diff --git a/Makefile b/Makefile
index 38ec5c7..1e45056 100644
--- a/Makefile
+++ b/Makefile
@@ -182,7 +182,7 @@ QEMUGDB = $(shell if $(QEMU) -help | grep -q '^-gdb'; \
 	then echo "-gdb tcp::$(GDBPORT)"; \
 	else echo "-s -p $(GDBPORT)"; fi)
 ifndef CPUS
-CPUS := 1
+CPUS := 2
 endif
 QEMUOPTS = -machine virt -kernel kernel -m 3G -smp $(CPUS) -nographic
 QEMUOPTS += -initrd fs.img
diff --git a/defs.h b/defs.h
index 0e8cb52..2353e4a 100644
--- a/defs.h
+++ b/defs.h
@@ -130,12 +130,11 @@ void            swtch(struct context*, struct context*);
 
 // spinlock.c
 void            acquire(struct spinlock*);
-void            getcallerpcs(void*, uint64*);
 int             holding(struct spinlock*);
 void            initlock(struct spinlock*, char*);
 void            release(struct spinlock*);
-void            pushcli(void);
-void            popcli(void);
+void            push_off(void);
+void            pop_off(void);
 
 // sleeplock.c
 void            acquiresleep(struct sleeplock*);
@@ -168,6 +167,7 @@ void            timerinit(void);
 // trap.c
 extern uint     ticks;
 void            trapinit(void);
+void            trapinithart(void);
 extern struct spinlock tickslock;
 void            usertrapret(void);
 
@@ -179,6 +179,7 @@ int             uartgetc(void);
 
 // vm.c
 void            kvminit(void);
+void            kvminithart(void);
 pagetable_t     uvmcreate(void);
 void            uvminit(pagetable_t, char *, uint);
 uint64          uvmalloc(pagetable_t, uint64, uint64);
@@ -194,6 +195,7 @@ int             copyinstr(pagetable_t pagetable, char *dst, uint64 srcva, uint64
 
 // plic.c
 void            plicinit(void);
+void            plicinithart(void);
 uint64          plic_pending(void);
 int             plic_claim(void);
 void            plic_complete(int);
diff --git a/main.c b/main.c
index 24793cd..2168b9f 100644
--- a/main.c
+++ b/main.c
@@ -4,25 +4,39 @@
 #include "riscv.h"
 #include "defs.h"
 
+volatile static int started = 0;
+
 // Bootstrap processor starts running C code here.
 // Allocate a real stack and switch to it, first
 // doing some setup required for memory allocator to work.
 void
-main(int hartid)
+main()
 {
-  w_tp(hartid);    // save hartid where cpuid() can find it
-  uartinit();      // serial port
-  consoleinit();
-  printf("entering main() on hart %d\n", hartid);
-  kinit();         // physical page allocator
-  kvminit();       // kernel page table
-  procinit();      // process table
-  trapinit();      // trap vectors
-  plicinit();      // set up interrupt controller
-  binit();         // buffer cache
-  fileinit();      // file table
-  ramdiskinit();   // disk
-  userinit();      // first user process
+  if(cpuid() == 0){
+    uartinit();      // serial port
+    consoleinit();
+    printf("hart %d starting\n", cpuid());
+    kinit();         // physical page allocator
+    kvminit();       // create kernel page table
+    kvminithart();   // turn on paging
+    procinit();      // process table
+    trapinit();      // trap vectors
+    trapinithart();  // install kernel trap vector
+    plicinit();      // set up interrupt controller
+    plicinithart();  // ask PLIC for device interrupts
+    binit();         // buffer cache
+    fileinit();      // file table
+    ramdiskinit();   // disk
+    userinit();      // first user process
+    started = 1;
+  } else {
+    while(started == 0)
+      ;
+    printf("hart %d starting\n", cpuid());
+    kvminithart();    // turn on paging
+    trapinithart();   // install kernel trap vector
+    plicinithart();   // ask PLIC for device interrupts
+  }
 
   scheduler();        
 }
diff --git a/memlayout.h b/memlayout.h
index 9c9cfdb..462986c 100644
--- a/memlayout.h
+++ b/memlayout.h
@@ -28,6 +28,14 @@
 
 // qemu puts programmable interrupt controller here.
 #define PLIC 0x0c000000L
+#define PLIC_PRIORITY (PLIC + 0x0)
+#define PLIC_PENDING (PLIC + 0x1000)
+#define PLIC_MENABLE(hart) (PLIC + 0x2000 + (hart)*0x100)
+#define PLIC_SENABLE(hart) (PLIC + 0x2080 + (hart)*0x100)
+#define PLIC_MPRIORITY(hart) (PLIC + 0x200000 + (hart)*0x2000)
+#define PLIC_SPRIORITY(hart) (PLIC + 0x201000 + (hart)*0x2000)
+#define PLIC_MCLAIM(hart) (PLIC + 0x200004 + (hart)*0x2000)
+#define PLIC_SCLAIM(hart) (PLIC + 0x201004 + (hart)*0x2000)
 
 #define RAMDISK 0x88000000L
 
diff --git a/proc.c b/proc.c
index 766fd93..28cac26 100644
--- a/proc.c
+++ b/proc.c
@@ -360,7 +360,7 @@ scheduler(void)
 {
   struct proc *p;
   struct cpu *c = mycpu();
-   
+
   c->proc = 0;
   for(;;){
     // Enable interrupts on this processor.
@@ -385,7 +385,6 @@ scheduler(void)
       c->proc = 0;
     }
     release(&ptable.lock);
-
   }
 }
 
@@ -393,7 +392,7 @@ scheduler(void)
 // and have changed proc->state. Saves and restores
 // intena because intena is a property of this
 // kernel thread, not this CPU. It should
-// be proc->intena and proc->ncli, but that would
+// be proc->intena and proc->noff, but that would
 // break in the few places where a lock is held but
 // there's no process.
 void
diff --git a/proc.h b/proc.h
index 00a1cb9..278e4cd 100644
--- a/proc.h
+++ b/proc.h
@@ -22,9 +22,8 @@ struct context {
 struct cpu {
   struct proc *proc;           // The process running on this cpu or null
   struct context scheduler;   // swtch() here to enter scheduler
-  volatile uint started;       // Has the CPU started?
-  int ncli;                    // Depth of pushcli nesting.
-  int intena;                  // Were interrupts enabled before pushcli?
+  int noff;                    // Depth of push_off() nesting.
+  int intena;                  // Were interrupts enabled before push_off()?
 };
 
 extern struct cpu cpus[NCPU];
diff --git a/spinlock.c b/spinlock.c
index 0377870..8b3c3f0 100644
--- a/spinlock.c
+++ b/spinlock.c
@@ -5,6 +5,7 @@
 #include "memlayout.h"
 #include "spinlock.h"
 #include "riscv.h"
+#include "proc.h"
 #include "defs.h"
 
 void
@@ -15,27 +16,6 @@ initlock(struct spinlock *lk, char *name)
   lk->cpu = 0;
 }
 
-void
-acquire(struct spinlock *lk)
-{
-  lk->locked = 1;
-  lk->cpu = mycpu();
-}
-
-void
-release(struct spinlock *lk)
-{
-  lk->locked = 0;
-  lk->cpu = 0;
-}
-
-int
-holding(struct spinlock *lk)
-{
-  return lk->locked && lk->cpu == mycpu();
-}
-
-#if 0
 // Acquire the lock.
 // Loops (spins) until the lock is acquired.
 // Holding a lock for a long time may cause
@@ -43,12 +23,14 @@ holding(struct spinlock *lk)
 void
 acquire(struct spinlock *lk)
 {
-  pushcli(); // disable interrupts to avoid deadlock.
+  push_off(); // disable interrupts to avoid deadlock.
   if(holding(lk))
     panic("acquire");
 
   // The xchg is atomic.
-  while(xchg(&lk->locked, 1) != 0)
+  //while(xchg(&lk->locked, 1) != 0)
+  //  ;
+  while(__sync_lock_test_and_set(&lk->locked, 1) != 0)
     ;
 
   // Tell the C compiler and the processor to not move loads or stores
@@ -58,7 +40,6 @@ acquire(struct spinlock *lk)
 
   // Record info about lock acquisition for holding() and debugging.
   lk->cpu = mycpu();
-  getcallerpcs(&lk, lk->pcs);
 }
 
 // Release the lock.
@@ -68,7 +49,6 @@ release(struct spinlock *lk)
   if(!holding(lk))
     panic("release");
 
-  lk->pcs[0] = 0;
   lk->cpu = 0;
 
   // Tell the C compiler and the processor to not move loads or stores
@@ -81,27 +61,10 @@ release(struct spinlock *lk)
   // Release the lock, equivalent to lk->locked = 0.
   // This code can't use a C assignment, since it might
   // not be atomic. A real OS would use C atomics here.
-  asm volatile("movl $0, %0" : "+m" (lk->locked) : );
+  //asm volatile("movl $0, %0" : "+m" (lk->locked) : );
+  __sync_lock_release(&lk->locked);
 
-  popcli();
-}
-
-// Record the current call stack in pcs[] by following the %ebp chain.
-void
-getcallerpcs(void *v, uint64 pcs[])
-{
-  uint64 *ebp;
-  int i;
-
-  asm volatile("mov %%rbp, %0" : "=r" (ebp));
-  for(i = 0; i < 10; i++){
-    if(ebp == 0 || ebp < (uint64*)KERNBASE || ebp == (uint64*)0xffffffff)
-      break;
-    pcs[i] = ebp[1];     // saved %eip
-    ebp = (uint64*)ebp[0]; // saved %ebp
-  }
-  for(; i < 10; i++)
-    pcs[i] = 0;
+  pop_off();
 }
 
 // Check whether this cpu is holding the lock.
@@ -109,37 +72,37 @@ int
 holding(struct spinlock *lk)
 {
   int r;
-  pushcli();
+  push_off();
   r = lk->locked && lk->cpu == mycpu();
-  popcli();
+  pop_off();
   return r;
 }
 
-
-// Pushcli/popcli are like cli/sti except that they are matched:
-// it takes two popcli to undo two pushcli.  Also, if interrupts
-// are off, then pushcli, popcli leaves them off.
+// push_off/pop_off are like intr_off()/intr_on() except that they are matched:
+// it takes two pop_off to undo two push_off.  Also, if interrupts
+// are initially off, then push_off, pop_off leaves them off.
 
 void
-pushcli(void)
+push_off(void)
 {
-  int eflags;
+  struct cpu *c = mycpu();
+  int old = intr_get();
 
-  eflags = readeflags();
-  cli();
-  if(mycpu()->ncli == 0)
-    mycpu()->intena = eflags & FL_IF;
-  mycpu()->ncli += 1;
+  intr_off();
+  if(c->noff == 0)
+    c->intena = old;
+  c->noff += 1;
 }
 
 void
-popcli(void)
+pop_off(void)
 {
-  if(readeflags()&FL_IF)
-    panic("popcli - interruptible");
-  if(--mycpu()->ncli < 0)
-    panic("popcli");
-  if(mycpu()->ncli == 0 && mycpu()->intena)
-    sti();
+  struct cpu *c = mycpu();
+  if(intr_get())
+    panic("pop_off - interruptible");
+  c->noff -= 1;
+  if(c->noff < 0)
+    panic("pop_off");
+  if(c->noff == 0 && c->intena)
+    intr_on();
 }
-#endif
diff --git a/spinlock.h b/spinlock.h
index 90bffdb..4392820 100644
--- a/spinlock.h
+++ b/spinlock.h
@@ -5,7 +5,5 @@ struct spinlock {
   // For debugging:
   char *name;        // Name of lock.
   struct cpu *cpu;   // The cpu holding the lock.
-  uint64 pcs[10];      // The call stack (an array of program counters)
-                     // that locked the lock.
 };
 
diff --git a/start.c b/start.c
index 53edb8e..ce22be8 100644
--- a/start.c
+++ b/start.c
@@ -9,12 +9,12 @@ void main();
 // entry.S needs one stack per CPU.
 __attribute__ ((aligned (16))) char stack0[4096 * NCPU];
 
-// assembly code in kernelvec for machine-mode timer interrupt.
-extern void machinevec();
-
 // scratch area for timer interrupt, one per CPU.
 uint64 mscratch0[NCPU * 32];
 
+// assembly code in kernelvec for machine-mode timer interrupt.
+extern void machinevec();
+
 // entry.S jumps here in machine mode on stack0.
 void
 mstart()
@@ -48,7 +48,9 @@ mstart()
   w_mstatus(r_mstatus() | MSTATUS_MIE);
   w_mie(r_mie() |  MIE_MTIE);
 
-  // call main(hartid) in supervisor mode.
-  asm("csrr a0, mhartid ; \
-       mret");
+  // keep each CPU's hartid in its tp register, for cpuid().
+  w_tp(id);
+
+  // call main() in supervisor mode.
+  asm("mret");
 }
diff --git a/trap.c b/trap.c
index 693c596..929f761 100644
--- a/trap.c
+++ b/trap.c
@@ -19,14 +19,16 @@ extern int devintr();
 void
 trapinit(void)
 {
-  int i;
-
-  // set up to take exceptions and traps while in the kernel.
-  w_stvec((uint64)kernelvec);
-
   initlock(&tickslock, "time");
 }
 
+// set up to take exceptions and traps while in the kernel.
+void
+trapinithart(void)
+{
+  w_stvec((uint64)kernelvec);
+}
+
 //
 // handle an interrupt, exception, or system call from user space.
 // called from trampoline.S
diff --git a/vm.c b/vm.c
index b24b271..1d3887a 100644
--- a/vm.c
+++ b/vm.c
@@ -54,9 +54,13 @@ kvminit()
   // the highest virtual address in the kernel.
   mappages(kernel_pagetable, TRAMPOLINE, PGSIZE,
            (uint64)trampout, PTE_R | PTE_X);
+}
 
-  // Switch h/w page table register to the kernel's page table,
-  // and enable paging.
+// Switch h/w page table register to the kernel's page table,
+// and enable paging.
+void
+kvminithart()
+{
   w_satp(MAKE_SATP(kernel_pagetable));
 }