Florian Weimer, le jeu. 02 janv. 2025 11:12:55 +0100, a ecrit: > * Samuel Thibault: > > > The old BSD 4.4 definition (not used by Linux) was not 64b-proof: the > > cmsg_data field is supposed to CMSG_ALIGN'ed (as can be also seen in the > > CMSG_LEN macro). > > > > Suggested-by: Diego Nieto Cid <dnie...@gmail.com> > > --- > > bits/socket.h | 12 ++++-------- > > sysdeps/mach/hurd/bits/socket.h | 12 ++++-------- > > 2 files changed, 8 insertions(+), 16 deletions(-) > > > > diff --git a/bits/socket.h b/bits/socket.h > > index 62276fb49d..e83ad5bc01 100644 > > --- a/bits/socket.h > > +++ b/bits/socket.h > > @@ -221,17 +221,13 @@ struct cmsghdr > > of cmsghdr structure. */ > > int cmsg_level; /* Originating protocol. */ > > int cmsg_type; /* Protocol specific type. */ > > -#if __glibc_c99_flexarr_available > > - __extension__ unsigned char __cmsg_data __flexarr; /* Ancillary data. > > */ > > -#endif > > + /* This field is to be aligned with CMSG_ALIGN */ > > + /* __extension__ unsigned char __cmsg_data __flexarr; */ /* Ancillary > > data. */ > > }; > > Does the Linux version have the same problem?
By luck, no, because cmsg_len is erroneously a size_t there, and thus cmsg_data happens to be size_t-aligned on both 32b and 64b archs. If we ever get a 128b arch which does require 128b alignment, it'll have the same problem, though. Samuel
