filedesc is already locked rather than having chroot() unlock the
filedesc so chroot_refuse_vdir_fds() can immediately relock it.
- Reorder chroot() a bitso that we do the namei lookup before checking
the process's struct filedesc. This closes at least one potential race
and allows us to only acquire the filedsec lock once in chroot().
- Push down Giant slightly into chroot().
_vm_map_lock_read(), and _vm_map_trylock(). Submitted by: tegge
o Remove GIANT_REQUIRED from kmem_alloc_wait() and kmem_free_wakeup().
(This clears the way for exec_map accesses to move outside of Giant.
The exec_map is not a system map.)
o Remove some premature MPSAFE comments.
Reviewed by: tegge
This was always broken in HEAD (the offending statement was introduced
in rev. 1.123 for HEAD, while RELENG_4 included this fix (in rev.
1.99.2.12 for RELENG_4) and I inadvertently deleted it in 1.99.2.30.
So I am also restoring these two lines in RELENG_4 now.
We might need another few things from 1.99.2.30.
page-zeroing code as well as from the general page-zeroing code and use a
lazy tlb page invalidation scheme based on a callback made at the end
of mi_switch.
A number of people came up with this idea at the same time so credit
belongs to Peter, John, and Jake as well.
Two-way SMP buildworld -j 5 tests (second run, after stabilization)
2282.76 real 2515.17 user 704.22 sys before peter's IPI commit
2266.69 real 2467.50 user 633.77 sys after peter's commit
2232.80 real 2468.99 user 615.89 sys after this commit
Reviewed by: peter, jhb
Approved by: peter
choosethread() in MI C code instead of doing it in in assembly in all the
various cpu_switch() functions. This fixes problems on ia64 and sparc64.
Reviewed by: julian, peter, benno
Tested on: i386, alpha, sparc64
(put the function stabs in traditional order on a.out, or gdb doesn't see
function local variables), I failed to remove the related knobs here.
Effectively were overrode the ELF-wide definition in elfos.h w/o providing
new infrastructure. This is what caused GDB to fail to debug applications
compiled and linked with -stabs. This is because GCC was unconditionally
inserts .stabs instruction for functions after the function body. GDB was
getting confused because what it thinks is function beginning address is
actually function ending address.
Submitted by: Alexander Kabaev <ak03@gte.com>
itself; this causes undefined behaviour on UltraSPARCs. In particular,
the interrupt packet data words will not necessarily be delivered
correctly, which would result in a crash.
This bug also caused the cache-flushing work to be done twice on the
triggering CPU (when it did not cause crashes).
Reviewed by: jake
Thus lines of any length can be handled, unlike before.
Don't assume that each line read from the files ends with a newline.
As a side effect in inithosts(), don't use automatic buffer at all,
utilize malloc(3) when getting local host name instead.
PR: misc/21494
Reviewed by: maxim, mikeh
MFC after: 1 month
to make it call `install' in the bsd.subdir.mk-driven makefiles
too. (share/examples/Makefile,v 1.29 changed the bsd.prog.mk
to bsd.subdir.mk and many stuff was lost during "make release".
I then merged this change in rev. 1.28.2.2 to work around the
namespace pollution (FILES) in this makefile.)
There was an added complexity here. Both the `distribute' and
`install' targets are recursive (they propagate to SUBDIRs).
So `distribute' first calls `install' in the ${.CURDIR}, then
calls `distribute' in each SUBDIR, etc. The problem is that
`install' (being also recursive) causes the stuff from SUBDIR
to be installed twice, first time thru `install' in ${.CURDIR}
triggered by `distribute', second time by `distribute' run in
the SUBDIR. This problem is not new, but it became apparent
only after I moved the `distribute' target from bsd.obj.mk to
bsd.subdir.mk. My first attempt testing the fix failed due to
this, because the whole world was distributed twice, causing
all the imaginable mess (kerberos5 stuff was installed into both
"base" and "krb5" dists, there was /sbin/init.bak, etc.)
I say the problem is not new because bsd.prog.mk and bsd.lib.mk
makefiles with SUBDIR (even without this fix) had this problem
for years. Try e.g. running ``make distribute DISTDIR=/foo''
from usr.bin/bzip2 or from lib/libcom_err (without the fix) and
watch the output.
So the solution was to make `install' behave non-recursive when
executed by `distribute'. My first attempt in passing SUBDIR=
to the `install' in the `distribute' body failed because of the
way how src/Makefile and src/Makefile.inc1 communicate with each
other. SUBDIR='s assignment precedence on the "make install
SUBDIR=" command line is lowered after src/Makefile wrapper calls
"make ... -f ${.CURDIR}/Makefile.inc1 install" because SUBDIR=
is moved into environment, and Makefile.inc1's assignments now
take higher precedence. This may be fixed someday when we merge
Makefile with Makefile.inc1. For now, this is implemented as a
NO_SUBDIR knob.
Spotted by: Dmitry Pryanishnikov <dmitry@atlantis.dp.ua>
Prodded by: des
MFC after: 3 days
is only defined after <bsd.prog.mk> is included, and .if make(1)
conditionals are evaluted on the first pass.
Spotted by: Michael Bretterklieber <mbretter@jawa.at>
- It actually works this time, honest!
- Fine grained TLB shootdowns for SMP on i386. IPI's are very expensive,
so try and optimize things where possible.
- Introduce ranged shootdowns that can be done as a single IPI.
- PG_G support for i386
- Specific-cpu targeted shootdowns. For example, there is no sense in
globally purging the TLB cache for where we are stealing a page from
the local unshared process on the local cpu. Use pm_active to track
this.
- Add some instrumentation for the tlb shootdown code.
- Rip out SMP code from <machine/cpufunc.h>
- Try and fix some very bogus PG_G and PG_PS interactions that were bad
enough to cause vm86 bios calls to break. vm86 depended on our existing
bugs and this was the cause of the VESA panics last time.
- Fix the silly one-line error that caused the 'panic: bad pte' last time.
- Fix a couple of other silly one-line errors that should have caused more
pain than they did.
Some more work is needed:
- pmap_{zero,copy}_page[_idle]. These can be done without IPI's if we
have a hook in cpu_switch.
- The IPI handlers need some cleanup. I have a bogus %ds load that can
be avoided.
- APTD handling is rather bogus and appears to be a large source of
global TLB IPI shootdowns for no really good reason.
I see speedups of between 1.5% and ~4% on buildworlds in a while 1 loop.
I expect to see a bigger difference when there is significant pageout
activity or the system otherwise has memory shortages.
I have backed out a few optimizations that I had been using over the last
few days in order to be a little more conservative. I'll revisit these
again over the next few days as the dust settles.
New option: DISABLE_PG_G - In case I missed something.
