I'm clearly biased, as I proposed this patch.
<br>
<br>
To reiterate my rational though: the chassis level product and serial are
<br>
already exposed, so this information is deemed to be important enough to
<br>
merit the code size cost.
<br>
<br>
The purpose of reading the smbios info at boot time is probably to identity
<br>
a node.
<br>
<br>
This chassis information is useless at it's intended purpose on blade
<br>
nodes, where it will only uniquely identify the parent chassis and not the
<br>
compute node itself.
<br>
<br>
Supplying the board level smbios information is more likely to achieve this
<br>
purpose, even on non blade nodes. Also, this information is much more
<br>
reliable than some other smbios attributes that are already exposed with
<br>
variable names, which may be defined and taking up space but have no actual
<br>
value.
<br>
<br>
Perhaps it's a slippery slope allowing any named smbios attributes at all?
<br>
What if we simply ifdef each class of smbios data, to make it easier to
<br>
override those which are named by default? Or, do away with all smbios
<br>
named variables and document more clearly how they values can be ready?
<br>
<br>
How can we really demonstrate if a naming a particular attribute can
<br>
justify its code size? To me this is really just a guess.
<br>
<br>
On Thursday, June 12, 2014, ipxe-devel <notifications@github.com> wrote:
<br>
<br>
> On 11/06/14 21:30, Dale Hamel wrote:
<br>
> > @mcb30 <https://github.com/mcb30> thoughts on exposing board level
<br>
> > serial / product since we already do for chassis, and board info is more
<br>
> > useful on blade nodes?
<br>
>
<br>
> All of the SMBIOS information is already exposed via constructed
<br>
> settings, e.g.
<br>
>
<br>
> ${smbios/2.7.0} # this is your "board-serial"
<br>
> ${smbios/2.5.0} # this is your "board-product"
<br>
>
<br>
> Since all of the information can already be accessed, the question
<br>
> becomes whether or not these settings will be sufficiently widely used
<br>
> (compared to all of the other information available via SMBIOS) to
<br>
> justify giving them names.
<br>
>
<br>
> There is a non-zero cost of naming a setting; each named setting costs
<br>
> approximately 20+<name_len>+<description_len> bytes: for example,
<br>
> "board-serial" will cost 20+13+18=51 bytes.
<br>
>
<br>
> So, to merge this patch, I need to know that it will be sufficiently
<br>
> useful to a sufficiently large number of people to justify the code size
<br>
> cost.
<br>
>
<br>
> Michael
<br>
>
<br>
> —
<br>
> Reply to this email directly or view it on GitHub
<br>
> <https://github.com/ipxe/ipxe/pull/21#issuecomment-45886529>.
<br>
>
<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">—<br>Reply to this email directly or <a href="https://github.com/ipxe/ipxe/pull/21#issuecomment-45888477">view it on GitHub</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/709460__eyJzY29wZSI6Ik5ld3NpZXM6QmVhY29uIiwiZXhwaXJlcyI6MTcxODE5NzYxOCwiZGF0YSI6eyJpZCI6MzQxMTQ4NTV9fQ==--eb4b151a6adb5c0f705970d2c926bf2aecf14ecc.gif" width="1" /></p>