If the architecture supports the batching of jump label updates, use it!
An easy way to see the benefits of this patch is switching the
schedstats on and off. For instance:
-------------------------- %< ----------------------------
#!/bin/sh
while [ true ]; do
sysctl -w kernel.sched_schedstats=1
sleep 2
sysctl -w kernel.sched_schedstats=0
sleep 2
done
-------------------------- >% ----------------------------
while watching the IPI count:
-------------------------- %< ----------------------------
# watch -n1 "cat /proc/interrupts | grep Function"
-------------------------- >% ----------------------------
With the current mode, it is possible to see +- 168 IPIs each 2 seconds,
while with this patch the number of IPIs goes to 3 each 2 seconds.
Regarding the performance impact of this patch set, I made two measurements:
The time to update a key (the task that is causing the change)
The time to run the int3 handler (the side effect on a thread that
hits the code being changed)
The schedstats static key was chosen as the key to being switched on and off.
The reason being is that it is used in more than 56 places, in a hot path. The
change in the schedstats static key will be done with the following command:
while [ true ]; do
sysctl -w kernel.sched_schedstats=1
usleep 500000
sysctl -w kernel.sched_schedstats=0
usleep 500000
done
In this way, they key will be updated twice per second. To force the hit of the
int3 handler, the system will also run a kernel compilation with two jobs per
CPU. The test machine is a two nodes/24 CPUs box with an Intel Xeon processor
@2.27GHz.
Regarding the update part, on average, the regular kernel takes 57 ms to update
the schedstats key, while the kernel with the batch updates takes just 1.4 ms
on average. Although it seems to be too good to be true, it makes sense: the
schedstats key is used in 56 places, so it was expected that it would take
around 56 times to update the keys with the current implementation, as the
IPIs are the most expensive part of the update.
Regarding the int3 handler, the non-batch handler takes 45 ns on average, while
the batch version takes around 180 ns. At first glance, it seems to be a high
value. But it is not, considering that it is doing 56 updates, rather than one!
It is taking four times more, only. This gain is possible because the patch
uses a binary search in the vector: log2(56)=5.8. So, it was expected to have
an overhead within four times.
(voice of tv propaganda) But, that is not all! As the int3 handler keeps on for
a shorter period (because the update part is on for a shorter time), the number
of hits in the int3 handler decreased by 10%.
The question then is: Is it worth paying the price of "135 ns" more in the int3
handler?
Considering that, in this test case, we are saving the handling of 53 IPIs,
that takes more than these 135 ns, it seems to be a meager price to be paid.
Moreover, the test case was forcing the hit of the int3, in practice, it
does not take that often. While the IPI takes place on all CPUs, hitting
the int3 handler or not!
For instance, in an isolated CPU with a process running in user-space
(nohz_full use-case), the chances of hitting the int3 handler is barely zero,
while there is no way to avoid the IPIs. By bounding the IPIs, we are improving
a lot this scenario.
Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Chris von Recklinghausen <crecklin@redhat.com>
Cc: Clark Williams <williams@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jason Baron <jbaron@akamai.com>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Scott Wood <swood@redhat.com>
Cc: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/acc891dbc2dbc9fd616dd680529a2337b1d1274c.1560325897.git.bristot@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
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
#endif
We 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 <yamada.masahiro@socionext.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Jump table entries are mostly read-only, with the exception of the
init and module loader code that defuses entries that point into init
code when the code being referred to is freed.
For robustness, it would be better to move these entries into the
ro_after_init section, but clearing the 'code' member of each jump
table entry referring to init code at module load time races with the
module_enable_ro() call that remaps the ro_after_init section read
only, so we'd like to do it earlier.
So given that whether such an entry refers to init code can be decided
much earlier, we can pull this check forward. Since we may still need
the code entry at this point, let's switch to setting a low bit in the
'key' member just like we do to annotate the default state of a jump
table entry.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Jessica Yu <jeyu@kernel.org>
Link: https://lkml.kernel.org/r/20180919065144.25010-8-ard.biesheuvel@linaro.org
Enabling HARDENED_USERCOPY may cause measurable regressions in networking
performance: up to 8% under UDP flood.
I ran a small packet UDP flood using pktgen vs. a host b2b connected. On
the receiver side the UDP packets are processed by a simple user space
process that just reads and drops them:
https://github.com/netoptimizer/network-testing/blob/master/src/udp_sink.c
Not very useful from a functional PoV, but it helps to pin-point
bottlenecks in the networking stack.
When running a kernel with CONFIG_HARDENED_USERCOPY=y, I see a 5-8%
regression in the receive tput, compared to the same kernel without this
option enabled.
With CONFIG_HARDENED_USERCOPY=y, perf shows ~6% of CPU time spent
cumulatively in __check_object_size (~4%) and __virt_addr_valid (~2%).
The call-chain is:
__GI___libc_recvfrom
entry_SYSCALL_64_after_hwframe
do_syscall_64
__x64_sys_recvfrom
__sys_recvfrom
inet_recvmsg
udp_recvmsg
__check_object_size
udp_recvmsg() actually calls copy_to_iter() (inlined) and the latters
calls check_copy_size() (again, inlined).
A generic distro may want to enable HARDENED_USERCOPY in their default
kernel config, but at the same time, such distro may want to be able to
avoid the performance penalties in with the default configuration and
disable the stricter check on a per-boot basis.
This change adds a boot parameter that conditionally disables
HARDENED_USERCOPY via "hardened_usercopy=off".
Signed-off-by: Chris von Recklinghausen <crecklin@redhat.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
With the following commit:
3335224470 ("jump_label: Explicitly disable jump labels in __init code")
... we explicitly disabled jump labels in __init code, so they could be
detected and not warned about in the following commit:
dc1dd184c2 ("jump_label: Warn on failed jump_label patching attempt")
In-kernel __exit code has the same issue. It's never used, so it's
freed along with the rest of initmem. But jump label entries in __exit
code aren't explicitly disabled, so we get the following warning when
enabling pr_debug() in __exit code:
can't patch jump_label at dmi_sysfs_exit+0x0/0x2d
WARNING: CPU: 0 PID: 22572 at kernel/jump_label.c:376 __jump_label_update+0x9d/0xb0
Fix the warning by disabling all jump labels in initmem (which includes
both __init and __exit code).
Reported-and-tested-by: Li Wang <liwang@redhat.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jason Baron <jbaron@akamai.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: dc1dd184c2 ("jump_label: Warn on failed jump_label patching attempt")
Link: http://lkml.kernel.org/r/7121e6e595374f06616c505b6e690e275c0054d1.1521483452.git.jpoimboe@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
After initmem has been freed, any jump labels in __init code are
prevented from being written to by the kernel_text_address() check in
__jump_label_update(). However, this check is quite broad. If
kernel_text_address() were to return false for any other reason, the
jump label write would fail silently with no warning.
For jump labels in module init code, entry->code is set to zero to
indicate that the entry is disabled. Do the same thing for core kernel
init code. This makes the behavior more consistent, and will also make
it more straightforward to detect non-init jump label write failures in
the next patch.
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jason Baron <jbaron@akamai.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/c52825c73f3a174e8398b6898284ec20d4deb126.1519051220.git.jpoimboe@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Pull tracing fixes from Steven Rostedt:
"There was some breakage with the changes for jump labels in the 4.11
merge window:
- powerpc broke as jump labels uses the two LSB bits as flags in
initialization.
A check was added to make sure that all jump label entries were 4
bytes aligned, but powerpc didn't work that way for modules. Adding
an alignment in the module linker script appeared to be the best
solution.
- Jump labels also added an anonymous union to access those LSB bits
as a normal long. But because this structure had static
initialization, it broke older compilers that could not statically
initialize anonymous unions without brackets.
- The command line parameter for setting function graph filter broke
the "EMPTY_HASH" descriptor by modifying it instead of creating a
new hash to hold the entries.
- The command line parameter ftrace_graph_max_depth was added to
allow its setting at boot time. It uses existing code and only the
command line hook was added.
This is not really a fix, but as it uses existing code without
affecting anything else, I added it to this release. It was ready
before the merge window closed, but I wanted to let it sit in
linux-next for a couple of days first"
* tag 'trace-v4.11-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
ftrace/graph: Add ftrace_graph_max_depth kernel parameter
tracing: Add #undef to fix compile error
jump_label: Add comment about initialization order for anonymous unions
jump_label: Fix anonymous union initialization
module: set __jump_table alignment to 8
ftrace/graph: Do not modify the EMPTY_HASH for the function_graph filter
tracing: Fix code comment for ftrace_ops_get_func()
Commit 3821fd35b5 ("jump_label: Reduce the size of struct static_key")
broke old compilers that could not handle static initialization of anonymous
unions. Boris fixed it with a patch that added brackets around the static
initializer. But this creates a dependency between those initializers and
the structure's order of its fields. Document this dependency in case new
fields are added to struct static_key in the future.
Noted-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Suggested-by: Chris Mason <clm@fb.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Pull tracing updates from Steven Rostedt:
"This release has no new tracing features, just clean ups, minor fixes
and small optimizations"
* tag 'trace-v4.11' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (25 commits)
tracing: Remove outdated ring buffer comment
tracing/probes: Fix a warning message to show correct maximum length
tracing: Fix return value check in trace_benchmark_reg()
tracing: Use modern function declaration
jump_label: Reduce the size of struct static_key
tracing/probe: Show subsystem name in messages
tracing/hwlat: Update old comment about migration
timers: Make flags output in the timer_start tracepoint useful
tracing: Have traceprobe_probes_write() not access userspace unnecessarily
tracing: Have COMM event filter key be treated as a string
ftrace: Have set_graph_function handle multiple functions in one write
ftrace: Do not hold references of ftrace_graph_{notrace_}hash out of graph_lock
tracing: Reset parser->buffer to allow multiple "puts"
ftrace: Have set_graph_functions handle write with RDWR
ftrace: Reset fgd->hash in ftrace_graph_write()
ftrace: Replace (void *)1 with a meaningful macro name FTRACE_GRAPH_EMPTY
ftrace: Create a slight optimization on searching the ftrace_hash
tracing: Add ftrace_hash_key() helper function
ftrace: Convert graph filter to use hash tables
ftrace: Expose ftrace_hash_empty and ftrace_lookup_ip
...
The static_key->next field goes mostly unused. The field is used for
associating module uses with a static key. Most uses of struct static_key
define a static key in the core kernel and make use of it entirely within
the core kernel, or define the static key in a module and make use of it
only from within that module. In fact, of the ~3,000 static keys defined,
I found only about 5 or so that did not fit this pattern.
Thus, we can remove the static_key->next field entirely and overload
the static_key->entries field. That is, when all the static_key uses
are contained within the same module, static_key->entries continues
to point to those uses. However, if the static_key uses are not contained
within the module where the static_key is defined, then we allocate a
struct static_key_mod, store a pointer to the uses within that
struct static_key_mod, and have the static key point at the static_key_mod.
This does incur some extra memory usage when a static_key is used in a
module that does not define it, but since there are only a handful of such
cases there is a net savings.
In order to identify if the static_key->entries pointer contains a
struct static_key_mod or a struct jump_entry pointer, bit 1 of
static_key->entries is set to 1 if it points to a struct static_key_mod and
is 0 if it points to a struct jump_entry. We were already using bit 0 in a
similar way to store the initial value of the static_key. This does mean
that allocations of struct static_key_mod and that the struct jump_entry
tables need to be at least 4-byte aligned in memory. As far as I can tell
all arches meet this criteria.
For my .config, the patch increased the text by 778 bytes, but reduced
the data + bss size by 14912, for a net savings of 14,134 bytes.
text data bss dec hex filename
8092427 5016512 790528 13899467 d416cb vmlinux.pre
8093205 5001600 790528 13885333 d3df95 vmlinux.post
Link: http://lkml.kernel.org/r/1486154544-4321-1-git-send-email-jbaron@akamai.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Joe Perches <joe@perches.com>
Signed-off-by: Jason Baron <jbaron@akamai.com>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Pull RAS updates from Ingo Molnar:
"The main changes were:
- Lots of enhancements for AMD SMCA (Scalable MCA
features/extensions) systems: extract, decode and print more
hardware error information and add matching support on the
injection/testing side as well. (Yazn Ghannam)
- Various MCE handling improvements on modern Intel Xeons. (Tony
Luck)
- Plus misc fixes and enhancements"
* 'ras-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (21 commits)
x86/RAS/mce_amd_inj: Remove debugfs dir recursively on exit
x86/RAS/mce_amd_inj: Fix signed wrap around when decrementing index 'i'
x86/RAS/mce_amd_inj: Fix some W= warnings
x86/MCE/AMD, EDAC: Handle reserved bank 4 on Fam17h properly
x86/mce/AMD: Extract the error address on SMCA systems
x86/mce, EDAC/mce_amd: Print MCA_SYND and MCA_IPID during MCE on SMCA systems
x86/mce/AMD: Save MCA_IPID in MCE struct on SMCA systems
x86/mce/AMD: Ensure the deferred error interrupt is of type APIC on SMCA systems
x86/mce/AMD: Update sysfs bank names for SMCA systems
x86/mce/AMD, EDAC/mce_amd: Define and use tables for known SMCA IP types
EDAC/mce_amd: Use SMCA prefix for error descriptions arrays
EDAC/mce_amd: Add missing SMCA error descriptions
x86/mce/AMD: Read MSRs on the CPU allocating the threshold blocks
x86/RAS: Add syndrome support to mce_amd_inj
EDAC/mce_amd: Print syndrome register value on SMCA systems
x86/mce: Add support for new MCA_SYND register
x86/mce/AMD: Use msr_ops.misc() in allocate_threshold_blocks()
x86/mce: Drop X86_FEATURE_MCE_RECOVERY and the related model string test
x86/mce: Improve memcpy_mcsafe()
x86/mce: Add PCI quirks to identify Xeons with machine check recovery
...
The static key API is currently designed around single variable
definitions. There are cases where an array of static keys is desirable,
so extend the API to allow this rather than using the internal static
key implementation directly.
Cc: Jason Baron <jbaron@akamai.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Suggested-by: Dave P Martin <Dave.Martin@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Daniel Bristot de Oliveira
Masahiro Yamada
Ard Biesheuvel
Chris von Recklinghausen
Josh Poimboeuf
Peter Zijlstra
Ingo Molnar
Greg Kroah-Hartman
Borislav Petkov
Marc Zyngier
Paolo Bonzini
Linus Torvalds
Steven Rostedt (VMware)
Boris Ostrovsky
Jason Baron
Luis R. Rodriguez
Catalin Marinas
Tony Luck