| Message ID | 20170206123956.20082-3-johannes@sipsolutions.net (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On 6-2-2017 13:39, Johannes Berg wrote: > From: Johannes Berg <johannes.berg@intel.com> > > Add a new semantic patch just like for dev_groups to > support class_groups. > > Change-Id: I114e1a6ed8db901b3aa738a5052721d1da45cae4 Does it make sense to keep the Change-Id: things here? > Signed-off-by: Johannes Berg <johannes.berg@intel.com> > --- > .../0001-group-attr/0003-class_groups.cocci | 54 ++++++++++++++++++++++ > 1 file changed, 54 insertions(+) > create mode 100644 patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci > > diff --git a/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci b/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci > new file mode 100644 > index 000000000000..a4694dd75763 > --- /dev/null > +++ b/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci > @@ -0,0 +1,54 @@ > +/* see upstream commit ced6473e74867 */ > + > +@ attribute_group @ > +identifier group; > +declarer name ATTRIBUTE_GROUPS; > +@@ > + > +ATTRIBUTE_GROUPS(group); > + > +@script:python attribute_groups_name@ > +group << attribute_group.group; > +groups; > +@@ > +coccinelle.groups = group + "_groups" > + > +@ class_group @ > +identifier group_class; > +identifier attribute_groups_name.groups; > +fresh identifier group_dev_attr = attribute_group.group ## "_dev_attrs"; So how will this work wrt naming scheme. Can we be sure all class_attrs occurrences in the files we copy are named as such, ie. with _dev_attrs post-fix? Regards, Arend -- To unsubscribe from this list: send the line "unsubscribe backports" in
> > Change-Id: I114e1a6ed8db901b3aa738a5052721d1da45cae4 > > Does it make sense to keep the Change-Id: things here? Not really. But eventually I got lazy. I've now removed the auto- generation for my tree, which means I'll have to add it manually internally, but whatever :) > > +@ class_group @ > > +identifier group_class; > > +identifier attribute_groups_name.groups; > > +fresh identifier group_dev_attr = attribute_group.group ## > > "_dev_attrs"; > > So how will this work wrt naming scheme. Can we be sure all > class_attrs occurrences in the files we copy are named as such, ie. > with _dev_attrs post-fix? No no, that's not what happens here. The _dev_attrs postfix is generated by the ATTRIBUTE_GROUPS_BACKPORT() macro. johannes -- To unsubscribe from this list: send the line "unsubscribe backports" in
diff --git a/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci b/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci new file mode 100644 index 000000000000..a4694dd75763 --- /dev/null +++ b/patches/collateral-evolutions/generic/0001-group-attr/0003-class_groups.cocci @@ -0,0 +1,54 @@ +/* see upstream commit ced6473e74867 */ + +@ attribute_group @ +identifier group; +declarer name ATTRIBUTE_GROUPS; +@@ + +ATTRIBUTE_GROUPS(group); + +@script:python attribute_groups_name@ +group << attribute_group.group; +groups; +@@ +coccinelle.groups = group + "_groups" + +@ class_group @ +identifier group_class; +identifier attribute_groups_name.groups; +fresh identifier group_dev_attr = attribute_group.group ## "_dev_attrs"; +@@ + +struct class group_class = { ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) + .class_groups = groups, ++#else ++ .class_attrs = group_dev_attr, ++#endif +}; + +@ attribute_group_mod depends on class_group @ +declarer name ATTRIBUTE_GROUPS_BACKPORT; +identifier attribute_group.group; +@@ + ++#if LINUX_VERSION_CODE >= KERNEL_VERSION(4,10,0) +ATTRIBUTE_GROUPS(group); ++#else ++#define BP_ATTR_GRP_STRUCT class_attribute ++ATTRIBUTE_GROUPS_BACKPORT(group); ++#endif + +@ class_registering @ +identifier class_register, ret; +identifier class_group.group_class; +fresh identifier group_class_init = "init_" ## attribute_group.group ## "_attrs"; +@@ + +( ++ group_class_init(); + return class_register(&group_class); +| ++ group_class_init(); + ret = class_register(&group_class); +)