bug-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH1/4] fully enable rpctrace to trace multitask programs.


From: olafBuddenhagen
Subject: Re: [PATCH1/4] fully enable rpctrace to trace multitask programs.
Date: Mon, 27 Jul 2009 08:11:39 +0200
User-agent: Mutt/1.5.19 (2009-01-05)

Hi,

First of all, when submitting a patch series, each individual message
should by named after the commit, i.e. the first line of the commit
message. git format-patch should in fact do that automatically -- though
I must admit that I have trouble using it properly myself :-)

The name of the patch series as a whole is only mentioned in the "cover
letter" (if there is one), i.e. a mail with [PATCH 0/x] in the subject,
and an explanation of the patch series in the body.

On Mon, Jul 20, 2009 at 07:38:27PM +0800, Da Zheng wrote:

> 2009-07-20  Zheng Da  <zhengda1936@gmail.com>
> 
> fix a bug #20612
> 
>       * rpctrace.c (print_contents): Check the number of elements in the port 
> array.

Note that we are using git now, which somewhat changes the requirements
regarding the commit message. Specifically, the author name and date
should *not* be mentioned any longer -- git am extracts that information
directly from the mail header...

Also, please try to make the first line as descriptive as possible.
(While keeping it a short sentence...) Just mentioning a bug number is
problematic, as people first need to find out what the bug is in order
to understand the commit... Mentioning the bug number is definitely
useful -- but please also describe what the commit actually does :-)

> 
> diff --git a/utils/rpctrace.c b/utils/rpctrace.c
> index d80f41d..b7379a7 100644
> --- a/utils/rpctrace.c
> +++ b/utils/rpctrace.c
> @@ -631,7 +631,7 @@ print_contents (mach_msg_header_t *inp,
>             else
>               type->msgt_name = name;
>           }
> -       else if (newtypes[0] != name)
> +       else if (nelt > 0 && newtypes[0] != name)
>           if (type->msgt_longform)
>             lt->msgtl_name = newtypes[0];
>           else

OK, this seems an obvious fix :-)

It doesn't seem related to tracing multiple tasks, though -- so it
shouldn't be part of the patch series, but rather applied independently.

-antrik-




reply via email to

[Prev in Thread] Current Thread [Next in Thread]