diff --git a/bin/regression-wally b/bin/regression-wally
index cc0588bf6..f4c3cc023 100755
--- a/bin/regression-wally
+++ b/bin/regression-wally
@@ -301,7 +301,8 @@ def addLockstepTestsByDir(dir, config, sim, fcovMode):
         for dirpath, dirnames, filenames in os.walk(os.path.abspath(dir)):
             for file in filenames:
                 # fcov lockstep only runs on WALLY-COV-ALL.elf files; other lockstep runs on all files
-                if (file.endswith(".elf") and fcovMode == 0 or file.endswith("ALL.elf") and fcovMode == 1):
+                if ((file.endswith(".elf") and (fcovMode == 0 or "tests/priv" in dir)) or 
+                    (file.endswith("ALL.elf") and fcovMode == 1)):
                     fullfile = os.path.join(dirpath, file)
                     fields = fullfile.rsplit('/', 3)
                     if (fields[2] == "ref"):
@@ -419,9 +420,11 @@ if (args.ccov):  # only run RV64GC tests on Questa in code coverage mode
     addTests(tests64gc_nofp, coveragesim)
     if (args.fp):
         addTests(tests64gc_fp, coveragesim)
-elif (args.fcov):  # only run RV64GC tests on Questa in lockstep in functional coverage mode
+elif (args.fcov):  # run tests in lockstep in functional coverage mode
     addLockstepTestsByDir(WALLY+"/addins/cvw-arch-verif/tests/rv32/", "rv32gc", coveragesim, 1)
     addLockstepTestsByDir(WALLY+"/addins/cvw-arch-verif/tests/rv64/", "rv64gc", coveragesim, 1)
+    addLockstepTestsByDir(WALLY+"/addins/cvw-arch-verif/tests/priv/rv32/", "rv32gc", coveragesim, 1)
+    addLockstepTestsByDir(WALLY+"/addins/cvw-arch-verif/tests/priv/rv64/", "rv64gc", coveragesim, 1)
     #addLockstepTestsByDir(WALLY+"/tests/riscof/work/wally-riscv-arch-test/rv64i_m/privilege/src/", "rv64gc", coveragesim, 0)
 
 else: 
diff --git a/config/rv32gc/imperas.ic b/config/rv32gc/imperas.ic
index 46d0d31c3..2ba3c1280 100644
--- a/config/rv32gc/imperas.ic
+++ b/config/rv32gc/imperas.ic
@@ -36,7 +36,7 @@
 --override cpu/Zicboz=T
 --override cmomp_bytes=64  # Zic64b
 --override cmoz_bytes=64   # Zic64b
---override lr_sc_grain=8   # Za64rs requires <=64; we use native word size
+--override lr_sc_grain=4   # Za64rs requires <=64; we use native word size
 
 # 64 KiB continuous huge pages supported
 #--override cpu/Svpbmt=F
diff --git a/config/shared/config-shared.vh b/config/shared/config-shared.vh
index 91e1d4100..4ccb24bf2 100644
--- a/config/shared/config-shared.vh
+++ b/config/shared/config-shared.vh
@@ -122,9 +122,3 @@ localparam FMALEN = 3*NF + 6;
 localparam NORMSHIFTSZ = `max(`max((CVTLEN+NF+1), (DIVb + 1 + NF + 1)), (FMALEN + 2));
 
 localparam LOGNORMSHIFTSZ = ($clog2(NORMSHIFTSZ));                  // log_2(NORMSHIFTSZ)
-
-// Disable spurious Verilator warnings
-
-/* verilator lint_off STMTDLY */
-/* verilator lint_off ASSIGNDLY */
-/* verilator lint_off PINCONNECTEMPTY */
diff --git a/config/shared/parameter-defs.vh b/config/shared/parameter-defs.vh
index c80b00232..88de34afc 100644
--- a/config/shared/parameter-defs.vh
+++ b/config/shared/parameter-defs.vh
@@ -95,11 +95,7 @@ localparam cvw_t P = '{
   PLIC_SPI_ID :        PLIC_SPI_ID,
   PLIC_SDC_ID :        PLIC_SDC_ID,
   BPRED_SUPPORTED :        BPRED_SUPPORTED,
-                       /* verilator lint_off ENUMVALUE */
-                       // *** definitely need to fix this.
-                       // it thinks we are casting from the enum type to BPRED_TYPE.
   BPRED_TYPE :        BPRED_TYPE,
-                       /* verilator lint_on ENUMVALUE */
   BPRED_SIZE :        BPRED_SIZE,
   BPRED_NUM_LHR : BPRED_NUM_LHR,                       
   BTB_SIZE :        BTB_SIZE,
diff --git a/src/fpu/fdivsqrt/fdivsqrtiter.sv b/src/fpu/fdivsqrt/fdivsqrtiter.sv
index 39de58855..ff03d29ec 100644
--- a/src/fpu/fdivsqrt/fdivsqrtiter.sv
+++ b/src/fpu/fdivsqrt/fdivsqrtiter.sv
@@ -38,7 +38,6 @@ module fdivsqrtiter import cvw::*;  #(parameter cvw_t P) (
   output logic [P.DIVb+3:0] FirstWS, FirstWC       // Q4.DIVb
 );
 
-  /* verilator lint_off UNOPTFLAT */
   logic [P.DIVb+3:0]      WSNext[P.DIVCOPIES-1:0]; // Q4.DIVb
   logic [P.DIVb+3:0]      WCNext[P.DIVCOPIES-1:0]; // Q4.DIVb
   logic [P.DIVb+3:0]      WS[P.DIVCOPIES:0];       // Q4.DIVb
@@ -56,7 +55,6 @@ module fdivsqrtiter import cvw::*;  #(parameter cvw_t P) (
   logic [P.DIVb+1:0]      NextC;                   // Q2.DIVb
   logic [P.DIVb:0]        UMux, UMMux;             // U1.DIVb
   logic [P.DIVb:0]        initU, initUM;           // U1.DIVb
-  /* verilator lint_on UNOPTFLAT */
 
   // Top Muxes and Registers
   // When start is asserted, the inputs are loaded into the divider.
diff --git a/src/fpu/fdivsqrt/fdivsqrtpreproc.sv b/src/fpu/fdivsqrt/fdivsqrtpreproc.sv
index 737d9089a..a399e4a4f 100644
--- a/src/fpu/fdivsqrt/fdivsqrtpreproc.sv
+++ b/src/fpu/fdivsqrt/fdivsqrtpreproc.sv
@@ -136,7 +136,7 @@ module fdivsqrtpreproc import cvw::*;  #(parameter cvw_t P) (
     // calculate right shift amount RightShiftX to complete in discrete number of steps
     if (P.RK > 1) begin // more than 1 bit per cycle
       logic [$clog2(P.RK)-1:0] RightShiftX;
-      /* verilator lint_offf WIDTH */
+      /* verilator lint_off WIDTH */
       assign RightShiftX = P.RK - 1 - ((IntResultBitsE - 1) % P.RK); // Right shift amount
       assign DivXShifted = DivX >> RightShiftX;                     // shift X by up to R*K-1 to complete in n steps
       /* verilator lint_on WIDTH */
diff --git a/src/fpu/fdivsqrt/fdivsqrtstage2.sv b/src/fpu/fdivsqrt/fdivsqrtstage2.sv
index fa13cadeb..06ac4ec82 100644
--- a/src/fpu/fdivsqrt/fdivsqrtstage2.sv
+++ b/src/fpu/fdivsqrt/fdivsqrtstage2.sv
@@ -28,7 +28,6 @@
 ////////////////////////////////////////////////////////////////////////////////////////////////
 
 
-/* verilator lint_off UNOPTFLAT */
 module fdivsqrtstage2 import cvw::*;  #(parameter cvw_t P) (
   input  logic [P.DIVb+3:0] D, DBar,        // Q4.DIVb
   input  logic [P.DIVb:0]   U, UM,          // U1.DIVb
@@ -40,7 +39,6 @@ module fdivsqrtstage2 import cvw::*;  #(parameter cvw_t P) (
   output logic [P.DIVb:0]   UNext, UMNext,  // U1.DIVb
   output logic [P.DIVb+3:0] WSNext, WCNext  // Q4.DIVb
 );
- /* verilator lint_on UNOPTFLAT */
 
   logic [P.DIVb+3:0]        Dsel;     // Q4.DIVb
   logic                     up, uz;
diff --git a/src/fpu/fround.sv b/src/fpu/fround.sv
index 2814c766b..519d55096 100644
--- a/src/fpu/fround.sv
+++ b/src/fpu/fround.sv
@@ -115,9 +115,9 @@ module fround import cvw::*;  #(parameter cvw_t P) (
   ///////////////////////////
 
   // Exact logic
-  // verilator lint_off WIDTHEXPAND
+  /* verilator lint_off WIDTHEXPAND */
   assign EgeNf = (E >= Nf) & Xe[P.NE-1]; // Check if E >= Nf.  Also check that Xe is positive to avoid wraparound problems
-  // verilator lint_on WIDTHEXPAND
+  /* verilator lint_on WIDTHEXPAND */
 
   // Rounding logic: determine whether to round up in magnitude
   always_comb begin
diff --git a/src/generic/prioritythermometer.sv b/src/generic/prioritythermometer.sv
index 23acfcfb3..ecf37e80e 100644
--- a/src/generic/prioritythermometer.sv
+++ b/src/generic/prioritythermometer.sv
@@ -39,11 +39,9 @@ module prioritythermometer #(parameter N = 8) (
   //  Rather than linear.
 
   // create thermometer code mask
-  /* verilator lint_off UNOPTFLAT */
   genvar i;
   assign y[0] = ~a[0];
   for (i=1; i<N; i++) begin:therm
     assign y[i] = y[i-1] & ~a[i];
   end
-  /* verilator lint_on UNOPTFLAT */
 endmodule
diff --git a/src/lsu/lsu.sv b/src/lsu/lsu.sv
index a1dd85d98..9f5d460f7 100644
--- a/src/lsu/lsu.sv
+++ b/src/lsu/lsu.sv
@@ -121,11 +121,9 @@ module lsu import cvw::*;  #(parameter cvw_t P) (
   logic                  DCacheCommittedM;                       // D$ memory operation started, delay interrupts
 
   logic [P.LLEN-1:0]     DTIMReadDataWordM;                      // DTIM read data
-  /* verilator lint_off WIDTHEXPAND */  
   logic [MLEN-1:0]       DCacheReadDataWordM;                    // D$ read data
   logic [MLEN-1:0]       LSUWriteDataSpillM;                     // Final write data
   logic [MLEN/8-1:0]     ByteMaskSpillM;                         // Selects which bytes within a word to write
-  /* verilator lint_on WIDTHEXPAND */
   logic [P.LLEN-1:0]     DCacheReadDataWordSpillM;               // D$ read data
   logic [P.LLEN-1:0]     ReadDataWordMuxM;                       // DTIM or D$ read data
   logic [P.LLEN-1:0]     LittleEndianReadDataWordM;              // Endian-swapped read data
diff --git a/src/mdu/divstep.sv b/src/mdu/divstep.sv
index d47742fa9..aa93d6767 100644
--- a/src/mdu/divstep.sv
+++ b/src/mdu/divstep.sv
@@ -27,8 +27,6 @@
 // and limitations under the License.
 ////////////////////////////////////////////////////////////////////////////////////////////////
 
-/* verilator lint_off UNOPTFLAT */
-
 module divstep #(parameter XLEN) (
   input  logic [XLEN-1:0] W,     // Residual in
   input  logic [XLEN-1:0] XQ,    // bits of dividend X and quotient Q in
@@ -46,5 +44,3 @@ module divstep #(parameter XLEN) (
   assign qi = ~qib;
   mux2 #(XLEN) wrestoremux(WShift, WPrime, qi, WOut); // if quotient is zero, restore W
 endmodule
-
-/* verilator lint_on UNOPTFLAT */
diff --git a/src/privileged/csrm.sv b/src/privileged/csrm.sv
index fd82106c1..1cdfd7d85 100644
--- a/src/privileged/csrm.sv
+++ b/src/privileged/csrm.sv
@@ -88,6 +88,7 @@ module csrm  import cvw::*;  #(parameter cvw_t P) (
   // .. up to 15 more at consecutive addresses
   localparam PMPADDR0      = 12'h3B0;
   // ... up to 63 more at consecutive addresses
+  /* verilator lint_off UNUSEDPARAM */
   localparam TSELECT       = 12'h7A0;
   localparam TDATA1        = 12'h7A1;
   localparam TDATA2        = 12'h7A2;
@@ -96,6 +97,7 @@ module csrm  import cvw::*;  #(parameter cvw_t P) (
   localparam DPC           = 12'h7B1;
   localparam DSCRATCH0     = 12'h7B2;
   localparam DSCRATCH1     = 12'h7B3;
+  /* verilator lint_off UNUSEDPARAM */
   // Constants
   localparam ZERO = {(P.XLEN){1'b0}};
   // when compressed instructions are supported, there can't be misaligned instructions
diff --git a/src/uncore/clint_apb.sv b/src/uncore/clint_apb.sv
index 76735aaa6..e03745194 100644
--- a/src/uncore/clint_apb.sv
+++ b/src/uncore/clint_apb.sv
@@ -115,7 +115,7 @@ module clint_apb import cvw::*;  #(parameter cvw_t P) (
     always_ff @(posedge PCLK) 
       if (~PRESETn) begin
         MSIP <= 1'b0;
-        MTIMECMP <= '0;
+        MTIMECMP <= 64'hFFFFFFFFFFFFFFFF; // Spec says MTIMECMP is not reset, but we reset to maximum value to prevent spurious timer interrupts
       end else if (memwrite) begin
         if (entry == 16'h0000) MSIP <= PWDATA[0];
         if (entry == 16'h4000) 
diff --git a/src/uncore/spi_apb.sv b/src/uncore/spi_apb.sv
index 04c3d3c97..4f4fcbc40 100644
--- a/src/uncore/spi_apb.sv
+++ b/src/uncore/spi_apb.sv
@@ -114,7 +114,7 @@ module spi_apb import cvw::*; #(parameter cvw_t P) (
   
   /* verilator lint_off UNDRIVEN */
   logic [2:0]  TransmitWriteWatermarkLevel, ReceiveReadWatermarkLevel; // unused generic FIFO outputs
-  /* verilator lint_off UNDRIVEN */
+  /* verilator lint_on UNDRIVEN */
   logic [7:0]  ReceiveShiftRegEndian;              // Reverses ReceiveShiftReg if Format[2] set (little endian transmission)
 
   // Shift reg signals
@@ -180,7 +180,7 @@ module spi_apb import cvw::*; #(parameter cvw_t P) (
           SPI_RXMARK:  ReceiveWatermark <= Din[2:0];
           SPI_IE:      InterruptEnable <= Din[1:0];
         endcase
-      /* verilator lint_off CASEINCOMPLETE */
+      /* verilator lint_on CASEINCOMPLETE */
       
       // According to FU540 spec: Once interrupt is pending, it will remain set until number 
       // of entries in tx/rx fifo is strictly more/less than tx/rxmark
diff --git a/src/uncore/uartPC16550D.sv b/src/uncore/uartPC16550D.sv
index 31172462d..3639fe8d5 100644
--- a/src/uncore/uartPC16550D.sv
+++ b/src/uncore/uartPC16550D.sv
@@ -335,10 +335,10 @@ module uartPC16550D #(parameter UART_PRESCALE) (
     end
 
   assign rxfifoempty = (rxfifohead == rxfifotail);
-  // verilator lint_off WIDTH
+  /* verilator lint_off WIDTH */
   assign rxfifoentries = (rxfifohead >= rxfifotail) ? (rxfifohead-rxfifotail) : 
                          (rxfifohead + 16 - rxfifotail);
-  // verilator lint_on WIDTH
+  /* verilator lint_on WIDTH */
   assign rxfifotriggered = rxfifoentries >= rxfifotriggerlevel;
   assign rxfifotimeout = rxtimeoutcnt == {rxbitsexpected, 6'b0}; // time out after 4 character periods; probably not right yet
   //assign rxfifotimeout = 0; // disabled pending fix