[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20091123212501.GA18917@elte.hu>
Date: Mon, 23 Nov 2009 22:25:01 +0100
From: Ingo Molnar <mingo@...e.hu>
To: Frederic Weisbecker <fweisbec@...il.com>
Cc: Arnaldo Carvalho de Melo <acme@...stprotocols.net>,
"K.Prasad" <prasad@...ux.vnet.ibm.com>,
LKML <linux-kernel@...r.kernel.org>,
Peter Zijlstra <peterz@...radead.org>,
Paul Mackerras <paulus@...ba.org>
Subject: Re: [PATCH 4/4] perf tools: Add support for breakpoint events in
perf tools
* Frederic Weisbecker <fweisbec@...il.com> wrote:
> > Using just /proc/kallsyms all we can do is find the size of a
> > variable by looking at its address and the address of the next one.
> >
> > - Arnaldo
>
> Hmm, but I worry a bit about alignment which would return us the wrong
> size.
>
> May be can we first try to get the address from /proc/kallsyms, and if
> we have dwarf, get the size from it, otherwise try some magic with
> /proc/kallsysms...
Can we extend /proc/kallsyms (or add /proc/kallsyms) to include a size
field?
Perhaps can we generate some sort of DSO-alike thing in /proc/vmlinux
(via a default-off debug option in .config), that perf could just
interpret the usual ELF way - which happens to be the symbol table of
the kernel? It would use up some RAM, but it would also be quite useful
for debugging purposes.
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/
Powered by blists - more mailing lists