[PING^5][PATCH v2] Fix 64-bit BFD detection causing build failures

[Maciej W. Rozycki]

On Tue, 15 Apr 2025, Luis Machado wrote:

> >>  Ping for the GDB side of:
> >> <https://inbox.sourceware.org/gdb-patches/e858fc66-13a2-f08d-c021-b92a757a5cd3@redhat.com/>.
> > 
> > I'll take a look at this and give this a go on my end. Sorry for the delay.

 Thank you.

> I tried building a 32-bit hosted variation, but I'm running into some 
> build issues.

 Well, 32-bit hosts don't get so much attention nowadays I'm afraid.  I 
only used a Canadian Cross to verify my fix too.

> Did you build this recently? I ask just to make sure this build issue is 
> not something specific to my system.

 No, I haven't rebuilt since the original submission.  From the error 
messages I infer it's a genuine issue with:

> commit c4fce3ef2927843f74df2b3c10f4347be6ff606f
> Author: Nick Clifton <nickc@redhat.com>
> Date:   Wed Apr 2 10:56:16 2025 +0100
> 
>     Add optional filename argument to the linker's --stats option, allowing extra resource use information to be reported.

 Thank you for tracking down the cause.  There are various ways to get it 
right, probably the simplest being the cast of `__time64_t' data items to 
`long long' and then the use of the `lld' conversion.  I'm not sure if 
it's the best here, but then I don't think there'll be any adverse effects 
from using the `long long' data type on 64-bit hosts; it certainly ends up 
the same as the `long' type at the machine code level for all the 64-bit 
psABIs I'm familiar with and the cast expands to nil.

> Maybe Nick has some thoughts on it?

 Nick?

  Maciej