Based on earlier discussions[1] we attempted to find a suitable
location for the omap DMA header in commit 2b6c4e73 (ARM: OMAP:
DMA: Move plat/dma.h to plat-omap/dma-omap.h) until the conversion
to dmaengine is complete.

Unfortunately that was before I was able to try to test compile
of the ARM multiplatform builds for omap2+, and the end result
was not very good.

So I'm creating yet another all over the place patch to cut the
last dependency for building omap2+ for ARM multiplatform. After
this, we have finally removed the driver dependencies to the
arch/arm code, except for few drivers that are being worked on.

The other option was to make the <plat-omap/dma-omap.h> path
to work, but we'd have to add some new header directory to for
multiplatform builds.

Or we would have to manually include arch/arm/plat-omap/include
again from arch/arm/Makefile for omap2+.

Neither of these alternatives sound appealing as they will
likely lead addition of various other headers exposed to the
drivers, which we want to avoid for the multiplatform kernels.

Since we already have a minimal include/linux/omap-dma.h,
let's just use that instead and add a note to it to not
use the custom omap DMA functions any longer where possible.

Note that converting omap DMA to dmaengine depends on
dmaengine supporting automatically incrementing the FIFO
address at the device end, and converting all the remaining
legacy drivers. So it's going to be few more merge windows.

[1] https://patchwork.kernel.org/patch/1519591/#



cc: Russell King <linux@arm.linux.org.uk>
cc: Kevin Hilman <khilman@ti.com>
cc: "Benoît Cousson" <b-cousson@ti.com>
cc: Herbert Xu <herbert@gondor.apana.org.au>
cc: "David S. Miller" <davem@davemloft.net>
cc: Vinod Koul <vinod.koul@intel.com>
cc: Dan Williams <djbw@fb.com>
cc: Mauro Carvalho Chehab <mchehab@infradead.org>
cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
cc: David Woodhouse <dwmw2@infradead.org>
cc: Kyungmin Park <kyungmin.park@samsung.com>
cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
cc: Florian Tobias Schandinat <FlorianSchandinat@gmx.de>
cc: Hans Verkuil <hans.verkuil@cisco.com>
cc: Vaibhav Hiremath <hvaibhav@ti.com>
cc: Lokesh Vutla <lokeshvutla@ti.com>
cc: Rusty Russell <rusty@rustcorp.com.au>
cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
cc: Afzal Mohammed <afzal@ti.com>
cc: linux-crypto@vger.kernel.org
cc: linux-media@vger.kernel.org
cc: linux-mtd@lists.infradead.org
cc: linux-usb@vger.kernel.org
cc: linux-fbdev@vger.kernel.org
Acked-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

We cannot include any plat or mach headers for the multiplatform
support.

Fix the issue by defining local mcbsp_omap1().

cc: Peter Ujfalusi <peter.ujfalusi@ti.com>
cc: Liam Girdwood <lrg@ti.com>
cc: Mark Brown <broonie@opensource.wolfsonmicro.com>
cc: Jaroslav Kysela <perex@perex.cz>
cc: Takashi Iwai <tiwai@suse.de>
cc: alsa-devel@alsa-project.org
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Recent changes to the omap_wdt.c removed the dependencies to
the core omap code, but forgot to remove mach/hardware.h.

We cannot include any plat headers with multiplatform
support enabled.

cc: Wim Van Sebroeck <wim@iguana.be>
cc: linux-watchdog@vger.kernel.org
Signed-off-by: Tony Lindgren <tony@atomide.com>

Merge tag 'tags/omap-for-v3.8/devel-prcm-signed' into omap-for-v3.8/cleanup-headers-prepare-multiplatform-v3

omap prcm changes via Paul Walmsley <paul@pwsan.com>:

Some miscellaneous OMAP hwmod changes for 3.8, along with a PRM
change needed for one of the hwmod patches to function.

Basic test logs for this branch on top of Tony's
omap-for-v3.8/clock branch at commit
558a0780 are here:

http://www.pwsan.com/omap/testlogs/hwmod_devel_a_3.8/20121121161522/

However, omap-for-v3.8/clock at 558a0780 does not include some fixes
that are needed for a successful test.  With several reverts,
fixes, and workarounds applied, the following test logs were
obtained:

http://www.pwsan.com/omap/testlogs/TEST_hwmod_devel_a_3.8/20121121162719/

which indicate that the series tests cleanly.

Conflicts:
	arch/arm/mach-omap2/cm33xx.c
	arch/arm/mach-omap2/io.c
	arch/arm/mach-omap2/prm_common.c

Merge tag 'omap-devel-a-for-3.8' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into omap-for-v3.8/devel-pcrm

Some miscellaneous OMAP hwmod changes for 3.8, along with a PRM
change needed for one of the hwmod patches to function.

Basic test logs for this branch on top of Tony's
omap-for-v3.8/clock branch at commit
558a0780 are here:

http://www.pwsan.com/omap/testlogs/hwmod_devel_a_3.8/20121121161522/

However, omap-for-v3.8/clock at 558a0780 does not include some fixes
that are needed for a successful test.  With several reverts,
fixes, and workarounds applied, the following test logs were
obtained:

http://www.pwsan.com/omap/testlogs/TEST_hwmod_devel_a_3.8/20121121162719/

which indicate that the series tests cleanly.

When booting with DT the OF core can fill up the resources provided within
the DT blob.
The current way of handling the DT boot prevents us from removing hwmod data
for platforms only suppose to boot with DT (OMAP5 for example) since we need
to keep the whole hwmod database intact in order to have more resources in
hwmod than in DT (to be able to append the DMA resource from hwmod).

To fix this issue we just examine the OF provided resources:
If we do not have resources we use hwmod to fill them.
If we have resources we check if we already able to recive DMA resource, if
no we only append the DMA resurce from hwmod to the OF provided ones.

In this way we can start removing hwmod data for devices which have their
resources correctly configured in DT without regressions.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Benoît Cousson <b-cousson@ti.com>
[paul@pwsan.com: fixed checkpatch problem; updated to apply]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Add flags parameter for omap_hwmod_count_resources() so users can tell which
type of resources they are interested when counting them in hwmod database.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Benoît Cousson <b-cousson@ti.com>
[paul@pwsan.com: updated to apply]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

OMAP4 has module specific context lost registers which makes it now
possible to have module level context loss count, instead of relying
on the powerdomain level context count.

Add 2 private hwmod api's to update/clear the hwmod/module specific
context lost counters/register.

Update the module specific context_lost_counter and clear the hardware
bits just after enabling the module.

omap_hwmod_get_context_loss_count() now returns the hwmod context loss
count them on platforms where they exist (OMAP4), else fall back on
the pwrdm level counters for older platforms.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: added function kerneldoc, fixed structure kerneldoc,
 rearranged structure to avoid memory waste, marked fns as OMAP4-specific,
 prevent fn entry on non-OMAP4 chips, reduced indentation, merged update
 and clear, merged patches]
[t-kristo@ti.com: added support for arch specific hwmod ops, and changed
 the no context offset indicator to USHRT_MAX]
Signed-off-by: Tero Kristo <t-kristo@ti.com>
[paul@pwsan.com: use NO_CONTEXT_LOSS_BIT flag rather than USHRT_MAX;
 convert unsigned context lost counter to int to match the return type;
 get rid of hwmod_ops in favor of the existing soc_ops mechanism;
 move context loss low-level accesses to the PRM code]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Some PRM functions will need to be called by the hwmod code early in
kernel init.  To handle this, split the PRM initialization code into
early and late phases.  The early init is handled via mach-omap2/io.c,
while the late init is handled by subsys_initcall().

Signed-off-by: Paul Walmsley <paul@pwsan.com>

Replaces several flags bearing the same meaning. There is no need
to set flags due to different omap types here, it can be checked
in appropriate places as well.

Cc: Tony Lindgren <tony@atomide.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: linux-omap@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Acked-by: Nishanth Menon <nm@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Commit 6e740f9a

 (ARM: OMAP: Move omap-pm-noop.c local to mach-omap2)
moved omap-pm-noop to be local to mach-omap2. However, the makefile
entry got placed within ifeq ($(CONFIG_PM),y) which was not the
case earlier.

Fix the issue by moving it out of the ifeq ($(CONFIG_PM),y) in
the makefile as these stubs are needed also when PM is not set.

Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Merge tag 'omap-cleanup-c-for-3.8' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into omap-for-v3.8/clock

Convert the OMAP2+ clock code and data to rely on the common
clock framework for internal bookkeeping and the driver API.

Basic test logs for this branch on top of Tony's cleanup-prcm branch
at commit c9d501e5 are here:

http://www.pwsan.com/omap/testlogs/common_clk_devel_3.8_rebase/20121112192516/

However, cleanup-prcm at c9d501e5 does not include some fixes
that are needed for a successful test.  With several reverts,
fixes, and workarounds applied, the following test logs were
obtained:

http://www.pwsan.com/omap/testlogs/TEST_common_clk_devel_3.8_rebase/20121112192300/

which indicate that the series tests cleanly.

N.B. The common clock data addition patches result in many
checkpatch warnings of the form "WARNING: static const char *
array should probably be static const char * const".  However, it
appears that resolving these would require changes to the CCF
itself.  So the resolution of these warnings is being postponed
until that can be coordinated.

These patches result in a ~55KiB increase in runtime kernel memory
usage when booting omap2plus_defconfig kernels.

Conflicts:
	arch/arm/mach-omap2/clock33xx_data.c
	arch/arm/mach-omap2/clock3xxx_data.c
	arch/arm/mach-omap2/clock44xx_data.c

Clean all #ifdef's added to common clock code.  This code is no longer
needed due to migration to the common clock framework.

Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: clean up new ifdefs added in clockdomain.c]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Drop the now-obsolete OMAP2420/2430 original OMAP clock data.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>

Clean all #ifdef's added to OMAP2 clock code to make it COMMON clk
ready, not that CONFIG_COMMON_CLK is enabled.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: also drop CONFIG_COMMON_CLK tests around APLL recalc_rate
 functions]
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: remove some ifdefs in mach-omap2/io.c]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Remove all of the code that is compiled when CONFIG_COMMON_CLK=n in the
OMAP3+ DPLL handling code.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Mike Turquette <mturquette@ti.com>
Cc: Rajendra Nayak <rnayak@ti.com>

Drop the now-obsolete AM33xx original OMAP clock data.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>

Drop the now-obsolete OMAP3xxx original OMAP clock data.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>

Clean all #ifdef's added to OMAP3 clock code to make it COMMON clk
ready, not that CONFIG_COMMON_CLK is enabled.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: remove some ifdefs in mach-omap2/io.c]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Drop the now-obsolete OMAP44xx original OMAP clock data.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>

Clean all #ifdef's added to OMAP4 clock code to make it COMMON clk
ready, now that CONFIG_COMMON_CLK is enabled.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: remove some ifdefs in mach-omap2/io.c]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Clean all #ifdef's added as part of fixing the clkdm
accesses from hwmod.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Select COMMON_CLK for all OMAP2+ SoCs and switch over to using new
data files for OMAP2/3/4.

The older data files will get removed in subsequent patches.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: move 'select COMMON_CLK' from ARCH_OMAP2PLUS_TYPICAL to
 the per-SoC and per-"arch" Kconfig sections]
[mturquette@ti.com: fixed up #ifdef mismatch in clock.h in previous
  patch which drops that change from this patch]
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

The patch is the output from a python script which converts
from the old OMAP clk format to COMMON clk format using a
JSON parser in between which was developed by Paul Walmsley.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: replace omap2_init_clksel_parent() with
 omap2_clksel_find_parent_index(); reflowed macros; dropped 243x clkdev
 aliases in 242x file; added recalc_rate fn ptrs to APLL clocks;
 fixed some checkpatch warnings]
[mturquette@ti.com: removed deprecated variables from omap24x0_clk_init]
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: fixed boot crash due to missing clock init code; added twl.fck
 alias; fix DPLL rate initialization; fix APLL clocks and virt_prcm_set
 initialization]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

The patch is the output from a python script which converts
from the old OMAP clk format to COMMON clk format using a
JSON parser in between which was developed by Paul Walmsley.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: AM3517/05: dropped bogus hsotgusb "ick" and "fck"
 clkdev aliases; added hsotgusb_fck alias; added emac_ick and emac_fck
 aliases; replace omap2_init_clksel_parent() with
 omap2_clksel_find_parent_index(); reflow macros and parent name
 lists; add clkdm_name argument to DEFINE_STRUCT_CLK_HW_OMAP macros]
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

OMAP2/3/4 clock-tree data is migrated to common-clock framework,
so it is needed to do same for AM33XX device.

Signed-off-by: Vaibhav Hiremath <hvaibhav@ti.com>
Cc: Rajendra Nayak <rnayak@ti.com>
Cc: Paul Walmsley <paul@pwsan.com>
Cc: Benoit Cousson <b-cousson@ti.com>
[paul@pwsan.com: replace omap2_init_clksel_parent() with
 omap2_clksel_find_parent_index(); modified to not use the AM33xx common
 clock data yet; updated patch description; reflowed the macros;
 updated DEFINE_STRUCT_CLK_HW_OMAP usage to include clkdm_name]
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

This patch is output from updated omap hw data autogeneration scripts
mostly contributed by Mike Turquette, with some later fixes from me.
All data is added into a new cclock44xx_data.c file which will be
switched with clock44xx_data.c file in a later patch.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: replace omap2_init_clksel_parent() with
 omap2_clksel_find_parent_index(); reflowed macros; updated
 DEFINE_STRUCT_CLK_HW_OMAP macro to include clkdm_name;
 use macros for clksel mux+gate clocks; many other fixes]
[mturquette@ti.com: converted DPLL outputs to HSDIVIDER macro; trace_clk_div_ck
 has clkdm ops]
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: fixed the omap-gpmc.fck alias per commit a2e5b90b

; fixed
 several checkpatch issues; moved the dpll3xxx.c clockdomain modifications to
 another patch]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Define four convenience macros to be used in the upcoming OMAP2+
common clock framework port.  Although the use of these macros will
make the data somewhat more difficult to read, they significantly reduce
the number of lines in the output patch data.

Most of these were created by Rajendra Nayak and Mike Turquette, as
far as I know.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
[mturquette@ti.com: added DEFINE_CLK_OMAP_HSDIVIDER macro]
Signed-off-by: Mike Turquette <mturquette@ti.com>
Cc: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

It's suspected that some of the clockdomain associations with clocks
can be removed from the clock data.  Drop several of these
associations to save diffstat and improve performance.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: removed most of the changes in this patch; modified patch
 description]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>

The OMAP port to the common clk framework[1] resulted in spurious WARNs
while disable unused clocks.  This is due to _clkdm_clk_hwmod_disable
catching clkdm->usecount's with a value of zero.  Even less desirable it
would not allow the clkdm_clk_disable function pointer to get called due
to an early return of -ERANGE.

This patch adds a check for such a corner case by skipping the WARN and
early return in the event that clkdm->usecount and clk->enable_usecount
are both zero.  Presumably this could only happen during the check for
unused clocks at boot-time.

[1] http://article.gmane.org/gmane.linux.ports.arm.omap/88824



Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: split the hwmod and clock disable cases; modified the
 code to skip the clockdomain handling during the disable-unused-clocks phase;
 added COMMON_CLK ifdef; removed include of clk-private.h at Mike's request]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Without this kernel would crash, since clkdm inside omap_hwmod
is accessed in some of the init functions like, _init_main_clk.

So call init_clkdm before _init_main_clk().

Signed-off-by: Vaibhav Hiremath <hvaibhav@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Platform code can use omap2_clk_enable_init_clocks() to enable a
list of clocks that are needed to be enabled at init.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: added kerneldoc to non-trivial new function]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Platforms can call omap2_init_clk_hw_omap_clocks() to register a clock
using clk_hw_omap. omap2_clk_enable_autoidle_all() and
omap2_clk_disable_autoidle_all() can then be used to run through
all the clocks which support autoidle to enable/disable them.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: added kerneldoc on non-trivial new functions]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

OMAP2420 and OMAP2430 chips each have two on-chip APLLs.  When locked,
one APLL generates a 96 MHz rate; the other, a 54 MHz rate.
Previously we treated these clocks as fixed-rate clocks at the locked
rates, but this isn't quite right.  The locked rate should be returned
when the APLL is locked, and a zero rate should be returned when the
APLL is stopped.  This patch adds the infrastructure that will be used
by the CCF changes.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
Cc: Rajendra Nayak <rnayak@ti.com>

Convert all OMAP2 specific platform files to use COMMON clk
and keep all the changes under the CONFIG_COMMON_CLK macro check
so it does not break any existing code. At a later point switch
to COMMON clk and get rid of all old/legacy code.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: updated to apply]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Convert all OMAP3 specific platform files to use COMMON clk
and keep all the changes under the CONFIG_COMMON_CLK macro check
so it does not break any existing code. At a later point switch
to COMMON clk and get rid of all old/legacy code.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Mike Turquette <mturquette@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

Remove duplicated include.

dpatch engine is used to auto generate this patch.
(https://github.com/weiyj/dpatch

)

Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Tony Lindgren <tony@atomide.com>

The debug-leds driver should not rely on hard coded macro for
the iomem size but use the resource size instead.

Signed-off-by: Benoit Cousson <b-cousson@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

Convert all OMAP4 specific platform files to use COMMON clk
and keep all the changes under the CONFIG_COMMON_CLK macro check
so it does not break any existing code. At a later point switch
to COMMON clk and get rid of all old/legacy code.

This converts all apis which will be called directly from COMMON
clk to take a struct clk_hw parameter, and all the internal platform
apis to take a struct clk_hw_omap parameter.

Changes are based off the original patch from Mike Turquette.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
[paul@pwsan.com: created new omap2_clksel_find_parent_index() rather than
 modifying omap2_init_clksel_parent(); moved clkhwops_iclk_wait to
 clkt_iclk.c to fix OMAP4-only builds; added clk-provider.h include to clock.h
 to try to fix some 3430-builds]
[mturquette@ti.com: squash patch for omap2_clkops_{en,dis}able_clkdm;
 omap2_dflt_clk_is_enabled should not enable clocks]
Signed-off-by: Mike Turquette <mturquette@ti.com>
[paul@pwsan.com: fix compiler warning; update to apply; added kerneldoc on
 non-trivial new functions; added the dpll3xxx clockdomain modifications]
Signed-off-by: Paul Walmsley <paul@pwsan.com>