[Lsb-infrastructure] possible toolchain issue on ppc32?

[Wichmann, Mats D]

Denis Silakov wrote:
> Wichmann, Mats D wrote:
>> I've crosschecked the build log with the missing_data.txt
>> files and adjusted them, so they are now representative
>> of the remaining Xt elements that don't have associated sizes
>> (merge is just pending now, will be revno 1453 in devel/build_env)
>> 
>> IA32 has a special exception case, four *Class elements
>> have size 8 in the database but are adjusted to 4 only for
>> this architecture, don't know what that's about
>> compositeWidgetClass, constraintWidgetClass, coreWidgetClass,
>> objectClass. 
>> 
> '4' seems to be correct value, since these interfaces return typedefs
> to some pointers.

ah, I see, so these are actually wrong on ppc32/s390 where we leave
them at 8, but they should be 4.  as you say:

> This is actually supposed to be devchk work - to correct ATsize, but
> it is not possible to check X11 headers at the moment. So I suggest
> to wait until X11 data is ready at least for devchk (for headers
> generation, huh, we've recently faced more mkheader problems).

yes, this is fine, there's no big rush - a "4.0 deliverable"
but I'm pointing out things as I notice them.