Use the new ChromeOS EC EC_CMD_PWM_{GET,SET}_DUTY commands to control
one or more PWMs attached to the Embedded Controller. Because the EC
allows us to modify the duty cycle (as a percentage, where U16_MAX is
100%) but not the period, we assign the period a fixed value of
EC_PWM_MAX_DUTY and reject all attempts to change it.
This driver supports only device tree at the moment, because that
provides a very flexible way of describing the relationship between PWMs
and their consumer devices (e.g., backlight). On a non-DT system, we'll
probably want to use the non-GENERIC addressing (i.e., we'll need to
make special device instances that will use EC_PWM_TYPE_KB_LIGHT or
EC_PWM_TYPE_DISPLAY_LIGHT), as well as the relatively inflexible
pwm_lookup infrastructure for matching devices. Defer that work for now.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The ChromeOS Embedded Controller can support controlling its attached
PWMs via its host-command interface. The number of supported PWMs varies
on a per-board basis, but we can autodetect this by checking the error
codes, so we don't need an extra property for this. And because the EC
only allows specifying the duty cycle and not the period, we don't
specify the period via pwm-cells, and instead have only support for one
cell -- to specify the index.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The EC_CMD_PWM_{GET,SET}_DUTY commands allow us to control a PWM that is
attached to the EC, rather than the main host SoC. The API provides
functionality-based (e.g., keyboard light, backlight) or index-based
addressing of the PWM(s). Duty cycles are represented by a 16-bit value,
where 0 maps to 0% duty cycle and U16_MAX maps to 100%. The period
cannot be controlled.
This command set is more generic than, e.g.,
EC_CMD_PWM_{GET,SET}_KEYBOARD_BACKLIGHT and could possibly used to
replace it on future products.
Let's update the command header to include the definitions.
Signed-off-by: Brian Norris <briannorris@chromium.org>
Acked-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
So that callers of cros_ec_cmd_xfer() don't have to repeat boilerplate
code when checking for errors from the EC side.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Reviewed-by: Benson Leung <bleung@chromium.org>
Signed-off-by: Brian Norris <briannorris@chromium.org>
Acked-by: Lee Jones <lee.jones@linaro.org>
Tested-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
When disabling a PWM channel, the PWM clock was being stopped
immediately after writing to PWM_DIS. As a result, the disabling
of the PWM channel did not complete properly, and the PWM output
might be left at the wrong level.
Fix this by waiting for the channel to be effectively disabled
(by checking the PWM_SR register) before disabling the clock.
Signed-off-by: Guillermo Rodriguez <guille.rodriguez@gmail.com>
Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Add device tree bindings for the PWM controller found on R-Car H3 SoCs.
The controller is compatible with the one found in earlier generations
of R-Car SoCs.
Signed-off-by: Ulrich Hecht <ulrich.hecht+renesas@gmail.com>
Acked-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Simon Horman <horms+renesas@verge.net.au>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Tegra186 has multiple PWM controllers with only one output instead of
one controller with four outputs in earlier SoC generations.
Add support for Tegra186 and detect the number of PWM outputs using
device tree match data.
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Reviewed-by: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Tegra186 has 8 different PWM controllers and each controller has only
one output. Earlier SoC generations have 4 PWM outputs per controller.
Add a device tree compatible string for Tegra186 to be able to
differentiate between the two.
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
duty_ns * (1 << PWM_DUTY_WIDTH) could overflow in integer calculation
when the PWM rate is low. Hence do all calculation on unsigned long long
to avoid overflow.
Signed-off-by: Hyong Bin Kim <hyongbink@nvidia.com>
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
To get 100 % duty cycle (always high), pulse width needs to be set to
256.
Signed-off-by: Victor(Weiguo) Pan <wpan@nvidia.com>
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Add reset control of the PWM controller to reset it before
accessing the PWM register.
Signed-off-by: Rohith Seelaboyina <rseelaboyina@nvidia.com>
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The former is much longer to type and is ambiguous because the value
stored in the field is not the (physical) base address of the memory-
mapped I/O registers, but the virtual address of those registers as
mapped through the MMU.
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
This macro is used to initialize the ->npwm field of the PWM chip. Use a
literal instead and make all other places rely on ->npwm.
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The PWM_PIN_LEVEL bit is leave unset by the kernel PWM driver.
Prior to commit 08ee77b5a5,
the PWM_PIN_LEVEL bit was always clear when the PWM was disable
and a 0 logic level was apply to the output.
According to the LPC32x0 User Manual [1],
the default value for bit 30 (PWM_PIN_LEVEL) is 0.
This change initialize the pin level to 0 (default value) and
update the register value accordingly.
[1] http://www.nxp.com/documents/user_manual/UM10326.pdf
Signed-off-by: Sylvain Lemieux <slemieux@tycoint.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
This adds a driver for the PWM block found in chips of the STMPE 24xx
series of multi-purpose I2C expanders. (I think STMPE means ST
Microelectronics Multi-Purpose Expander.) This PWM was designed in
accordance with Nokia specifications and is kind of weird and usually
just switched between max and zero duty cycle. However it is indeed a
PWM so it needs to live in the PWM subsystem.
This PWM is mostly used for white LED backlight.
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
This adds fairly standard device tree bindings for the STMPE PWM found
in STMPE24xx multi-purpose expanders.
Cc: devicetree@vger.kernel.org
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Implement the ->apply() function to add support for atomic update.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Tested-by: Brian Norris <briannorris@chromium.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The current logic will disable the PWM clk even if the PWM was left
enabled by the bootloader (because it's controlling a critical device
like a regulator for example).
Keep the PWM clk enabled if the PWM is enabled to avoid any glitches.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Tested-by: Brian Norris <briannorris@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Implement the ->get_state() function to expose initial state.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Tested-by: Brian Norris <briannorris@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The current implementation always round down the duty and period values,
while it would be better to round them to the closest integer.
These changes are needed in preparation of atomic update support to
prevent a period/duty cycle drift when executing several times the
'pwm_get_state() / modify / pwm_apply_state()' sequence.
Say you have an expected period of 3.333 us and a clk rate of
112.666667 MHz -- the clock frequency doesn't divide evenly, so the
period (stashed in nanoseconds) shrinks when we convert to the register
value and back, as follows:
pwm_apply_state(): register = period * 112666667 / 1000000000;
pwm_get_state(): period = register * 1000000000 / 112666667;
or in other words:
period = period * 112666667 / 1000000000 * 1000000000 / 112666667;
which yields a sequence like:
3333 -> 3328
3328 -> 3319
3319 -> 3310
3310 -> 3301
3301 -> 3292
3292 -> ... (etc) ...
With this patch, we'd see instead:
period = div_round_closest(period * 112666667, 1000000000) *
1000000000 / 112666667;
which yields a stable sequence:
3333 -> 3337
3337 -> 3337
3337 -> ... (etc) ...
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Brian Norris <briannorris@chromium.org>
Tested-by: Brian Norris <briannorris@chromium.org>
Tested-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Add support for the PWM controller present in Broadcom's iProc family of
SoCs. It has been tested on the Northstar+ bcm958625HR board.
Signed-off-by: Yendapally Reddy Dhananjaya Reddy <yendapally.reddy@broadcom.com>
[thierry.reding@gmail.com: bunch of coding style fixes, cleanups]
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
There is no need to check each time if the clk_rate defined or not when we call
pwm_lpss_config(). Move the check to ->probe() instead.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Intel Edison has 4 PWM channels on the die with the same IP as in
Broxton. Enable it.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
If duty_ns is large enough multiplying it by 255 overflows and results
wrong duty cycle value being programmed. For example with 10ms duty when
period is 20ms (50%) we get
255 * 10000000 / 20000000 = -87
because 255 * 10000000 overlows int. Whereas correct value should be
255 * 10000000 / 20000000 = 127
Fix this by using unsigned long long as type for on_time_div and changing
integer literals to use proper type annotation.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The base_unit calculation applies an offset of 0x2 which adds
significant error for lower frequencies and doesn't appear to be
warranted - rounding the division result gives a correct value.
Also, the upper limit check for base_unit is off-by-one; the upper
nibble of base_unit is invalid if >=128 according to the Table 88
in the Z8000 Processor Series Datasheet Volume 1 (Rev. 2).
Verified on UP Board (Cherry Trail) and Minnowboard Max (Bay Trail).
Signed-off-by: Dan O'Donovan <dan@emutex.com>
Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
The PWMSS local clock gating registers have no real purpose on OMAP ARM
devices. These registers were left over registers from DSP IP where the
PRCM doesn't exist. There is a silicon bug where gating and ungating clocks
don't function properly. TRMs will be update to indicate that these
registers shouldn't be touched.
Therefore, all code that accesses the PWMSS_CLKCONFIG or PWMSS_CLKSTATUS
will be removed by this patch with zero loss of functionality by the ECAP
and EPWM drivers.
Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
When using the old eCAP and ePWM bindings for AM335x and AM437x the clock
can be retrieved from the PWMSS parent. Newer bindings will insure that
this clock is provided via device tree.
Therefore, update this driver to support the newer and older bindings. In
the case of the older binding being used give a warning.
Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
[thierry.reding@gmail.com: rewrite slightly for readability]
Signed-off-by: Thierry Reding <thierry.reding@gmail.com>
Now that the node name has been changed from ehrpwm to pwm the document
should show this proper usage. Also change the unit address in the example
from 0 to the proper physical address value that should be used.
Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Thierry Reding <thierry.reding@gmail.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>
That's all that's ever asked for, and it makes the return
type of hash_long() consistent.
It also allows (upcoming patch) an optimized implementation
of hash_64 on 32-bit machines.
I tried adding a BUILD_BUG_ON to ensure the number of bits requested
was never more than 32 (most callers use a compile-time constant), but
adding <linux/bug.h> to <linux/hash.h> breaks the tools/perf compiler
unless tools/perf/MANIFEST is updated, and understanding that code base
well enough to update it is too much trouble. I did the rest of an
allyesconfig build with such a check, and nothing tripped.
Signed-off-by: George Spelvin <linux@sciencehorizons.net>
