Hi,
commit b05bbe3ea2db ("Reimplement IDR and IDA using the radix tree") seems to
break thermal zone allocation. This happens both on todays mainline and
linux-next-20161216 and produces errors like:
[ 9.397201] WARNING: CPU: 1 PID: 165 at ../fs/sysfs/dir.c:31
sysfs_warn_dup+0x64/0x74
[ 9.405145] sysfs: cannot create duplicate filename
'/devices/virtual/thermal/thermal_zone1'
[ 9.514951] Hardware name: Rockchip (Device Tree)
[ 9.522611] [<c03104a4>] (unwind_backtrace) from [<c030ba50>]
(show_stack+0x10/0x14)
[ 9.532925] [<c030ba50>] (show_stack) from [<c05a11fc>]
(dump_stack+0x88/0x9c)
[ 9.542695] [<c05a11fc>] (dump_stack) from [<c0340c68>] (__warn+0xe8/0x100)
[ 9.552166] [<c0340c68>] (__warn) from [<c0340cb8>]
(warn_slowpath_fmt+0x38/0x48)
[ 9.562154] [<c0340cb8>] (warn_slowpath_fmt) from [<c047e944>]
(sysfs_warn_dup+0x64/0x74)
[ 9.572834] [<c047e944>] (sysfs_warn_dup) from [<c047ea20>]
(sysfs_create_dir_ns+0x84/0x94)
[ 9.583687] [<c047ea20>] (sysfs_create_dir_ns) from [<c05a2cb8>]
(kobject_add_internal+0xac/0x2fc)
[ 9.595149] [<c05a2cb8>] (kobject_add_internal) from [<c05a2f50>]
(kobject_add+0x48/0x98)
[ 9.605829] [<c05a2f50>] (kobject_add) from [<c07ee340>]
(device_add+0xd4/0x5b0)
[ 9.615714] [<c07ee340>] (device_add) from [<c0a43a98>]
(thermal_zone_device_register+0x174/0x5cc)
[ 9.627173] [<c0a43a98>] (thermal_zone_device_register) from [<c0a3c588>]
(__power_supply_register+0x374/0x44c)
[ 9.639772] [<c0a3c588>] (__power_supply_register) from [<c0a3c6bc>]
(devm_power_supply_register+0x4c/0x7c)
System is a Rockchip rk3288-veyron and it has 4 thermal zones (3 from
the thermal ip block of the soc and one from the battery).
Reverting the patch mentioned above results in thermal zones being
registered again.
While I haven't looked to deeply into what idr exactly does, some findings:
- thermal_zone0 and thermal_zone1 are allocated correctly
- every further thermal_zone always gets allocated the number "1"
- thermal core calls idr_alloc with 0 for both start and end
- the rewrite-patch seems to change the semantics of idr_alloc
where it orignally said "@end: the maximum id (exclusive, <= 0 for max)"
the "<= 0" part is gone now, but I checked, simply setting INT_MAX
as end in the thermal_core does not help
Heiko
