126 lines · plain
1======================================================2HiSilicon SoC uncore Performance Monitoring Unit (PMU)3======================================================4 5The HiSilicon SoC chip includes various independent system device PMUs6such as L3 cache (L3C), Hydra Home Agent (HHA) and DDRC. These PMUs are7independent and have hardware logic to gather statistics and performance8information.9 10The HiSilicon SoC encapsulates multiple CPU and IO dies. Each CPU cluster11(CCL) is made up of 4 cpu cores sharing one L3 cache; each CPU die is12called Super CPU cluster (SCCL) and is made up of 6 CCLs. Each SCCL has13two HHAs (0 - 1) and four DDRCs (0 - 3), respectively.14 15HiSilicon SoC uncore PMU driver16-------------------------------17 18Each device PMU has separate registers for event counting, control and19interrupt, and the PMU driver shall register perf PMU drivers like L3C,20HHA and DDRC etc. The available events and configuration options shall21be described in the sysfs, see:22 23/sys/bus/event_source/devices/hisi_sccl{X}_<l3c{Y}/hha{Y}/ddrc{Y}>.24The "perf list" command shall list the available events from sysfs.25 26Each L3C, HHA and DDRC is registered as a separate PMU with perf. The PMU27name will appear in event listing as hisi_sccl<sccl-id>_module<index-id>.28where "sccl-id" is the identifier of the SCCL and "index-id" is the index of29module.30 31e.g. hisi_sccl3_l3c0/rd_hit_cpipe is READ_HIT_CPIPE event of L3C index #0 in32SCCL ID #3.33 34e.g. hisi_sccl1_hha0/rx_operations is RX_OPERATIONS event of HHA index #0 in35SCCL ID #1.36 37The driver also provides a "cpumask" sysfs attribute, which shows the CPU core38ID used to count the uncore PMU event.39 40Example usage of perf::41 42 $# perf list43 hisi_sccl3_l3c0/rd_hit_cpipe/ [kernel PMU event]44 ------------------------------------------45 hisi_sccl3_l3c0/wr_hit_cpipe/ [kernel PMU event]46 ------------------------------------------47 hisi_sccl1_l3c0/rd_hit_cpipe/ [kernel PMU event]48 ------------------------------------------49 hisi_sccl1_l3c0/wr_hit_cpipe/ [kernel PMU event]50 ------------------------------------------51 52 $# perf stat -a -e hisi_sccl3_l3c0/rd_hit_cpipe/ sleep 553 $# perf stat -a -e hisi_sccl3_l3c0/config=0x02/ sleep 554 55For HiSilicon uncore PMU v2 whose identifier is 0x30, the topology is the same56as PMU v1, but some new functions are added to the hardware.57 581. L3C PMU supports filtering by core/thread within the cluster which can be59specified as a bitmap::60 61 $# perf stat -a -e hisi_sccl3_l3c0/config=0x02,tt_core=0x3/ sleep 562 63This will only count the operations from core/thread 0 and 1 in this cluster.64 652. Tracetag allow the user to chose to count only read, write or atomic66operations via the tt_req parameeter in perf. The default value counts all67operations. tt_req is 3bits, 3'b100 represents read operations, 3'b10168represents write operations, 3'b110 represents atomic store operations and693'b111 represents atomic non-store operations, other values are reserved::70 71 $# perf stat -a -e hisi_sccl3_l3c0/config=0x02,tt_req=0x4/ sleep 572 73This will only count the read operations in this cluster.74 753. Datasrc allows the user to check where the data comes from. It is 5 bits.76Some important codes are as follows:77 78- 5'b00001: comes from L3C in this die;79- 5'b01000: comes from L3C in the cross-die;80- 5'b01001: comes from L3C which is in another socket;81- 5'b01110: comes from the local DDR;82- 5'b01111: comes from the cross-die DDR;83- 5'b10000: comes from cross-socket DDR;84 85etc, it is mainly helpful to find that the data source is nearest from the CPU86cores. If datasrc_cfg is used in the multi-chips, the datasrc_skt shall be87configured in perf command::88 89 $# perf stat -a -e hisi_sccl3_l3c0/config=0xb9,datasrc_cfg=0xE/,90 hisi_sccl3_l3c0/config=0xb9,datasrc_cfg=0xF/ sleep 591 924. Some HiSilicon SoCs encapsulate multiple CPU and IO dies. Each CPU die93contains several Compute Clusters (CCLs). The I/O dies are called Super I/O94clusters (SICL) containing multiple I/O clusters (ICLs). Each CCL/ICL in the95SoC has a unique ID. Each ID is 11bits, include a 6-bit SCCL-ID and 5-bit96CCL/ICL-ID. For I/O die, the ICL-ID is followed by:97 98- 5'b00000: I/O_MGMT_ICL;99- 5'b00001: Network_ICL;100- 5'b00011: HAC_ICL;101- 5'b10000: PCIe_ICL;102 1035. uring_channel: UC PMU events 0x47~0x59 supports filtering by tx request104uring channel. It is 2 bits. Some important codes are as follows:105 106- 2'b11: count the events which sent to the uring_ext (MATA) channel;107- 2'b01: is the same as 2'b11;108- 2'b10: count the events which sent to the uring (non-MATA) channel;109- 2'b00: default value, count the events which sent to the both uring and110 uring_ext channel;111 112Users could configure IDs to count data come from specific CCL/ICL, by setting113srcid_cmd & srcid_msk, and data desitined for specific CCL/ICL by setting114tgtid_cmd & tgtid_msk. A set bit in srcid_msk/tgtid_msk means the PMU will not115check the bit when matching against the srcid_cmd/tgtid_cmd.116 117If all of these options are disabled, it can works by the default value that118doesn't distinguish the filter condition and ID information and will return119the total counter values in the PMU counters.120 121The current driver does not support sampling. So "perf record" is unsupported.122Also attach to a task is unsupported as the events are all uncore.123 124Note: Please contact the maintainer for a complete list of events supported for125the PMU devices in the SoC and its information if needed.126