linux-kernel - RE: [PATCH RFC V2 05/10] perf tools: lock to protect thread list

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: <37D7C6CF3E00A74B8858931C1DB2F077537C389A@SHSMSX103.ccr.corp.intel.com>
Date:   Mon, 18 Sep 2017 16:18:26 +0000
From:   "Liang, Kan" <kan.liang@...el.com>
To:     Jiri Olsa <jolsa@...hat.com>
CC:     "acme@...nel.org" <acme@...nel.org>,
        "peterz@...radead.org" <peterz@...radead.org>,
        "mingo@...hat.com" <mingo@...hat.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "jolsa@...nel.org" <jolsa@...nel.org>,
        "namhyung@...nel.org" <namhyung@...nel.org>,
        "Hunter, Adrian" <adrian.hunter@...el.com>,
        "Odzioba, Lukasz" <lukasz.odzioba@...el.com>,
        "ak@...ux.intel.com" <ak@...ux.intel.com>
Subject: RE: [PATCH RFC V2 05/10] perf tools: lock to protect thread list

> 
> SNIP
> 
> > +	pthread_mutex_unlock(&thread->namespaces_lock);
> > +
> >  	return 0;
> >  }
> >
> > -void thread__namespaces_id(const struct thread *thread,
> > +void thread__namespaces_id(struct thread *thread,
> >  			   u64 *dev, u64 *ino)
> >  {
> >  	struct namespaces *ns;
> >
> > +	pthread_mutex_lock(&thread->namespaces_lock);
> >  	ns = thread__namespaces(thread);
> 
> isn't it just thread__namespaces that needs this lock?

I also wanted to protect
*dev = ns ? ns->link_info[CGROUP_NS_INDEX].dev : 0;
*ino = ns ? ns->link_info[CGROUP_NS_INDEX].ino : 0;
Because I was not sure if ns is still accurate when we try to use
it later.
But for our case (perf top event synthesizing), it looks I worried too much.
Namespaces event isn't processed at all.  
So yes, we don't need patch 4 for the optimization.

Based on the same reason, I used comm_str in patch 3.
It's not help for the optimization either, but should be useful for future.

Anyway, I think I will drop patch 3 & 4 for V3.

Thanks,
Kan

> 
> if that's the case we don't need the change for __hists__add_entry in
> previous patch
> 
> jirka