- A series from Arnd to clean up cpu_is_mx*() from i.MX platform
- A series from Andrey to clean up i.MX L2-cache code by using
core support as much as possible
- Remove the orphan header eukrea-baseboards.h from i.MX platform
- Remove boilerplate code from TZIC driver by using IRQCHIP_DECLARE
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJXcje3AAoJEFBXWFqHsHzO1DcH/2DOFJOaTTAPLJKI4rChNw4/
04iNpUulFoFuPOI6xmjfMM8lxNQKEiQ/VLE6NuuV9OLiBSSkblEsZ7s50b0hubOC
78bVsbM4UvWU1Sc7LIdxe/OZJFT3RaESIWXfeSrZjak6GAbZDzW5VVcovE12ih87
NY/PcS6aAxASjO6PiQuQ+LTlhXxxiEUzyseLwGcLl2SUnqoL4pwN4KYlg93nJUDy
Fli6+ZDBkM5aCYWZG0xcUiueWcbRGXHUbEe7QYy7czV5jNGfLZOgWSeRli2H++VR
G20NqcRcsU+/BrBOiQjHVS/QND4A4jPksiDH0joWYQH9HXC4GJpSHwW11pOTifU=
=Vena
-----END PGP SIGNATURE-----
Merge tag 'imx-cleanup-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into next/cleanup
The i.MX cleanup for 4.8:
- A series from Arnd to clean up cpu_is_mx*() from i.MX platform
- A series from Andrey to clean up i.MX L2-cache code by using
core support as much as possible
- Remove the orphan header eukrea-baseboards.h from i.MX platform
- Remove boilerplate code from TZIC driver by using IRQCHIP_DECLARE
* tag 'imx-cleanup-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
ARM: imx: remove cpu_is_mx*()
ARM: imx: remove last call to cpu_is_mx5*
ARM: imx: rework mx27_pm_init() call
ARM: imx: deconstruct mx3_idle
ARM: imx: deconstruct mxc_rnga initialization
ARM: imx: remove cpu_is_mx1 check
ARM: i.MX: Do not explicitly call l2x0_of_init()
ARM: i.MX: system.c: Tweak prefetch settings for performance
ARM: i.MX: system.c: Replace magic numbers
ARM: i.MX: system.c: Remove redundant errata 752271 code
ARM: i.MX: system.c: Convert goto to if statement
ARM: imx: Use IRQCHIP_DECLARE for TZIC
ARM: imx: Remove orphan header
Signed-off-by: Olof Johansson <olof@lixom.net>
Various warning fixes
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iEYEABECAAYFAldpAwMACgkQCwYYjhRyO9WIhACeJQPPSTdBX/Gaa23T1aQT9gwq
/g8AoJeSQumzfxqnQLCuz86ba4sXQkyx
=JyAq
-----END PGP SIGNATURE-----
Merge tag 'mvebu-cleanup-4.8-1' of git://git.infradead.org/linux-mvebu into next/cleanup
mvebu cleanup for 4.8 (part 1)
Various warning fixes
* tag 'mvebu-cleanup-4.8-1' of git://git.infradead.org/linux-mvebu:
ARM: Kirkwood: fix kirkwood_pm_init() declaration/type
ARM: Kirkwood: make kirkwood_disable_mbus_error_propagation() static
ARM: orion5x: make orion5x_legacy_handle_irq static
mvebu: add definition for coherency_base
mvebu: make mvebu_armada375_smp_wa_init() static
mvebu: fix missing include of common.h in cpu-reset.c
mvebu: fix missing include of common.h in pm.c
mvebu: fix missing includes in pmsu.c
Signed-off-by: Olof Johansson <olof@lixom.net>
The mxc_cpu_type and cpu_is_mx() logic is largely unused, and the
few remaining users were easy to convert into simpler code. Now that
they are gone, we can remove all those macros as well.
The related cpu_is_imx6*() set of function unfortunately is harder
to remove, so those are staying around for now.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The check for cpu_is_mx51/cpu_is_mx53() in mx51_revision()/mx53_revision()
is just a safety precaution, but there are only two callers of this
are using it only on the correct CPUs, and none of the other respective
functions have this extra check.
Removing these lets us kill off the cpu_is_* functions.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
mx27_pm_init() uses its own initcall, unlike all of the other
functions like it. Replacing the initcall with a .init_late()
callback makes imx27 more like the others and lets us remove
the last caller of cpu_is_mx27().
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The imx31 and imx35 idle functions are almost the same, but we
currently have to check the cpu type every time. This can be
simplified by moving the logic from mx3_cpu_lp_set() into
two separate idle functions, removing the last user of
cpu_is_mx35.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The rnga platform device is initialized for all imx31 machines
from its own initcall, but is never initialized anywhere else.
This moves the platform device creation into both the imx31
dt and non-dt machine init sequences, which has basically the
exact same effect as before, but makes it more obvious what
is going on, while reducing the amount of code and removing
the last user of cpu_is_mx31().
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
There is only one call site for this, and it's easily replaced
by initializing the reset value at boot time.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
There's no need to explicitly call l2x0_of_init() since it will be
called as a part of init_IRQ() (see arch/arm/kernel/irq.c for
details). This way we can simplify imx_init_l2cache() and ditch the call
to it on i.MX35 (which does not claim compatibility with
"arm,pl310-cache") alltogether.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Suggested-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Update Prefetch Control Register settings to match that of Freescale's
Linux tree. As the commit e3addf1b773964eac7f797e8538c69481be4279c
states (author Nitin Garg):
"... set Prefetch offset to 15, since it improves memcpy performance by
35%. Don't enable Incr double Linefill enable since it adversely affects
memcpy performance by about 32MB/s and reads by 90MB/s. Tested with 4K
to 16MB sized src and dst aligned buffer..."
Those results are also corroborated by our own testing.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Tested-by: Chris Healy <cphealy@gmail.com>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Applying a fix for ARM errata 752271 would already be taken care by a
call to a 'fixup' hook as a part of l2x0_of_init() -> __l2c_init() call
chain. Moreso the code in 'fixup' function would do that based on the
PL310's revsion information, whereas removed code does so based on SoC
version which does not work very well on i.MX6Q+ which identifies itself
as i.MX6Q as well but is not affected by 752271.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Using goto here doesn't bring any advantages and only makes the code
flow less clear. No functional changes.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
Few fixes to remove build warnings with W=1
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=Izhy
-----END PGP SIGNATURE-----
Merge tag 'vexpress-fixes-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux into next/cleanup
ARMv7 VExpress fixes for v4.8
Few fixes to remove build warnings with W=1
* tag 'vexpress-fixes-4.8' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux:
power: vexpress: make dev_attr_active static
ARM: vexpress/spc: fix missing include of spc.h
ARM: versatile: fix missing <plat/platsmp.h> include
ARM: vexpress/hotplug: fix missing core.h include
ARM: vexpress/spc: remove unused variable perf_stat_reg
Signed-off-by: Olof Johansson <olof@lixom.net>
The kirkwood-pm.c was missing the include of kirkwood-pm.h to
define the kirkwood_pm_init() function. However once this is
included, the types do not match.
Fixup the include, and then the prototype to avoid the following
warning:
arch/arm/mach-mvebu/kirkwood-pm.c:69:12: warning: symbol 'kirkwood_pm_init' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The kirkwood_disable_mbus_error_propagation is not exported or declared
elsewhere, so make it static to avoid the following warning:
arch/arm/mach-mvebu/kirkwood.c:153:6: warning: symbol 'kirkwood_disable_mbus_error_propagation' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The orion5x_legacy_handle_irq() is not used or declared outside
of irq.c so make it static to avoid the following warning:
arch/arm/mach-orion5x/irq.c:30:23: warning: symbol 'orion5x_legacy_handle_irq' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Acked-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
Fix the warning that coherency_base is not defined by adding
it to coherency.h (it is only used in the coherency_ll.S):
arch/arm/mach-mvebu/coherency.c:41:14: warning: symbol 'coherency_base' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The mvebu_armada375_smp_wa_init() is not exported or declared
anywhere, so make it static to fix the following warning:
arch/arm/mach-mvebu/system-controller.c:130:6: warning: symbol 'mvebu_armada375_smp_wa_init' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The mvebu_cpu_reset_deassert() is missing the definition for
it, so include common.h where it is defined to fix the warning:
arch/arm/mach-mvebu/cpu-reset.c:25:5: warning: symbol 'mvebu_cpu_reset_deassert' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The mvebu_pm_suspend_init() is missing a definition, so
include common.h which defines this function into pm.c to
fix the following warning:
arch/arm/mach-mvebu/pm.c:235:12: warning: symbol 'mvebu_pm_suspend_init' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The file is missing definitions for some functions due to not
including two header files. Fix the following warnings by
including "pmsu.h" and <linux/mvebu-pmsu.h> in pmsu.c:
arch/arm/mach-mvebu/pmsu.c:127:5: warning: symbol 'mvebu_setup_boot_addr_wa' was not declared. Should it be static?
arch/arm/mach-mvebu/pmsu.c:267:5: warning: symbol 'armada_370_xp_pmsu_idle_enter' was not declared. Should it be static?
arch/arm/mach-mvebu/pmsu.c:313:5: warning: symbol 'armada_38x_do_cpu_suspend' was not declared. Should it be static?
arch/arm/mach-mvebu/pmsu.c:340:6: warning: symbol 'mvebu_v7_pmsu_idle_exit' was not declared. Should it be static?
arch/arm/mach-mvebu/pmsu.c:570:5: warning: symbol 'mvebu_pmsu_dfs_request' was not declared. Should it be static?
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Gregory CLEMENT <gregory.clement@free-electrons.com>
The dev_attr_active is not exported or defined to be used
outside the driver, so make it static to avoid the following
warning:
drivers/power/reset/vexpress-poweroff.c:77:1: warning: symbol 'dev_attr_active' was not declared. Should it be static?
Cc: Liviu Dudau <liviu.dudau@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Fix missing function prototypes found in spc.h by including
the file to remove the following warnings:
arch/arm/mach-vexpress/spc.c:131:6: warning: symbol 've_spc_global_wakeup_irq' was not declared. Should it be static?
arch/arm/mach-vexpress/spc.c:156:6: warning: symbol 've_spc_cpu_wakeup_irq' was not declared. Should it be static?
arch/arm/mach-vexpress/spc.c:185:6: warning: symbol 've_spc_set_resume_addr' was not declared. Should it be static?
arch/arm/mach-vexpress/spc.c:210:6: warning: symbol 've_spc_powerdown' was not declared. Should it be static?
arch/arm/mach-vexpress/spc.c:240:5: warning: symbol 've_spc_cpu_in_wfi' was not declared. Should it be static?
arch/arm/mach-vexpress/spc.c:450:12: warning: symbol 've_spc_init' was not declared. Should it be static?
Cc: Liviu Dudau <liviu.dudau@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Fix two missing function declarations by including the
<plat/platsmp.h> file where they are defined. Fixes:
arch/arm/plat-versatile/platsmp.c:35:6: warning: symbol 'versatile_secondary_init' was not declared. Should it be static?
arch/arm/plat-versatile/platsmp.c:50:5: warning: symbol 'versatile_boot_secondary' was not declared. Should it be static?
Cc: Liviu Dudau <liviu.dudau@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Fix the missing declaration of vexpress_cpu_die() by
including core.h where it is defined. Fixes:
arch/arm/mach-vexpress/hotplug.c:88:6: warning: symbol 'vexpress_cpu_die' was not declared. Should it be static?
Cc: Liviu Dudau <liviu.dudau@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Ben Dooks <ben.dooks@codethink.co.uk>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
The variable 'perf_stat_reg' in ve_spc_set_performance is assigned a
value but that is never used.
So let's remove the variable 'perf_stat_reg'
Reported-by: David Binderman <dcb314@hotmail.com>
Cc: Liviu Dudau <liviu.dudau@arm.com>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Remove boilerplate code by using IRQCHIP_DECLARE macro.
Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
The header eukrea-baseboards.h is not used, remove it.
Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Signed-off-by: Shawn Guo <shawnguo@kernel.org>
This replaces:
- "select ARCH_REQUIRE_GPIOLIB" with "select GPIOLIB" as this can
now be selected directly.
- "select ARCH_WANT_OPTIONAL_GPIOLIB" with no dependency: GPIOLIB
is now selectable by everyone, so we need not declare our
intent to select it.
When ordering the symbols the following rationale was used:
if the selects were in alphabetical order, I moved select GPIOLIB
to be in alphabetical order, but if the selects were not
maintained in alphabetical order, I just replaced
"select ARCH_REQUIRE_GPIOLIB" with "select GPIOLIB".
Cc: Michael Büsch <m@bues.ch>
Cc: arm@kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Olof Johansson <olof@lixom.net>
Commit 307d40c56b ("ARM: uniphier: rework SMP code to support new
System Bus binding") added a new DT binding for SMP code, but still
kept old code for the backward compatibility.
Linux 4.6 was out with both bindings supported, so it should not
hurt to drop the old code now. Moreover, the mainline code are
currently not used for any of our products, so this change has
no impact on our customers in any way.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Olof Johansson <olof@lixom.net>
Some later changes in cpufreq core require these tables to be sorted
based on ascending order of their frequencies. There was only one
offender. Fix it and add comments over the arrays.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Fix some language typos in comments.
Signed-off-by: Andrea Gelmini <andrea.gelmini@gelma.net>
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
The self-test was updated to cover zero-length strings; the function
needs to be updated, too.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Fixes: fcfd2fbf22 ("fs/namei.c: Add hashlen_string() function")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The original name was simply hash_string(), but that conflicted with a
function with that name in drivers/base/power/trace.c, and I decided
that calling it "hashlen_" was better anyway.
But you have to do it in two places.
[ This caused build errors for architectures that don't define
CONFIG_DCACHE_WORD_ACCESS - Linus ]
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Reported-by: Guenter Roeck <linux@roeck-us.net>
Fixes: fcfd2fbf22 ("fs/namei.c: Add hashlen_string() function")
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The HPFS filesystem used generic_show_options to produce string that is
displayed in /proc/mounts. However, there is a problem that the options
may disappear after remount. If we mount the filesystem with option1
and then remount it with option2, /proc/mounts should show both option1
and option2, however it only shows option2 because the whole option
string is replaced with replace_mount_options in hpfs_remount_fs.
To fix this bug, implement the hpfs_show_options function that prints
options that are currently selected.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit c8f33d0bec ("affs: kstrdup() memory handling") checks if the
kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: c8f33d0bec ("affs: kstrdup() memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # v4.1+
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit ce657611ba ("hpfs: kstrdup() out of memory handling") checks if
the kstrdup function returns NULL due to out-of-memory condition.
However, if we are remounting a filesystem with no change to
filesystem-specific options, the parameter data is NULL. In this case,
kstrdup returns NULL (because it was passed NULL parameter), although no
out of memory condition exists. The mount syscall then fails with
ENOMEM.
This patch fixes the bug. We fail with ENOMEM only if data is non-NULL.
The patch also changes the call to replace_mount_options - if we didn't
pass any filesystem-specific options, we don't call
replace_mount_options (thus we don't erase existing reported options).
Fixes: ce657611ba ("hpfs: kstrdup() out of memory handling")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Various builds (such as i386:allmodconfig) fail with
fs/binfmt_aout.c:133:2: error: expected identifier or '(' before 'return'
fs/binfmt_aout.c:134:1: error: expected identifier or '(' before '}' token
[ Oops. My bad, I had stupidly thought that "allmodconfig" covered this
on x86-64 too, but it obviously doesn't. Egg on my face. - Linus ]
Fixes: 5d22fc25d4 ("mm: remove more IS_ERR_VALUE abuses")
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull string hash improvements from George Spelvin:
"This series does several related things:
- Makes the dcache hash (fs/namei.c) useful for general kernel use.
(Thanks to Bruce for noticing the zero-length corner case)
- Converts the string hashes in <linux/sunrpc/svcauth.h> to use the
above.
- Avoids 64-bit multiplies in hash_64() on 32-bit platforms. Two
32-bit multiplies will do well enough.
- Rids the world of the bad hash multipliers in hash_32.
This finishes the job started in commit 689de1d6ca ("Minimal
fix-up of bad hashing behavior of hash_64()")
The vast majority of Linux architectures have hardware support for
32x32-bit multiply and so derive no benefit from "simplified"
multipliers.
The few processors that do not (68000, h8/300 and some models of
Microblaze) have arch-specific implementations added. Those
patches are last in the series.
- Overhauls the dcache hash mixing.
The patch in commit 0fed3ac866 ("namei: Improve hash mixing if
CONFIG_DCACHE_WORD_ACCESS") was an off-the-cuff suggestion.
Replaced with a much more careful design that's simultaneously
faster and better. (My own invention, as there was noting suitable
in the literature I could find. Comments welcome!)
- Modify the hash_name() loop to skip the initial HASH_MIX(). This
would let us salt the hash if we ever wanted to.
- Sort out partial_name_hash().
The hash function is declared as using a long state, even though
it's truncated to 32 bits at the end and the extra internal state
contributes nothing to the result. And some callers do odd things:
- fs/hfs/string.c only allocates 32 bits of state
- fs/hfsplus/unicode.c uses it to hash 16-bit unicode symbols not bytes
- Modify bytemask_from_count to handle inputs of 1..sizeof(long)
rather than 0..sizeof(long)-1. This would simplify users other
than full_name_hash"
Special thanks to Bruce Fields for testing and finding bugs in v1. (I
learned some humbling lessons about "obviously correct" code.)
On the arch-specific front, the m68k assembly has been tested in a
standalone test harness, I've been in contact with the Microblaze
maintainers who mostly don't care, as the hardware multiplier is never
omitted in real-world applications, and I haven't heard anything from
the H8/300 world"
* 'hash' of git://ftp.sciencehorizons.net/linux:
h8300: Add <asm/hash.h>
microblaze: Add <asm/hash.h>
m68k: Add <asm/hash.h>
<linux/hash.h>: Add support for architecture-specific functions
fs/namei.c: Improve dcache hash function
Eliminate bad hash multipliers from hash_32() and hash_64()
Change hash_64() return value to 32 bits
<linux/sunrpc/svcauth.h>: Define hash_str() in terms of hashlen_string()
fs/namei.c: Add hashlen_string() function
Pull out string hash to <linux/stringhash.h>
This will improve the performance of hash_32() and hash_64(), but due
to complete lack of multi-bit shift instructions on H8, performance will
still be bad in surrounding code.
Designing H8-specific hash algorithms to work around that is a separate
project. (But if the maintainers would like to get in touch...)
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: uclinux-h8-devel@lists.sourceforge.jp
Microblaze is an FPGA soft core that can be configured various ways.
If it is configured without a multiplier, the standard __hash_32()
will require a call to __mulsi3, which is a slow software loop.
Instead, use a shift-and-add sequence for the constant multiply.
GCC knows how to do this, but it's not as clever as some.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Alistair Francis <alistair.francis@xilinx.com>
Cc: Michal Simek <michal.simek@xilinx.com>
This provides a multiply by constant GOLDEN_RATIO_32 = 0x61C88647
for the original mc68000, which lacks a 32x32-bit multiply instruction.
Yes, the amount of optimization effort put in is excessive. :-)
Shift-add chain found by Yevgen Voronenko's Hcub algorithm at
http://spiral.ece.cmu.edu/mcm/gen.html
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Greg Ungerer <gerg@linux-m68k.org>
Cc: Andreas Schwab <schwab@linux-m68k.org>
Cc: Philippe De Muyter <phdm@macq.eu>
Cc: linux-m68k@lists.linux-m68k.org
This is just the infrastructure; there are no users yet.
This is modelled on CONFIG_ARCH_RANDOM; a CONFIG_ symbol declares
the existence of <asm/hash.h>.
That file may define its own versions of various functions, and define
HAVE_* symbols (no CONFIG_ prefix!) to suppress the generic ones.
Included is a self-test (in lib/test_hash.c) that verifies the basics.
It is NOT in general required that the arch-specific functions compute
the same thing as the generic, but if a HAVE_* symbol is defined with
the value 1, then equality is tested.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Greg Ungerer <gerg@linux-m68k.org>
Cc: Andreas Schwab <schwab@linux-m68k.org>
Cc: Philippe De Muyter <phdm@macq.eu>
Cc: linux-m68k@lists.linux-m68k.org
Cc: Alistair Francis <alistai@xilinx.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: uclinux-h8-devel@lists.sourceforge.jp
Patch 0fed3ac866 improved the hash mixing, but the function is slower
than necessary; there's a 7-instruction dependency chain (10 on x86)
each loop iteration.
Word-at-a-time access is a very tight loop (which is good, because
link_path_walk() is one of the hottest code paths in the entire kernel),
and the hash mixing function must not have a longer latency to avoid
slowing it down.
There do not appear to be any published fast hash functions that:
1) Operate on the input a word at a time, and
2) Don't need to know the length of the input beforehand, and
3) Have a single iterated mixing function, not needing conditional
branches or unrolling to distinguish different loop iterations.
One of the algorithms which comes closest is Yann Collet's xxHash, but
that's two dependent multiplies per word, which is too much.
The key insights in this design are:
1) Barring expensive ops like multiplies, to diffuse one input bit
across 64 bits of hash state takes at least log2(64) = 6 sequentially
dependent instructions. That is more cycles than we'd like.
2) An operation like "hash ^= hash << 13" requires a second temporary
register anyway, and on a 2-operand machine like x86, it's three
instructions.
3) A better use of a second register is to hold a two-word hash state.
With careful design, no temporaries are needed at all, so it doesn't
increase register pressure. And this gets rid of register copying
on 2-operand machines, so the code is smaller and faster.
4) Using two words of state weakens the requirement for one-round mixing;
we now have two rounds of mixing before cancellation is possible.
5) A two-word hash state also allows operations on both halves to be
done in parallel, so on a superscalar processor we get more mixing
in fewer cycles.
I ended up using a mixing function inspired by the ChaCha and Speck
round functions. It is 6 simple instructions and 3 cycles per iteration
(assuming multiply by 9 can be done by an "lea" instruction):
x ^= *input++;
y ^= x; x = ROL(x, K1);
x += y; y = ROL(y, K2);
y *= 9;
Not only is this reversible, two consecutive rounds are reversible:
if you are given the initial and final states, but not the intermediate
state, it is possible to compute both input words. This means that at
least 3 words of input are required to create a collision.
(It also has the property, used by hash_name() to avoid a branch, that
it hashes all-zero to all-zero.)
The rotate constants K1 and K2 were found by experiment. The search took
a sample of random initial states (I used 1023) and considered the effect
of flipping each of the 64 input bits on each of the 128 output bits two
rounds later. Each of the 8192 pairs can be considered a biased coin, and
adding up the Shannon entropy of all of them produces a score.
The best-scoring shifts also did well in other tests (flipping bits in y,
trying 3 or 4 rounds of mixing, flipping all 64*63/2 pairs of input bits),
so the choice was made with the additional constraint that the sum of the
shifts is odd and not too close to the word size.
The final state is then folded into a 32-bit hash value by a less carefully
optimized multiply-based scheme. This also has to be fast, as pathname
components tend to be short (the most common case is one iteration!), but
there's some room for latency, as there is a fair bit of intervening logic
before the hash value is used for anything.
(Performance verified with "bonnie++ -s 0 -n 1536:-2" on tmpfs. I need
a better benchmark; the numbers seem to show a slight dip in performance
between 4.6.0 and this patch, but they're too noisy to quote.)
Special thanks to Bruce fields for diligent testing which uncovered a
nasty fencepost error in an earlier version of this patch.
[checkpatch.pl formatting complaints noted and respectfully disagreed with.]
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Tested-by: J. Bruce Fields <bfields@redhat.com>
The "simplified" prime multipliers made very bad hash functions, so get rid
of them. This completes the work of 689de1d6ca.
To avoid the inefficiency which was the motivation for the "simplified"
multipliers, hash_64() on 32-bit systems is changed to use a different
algorithm. It makes two calls to hash_32() instead.
drivers/media/usb/dvb-usb-v2/af9015.c uses the old GOLDEN_RATIO_PRIME_32
for some horrible reason, so it inherits a copy of the old definition.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
Cc: Antti Palosaari <crope@iki.fi>
Cc: Mauro Carvalho Chehab <m.chehab@samsung.com>
Jungseung Lee