On Fri, Jul 10, 2020 at 11:58:06AM +0100, Nick Clifton wrote:
> Hi Guys,
> 
>   I have just updated PR 26047 with a suggested patch which I think could 
> resolve
>   this situation.  (Patch attached here as well in order to save time).  It 
> adds
>   a new linker command line option: -z allowexec  which will disable the 
> warnings
>   about linking in executable files.  The default is still to have these 
> warnings
>   as I think that in most cases this behaviour makes sense.
> 
>   Will this solve the problem for you ?

Hi Nick,

Thanks for looking at this.

I think the main issue is going to be the breaking of existing
software builds.  In particular, as distros pull in the new version of
binutils, they'll run into errors building their current version of
SeaBIOS.  Even if we update SeaBIOS today, it will be some time before
the distros will pull that change in.  It'll also be a problem for
those trying to build older versions of SeaBIOS on newer toolchains.

Unless I'm missing something, this will also be a problem for those
building recent versions of the Linux kernel.  It's build used this
ability of ld up until a few months ago (see
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=90ceddcb495008ac8ba7a3dce297841efcd7d584
).  Although the latest Linux has changed, building older versions is
fairly common.  It's unclear to me if there are other common packages
that use this ability of ld.

Also, you mentioned "which will disable the warnings about linking in
executable files".  I thought PR 26047 created a fatal error in this
situation.  Is the proposed change to make the error a warning or am I
missing something?

As for the particular workaround, "-z allowexec" generates a warning
on current versions of ld.  Given that, for SeaBIOS, my first reaction
would probably be to change the SeaBIOS build to unconditionally clear
the flag in the intermediate binary than to deploy a more complicated
ld version check to set the given command-line flag.  I'd like to give
that some more thought though.

Thanks again,
-Kevin
_______________________________________________
SeaBIOS mailing list -- seabios@seabios.org
To unsubscribe send an email to seabios-le...@seabios.org

Reply via email to