Age | Commit message (Collapse) | Author |
|
mmap should return -1 on failure, not EOPNOTSUPP.
As it was now, an mmap of /dev/mem always returned page 45.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
we will accept.
|
|
|
|
|
|
messages; bug report from edmundo@rano.org
|
|
|
|
|
|
|
|
|
|
As it was now, an mmap of /dev/mem always returned page 45.
|
|
Correct line numbers. Note that autoconf 2.13 is slightly different from
the version originally used in binutils 2.10 (namely the mmap and alloca
tests have been slightly modified, with no consequences for us... the
changes are mostly for C++ compatibility).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
libkrb.
XXX - we really, really, really want to unify libroken between arla and krb4
(and krb5 in the future).
|
|
|
|
This works around the fact that M_PREPEND() with a non-word sized length
can leave m->m_data pointing to a non-word aligned address.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- Remove static.
|
|
|