[U-Boot] [PATCH v2 REPOST 2/3] image: Implement IH_TYPE_KERNEL_REL

[Simon Glass]

Hi Wolfgang,

On Thu, Oct 20, 2011 at 12:03 PM, Stephen Warren <swarren at nvidia.com> wrote:
> Wolfgang Denk wrote at Thursday, October 20, 2011 12:57 PM:
>> In message <74CDBE0F657A3D45AFBB94109FB122FF173D51C136 at HQMAIL01.nvidia.com> you wrote:
>> >
>> > > > > Please give me a bit more context - how much is "a few 10s of bytes",
>> > > > > and which exact parts of the code are we talkign about?
>> > > >
>> > > > How much? However much adding an extra value into an if check compiles to.
>> > >
>> > > As much as needed so I can see what you mean without having to go
>> > > through the whole thread.
>> >
>> > 'How much' above was a quote from where you were asking 'how much is "a
>> > few bytes"', not 'how much' explanation you want.
>>
>> You wrote "a few 10s of bytes" - this is obviously more than "adding
>
> No, Simon Glass wrote that.

Yes that was my comment. There is a bit of confusion here, sorry for
not jumping in earlier.

My comment was that if code size could increase slightly the code
could have less #ifdefs. Wolfgang said that it could not, and
therefore my comments do not apply. Please ignore them Wolfgang and
Stephen.

That takes us back to Stephen's original patch which otherwise looks
fine to me (and does not increase code size).

Regards,
Simon

>
>> an extra value into an if check compiles to", so I guess you ar enot
>> talking about doing this just once, but more frequently.  So please be
>> precise.
>>
>> > OK then, please just disregard the question and review the patch itself
>> > as I posted it. Just by reviewing and hopefully accepting the patch,
>> > you'll see the number of ifdefs and whether they're appropriate.
>>
>> Please provide this information in advance, before I start looking at
>> the patch.
>
> Sorry, but why exactly?
>
> The patch I posted doesn't add size to the U-Boot binary without the
> configuration option turned on, and I believe should be reviewable and
> hopefully even in a state ready to be replied without any further change
> or explanation.
>
> Simon Glass questioned whether the change should be allowed to increase
> binary size even when disabled, so as to avoid adding at least some of
> the ifdefs. At present, I'm not going to pursue that since it's obvious
> that's not what you want.
>
> Please, again, just take a look at the patch itself and ignore this part
> of the thread. I *really* don't think the patch needs any explanation in
> its current form.
>
> At this point, I'm honestly tempted to just give up on these U-Boot patches.
>
> --
> nvpublic
>
>