linux-kernel - Re: [PATCH] perf report: auto-detect branch stack sampling 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: <CABPqkBQ6rYstBwLYsGrRZEj4faPR2i0k7b4QasrhRD-9JH9DTA@mail.gmail.com>
Date:	Wed, 7 Mar 2012 13:49:30 +0100
From:	Stephane Eranian <eranian@...gle.com>
To:	Ingo Molnar <mingo@...e.hu>
Cc:	linux-kernel@...r.kernel.org, acme@...hat.com,
	peterz@...radead.org, dsahern@...il.com, ravitillo@....gov,
	khandual@...ux.vnet.ibm.com, asharma@...com,
	robert.richter@....com, ming.m.lin@...el.com,
	vweaver1@...s.utk.edu, andi@...stfloor.org,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>
Subject: Re: [PATCH] perf report: auto-detect branch stack sampling mode

On Mon, Mar 5, 2012 at 4:47 PM, Ingo Molnar <mingo@...e.hu> wrote:
>
> Another usability bug of the new branch sampling feature is that
> typing the obvious:
>
>  $ perf record -b ./myprog
>
> does not do the obvious thing and default to 'any' but throws a
> *very* unhelpful generic perf error:
>
Ok, so looked at this. Given the option parsing code, looks like
we would have to split the option in two: -b and something else
to disambiguate cmdline such as:

$ perf record -b foo
$ perf record -b any_call foo
any_call: no such file or directory

The code cannot disambiguate between any_call being a branch filter
vs. the command
to run, given that -b has now an optional argument.

What we can do is split: -b and --branch-filter for instance. The former
takes no argument and sets up branch-stack to ANY. The other one requires
a parameter.

I am almost done with  TUI navigation of branch samples.

>  usage: perf record [<options>] [<command>]
>    or: perf record [<options>] -- <command> [<options>]
>
>  ...
>
> Really, we should do better than this. Only people who *want* to
> specify finer LBR filters should be forced to specify them.
>
> Thanks,
>
>        Ingo
--
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/