15 May, 2019
1 commit
-
Patch series "mm: Randomize free memory", v10.
This patch (of 3):
Randomization of the page allocator improves the average utilization of
a direct-mapped memory-side-cache. Memory side caching is a platform
capability that Linux has been previously exposed to in HPC
(high-performance computing) environments on specialty platforms. In
that instance it was a smaller pool of high-bandwidth-memory relative to
higher-capacity / lower-bandwidth DRAM. Now, this capability is going
to be found on general purpose server platforms where DRAM is a cache in
front of higher latency persistent memory [1].Robert offered an explanation of the state of the art of Linux
interactions with memory-side-caches [2], and I copy it here:It's been a problem in the HPC space:
http://www.nersc.gov/research-and-development/knl-cache-mode-performance-coe/A kernel module called zonesort is available to try to help:
https://software.intel.com/en-us/articles/xeon-phi-softwareand this abandoned patch series proposed that for the kernel:
https://lkml.kernel.org/r/20170823100205.17311-1-lukasz.daniluk@intel.comDan's patch series doesn't attempt to ensure buffers won't conflict, but
also reduces the chance that the buffers will. This will make performance
more consistent, albeit slower than "optimal" (which is near impossible
to attain in a general-purpose kernel). That's better than forcing
users to deploy remedies like:
"To eliminate this gradual degradation, we have added a Stream
measurement to the Node Health Check that follows each job;
nodes are rebooted whenever their measured memory bandwidth
falls below 300 GB/s."A replacement for zonesort was merged upstream in commit cc9aec03e58f
("x86/numa_emulation: Introduce uniform split capability"). With this
numa_emulation capability, memory can be split into cache sized
("near-memory" sized) numa nodes. A bind operation to such a node, and
disabling workloads on other nodes, enables full cache performance.
However, once the workload exceeds the cache size then cache conflicts
are unavoidable. While HPC environments might be able to tolerate
time-scheduling of cache sized workloads, for general purpose server
platforms, the oversubscribed cache case will be the common case.The worst case scenario is that a server system owner benchmarks a
workload at boot with an un-contended cache only to see that performance
degrade over time, even below the average cache performance due to
excessive conflicts. Randomization clips the peaks and fills in the
valleys of cache utilization to yield steady average performance.Here are some performance impact details of the patches:
1/ An Intel internal synthetic memory bandwidth measurement tool, saw a
3X speedup in a contrived case that tries to force cache conflicts.
The contrived cased used the numa_emulation capability to force an
instance of the benchmark to be run in two of the near-memory sized
numa nodes. If both instances were placed on the same emulated they
would fit and cause zero conflicts. While on separate emulated nodes
without randomization they underutilized the cache and conflicted
unnecessarily due to the in-order allocation per node.2/ A well known Java server application benchmark was run with a heap
size that exceeded cache size by 3X. The cache conflict rate was 8%
for the first run and degraded to 21% after page allocator aging. With
randomization enabled the rate levelled out at 11%.3/ A MongoDB workload did not observe measurable difference in
cache-conflict rates, but the overall throughput dropped by 7% with
randomization in one case.4/ Mel Gorman ran his suite of performance workloads with randomization
enabled on platforms without a memory-side-cache and saw a mix of some
improvements and some losses [3].While there is potentially significant improvement for applications that
depend on low latency access across a wide working-set, the performance
may be negligible to negative for other workloads. For this reason the
shuffle capability defaults to off unless a direct-mapped
memory-side-cache is detected. Even then, the page_alloc.shuffle=0
parameter can be specified to disable the randomization on those systems.Outside of memory-side-cache utilization concerns there is potentially
security benefit from randomization. Some data exfiltration and
return-oriented-programming attacks rely on the ability to infer the
location of sensitive data objects. The kernel page allocator, especially
early in system boot, has predictable first-in-first out behavior for
physical pages. Pages are freed in physical address order when first
onlined.Quoting Kees:
"While we already have a base-address randomization
(CONFIG_RANDOMIZE_MEMORY), attacks against the same hardware and
memory layouts would certainly be using the predictability of
allocation ordering (i.e. for attacks where the base address isn't
important: only the relative positions between allocated memory).
This is common in lots of heap-style attacks. They try to gain
control over ordering by spraying allocations, etc.I'd really like to see this because it gives us something similar
to CONFIG_SLAB_FREELIST_RANDOM but for the page allocator."While SLAB_FREELIST_RANDOM reduces the predictability of some local slab
caches it leaves vast bulk of memory to be predictably in order allocated.
However, it should be noted, the concrete security benefits are hard to
quantify, and no known CVE is mitigated by this randomization.Introduce shuffle_free_memory(), and its helper shuffle_zone(), to perform
a Fisher-Yates shuffle of the page allocator 'free_area' lists when they
are initially populated with free memory at boot and at hotplug time. Do
this based on either the presence of a page_alloc.shuffle=Y command line
parameter, or autodetection of a memory-side-cache (to be added in a
follow-on patch).The shuffling is done in terms of CONFIG_SHUFFLE_PAGE_ORDER sized free
pages where the default CONFIG_SHUFFLE_PAGE_ORDER is MAX_ORDER-1 i.e. 10,
4MB this trades off randomization granularity for time spent shuffling.
MAX_ORDER-1 was chosen to be minimally invasive to the page allocator
while still showing memory-side cache behavior improvements, and the
expectation that the security implications of finer granularity
randomization is mitigated by CONFIG_SLAB_FREELIST_RANDOM. The
performance impact of the shuffling appears to be in the noise compared to
other memory initialization work.This initial randomization can be undone over time so a follow-on patch is
introduced to inject entropy on page free decisions. It is reasonable to
ask if the page free entropy is sufficient, but it is not enough due to
the in-order initial freeing of pages. At the start of that process
putting page1 in front or behind page0 still keeps them close together,
page2 is still near page1 and has a high chance of being adjacent. As
more pages are added ordering diversity improves, but there is still high
page locality for the low address pages and this leads to no significant
impact to the cache conflict rate.[1]: https://itpeernetwork.intel.com/intel-optane-dc-persistent-memory-operating-modes/
[2]: https://lkml.kernel.org/r/AT5PR8401MB1169D656C8B5E121752FC0F8AB120@AT5PR8401MB1169.NAMPRD84.PROD.OUTLOOK.COM
[3]: https://lkml.org/lkml/2018/10/12/309[dan.j.williams@intel.com: fix shuffle enable]
Link: http://lkml.kernel.org/r/154943713038.3858443.4125180191382062871.stgit@dwillia2-desk3.amr.corp.intel.com
[cai@lca.pw: fix SHUFFLE_PAGE_ALLOCATOR help texts]
Link: http://lkml.kernel.org/r/20190425201300.75650-1-cai@lca.pw
Link: http://lkml.kernel.org/r/154899811738.3165233.12325692939590944259.stgit@dwillia2-desk3.amr.corp.intel.com
Signed-off-by: Dan Williams
Signed-off-by: Qian Cai
Reviewed-by: Kees Cook
Acked-by: Michal Hocko
Cc: Dave Hansen
Cc: Keith Busch
Cc: Robert Elliott
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
08 May, 2019
1 commit
-
Pull driver core/kobject updates from Greg KH:
"Here is the "big" set of driver core patches for 5.2-rc1There are a number of ACPI patches in here as well, as Rafael said
they should go through this tree due to the driver core changes they
required. They have all been acked by the ACPI developers.There are also a number of small subsystem-specific changes in here,
due to some changes to the kobject core code. Those too have all been
acked by the various subsystem maintainers.As for content, it's pretty boring outside of the ACPI changes:
- spdx cleanups
- kobject documentation updates
- default attribute groups for kobjects
- other minor kobject/driver core fixesAll have been in linux-next for a while with no reported issues"
* tag 'driver-core-5.2-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core: (47 commits)
kobject: clean up the kobject add documentation a bit more
kobject: Fix kernel-doc comment first line
kobject: Remove docstring reference to kset
firmware_loader: Fix a typo ("syfs" -> "sysfs")
kobject: fix dereference before null check on kobj
Revert "driver core: platform: Fix the usage of platform device name(pdev->name)"
init/config: Do not select BUILD_BIN2C for IKCONFIG
Provide in-kernel headers to make extending kernel easier
kobject: Improve doc clarity kobject_init_and_add()
kobject: Improve docs for kobject_add/del
driver core: platform: Fix the usage of platform device name(pdev->name)
livepatch: Replace klp_ktype_patch's default_attrs with groups
cpufreq: schedutil: Replace default_attrs field with groups
padata: Replace padata_attr_type default_attrs field with groups
irqdesc: Replace irq_kobj_type's default_attrs field with groups
net-sysfs: Replace ktype default_attrs field with groups
block: Replace all ktype default_attrs with groups
samples/kobject: Replace foo_ktype's default_attrs field with groups
kobject: Add support for default attribute groups to kobj_type
driver core: Postpone DMA tear-down until after devres release for probe failure
...
29 Apr, 2019
2 commits
-
Since commit 13610aa908dc ("kernel/configs: use .incbin directive to
embed config_data.gz"), IKCONFIG no longer uses BUILD_BIN2C so prevent
it from being selected in Kconfig.Reviewed-by: Masahiro Yamada
Signed-off-by: Joel Fernandes (Google)
Signed-off-by: Greg Kroah-Hartman -
Introduce in-kernel headers which are made available as an archive
through proc (/proc/kheaders.tar.xz file). This archive makes it
possible to run eBPF and other tracing programs that need to extend the
kernel for tracing purposes without any dependency on the file system
having headers.A github PR is sent for the corresponding BCC patch at:
https://github.com/iovisor/bcc/pull/2312On Android and embedded systems, it is common to switch kernels but not
have kernel headers available on the file system. Further once a
different kernel is booted, any headers stored on the file system will
no longer be useful. This is an issue even well known to distros.
By storing the headers as a compressed archive within the kernel, we can
avoid these issues that have been a hindrance for a long time.The best way to use this feature is by building it in. Several users
have a need for this, when they switch debug kernels, they do not want to
update the filesystem or worry about it where to store the headers on
it. However, the feature is also buildable as a module in case the user
desires it not being part of the kernel image. This makes it possible to
load and unload the headers from memory on demand. A tracing program can
load the module, do its operations, and then unload the module to save
kernel memory. The total memory needed is 3.3MB.By having the archive available at a fixed location independent of
filesystem dependencies and conventions, all debugging tools can
directly refer to the fixed location for the archive, without concerning
with where the headers on a typical filesystem which significantly
simplifies tooling that needs kernel headers.The code to read the headers is based on /proc/config.gz code and uses
the same technique to embed the headers.Other approaches were discussed such as having an in-memory mountable
filesystem, but that has drawbacks such as requiring an in-kernel xz
decompressor which we don't have today, and requiring usage of 42 MB of
kernel memory to host the decompressed headers at anytime. Also this
approach is simpler than such approaches.Reviewed-by: Masahiro Yamada
Signed-off-by: Joel Fernandes (Google)
Signed-off-by: Greg Kroah-Hartman
19 Apr, 2019
1 commit
-
Make the anon_inodes facility unconditional so that it can be used by core
VFS code and pidfd code.Signed-off-by: David Howells
Signed-off-by: Al Viro
[christian@brauner.io: adapt commit message to mention pidfds]
Signed-off-by: Christian Brauner
11 Mar, 2019
2 commits
-
Pull Kbuild updates from Masahiro Yamada:
- do not generate unneeded top-level built-in.a
- let git ignore O= directory entirely
- optimize scripts/kallsyms slightly
- exclude DWARF info from *.s regardless of config options
- fix GCC toolchain search path for Clang to prepare ld.lld support
- do not generate modules.order when CONFIG_MODULES is disabled
- simplify single target rules and remove VPATH for external module
build- allow to add optional flags to dpkg-buildpackage when building
deb-pkg- move some compiler option tests from Makefile to Kconfig
- various Makefile cleanups
* tag 'kbuild-v5.1' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (40 commits)
kbuild: remove scripts/basic/% build target
kbuild: use -Werror=implicit-... instead of -Werror-implicit-...
kbuild: clean up scripts/gcc-version.sh
kbuild: remove cc-version macro
kbuild: update comment block of scripts/clang-version.sh
kbuild: remove commented-out INITRD_COMPRESS
kbuild: move -gsplit-dwarf, -gdwarf-4 option tests to Kconfig
kbuild: [bin]deb-pkg: add DPKG_FLAGS variable
kbuild: move ".config not found!" message from Kconfig to Makefile
kbuild: invoke syncconfig if include/config/auto.conf.cmd is missing
kbuild: simplify single target rules
kbuild: remove empty rules for makefiles
kbuild: make -r/-R effective in top Makefile for old Make versions
kbuild: move tools_silent to a more relevant place
kbuild: compute false-positive -Wmaybe-uninitialized cases in Kconfig
kbuild: refactor cc-cross-prefix implementation
kbuild: hardcode genksyms path and remove GENKSYMS variable
scripts/gdb: refactor rules for symlink creation
kbuild: create symlink to vmlinux-gdb.py in scripts_gdb target
scripts/gdb: do not descend into scripts/gdb from scripts
... -
Pull timer fix from Thomas Gleixner:
"A single fix to prevent a unmet dependencies warning in Kconfig"* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
time: Make VIRT_CPU_ACCOUNTING_GEN depend on GENERIC_CLOCKEVENTS
07 Mar, 2019
1 commit
-
Moving the CONTEXT_TRACKING Kconfig option into kernel/time/Kconfig added
an implicit dependency on the surrounding GENERIC_CLOCKEVENTS option, but
this is not always enabled when it is possible to select
VIRT_CPU_ACCOUNTING_GEN:WARNING: unmet direct dependencies detected for CONTEXT_TRACKING
Depends on [n]: GENERIC_CLOCKEVENTS [=n]
Selected by [y]:
- VIRT_CPU_ACCOUNTING_GEN [=y] && && HAVE_CONTEXT_TRACKING [=y] && HAVE_VIRT_CPU_ACCOUNTING_GEN [=y]Platforms without GENERIC_CLOCKEVENTS are rare enough so that corner case
can be just ignored. Make it a dependency for VIRT_CPU_ACCOUNTING_GEN to
simplify the configuration.Fixes: a4cffdad7314 ("time: Move CONTEXT_TRACKING to kernel/time/Kconfig")
Signed-off-by: Arnd Bergmann
Signed-off-by: Thomas Gleixner
Cc: "Paul E . McKenney"
Cc: Frederic Weisbecker
Link: https://lkml.kernel.org/r/20190304200202.1163250-1-arnd@arndb.de
04 Mar, 2019
1 commit
-
Now that the Kconfig is the only user of this script, we can drop
unneeded code.Remove the -p option, and stop prepending the output with zero,
so that Kconfig can directly use the output from this script.Signed-off-by: Masahiro Yamada
28 Feb, 2019
1 commit
-
The submission queue (SQ) and completion queue (CQ) rings are shared
between the application and the kernel. This eliminates the need to
copy data back and forth to submit and complete IO.IO submissions use the io_uring_sqe data structure, and completions
are generated in the form of io_uring_cqe data structures. The SQ
ring is an index into the io_uring_sqe array, which makes it possible
to submit a batch of IOs without them being contiguous in the ring.
The CQ ring is always contiguous, as completion events are inherently
unordered, and hence any io_uring_cqe entry can point back to an
arbitrary submission.Two new system calls are added for this:
io_uring_setup(entries, params)
Sets up an io_uring instance for doing async IO. On success,
returns a file descriptor that the application can mmap to
gain access to the SQ ring, CQ ring, and io_uring_sqes.io_uring_enter(fd, to_submit, min_complete, flags, sigset, sigsetsize)
Initiates IO against the rings mapped to this fd, or waits for
them to complete, or both. The behavior is controlled by the
parameters passed in. If 'to_submit' is non-zero, then we'll
try and submit new IO. If IORING_ENTER_GETEVENTS is set, the
kernel will wait for 'min_complete' events, if they aren't
already available. It's valid to set IORING_ENTER_GETEVENTS
and 'min_complete' == 0 at the same time, this allows the
kernel to return already completed events without waiting
for them. This is useful only for polling, as for IRQ
driven IO, the application can just check the CQ ring
without entering the kernel.With this setup, it's possible to do async IO with a single system
call. Future developments will enable polled IO with this interface,
and polled submission as well. The latter will enable an application
to do IO without doing ANY system calls at all.For IRQ driven IO, an application only needs to enter the kernel for
completions if it wants to wait for them to occur.Each io_uring is backed by a workqueue, to support buffered async IO
as well. We will only punt to an async context if the command would
need to wait for IO on the device side. Any data that can be accessed
directly in the page cache is done inline. This avoids the slowness
issue of usual threadpools, since cached data is accessed as quickly
as a sync interface.Sample application: http://git.kernel.dk/cgit/fio/plain/t/io_uring.c
Reviewed-by: Hannes Reinecke
Signed-off-by: Jens Axboe
27 Feb, 2019
1 commit
-
Since -Wmaybe-uninitialized was introduced by GCC 4.7, we have patched
various false positives:- commit e74fc973b6e5 ("Turn off -Wmaybe-uninitialized when building
with -Os") turned off this option for -Os.- commit 815eb71e7149 ("Kbuild: disable 'maybe-uninitialized' warning
for CONFIG_PROFILE_ALL_BRANCHES") turned off this option for
CONFIG_PROFILE_ALL_BRANCHES- commit a76bcf557ef4 ("Kbuild: enable -Wmaybe-uninitialized warning
for "make W=1"") turned off this option for GCC < 4.9
Arnd provided more explanation in https://lkml.org/lkml/2017/3/14/903I think this looks better by shifting the logic from Makefile to Kconfig.
Link: https://github.com/ClangBuiltLinux/linux/issues/350
Signed-off-by: Masahiro Yamada
Reviewed-by: Nathan Chancellor
Tested-by: Nick Desaulniers
02 Feb, 2019
2 commits
-
The current help text caused some confusion in online forums about
whether or not to default-enable or default-disable psi in vendor
kernels. This is because it doesn't communicate the reason for why we
made this setting configurable in the first place: that the overhead is
non-zero in an artificial scheduler stress test.Since this isn't representative of real workloads, and the effect was
not measurable in scheduler-heavy real world applications such as the
webservers and memcache installations at Facebook, it's fair to point
out that this is a pretty cautious option to select.Link: http://lkml.kernel.org/r/20190129233617.16767-1-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner
Reviewed-by: Andrew Morton
Cc: Mel Gorman
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Link: http://lkml.kernel.org/r/20190129150813.15785-1-j.neuschaefer@gmx.net
Signed-off-by: Jonathan Neuschäfer
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
14 Jan, 2019
1 commit
-
When building using GCC 4.7 or older, -ffunction-sections & the -pg flag
used by ftrace are incompatible. This causes warnings or build failures
(where -Werror applies) such as the following:arch/mips/generic/init.c:
error: -ffunction-sections disabled; it makes profiling impossibleThis used to be taken into account by the ordering of calls to cc-option
from within the top-level Makefile, which was introduced by commit
90ad4052e85c ("kbuild: avoid conflict between -ffunction-sections and
-pg on gcc-4.7"). Unfortunately this was broken when the
CONFIG_LD_DEAD_CODE_DATA_ELIMINATION cc-option check was moved to
Kconfig in commit e85d1d65cd8a ("kbuild: test dead code/data elimination
support in Kconfig"), because the flags used by this check no longer
include -pg.Fix this by not allowing CONFIG_LD_DEAD_CODE_DATA_ELIMINATION to be
enabled at the same time as ftrace/CONFIG_FUNCTION_TRACER when building
using GCC 4.7 or older.Signed-off-by: Paul Burton
Fixes: e85d1d65cd8a ("kbuild: test dead code/data elimination support in Kconfig")
Reported-by: Geert Uytterhoeven
Cc: Nicholas Piggin
Cc: stable@vger.kernel.org # v4.19+
Signed-off-by: Masahiro Yamada
06 Jan, 2019
1 commit
-
Currently, CONFIG_JUMP_LABEL just means "I _want_ to use jump label".
The jump label is controlled by HAVE_JUMP_LABEL, which is defined
like this:#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
# define HAVE_JUMP_LABEL
#endifWe can improve this by testing 'asm goto' support in Kconfig, then
make JUMP_LABEL depend on CC_HAS_ASM_GOTO.Ugly #ifdef HAVE_JUMP_LABEL will go away, and CONFIG_JUMP_LABEL will
match to the real kernel capability.Signed-off-by: Masahiro Yamada
Acked-by: Michael Ellerman (powerpc)
Tested-by: Sedat Dilek
28 Dec, 2018
1 commit
-
Pull audit updates from Paul Moore:
"In the finest of holiday of traditions, I have a number of gifts to
share today. While most of them are re-gifts from others, unlike the
typical re-gift, these are things you will want in and around your
tree; I promise.This pull request is perhaps a bit larger than our typical PR, but
most of it comes from Jan's rework of audit's fanotify code; a very
welcome improvement. We ran this through our normal regression tests,
as well as some newly created stress tests and everything looks good.Richard added a few patches, mostly cleaning up a few things and and
shortening some of the audit records that we send to userspace; a
change the userspace folks are quite happy about.Finally YueHaibing and I kick in a few patches to simplify things a
bit and make the code less prone to errors.Lastly, I want to say thanks one more time to everyone who has
contributed patches, testing, and code reviews for the audit subsystem
over the past year. The project is what it is due to your help and
contributions - thank you"* tag 'audit-pr-20181224' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/audit: (22 commits)
audit: remove duplicated include from audit.c
audit: shorten PATH cap values when zero
audit: use current whenever possible
audit: minimize our use of audit_log_format()
audit: remove WATCH and TREE config options
audit: use session_info helper
audit: localize audit_log_session_info prototype
audit: Use 'mark' name for fsnotify_mark variables
audit: Replace chunk attached to mark instead of replacing mark
audit: Simplify locking around untag_chunk()
audit: Drop all unused chunk nodes during deletion
audit: Guarantee forward progress of chunk untagging
audit: Allocate fsnotify mark independently of chunk
audit: Provide helper for dropping mark's chunk reference
audit: Remove pointless check in insert_hash()
audit: Factor out chunk replacement code
audit: Make hash table insertion safe against concurrent lookups
audit: Embed key into chunk
audit: Fix possible tagging failures
audit: Fix possible spurious -ENOSPC error
...
15 Dec, 2018
1 commit
-
The kernel commandline parameter named in CONFIG_PSI_DEFAULT_DISABLED
help text contradicts the documentation in kernel-parameters.txt, and
the code. Fix that.Link: http://lkml.kernel.org/r/20181203213416.GA12627@cmpxchg.org
Fixes: e0c274472d ("psi: make disabling/enabling easier for vendor kernels")
Signed-off-by: Baruch Siach
Acked-by: Johannes Weiner
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
01 Dec, 2018
1 commit
-
Mel Gorman reports a hackbench regression with psi that would prohibit
shipping the suse kernel with it default-enabled, but he'd still like
users to be able to opt in at little to no cost to others.With the current combination of CONFIG_PSI and the psi_disabled bool set
from the commandline, this is a challenge. Do the following things to
make it easier:1. Add a config option CONFIG_PSI_DEFAULT_DISABLED that allows distros
to enable CONFIG_PSI in their kernel but leave the feature disabled
unless a user requests it at boot-time.To avoid double negatives, rename psi_disabled= to psi=.
2. Make psi_disabled a static branch to eliminate any branch costs
when the feature is disabled.In terms of numbers before and after this patch, Mel says:
: The following is a comparision using CONFIG_PSI=n as a baseline against
: your patch and a vanilla kernel
:
: 4.20.0-rc4 4.20.0-rc4 4.20.0-rc4
: kconfigdisable-v1r1 vanilla psidisable-v1r1
: Amean 1 1.3100 ( 0.00%) 1.3923 ( -6.28%) 1.3427 ( -2.49%)
: Amean 3 3.8860 ( 0.00%) 4.1230 * -6.10%* 3.8860 ( -0.00%)
: Amean 5 6.8847 ( 0.00%) 8.0390 * -16.77%* 6.7727 ( 1.63%)
: Amean 7 9.9310 ( 0.00%) 10.8367 * -9.12%* 9.9910 ( -0.60%)
: Amean 12 16.6577 ( 0.00%) 18.2363 * -9.48%* 17.1083 ( -2.71%)
: Amean 18 26.5133 ( 0.00%) 27.8833 * -5.17%* 25.7663 ( 2.82%)
: Amean 24 34.3003 ( 0.00%) 34.6830 ( -1.12%) 32.0450 ( 6.58%)
: Amean 30 40.0063 ( 0.00%) 40.5800 ( -1.43%) 41.5087 ( -3.76%)
: Amean 32 40.1407 ( 0.00%) 41.2273 ( -2.71%) 39.9417 ( 0.50%)
:
: It's showing that the vanilla kernel takes a hit (as the bisection
: indicated it would) and that disabling PSI by default is reasonably
: close in terms of performance for this particular workload on this
: particular machine so;Link: http://lkml.kernel.org/r/20181127165329.GA29728@cmpxchg.org
Signed-off-by: Johannes Weiner
Tested-by: Mel Gorman
Reported-by: Mel Gorman
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
20 Nov, 2018
1 commit
-
Remove the CONFIG_AUDIT_WATCH and CONFIG_AUDIT_TREE config options since
they are both dependent on CONFIG_AUDITSYSCALL and force
CONFIG_FSNOTIFY.Signed-off-by: Richard Guy Briggs
Signed-off-by: Paul Moore
27 Oct, 2018
2 commits
-
On a system that executes multiple cgrouped jobs and independent
workloads, we don't just care about the health of the overall system, but
also that of individual jobs, so that we can ensure individual job health,
fairness between jobs, or prioritize some jobs over others.This patch implements pressure stall tracking for cgroups. In kernels
with CONFIG_PSI=y, cgroup2 groups will have cpu.pressure, memory.pressure,
and io.pressure files that track aggregate pressure stall times for only
the tasks inside the cgroup.Link: http://lkml.kernel.org/r/20180828172258.3185-10-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner
Acked-by: Tejun Heo
Acked-by: Peter Zijlstra (Intel)
Tested-by: Daniel Drake
Tested-by: Suren Baghdasaryan
Cc: Christopher Lameter
Cc: Ingo Molnar
Cc: Johannes Weiner
Cc: Mike Galbraith
Cc: Peter Enderborg
Cc: Randy Dunlap
Cc: Shakeel Butt
Cc: Vinayak Menon
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
When systems are overcommitted and resources become contended, it's hard
to tell exactly the impact this has on workload productivity, or how close
the system is to lockups and OOM kills. In particular, when machines work
multiple jobs concurrently, the impact of overcommit in terms of latency
and throughput on the individual job can be enormous.In order to maximize hardware utilization without sacrificing individual
job health or risk complete machine lockups, this patch implements a way
to quantify resource pressure in the system.A kernel built with CONFIG_PSI=y creates files in /proc/pressure/ that
expose the percentage of time the system is stalled on CPU, memory, or IO,
respectively. Stall states are aggregate versions of the per-task delay
accounting delays:cpu: some tasks are runnable but not executing on a CPU
memory: tasks are reclaiming, or waiting for swapin or thrashing cache
io: tasks are waiting for io completionsThese percentages of walltime can be thought of as pressure percentages,
and they give a general sense of system health and productivity loss
incurred by resource overcommit. They can also indicate when the system
is approaching lockup scenarios and OOMs.To do this, psi keeps track of the task states associated with each CPU
and samples the time they spend in stall states. Every 2 seconds, the
samples are averaged across CPUs - weighted by the CPUs' non-idle time to
eliminate artifacts from unused CPUs - and translated into percentages of
walltime. A running average of those percentages is maintained over 10s,
1m, and 5m periods (similar to the loadaverage).[hannes@cmpxchg.org: doc fixlet, per Randy]
Link: http://lkml.kernel.org/r/20180828205625.GA14030@cmpxchg.org
[hannes@cmpxchg.org: code optimization]
Link: http://lkml.kernel.org/r/20180907175015.GA8479@cmpxchg.org
[hannes@cmpxchg.org: rename psi_clock() to psi_update_work(), per Peter]
Link: http://lkml.kernel.org/r/20180907145404.GB11088@cmpxchg.org
[hannes@cmpxchg.org: fix build]
Link: http://lkml.kernel.org/r/20180913014222.GA2370@cmpxchg.org
Link: http://lkml.kernel.org/r/20180828172258.3185-9-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner
Acked-by: Peter Zijlstra (Intel)
Tested-by: Daniel Drake
Tested-by: Suren Baghdasaryan
Cc: Christopher Lameter
Cc: Ingo Molnar
Cc: Johannes Weiner
Cc: Mike Galbraith
Cc: Peter Enderborg
Cc: Randy Dunlap
Cc: Shakeel Butt
Cc: Tejun Heo
Cc: Vinayak Menon
Cc: Randy Dunlap
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
02 Oct, 2018
1 commit
-
Create a config for enabling irq load tracking in the scheduler.
irq load tracking is useful only when irq or paravirtual time is
accounted but it's only possible with SMP for now.Also use __maybe_unused to remove the compilation warning in
update_rq_clock_task() that has been introduced by:2e62c4743adc ("sched/fair: Remove #ifdefs from scale_rt_capacity()")
Suggested-by: Ingo Molnar
Reported-by: Dou Liyang
Reported-by: Miguel Ojeda
Signed-off-by: Vincent Guittot
Signed-off-by: Peter Zijlstra (Intel)
Cc: Linus Torvalds
Cc: Peter Zijlstra
Cc: Thomas Gleixner
Cc: bp@alien8.de
Cc: dou_liyang@163.com
Fixes: 2e62c4743adc ("sched/fair: Remove #ifdefs from scale_rt_capacity()")
Link: http://lkml.kernel.org/r/1537867062-27285-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar
26 Aug, 2018
1 commit
-
Pull more Kbuild updates from Masahiro Yamada:
- add build_{menu,n,g,x}config targets for compile-testing Kconfig
- fix and improve recursive dependency detection in Kconfig
- fix parallel building of menuconfig/nconfig
- fix syntax error in clang-version.sh
- suppress distracting log from syncconfig
- remove obsolete "rpm" target
- remove VMLINUX_SYMBOL(_STR) macro entirely
- fix microblaze build with CONFIG_DYNAMIC_FTRACE
- move compiler test for dead code/data elimination to Kconfig
- rename well-known LDFLAGS variable to KBUILD_LDFLAGS
- misc fixes and cleanups
* tag 'kbuild-v4.19-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: rename LDFLAGS to KBUILD_LDFLAGS
kbuild: pass LDFLAGS to recordmcount.pl
kbuild: test dead code/data elimination support in Kconfig
initramfs: move gen_initramfs_list.sh from scripts/ to usr/
vmlinux.lds.h: remove stale include
export.h: remove VMLINUX_SYMBOL() and VMLINUX_SYMBOL_STR()
Coccinelle: remove pci_alloc_consistent semantic to detect in zalloc-simple.cocci
kbuild: make sorting initramfs contents independent of locale
kbuild: remove "rpm" target, which is alias of "rpm-pkg"
kbuild: Fix LOADLIBES rename in Documentation/kbuild/makefiles.txt
kconfig: suppress "configuration written to .config" for syncconfig
kconfig: fix "Can't open ..." in parallel build
kbuild: Add a space after `!` to prevent parsing as file pattern
scripts: modpost: check memory allocation results
kconfig: improve the recursive dependency report
kconfig: report recursive dependency involving 'imply'
kconfig: error out when seeing recursive dependency
kconfig: add build-only configurator targets
scripts/dtc: consolidate include path options in Makefile
24 Aug, 2018
1 commit
-
This config option should be enabled only when both the compiler and
the linker support necessary flags. Add proper dependencies to Kconfig.Signed-off-by: Masahiro Yamada
23 Aug, 2018
2 commits
-
The CHECKPOINT_RESTORE configuration option was introduced in 2012 and
combined with EXPERT. CHECKPOINT_RESTORE is already enabled in many
distribution kernels and also part of the defconfigs of various
architectures.To make it easier for distributions to enable CHECKPOINT_RESTORE this
removes EXPERT and moves the configuration option out of the EXPERT block.Link: http://lkml.kernel.org/r/20180712130733.11510-1-adrian@lisas.de
Signed-off-by: Adrian Reber
Acked-by: Oleg Nesterov
Reviewed-by: Hendrik Brueckner
Acked-by: Pavel Emelyanov
Cc: Eric W. Biederman
Cc: Andrei Vagin
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds -
Correct typos of "it's" to "its.
Link: http://lkml.kernel.org/r/0ac627b6-5527-55f4-0489-1631aa34fc11@infradead.org
Signed-off-by: Randy Dunlap
Cc: Ingo Molnar
Cc: Thomas Gleixner
Cc: Masahiro Yamada
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
18 Aug, 2018
1 commit
-
Introduce new config option, which is used to replace repeating
CONFIG_MEMCG && !CONFIG_SLOB pattern. Next patches add a little more
memcg+kmem related code, so let's keep the defines more clearly.Link: http://lkml.kernel.org/r/153063053670.1818.15013136946600481138.stgit@localhost.localdomain
Signed-off-by: Kirill Tkhai
Acked-by: Vladimir Davydov
Tested-by: Shakeel Butt
Cc: Al Viro
Cc: Andrey Ryabinin
Cc: Chris Wilson
Cc: Greg Kroah-Hartman
Cc: Guenter Roeck
Cc: "Huang, Ying"
Cc: Johannes Weiner
Cc: Josef Bacik
Cc: Li RongQing
Cc: Matthew Wilcox
Cc: Matthias Kaehlcke
Cc: Mel Gorman
Cc: Michal Hocko
Cc: Minchan Kim
Cc: Philippe Ombredanne
Cc: Roman Gushchin
Cc: Sahitya Tummala
Cc: Stephen Rothwell
Cc: Tetsuo Handa
Cc: Thomas Gleixner
Cc: Waiman Long
Signed-off-by: Andrew Morton
Signed-off-by: Linus Torvalds
16 Aug, 2018
3 commits
-
Pull Kconfig consolidation from Masahiro Yamada:
"Consolidation of Kconfig files by Christoph Hellwig.Move the source statements of arch-independent Kconfig files instead
of duplicating the includes in every arch/$(SRCARCH)/Kconfig"* tag 'kconfig-v4.19-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: add a Memory Management options" menu
kconfig: move the "Executable file formats" menu to fs/Kconfig.binfmt
kconfig: use a menu in arch/Kconfig to reduce clutter
kconfig: include kernel/Kconfig.preempt from init/Kconfig
Kconfig: consolidate the "Kernel hacking" menu
kconfig: include common Kconfig files from top-level Kconfig
kconfig: remove duplicate SWAP symbol defintions
um: create a proper drivers Kconfig
um: cleanup Kconfig files
um: stop abusing KBUILD_KCONFIG -
Pull Kconfig updates from Masahiro Yamada:
- show clearer error messages where pkg-config is needed, but not
installed- rename SYMBOL_AUTO to SYMBOL_NO_WRITE to reflect its semantics
- create all necessary directories by Kconfig tool itself instead of
Makefile- update the .config unconditionally when syncconfig is invoked
- use 'include' directive instead of '-include' where
include/config/{auto,tristate}.conf is mandatory- do not try to update the .config when running install targets
- add .DELETE_ON_ERROR to delete partially updated files
- misc cleanups and fixes
* tag 'kconfig-v4.19' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: remove P_ENV property type
kconfig: remove unused sym_get_env_prop() function
kconfig: fix the rule of mainmenu_stmt symbol
init/Kconfig: Use short unix-style option instead of --longname
Kbuild: Makefile.modbuiltin: include auto.conf and tristate.conf mandatory
kbuild: remove auto.conf from prerequisite of phony targets
kbuild: do not update config for 'make kernelrelease'
kbuild: do not update config when running install targets
kbuild: add .DELETE_ON_ERROR special target
kbuild: use 'include' directive to load auto.conf from top Makefile
kconfig: allow all config targets to write auto.conf if missing
kconfig: make syncconfig update .config regardless of sym_change_count
kconfig: create directories needed for syncconfig by itself
kconfig: remove unneeded directory generation from local*config
kconfig: split out useful helpers in confdata.c
kconfig: rename file_write_dep and move it to confdata.c
kconfig: fix typos in description of "choice" in kconfig-language.txt
kconfig: handle format string before calling conf_message_callback()
kconfig: rename SYMBOL_AUTO to SYMBOL_NO_WRITE
kconfig: check for pkg-config on make {menu,n,g,x}config -
Pull Kbuild updates from Masahiro Yamada:
- verify depmod is installed before modules_install
- support build salt in case build ids must be unique between builds
- allow users to specify additional host compiler flags via HOST*FLAGS,
and rename internal variables to KBUILD_HOST*FLAGS- update buildtar script to drop vax support, add arm64 support
- update builddeb script for better debarch support
- document the pit-fall of if_changed usage
- fix parallel build of UML with O= option
- make 'samples' target depend on headers_install to fix build errors
- remove deprecated host-progs variable
- add a new coccinelle script for refcount_t vs atomic_t check
- improve double-test coccinelle script
- misc cleanups and fixes
* tag 'kbuild-v4.19' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (41 commits)
coccicheck: return proper error code on fail
Coccinelle: doubletest: reduce side effect false positives
kbuild: remove deprecated host-progs variable
kbuild: make samples really depend on headers_install
um: clean up archheaders recipe
kbuild: add %asm-generic to no-dot-config-targets
um: fix parallel building with O= option
scripts: Add Python 3 support to tracing/draw_functrace.py
builddeb: Add automatic support for sh{3,4}{,eb} architectures
builddeb: Add automatic support for riscv* architectures
builddeb: Add automatic support for m68k architecture
builddeb: Add automatic support for or1k architecture
builddeb: Add automatic support for sparc64 architecture
builddeb: Add automatic support for mips{,64}r6{,el} architectures
builddeb: Add automatic support for mips64el architecture
builddeb: Add automatic support for ppc64 and powerpcspe architectures
builddeb: Introduce functions to simplify kconfig tests in set_debarch
builddeb: Drop check for 32-bit s390
builddeb: Change architecture detection fallback to use dpkg-architecture
builddeb: Skip architecture detection when KBUILD_DEBARCH is set
...
14 Aug, 2018
1 commit
-
Pull s390 updates from Heiko Carstens:
"Since Martin is on vacation you get the s390 pull request from me:- Host large page support for KVM guests. As the patches have large
impact on arch/s390/mm/ this series goes out via both the KVM and
the s390 tree.- Add an option for no compression to the "Kernel compression mode"
menu, this will come in handy with the rework of the early boot
code.- A large rework of the early boot code that will make life easier
for KASAN and KASLR. With the rework the bootable uncompressed
image is not generated anymore, only the bzImage is available. For
debuggung purposes the new "no compression" option is used.- Re-enable the gcc plugins as the issue with the latent entropy
plugin is solved with the early boot code rework.- More spectre relates changes:
+ Detect the etoken facility and remove expolines automatically.
+ Add expolines to a few more indirect branches.- A rewrite of the common I/O layer trace points to make them
consumable by 'perf stat'.- Add support for format-3 PCI function measurement blocks.
- Changes for the zcrypt driver:
+ Add attributes to indicate the load of cards and queues.
+ Restructure some code for the upcoming AP device support in KVM.- Build flags improvements in various Makefiles.
- A few fixes for the kdump support.
- A couple of patches for gcc 8 compile warning cleanup.
- Cleanup s390 specific proc handlers.
- Add s390 support to the restartable sequence self tests.
- Some PTR_RET vs PTR_ERR_OR_ZERO cleanup.
- Lots of bug fixes"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (107 commits)
s390/dasd: fix hanging offline processing due to canceled worker
s390/dasd: fix panic for failed online processing
s390/mm: fix addressing exception after suspend/resume
rseq/selftests: add s390 support
s390: fix br_r1_trampoline for machines without exrl
s390/lib: use expoline for all bcr instructions
s390/numa: move initial setup of node_to_cpumask_map
s390/kdump: Fix elfcorehdr size calculation
s390/cpum_sf: save TOD clock base in SDBs for time conversion
KVM: s390: Add huge page enablement control
s390/mm: Add huge page gmap linking support
s390/mm: hugetlb pages within a gmap can not be freed
KVM: s390: Add skey emulation fault handling
s390/mm: Add huge pmd storage key handling
s390/mm: Clear skeys for newly mapped huge guest pmds
s390/mm: Clear huge page storage keys on enable_skey
s390/mm: Add huge page dirty sync support
s390/mm: Add gmap pmd invalidation and clearing
s390/mm: Add gmap pmd notification bit setting
s390/mm: Add gmap pmd linking
...
09 Aug, 2018
1 commit
-
Avoids warning messages with the latest release of toybox, which never
bothered to implement the --longopts nothing was using.Signed-off-by: Rob Landley
Signed-off-by: Masahiro Yamada
02 Aug, 2018
3 commits
-
Almost all architectures include it. Add a ARCH_NO_PREEMPT symbol to
disable preempt support for alpha, hexagon, non-coldfire m68k and
user mode Linux.Signed-off-by: Christoph Hellwig
Signed-off-by: Masahiro Yamada -
Instead of duplicating the source statements in every architecture just
do it once in the toplevel Kconfig file.Note that with this the inclusion of arch/$(SRCARCH/Kconfig moves out of
the top-level Kconfig into arch/Kconfig so that don't violate ordering
constraits while keeping a sensible menu structure.Signed-off-by: Christoph Hellwig
Signed-off-by: Masahiro Yamada -
microblaze and nios2 define their own always n SWAP symbols. Remove those
and let the generic defintion do the right thing by adding a new symbol
to disable swap entirely.Signed-off-by: Christoph Hellwig
Signed-off-by: Masahiro Yamada
18 Jul, 2018
1 commit
-
In Fedora, the debug information is packaged separately (foo-debuginfo) and
can be installed separately. There's been a long standing issue where only
one version of a debuginfo info package can be installed at a time. There's
been an effort for Fedora for parallel debuginfo to rectify this problem.Part of the requirement to allow parallel debuginfo to work is that build ids
are unique between builds. The existing upstream rpm implementation ensures
this by re-calculating the build-id using the version and release as a
seed. This doesn't work 100% for the kernel because of the vDSO which is
its own binary and doesn't get updated when embedded.Fix this by adding some data in an ELF note for both the kernel and modules.
The data is controlled via a Kconfig option so distributions can set it
to an appropriate value to ensure uniqueness between builds.Suggested-by: Masahiro Yamada
Signed-off-by: Laura Abbott
Signed-off-by: Masahiro Yamada
01 Jul, 2018
1 commit
-
…masahiroy/linux-kbuild
Pull Kbuild fixes from Masahiro Yamada:
- introduce __diag_* macros and suppress -Wattribute-alias warnings
from GCC 8- fix stack protector test script for x86_64
- fix line number handling in Kconfig
- document that '#' starts a comment in Kconfig
- handle P_SYMBOL property in dump debugging of Kconfig
- correct help message of LD_DEAD_CODE_DATA_ELIMINATION
- fix occasional segmentation faults in Kconfig
* tag 'kbuild-fixes-v4.18' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: loop boundary condition fix
kbuild: reword help of LD_DEAD_CODE_DATA_ELIMINATION
kconfig: handle P_SYMBOL in print_symbol()
kconfig: document Kconfig source file comments
kconfig: fix line numbers for if-entries in menu tree
stack-protector: Fix test with 32-bit userland and CONFIG_64BIT=y
powerpc: Remove -Wattribute-alias pragmas
disable -Wattribute-alias warning for SYSCALL_DEFINEx()
kbuild: add macro for controlling warnings to linux/compiler.h
28 Jun, 2018
1 commit
-
Since commit 5d20ee3192a5 ("kbuild: Allow LD_DEAD_CODE_DATA_ELIMINATION
to be selectable if enabled"), HAVE_LD_DEAD_CODE_DATA_ELIMINATION is
supposed to be selected by architectures that are capable of this
functionality. LD_DEAD_CODE_DATA_ELIMINATION is now users' selection.
Update the help message.Signed-off-by: Masahiro Yamada
25 Jun, 2018
1 commit
-
Add "None" as the kernel compression mode.
This option is useful for debugging the kernel in slow simulation
environments, where decompressing and moving the kernel is awfully slow.Uncompressed kernel implementation might allow early boot code to skip the
decompressor and jump right at uncompressed kernel image entry point.Platforms implementing that should define HAVE_KERNEL_UNCOMPRESSED.
Reviewed-by: Heiko Carstens
Signed-off-by: Vasily Gorbik
Signed-off-by: Martin Schwidefsky
14 Jun, 2018
1 commit
-
Currently the code is split over various files with dma- prefixes in the
lib/ and drives/base directories, and the number of files keeps growing.
Move them into a single directory to keep the code together and remove
the file name prefixes. To match the irq infrastructure this directory
is placed under the kernel/ directory.Signed-off-by: Christoph Hellwig