>> + [RDT_RESOURCE_NODE] =
>> + {
>> + .r_resctrl = {
>> + .rid = RDT_RESOURCE_NODE,
>> + .name = "L3",
> "L3" was named as RDT_RESOURCE_L3 already. The duplicate name here may
> cause duplicate file names in info dir. Maybe rename it as "L3_NODE"?
I thought the same, and my first implementation used a different string here (I picked
"NODE" rather than "L3_NODE").
But my testers complained that this broke all their existing infrastructure that reads
cache occupancy and memory bandwidth. This string is not just used in the info/
directory, it is also the basis for the directory names in mon_data/
$ tree /sys/fs/resctrl/mon_data
/sys/fs/resctrl/mon_data
├── mon_L3_00
│ ├── llc_occupancy
│ ├── mbm_local_bytes
│ └── mbm_total_bytes
├── mon_L3_01
│ ├── llc_occupancy
│ ├── mbm_local_bytes
│ └── mbm_total_bytes
├── mon_L3_02
│ ├── llc_occupancy
│ ├── mbm_local_bytes
│ └── mbm_total_bytes
└── mon_L3_03
├── llc_occupancy
├── mbm_local_bytes
└── mbm_total_bytes
The name using "L3" is still appropriate and accurate.
There isn't a "duplicate file names" problem in the info/ directory because a system
either has SNC disabled, and uses the L3-scoped resource, or has SNC enabled and
uses the node-scoped resource.
-Tony
