This series adds tracing events for the chips IRQ and registers that are useful to figure out the current data and power status. This came about since diagnosing why a certain usb-c hub or dp-alt-mode adapter fails is hard with the information in /sys/class/typec alone since this does not have a timeline of events (and we don't want every typec user having to also buy a PD analyzer). With this series debugging these kinds of things starts to become fun: # echo 1 > /sys/kernel/debug/tracing/events/tps6598x/enable # cat /sys/kernel/debug/tracing/trace_pipe irq/79-0-003f-526 [003] .... 512.717871: tps6598x_irq: event1=PLUG_EVENT|DATA_STATUS_UPDATE|STATUS_UPDATE, event2= irq/79-0-003f-526 [003] .... 512.722408: tps6598x_status: conn: conn-Ra, pp_5v0: off, pp_hv: off, pp_ext: off, pp_cable: off, pwr-src: vin-3p3, vbus: vSafe0V, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE irq/79-0-003f-526 [003] .... 512.727127: tps6598x_data_status: DATA_CONNECTION|USB2_CONNECTION|USB3_CONNECTION irq/79-0-003f-526 [003] .... 512.769571: tps6598x_irq: event1=PP_SWITCH_CHANGED|STATUS_UPDATE, event2= irq/79-0-003f-526 [003] .... 512.773380: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: vSafe0V, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 512.872450: tps6598x_irq: event1=POWER_STATUS_UPDATE|PD_STATUS_UPDATE, event2= irq/79-0-003f-526 [003] .... 512.876311: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: vSafe0V, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 512.880237: tps6598x_power_status: conn: 1, pwr-role: source, typec: usb, bc: sdp irq/79-0-003f-526 [003] .... 513.072682: tps6598x_irq: event1=STATUS_UPDATE, event2= irq/79-0-003f-526 [003] .... 513.076390: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: vSafe5V, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.090676: tps6598x_irq: event1=ERROR_CANNOT_PROVIDE_PWR, event2= irq/79-0-003f-526 [003] .... 513.094368: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: vSafe5V, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.109606: tps6598x_irq: event1=NEW_CONTRACT_AS_PROVIDER|POWER_STATUS_UPDATE|STATUS_UPDATE|SRC_TRANSITION, event2= irq/79-0-003f-526 [003] .... 513.113777: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: pd, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.117475: tps6598x_power_status: conn: 1, pwr-role: source, typec: pd, bc: sdp irq/79-0-003f-526 [003] .... 513.137469: tps6598x_irq: event1=VDM_RECEIVED, event2= irq/79-0-003f-526 [003] .... 513.141570: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: pd, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.281926: tps6598x_irq: event1=VDM_RECEIVED, event2= irq/79-0-003f-526 [003] .... 513.285638: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: pd, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.300515: tps6598x_irq: event1=VDM_RECEIVED|DATA_STATUS_UPDATE, event2= irq/79-0-003f-526 [003] .... 513.304226: tps6598x_status: conn: conn-Ra, pp_5v0: out, pp_hv: off, pp_ext: off, pp_cable: in, pwr-src: vin-3p3, vbus: pd, usb-host: no, legacy: no, flags: PLUG_PRESENT|PORTROLE|DATAROLE|VCONN irq/79-0-003f-526 [003] .... 513.308302: tps6598x_data_status: DATA_CONNECTION|USB2_CONNECTION|USB3_CONNECTION|DP_CONNECTION, DP pinout D It should not impose any problems for firmwares that don't have IRQs for the registers enabled. The trace will then just be missing those events. Patch is against next-20210210. changes from v1: - Fix issues on 32bit arches and missing include spotted by kbuild (build testes on mips) https://lore.kernel.org/linux-usb/202102120644.IXu75GmJ-lkp@xxxxxxxxx/ https://lore.kernel.org/linux-usb/202102120159.Lyh8AGhQ-lkp@xxxxxxxxx/ changes from v2: - Reduce macro expansion size by using a custom FIELD_GET. This avoids a sparse warning. https://lore.kernel.org/linux-usb/20210213031237.GP219708@shao2-debian/ Guido Günther (4): usb: typec: tps6598x: Add trace event for IRQ events usb: typec: tps6598x: Add trace event for status register usb: typec: tps6598x: Add trace event for power status register usb: typec: tps6598x: Add trace event for data status drivers/usb/typec/Makefile | 3 + drivers/usb/typec/tps6598x.c | 66 ++++--- drivers/usb/typec/tps6598x.h | 189 +++++++++++++++++++ drivers/usb/typec/tps6598x_trace.h | 283 +++++++++++++++++++++++++++++ 4 files changed, 515 insertions(+), 26 deletions(-) create mode 100644 drivers/usb/typec/tps6598x.h create mode 100644 drivers/usb/typec/tps6598x_trace.h -- 2.30.0
