Hello.
On 03/02/2017 06:42 PM, Harry Morris wrote:
Hello
On 02/03/2017 15:24, Stefan Schmidt wrote:
Hello.
On 03/02/2017 04:08 PM, Marcel Holtmann wrote:
Hi Stefan,
This patchset adds a new device driver, documentation and build
support for
Cascoda's CA8210 IEEE 802.15.4 radio transceiver:
http://www.cascoda.com/products/ca-821x/
v8:
Use module_spi_driver helper
Remove async tx timeout at driver level
Style conformance change
Rework synchronous command handling to remove mutex
v7:
Remove use of cs_change property, always transmit full length
buffers
Implement fully async spi transfers
Combine spi transmission and reception into one function
Use completion objects when waiting for events
Rework a lot of locking
Remove deprecated xmit_sync
Remove use of workqueue for IRQ handling
v6:
Kernel style conformance changes
Additional documentation
More meaningful register structs/macros
Removal of old TODOs
Removal of unnecessary macros
Check for safe removal of test interface
v5:
+Cc: Marcel Holtmann
v4:
Added allowable tx_powers and cca_ed_levels
Changed power units to mbm (tx_power & cca_ed_level)
Harry Morris (3):
ieee802154: Add CA8210 IEEE 802.15.4 device driver
ieee802154: Add device tree documentation for CA8210
ieee802154: Add entry in MAINTAINTERS for CA8210 driver
.../devicetree/bindings/net/ieee802154/ca8210.txt | 28 +
.../devicetree/bindings/vendor-prefixes.txt | 1 +
MAINTAINERS | 9 +
drivers/net/ieee802154/Kconfig | 21 +
drivers/net/ieee802154/Makefile | 1 +
drivers/net/ieee802154/ca8210.c | 3155
++++++++++++++++++++
6 files changed, 3215 insertions(+)
create mode 100644
Documentation/devicetree/bindings/net/ieee802154/ca8210.txt
create mode 100644 drivers/net/ieee802154/ca8210.c
CC drivers/net/ieee802154/ca8210.o
drivers/net/ieee802154/ca8210.c: In function
‘ca8210_register_ext_clock’:
drivers/net/ieee802154/ca8210.c:2698:14: error: implicit declaration
of function ‘clk_register_fixed_rate’
[-Werror=implicit-function-declaratio]
priv->clk = clk_register_fixed_rate(
^~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:2698:12: warning: assignment makes
pointer from integer without a cast [-Wint-conversion]
priv->clk = clk_register_fixed_rate(
^
drivers/net/ieee802154/ca8210.c:2710:8: error: implicit
declaration of
function ‘of_clk_add_provider’
[-Werror=implicit-function-declaration]
ret = of_clk_add_provider(np, of_clk_src_simple_get, priv->clk);
^~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:2710:32: error:
‘of_clk_src_simple_get’ undeclared (first use in this function)
ret = of_clk_add_provider(np, of_clk_src_simple_get, priv->clk);
^~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:2710:32: note: each undeclared
identifier is reported only once for each function it appears in
drivers/net/ieee802154/ca8210.c:2712:3: error: implicit
declaration of
function ‘clk_unregister’ [-Werror=implicit-function-declaration]
clk_unregister(priv->clk);
^~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c: In function
‘ca8210_unregister_ext_clock’:
drivers/net/ieee802154/ca8210.c:2736:2: error: implicit
declaration of
function ‘of_clk_del_provider’
[-Werror=implicit-function-declaration]
of_clk_del_provider(spi->dev.of_node);
^~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:2736:2: warning: ‘return’ with a
value, in function returning void
of_clk_del_provider(spi->dev.of_node);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:2729:13: note: declared here
static void ca8210_unregister_ext_clock(struct spi_device *spi)
^~~~~~~~~~~~~~~~~~~~~~~~~~~
I remember these same issues being raised previously and found the old
exchange:
Hi Stefan, Marcel,
Apologies for the slow reply, I've been snowed under the past week or
so...
On 09/12/2016 11:52, Stefan Schmidt wrote:
On 09/12/16 09:22, Marcel Holtmann wrote:
drivers/net/ieee802154/ca8210.c: In function
‘ca8210_register_ext_clock’:
drivers/net/ieee802154/ca8210.c:3159:14: error: implicit
declaration
of function ‘clk_register_fixed_rate’
[-Werror=implicit-function-declaratio]
priv->clk = clk_register_fixed_rate(
^~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:3159:12: warning: assignment makes
pointer from integer without a cast [-Wint-conversion]
priv->clk = clk_register_fixed_rate(
^
drivers/net/ieee802154/ca8210.c:3171:8: error: implicit declaration
of function ‘of_clk_add_provider’
[-Werror=implicit-function-declaration]
ret = of_clk_add_provider(np, of_clk_src_simple_get, priv->clk);
^~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:3171:32: error:
‘of_clk_src_simple_get’ undeclared (first use in this function)
ret = of_clk_add_provider(np, of_clk_src_simple_get, priv->clk);
^~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:3171:32: note: each undeclared
identifier is reported only once for each function it appears in
drivers/net/ieee802154/ca8210.c:3173:3: error: implicit declaration
of function ‘clk_unregister’
[-Werror=implicit-function-declaration]
clk_unregister(priv->clk);
^~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c: In function
‘ca8210_unregister_ext_clock’:
drivers/net/ieee802154/ca8210.c:3197:2: error: implicit declaration
of function ‘of_clk_del_provider’
[-Werror=implicit-function-declaration]
of_clk_del_provider(spi->dev.of_node);
^~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:3197:2: warning: ‘return’ with a
value, in function returning void
of_clk_del_provider(spi->dev.of_node);
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ieee802154/ca8210.c:3190:13: note: declared here
static void ca8210_unregister_ext_clock(struct spi_device *spi)
^~~~~~~~~~~~~~~~~~~~~~~~~~~
None of these showed up here. A #include <linux/clk-provider.h>
should fix those imho.
I'm also not seeing these, I am including <linux/clk-provider.h>
in my
file, is there any action needed on my part?
I couldn't find any further discussion but again I'm including
clk-provider.h which should cover all this?
I had another look at this and I would suspect that the config
Marcel used does no have CONFIG_COMMON_CLK enabled. Marcel, could
you confirm this?
In that case we need either a select COMMON_CLK in the KConfig entry
of your driver to make sure it really is enabled.
I tried to select COMMON_CLK, but somehow that is not as easy as you
think on an x86 kernel.
Hmm, strange. Any special configuration that you have there? I have it
enabled in a minnow board config as well as a fedora derived config
for my x86_64 laptop. Both configs compile the driver just fine, which
is the reason I did not spot the build breakage before.
Looking into arch/x86/Kconfig a few X86 options actually use "select
COMMON_CLK" already.
However I want things to compile test at least on x86.
Fully agreed. Our drivers should at least be able to compile for it.
If we want to avoid such build breaks for other situation like yours
the driver should really either use "select COMMON_CLK" or "depends on
OMMON_CLK" in its own KConfig entry. That way we enforce that its
enabled when compiling the driver.
I would prefer the select approach as that avoids the user digging
around through there config to find and enable it manually. I would
think that selecting COMMON_CLK directly is safe enough but if you
would like to avoid that we can use depends on.
I was looking into this "select vs depends on" and it seems select is by
far the more widely used approach when it comes to COMMON_CLK. I guess
this doesn't exactly justify it but also I agree having to manually
enable it feels a bit unintuitive. I'd appreciate any further thoughts
though.
It seems Marcel is rather busy right now.
Please go ahead and ad the select COMMON_CLK bit and respin your
patchset. Hopefully we finally can get this merged. :)
regards
Stefan Schmidt
--
To unsubscribe from this list: send the line "unsubscribe linux-wpan" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
