Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Mon, 08 Jun 2020 15:12:03 -0300
User-Agent: K-9 Mail for Android
In-Reply-To: <CAEf4BzbEcV6YaezP4yY8J=kYSBhh0cRHCvgCUe9xvB12mF08qg@mail.gmail.com>
References: <20200608161150.GA3073@kernel.org> <CAEf4BzbEcV6YaezP4yY8J=kYSBhh0cRHCvgCUe9xvB12mF08qg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable
Subject: Re: libbpf's hashmap use of __WORDSIZE
To:     Andrii Nakryiko <andrii.nakryiko@gmail.com>
CC:     Peter Zijlstra <a.p.zijlstra@chello.nl>,
        Ian Rogers <irogers@google.com>,
        Namhyung Kim <namhyung@kernel.org>,
        Pekka Enberg <penberg@kernel.org>,
        Jiri Olsa <jolsa@kernel.org>,
        Irina Tirdea <irina.tirdea@intel.com>,
        bpf <bpf@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
From:   Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>
Message-ID: <1BB90A3B-1372-487E-9E96-193AAAEBC095@gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Mi

On June 8, 2020 2:32:16 PM GMT-03:00, Andrii Nakryiko <andrii=2Enakryiko@g=
mail=2Ecom> wrote:
>On Mon, Jun 8, 2020 at 9:11 AM Arnaldo Carvalho de Melo
><arnaldo=2Emelo@gmail=2Ecom> wrote:
>>
>> Hi Andrii,
>>
>>         We've got that hashmap=2E[ch] copy from libbpf so that we can
>> build perf in systems where libbpf isn't available, and to make it
>build
>> in all the containers I regularly test build perf I had to add the
>patch
>> below, I test build with many versions of both gcc and clang and
>> multiple libcs=2E
>>
>>   https://gcc=2Egnu=2Eorg/onlinedocs/cpp/Common-Predefined-Macros=2Ehtm=
l
>>
>> The way that tools/include/linux/bitops=2Eh has been doing since 2012
>is
>> explained in:
>>
>>   http://git=2Ekernel=2Eorg/torvalds/c/3f34f6c0233ae055b5
>>
>> Please take a look and see if you find it acceptable,
>>
>> Thanks,
>>
>> - Arnaldo
>>
>>   Warning: Kernel ABI header at 'tools/perf/util/hashmap=2Eh' differs
>from latest version at 'tools/lib/bpf/hashmap=2Eh'
>>   diff -u tools/perf/util/hashmap=2Eh tools/lib/bpf/hashmap=2Eh
>>
>> $ diff -u tools/lib/bpf/hashmap=2Eh tools/perf/util/hashmap=2Eh
>> --- tools/lib/bpf/hashmap=2Eh     2020-06-05 13:25:27=2E822079838 -0300
>> +++ tools/perf/util/hashmap=2Eh   2020-06-05 13:25:27=2E838079794 -0300
>> @@ -10,10 +10,9 @@
>>
>>  #include <stdbool=2Eh>
>>  #include <stddef=2Eh>
>> -#ifdef __GLIBC__
>> -#include <bits/wordsize=2Eh>
>> -#else
>> -#include <bits/reg=2Eh>
>> +#include <limits=2Eh>
>> +#ifndef __WORDSIZE
>> +#define __WORDSIZE (__SIZEOF_LONG__ * 8)
>>  #endif
>
>This looks fine, I also build-tested it in Travis CI, so all good=2E
>There is actually __SIZEOF_SIZE_T__, which is more directly what
>hash_bits work with, but I don't think it matters for any reasonable
>system in use :)
>
>So yeah,
>
>Acked-by: Andrii Nakryiko <andriin@fb=2Ecom>
>
>Are you going to do this change for libbpf's variant, or should I
>submit a separate patch?

I'll send the patch later,

Thanks for checking,

- Arnaldo
>
>>
>>  static inline size_t hash_bits(size_t h, int bits)

--=20
Sent from my Android device with K-9 Mail=2E Please excuse my brevity=2E