lkml.org 
[lkml]   [2015]   [Jun]   [27]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [RFC PATCH 1/3] perf probe: Init symbol as kprobe if any pev is kprobe


On 2015/6/27 15:29, Masami Hiramatsu wrote:
> On 2015/06/25 19:37, Wang Nan wrote:
>> Before this patch, add_perf_probe_events() init symbol maps only for
>> uprobe if the first pev passed to it is a uprobe event. However, with
>> the incoming BPF uprobe support, now it will be possible to pass an
>> array with combined kprobe and uprobe events to add_perf_probe_events().
> This description is not correct. Actually, add_perf_probe_events already
> supports mix of uprobes and kprobes. However, from the command line syntax
> constrains the first elements of the probe_event arrays must be kprobes.
> So, if the array starts with uprobes, no kprobes should be there.
>
>> This patch check all pevs instead of the first one, and init kernel
>> symbol if any events is not uprobe.
> Anyway, I prefer to call init_symbol_maps() with "false" :)

I also prefer "false", so I try to find whether all events are uprobe
instead
of init_symbol_maps(true).

So for this patch only commit message needs to be corrected, the code is
no problem, right?

Thank you.

> Thank you,
>
>> Signed-off-by: Wang Nan <wangnan0@huawei.com>
>> ---
>> tools/perf/util/probe-event.c | 15 ++++++++++++++-
>> 1 file changed, 14 insertions(+), 1 deletion(-)
>>
>> diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c
>> index b386d2f..a2b3026 100644
>> --- a/tools/perf/util/probe-event.c
>> +++ b/tools/perf/util/probe-event.c
>> @@ -2802,8 +2802,21 @@ int cleanup_perf_probe_event(struct perf_probe_event *pev)
>> int add_perf_probe_events(struct perf_probe_event *pevs, int npevs, bool cleanup)
>> {
>> int i, ret;
>> + bool user_only = true;
>>
>> - ret = init_symbol_maps(pevs->uprobes);
>> + /* If any pev is kprobe, init kernel symbols. */
>> + for (i = 0; i < npevs; i++) {
>> + if (!pevs[i].uprobes) {
>> + user_only = false;
>> + break;
>> + }
>> + }
>> +
>> + /*
>> + * Compiler can drop user_only:
>> + * ret = init_symbol_maps(i >= npevs);
>> + */
>> + ret = init_symbol_maps(user_only);
>> if (ret < 0)
>> return ret;
>>
>>
>




\
 
 \ /
  Last update: 2015-06-27 10:01    [W:0.629 / U:0.008 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site