Re: [PATCH -tip] drivers/serial/8250.c: 'i' may be used uninitialized

[Steven Noonan]

On Wed, Oct 1, 2008 at 1:29 AM, Ingo Molnar <mingo@xxxxxxx> wrote:
>
> * Steven Noonan <steven@xxxxxxxxxxxxxx> wrote:
>
>> Signed-off-by: Steven Noonan <steven@xxxxxxxxxxxxxx>
>> ---
>>  drivers/serial/8250.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> the change is obvious to you, but it's useful to put an analysis into
> the changelog. Something like:
>
>  serial_unlink_irq_chain() does not initialize iterator 'i', and that is
>  correct logically because it is always initialized due to XYZ. Gcc does
>  not realize this connection and emits a false warning. Annotate it with
>  uninitialized_var().
>
> and fill in XYZ.
>
> Doing such changelogs is useful to maintainers: they'll see that you
> havent just squashed a warning you noticed, you understood the code and
> determined it via review that the warning is GCC's fault, not the
> kernel's.
>
> with an empty changelog the maintainer will have to do this himself.
> (and can easily put your patch to the tail of a very long TODO list, or
> outright skip your patch.)
>
>        Ingo
>

I was kind of worried about being exceedingly verbose, but I
understand your point perfectly. I'm going to resend the patch with an
appropriately verbose comment.

As always, I appreciate the criticism. Thank you! :)

- Steven
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/