Re: [PATCH v0 01/11] stm class: Introduce an abstraction for System Trace Module devices

Paul Bolle [off-list ref] writes:

On Sat, 2015-03-07 at 13:35 +0200, Alexander Shishkin wrote:

quoted

 Documentation/ABI/testing/configfs-stp-policy    |  44 ++

git am whined about this file when I tried to apply this patch:
    Applying: stm class: Introduce an abstraction for System Trace Module devices
    [...]/.git/rebase-apply/patch:77: new blank line at EOF.

quoted

 Documentation/ABI/testing/sysfs-class-stm        |  14 +
 Documentation/ABI/testing/sysfs-class-stm_source |  11 +
 Documentation/trace/stm.txt                      |  77 +++
 drivers/Kconfig                                  |   2 +
 drivers/Makefile                                 |   1 +
 drivers/stm/Kconfig                              |   8 +
 drivers/stm/Makefile                             |   3 +
 drivers/stm/core.c                               | 839 +++++++++++++++++++++++
 drivers/stm/policy.c                             | 470 +++++++++++++
 drivers/stm/stm.h                                |  77 +++
 include/linux/stm.h                              |  87 +++
 include/uapi/linux/stm.h                         |  47 ++

quoted

--- /dev/null
+++ b/drivers/stm/Kconfig

@@ -0,0 +1,8 @@
+config STM
+	tristate "System Trace Module devices"
+	help
+	  A System Trace Module (STM) is a device exporting data in System
+	  Trace Protocol (STP) format as defined by MIPI STP standards.
+	  Examples of such devices are Intel Trace Hub and Coresight STM.
+
+	  Say Y here to enable System Trace Module device support.

diff --git a/drivers/stm/Makefile b/drivers/stm/Makefile
new file mode 100644
index 0000000000..adec701649
--- /dev/null
+++ b/drivers/stm/Makefile

@@ -0,0 +1,3 @@
+obj-$(CONFIG_STM)	+= stm_core.o
+
+stm_core-y		:= core.o policy.o

I tried to compile this as a module:
    $ make -C ../.. M=$PWD CONFIG_STM=m stm_core.ko
    make: Entering directory `[...]'
      LD [M]  [...]/drivers/stm/stm_core.o
    [...]/drivers/stm/policy.o: In function `stp_configfs_init':
    policy.c:(.text+0x5f0): multiple definition of `init_module'
    [...]/drivers/stm/core.o:core.c:(.init.text+0x0): first defined here
    make[1]: *** [[...]/drivers/stm/stm_core.o] Error 1
    make: *** [stm_core.ko] Error 2
    make: Leaving directory `[...]'

I think that's because
    postcore_initcall(stm_core_init);

in core.c becomes
    module_init(stm_core_init);

if this driver is compiled as a module. And that will clash with
    module_init(stp_configfs_init);

in policy.c. Am I missing something obvious or should STM not be a
tristate symbol?

My mistake, I think I fancied the policy to be a separate module in the
beginning. But I don't see why STM shouldn't be a tristate after the
above is fixed.

Thanks,
--
Alex