netdev - Re: [RFC perf,bpf 1/5] perf, bpf: Introduce PERF_RECORD_BPF

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: <050ABAC6-6C3B-4B6B-BB68-727127E00B51@fb.com>
Date:   Thu, 8 Nov 2018 18:49:19 +0000
From:   Song Liu <songliubraving@...com>
To:     David Ahern <dsahern@...il.com>
CC:     Peter Zijlstra <peterz@...radead.org>,
        Netdev <netdev@...r.kernel.org>,
        lkml <linux-kernel@...r.kernel.org>,
        Kernel Team <Kernel-team@...com>,
        "ast@...nel.org" <ast@...nel.org>,
        "daniel@...earbox.net" <daniel@...earbox.net>,
        "acme@...nel.org" <acme@...nel.org>
Subject: Re: [RFC perf,bpf 1/5] perf, bpf: Introduce PERF_RECORD_BPF_EVENT

Hi David,

> On Nov 8, 2018, at 10:26 AM, David Ahern <dsahern@...il.com> wrote:
> 
> On 11/8/18 11:04 AM, Song Liu wrote:
>> On the other hand, processing BPF load/unload events synchronously should
>> not introduce too much overhead for meaningful use cases. If many BPF progs
>> are being loaded/unloaded within short period of time, it is not the steady
>> state that profiling works care about. 
> 
> but, profiling is not the only use case, and perf-record is common with
> those other use cases.
> 
> I think that answers why your RFC set does not fork a thread for the bpf
> events. You are focused on profiling and for already loaded programs or
> for a small number of programs loaded by a specific workload started by
> perf.

We sure can fork a thread for the BPF event. But I guess that's not Peter's
main concern here...

Could you please point me to more information about the use cases you worry 
about? I am more than happy to optimize the logic for those use cases. 

Thanks,
Song