This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH 10/13] Installed-header hygiene (BZ#20366): struct ether_addr.


On Wed, Sep 21, 2016 at 3:51 PM, Carlos O'Donell <carlos@redhat.com> wrote:
> On 08/29/2016 09:16 PM, Zack Weinberg wrote:
>> Some versions of netinet/if_ether.h only define struct ether_addr when
>> __USE_MISC.  This breaks function prototypes in netinet/ether.h that
>> depend on it.  The path of least resistance is to condition those
>> declarations on __USE_MISC too.
>>
>>       * inet/netinet/if_ether.h: Condition all function prototypes
>>         on __USE_MISC, to ensure struct ether_addr is declared.
>
> Wrong ChangeLog path?

Yes, that should've been inet/netinet/ether.h.

You may have already figured this out for yourself, but for the
record: There are two versions of netinet/if_ether.h in the tree:

./sysdeps/generic/netinet/if_ether.h
./sysdeps/unix/sysv/linux/netinet/if_ether.h

(There is also sysdeps/mach/hurd/net/if_ether.h but that appears to be
an unrelated header - note it's in net/ instead of netinet/ - it
defines approximately the same things that uapi <linux/if_ether.h>
does.)

The generic netinet/if_ether.h defines struct ether_addr
unconditionally.  The Linux-specific version, however, defines it only
under __USE_MISC (by conditionally including <net/ethernet.h> ... this
sure is a twisty little maze of subtly different headers, isn't it?)

inet/netinet/ether.h includes <netinet/if_ether.h> and then expects
struct ether_addr to be available.  And that's where we came in.

zw


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]