Add initial configfs support for IIO
Subsystems (4)
v10 cover Add initial configfs support for IIO
This patchset introduces IIO software triggers, offers a way of configuring them via configfs and adds the IIO hrtimer based interrupt source to be used with software triggers.
The arhitecture is now split in 3 parts, to remove all IIO trigger specific parts from IIO configfs core:
(1) IIO configfs - creates the root of the IIO configfs subsys.
(2) IIO software triggers - software trigger implementation, dynamically creating /config/iio/triggers group.
(3) IIO hrtimer trigger - is the first interrupt source for software triggers (with syfs to follow). Each trigger type can implement its own set of attributes.
Lockdep seems to be happy with the locking in configfs patch.
Changes since v9:
* introduce configfs_{un}register_default_group boilerplate code, that does group allocation and registration
* remove configfs_alloc_group/configfs_free_group as we can directly code them in.
Changes since v8:
* patch 1/5
* dynamically allocate groups as suggested by Christoph
* add kernel-doc documentation to the newly introduced functions
* patch 2/5
* better isolation, here we only create the root of the IIO configfs subys, and let the other modules (e.g software triggers) to create their group.
* patch 3/5
* dynamically create triggers/ group to better isolate the code
* fixes 0-day build warnings and compilation errors.
* patch 4/5
* none
* patch 5/5
* none
Changes since v7:
* patch 1/5
* do not add dentry parameter for create_default_group, we can retrieve it from group.
* call configfs_detach_prep before configfs_detach_group in configfs_unregister_group.
* call link_group/unlink_group under su_mutex lock
* take configfs_dirent_lock when calling configfs_dir_set_ready
* patch 2/5
* export iio_triggers_group, this will be later used by sw triggers
* patch 3/5
* make iio_sw_trigger part of industrialio core
* fix compilation problem reported by Vladimir and Lars. Things didn't get well when having IIO_SW_TRIGGER enabled and IIO_CONFIGFS disabled.
* use format string in config_item_set_name
* make iio_unregister_sw_trigger_type return type void.
* patch 4/5
* none
* patch 5/5
* include a small trigger type sample example.
Changes since v6:
* implemented Lars-Peter's idea (https://lkml.org/lkml/2015/5/13/302)
to switch from /config/iio/triggers/hrtimer-instance1 to /config/iio/triggers/instance1.
Changes since v5: (after Lars comments)
* the most important change is that we moved sampling_frequency attribute from configfs to trigger's directory in /sys.
* couple of const added to strings
* documentation to public API in sw_trigger.h
* replace pr_err with WARN_ONCE in trigger_make_group to avoid spamming kernel log, but without leaving user clueless in case of errors.
* we still need to decide if we get a real gain by adding min/max limits for sampling frequency in /config dir. Anyhow, this can be done in a later patch.
* fix race in hrtimer_remove
Changes since v4:
* patch 1/4
- fixed "new line" nit in industrialio-sw-trigger.c
- added license header in sw_trigger.h o
* patch 2/4
- none
* patch 3/4
- none
* patch 4/4
- removed "Further work" chapter in iio_configfs.txt
- added configfs-iio file in Documentation/ABI/testing
Daniel Baluta (5):
configfs: Allow dynamic group creation
iio: core: Introduce IIO configfs support
iio: core: Introduce IIO software triggers
iio: trigger: Introduce IIO hrtimer based trigger
iio: Documentation: Add IIO configfs documentation
Documentation/ABI/testing/configfs-iio | 21 ++++ Documentation/iio/iio_configfs.txt | 93 ++++++++++++++++ drivers/iio/Kconfig | 16 +++ drivers/iio/Makefile | 2 + drivers/iio/industrialio-configfs.c | 50 +++++++++ drivers/iio/industrialio-sw-trigger.c | 183 +++++++++++++++++++++++++++++++ drivers/iio/trigger/Kconfig | 10 ++ drivers/iio/trigger/Makefile | 2 + drivers/iio/trigger/iio-trig-hrtimer.c | 193 +++++++++++++++++++++++++++++++++ fs/configfs/dir.c | 110 +++++++++++++++++++ include/linux/configfs.h | 10 ++ include/linux/iio/sw_trigger.h | 71 ++++++++++++ 12 files changed, 761 insertions(+) create mode 100644 Documentation/ABI/testing/configfs-iio create mode 100644 Documentation/iio/iio_configfs.txt create mode 100644 drivers/iio/industrialio-configfs.c create mode 100644 drivers/iio/industrialio-sw-trigger.c create mode 100644 drivers/iio/trigger/iio-trig-hrtimer.c create mode 100644 include/linux/iio/sw_trigger.h
Signature · 1 line
1.9.1
PATCH v10 1/5 configfs: Allow dynamic group creation
We don't want to hardcode default groups at subsystem creation time. We export:
* configfs_register_group
* configfs_unregister_group
to allow drivers to programatically create/destroy groups later, after module init time.
This is needed for IIO configfs support.
--- fs/configfs/dir.c | 110 +++++++++++++++++++++++++++++++++++++++++++++++ include/linux/configfs.h | 10 +++++ 2 files changed, 120 insertions(+)
fs/configfs/dir.c +110 -0
@@ -1636,6 +1636,116 @@ const struct file_operations configfs_dir_operations = { .iterate = configfs_readdir, }; +/** +* configfs_register_group - creates a parent-child relation between two groups +* @parent_group: parent group +* @group: child group +* +* link groups, creates dentry for the child and attaches it to the +* parent dentry. +* +* Return: 0 on success, negative errno code on error +*/ +int configfs_register_group (struct config_group *parent_group, +struct config_group *group) +{ +struct configfs_subsystem *subsys = parent_group->cg_subsys; +struct dentry *parent; +int ret; + +mutex_lock (&subsys->su_mutex); +link_group (parent_group, group); +mutex_unlock (&subsys->su_mutex); + + parent = parent_group->cg_item.ci_dentry; + +mutex_lock_nested (&d_inode (parent)->i_mutex,I_MUTEX_PARENT ); + ret =create_default_group (parent_group, group); +if (!ret) { +spin_lock (&configfs_dirent_lock); +configfs_dir_set_ready (group->cg_item.ci_dentry->d_fsdata); +spin_unlock (&configfs_dirent_lock); + } +mutex_unlock (&d_inode (parent)->i_mutex); +return ret; +} +EXPORT_SYMBOL (configfs_register_group); + +/** +* configfs_unregister_group() - unregisters a child group from its parent +* @group: parent group to be unregistered +* +* Undoes configfs_register_group() +*/ +void configfs_unregister_group (struct config_group *group) +{ +struct configfs_subsystem *subsys = group->cg_subsys; +struct dentry *dentry = group->cg_item.ci_dentry; +struct dentry *parent = group->cg_item.ci_parent->ci_dentry; + +mutex_lock_nested (&d_inode (parent)->i_mutex,I_MUTEX_PARENT ); +spin_lock (&configfs_dirent_lock); +configfs_detach_prep (dentry,NULL ); +spin_unlock (&configfs_dirent_lock); + +configfs_detach_group (&group->cg_item); +d_inode (dentry)->i_flags |=S_DEAD ; +dont_mount (dentry); +d_delete (dentry); +mutex_unlock (&d_inode (parent)->i_mutex); + +dput (dentry); + +mutex_lock (&subsys->su_mutex); +unlink_group (group); +mutex_unlock (&subsys->su_mutex); +} +EXPORT_SYMBOL (configfs_unregister_group); + +/** +* configfs_register_default_group() - allocates and registers a child group +* @parent_group: parent group +* @name: child group name +* @item_type: child item type description +* +* boilerplate to allocate and register a child group with its parent. We need +* kzalloc'ed memory because child's default_group is initially empty. +* +* Return: allocated config group or ERR_PTR() on error +*/ +struct config_group * +configfs_register_default_group (struct config_group *parent_group, +const char *name, +struct config_item_type *item_type) +{ +int ret; +struct config_group *group; + + group =kzalloc (sizeof (*group),GFP_KERNEL ); +if (!group) +return ERR_PTR (-ENOMEM ); +config_group_init_type_name (group, name, item_type); + + ret =configfs_register_group (parent_group, group); +if (ret) { +kfree (group); +return ERR_PTR (ret); + } +return group; +} +EXPORT_SYMBOL (configfs_register_default_group); + +/** +* configfs_unregister_default_group() - unregisters and frees a child group +* @group: the group to act on +*/ +void configfs_unregister_default_group (struct config_group *group) +{ +configfs_unregister_group (group); +kfree (group); +} +EXPORT_SYMBOL (configfs_unregister_default_group); +int configfs_register_subsystem (struct configfs_subsystem *subsys) {int err;
include/linux/configfs.h +10 -0
@@ -252,6 +252,16 @@ static inline struct configfs_subsystem *to_configfs_subsystem(struct config_groint configfs_register_subsystem (struct configfs_subsystem *subsys);void configfs_unregister_subsystem (struct configfs_subsystem *subsys); +int configfs_register_group (struct config_group *parent_group, +struct config_group *group); +void configfs_unregister_group (struct config_group *group); + +struct config_group * +configfs_register_default_group (struct config_group *parent_group, +const char *name, +struct config_item_type *item_type); +void configfs_unregister_default_group (struct config_group *group); +/* These functions can sleep and can alloc with GFP_KERNEL */ /* WARNING: These cannot be called underneath configfs callbacks!! */ int configfs_depend_item (struct configfs_subsystem *subsys,struct config_item *target);
Signature · 1 line
1.9.1
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Is this going into 4.4 through the iio tree? If not is there any chance to get it in through some other tree? While we're not past the merge window is trivial new functionality that doesn't change code, and I'd like to move existing configfs users over to it ASAP, so getting it into a baseline tree ASAP would be immensely helpful.
Oh, and:
(I though we already had that, but oh well)
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 5 lines · Christoph Hellwig
Is this going into 4.4 through the iio tree? If not is there any chance to get it in through some other tree? While we're not past the merge window is trivial new functionality that doesn't change code, and I'd like to move existing configfs users over to it ASAP, so getting it into a baseline tree ASAP would be immensely helpful.
I think Jonathan could take this via linux-iio around Sunday.
Could anyone do it faster :)? Andrew?
Quoted 4 lines
Oh, and:
Reviewed-by Christoph Hellwig <[email protected]>
I didn't add your Reviewed-by tag because I removed configfs_alloc_group/configfs_free_group functions.
Thanks Christoph!
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 9 lines · Daniel Baluta
Quoted 5 lines · Christoph Hellwig
Is this going into 4.4 through the iio tree? If not is there any chance to get it in through some other tree? While we're not past the merge window is trivial new functionality that doesn't change code, and I'd like to move existing configfs users over to it ASAP, so getting it into a baseline tree ASAP would be immensely helpful.
I think Jonathan could take this via linux-iio around Sunday.
Since the IIO tree goes through the staging tree, which is already closed for 4.4, we won't be able to get it into 4.4 through the IIO tree. The whole configfs support for IIO will have to wait for 4.5.
Quoted 1 line
Could anyone do it faster :)? Andrew?
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 12 lines · Lars-Peter Clausen
Quoted 2 lines · Daniel Baluta
Quoted 1 line · Christoph Hellwig
Is this going into 4.4 through the iio tree? If not is there any
chance
Quoted 1 line
Quoted 1 line
to get it in through some other tree? While we're not past the
merge
Quoted 1 line
Quoted 1 line
window is trivial new functionality that doesn't change code, and
I'd like to
Quoted 1 line
Quoted 1 line
move existing configfs users over to it ASAP, so getting it into a
baseline
Quoted 2 lines
Quoted 1 line
tree ASAP would be immensely helpful.
Quoted 8 lines
Quoted 2 lines
I think Jonathan could take this via linux-iio around Sunday.
Since the IIO tree goes through the staging tree, which is already closed
for 4.4, we won't be able to get it into 4.4 through the IIO tree. The whole
configfs support for IIO will have to wait for 4.5.
Yup. I'd have no objection to a direct request to Linus to take this as a one off.
Or we can do an immutable branch and pull it into all relevant subtrees so all
users can hit mainline all in 4.5 without waiting for a cycle. That tends to work well for bits like this.
Something odd going on with my email as I missed this version entirely. Will investigate. Reading up on the archive.
Jonathan
Quoted 6 lines
Quoted 2 lines
Could anyone do it faster :)? Andrew?
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Signature · 1 line
Sent from my Android device with K-9 Mail. Please excuse my brevity.
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 1 line · Jonathan Cameron
Yup. I'd have no objection to a direct request to Linus to take this as a one off.
I'd appreciate if you could give it a try.
Quoted 2 lines
Or we can do an immutable branch and pull it into all relevant subtrees so all
users can hit mainline all in 4.5 without waiting for a cycle. That tends to work well for bits like this.
That would be the second best option.
Jonathan Cameron
maintainerRe: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 4 lines · Christoph Hellwig
Quoted 1 line · Jonathan Cameron
Yup. I'd have no objection to a direct request to Linus to take this as a one off.
I'd appreciate if you could give it a try.
Unfortunately this would fall out of scope for me to send. The bit we care about is definitely a file system code change in configfs. Whilst the first user will probably be IIO, I can't justify to Linus why it is necessary to put this in place ASAP rather than via normal paths (which in my case is via Greg)
If Joel was active I'd suggest it should come from him if we want to try this, as he isn't (as a side note, we should probably update the MAINTAINERS entry to mark it as unmaintained I guess), who ends up being maintainer of last resort for file systems?
<quick git log browse later>
I'd say you need to make a case to Al Viro or Andrew Morton.
Quoted 5 lines
Quoted 3 lines
Or we can do an immutable branch and pull it into all relevant subtrees so all
users can hit mainline all in 4.5 without waiting for a cycle. That tends to work well for bits like this.That would be the second best option.
I can do this easily enough and it's probably my preferred option.
Jonathan
Quoted 5 lines
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
In reply to Daniel BalutaAndrew,
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
Thanks,
Christoph
Quoted 160 lines · Daniel Baluta
We don't want to hardcode default groups at subsystem creation time. We export:
*configfs_register_group
*configfs_unregister_group
to allow drivers to programatically create/destroy groups later, after module init time.This is needed for IIO configfs support.
Suggested-by Lars-Peter Clausen <[email protected]>Signed-off-by Daniel Baluta <[email protected]>--- fs/configfs/dir.c | 110 +++++++++++++++++++++++++++++++++++++++++++++++ include/linux/configfs.h | 10 +++++ 2 files changed, 120 insertions(+)fs/configfs/dir.c +110
-0@@ -1636,6 +1636,116 @@ const struct file_operations configfs_dir_operations = { .iterate = configfs_readdir, }; +/** +* configfs_register_group - creates a parent-child relation between two groups +* @parent_group: parent group +* @group: child group +* +* link groups, creates dentry for the child and attaches it to the +* parent dentry. +* +* Return: 0 on success, negative errno code on error +*/ +int configfs_register_group (struct config_group *parent_group, +struct config_group *group) +{ +struct configfs_subsystem *subsys = parent_group->cg_subsys; +struct dentry *parent; +int ret; + +mutex_lock (&subsys->su_mutex); +link_group (parent_group, group); +mutex_unlock (&subsys->su_mutex); + + parent = parent_group->cg_item.ci_dentry; + +mutex_lock_nested (&d_inode (parent)->i_mutex,I_MUTEX_PARENT ); + ret =create_default_group (parent_group, group); +if (!ret) { +spin_lock (&configfs_dirent_lock); +configfs_dir_set_ready (group->cg_item.ci_dentry->d_fsdata); +spin_unlock (&configfs_dirent_lock); + } +mutex_unlock (&d_inode (parent)->i_mutex); +return ret; +} +EXPORT_SYMBOL (configfs_register_group); + +/** +* configfs_unregister_group() - unregisters a child group from its parent +* @group: parent group to be unregistered +* +* Undoes configfs_register_group() +*/ +void configfs_unregister_group (struct config_group *group) +{ +struct configfs_subsystem *subsys = group->cg_subsys; +struct dentry *dentry = group->cg_item.ci_dentry; +struct dentry *parent = group->cg_item.ci_parent->ci_dentry; + +mutex_lock_nested (&d_inode (parent)->i_mutex,I_MUTEX_PARENT ); +spin_lock (&configfs_dirent_lock); +configfs_detach_prep (dentry,NULL ); +spin_unlock (&configfs_dirent_lock); + +configfs_detach_group (&group->cg_item); +d_inode (dentry)->i_flags |=S_DEAD ; +dont_mount (dentry); +d_delete (dentry); +mutex_unlock (&d_inode (parent)->i_mutex); + +dput (dentry); + +mutex_lock (&subsys->su_mutex); +unlink_group (group); +mutex_unlock (&subsys->su_mutex); +} +EXPORT_SYMBOL (configfs_unregister_group); + +/** +* configfs_register_default_group() - allocates and registers a child group +* @parent_group: parent group +* @name: child group name +* @item_type: child item type description +* +* boilerplate to allocate and register a child group with its parent. We need +* kzalloc'ed memory because child's default_group is initially empty. +* +* Return: allocated config group or ERR_PTR() on error +*/ +struct config_group * +configfs_register_default_group (struct config_group *parent_group, +const char *name, +struct config_item_type *item_type) +{ +int ret; +struct config_group *group; + + group =kzalloc (sizeof (*group),GFP_KERNEL ); +if (!group) +return ERR_PTR (-ENOMEM ); +config_group_init_type_name (group, name, item_type); + + ret =configfs_register_group (parent_group, group); +if (ret) { +kfree (group); +return ERR_PTR (ret); + } +return group; +} +EXPORT_SYMBOL (configfs_register_default_group); + +/** +* configfs_unregister_default_group() - unregisters and frees a child group +* @group: the group to act on +*/ +void configfs_unregister_default_group (struct config_group *group) +{ +configfs_unregister_group (group); +kfree (group); +} +EXPORT_SYMBOL (configfs_unregister_default_group); +int configfs_register_subsystem (struct configfs_subsystem *subsys) {int err;include/linux/configfs.h +10
-0@@ -252,6 +252,16 @@ static inline struct configfs_subsystem *to_configfs_subsystem(struct config_groint configfs_register_subsystem (struct configfs_subsystem *subsys);void configfs_unregister_subsystem (struct configfs_subsystem *subsys); +int configfs_register_group (struct config_group *parent_group, +struct config_group *group); +void configfs_unregister_group (struct config_group *group); + +struct config_group * +configfs_register_default_group (struct config_group *parent_group, +const char *name, +struct config_item_type *item_type); +void configfs_unregister_default_group (struct config_group *group); +/* These functions can sleep and can alloc with GFP_KERNEL */ /* WARNING: These cannot be called underneath configfs callbacks!! */ int configfs_depend_item (struct configfs_subsystem *subsys,struct config_item *target);Signature · 1 line
1.9.1
---end quoted text---
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 4 lines · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for 4.5-rc1, or something else?
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 8 lines · Andrew Morton
Quoted 4 lines · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
Quoted 5 lines
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for 4.5-rc1, or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg KH for the 4.5 merge window.
Jonathan
Quoted 5 lines
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Signature · 1 line
Sent from my Android device with K-9 Mail. Please excuse my brevity.
Jonathan Cameron
maintainerRe: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 11 lines · Jonathan Cameron
Quoted 8 lines · Andrew Morton
Quoted 4 lines · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
Quoted 8 lines
Quoted 5 lines
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for 4.5-rc1, or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg KH for the 4.5 merge window.
Hi Andrew,
Just taken a quick look at your mmotm list and see this ended up in the mainline later group (fair enough given the timing!).
As such shall we fall back to plan b) a special git branch pulled into the trees of anyone who cares?
I'll base such a tree on some obvious point in Linus' tree (either 4.4 or 4.5-rc1) That way I can get the IIO stuff queued up asap and we can build on that going forward during this cycle.
Thanks,
Jonathan
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 22 lines · On Sun, 15 Nov 2015 10:39:08 +0000 Jonathan Cameron <[email protected]> wrote:
Quoted 11 lines · Jonathan Cameron
Quoted 8 lines · Andrew Morton
Quoted 4 lines · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
Quoted 8 lines
Quoted 5 lines
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for 4.5-rc1, or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg KH for the 4.5 merge window.
Hi Andrew,
Quoted 8 lines
Just taken a quick look at your mmotm list and see this ended up in the mainline later group (fair enough given the timing!).
As such shall we fall back to plan b) a special git branch pulled into the trees of anyone who cares?I'll base such a tree on some obvious point in Linus' tree (either 4.4 or 4.5-rc1) That way I can get the IIO stuff queued up asap and we can build on that going forward during this cycle.
I plan to send configfs-allow-dynamic-group-creation.patch to Linus this week. I'll retain
iio-core-introduce-iio-configfs-support.patch iio-core-introduce-iio-software-triggers.patch iio-core-introduce-iio-software-triggers-fix.patch iio-trigger-introduce-iio-hrtimer-based-trigger.patch iio-documentation-add-iio-configfs-documentation.patch
with a view to dropping them once I see them turn up in linux-next.
Re: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 46 lines · Andrew Morton
Quoted 4 lines · On Sun, 15 Nov 2015 10:39:08 +0000 Jonathan Cameron <[email protected]> wrote:
Quoted 3 lines · Jonathan Cameron
On 10 November 2015 21:12:37 GMT+00:00, Andrew Morton
Quoted 4 lines · <[email protected]> wrote:
Quoted 4 lines
Quoted 4 lines
Quoted 1 line · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to
Linus
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the
next
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 1 line
merge window.
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for
4.5-rc1,
Quoted 2 lines
Quoted 2 lines
Quoted 1 line
or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg
KH for the 4.5
Quoted 5 lines
Quoted 2 lines
merge window.
Hi Andrew,
Just taken a quick look at your mmotm list and see this ended up in
the
Quoted 2 lines
mainline later group (fair enough given the timing!). As such shall we fall back to plan b) a special git branch pulled
into the trees
Quoted 3 lines
of anyone who cares?
I'll base such a tree on some obvious point in Linus' tree (either
4.4 or 4.5-rc1)
Quoted 1 line
That way I can get the IIO stuff queued up asap and we can build on
that going
Quoted 1 line
forward during this cycle.
I plan to send configfs-allow-dynamic-group-creation.patch to Linus this week. I'll retain
Quoted 7 lines
iio-core-introduce-iio-configfs-support.patch iio-core-introduce-iio-software-triggers.patch iio-core-introduce-iio-software-triggers-fix.patch iio-trigger-introduce-iio-hrtimer-based-trigger.patch iio-documentation-add-iio-configfs-documentation.patch
with a view to dropping them once I see them turn up in linux-next.
That's great. Thanks.
Jonathan
Signature · 1 line
Sent from my Android device with K-9 Mail. Please excuse my brevity.
Jonathan Cameron
maintainerRe: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 55 lines · Jonathan Cameron
Quoted 52 lines · Andrew Morton
Quoted 4 lines · On Sun, 15 Nov 2015 10:39:08 +0000 Jonathan Cameron <[email protected]> wrote:
Quoted 3 lines · Jonathan Cameron
On 10 November 2015 21:12:37 GMT+00:00, Andrew Morton
Quoted 4 lines · <[email protected]> wrote:
Quoted 4 lines
Quoted 4 lines
Quoted 1 line · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to
Linus
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the
next
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 1 line
merge window.
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for
4.5-rc1,
Quoted 2 lines
Quoted 2 lines
Quoted 1 line
or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg
KH for the 4.5
Quoted 5 lines
Quoted 2 lines
merge window.
Hi Andrew,
Just taken a quick look at your mmotm list and see this ended up in
the
Quoted 2 lines
mainline later group (fair enough given the timing!). As such shall we fall back to plan b) a special git branch pulled
into the trees
Quoted 3 lines
of anyone who cares?
I'll base such a tree on some obvious point in Linus' tree (either
4.4 or 4.5-rc1)
Quoted 1 line
That way I can get the IIO stuff queued up asap and we can build on
that going
Quoted 1 line
forward during this cycle.
I plan to send configfs-allow-dynamic-group-creation.patch to Linus this week. I'll retain
iio-core-introduce-iio-configfs-support.patch iio-core-introduce-iio-software-triggers.patch iio-core-introduce-iio-software-triggers-fix.patch iio-trigger-introduce-iio-hrtimer-based-trigger.patch iio-documentation-add-iio-configfs-documentation.patch
Quoted 3 lines
Quoted 1 line
with a view to dropping them once I see them turn up in linux-next.
That's great. Thanks.
I've now applied the rest of the series to my local togreg branch and pushed out as testing for the autobuilders to play with them.
This branch will get rebased once Greg has picked up the previous PULL request.
Thanks,
Jonathan
Quoted 3 lines
Jonathan
Jonathan Cameron
maintainerRe: [PATCH v10 1/5] configfs: Allow dynamic group creation
Quoted 56 lines · Jonathan Cameron
Quoted 55 lines · Jonathan Cameron
Quoted 52 lines · Andrew Morton
Quoted 4 lines · On Sun, 15 Nov 2015 10:39:08 +0000 Jonathan Cameron <[email protected]> wrote:
Quoted 3 lines · Jonathan Cameron
On 10 November 2015 21:12:37 GMT+00:00, Andrew Morton
Quoted 4 lines · <[email protected]> wrote:
Quoted 4 lines
Quoted 4 lines
Quoted 1 line · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to
Linus
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the
next
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 1 line
merge window.
It's super late, but the configfs changes are obviously safe to existing code.
What about the IIO changes? Will someone be merging them for
4.5-rc1,
Quoted 2 lines
Quoted 2 lines
Quoted 1 line
or something else?
Yes. I'll take the IIO bits and ultimately they'll go through Greg
KH for the 4.5
Quoted 5 lines
Quoted 2 lines
merge window.
Hi Andrew,
Just taken a quick look at your mmotm list and see this ended up in
the
Quoted 2 lines
mainline later group (fair enough given the timing!). As such shall we fall back to plan b) a special git branch pulled
into the trees
Quoted 3 lines
of anyone who cares?
I'll base such a tree on some obvious point in Linus' tree (either
4.4 or 4.5-rc1)
Quoted 1 line
That way I can get the IIO stuff queued up asap and we can build on
that going
Quoted 1 line
forward during this cycle.
I plan to send configfs-allow-dynamic-group-creation.patch to Linus this week. I'll retain
iio-core-introduce-iio-configfs-support.patch iio-core-introduce-iio-software-triggers.patch iio-core-introduce-iio-software-triggers-fix.patch iio-trigger-introduce-iio-hrtimer-based-trigger.patch iio-documentation-add-iio-configfs-documentation.patch
Quoted 7 lines
Quoted 3 lines
Quoted 1 line
with a view to dropping them once I see them turn up in linux-next.
That's great. Thanks.
I've now applied the rest of the series to my local togreg branch and pushed out as testing for the autobuilders to play with them.
This branch will get rebased once Greg has picked up the previous PULL request.
Now rebased - I also made a tiny change to take the iio_configfs_subsys structure static in response to a sparse warning - shout if I've done anything silly.
Quoted 13 lines
Thanks,
Jonathan
Quoted 3 lines
Jonathan
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Joel Becker
Acked-byRe: [PATCH v10 1/5] configfs: Allow dynamic group creation
In reply to Andrew MortonQuoted 9 lines · Andrew Morton
Quoted 4 lines · On Tue, 10 Nov 2015 07:51:26 +0100 Christoph Hellwig <[email protected]> wrote:
is this simple addition something you could still send on to Linus for this merge window? I would make my life easier to have it in so I could start using it in patches for various trees in the next merge window.
It's super late, but the configfs changes are obviously safe to existing code.
Loving this change, even if this is merely a drive-by:
Quoted 4 lines
What about the IIO changes? Will someone be merging them for 4.5-rc1, or something else?
Signature · 5 lines
"Copy from one, it's plagiarism; copy from two, it's research."
- Wilson Mizner
http://www.jlbec.org/
[email protected]PATCH v10 2/5 iio: core: Introduce IIO configfs support
In reply to Daniel BalutaThis patch creates the IIO configfs root group. The group will appear under <mount-point>/iio/, usually /config/iio.
We introduce configfs support in IIO in order to be able to easily create IIO objects from userspace. The first supported IIO objects are triggers introduced with next patches.
--- drivers/iio/Kconfig | 8 ++++++ drivers/iio/Makefile | 1 + drivers/iio/industrialio-configfs.c | 50 +++++++++++++++++++++++++++++++++++++ 3 files changed, 59 insertions(+) create mode 100644 drivers/iio/industrialio-configfs.c
drivers/iio/Kconfig +8 -0
@@ -22,6 +22,14 @@ if IIO_BUFFER source "drivers/iio/buffer/Kconfig" endif # IIO_BUFFER +config IIO_CONFIGFS + tristate "Enable IIO configuration via configfs" + select CONFIGFS_FS + help + This allows configuring various IIO bits through configfs + (e.g. software triggers). For more info see + Documentation/iio/iio_configfs.txt. + config IIO_TRIGGER bool "Enable triggered sampling support" help
drivers/iio/Makefile +1 -0
@@ -7,6 +7,7 @@ industrialio-y := industrialio-core.o industrialio-event.o inkern.o industrialio-$(CONFIG_IIO_BUFFER) += industrialio-buffer.o industrialio-$(CONFIG_IIO_TRIGGER) += industrialio-trigger.o +obj-$(CONFIG_IIO_CONFIGFS) += industrialio-configfs.o obj-$(CONFIG_IIO_TRIGGERED_EVENT) += industrialio-triggered-event.o obj-y += accel/
drivers/iio/industrialio-configfs.c new file +50 -0
50 lines not on this page · raw message
Signature · 1 line
1.9.1
Re: [PATCH v10 2/5] iio: core: Introduce IIO configfs support
Quoted 110 lines · Daniel Baluta
This patch creates the IIO configfs root group. The group will appear under <mount-point>/iio/, usually /config/iio.
We introduce configfs support in IIO in order to be able to easily create IIO objects from userspace. The first supported IIO objects are triggers introduced with next patches.
Signed-off-by Daniel Baluta <[email protected]>--- drivers/iio/Kconfig | 8 ++++++ drivers/iio/Makefile | 1 + drivers/iio/industrialio-configfs.c | 50 +++++++++++++++++++++++++++++++++++++ 3 files changed, 59 insertions(+) create mode 100644 drivers/iio/industrialio-configfs.cdrivers/iio/Kconfig +8
-0@@ -22,6 +22,14 @@ if IIO_BUFFER source "drivers/iio/buffer/Kconfig" endif # IIO_BUFFER +config IIO_CONFIGFS + tristate "Enable IIO configuration via configfs" + select CONFIGFS_FS + help + This allows configuring various IIO bits through configfs + (e.g. software triggers). For more info see + Documentation/iio/iio_configfs.txt. + config IIO_TRIGGER bool "Enable triggered sampling support" helpdrivers/iio/Makefile +1
-0@@ -7,6 +7,7 @@ industrialio-y := industrialio-core.o industrialio-event.o inkern.o industrialio-$(CONFIG_IIO_BUFFER) += industrialio-buffer.o industrialio-$(CONFIG_IIO_TRIGGER) += industrialio-trigger.o +obj-$(CONFIG_IIO_CONFIGFS) += industrialio-configfs.o obj-$(CONFIG_IIO_TRIGGERED_EVENT) += industrialio-triggered-event.o obj-y += accel/drivers/iio/industrialio-configfs.c new file +50
-1@@ -0,0 +1,50 @@ +/* +* Industrial I/O configfs bits +* +* Copyright (c) 2015 Intel Corporation +* +* This program is free software; you can redistribute it and/or modify it +* under the terms of the GNU General Public License version 2 as published by +* the Free Software Foundation. +*/ + +#include <linux/configfs.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/kmod.h> +#include <linux/slab.h> + +#include <linux/iio/iio.h> + +static struct config_item_type iio_root_group_type = { + .ct_owner =THIS_MODULE , +}; + +struct configfs_subsystem iio_configfs_subsys = { + .su_group = { + .cg_item = { + .ci_namebuf ="iio" , + .ci_type = &iio_root_group_type, + }, + }, + .su_mutex =__MUTEX_INITIALIZER (iio_configfs_subsys.su_mutex), +}; +EXPORT_SYMBOL (iio_configfs_subsys); + +static int __init iio_configfs_init (void ) +{ +config_group_init (&iio_configfs_subsys.su_group); + +return configfs_register_subsystem (&iio_configfs_subsys); +} +module_init (iio_configfs_init); + +static void __exit iio_configfs_exit(void) +{ +configfs_unregister_subsystem (&iio_configfs_subsys); +} +module_exit (iio_configfs_exit); + +MODULE_AUTHOR ("Daniel Baluta <[email protected]>" ); +MODULE_DESCRIPTION ("Industrial I/O configfs support" ); +MODULE_LICENSE ("GPL v2" ); --1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
PATCH v10 3/5 iio: core: Introduce IIO software triggers
In reply to Daniel BalutaA software trigger associates an IIO device trigger with a software interrupt source (e.g: timer, sysfs). This patch adds the generic infrastructure for handling software triggers.
Software interrupts sources are kept in a iio_trigger_types_list and registered separately when the associated kernel module is loaded.
Software triggers can be created directly from drivers or from user space via configfs interface.
To sum up, this dynamically creates "triggers" group to be found under /config/iio/triggers and offers the possibility of dynamically creating trigger types groups. The first supported trigger type is "hrtimer" found under /config/iio/triggers/hrtimer.
--- drivers/iio/Kconfig | 8 ++ drivers/iio/Makefile | 1 + drivers/iio/industrialio-sw-trigger.c | 183 ++++++++++++++++++++++++++++++++++ include/linux/iio/sw_trigger.h | 71 +++++++++++++ 4 files changed, 263 insertions(+) create mode 100644 drivers/iio/industrialio-sw-trigger.c create mode 100644 include/linux/iio/sw_trigger.h
drivers/iio/Kconfig +8 -0
@@ -46,6 +46,14 @@ config IIO_CONSUMERS_PER_TRIGGER This value controls the maximum number of consumers that a given trigger may handle. Default is 2. +config IIO_SW_TRIGGER + tristate "Enable software triggers support" + select IIO_CONFIGFS + help + Provides IIO core support for software triggers. A software + trigger can be created via configfs or directly by a driver + using the API provided. + config IIO_TRIGGERED_EVENT tristate select IIO_TRIGGER
drivers/iio/Makefile +1 -0
@@ -8,6 +8,7 @@ industrialio-$(CONFIG_IIO_BUFFER) += industrialio-buffer.o industrialio-$(CONFIG_IIO_TRIGGER) += industrialio-trigger.o obj-$(CONFIG_IIO_CONFIGFS) += industrialio-configfs.o +obj-$(CONFIG_IIO_SW_TRIGGER) += industrialio-sw-trigger.o obj-$(CONFIG_IIO_TRIGGERED_EVENT) += industrialio-triggered-event.o obj-y += accel/
drivers/iio/industrialio-sw-trigger.c new file +183 -0
183 lines not on this page · raw message
include/linux/iio/sw_trigger.h new file +71 -0
71 lines not on this page · raw message
Signature · 1 line
1.9.1
kbuild test robot
build bot[PATCH] iio: core: fix ptr_ret.cocci warnings
drivers/iio/industrialio-sw-trigger.c:169:1-3: WARNING: PTR_ERR_OR_ZERO can be used
drivers/iio/industrialio-sw-trigger.c:169:1-3: WARNING: PTR_ERR_OR_ZERO can be used
Use PTR_ERR_OR_ZERO rather than if(IS_ERR(...)) + PTR_ERR
Generated by: scripts/coccinelle/api/ptr_ret.cocci
---
industrialio-sw-trigger.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/iio/industrialio-sw-trigger.c +1 -3
@@ -166,9 +166,7 @@ static int __init iio_sw_trigger_init(voconfigfs_register_default_group (&iio_configfs_subsys.su_group,"triggers" , &iio_triggers_group_type); -if (IS_ERR (iio_triggers_group)) -return PTR_ERR (iio_triggers_group); -return 0 ; +return PTR_ERR_OR_ZERO (iio_triggers_group); }module_init (iio_sw_trigger_init);
kbuild test robot
build botRe: [PATCH v10 3/5] iio: core: Introduce IIO software triggers
In reply to Daniel Balutaurl: https://github.com/0day-ci/linux/commits/Daniel-Baluta/Add-initial-configfs-support-for-IIO/20151023-233520
Hi Daniel,
[auto build test WARNING on next-20151022 -- if it's inappropriate base, please suggest rules for selecting the more suitable base]
coccinelle warnings: (new ones prefixed by >>)
>> drivers/iio/industrialio-sw-trigger.c:169:1-3: WARNING: PTR_ERR_OR_ZERO can be used
Please review and possibly fold the followup patch.
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
PATCH v10 4/5 iio: trigger: Introduce IIO hrtimer based trigger
In reply to Daniel BalutaThis patch registers a new IIO software trigger interrupt source based on high resolution timers.
Notice that if configfs is enabled we create sampling_frequency attribute allowing users to change hrtimer period (1/sampling_frequency).
The IIO hrtimer trigger has a long history, this patch is based on an older version from Marten and Lars-Peter.
--- drivers/iio/trigger/Kconfig | 10 ++ drivers/iio/trigger/Makefile | 2 + drivers/iio/trigger/iio-trig-hrtimer.c | 193 +++++++++++++++++++++++++++++++++ 3 files changed, 205 insertions(+) create mode 100644 drivers/iio/trigger/iio-trig-hrtimer.c
drivers/iio/trigger/Kconfig +10 -0
@@ -5,6 +5,16 @@ menu "Triggers - standalone" +config IIO_HRTIMER_TRIGGER + tristate "High resolution timer trigger" + depends on IIO_SW_TRIGGER + help + Provides a frequency based IIO trigger using high resolution + timers as interrupt source. + + To compile this driver as a module, choose M here: the + module will be called iio-trig-hrtimer. + config IIO_INTERRUPT_TRIGGER tristate "Generic interrupt trigger" help
drivers/iio/trigger/Makefile +2 -0
@@ -3,5 +3,7 @@ # # When adding new entries keep the list in alphabetical order + +obj-$(CONFIG_IIO_HRTIMER_TRIGGER) += iio-trig-hrtimer.o obj-$(CONFIG_IIO_INTERRUPT_TRIGGER) += iio-trig-interrupt.o obj-$(CONFIG_IIO_SYSFS_TRIGGER) += iio-trig-sysfs.o
drivers/iio/trigger/iio-trig-hrtimer.c new file +193 -0
193 lines not on this page · raw message
Signature · 1 line
1.9.1
PATCH v10 5/5 iio: Documentation: Add IIO configfs documentation
In reply to Daniel Baluta--- Documentation/ABI/testing/configfs-iio | 21 ++++++++ Documentation/iio/iio_configfs.txt | 93 ++++++++++++++++++++++++++++++++++ 2 files changed, 114 insertions(+) create mode 100644 Documentation/ABI/testing/configfs-iio create mode 100644 Documentation/iio/iio_configfs.txt
Documentation/ABI/testing/configfs-iio new file +21 -0
@@ -0,0 +1,21 @@ +What: /config/iio +Date: October 2015 +KernelVersion: 4.4 +Contact: [email protected] +Description: + This represents Industrial IO configuration entry point + directory. It contains sub-groups corresponding to IIO + objects. + +What: /config/iio/triggers +Date: October 2015 +KernelVersion: 4.4 +Description: + Industrial IO software triggers directory. + +What: /config/iio/triggers/hrtimers +Date: October 2015 +KernelVersion: 4.4 +Description: + High resolution timers directory. Creating a directory here + will result in creating a hrtimer trigger in the IIO subsystem.
Documentation/iio/iio_configfs.txt new file +93 -0
93 lines not on this page · raw message
Signature · 1 line
1.9.1
Re: [PATCH v10 5/5] iio: Documentation: Add IIO configfs documentation
(as my last comment was sufficiently explained and kinda forgot to add it)
Quoted 141 lines · Daniel Baluta
Signed-off-by Daniel Baluta <[email protected]>--- Documentation/ABI/testing/configfs-iio | 21 ++++++++ Documentation/iio/iio_configfs.txt | 93 ++++++++++++++++++++++++++++++++++ 2 files changed, 114 insertions(+) create mode 100644 Documentation/ABI/testing/configfs-iio create mode 100644 Documentation/iio/iio_configfs.txtDocumentation/ABI/testing/configfs-iio new file +21
-0@@ -0,0 +1,21 @@ +What: /config/iio +Date: October 2015 +KernelVersion: 4.4 +Contact: [email protected] +Description: + This represents Industrial IO configuration entry point + directory. It contains sub-groups corresponding to IIO + objects. + +What: /config/iio/triggers +Date: October 2015 +KernelVersion: 4.4 +Description: + Industrial IO software triggers directory. + +What: /config/iio/triggers/hrtimers +Date: October 2015 +KernelVersion: 4.4 +Description: + High resolution timers directory. Creating a directory here + will result in creating a hrtimer trigger in the IIO subsystem.Documentation/iio/iio_configfs.txt new file +93
-1@@ -0,0 +1,93 @@ +Industrial IIO configfs support + +1. Overview + +Configfs is a filesystem-based manager of kernel objects. IIO uses some +objects that could be easily configured using configfs (e.g.: devices, +triggers). + +See Documentation/filesystems/configfs/configfs.txt for more information +about how configfs works. + +2. Usage + +In order to use configfs support in IIO we need to select it at compile +time via CONFIG_IIO_CONFIGFS config option. + +Then, mount the configfs filesystem (usually under /config directory): + +$ mkdir /config +$ mount -t configfs none /config + +At this point, all default IIO groups will be created and can be accessed +under /config/iio. Next chapters will describe available IIO configuration +objects. + +3. Software triggers + +One of the IIO default configfs groups is the "triggers" group. It is +automagically accessible when the configfs is mounted and can be found +under /config/iio/triggers. + +IIO software triggers implementation offers support for creating multiple +trigger types. A new trigger type is usually implemented as a separate +kernel module following the interface in include/linux/iio/sw_trigger.h: + +/* + * drivers/iio/trigger/iio-trig-sample.c + * sample kernel module implementing a new trigger type + */ +#include <linux/iio/sw_trigger.h> + + +static struct iio_sw_trigger *iio_trig_sample_probe(const char *name) +{ + /* + * This allocates and registers an IIO trigger plus other + * trigger type specific initialization. + */ +} + +static int iio_trig_hrtimer_remove(struct iio_sw_trigger *swt) +{ + /* + * This undoes the actions in iio_trig_sample_probe + */ +} + +static const struct iio_sw_trigger_ops iio_trig_sample_ops = { + .probe = iio_trig_sample_probe, + .remove = iio_trig_sample_remove, +}; + +static struct iio_sw_trigger_type iio_trig_sample = { + .name = "trig-sample", + .owner = THIS_MODULE, + .ops = &iio_trig_sample_ops, +}; + +module_iio_sw_trigger_driver(iio_trig_sample); + +Each trigger type has its own directory under /config/iio/triggers. Loading +iio-trig-sample module will create 'trig-sample' trigger type directory +/config/iio/triggers/trig-sample. + +We support the following interrupt sources (trigger types): + * hrtimer, uses high resolution timers as interrupt source + +3.1 Hrtimer triggers creation and destruction + +Loading iio-trig-hrtimer module will register hrtimer trigger types allowing +users to create hrtimer triggers under /config/iio/triggers/hrtimer. + +e.g: + +$ mkdir /config/triggers/hrtimer/instance1 +$ rmdir /config/triggers/hrtimer/instance1 + +Each trigger can have one or more attributes specific to the trigger type. + +3.2 "hrtimer" trigger types attributes + +"hrtimer" trigger type doesn't have any configurable attribute from /config dir. +It does introduce the sampling_frequency attribute to trigger directory. --1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html