[Bug binutils/25803] cross compilation of glibc using mips64el-none-linux-gnu as the host

[broly at mac dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=25803

--- Comment #31 from gagan singh sidhu (gagz, broly, w/e u want) <broly at mac 
dot com> ---
that is fair.

and you are right. if those messages were available, it would have been
helpful.

and yeah my change doesn’t really provide anything in terms of output unless
someone pores through config.log and sees the output, but otherwise there is no
hint.

no problem. my apologies for being an armchair quarterback. i just try to
exercise extreme caution when making changes to probably one of the most
critical (and free!) tools in the world.

thank you sire

Thanks,
Gagan

> On Apr 17, 2020, at 9:27 AM, nickc at redhat dot com <address@hidden> wrote:
> 
> https://sourceware.org/bugzilla/show_bug.cgi?id=25803
> 
> --- Comment #30 from Nick Clifton <nickc at redhat dot com> ---
> (In reply to gagan singh sidhu (gagz, broly, w/e u want) from comment #29)
> 
>> i saw your commit and while it’s not wrong, i was hoping you would be open
>> to suggestions:
>> 
>> 1. we don’t need both the changes to elfxx-mips.c and obj-elf.c. 
>>      -the change to obj-elf.c will avoid us ever having the situation again 
>> in
>> elfxx-mips.c
>> 
>> so in terms of a minimalist aspect, i was hoping you would pick whichever
>> change you wanted, and disregard the other.
>>      -the whole “less changes is a good thing” paradigm
> 
> Except that in this case we are fixing two bugs.  The first was that
> the linker was not providing a helpful error message that would allow
> the user to investigate the problem with their link.  Ie if the linker
> had provided the name(s) of the invalid symbols when this problem
> first arose then someone may have worked out that the issue was ifunc
> symbols a long time ago.  Hence I think that keeping the change to
> elfxx-mips.c is a good idea.
> 
> The change to obj-elf.c fixes the underlying problem, hence it needs
> to remain.  But being more helpful to users is always a good thing,
> so I am going to leave both patches in.
> 
> -- 
> You are receiving this mail because:
> You reported the bug.

-- 
You are receiving this mail because:
You are on the CC list for the bug.