PWRMGR DV document

Goals

  • DV
    • Verify all PWRMGR IP features by running dynamic simulations with a SV/UVM based testbench.
    • Develop and run all tests based on the testplan below towards closing code and functional coverage on the IP and all of its sub-modules.
  • FPV
    • Verify TileLink device protocol compliance with an SVA based testbench.

Current status

Design features

For detailed information on PWRMGR design features, please see the PWRMGR HWIP technical specification.

Testbench architecture

PWRMGR testbench has been constructed based on the CIP testbench architecture.

Block diagram

Block diagram

Top level testbench

Top level testbench is located at hw/ip/pwrmgr/dv/tb.sv. It instantiates the PWRMGR DUT module hw/ip/pwrmgr/rtl/pwrmgr.sv. In addition, it instantiates the following interfaces, connects them to the DUT and sets their handle into uvm_config_db:

Common DV utility components

The following utilities provide generic helper tasks and functions to perform activities that are common across the project:

Global types & methods

All common types and methods defined at the package level can be found in pwrmgr_env_pkg. Some of them in use are:

[list a few parameters, types & methods; no need to mention all]

TL_agent

PWRMGR testbench instantiates (already handled in CIP base env) tl_agent which provides the ability to drive and independently monitor random traffic via TL host interface into PWRMGR device.

UVM RAL Model

The PWRMGR RAL model is created with the ralgen FuseSoC generator script automatically when the simulation is at the build stage.

It can be created manually by invoking regtool:

Stimulus strategy

The sequences are closely related to the testplan’s testpoints. Testpoints and coverage are described in more detail in the testplan.

Test sequences

All test sequences reside in hw/ip/pwrmgr/dv/env/seq_lib, and extend pwrmgr_base_vseq. The pwrmgr_base_vseq virtual sequence is extended from cip_base_vseq and serves as a starting point. It provides commonly used handles, variables, functions and tasks that the test sequences can simple use / call. Some of the most commonly used tasks / functions are as follows:

  • task start_slow_fsm: Sets the AST main_pok input to 1 after some random slow clock cycles.
  • task wait_for_fast_fsm_active: Waits for the fetch_en_o output to become 1, indicating the fast fsm is active and the cpu can fetch instructions. We wait for this before the tests can start, since any CSR accesses require the CPU to be running. Due to complexities in the UVM sequences this task is called in the virtual post_apply_reset task of dv_base_vseq.
  • task wait_for_csr_to_propagate_to_slow_domain: Waits for cfg_cdc_sync CSR to be clear, indicating the CDC to the slow clock has completed.
  • task wait_for_reset_cause: Waits for the pwr_rst_req.reset_cause output to match an expected cause.

In addition, the base sequence provides two tasks that provide expected inputs based on the pwrmgr outputs. In the absence of these inputs the pwrmgr will be stuck waiting forever. Being based on outputs means the inputs are in accordance to the implicit protocol. The tasks in question are:

  • task slow_responder: Handles required input changes for the slow state machine. For the various *clk_en outputs it changes the *clk_val as required.
  • task fast_responder: Handles input changes for the fast state machine.
    • Completes the handshake with rstmgr for lc and sys resets: some random cycles after an output reset is requested the corresponding reset src input must go low.
    • Completes the handshake with clkmgr: the clk_status input needs to match the ip_clk_en output after some cycles.
    • Completes the handshake with lc and otp: both *_done input must match the corresponding *_init output after some cycles.

These tasks are started by the parent sequence’s pre_start task, and terminated gracefully in the parent sequence’s post_start task.

The test sequences besides the base are as follows:

  • pwrmgr_smoke_vseq tests the pwrmgr through POR, entry and exit from software initiated low power and reset.
  • pwrmgr_wakeup_vseq checks the transitions to low power and the wakeup settings. It randomizes wakeup inputs, wakeup enables, the wakeup info capture enable, and the interrupt enable.
  • pwrmgr_clks_en_vseq checks that the peripheral clock enables match the settings of the control CSR during low power. It uses a subset of the wakeup sequence.
  • pwrmgr_aborted_lowpower_vseq creates scenarios that lead to aborting a lowpower transition. The abort can be due to the processor waking up very soon, or otp, lc, or flash being busy.
  • pwrmgr_reset_vseq checks the pwrmgr response to resets and reset enables.
  • pwrmgr_escalation_reset_vseq checks the response to an escalation reset.
  • pwrmgr_reset_wakeup_vseq aligns reset and wakeup from low power.
  • pwrmgr_lowpower_wakeup_race_vseq aligns a wakeup event coming in proximity to low power entry. Notice the wakeup is not expected to impact low power entry, since it is not sampled at this time.

Functional coverage

To ensure high quality constrained random stimulus, it is necessary to develop a functional coverage model. The following covergroups have been developed to prove that the test intent has been adequately met:

  • wakeup_ctrl_cg covers wakeup and capture control.
  • wakeup_intr_cg covers control of the interrupt due to a wakeup.
  • clock_enables_cg covers clock controls.
  • reset_cg
  • reset_lowpower_distance_cg

More details about these sequences and covergroups can be found at hw/ip/pwrmgr/data/pwrmgr_testplan.hjson.

Self-checking strategy

Scoreboard

The pwrmgr_scoreboard is primarily used for end to end checking.

It creates the following analysis ports to retrieve the data monitored by corresponding interface agents:

  • analysis port1:
  • analysis port2:

Many inputs must have specific transitions to prevent the pwrmgr fsms from wait forever. When possible the transitions are triggered by pwrmgr output changes. These are described according to the unit that originates or is the recepient ot the ports. See also the test plan for specific ways these are driven to trigger different testpoints.

AST
  • Output slow_clk_en is always on.
  • Input slow_clk_val is unused.
  • Outputs core_clk_en, io_clk_en, and usb_clk_en reset low, and go high prior to the slow fsm requesting the fast fsm to wakeup. Notice the usb clock can be programmed to stay low on wakeup via the control CSR. These clock enables are cleared on reset, and should match their corresponding enables in the control CSR on low power transitions. These clock enables are checked via SVAs in hw/ip/pwrmgr/dv/sva/pwrmgr_clock_enables_sva_if.sv. When slow fsm transitions to SlowPwrStateReqPwrUp the clock enables should be on (except usb should match control.usb_clk_en_active). When slow fsm transitions to SlowPwrStatePwrClampOn the clock enables should match their bits in the control CSR.
  • Inputs core_clk_val, io_clk_val, and usb_clk_val track the corresponding enables. They are driven by slow_responder, which turn them off when their enables go off, and turn them back on a few random slow clock cycles after their enables go on. Slow fsm waits for them to go high prior to requesting fast fsm wakeup. Lack of a high transition when needed is detected via timeout. Such timeout would be due to the corresponding enables being set incorrectly. These inputs are checked via SVAs in hw/ip/pwrmgr/dv/sva/pwrmgr_ast_sva_if.sv.
  • Output main_pd_n should go high when slow fsm transitions to SlowPwrStateMainPowerOn, and should match control.main_pd_n CSR when slow fsm transitions to SlowPwrStateMainPowerOff.
  • Input main_pok should turn on for the slow fsm to start power up sequence. This is also driven by slow_responder, which turn this off in response to main_pd_n going low, and turn it back on after a few random slow clock cycles from main_pd_n going high. Lack of a high transition causes a timeout, and would point to main_pd_n being set incorrectly.
  • Output transitions of pwr_clamp_env must always precede transitions of pwr_clamp output. Output transitions of pwr_clamp to active must always precede transitions of main_pd_n output to active. Output transitions of pwr_clamp to inactive must always follow transitions of main_pd_n output to inactive.
RSTMGR
  • Output rst_lc_req resets to 1, also set on reset transition, and on low power transitions that turn off main clock. Cleared early on during the steps to fast fsm active.
  • Input rst_lc_src_n go low in response to rst_lc_req high, go high when rst_lc_req clears (and lc is reset). Driven by fast_responder in response to rst_lc_req, waiting a few random cycles prior to transitions. Fast fsm waits for it to go low before deactivating, and for it to go high before activating. Checked implicitly by lack of timeout: a timeout would be due to rst_lc_req being set incorrectly.
  • Output rst_sys_req resets to 1, also set to on reset, and on low power transitions that turn off main clock. Cleared right before the fast fsm goes active.
  • Input rst_sys_src_n go low in response to rst_sys_req high. Transitions go high when rst_sysd_req clears (and lc is reset). Fast fsm waits for it to go low before deactivating. Also driver by fast_responder. Checked implicitly by lack of timeout.
  • Output rstreqs correspond to the enabled pwrmgr rstreqs inputs plus escalation reset. Checked in scoreboard.
  • Output reset_cause indicates a reset is due to low power entry or a reset request. Checked in scoreboard.
CLKMGR
  • Output ip_clk_en resets low, is driven high by fast fsm when going active, and driven low when going inactive.
  • Input clk_status is expected to track ip_clk_en. Fast fsm waits for it going high prior to going active, and for it to go low prior to deactivating. Driven by fast_responder, which turns it off when ip_clk_en goes low, and rurn it back on a few random cycles after ip_clk_en going high. Checked by lack of a timeout: such timeout would be due to ip_clk_en being set incorrectly.
OTP
  • Output otp_init resets low, goes high when the fast fsm is going active, and low after the otp_done input goes high.
  • Input otp_done is driven by fast_responder. It is initialized low, and goes high some random cycles after otp_init goes high. The sequencer will timeout if otp_init is not driven high.
  • Input otp_idle will normally be set high, but will be set low by the pwrmgr_aborted_lowpower_vseq sequence.
LC

The pins connecting to LC behave pretty much the same way as those to OTP.

FLASH
  • Input flash_idle is handled much like lc_idle and otp_idle.
CPU
  • Input core_sleeping is driven by sequences. It is driven low to enable a transition to low power. After the transition is under way it is a don’t care. The pwrmgr_aborted_lowpower_vseq sequence sets it carefully to abort a low power entry soon after the attempt because the processor wakes up.
Wakepus and Resets

There are a number of wakeup and reset requests. They are driven by sequences as they need to.

Assertions

  • TLUL assertions: The hw/ip/pwrmgr/dv/sva/pwrmgr_bind.sv module binds the tlul_assert assertions to the IP to ensure TileLink interface protocol compliance.
  • Unknown checks on DUT outputs: The RTL has assertions to ensure all outputs are initialized to known values after coming out of reset.
  • Clock enables assertions: The hw/ip/pwrmgr/dv/sva/pwrmgr_bind.sv module binds pwrmgr_clock_enables_if to the ip. It contains assertions checking that the various clk_en outputs correspond to the settings in the control CSR.
  • AST input/output handshake assertions: The hw/ip/pwrmgr/dv/sva/pwrmgr_bind.sv module binds pwrmgr_ast_if to the ip. It contains assertions checking that the inputs from the AST respond to the pwrmgr outputs.
  • RSTMGR input/output handshake assertions: The hw/ip/pwrmgr/dv/sva/pwrmgr_bind.sv module binds pwrmgr_rstmgr_if to the ip. It contains assertions checking that the inputs from the RSTMGR respond to the pwrmgr outputs.

More assertions will be added according to the different unit’s descriptions above.

Building and running tests

We are using our in-house developed regression tool for building and running our tests and regressions. Please take a look at the link for detailed information on the usage, capabilities, features and known issues. Here’s how to run a smoke test:

$ $REPO_TOP/util/dvsim/dvsim.py $REPO_TOP/hw/ip/pwrmgr/dv/pwrmgr_sim_cfg.hjson -i pwrmgr_smoke

Testplan

Testpoints

Milestone Name Tests Description
V1 smoke pwrmgr_smoke

Smoke test exercising the pwrmgr state transitions.

  • Brings pwrmgr out of POR.
  • Enables wakeup.
  • Triggers SW initiated low power transition with reset settings in control CSR.
  • Triggers wakeup.
  • Enables reset.
  • Triggers a reset.
  • Waits for pwrmgr to be out of reset.

Stimulus:

  • CSR writes to wakeup_en, reset_en, and low_power_hint.
  • Needs many input pins to line up correctly in order to prevent the pwrmgr from waiting forever.

Checks:

  • Checks that fast fsm is active checking the CSR ctrl_cfg_regwen is set.
  • Checks that the wakeup and reset causes are as expected reading CSRs wake_status and reset_status.
  • Checks the output pwr_rst_req.reset_cause matches a low power or reset cause.
  • Checks the output pwr_rst_req.rstreqs matches the enabled resets.
V1 csr_hw_reset pwrmgr_csr_hw_reset

Verify the reset values as indicated in the RAL specification.

  • Write all CSRs with a random value.
  • Apply reset to the DUT as well as the RAL model.
  • Read each CSR and compare it against the reset value. it is mandatory to replicate this test for each reset that affects all or a subset of the CSRs.
  • It is mandatory to run this test for all available interfaces the CSRs are accessible from.
  • Shuffle the list of CSRs first to remove the effect of ordering.
V1 csr_rw pwrmgr_csr_rw

Verify accessibility of CSRs as indicated in the RAL specification.

  • Loop through each CSR to write it with a random value.
  • Read the CSR back and check for correctness while adhering to its access policies.
  • It is mandatory to run this test for all available interfaces the CSRs are accessible from.
  • Shuffle the list of CSRs first to remove the effect of ordering.
V1 csr_bit_bash pwrmgr_csr_bit_bash

Verify no aliasing within individual bits of a CSR.

  • Walk a 1 through each CSR by flipping 1 bit at a time.
  • Read the CSR back and check for correctness while adhering to its access policies.
  • This verify that writing a specific bit within the CSR did not affect any of the other bits.
  • It is mandatory to run this test for all available interfaces the CSRs are accessible from.
  • Shuffle the list of CSRs first to remove the effect of ordering.
V1 csr_aliasing pwrmgr_csr_aliasing

Verify no aliasing within the CSR address space.

  • Loop through each CSR to write it with a random value
  • Shuffle and read ALL CSRs back.
  • All CSRs except for the one that was written in this iteration should read back the previous value.
  • The CSR that was written in this iteration is checked for correctness while adhering to its access policies.
  • It is mandatory to run this test for all available interfaces the CSRs are accessible from.
  • Shuffle the list of CSRs first to remove the effect of ordering.
V1 csr_mem_rw_with_rand_resetpwrmgr_csr_mem_rw_with_rand_reset

Verify random reset during CSR/memory access.

  • Run csr_rw sequence to randomly access CSRs
  • If memory exists, run mem_partial_access in parallel with csr_rw
  • Randomly issue reset and then use hw_reset sequence to check all CSRs are reset to default value
  • It is mandatory to run this test for all available interfaces the CSRs are accessible from.
V2 wakeup

Test random wakeup, wakeup_en, wake_info_capture_dis, and interrupt.

The different wakeup inputs can be disabled via bits in the wakeup_en CSR. Update of wakeup_info can be disabled via the wake_info_capture_dis CSR. Any wakeup causes an interrupt unless interrupts are disabled.

Stimulus:

  • Sets wakeup_en randomly but don't set it to zero, or the test will timeout.
  • Bring pwrmgr to low power.
  • Set wakeups_i inputs randomly.
  • Set intr_enable randomly.

Checks:

  • Waits for fast fsm to become active reading ctrl_cfg_regwen until it becomes 1.
  • Depending on wakeups_i:
    • If all wakeups are disabled, wait some time checking the state remains inactive.
    • Set wakeups_i so at least one is enabled.
  • Checks ip_clk_en output has a low transition.
  • Checks the wakeup_info and wakeup_status CSRs.
  • Checks the output pwr_rst_req.reset_cause matches a low power cause.
  • Check that intr_wakeup_o is set according to intr_enable CSR.
  • Coverage collected by wakeup_cg and wakeup_intr_cg.
V2 control_clks

Test CSR control of peripheral clocks during low power.

The peripheral clocks can be configured to remain on or be turned off during low power with bits in the control CSR register. The usb clock can also be configured off in active mode.

Stimulus:

  • Sets these control bits at random.
  • Cause a low power transition and wakeup.

Checks:

  • The clock enable outputs to the AST clocks during a low power transition match the control bits.
  • The usb clock enable is also checked during active mode against the control register.
V2 aborted_lowpower

Test aborted low power transitions.

Low power transitions can be aborted in two cases:

  • The processor gets an interrupt soon after a low power entry is triggered.
  • OTP, LC, or FLASH are not idle. This test aborts low power transitions, and disables any wakeups, so the test would timeout if low power was entered.

Stimulus:

  • Bring pwrmgr to low power.
  • Either disable pwr_cpu.core_sleeping or keep some of lc_idle, otp_idle, or flash_idle inputs off.
  • Disable all wakeup enables.
  • Randomly set wakeup_info_capture_dis CSR.

Checks:

  • The ctrl_cfg_regwen CSR reads as 1 on the first attempt.
  • Checks the output pwr_rst_req.reset_cause doesn't change for a bounded amount of time.
  • Check that the wakeup_info CSR flags either fall_through or abort events when capture is enabled.
V2 reset

Test random reset and reset_en.

The different reset inputs can be disabled via bits in the reset_en CSR. Resets can be triggered either in active or low power state.

Stimulus:

  • Sets reset_en randomly.
  • Randomly choose whether to put the unit in low power mode.
  • Set rstreqs_i inputs randomly, both in value and in time.

Checks:

  • Waits for fast fsm to become active reading ctrl_cfg_regwen until it becomes 1.
  • Checks the reset_status CSRs.
  • Checks ip_clk_en output has a low transition.
  • Checks the output pwr_rst_req.reset_cause matches a reset cause.
  • Checks the output pwr_rst_req.rstreqs matches the enabled resets.
V2 escalation_reset

Test escalation reset.

An escalation reset cannot be disabled.

Stimulus:

  • Set esc_rst_tx_i inputs randomly.

Checks:

  • Waits for fast fsm to become active reading ctrl_cfg_regwen until it becomes 1.
  • Checks the reset_status CSRs.
  • Checks the output pwr_rst_req.reset_cause matches a reset cause.
  • Checks the output pwr_rst_req.rstreqs is not set.
  • Check the output esc_rst_rx_o indicates an alert escalation.
V2 reset_wakeup_race

Test wakeup from low power and reset request almost coinciding.

If a wakeup from low power and a reset occur at nearly the same time the system handles them one at a time.

Stimulus:

  • Trigger reset and wakeup from low power as described for other testpoints.
  • Make sure to test them coinciding also.

Check:

  • Similar tests as for the wakeup and reset testpoints, except making sure they happen per the triggering order.
V2 lowpower_wakeup_race

Test wakeups coming close to lowpower entry.

If low power entry and a wakeup are closely aligned the hardware could get confused. Notice this is very unlikely, since wakeup is only sensed when the slow fsm is in LowPower state.

Stimulus:

  • Trigger low power entry as described for other testpoints.
  • Have all wakeups enabled.
  • Assert wakeups_i in the temporal neighborhood of low power entry.

Check:

  • No timeout occurs.
  • Either pwrmgr remains active or a full low power cycle occurs.
V2 intr_test pwrmgr_intr_test

Verify common intr_test CSRs that allows SW to mock-inject interrupts.

  • Enable a random set of interrupts by writing random value(s) to intr_enable CSR(s).
  • Randomly "turn on" interrupts by writing random value(s) to intr_test CSR(s).
  • Read all intr_state CSR(s) back to verify that it reflects the same value as what was written to the corresponding intr_test CSR.
  • Check the cfg.intr_vif pins to verify that only the interrupts that were enabled and turned on are set.
  • Clear a random set of interrupts by writing a randomly value to intr_state CSR(s).
  • Repeat the above steps a bunch of times.
V2 tl_d_oob_addr_access pwrmgr_tl_errors

Access out of bounds address and verify correctness of response / behavior

V2 tl_d_illegal_access pwrmgr_tl_errors

Drive unsupported requests via TL interface and verify correctness of response / behavior. Below error cases are tested bases on the [TLUL spec]({{< relref "hw/ip/tlul/doc/_index.md#explicit-error-cases" >}})

  • TL-UL protocol error cases
    • invalid opcode
    • some mask bits not set when opcode is PutFullData
    • mask does not match the transfer size, e.g. a_address = 0x00, a_size = 0, a_mask = 'b0010
    • mask and address misaligned, e.g. a_address = 0x01, a_mask = 'b0001
    • address and size aren't aligned, e.g. a_address = 0x01, a_size != 0
    • size is greater than 2
  • OpenTitan defined error cases
    • access unmapped address, expect d_error = 1 when devmode_i == 1
    • write a CSR with unaligned address, e.g. a_address[1:0] != 0
    • write a CSR less than its width, e.g. when CSR is 2 bytes wide, only write 1 byte
    • write a memory with a_mask != '1 when it doesn't support partial accesses
    • read a WO (write-only) memory
    • write a RO (read-only) memory
V2 tl_d_outstanding_access pwrmgr_csr_hw_reset
pwrmgr_csr_rw
pwrmgr_csr_aliasing
pwrmgr_same_csr_outstanding

Drive back-to-back requests without waiting for response to ensure there is one transaction outstanding within the TL device. Also, verify one outstanding when back- to-back accesses are made to the same address.

V2 tl_d_partial_access pwrmgr_csr_hw_reset
pwrmgr_csr_rw
pwrmgr_csr_aliasing
pwrmgr_same_csr_outstanding

Access CSR with one or more bytes of data. For read, expect to return all word value of the CSR. For write, enabling bytes should cover all CSR valid fields.

V3 stress

The standard stress test.

V3 tl_intg_err pwrmgr_tl_intg_err

Verify that the data integrity check violation generates an alert.

Randomly inject errors on the control, data, or the ECC bits during CSR accesses. Verify that triggers the correct fatal alert.

V3 stress_all_with_rand_resetpwrmgr_stress_all_with_rand_reset

This test runs 3 parallel threads - stress_all, tl_errors and random reset. After reset is asserted, the test will read and check all valid CSR registers.

Covergroups

Name Description
clock_control_cg

Collects coverage on clock enable bits from control CSR during a lowpower transition and active state.

reset_cg

Collects coverage related to reset reporting.

Covergroup contains coverpoints for the rstreqs_i and alert escalation inputs, reset_en, reset_status, escalate_reset_status and whether reset or escalation is asserted during low power state, and suitable crosses.

reset_lowpower_distance_cg

This is a temporal cover statement that records the temporal distance between when lowpower entry and reset trigger, in order to measure coverage of the reset_wakeup_race sequence.

tl_errors_cg

Cover the following error cases on TL-UL bus:

  • TL-UL protocol error cases.
  • OpenTitan defined error cases, refer to testpoint tl_d_illegal_access.
tl_intg_err_cg

Cover all kinds of integrity errors (command, data or both) and cover number of error bits on each integrity check.

tl_intg_err_mem_subword_cg

Cover the kinds of integrity errors with byte enabled write on memory.

Some memories store the integrity values. When there is a subword write, design re-calculate the integrity with full word data and update integrity in the memory. This coverage ensures that memory byte write has been issued and the related design logic has been verfied.

wakeup_ctrl_cg

Collects coverage on wakeup enable and capture functionality.

This is collected per individual wakeup bit. Covergroup contains coverpoints for the wakeup_en CSR bit, wakeup_info_capture_dis CSR, wakeups_i input bit, and wakeup_status CSR bit, and their cross.

wakeup_intr_cg

Collects coverage on interrupts for wakeup functionality.

This is collected per individual wakeup bit. Covergroup contains coverpoints for the intr_en CSR, the wakeup_status CSR bit, and intr_status CSR, and their cross.