Re: Weak x86 aliases

[Cherry G.Mathew <cherry%zyx.in@localhost>]

John Nemeth <jnemeth%cue.bc.ca@localhost> writes:

> On Dec 28, 11:52pm, "Mathew, Cherry G." wrote:
> } On December 28, 2018 9:54:11 PM GMT+05:30, John Nemeth <jnemeth%cue.bc.ca@localhost> wrote:
> } >On Dec 28,  7:36pm, "Cherry G.Mathew" wrote:
> } >} Maxime Villard <max%m00nbsd.net@localhost> writes:
> } >} > Le 28/12/2018 �� 14:57, Cherry G.Mathew a ��crit :
> } >} >> Maxime Villard <max%m00nbsd.net@localhost> writes:
> } >} >>>> Introduce a weak alias method of exporting different
> } >implementations
> } >} >>>> of the same API.
> } >} >>>
> } >} >>> Please revert or fix this change.
> } >} >>
> } >} >> I'm not sure what the fix is - do you have a suggestion ?
> } >} >
> } >} > either re-apply it without using weak symbols, or change the
> } >modloader
> } >} > to accept weak symbols
> } >} 
> } >} I don't like the imperative in your tone. NVMM is the user of
> } >modloader,
> } >} not PVHVM. So if you feel like your usecase needs fixing, I'd say
> } >it's
> } >} your problem - or don't use modules, but see below.
> } >
> } >     I suspect there's a language issue here due to people using
> } >English as a second language.  However, I don't see an imperative
> } >(command) here.  You asked for suggestions on how to fix a problem.
> } >He answered your question with a couple of suggestions.  That's
> } >all.
> } >
> } >     Also, I would argue that the kernel uses modloader, not the
> } >module.  In any event, as mentioned, it is your change that broke
> } >things...
> } 
> } Based on Jason's reply I suspect I've broken modules on Xen too.
> } ISTR that you did some work in this area. If you did, can you
> } comment?
>
>      I mostly worked on the infrastructure to build a seperate set
> of modules for xen and xen-pae.  On amd64, Xen modules didn't work
> due to missing modmap, which maxv fixed.  On i386, only simple
> modules worked due to not being able to find symbols.  I really
> don't know a whole lot about linkers and loaders (manipulating
> makefiles is much simpler).
>
>      However, having said that, I suspect that your work with PVHVM
> (and presumably PVH after that) will make the idea of having seperate
> modules for Xen obsolete.  I.e. it appears to me that we're headed
> to a world where a single kernel will work both on raw iron and
> under Xen.
>

I would like to maintain PV as is. It has advantages that the PVHVM
stuff doesn't have.

-- 
~cherry