target kvm support for amd64

gnu/dist/gdb6/gdb/amd64nbsd-nat.c

@@ -28,6 +28,9 @@
 #include "amd64-tdep.h"
 #include "amd64-nat.h"
 
+#include <machine/frame.h>
+#include <machine/pcb.h>
+
 /* Mapping between the general-purpose registers in NetBSD/amd64
    `struct reg' format and GDB's register cache layout for
    NetBSD/i386.
@@ -58,6 +61,68 @@ static int amd64nbsd32_r_reg_offset[] =
 };
 
 
+static int
+amd64nbsd_supply_pcb (struct regcache *regcache, struct pcb *pcb)
+{
+  struct switchframe sf;
+  int regnum;
+
+  /* The following is true for NetBSD/amd64:
+
+     The pcb contains the stack pointer at the point of the context
+     switch in cpu_switch().  At that point we have a stack frame as
+     described by `struct switchframe', which for NetBSD/amd64 has the
+     following layout:
+
+     interrupt level
+     %r15
+     %r14
+     %r13
+     %r12
+     %rbp
+     %rbx
+     return address
+
+     Together with %rsp in the pcb, this accounts for all callee-saved
+     registers specified by the psABI.  From this information we
+     reconstruct the register state as it would look when we just
+     returned from cpu_switch().
+
+     For kernel core dumps the pcb is saved by savectx().  In that case
+     the stack frame only contains the return address, and there is no way
+     to recover the other registers.  */
+
+  /* The stack pointer shouldn't be zero.  */
+  if (pcb->pcb_rsp == 0)
+    return 0;
+
+  /* Read the stack frame, and check its validity.  */
+  read_memory (pcb->pcb_rsp, (gdb_byte *) &sf, sizeof sf);
+  if (sf.sf_rbp == pcb->pcb_rbp)
+    {
+      /* Yes, we have a frame that matches cpu_switch().  */
+      pcb->pcb_rsp += sizeof (struct switchframe);
+      regcache_raw_supply (regcache, 12, &sf.sf_r12);
+      regcache_raw_supply (regcache, 13, &sf.sf_r13);
+      regcache_raw_supply (regcache, 14, &sf.sf_r14);
+      regcache_raw_supply (regcache, 15, &sf.sf_r15);
+      regcache_raw_supply (regcache, AMD64_RBX_REGNUM, &sf.sf_rbx);
+      regcache_raw_supply (regcache, AMD64_RIP_REGNUM, &sf.sf_rip);
+    }
+  else
+    {
+      /* No, the pcb must have been last updated by savectx().  */
+      pcb->pcb_rsp += 8;
+      regcache_raw_supply (regcache, AMD64_RIP_REGNUM, &sf);
+    }
+
+  regcache_raw_supply (regcache, AMD64_RSP_REGNUM, &pcb->pcb_rsp);
+  regcache_raw_supply (regcache, AMD64_RBP_REGNUM, &pcb->pcb_rbp);
+  regcache_raw_supply (regcache, AMD64_FS_REGNUM, &pcb->pcb_fs);
+  regcache_raw_supply (regcache, AMD64_GS_REGNUM, &pcb->pcb_gs);
+
+  return 1;
+}
 /* Provide a prototype to silence -Wmissing-prototypes.  */
 void _initialize_amd64nbsd_nat (void);
 
@@ -74,4 +139,8 @@ _initialize_amd64nbsd_nat (void)
   t = amd64bsd_target ();
   t->to_pid_to_exec_file = nbsd_pid_to_exec_file;
   add_target (t);
+
+  /* Support debugging kernel virtual memory images.  */
+  bsd_kvm_add_target (amd64nbsd_supply_pcb);
+
 }

gnu/dist/gdb6/gdb/amd64nbsd-tdep.c

@@ -31,6 +31,8 @@
 #include "amd64-tdep.h"
 #include "nbsd-tdep.h"
 #include "solib-svr4.h"
+#include "trad-frame.h"
+#include "frame-unwind.h"
 
 /* Support for signal handlers.  */
 
@@ -98,6 +100,119 @@ int amd64nbsd_r_reg_offset[] =
   15 * 8			/* %gs */
 };
 
+/* Kernel debuging support */
+
+#define amd64nbsd_tf_reg_offset amd64nbsd_r_reg_offset
+
+static struct trad_frame_cache *
+amd64nbsd_trapframe_cache(struct frame_info *next_frame, void **this_cache)
+{
+  struct trad_frame_cache *cache;
+  CORE_ADDR func, sp, addr;
+  ULONGEST cs;
+  char *name;
+  int i;
+
+  if (*this_cache)
+    return *this_cache;
+
+  cache = trad_frame_cache_zalloc (next_frame);
+  *this_cache = cache;
+
+  func = frame_func_unwind (next_frame);
+  sp = frame_unwind_register_unsigned (next_frame, AMD64_RSP_REGNUM);
+
+  find_pc_partial_function (func, &name, NULL, NULL);
+  if (name && strncmp (name, "Xintr", 5) == 0)
+    addr = sp + 8;		/* It's an interrupt frame.  */
+  else
+    addr = sp;
+
+  for (i = 0; i < ARRAY_SIZE (amd64nbsd_tf_reg_offset); i++)
+    if (amd64nbsd_tf_reg_offset[i] != -1)
+      trad_frame_set_reg_addr (cache, i, addr + amd64nbsd_tf_reg_offset[i]);
+
+  /* Read %cs from trap frame.  */
+  addr += amd64nbsd_tf_reg_offset[AMD64_CS_REGNUM];
+  cs = read_memory_unsigned_integer (addr, 8); 
+  if ((cs & I386_SEL_RPL) == I386_SEL_UPL)
+    {
+      /* Trap from user space; terminate backtrace.  */
+      trad_frame_set_id (cache, null_frame_id);
+    }
+  else
+    {
+      /* Construct the frame ID using the function start.  */
+      trad_frame_set_id (cache, frame_id_build (sp + 16, func));
+    }
+
+  return cache;
+}
+
+static void
+amd64nbsd_trapframe_this_id (struct frame_info *next_frame,
+			     void **this_cache, struct frame_id *this_id)
+{
+  struct trad_frame_cache *cache =
+    amd64nbsd_trapframe_cache (next_frame, this_cache);
+  
+  trad_frame_get_id (cache, this_id);
+}
+
+static void
+amd64nbsd_trapframe_prev_register (struct frame_info *next_frame,
+				   void **this_cache, int regnum,
+				   int *optimizedp, enum lval_type *lvalp,
+				   CORE_ADDR *addrp, int *realnump,
+				   gdb_byte *valuep)
+{
+  struct trad_frame_cache *cache =
+    amd64nbsd_trapframe_cache (next_frame, this_cache);
+
+  trad_frame_get_register (cache, next_frame, regnum,
+			   optimizedp, lvalp, addrp, realnump, valuep);
+}
+
+static int
+amd64nbsd_trapframe_sniffer (const struct frame_unwind *self,
+			     struct frame_info *next_frame,
+			     void **this_prologue_cache)
+{
+  ULONGEST cs;
+  char *name;
+
+  /* Check Current Privilege Level and bail out if we're not executing
+     in kernel space.  */
+  cs = frame_unwind_register_unsigned (next_frame, AMD64_CS_REGNUM);
+  if ((cs & I386_SEL_RPL) == I386_SEL_UPL)
+    return 0;
+
+  find_pc_partial_function (frame_pc_unwind (next_frame), &name, NULL, NULL);
+  return (name && ((strcmp (name, "alltraps") == 0)
+		   || (strcmp (name, "calltrap") == 0)
+		   || (strncmp (name, "Xtrap", 5) == 0)
+		   || (strcmp (name, "osyscall1") == 0)
+		   || (strcmp (name, "Xsyscall") == 0)
+		   || (strncmp (name, "Xintr", 5) == 0)
+		   || (strncmp (name, "Xresume", 7) == 0)
+		   || (strncmp (name, "Xstray", 6) == 0)
+		   || (strncmp (name, "Xrecurse", 8) == 0)
+		   || (strncmp (name, "Xsoft", 5) == 0)
+		   || (strcmp (name, "Xdoreti", 5) == 0)));
+}
+
+static const struct frame_unwind amd64nbsd_trapframe_unwind = {
+  /* FIXME: kettenis/20051219: This really is more like an interrupt
+     frame, but SIGTRAMP_FRAME would print <signal handler called>,
+     which really is not what we want here.  */
+  NORMAL_FRAME,
+  amd64nbsd_trapframe_this_id,
+  amd64nbsd_trapframe_prev_register,
+  NULL,
+  amd64nbsd_trapframe_sniffer
+};
+
+
 static void
 amd64nbsd_init_abi (struct gdbarch_info info, struct gdbarch *gdbarch)
 {
@@ -121,6 +236,9 @@ amd64nbsd_init_abi (struct gdbarch_info info, struct gdbarch *gdbarch)
   /* NetBSD uses SVR4-style shared libraries.  */
   set_solib_svr4_fetch_link_map_offsets
     (gdbarch, svr4_lp64_fetch_link_map_offsets);
+
+  /* Unwind kernel trap frames correctly.  */
+  frame_unwind_prepend_unwinder (gdbarch, &amd64nbsd_trapframe_unwind);
 }
 
 

gnu/dist/gdb6/gdb/config/i386/nbsd64.mh

@@ -1,3 +1,3 @@
 # Host: NetBSD/amd64
 NATDEPFILES= fork-child.o inf-ptrace.o \
-	amd64-nat.o amd64bsd-nat.o nbsd-nat.o amd64nbsd-nat.o
+	amd64-nat.o amd64bsd-nat.o nbsd-nat.o amd64nbsd-nat.o bsd-kvm.o