linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Wangnan (F)" <wangnan0@huawei.com>
To: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: <nilayvaish@gmail.com>, <linux-kernel@vger.kernel.org>,
	<lizefan@huawei.com>, Hou Pengyang <houpengyang@huawei.com>,
	He Kuang <hekuang@huawei.com>,
	Arnaldo Carvalho de Melo <acme@redhat.com>
Subject: Re: [PATCH 1/3 (fix commit message)] perf tools: Recognize hugetlb mapping as anon mapping
Date: Tue, 6 Sep 2016 09:22:03 +0800	[thread overview]
Message-ID: <57CE1A3B.5050905@huawei.com> (raw)
In-Reply-To: <20160905191805.GA12939@kernel.org>



On 2016/9/6 3:18, Arnaldo Carvalho de Melo wrote:
> Em Sun, Sep 04, 2016 at 05:04:38AM +0000, Wang Nan escreveu:
>> Hugetlbfs mapping should be recognized as anon mapping so user has
>> a chance to create /tmp/perf-<pid>.map file for symbol resolving. This
>> patch utilizes MAP_HUGETLB to identify hugetlb mapping.
>>
>> After this patch, if perf is started before the program starts using
>> huge pages (so perf gets MMAP2 events from kernel), perf is able to
>> recognize hugetlb mapping as anon mapping.
> Fixing build on older distros (Centos5, Ubuntu 12.04.5, etc)
>
>
>    CC       /tmp/build/perf/arch/x86/util/perf_regs.o
> util/event.c: In function 'perf_event__synthesize_mmap_events':
> util/event.c:350: error: 'MAP_HUGETLB' undeclared (first use in this function)
> util/event.c:350: error: (Each undeclared identifier is reported only once
> util/event.c:350: error: for each function it appears in.)
>
>   

Fixing MAP_HUGETLB is not as easy as it seems to be. Its value
is different in each arch:

./arch/powerpc/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x40000    
     /* create a huge page mapping */
./arch/alpha/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x100000    /* 
create a huge page mapping */
./arch/xtensa/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x80000        
/* create a huge page mapping */
./arch/tile/include/uapi/asm/mman.h:#define MAP_HUGETLB    0x4000     /* 
create a huge page mapping */
./arch/parisc/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x80000        
/* create a huge page mapping */
./arch/sparc/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x40000        
/* create a huge page mapping */
./arch/mips/include/uapi/asm/mman.h:#define MAP_HUGETLB 0x80000        
/* create a huge page mapping */
./include/uapi/asm-generic/mman.h:#define MAP_HUGETLB    0x40000     /* 
create a huge page mapping */

So we need to introduce another header into tools for this macro. Sigh...

Thank you.

>> Signed-off-by: Wang Nan <wangnan0@huawei.com>
>> Signed-off-by: Hou Pengyang <houpengyang@huawei.com>
>> Cc: He Kuang <hekuang@huawei.com>
>> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
>> Cc: Nilay Vaish <nilayvaish@gmail.com>
>> ---
>>   tools/perf/util/map.c | 8 +++++---
>>   1 file changed, 5 insertions(+), 3 deletions(-)
>>
>> diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
>> index 728129a..a42010d 100644
>> --- a/tools/perf/util/map.c
>> +++ b/tools/perf/util/map.c
>> @@ -6,6 +6,7 @@
>>   #include <string.h>
>>   #include <stdio.h>
>>   #include <unistd.h>
>> +#include <sys/mman.h>
>>   #include "map.h"
>>   #include "thread.h"
>>   #include "strlist.h"
>> @@ -24,9 +25,10 @@ const char *map_type__name[MAP__NR_TYPES] = {
>>   	[MAP__VARIABLE] = "Variables",
>>   };
>>   
>> -static inline int is_anon_memory(const char *filename)
>> +static inline int is_anon_memory(const char *filename, u32 flags)
>>   {
>> -	return !strcmp(filename, "//anon") ||
>> +	return flags & MAP_HUGETLB ||
>> +	       !strcmp(filename, "//anon") ||
>>   	       !strncmp(filename, "/dev/zero", sizeof("/dev/zero") - 1) ||
>>   	       !strncmp(filename, "/anon_hugepage", sizeof("/anon_hugepage") - 1);
>>   }
>> @@ -155,7 +157,7 @@ struct map *map__new(struct machine *machine, u64 start, u64 len,
>>   		int anon, no_dso, vdso, android;
>>   
>>   		android = is_android_lib(filename);
>> -		anon = is_anon_memory(filename);
>> +		anon = is_anon_memory(filename, flags);
>>   		vdso = is_vdso_map(filename);
>>   		no_dso = is_no_dso_memory(filename);
>>   
>> -- 
>> 1.8.3.4

  reply	other threads:[~2016-09-06  1:22 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-09-02 13:59 [PATCH 0/3] perf report: Recognize hugetlb mapping as anon mapping Wang Nan
2016-09-02 13:59 ` [PATCH 1/3] perf tools: " Wang Nan
2016-09-03 15:55   ` Nilay Vaish
2016-09-04  5:04     ` [PATCH 1/3 (fix commit message)] " Wang Nan
2016-09-05 19:18       ` Arnaldo Carvalho de Melo
2016-09-06  1:22         ` Wangnan (F) [this message]
2016-09-02 13:59 ` [PATCH 2/3] tools lib api fs: Add hugetlbfs filesystem detector Wang Nan
2016-09-02 13:59 ` [PATCH 3/3] perf record: Mark MAP_HUGETLB during synthesizing mmap events Wang Nan
2016-09-03 15:59 ` [PATCH 0/3] perf report: Recognize hugetlb mapping as anon mapping Nilay Vaish

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=57CE1A3B.5050905@huawei.com \
    --to=wangnan0@huawei.com \
    --cc=acme@kernel.org \
    --cc=acme@redhat.com \
    --cc=hekuang@huawei.com \
    --cc=houpengyang@huawei.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lizefan@huawei.com \
    --cc=nilayvaish@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).