/* Live and postmortem kernel debugging functions for FreeBSD. Copyright 1996 Free Software Foundation, Inc. This file is part of GDB. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ #include "defs.h" #include #include #include #include #include #include #include #include #include "frame.h" /* required by inferior.h */ #include "inferior.h" #include "symtab.h" #include "symfile.h" #include "objfiles.h" #include "command.h" #include "bfd.h" #include "target.h" #include "gdbcore.h" #include #include #include #include #include #include #include #include static void kcore_files_info PARAMS ((struct target_ops *)); static void kcore_close PARAMS ((int)); static void get_kcore_registers PARAMS ((int)); static int kcore_xfer_kmem PARAMS ((CORE_ADDR, char *, int, int, struct target_ops *)); static int xfer_umem PARAMS ((CORE_ADDR, char *, int, int)); static CORE_ADDR ksym_lookup PARAMS ((const char *)); static int read_pcb PARAMS ((int, CORE_ADDR)); static struct proc * curProc PARAMS ((void)); static int set_proc_context PARAMS ((CORE_ADDR paddr)); static void kcore_open PARAMS ((char *filename, int from_tty)); static void kcore_detach PARAMS ((char *args, int from_tty)); static void set_proc_cmd PARAMS ((char *arg, int from_tty)); static void set_cpu_cmd PARAMS ((char *arg, int from_tty)); static CORE_ADDR kvtophys PARAMS ((int, CORE_ADDR)); static int physrd PARAMS ((int, u_int, char*, int)); static int kvm_open PARAMS ((const char *efile, char *cfile, char *sfile, int perm, char *errout)); static int kvm_close PARAMS ((int fd)); static int kvm_write PARAMS ((int core_kd, CORE_ADDR memaddr, char *myaddr, int len)); static int kvm_read PARAMS ((int core_kd, CORE_ADDR memaddr, char *myaddr, int len)); static int kvm_uread PARAMS ((int core_kd, struct proc *p, CORE_ADDR memaddr, char *myaddr, int len)); static int kernel_core_file_hook PARAMS ((int fd, CORE_ADDR addr, char *buf, int len)); static struct kinfo_proc * kvm_getprocs PARAMS ((int cfd, int op, CORE_ADDR proc, int *cnt)); extern struct target_ops kcore_ops; /* Forward decl */ /* Non-zero means we are debugging a kernel core file */ int kernel_debugging = 0; int kernel_writablecore = 0; static char *core_file; static int core_kd = -1; static struct proc *cur_proc; static CORE_ADDR kernel_start; static int ncpus; static int cpuid; static CORE_ADDR prv_space; /* per-cpu private space */ static int prv_space_size; #define prv_start (prv_space + cpuid * prv_space_size) /* * Read the "thing" at kernel address 'addr' into the space pointed to * by point. The length of the "thing" is determined by the type of p. * Result is non-zero if transfer fails. */ #define kvread(addr, p) \ (target_read_memory ((CORE_ADDR)(addr), (char *)(p), sizeof(*(p)))) /* * The following is FreeBSD-specific hackery to decode special frames * and elide the assembly-language stub. This could be made faster by * defining a frame_type field in the machine-dependent frame information, * but we don't think that's too important right now. */ enum frametype { tf_normal, tf_trap, tf_interrupt, tf_syscall }; CORE_ADDR fbsd_kern_frame_saved_pc (fr) struct frame_info *fr; { struct minimal_symbol *sym; CORE_ADDR this_saved_pc; enum frametype frametype; this_saved_pc = read_memory_integer (fr->frame + 4, 4); sym = lookup_minimal_symbol_by_pc (this_saved_pc); frametype = tf_normal; if (sym != NULL) { if (strcmp (SYMBOL_NAME(sym), "calltrap") == 0) frametype = tf_trap; else if (strncmp (SYMBOL_NAME(sym), "Xresume", 7) == 0) frametype = tf_interrupt; else if (strcmp (SYMBOL_NAME(sym), "Xsyscall") == 0) frametype = tf_syscall; } switch (frametype) { case tf_normal: return (this_saved_pc); #define oEIP offsetof(struct trapframe, tf_eip) case tf_trap: return (read_memory_integer (fr->frame + 8 + oEIP, 4)); case tf_interrupt: return (read_memory_integer (fr->frame + 16 + oEIP, 4)); case tf_syscall: return (read_memory_integer (fr->frame + 8 + oEIP, 4)); #undef oEIP } } static CORE_ADDR ksym_lookup (name) const char *name; { struct minimal_symbol *sym; sym = lookup_minimal_symbol (name, NULL, NULL); if (sym == NULL) error ("kernel symbol `%s' not found.", name); return SYMBOL_VALUE_ADDRESS (sym); } static struct proc * curProc () { struct proc *p; CORE_ADDR addr = ksym_lookup ("gd_curproc") + prv_start; if (kvread (addr, &p)) error ("cannot read proc pointer at %x\n", addr); return p; } /* * Set the process context to that of the proc structure at * system address paddr. */ static int set_proc_context (paddr) CORE_ADDR paddr; { struct proc p; if (paddr < kernel_start) return (1); cur_proc = (struct proc *)paddr; #ifdef notyet set_kernel_boundaries (cur_proc); #endif /* Fetch all registers from core file */ target_fetch_registers (-1); /* Now, set up the frame cache, and print the top of stack */ flush_cached_frames (); set_current_frame (create_new_frame (read_fp (), read_pc ())); select_frame (get_current_frame (), 0); return (0); } /* Discard all vestiges of any previous core file and mark data and stack spaces as empty. */ /* ARGSUSED */ static void kcore_close (quitting) int quitting; { inferior_pid = 0; /* Avoid confusion from thread stuff */ if (core_kd) { kvm_close (core_kd); free (core_file); core_file = NULL; core_kd = -1; } } /* This routine opens and sets up the core file bfd */ static void kcore_open (filename, from_tty) char *filename; int from_tty; { const char *p; struct cleanup *old_chain; char buf[256], *cp; int ontop; CORE_ADDR addr; struct pcb pcb; target_preopen (from_tty); unpush_target (&kcore_ops); if (!filename) { /*error (core_kd?*/ error ( (core_kd >= 0)? "No core file specified. (Use `detach' to stop debugging a core file.)" : "No core file specified."); } filename = tilde_expand (filename); if (filename[0] != '/') { cp = concat (current_directory, "/", filename, NULL); free (filename); filename = cp; } old_chain = make_cleanup (free, filename); /* * gdb doesn't really do anything if the exec-file couldn't * be opened (in that case exec_bfd is NULL). Usually that's * no big deal, but kvm_open needs the exec-file's name, * which results in dereferencing a NULL pointer, a real NO-NO ! * So, check here if the open of the exec-file succeeded. */ if (exec_bfd == NULL) /* the open failed */ error ("kgdb could not open the exec-file, please check the name you used !"); core_kd = kvm_open (exec_bfd->filename, filename, NULL, kernel_writablecore? O_RDWR : O_RDONLY, "kgdb: "); if (core_kd < 0) perror_with_name (filename); /* Looks semi-reasonable. Toss the old core file and work on the new. */ discard_cleanups (old_chain); /* Don't free filename any more */ core_file = filename; ontop = !push_target (&kcore_ops); kernel_start = bfd_get_start_address (exec_bfd); /* XXX */ /* print out the panic string if there is one */ if (kvread (ksym_lookup ("panicstr"), &addr) == 0 && addr != 0 && target_read_memory (addr, buf, sizeof (buf)) == 0) { for (cp = buf; cp < &buf[sizeof (buf)] && *cp; cp++) if (!isascii (*cp) || (!isprint (*cp) && !isspace (*cp))) *cp = '?'; *cp = '\0'; if (buf[0] != '\0') printf ("panicstr: %s\n", buf); } /* Print all the panic messages if possible. */ if (symfile_objfile != NULL) { printf ("panic messages:\n---\n"); snprintf (buf, sizeof buf, "/sbin/dmesg -N %s -M %s | \ /usr/bin/awk '/^(panic:|Fatal trap) / { printing = 1 } \ { if (printing) print $0 }'", symfile_objfile->name, filename); fflush(stdout); system (buf); printf ("---\n"); } if (!ontop) { warning ("you won't be able to access this core file until you terminate\n\ your %s; do ``info files''", target_longname); return; } /* we may need this later */ cur_proc = (struct proc *)curProc (); /* Now, set up the frame cache, and print the top of stack */ flush_cached_frames (); set_current_frame (create_new_frame (read_fp (), read_pc ())); select_frame (get_current_frame (), 0); print_stack_frame (selected_frame, selected_frame_level, 1); } static void kcore_detach (args, from_tty) char *args; int from_tty; { if (args) error ("Too many arguments"); unpush_target (&kcore_ops); reinit_frame_cache (); if (from_tty) printf_filtered ("No kernel core file now.\n"); } /* Get the registers out of a core file. This is the machine- independent part. Fetch_core_registers is the machine-dependent part, typically implemented in the xm-file for each architecture. */ /* We just get all the registers, so we don't use regno. */ /* ARGSUSED */ static void get_kcore_registers (regno) int regno; { struct user *uaddr; /* find the pcb for the current process */ if (cur_proc == NULL || kvread (&cur_proc->p_addr, &uaddr)) error ("cannot read u area ptr for proc at %#x", cur_proc); if (read_pcb (core_kd, (CORE_ADDR)&uaddr->u_pcb) < 0) error ("cannot read pcb at %#x", &uaddr->u_pcb); } static void kcore_files_info (t) struct target_ops *t; { printf ("\t`%s'\n", core_file); } static int kcore_xfer_kmem (memaddr, myaddr, len, write, target) CORE_ADDR memaddr; char *myaddr; int len; int write; struct target_ops *target; { int ns; int nu; if (memaddr >= (CORE_ADDR)VM_MAXUSER_ADDRESS) nu = 0; else { nu = xfer_umem (memaddr, myaddr, len, write); if (nu <= 0) return (0); if (nu == len) return (nu); memaddr += nu; if (memaddr != (CORE_ADDR)VM_MAXUSER_ADDRESS) return (nu); myaddr += nu; len -= nu; } ns = (write ? kvm_write : kvm_read) (core_kd, memaddr, myaddr, len); if (ns < 0) ns = 0; return (nu + ns); } static int xfer_umem (memaddr, myaddr, len, write) CORE_ADDR memaddr; char *myaddr; int len; int write; /* ignored */ { int n; struct proc proc; if (cur_proc == NULL || kvread (cur_proc, &proc)) error ("cannot read proc at %#x", cur_proc); n = kvm_uread (core_kd, &proc, memaddr, myaddr, len) ; if (n < 0) return 0; return n; } #define KERNOFF ((unsigned)KERNBASE) #define INKERNEL(x) ((x) >= KERNOFF) static CORE_ADDR sbr; static CORE_ADDR curpcb; static int found_pcb; static int devmem; static int kfd; static struct pcb pcb; static void set_proc_cmd (arg, from_tty) char *arg; int from_tty; { CORE_ADDR paddr; struct kinfo_proc *kp; int cnt = 0; if (!arg) error_no_arg ("proc address for new current process"); if (!kernel_debugging) error ("not debugging kernel"); paddr = (CORE_ADDR)parse_and_eval_address (arg); /* assume it's a proc pointer if it's in the kernel */ if (paddr >= kernel_start) { if (set_proc_context(paddr)) error("invalid proc address"); } else { kp = kvm_getprocs(core_kd, KERN_PROC_PID, paddr, &cnt); if (!cnt) error("invalid pid"); if (set_proc_context((CORE_ADDR)kp->kp_eproc.e_paddr)) error("invalid proc address"); } } static void set_cpu_cmd (arg, from_tty) char *arg; int from_tty; { CORE_ADDR paddr; struct kinfo_proc *kp; int cpu, cfd; if (!arg) error_no_arg ("cpu number"); if (!kernel_debugging) error ("not debugging kernel"); if (!ncpus) error ("not debugging SMP kernel"); cpu = (int)parse_and_eval_address (arg); if (cpu < 0 || cpu > ncpus) error ("cpu number out of range"); cpuid = cpu; cfd = core_kd; curpcb = kvtophys(cfd, ksym_lookup ("gd_curpcb") + prv_start); physrd (cfd, curpcb, (char*)&curpcb, sizeof curpcb); if (!devmem) paddr = ksym_lookup ("dumppcb") - KERNOFF; else paddr = kvtophys (cfd, curpcb); read_pcb (cfd, paddr); printf ("initial pcb at %lx\n", (unsigned long)paddr); if ((cur_proc = curProc())) target_fetch_registers (-1); /* Now, set up the frame cache, and print the top of stack */ flush_cached_frames (); set_current_frame (create_new_frame (read_fp (), read_pc ())); select_frame (get_current_frame (), 0); print_stack_frame (selected_frame, selected_frame_level, 1); } /* substitutes for the stuff in libkvm which doesn't work */ /* most of this was taken from the old kgdb */ /* we don't need all this stuff, but the call should look the same */ static int kvm_open (efile, cfile, sfile, perm, errout) const char *efile; char *cfile; char *sfile; /* makes this kvm_open more compatible to the one in libkvm */ int perm; char *errout; /* makes this kvm_open more compatible to the one in libkvm */ { struct stat stb; int cfd; CORE_ADDR paddr; if ((cfd = open (cfile, perm, 0)) < 0) return (cfd); fstat (cfd, &stb); if ((stb.st_mode & S_IFMT) == S_IFCHR && stb.st_rdev == makedev (2, 0)) { devmem = 1; kfd = open ("/dev/kmem", perm, 0); } if (lookup_minimal_symbol("mp_ncpus", NULL, NULL)) { physrd(cfd, ksym_lookup("mp_ncpus") - KERNOFF, (char*)&ncpus, sizeof(ncpus)); prv_space = ksym_lookup("SMP_prvspace"); prv_space_size = (int)ksym_lookup("gd_idlestack_top"); printf ("SMP %d cpus\n", ncpus); } else { ncpus = 0; prv_space = 0; prv_space_size = 0; } cpuid = 0; physrd (cfd, ksym_lookup ("IdlePTD") - KERNOFF, (char*)&sbr, sizeof sbr); printf ("IdlePTD %lu\n", (unsigned long)sbr); curpcb = kvtophys(cfd, ksym_lookup ("gd_curpcb") + prv_start); physrd (cfd, curpcb, (char*)&curpcb, sizeof curpcb); found_pcb = 1; /* for vtophys */ if (!devmem) paddr = ksym_lookup ("dumppcb") - KERNOFF; else paddr = kvtophys (cfd, curpcb); read_pcb (cfd, paddr); printf ("initial pcb at %lx\n", (unsigned long)paddr); return (cfd); } static int kvm_close (fd) int fd; { return (close (fd)); } static int kvm_write (core_kd, memaddr, myaddr, len) int core_kd; CORE_ADDR memaddr; char *myaddr; { int cc; if (devmem) { if (kfd > 0) { /* * Just like kvm_read, only we write. */ errno = 0; if (lseek (kfd, (off_t)memaddr, 0) < 0 && errno != 0) { error ("kvm_write:invalid address (%x)", memaddr); return (0); } cc = write (kfd, myaddr, len); if (cc < 0) { error ("kvm_write:write failed"); return (0); } else if (cc < len) error ("kvm_write:short write"); return (cc); } else return (0); } else { printf ("kvm_write not implemented for dead kernels\n"); return (0); } /* NOTREACHED */ } static int kvm_read (core_kd, memaddr, myaddr, len) int core_kd; CORE_ADDR memaddr; char *myaddr; { return (kernel_core_file_hook (core_kd, memaddr, myaddr, len)); } static int kvm_uread (core_kd, p, memaddr, myaddr, len) int core_kd; register struct proc *p; CORE_ADDR memaddr; char *myaddr; int len; { register char *cp; char procfile[MAXPATHLEN]; ssize_t amount; int fd; if (devmem) { sprintf (procfile, "/proc/%d/mem", p->p_pid); fd = open (procfile, O_RDONLY, 0); if (fd < 0) { error ("cannot open %s", procfile); close (fd); return (0); } cp = myaddr; while (len > 0) { errno = 0; if (lseek (fd, (off_t)memaddr, 0) == -1 && errno != 0) { error ("invalid address (%x) in %s", memaddr, procfile); break; } amount = read (fd, cp, len); if (amount < 0) { error ("error reading %s", procfile); break; } if (amount == 0) { error ("EOF reading %s", procfile); break; } cp += amount; memaddr += amount; len -= amount; } close (fd); return ((ssize_t) (cp - myaddr)); } else return (kernel_core_file_hook (core_kd, memaddr, myaddr, len)); } static struct kinfo_proc kp; /* * try to do what kvm_proclist in libkvm would do */ static int kvm_proclist (cfd, pid, p, cnt) int cfd, pid, *cnt; struct proc *p; { struct proc lp; for (; p != NULL; p = lp.p_list.le_next) { if (!kvm_read(cfd, (CORE_ADDR)p, (char *)&lp, sizeof (lp))) return (0); if (lp.p_pid != pid) continue; kp.kp_eproc.e_paddr = p; *cnt = 1; return (1); } *cnt = 0; return (0); } /* * try to do what kvm_deadprocs in libkvm would do */ static struct kinfo_proc * kvm_deadprocs (cfd, pid, cnt) int cfd, pid, *cnt; { CORE_ADDR allproc, zombproc; struct proc *p; allproc = ksym_lookup("allproc"); if (kvm_read(cfd, allproc, (char *)&p, sizeof (p)) == 0) return (NULL); kvm_proclist (cfd, pid, p, cnt); if (!*cnt) { zombproc = ksym_lookup("zombproc"); if (kvm_read(cfd, zombproc, (char *)&p, sizeof (p)) == 0) return (NULL); kvm_proclist (cfd, pid, p, cnt); } return (&kp); } /* * try to do what kvm_getprocs in libkvm would do */ static struct kinfo_proc * kvm_getprocs (cfd, op, proc, cnt) int cfd, op, *cnt; CORE_ADDR proc; { int mib[4], size; *cnt = 0; /* assume it's a pid */ if (devmem) { /* "live" kernel, use sysctl */ mib[0] = CTL_KERN; mib[1] = KERN_PROC; mib[2] = KERN_PROC_PID; mib[3] = (int)proc; size = sizeof (kp); if (sysctl (mib, 4, &kp, &size, NULL, 0) < 0) { perror("sysctl"); *cnt = 0; return (NULL); } if (!size) *cnt = 0; else *cnt = 1; return (&kp); } else return (kvm_deadprocs (cfd, (int)proc, cnt)); } static int physrd (cfd, addr, dat, len) int cfd; u_int addr; char *dat; int len; { if (lseek (cfd, (off_t)addr, L_SET) == -1) return (-1); return (read (cfd, dat, len)); } static CORE_ADDR kvtophys (fd, addr) int fd; CORE_ADDR addr; { CORE_ADDR v; unsigned int pte; static CORE_ADDR PTD = -1; CORE_ADDR current_ptd; /* * We may no longer have a linear system page table... * * Here's the scoop. IdlePTD contains the physical address * of a page table directory that always maps the kernel. * IdlePTD is in memory that is mapped 1-to-1, so we can * find it easily given its 'virtual' address from ksym_lookup(). * For hysterical reasons, the value of IdlePTD is stored in sbr. * * To look up a kernel address, we first convert it to a 1st-level * address and look it up in IdlePTD. This gives us the physical * address of a page table page; we extract the 2nd-level part of * VA and read the 2nd-level pte. Finally, we add the offset part * of the VA into the physical address from the pte and return it. * * User addresses are a little more complicated. If we don't have * a current PCB from read_pcb(), we use PTD, which is the (fixed) * virtual address of the current ptd. Since it's NOT in 1-to-1 * kernel space, we must look it up using IdlePTD. If we do have * a pcb, we get the ptd from pcb_ptd. */ if (INKERNEL (addr)) current_ptd = sbr; else if (found_pcb == 0) { if (PTD == -1) PTD = kvtophys (fd, ksym_lookup ("PTD")); current_ptd = PTD; } else current_ptd = pcb.pcb_cr3; /* * Read the first-level page table (ptd). */ v = current_ptd + ( (unsigned)addr >> PDRSHIFT) * sizeof pte; if (physrd (fd, v, (char *)&pte, sizeof pte) < 0 || (pte&PG_V) == 0) return (~0); if (pte & PG_PS) { /* * No second-level page table; ptd describes one 4MB page. * (We assume that the kernel wouldn't set PG_PS without enabling * it cr0, and that the kernel doesn't support 36-bit physical * addresses). */ #define PAGE4M_MASK (NBPDR - 1) #define PG_FRAME4M (~PAGE4M_MASK) addr = (pte & PG_FRAME4M) + (addr & PAGE4M_MASK); } else { /* * Read the second-level page table. */ v = (pte&PG_FRAME) + ((addr >> PAGE_SHIFT)&(NPTEPG-1)) * sizeof pte; if (physrd (fd, v, (char *) &pte, sizeof (pte)) < 0 || (pte&PG_V) == 0) return (~0); addr = (pte & PG_FRAME) + (addr & PAGE_MASK); } #if 0 printf ("vtophys (%x) -> %x\n", oldaddr, addr); #endif return (addr); } static int read_pcb (fd, uaddr) int fd; CORE_ADDR uaddr; { int i; int noreg; CORE_ADDR nuaddr = uaddr; /* need this for the `proc' command to work */ if (INKERNEL(uaddr)) nuaddr = kvtophys(fd, uaddr); if (physrd (fd, nuaddr, (char *)&pcb, sizeof pcb) < 0) { error ("cannot read pcb at %x\n", uaddr); return (-1); } /* * get the register values out of the sys pcb and * store them where `read_register' will find them. */ /* * XXX many registers aren't available. * XXX for the non-core case, the registers are stale - they are for * the last context switch to the debugger. * XXX gcc's register numbers aren't all #defined in tm-i386.h. */ noreg = 0; for (i = 0; i < 3; ++i) /* eax,ecx,edx */ supply_register (i, (char *)&noreg); supply_register (3, (char *)&pcb.pcb_ebx); supply_register (SP_REGNUM, (char *)&pcb.pcb_esp); supply_register (FP_REGNUM, (char *)&pcb.pcb_ebp); supply_register (6, (char *)&pcb.pcb_esi); supply_register (7, (char *)&pcb.pcb_edi); supply_register (PC_REGNUM, (char *)&pcb.pcb_eip); for (i = 9; i < 14; ++i) /* eflags, cs, ss, ds, es, fs */ supply_register (i, (char *)&noreg); supply_register (15, (char *)&pcb.pcb_gs); /* XXX 80387 registers? */ } /* * read len bytes from kernel virtual address 'addr' into local * buffer 'buf'. Return numbert of bytes if read ok, 0 otherwise. On read * errors, portion of buffer not read is zeroed. */ static int kernel_core_file_hook (fd, addr, buf, len) int fd; CORE_ADDR addr; char *buf; int len; { int i; CORE_ADDR paddr; register char *cp; int cc; cp = buf; while (len > 0) { paddr = kvtophys (fd, addr); if (paddr == ~0) { memset (buf, '\000', len); break; } /* we can't read across a page boundary */ i = min (len, PAGE_SIZE - (addr & PAGE_MASK)); if ( (cc = physrd (fd, paddr, cp, i)) <= 0) { memset (cp, '\000', len); return (cp - buf); } cp += cc; addr += cc; len -= cc; } return (cp - buf); } static struct target_ops kcore_ops; void _initialize_kcorelow() { kcore_ops.to_shortname = "kcore"; kcore_ops.to_longname = "Kernel core dump file"; kcore_ops.to_doc = "Use a core file as a target. Specify the filename of the core file."; kcore_ops.to_open = kcore_open; kcore_ops.to_close = kcore_close; kcore_ops.to_attach = find_default_attach; kcore_ops.to_detach = kcore_detach; kcore_ops.to_fetch_registers = get_kcore_registers; kcore_ops.to_xfer_memory = kcore_xfer_kmem; kcore_ops.to_files_info = kcore_files_info; kcore_ops.to_create_inferior = find_default_create_inferior; kcore_ops.to_stratum = kcore_stratum; kcore_ops.to_has_memory = 1; kcore_ops.to_has_stack = 1; kcore_ops.to_has_registers = 1; kcore_ops.to_magic = OPS_MAGIC; add_target (&kcore_ops); add_com ("proc", class_obscure, set_proc_cmd, "Set current process context"); add_com ("cpu", class_obscure, set_cpu_cmd, "Set current cpu"); }