[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <A489A0C6-D2FD-4FA5-8682-DD3394A6306A@fb.com>
Date: Wed, 6 Feb 2019 04:07:27 +0000
From: Song Liu <songliubraving@...com>
To: Alexei Starovoitov <alexei.starovoitov@...il.com>
CC: Daniel Borkmann <daniel@...earbox.net>,
Alexei Starovoitov <ast@...nel.org>,
Kernel Team <Kernel-team@...com>,
Netdev <netdev@...r.kernel.org>
Subject: Re: Pull patches from tip/perf/core to bpf-next
> On Feb 5, 2019, at 7:47 PM, Alexei Starovoitov <alexei.starovoitov@...il.com> wrote:
>
> On Wed, Feb 06, 2019 at 03:44:29AM +0000, Song Liu wrote:
>>
>>
>>> On Feb 5, 2019, at 7:36 PM, Alexei Starovoitov <alexei.starovoitov@...il.com> wrote:
>>>
>>> On Tue, Feb 05, 2019 at 10:47:06PM +0000, Song Liu wrote:
>>>> Hi Alexei and Daniel,
>>>>
>>>> The following patches are required for BPF introspection in perf tools.
>>>> Please pull them to bpf-next, so that we get all the dependencies in one
>>>> tree.
>>>>
>>>> Thanks,
>>>> Song
>>>>
>>>> (from 1/10 to 10/10)
>>>> 76193a94522f perf, bpf: Introduce PERF_RECORD_KSYMBOL
>>>> d764ac646491 tools headers uapi: Sync tools/include/uapi/linux/perf_event.h
>>>> 6ee52e2a3fe4 perf, bpf: Introduce PERF_RECORD_BPF_EVENT
>>>> df063c83aa2c tools headers uapi: Sync tools/include/uapi/linux/perf_event.h
>>>> 9aa0bfa370b2 perf tools: Handle PERF_RECORD_KSYMBOL
>>>> 45178a928a4b perf tools: Handle PERF_RECORD_BPF_EVENT
>>>> 7b612e291a5a perf tools: Synthesize PERF_RECORD_* for loaded BPF programs
>>>> a40b95bcd30c perf top: Synthesize BPF events for pre-existing loaded BPF programs
>>>> 6934058d9fb6 bpf: Add module name [bpf] to ksymbols for bpf programs
>>>> 811184fb6977 perf bpf: Fix synthesized PERF_RECORD_KSYMBOL/BPF_EVENT
>>>
>>> yes. we can certainly do that.
>>> Do you have bpf specific patches that depend on that ?
>>> Since it's rc5 already. Are you planning to send them within next week?
>>
>> BPF introspection work depends on these patches. I have been hopping
>> between perf tree and bpf-next tree. I think basing the series up on
>> bpf-next plus these patches leads least conflicts.
>>
>> I do plan to send the series within next week.
>>
>> On a second thought, maybe I should send based on perf tree, and worry
>> about the conflicts later? It is really heavier on perf side.
>
> whichever way is easier.
> if bpf-next is the best use that as a base with above patches.
> Once your set gets Acks from perf folks we can push above patches
> from tip first and then apply your set.
I see. Let's wait until the patchset is acked.
Thanks,
Song
Powered by blists - more mailing lists