This requires adjusting the CPU voltage, so enable the axp20x driver for
A20 boards.
In evbarm/awin/awin_machdep.c, the DCDC powering up the CPU cores needs to
be defined for each board. The board machine-dependant code (at this time,
only awin_machdep.c) has to provide a awin_set_mpu_volt() function
which calls the right PMU driver to change the CPU core voltage.
The CPU frequency/voltage table in awin_board.c comes from various
sources: linux kernel, device tree and fex scripts, and experiments on
olimex lime2 and cubieboard 2.
The following sysctls are provided (compatible with estd):
machdep.frequency.min,machdep.frequency.max: writable lower and upper
bounds of the useable frequencies. Affects machdep.frequency.available.
Lower bound defaults to 700Mhz, as does linux.
Upper bound defaults to 960Mhz, which is the boot frequency
on the boards I tested. There is a 1008Mhz entry available,
but requires an out of spec CPU voltage (more than 1.4V) so it's
not available by default.
machdep.frequency.available: list of available frequencies. This is
the CPU frequency/voltage table, bound by machdep.frequency.{min,max}.
machdep.frequency.current: current CPU speed. Write a new value to change
the CPU speed, only values from machdep.frequency.available are
accepted.
not (AWIN_REF_FREQ * n * k) / m (m is only used for the SATA clock).
On the boards I tested, m happens to be 2 so the correct value was returned
anyway.
by a shiftout ends up reading bits(0,1) instead of AWIN_CPU_CLK_SRC_SEL.
It happens that these bits (AWIN_AXI_CLK_DIV_RATIO) are 2 (divide by 3) at boot
(at last on cubieboard2 and olimex lime2), which matches
AWIN_CPU_CLK_SRC_SEL_PLL1, so this has gone unnoticed.
https://mail-index.netbsd.org/tech-kern/2015/10/14/msg019511.html
don't sleep on sme->sme_mtx in the callout but use mutex_tryenter()
and just reschedule the callout if we can't get the mutex now.
This fixes a deadlock which can happen if the backed wants to
sleep with timeout (e.g. cv_timedwait()) as the backed is called with
sme->sme_mtx held.
This is a stopgap measure for netbsd-7; sysmon should be changed to not
sleep (or call a backend which will sleep) with mutexes held.
address in the TEST-NET-2 network as noted in RFC5737 instead of using
the 1.1.1.1 address. Also, use port 7 (echo) for better style.
Fixes PR bin/46758 thanks to Lloyd Parkes.
While here, if we do fail to connect to the test address, don't spam
this to the console as it's a common case during the boot sequence.
For now at least it will refuse to touch it though as it needs to be
taught more abstraction about directory entries; currently it blindly
uses struct direct from ffs and will croak on the lfs64 64-bit inode
numbers.
To create an lfs64 volume, use -w 64. You can also force a 32-bit
volume with -w 32, but this will fail on devices thta are too large.
lfs64 is the default for devices > 1TB. (1TB rather than 2TB because
daddr_t is signed and negative block numbers can cause interesting
complications.)
For now print a warning that the lfs64 format isn't finalized, because
it isn't. For now any lfs64 use should be limited to test data,
benchmarking, and so forth.
This will be unflipped when the format is finalized - right now I
still have pending changes to the superblock in mind (to reduce the
number of redundant fields) so anything created now is not future-
proof. However, the code's also nearing being ready for testing; so
I'm doing this before turning it on as a precaution.
unconditionally, not only #ifndef _KERNEL. The kernel declarations
require cdefs.h and standard practice is to include cdefs.h where it's
used; they also require sys/ansi.h; and while they don't require
featuretest.h it's also harmless... and includes should be at the top
anyhow.
PR 46557 from Richard Hansen.
