2009-12-21 09:26:22

by tip-bot for Wenji Huang

Subject: [PATCH] perf: fix incorrect return value for else case

Return original cmd instead of adding prefix.

Signed-off-by: Wenji Huang <[email protected]>
---
tools/perf/builtin-help.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
index 9f810b1..ca77df5 100644
--- a/tools/perf/builtin-help.c
+++ b/tools/perf/builtin-help.c
@@ -317,7 +317,7 @@ static const char *cmd_to_page(const char *perf_cmd)
else if (is_perf_command(perf_cmd))
return prepend("perf-", perf_cmd);
else
- return prepend("perf-", perf_cmd);
+ return perf_cmd;
}

static void setup_man_path(void)
--
1.5.6

2009-12-21 15:23:23

by John Kacur

Subject: Re: [PATCH] perf: fix incorrect return value for else case

On Mon, Dec 21, 2009 at 10:22 AM, Wenji Huang <[email protected]> wrote:
> Return original cmd instead of adding prefix.
>
> Signed-off-by: Wenji Huang <[email protected]>
> ---
> ?tools/perf/builtin-help.c | ? ?2 +-
> ?1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
> index 9f810b1..ca77df5 100644
> --- a/tools/perf/builtin-help.c
> +++ b/tools/perf/builtin-help.c
> @@ -317,7 +317,7 @@ static const char *cmd_to_page(const char *perf_cmd)
> ? ? ? ?else if (is_perf_command(perf_cmd))
> ? ? ? ? ? ? ? ?return prepend("perf-", perf_cmd);
> ? ? ? ?else
> - ? ? ? ? ? ? ? return prepend("perf-", perf_cmd);
> + ? ? ? ? ? ? ? return perf_cmd;
> ?}
>
> ?static void setup_man_path(void)
> --
> 1.5.6

Sorry - I believe we should NAK this patch.
It would turn the following

./perf nonsuchcmd --help
No manual entry for perf-nonsuchcmd

into

./perf nonsuchcmd --help
No manual entry for nonsuchcmd

The former is correct, the name of the man page includes the prefix "perf-"

NAK

(cc-ing Frederic in case he sees it differently)

Thank You

John Kacur

2009-12-22 01:32:47

by tip-bot for Wenji Huang

Subject: Re: [PATCH] perf: fix incorrect return value for else case

John Kacur wrote:
> On Mon, Dec 21, 2009 at 10:22 AM, Wenji Huang <[email protected]> wrote:
>> Return original cmd instead of adding prefix.
>>
>> Signed-off-by: Wenji Huang <[email protected]>
>> ---
>> tools/perf/builtin-help.c | 2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
>> index 9f810b1..ca77df5 100644
>> --- a/tools/perf/builtin-help.c
>> +++ b/tools/perf/builtin-help.c
>> @@ -317,7 +317,7 @@ static const char *cmd_to_page(const char *perf_cmd)
>> else if (is_perf_command(perf_cmd))
>> return prepend("perf-", perf_cmd);
>> else
>> - return prepend("perf-", perf_cmd);
>> + return perf_cmd;
>> }
>>
>> static void setup_man_path(void)
>> --
>> 1.5.6
>
> Sorry - I believe we should NAK this patch.
> It would turn the following
>
> ./perf nonsuchcmd --help
> No manual entry for perf-nonsuchcmd
>
> into
>
> ./perf nonsuchcmd --help
> No manual entry for nonsuchcmd
>
> The former is correct, the name of the man page includes the prefix "perf-"
>
> NAK
>
> (cc-ing Frederic in case he sees it differently)
>
Thanks. Since we think the former is better, why not make
the code compact? Like,

diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
index 9f810b1..65e2691 100644
--- a/tools/perf/builtin-help.c
+++ b/tools/perf/builtin-help.c
@@ -314,8 +314,6 @@ static const char *cmd_to_page(const char *perf_cmd)
return "perf";
else if (!prefixcmp(perf_cmd, "perf"))
return perf_cmd;
- else if (is_perf_command(perf_cmd))
- return prepend("perf-", perf_cmd);
else
return prepend("perf-", perf_cmd);
}

2009-12-30 22:12:21

by Frederic Weisbecker

Subject: Re: [PATCH] perf: fix incorrect return value for else case

On Tue, Dec 22, 2009 at 09:27:12AM +0800, Wenji Huang wrote:
> John Kacur wrote:
>> On Mon, Dec 21, 2009 at 10:22 AM, Wenji Huang <[email protected]> wrote:
>>> Return original cmd instead of adding prefix.
>>>
>>> Signed-off-by: Wenji Huang <[email protected]>
>>> ---
>>> tools/perf/builtin-help.c | 2 +-
>>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
>>> index 9f810b1..ca77df5 100644
>>> --- a/tools/perf/builtin-help.c
>>> +++ b/tools/perf/builtin-help.c
>>> @@ -317,7 +317,7 @@ static const char *cmd_to_page(const char *perf_cmd)
>>> else if (is_perf_command(perf_cmd))
>>> return prepend("perf-", perf_cmd);
>>> else
>>> - return prepend("perf-", perf_cmd);
>>> + return perf_cmd;
>>> }
>>>
>>> static void setup_man_path(void)
>>> --
>>> 1.5.6
>>
>> Sorry - I believe we should NAK this patch.
>> It would turn the following
>>
>> ./perf nonsuchcmd --help
>> No manual entry for perf-nonsuchcmd
>>
>> into
>>
>> ./perf nonsuchcmd --help
>> No manual entry for nonsuchcmd
>>
>> The former is correct, the name of the man page includes the prefix "perf-"
>>
>> NAK
>>
>> (cc-ing Frederic in case he sees it differently)


I personally don't mind.
Having either

./perf nonsuchcmd --help
No manual entry for perf-nonsuchcmd
or

./perf nonsuchcmd --help
No manual entry for nonsuchcmd

both make sense for the user.


>>
> Thanks. Since we think the former is better, why not make
> the code compact? Like,
>
> diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
> index 9f810b1..65e2691 100644
> --- a/tools/perf/builtin-help.c
> +++ b/tools/perf/builtin-help.c
> @@ -314,8 +314,6 @@ static const char *cmd_to_page(const char *perf_cmd)
> return "perf";
> else if (!prefixcmp(perf_cmd, "perf"))
> return perf_cmd;
> - else if (is_perf_command(perf_cmd))
> - return prepend("perf-", perf_cmd);
> else
> return prepend("perf-", perf_cmd);
> }


Agreed!
Thanks.

2010-01-13 08:47:17

by Ingo Molnar

Subject: Re: [PATCH] perf: fix incorrect return value for else case

* Frederic Weisbecker <[email protected]> wrote:

> On Tue, Dec 22, 2009 at 09:27:12AM +0800, Wenji Huang wrote:
> > John Kacur wrote:
> >> On Mon, Dec 21, 2009 at 10:22 AM, Wenji Huang <[email protected]> wrote:
> >>> Return original cmd instead of adding prefix.
> >>>
> >>> Signed-off-by: Wenji Huang <[email protected]>
> >>> ---
> >>> tools/perf/builtin-help.c | 2 +-
> >>> 1 files changed, 1 insertions(+), 1 deletions(-)
> >>>
> >>> diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
> >>> index 9f810b1..ca77df5 100644
> >>> --- a/tools/perf/builtin-help.c
> >>> +++ b/tools/perf/builtin-help.c
> >>> @@ -317,7 +317,7 @@ static const char *cmd_to_page(const char *perf_cmd)
> >>> else if (is_perf_command(perf_cmd))
> >>> return prepend("perf-", perf_cmd);
> >>> else
> >>> - return prepend("perf-", perf_cmd);
> >>> + return perf_cmd;
> >>> }
> >>>
> >>> static void setup_man_path(void)
> >>> --
> >>> 1.5.6
> >>
> >> Sorry - I believe we should NAK this patch.
> >> It would turn the following
> >>
> >> ./perf nonsuchcmd --help
> >> No manual entry for perf-nonsuchcmd
> >>
> >> into
> >>
> >> ./perf nonsuchcmd --help
> >> No manual entry for nonsuchcmd
> >>
> >> The former is correct, the name of the man page includes the prefix "perf-"
> >>
> >> NAK
> >>
> >> (cc-ing Frederic in case he sees it differently)
>
>
> I personally don't mind.
> Having either
>
> ./perf nonsuchcmd --help
> No manual entry for perf-nonsuchcmd
> or
>
> ./perf nonsuchcmd --help
> No manual entry for nonsuchcmd
>
> both make sense for the user.
>
>
> >>
> > Thanks. Since we think the former is better, why not make
> > the code compact? Like,
> >
> > diff --git a/tools/perf/builtin-help.c b/tools/perf/builtin-help.c
> > index 9f810b1..65e2691 100644
> > --- a/tools/perf/builtin-help.c
> > +++ b/tools/perf/builtin-help.c
> > @@ -314,8 +314,6 @@ static const char *cmd_to_page(const char *perf_cmd)
> > return "perf";
> > else if (!prefixcmp(perf_cmd, "perf"))
> > return perf_cmd;
> > - else if (is_perf_command(perf_cmd))
> > - return prepend("perf-", perf_cmd);
> > else
> > return prepend("perf-", perf_cmd);
> > }
>
>
> Agreed!
> Thanks.

if this looks good to everyone please re-send the patch with a changelog and
with acks added.

Thanks,

Ingo