lkml.org 
[lkml]   [2013]   [Mar]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 0/7] uretprobes: return probes implementation
On Fri, Mar 22, 2013 at 02:40:42PM -0700, Josh Stone wrote:
> On 03/22/2013 08:10 AM, Oleg Nesterov wrote:
> > This looks simple, but probably we need to add the additional "ulong bp_vaddr"
> > argument for rp_handler().
>
> FWIW, in SystemTap we don't currently do anything that would need
> bp_vaddr. The uprobe_consumer already gives the context, so I'm not
> sure what else you would do with the *entry* address of the function
> that just returned. Perhaps you could emulate an additional top line on
> a stacktrace, but it doesn't tell you where in the function you were.

IIUC, Oleg have stressed that perf and trace events requires
additional code to leverage the return probes. IOW this patchset
implements core only and to use it we need to teach the perf about
rp_handler().
And I don't see why we would need additional parameter for
rp_handler() as well.

Anton


\
 
 \ /
  Last update: 2013-03-23 08:01    [W:0.309 / U:0.012 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site