Remove our definition for FUNCTION_PROFILER as it is wrong.
Note that "jsr $28,_mcount" is a macro for
ldq $27, _mcount($29) !literal!1
jsr $28, ($27), _mcount !lituse_jsr!1
1. The call to _mcount is added by alpha_expand_prologue after we load the gp.
Our _mcount uses $27 for the incoming address, unlike OSF/1 and Linux,
which use $28. This probably doesn't matter since we probably don't use $27
within _mcount itself.
2. You can't use this insn with _mcount because it uses the PLT, which clobbers
the return address in $28. Note that the prologue_mcount pattern carefully
avoids adding the lituse_jsr relocation so that we call through the GOT
directly.
Submitted by: Richard Henderson <rth@redhat.com>
for SMP in the plain profiling case. It seems to work too.
This error was not detected by LINT because LINT only compiles the
GUPROF profiling case, which is is a superset of the plain profiling
case for !SMP but which is so broken for SMP that the buggy code is
not compiled.
The nvi maintainer expects this patch to be included in the upcoming
nvi-1.81.6 unless a better fix presents itself.
The MFC reminder below should be taken as a cue to MFC rev 1.1.1.2
of vs_relative.c as well.
PR: bin/26869
Reported by: Alex Semenyaka <alexs@snark.ratmir.ru>
Obtained from: skimo@kotnet.org (via "Alexey V. Neyman" <alex.neyman@auriga.ru>)
MFC after: 1 month
- explictly say not to edit infrastructure for vendor ids (not just the
ids).
- say to enclose vendor ids and their infrastucture in ``#if 0'', and
partly explain why.
- don't set a bad example by mangling the Berkeley id infrastructure from
``static char sccsid[] ...'' to ``__RCSID(...)''.
- show a blank line between the vendor id cruft and the FreeeBSD if cruft
in the example.
- relaxed the rule about adding "From: " to say that "From: " is actually
useful if the file has been renamed.
- minor English improvements.
Discussed with: obrien
commit.
Fixed related style bugs:
basename.c: misplaced '#if 0'
dirname.c: misplaced '#if 0'
getgrent.c: missing '#if 0', and tab lossage in vendor id (the previous
commit fixed the complete corruption of the vendor id but
lost a tab)
getpwent.c: missing '#if 0'
are removed, because ports.tgz includes distfiles by accident.
However, removing files/directories is *not* mandatory. The only
requirement is "ports.tgz doesn't include distfiles/* but distfiles
directory itself". If this is correct, the right way is to set an
option to tar(1), not to include these files.
This commit is based on PR: 34194, but a little bit of modification
is made by me (original patch doesn't include 'distfiles' directory).
PR: 34194
Reviewed by: Adrian Steinmann <ast@marabu.ch> (PR submitter)
MFC after: 9 days
SIOCSIFMAC, which are for the Mandatory Access Control
implementation. This will prevent collisions on the p4 development
branch.
Obtained from: TrustedBSD Project
pam_login_access(8) and pam_securetty(8) to enforce various checks
previously done by login(1) but now handled by PAM, and pam_lastlog(8) to
record login sessions in utmp / wtmp / lastlog.
Sponsored by: DARPA, NAI Labs
it a little and try to make it more resilient to various possible failure
conditions. Change the man page accordingly, and take advantage of this
opportunity to simplify its language.
Sponsored by: DARPA, NAI Labs
the packet transfer routines, since rev.1.468 of machdep.c does this
better. I'm surprised that disabling interrupts helped much. Disabling
them in the packet receive routine is too late.
Fixed some minor style bugs in rev.1.14.
to fetch the magic word instead of useracc() plus a direct access.
This is more efficient as well as simpler and less incorrect:
- it was inefficent because useracc() takes much longer than just
accessing the data using a correct access method, at least on i386's.
- it was incorrect because direct access is incorrect unless the address
has been mapped. This and nearby direct accesses are mostly handled
better for other arches because they have to be (direct accesses don't
work).
- using magic in sigreturn is still fundamentally broken because false
matches are possible. On i386's, a false match occurs when %eip in a
new signal context happens to equal the magic value. This is not
handled better for other arches.
npx is no more mandatory than sc. Its mandatoryness went away in
rev.1.226 of i386/machdep.c 9 months before it was made mandatory in
rev.1.24 of config/mkmakefile.c.
This change is mainly to test building of minimal kernel configurations.
npx should really be even more standard than clk. It was optional mainly
so that the usual device driver configuration info could be specified in
the usual way in config files, but this hasn't been necessary for a few
years.
