freebsd-skq/lib/libkvm/kvm_i386.c
John Baldwin 7f911abe54 Add support to libkvm for reading vmcores from other architectures.
- Add a kvaddr_type to represent kernel virtual addresses instead of
  unsigned long.
- Add a struct kvm_nlist which is a stripped down version of struct nlist
  that uses kvaddr_t for n_value.
- Add a kvm_native() routine that returns true if an open kvm descriptor
  is for a native kernel and memory image.
- Add a kvm_open2() function similar to kvm_openfiles().  It drops the
  unused 'swapfile' argument and adds a new function pointer argument for
  a symbol resolving function.  Native kernels still use _fdnlist() from
  libc to resolve symbols if a resolver function is not supplied, but cross
  kernels require a resolver.
- Add a kvm_nlist2() function similar to kvm_nlist() except that it uses
  struct kvm_nlist instead of struct nlist.
- Add a kvm_read2() function similar to kvm_read() except that it uses
  kvaddr_t instead of unsigned long for the kernel virtual address.
- Add a new kvm_arch switch of routines needed by a vmcore backend.
  Each backend is responsible for implementing kvm_read2() for a given
  vmcore format.
- Use libelf to read headers from ELF kernels and cores (except for
  powerpc cores).
- Add internal helper routines for the common page offset hash table used
  by the minidump backends.
- Port all of the existing kvm backends to implement a kvm_arch switch and
  to be cross-friendly by using private constants instead of ones that
  vary by platform (e.g. PAGE_SIZE).  Static assertions are present when
  a given backend is compiled natively to ensure the private constants
  match the real ones.
- Enable all of the existing vmcore backends on all platforms.  This means
  that libkvm on any platform should be able to perform KVA translation
  and read data from a vmcore of any platform.

Tested on:	amd64, i386, sparc64 (marius)
Differential Revision:	https://reviews.freebsd.org/D3341
2015-11-27 18:58:26 +00:00

423 lines
10 KiB
C

/*-
* Copyright (c) 1989, 1992, 1993
* The Regents of the University of California. All rights reserved.
*
* This code is derived from software developed by the Computer Systems
* Engineering group at Lawrence Berkeley Laboratory under DARPA contract
* BG 91-66 and contributed to Berkeley.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#if defined(LIBC_SCCS) && !defined(lint)
#if 0
static char sccsid[] = "@(#)kvm_hp300.c 8.1 (Berkeley) 6/4/93";
#endif
#endif /* LIBC_SCCS and not lint */
/*
* i386 machine dependent routines for kvm. Hopefully, the forthcoming
* vm code will one day obsolete this module.
*/
#include <sys/param.h>
#include <sys/endian.h>
#include <stdint.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <kvm.h>
#ifdef __i386__
#include <machine/vmparam.h> /* For KERNBASE. */
#endif
#include <limits.h>
#include "kvm_private.h"
#include "kvm_i386.h"
struct vmstate {
void *PTD;
int pae;
size_t phnum;
GElf_Phdr *phdr;
};
/*
* Translate a physical memory address to a file-offset in the crash-dump.
*/
static size_t
_kvm_pa2off(kvm_t *kd, uint64_t pa, off_t *ofs)
{
struct vmstate *vm = kd->vmst;
GElf_Phdr *p;
size_t n;
if (kd->rawdump) {
*ofs = pa;
return (I386_PAGE_SIZE - (pa & I386_PAGE_MASK));
}
p = vm->phdr;
n = vm->phnum;
while (n && (pa < p->p_paddr || pa >= p->p_paddr + p->p_memsz))
p++, n--;
if (n == 0)
return (0);
*ofs = (pa - p->p_paddr) + p->p_offset;
return (I386_PAGE_SIZE - (pa & I386_PAGE_MASK));
}
static void
_i386_freevtop(kvm_t *kd)
{
struct vmstate *vm = kd->vmst;
if (vm->PTD)
free(vm->PTD);
free(vm->phdr);
free(vm);
kd->vmst = NULL;
}
static int
_i386_probe(kvm_t *kd)
{
return (_kvm_probe_elf_kernel(kd, ELFCLASS32, EM_386) &&
!_kvm_is_minidump(kd));
}
static int
_i386_initvtop(kvm_t *kd)
{
struct kvm_nlist nl[2];
i386_physaddr_t pa;
kvaddr_t kernbase;
char *PTD;
int i;
kd->vmst = (struct vmstate *)_kvm_malloc(kd, sizeof(struct vmstate));
if (kd->vmst == NULL) {
_kvm_err(kd, kd->program, "cannot allocate vm");
return (-1);
}
kd->vmst->PTD = 0;
if (kd->rawdump == 0) {
if (_kvm_read_core_phdrs(kd, &kd->vmst->phnum,
&kd->vmst->phdr) == -1)
return (-1);
}
nl[0].n_name = "kernbase";
nl[1].n_name = 0;
if (kvm_nlist2(kd, nl) != 0) {
#ifdef __i386__
kernbase = KERNBASE; /* for old kernels */
#else
_kvm_err(kd, kd->program, "cannot resolve kernbase");
return (-1);
#endif
} else
kernbase = nl[0].n_value;
nl[0].n_name = "IdlePDPT";
nl[1].n_name = 0;
if (kvm_nlist2(kd, nl) == 0) {
i386_physaddr_pae_t pa64;
if (kvm_read2(kd, (nl[0].n_value - kernbase), &pa,
sizeof(pa)) != sizeof(pa)) {
_kvm_err(kd, kd->program, "cannot read IdlePDPT");
return (-1);
}
pa = le32toh(pa);
PTD = _kvm_malloc(kd, 4 * I386_PAGE_SIZE);
for (i = 0; i < 4; i++) {
if (kvm_read2(kd, pa + (i * sizeof(pa64)), &pa64,
sizeof(pa64)) != sizeof(pa64)) {
_kvm_err(kd, kd->program, "Cannot read PDPT");
free(PTD);
return (-1);
}
pa64 = le64toh(pa64);
if (kvm_read2(kd, pa64 & I386_PG_FRAME_PAE,
PTD + (i * I386_PAGE_SIZE), I386_PAGE_SIZE) !=
I386_PAGE_SIZE) {
_kvm_err(kd, kd->program, "cannot read PDPT");
free(PTD);
return (-1);
}
}
kd->vmst->PTD = PTD;
kd->vmst->pae = 1;
} else {
nl[0].n_name = "IdlePTD";
nl[1].n_name = 0;
if (kvm_nlist2(kd, nl) != 0) {
_kvm_err(kd, kd->program, "bad namelist");
return (-1);
}
if (kvm_read2(kd, (nl[0].n_value - kernbase), &pa,
sizeof(pa)) != sizeof(pa)) {
_kvm_err(kd, kd->program, "cannot read IdlePTD");
return (-1);
}
pa = le32toh(pa);
PTD = _kvm_malloc(kd, I386_PAGE_SIZE);
if (kvm_read2(kd, pa, PTD, I386_PAGE_SIZE) != I386_PAGE_SIZE) {
_kvm_err(kd, kd->program, "cannot read PTD");
return (-1);
}
kd->vmst->PTD = PTD;
kd->vmst->pae = 0;
}
return (0);
}
static int
_i386_vatop(kvm_t *kd, kvaddr_t va, off_t *pa)
{
struct vmstate *vm;
i386_physaddr_t offset;
i386_physaddr_t pte_pa;
i386_pde_t pde;
i386_pte_t pte;
kvaddr_t pdeindex;
kvaddr_t pteindex;
size_t s;
i386_physaddr_t a;
off_t ofs;
i386_pde_t *PTD;
vm = kd->vmst;
PTD = (i386_pde_t *)vm->PTD;
offset = va & I386_PAGE_MASK;
/*
* If we are initializing (kernel page table descriptor pointer
* not yet set) then return pa == va to avoid infinite recursion.
*/
if (PTD == 0) {
s = _kvm_pa2off(kd, va, pa);
if (s == 0) {
_kvm_err(kd, kd->program,
"_i386_vatop: bootstrap data not in dump");
goto invalid;
} else
return (I386_PAGE_SIZE - offset);
}
pdeindex = va >> I386_PDRSHIFT;
pde = le32toh(PTD[pdeindex]);
if ((pde & I386_PG_V) == 0) {
_kvm_err(kd, kd->program, "_i386_vatop: pde not valid");
goto invalid;
}
if (pde & I386_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).
*/
offset = va & I386_PAGE_PS_MASK;
a = (pde & I386_PG_PS_FRAME) + offset;
s = _kvm_pa2off(kd, a, pa);
if (s == 0) {
_kvm_err(kd, kd->program,
"_i386_vatop: 4MB page address not in dump");
goto invalid;
}
return (I386_NBPDR - offset);
}
pteindex = (va >> I386_PAGE_SHIFT) & (I386_NPTEPG - 1);
pte_pa = (pde & I386_PG_FRAME) + (pteindex * sizeof(pte));
s = _kvm_pa2off(kd, pte_pa, &ofs);
if (s < sizeof(pte)) {
_kvm_err(kd, kd->program, "_i386_vatop: pte_pa not found");
goto invalid;
}
/* XXX This has to be a physical address read, kvm_read is virtual */
if (pread(kd->pmfd, &pte, sizeof(pte), ofs) != sizeof(pte)) {
_kvm_syserr(kd, kd->program, "_i386_vatop: pread");
goto invalid;
}
pte = le32toh(pte);
if ((pte & I386_PG_V) == 0) {
_kvm_err(kd, kd->program, "_kvm_kvatop: pte not valid");
goto invalid;
}
a = (pte & I386_PG_FRAME) + offset;
s = _kvm_pa2off(kd, a, pa);
if (s == 0) {
_kvm_err(kd, kd->program, "_i386_vatop: address not in dump");
goto invalid;
} else
return (I386_PAGE_SIZE - offset);
invalid:
_kvm_err(kd, 0, "invalid address (0x%jx)", (uintmax_t)va);
return (0);
}
static int
_i386_vatop_pae(kvm_t *kd, kvaddr_t va, off_t *pa)
{
struct vmstate *vm;
i386_physaddr_pae_t offset;
i386_physaddr_pae_t pte_pa;
i386_pde_pae_t pde;
i386_pte_pae_t pte;
kvaddr_t pdeindex;
kvaddr_t pteindex;
size_t s;
i386_physaddr_pae_t a;
off_t ofs;
i386_pde_pae_t *PTD;
vm = kd->vmst;
PTD = (i386_pde_pae_t *)vm->PTD;
offset = va & I386_PAGE_MASK;
/*
* If we are initializing (kernel page table descriptor pointer
* not yet set) then return pa == va to avoid infinite recursion.
*/
if (PTD == 0) {
s = _kvm_pa2off(kd, va, pa);
if (s == 0) {
_kvm_err(kd, kd->program,
"_i386_vatop_pae: bootstrap data not in dump");
goto invalid;
} else
return (I386_PAGE_SIZE - offset);
}
pdeindex = va >> I386_PDRSHIFT_PAE;
pde = le64toh(PTD[pdeindex]);
if ((pde & I386_PG_V) == 0) {
_kvm_err(kd, kd->program, "_kvm_kvatop_pae: pde not valid");
goto invalid;
}
if (pde & I386_PG_PS) {
/*
* No second-level page table; ptd describes one 2MB
* page. (We assume that the kernel wouldn't set
* PG_PS without enabling it cr0).
*/
offset = va & I386_PAGE_PS_MASK_PAE;
a = (pde & I386_PG_PS_FRAME_PAE) + offset;
s = _kvm_pa2off(kd, a, pa);
if (s == 0) {
_kvm_err(kd, kd->program,
"_i386_vatop: 2MB page address not in dump");
goto invalid;
}
return (I386_NBPDR_PAE - offset);
}
pteindex = (va >> I386_PAGE_SHIFT) & (I386_NPTEPG_PAE - 1);
pte_pa = (pde & I386_PG_FRAME_PAE) + (pteindex * sizeof(pde));
s = _kvm_pa2off(kd, pte_pa, &ofs);
if (s < sizeof(pte)) {
_kvm_err(kd, kd->program, "_i386_vatop_pae: pdpe_pa not found");
goto invalid;
}
/* XXX This has to be a physical address read, kvm_read is virtual */
if (pread(kd->pmfd, &pte, sizeof(pte), ofs) != sizeof(pte)) {
_kvm_syserr(kd, kd->program, "_i386_vatop_pae: read");
goto invalid;
}
pte = le64toh(pte);
if ((pte & I386_PG_V) == 0) {
_kvm_err(kd, kd->program, "_i386_vatop_pae: pte not valid");
goto invalid;
}
a = (pte & I386_PG_FRAME_PAE) + offset;
s = _kvm_pa2off(kd, a, pa);
if (s == 0) {
_kvm_err(kd, kd->program,
"_i386_vatop_pae: address not in dump");
goto invalid;
} else
return (I386_PAGE_SIZE - offset);
invalid:
_kvm_err(kd, 0, "invalid address (0x%jx)", (uintmax_t)va);
return (0);
}
static int
_i386_kvatop(kvm_t *kd, kvaddr_t va, off_t *pa)
{
if (ISALIVE(kd)) {
_kvm_err(kd, 0, "vatop called in live kernel!");
return (0);
}
if (kd->vmst->pae)
return (_i386_vatop_pae(kd, va, pa));
else
return (_i386_vatop(kd, va, pa));
}
int
_i386_native(kvm_t *kd)
{
#ifdef __i386__
return (1);
#else
return (0);
#endif
}
struct kvm_arch kvm_i386 = {
.ka_probe = _i386_probe,
.ka_initvtop = _i386_initvtop,
.ka_freevtop = _i386_freevtop,
.ka_kvatop = _i386_kvatop,
.ka_native = _i386_native,
};
KVM_ARCH(kvm_i386);