linux-kernel - Re: [PATCH v4 0/4] perf timechart io mode

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <53A9B7A4.9050805@gmail.com>
Date:	Tue, 24 Jun 2014 11:38:44 -0600
From:	David Ahern <dsahern@...il.com>
To:	Stanislav Fomichev <stfomichev@...dex-team.ru>,
	linux-kernel@...r.kernel.org
CC:	a.p.zijlstra@...llo.nl, paulus@...ba.org, mingo@...hat.com,
	acme@...nel.org, namhyung@...nel.org, artagnon@...il.com,
	rusty@...tcorp.com.au, bp@...e.de, Jiri Olsa <jolsa@...hat.com>
Subject: Re: [PATCH v4 0/4] perf timechart io mode

On 6/24/14, 10:57 AM, Stanislav Fomichev wrote:
> On Tue, Jun 24, 2014 at 01:10:22PM +0400, Stanislav Fomichev wrote:
>>> hum, got this when trying:
>>>
>>> [jolsa@...va perf]$ sudo ./perf timechart record -I
>>> ^C[ perf record: Woken up 2 times to write data ]
>>> [ perf record: Captured and wrote 1.071 MB perf.data (~46806 samples) ]
>>> [jolsa@...va perf]$ ./perf timechart
>>> Invalid previous event (non-zero)!
>>> 0x113f80 [0x8]: failed to process type: 68

I see this kind of message a lot, but typically in high event use cases 
(e.g., 100+MB files in a run of a few seconds). I usually increase the 
map size (-m 4096) and/or bump the priority of perf (-r1) and re-run the 
test.

>> Hm, I added a bunch of consistency checks which I couldn't actually hit
>> myself. Nice work :-) Can I get you perf.data?
>
> For some reason, in Jiri's trace we see exit from sys_read with 2,
> and then again exit from sys_read with -EAGAIN:
>              xterm 32028 [000] 43390.908909: syscalls:sys_enter_read: fd: 0x00000004, buf: 0x02076e8c, count: 0x0
>              xterm 32028 [000] 43390.908917: syscalls:sys_exit_read: 0x2
>              xterm 32028 [000] 43390.908982: syscalls:sys_exit_read: 0xfffffffffffffff5
>
> I'm not sure how it is possible (either we lost sys_enter, or timestamp
> of seconds -EAGAIN is wrong?). I think we might just convert this error
> to a warning (we couldn't reproduce it again).

Any lost events?

David
--
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/