Clang emits SSE instructions on amd64 in the common path of
pthread_mutex_unlock. If the thread does not otherwise use SSE,
this usage incurs a context-switch of the FPU/SSE state, which
reduces the performance of multiple real-world applications by a
non-trivial amount (3-5% in one application).
Instead of this change, I experimented with eagerly switching the
FPU state at context-switch time. This did not help. Most of the
cost seems to be in the read/write of memory--as kib@ stated--and
not in the #NM handling. I tested on machines with and without
XSAVEOPT.
One counter-argument to this change is that most applications already
use SIMD, and the number of applications and amount of SIMD usage
are only increasing. This is absolutely true. I agree that--in
general and in principle--this change is in the wrong direction.
However, there are applications that do not use enough SSE to offset
the extra context-switch cost. SSE does not provide a clear benefit
in the current libthr code with the current compiler, but it does
provide a clear loss in some cases. Therefore, disabling SSE in
libthr is a non-loss for most, and a gain for some.
I refrained from disabling SSE in libc--as was suggested--because
I can't make the above argument for libc. It provides a wide variety
of code; each case should be analyzed separately.
https://lists.freebsd.org/pipermail/freebsd-current/2015-March/055193.html
Suggestions from: dim, jmg, rpaulo
Approved by: kib (mentor)
MFC after: 2 weeks
Sponsored by: Dell Inc.
CloudABI has two separate kernel modules: cloudabi and cloudabi64. The
first module contains all the pointer size independent code, whereas
cloudabi64 contains the actual 64-bits specific system calls and the ELF
loader.
Reviewed by: wblock
Obtained from: https://github.com/NuxiNL/freebsd
Differential Revision: https://reviews.freebsd.org/D3258
doubt most people will read to the end... Note the use of sys/cdefs.h
for pre-C11 compilers...
I didn't included a note about being compatibile w/ userland since a
C11 feature should be obviously usable in userland...
Suggested by: imp
We previously disabled CLANG_FULL on (little-endian) ARM because the
build failed. This is no longer the case and as of Clang 3.5 we cannot
build any part of the in-tree Clang with in-tree GCC, so it's no longer
necessary to disable CLANG_FULL.
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D2525
ELF Tool Chain elfcopy is nearly a drop-in replacement for GNU objcopy,
but does not currently support PE output which is needed for building
x86 UEFI bits.
Add a src.conf knob to allow installing it as objcopy and set it by
default for aarch64 only, where we don't have a native binutils.
Reviewed by: bapt
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D2887
to no longer claim they are experimental.
Reviewed by: rwatson@, wblock@
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D2985
The random_get() system call works similar to getentropy()/getrandom()
on OpenBSD/Linux. It fills a buffer with random data.
This change introduces a new function, read_random_uio(), that is used
to implement read() on the random devices. We can call into this
function from within the CloudABI compatibility layer.
Approved by: secteam
Reviewed by: jmg, markm, wblock
Obtained from: https://github.com/NuxiNL/freebsd
Differential Revision: https://reviews.freebsd.org/D3053
- Tweek man page.
- Remove all mention of RANDOM_FORTUNA. If the system owner wants YARROW or DUMMY, they ask for it, otherwise they get FORTUNA.
- Tidy up headers a bit.
- Tidy up declarations a bit.
- Make static in a couple of places where needed.
- Move Yarrow/Fortuna SYSINIT/SYSUNINIT to randomdev.c, moving us towards a single file where the algorithm context is used.
- Get rid of random_*_process_buffer() functions. They were only used in one place each, and are better subsumed into those places.
- Remove *_post_read() functions as they are stubs everywhere.
- Assert against buffer size illegalities.
- Clean up some silly code in the randomdev_read() routine.
- Make the harvesting more consistent.
- Make some requested argument name changes.
- Tidy up and clarify a few comments.
- Make some requested comment changes.
- Make some requested macro changes.
* NOTE: the thing calling itself a 'unit test' is not yet a proper
unit test, but it helps me ensure things work. It may be a proper
unit test at some time in the future, but for now please don't make
any assumptions or hold any expectations.
Differential Revision: https://reviews.freebsd.org/D2025
Approved by: so (/dev/random blanket)
This is based on work done by jeff@ and jhb@, as well as the numa.diff
patch that has been circulating when someone asks for first-touch NUMA
on -10 or -11.
* Introduce a simple set of VM policy and iterator types.
* tie the policy types into the vm_phys path for now, mirroring how
the initial first-touch allocation work was enabled.
* add syscalls to control changing thread and process defaults.
* add a global NUMA VM domain policy.
* implement a simple cascade policy order - if a thread policy exists, use it;
if a process policy exists, use it; use the default policy.
* processes inherit policies from their parent processes, threads inherit
policies from their parent threads.
* add a simple tool (numactl) to query and modify default thread/process
policities.
* add documentation for the new syscalls, for numa and for numactl.
* re-enable first touch NUMA again by default, as now policies can be
set in a variety of methods.
This is only relevant for very specific workloads.
This doesn't pretend to be a final NUMA solution.
The previous defaults in -HEAD (with MAXMEMDOM set) can be achieved by
'sysctl vm.default_policy=rr'.
This is only relevant if MAXMEMDOM is set to something other than 1.
Ie, if you're using GENERIC or a modified kernel with non-NUMA, then
this is a glorified no-op for you.
Thank you to Norse Corp for giving me access to rather large
(for FreeBSD!) NUMA machines in order to develop and verify this.
Thank you to Dell for providing me with dual socket sandybridge
and westmere v3 hardware to do NUMA development with.
Thank you to Scott Long at Netflix for providing me with access
to the two-socket, four-domain haswell v3 hardware.
Thank you to Peter Holm for running the stress testing suite
against the NUMA branch during various stages of development!
Tested:
* MIPS (regression testing; non-NUMA)
* i386 (regression testing; non-NUMA GENERIC)
* amd64 (regression testing; non-NUMA GENERIC)
* westmere, 2 socket (thankyou norse!)
* sandy bridge, 2 socket (thankyou dell!)
* ivy bridge, 2 socket (thankyou norse!)
* westmere-EX, 4 socket / 1TB RAM (thankyou norse!)
* haswell, 2 socket (thankyou norse!)
* haswell v3, 2 socket (thankyou dell)
* haswell v3, 2x18 core (thankyou scott long / netflix!)
* Peter Holm ran a stress test suite on this work and found one
issue, but has not been able to verify it (it doesn't look NUMA
related, and he only saw it once over many testing runs.)
* I've tested bhyve instances running in fixed NUMA domains and cpusets;
all seems to work correctly.
Verified:
* intel-pcm - pcm-numa.x and pcm-memory.x, whilst selecting different
NUMA policies for processes under test.
Review:
This was reviewed through phabricator (https://reviews.freebsd.org/D2559)
as well as privately and via emails to freebsd-arch@. The git history
with specific attributes is available at https://github.com/erikarn/freebsd/
in the NUMA branch (https://github.com/erikarn/freebsd/compare/local/adrian_numa_policy).
This has been reviewed by a number of people (stas, rpaulo, kib, ngie,
wblock) but not achieved a clear consensus. My hope is that with further
exposure and testing more functionality can be implemented and evaluated.
Notes:
* The VM doesn't handle unbalanced domains very well, and if you have an overly
unbalanced memory setup whilst under high memory pressure, VM page allocation
may fail leading to a kernel panic. This was a problem in the past, but it's
much more easily triggered now with these tools.
* This work only controls the path through vm_phys; it doesn't yet strongly/predictably
affect contigmalloc, KVA placement, UMA, etc. So, driver placement of memory
isn't really guaranteed in any way. That's next on my plate.
Sponsored by: Norse Corp, Inc.; Dell
OCF w/o documentation...
Document the new (8+ year old) device_t way of handling things, that
_unregister_all will leave no threads in newsession, the _SYNC flag,
the requirement that a flag be specified...
Other minor changes like breaking up a wall of text into paragraphs...
line statements inside of braces is recognized as an acceptable
style.
http://reviews.freebsd.org/V3
As always, this isn't license for wholesale change, etc.