Mohamed KORBOSLI korbosli@webmails.com
Sat Jul 16 07:34:00 GMT 2005

• Previous message (by thread): size of bfd_vma
• Next message (by thread): Paranoid NOP padding?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

---

Thanks

On Fri Jul 15 19:29 , Ian Lance Taylor <ian@airs.com> sent:

>Nick Clifton nickc@redhat.com> writes:
>
>> Hi Mohamed,
>> 
>> > I have already dad a look to bfd/bfd-in.h and what I understood that
>> > the bfd_vma type is defined as an "unsigned long" on 32 bits machine
>> > and as an "unsigned long long" on 64 ones;
>> > Am I correct?
>> 
>> Not quite.  A bfd_vma is defined as a BFD_HOST_U_64_BIT type which is
>> computed by the configure script as whatever C type can hold an
>> unsigned 64-bit value.  Thus on a 32-bit host it will be a "unsigned
>> long long" and on a 64-bit host it will be an "unsigned long".  Either
>> way a bfd_vma is a 64-bit unsigned value.
>
>Not always.  bfd_vma will be a 32-bit unsigned value if:
>
>1) unsigned long on the host is 32 bits;
>2) none of the targets require 64 bits, as set by either:
>   a) want64=true in config.bfd;
>   b) target_size=64 in configure.in
>3) the user did not configure with --enable-64-bit-bfd
>
>Ian


---- Msg sent via WebMails.com - http://www.webmails.com/

---

• Previous message (by thread): size of bfd_vma
• Next message (by thread): Paranoid NOP padding?
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

---

More information about the Binutils mailing list