	if (pevent == NULL) {
>
> Shouldn't we still set *ppevent to NULL? Or will the user now need to
> make sure that its pevent is NULL and always check for error?

I thought it's impossible to check return value for error since we don't
know how large a tracing data is, so I decided to init *ppevent to NULL
at the beginning and set it to a valid pevent right before return.  Thus
user need to check NULL before using it.

In the above case of perf_event__process_tracing_data(), it's a pipe
mode and we can know its size since it's saved in a temp file.

Thanks,
Namhyung
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/