This is taken from the work I did on lguest64.
When killing a guest, we read the guest stack to do a nice back trace of
the guest and send it via printk to the host.
So instead of just getting an error message from the lguest launcher of:
lguest: bad read address 537012178 len 1
I also get in my dmesg:
called from [<c0405f30>] show_trace_log_lvl+0x1a/0x2f
[<c04069aa>] show_trace+0x12/0x14
[<c0406a03>] dump_stack+0x16/0x18
[<f8c746de>] lguest_dump_lg_regs+0x22/0x13c [lg]
[<f8c7131b>] lgread+0x59/0x90 [lg]
[<f8c715bd>] run_guest+0x26b/0x406 [lg]
[<f8c739be>] read+0x73/0x7d [lg]
[<c04825e9>] vfs_read+0xad/0x161
[<c0482a75>] sys_read+0x3d/0x61
[<c0404f34>] syscall_call+0x7/0xb
======================[<f8c7131b>] lgread+0x59/0x90 [lg]
Printing LG 0 regs cr3: 021eb000
EIP: 0061: [<00000000e00227d2>]
ESP: 0069:c236fe3c EFLAGS: 00010202
EAX: 00000004 EBX: e001fb20 ECX: 00000008 EDX: 000003f2
ESI: e001ee00 EDI: e001fb60 EBP: c236fea0
CR2: 1278000 lguest_data->cr2: 80011380
errcode: 0 trapnum: d
Stack Dump:
[<00000000c1042b7a>] trace_hardirqs_on+0x125/0x149
[<00000000c123b0ea>] wait_for_completion+0x90/0x98
[<00000000c123bddc>] __mutex_unlock_slowpath+0x129/0x13e
[<00000000c1048769>] unlock_cpu_hotplug+0x62/0x64
[<00000000c104b5b6>] sys_init_module+0x14e3/0x162c
[<00000000c1081d44>] do_sync_read+0xc2/0xff
[<00000000c1004f85>] restore_nocheck+0x12/0x15
[<00000000c1004f34>] syscall_call+0x7/0xb
TODO:
- Clean up a little (still has stuff from lguest64 in it).
- Perhaps make a config option or runtime switch to turn it off.
- Send to the launcher the dump instead of printk.
- make modules work too.
Also I need to change the %u of the bad read print to a %x, because
seeing 0x200227d2 is better than seeing 537012178 for addresses.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Index: linux-2.6.21-rc5-mm2/drivers/lguest/Makefile
==================================================================---
linux-2.6.21-rc5-mm2.orig/drivers/lguest/Makefile
+++ linux-2.6.21-rc5-mm2/drivers/lguest/Makefile
@@ -4,4 +4,4 @@ obj-$(CONFIG_LGUEST_GUEST) += lguest.o l
# Host requires the other files, which can be a module.
obj-$(CONFIG_LGUEST) += lg.o
lg-objs := core.o hypercalls.o page_tables.o interrupts_and_traps.o \
- segments.o io.o lguest_user.o hypervisor.o
+ segments.o io.o lguest_user.o hypervisor.o lguest_debug.o
Index: linux-2.6.21-rc5-mm2/drivers/lguest/core.c
==================================================================---
linux-2.6.21-rc5-mm2.orig/drivers/lguest/core.c
+++ linux-2.6.21-rc5-mm2/drivers/lguest/core.c
@@ -210,6 +210,28 @@ int lguest_address_ok(const struct lgues
}
/* Just like get_user, but don't let guest access lguest binary. */
+u8 lgread_u8(struct lguest *lg, u32 addr)
+{
+ u8 val = 0;
+
+ /* Don't let them access lguest binary */
+ if (!lguest_address_ok(lg, addr)
+ || get_user(val, (u32 __user *)addr) != 0)
+ kill_guest(lg, "bad read address %u", addr);
+ return val;
+}
+
+u16 lgread_u16(struct lguest *lg, u32 addr)
+{
+ u16 val = 0;
+
+ /* Don't let them access lguest binary */
+ if (!lguest_address_ok(lg, addr)
+ || get_user(val, (u32 __user *)addr) != 0)
+ kill_guest(lg, "bad read address %u", addr);
+ return val;
+}
+
u32 lgread_u32(struct lguest *lg, u32 addr)
{
u32 val = 0;
Index: linux-2.6.21-rc5-mm2/drivers/lguest/lg.h
==================================================================---
linux-2.6.21-rc5-mm2.orig/drivers/lguest/lg.h
+++ linux-2.6.21-rc5-mm2/drivers/lguest/lg.h
@@ -176,6 +176,8 @@ extern struct mutex lguest_lock;
/* core.c: */
/* Entry points in hypervisor */
const unsigned long *__lguest_default_idt_entries(void);
+u8 lgread_u8(struct lguest *lg, u32 addr);
+u16 lgread_u16(struct lguest *lg, u32 addr);
u32 lgread_u32(struct lguest *lg, u32 addr);
void lgwrite_u32(struct lguest *lg, u32 val, u32 addr);
void lgread(struct lguest *lg, void *buf, u32 addr, unsigned bytes);
@@ -238,6 +240,7 @@ int hypercall(struct lguest *info, struc
#define kill_guest(lg, fmt...) \
do { \
if (!(lg)->dead) { \
+ lguest_dump_lg_regs(lg); \
(lg)->dead = kasprintf(GFP_ATOMIC, fmt); \
if (!(lg)->dead) \
(lg)->dead = (void *)1; \
@@ -248,5 +251,11 @@ static inline unsigned long guest_pa(str
{
return vaddr - lg->page_offset;
}
+
+/* lguest_debug.c */
+void lguest_print_address(struct lguest *lg, unsigned long address);
+void lguest_dump_trace(struct lguest *lg, struct lguest_regs *regs);
+void lguest_dump_lg_regs(struct lguest *lg);
+
#endif /* __ASSEMBLY__ */
#endif /* _LGUEST_H */
Index: linux-2.6.21-rc5-mm2/drivers/lguest/lguest.c
==================================================================---
linux-2.6.21-rc5-mm2.orig/drivers/lguest/lguest.c
+++ linux-2.6.21-rc5-mm2/drivers/lguest/lguest.c
@@ -48,6 +48,14 @@ extern const char lgstart_iret[], lgend_
extern asmlinkage void lguest_maybe_init(void);
extern void lguest_iret(void);
+/* For back tracing from the HV */
+extern const unsigned long kallsyms_addresses[] __attribute__((weak));
+extern const unsigned long kallsyms_num_syms __attribute__((weak));
+extern const u8 kallsyms_names[] __attribute__((weak));
+extern const u8 kallsyms_token_table[] __attribute__((weak));
+extern const u16 kallsyms_token_index[] __attribute__((weak));
+extern const unsigned long kallsyms_markers[] __attribute__((weak));
+
struct lguest_data lguest_data = {
.hcall_status = { [0 ... LHCALL_RING_SIZE-1] = 0xFF },
.noirq_start = (u32)lguest_noirq_start,
@@ -56,6 +64,7 @@ struct lguest_data lguest_data = {
};
struct lguest_device_desc *lguest_devices;
static __initdata const struct lguest_boot_info *boot = __va(0);
+static struct lguest_text_ptr code_stack[2];
static int lazy_mode;
static void lguest_lazy_mode(int mode)
@@ -468,6 +477,22 @@ __init void lguest_init(void)
paravirt_ops.set_lazy_mode = lguest_lazy_mode;
paravirt_ops.wbinvd = lguest_wbinvd;
+ code_stack[0].next = __pa(&code_stack[1]);
+ code_stack[0].start = (unsigned long)_stext;
+ code_stack[0].end = (unsigned long)_etext;
+ code_stack[1].next = 0;
+ code_stack[1].start = (unsigned long)_sinittext;
+ code_stack[1].end = (unsigned long)_einittext;
+
+ lguest_data.text = __pa(&code_stack[0]);
+
+ lguest_data.kallsyms_addresses = __pa(&kallsyms_addresses);
+ lguest_data.kallsyms_num_syms = kallsyms_num_syms;
+ lguest_data.kallsyms_names = __pa(&kallsyms_names);
+ lguest_data.kallsyms_token_table = __pa(&kallsyms_token_table);
+ lguest_data.kallsyms_token_index = __pa(&kallsyms_token_index);
+ lguest_data.kallsyms_markers = __pa(&kallsyms_markers);
+
hcall(LHCALL_LGUEST_INIT, __pa(&lguest_data), 0, 0);
strncpy(boot_command_line, boot->cmdline, COMMAND_LINE_SIZE);
Index: linux-2.6.21-rc5-mm2/drivers/lguest/lguest_debug.c
==================================================================--- /dev/null
+++ linux-2.6.21-rc5-mm2/drivers/lguest/lguest_debug.c
@@ -0,0 +1,458 @@
+/*
+ lguest debug utils. Modified from various other parts of Linux.
+ What was modified is Copyright 2007 Steven Rostedt, Red Hat
+
+ 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+*/
+#include <linux/module.h>
+#include <linux/sched.h>
+#include <linux/freezer.h>
+#include <linux/kallsyms.h>
+#include <asm/paravirt.h>
+#include <asm/pgtable.h>
+#include <asm/uaccess.h>
+#include "lg.h"
+
+#if 1
+#define DEBUG_DUMP_STACK() dump_stack()
+#else
+#define DEBUG_DUMP_STACK() do { } while(0)
+#endif
+
+int lguest_debug;
+
+static DEFINE_SPINLOCK(lgdebug_print_lock);
+#define LGDEBUG_BUF_SIZ 1024
+static char lgdebug_print_buf[LGDEBUG_BUF_SIZ];
+
+void lgdebug_vprint(const char *fmt, va_list ap)
+{
+ unsigned long flags;
+
+ if (!lguest_debug)
+ return;
+
+ spin_lock_irqsave(&lgdebug_print_lock, flags);
+ vsnprintf(lgdebug_print_buf, LGDEBUG_BUF_SIZ-1, fmt, ap);
+ printk("%s", lgdebug_print_buf);
+ spin_unlock_irqrestore(&lgdebug_print_lock, flags);
+}
+
+void lgdebug_print(const char *fmt, ...)
+{
+ va_list ap;
+
+ if (!lguest_debug)
+ return;
+
+ /* irq save? */
+ va_start(ap, fmt);
+ lgdebug_vprint(fmt, ap);
+ va_end(ap);
+}
+
+void lgdebug_lprint(unsigned flags, const char *fmt, ...)
+{
+ va_list ap;
+
+ if (!(lguest_debug & flags))
+ return;
+
+ va_start(ap, fmt);
+ lgdebug_vprint(fmt, ap);
+ va_end(ap);
+}
+
+void lgdebug_lvprint(unsigned flags, const char *fmt, va_list ap)
+{
+ if (!(lguest_debug & flags))
+ return;
+
+ lgdebug_vprint(fmt, ap);
+}
+
+void lguest_dump_lg_regs(struct lguest *lg)
+{
+ struct lguest_regs *regs = lg->regs;
+ struct pgdir *pgdir = &lg->pgdirs[lg->pgdidx];
+ unsigned long cr2, data_cr2;
+
+ printk("called from ");
+ DEBUG_DUMP_STACK();
+ print_ip_sym((unsigned long)__builtin_return_address(0));
+ printk("Printing LG %d regs cr3: %08lx\n",
+ lg->guestid, pgdir->cr3);
+ printk("EIP: %04x: ", regs->cs & 0xffff);
+ lguest_print_address(lg, regs->eip);
+ printk("ESP: %04x:%08x EFLAGS: %08x\n", regs->ss, regs->esp,
+ regs->eflags);
+ printk("EAX: %08x EBX: %08x ECX: %08x EDX: %08x\n",
+ regs->eax, regs->ebx, regs->ecx, regs->edx);
+ printk("ESI: %08x EDI: %08x EBP: %08x\n",
+ regs->esi, regs->edi, regs->ebp);
+#if 0
+ printk("FS %04llx: Base: %016lx Desc: %08lx\n",
+ regs->fs, vcpu->guest_fs_base, vcpu->guest_fs_desc);
+#endif
+
+ if (get_user(data_cr2, &lg->lguest_data->cr2))
+ data_cr2 = 0;
+ cr2 = read_cr2();
+ printk(" CR2: %lx lguest_data->cr2: %lx\n", cr2, data_cr2);
+
+ printk("errcode: %x trapnum: %x\n",
+ regs->errcode, regs->trapnum);
+
+ lguest_dump_trace(lg, regs);
+}
+
+struct guest_ksym_stuff {
+ unsigned long *addresses;
+ unsigned long num_syms;
+ u8 *names;
+ u8 *token_table;
+ u16 *token_index;
+ unsigned long *markers;
+};
+
+static struct lguest_text_ptr *get_text_segs(struct lguest *lg)
+{
+ struct lguest_text_ptr *segs, **p;
+ struct lguest_text_ptr *g;
+ unsigned long addr;
+ int i;
+
+ if (!lg->lguest_data)
+ return NULL;
+
+ addr = lgread_u32(lg, (unsigned long)&lg->lguest_data->text);
+ if (!addr)
+ return NULL;
+
+ g = (struct lguest_text_ptr*)addr;
+
+ p = &segs;
+
+ /* only allow for 10 segs */
+ for (i=0; i < 10; i++) {
+ *p = kmalloc(sizeof(*segs), GFP_KERNEL);
+ if (!*p)
+ goto free_me;
+ (*p)->start = lgread_u32(lg, (u32)&g->start);
+ (*p)->end = lgread_u32(lg, (u32)&g->end);
+ addr = lgread_u32(lg, (u32)&g->next);
+ p = (struct lguest_text_ptr**)&((*p)->next);
+ if (!addr)
+ break;
+ g = (struct lguest_text_ptr*)addr;
+ }
+ *p = NULL;
+
+ return segs;
+
+free_me:
+ while (segs) {
+ g = (struct lguest_text_ptr*)segs->next;
+ kfree(segs);
+ segs = g;
+ }
+ return NULL;
+}
+
+static int is_text_seg(struct lguest_text_ptr *segs, unsigned long addr)
+{
+ while (segs) {
+ if (addr >= segs->start &&
+ addr <= segs->end)
+ return 1;
+ segs = (struct lguest_text_ptr*)segs->next;
+ }
+ return 0;
+}
+
+static void put_text_segs(struct lguest_text_ptr *segs)
+{
+ struct lguest_text_ptr *p;
+
+ while (segs) {
+ p = (struct lguest_text_ptr*)segs->next;
+ kfree(segs);
+ segs = p;
+ }
+}
+
+static unsigned int expand_symbol(struct lguest *lg,
+ struct guest_ksym_stuff *kstuff,
+ unsigned int off, char *result)
+{
+ int len, skipped_first = 0;
+ const u8 *tptr, *data;
+
+ /* get the compressed symbol length from the first symbol byte */
+ data = &kstuff->names[off];
+
+ len = lgread_u8(lg, (unsigned long)data);
+
+ data++;
+
+ /* update the offset to return the offset for the next symbol on
+ * the compressed stream */
+ off += len + 1;
+
+ /* for every byte on the compressed symbol data, copy the table
+ entry for that byte */
+ while(len) {
+ u8 idx;
+ u16 tok;
+ idx = lgread_u8(lg, (unsigned long)data);
+ tok = lgread_u16(lg, (unsigned long)(&kstuff->token_index[idx]));
+ tptr = &kstuff->token_table[ tok ];
+ data++;
+ len--;
+
+ idx = lgread_u8(lg, (unsigned long)tptr);
+ while (idx) {
+ if(skipped_first) {
+ *result = idx;
+ result++;
+ } else
+ skipped_first = 1;
+ tptr++;
+ idx = lgread_u8(lg, (unsigned long)tptr);
+ }
+ }
+
+ *result = '\0';
+
+ /* return to offset to the next symbol */
+ return off;
+}
+
+static unsigned long get_symbol_pos(struct lguest *lg,
+ struct guest_ksym_stuff *kstuff,
+ unsigned long addr,
+ unsigned long *symbolsize,
+ unsigned long *offset)
+{
+ unsigned long symbol_start = 0, symbol_end = 0;
+ unsigned long i, low, high, mid;
+
+ /* do a binary search on the sorted kallsyms_addresses array */
+ low = 0;
+ high = kstuff->num_syms;
+
+ while (high - low > 1) {
+ mid = (low + high) / 2;
+ if (kstuff->addresses[mid] <= addr)
+ low = mid;
+ else
+ high = mid;
+ }
+
+ /*
+ * search for the first aliased symbol. Aliased
+ * symbols are symbols with the same address
+ */
+ while (low && kstuff->addresses[low-1] ==
kstuff->addresses[low])
+ --low;
+
+ symbol_start = kstuff->addresses[low];
+
+ /* Search for next non-aliased symbol */
+ for (i = low + 1; i < kstuff->num_syms; i++) {
+ if (kstuff->addresses[i] > symbol_start) {
+ symbol_end = kstuff->addresses[i];
+ break;
+ }
+ }
+
+ /* if we found no next symbol, we use the end of the section */
+ if (!symbol_end) {
+ return (unsigned long)(-1UL);
+#if 0
+ if (is_kernel_inittext(addr))
+ symbol_end = (unsigned long)_einittext;
+ else if (all_var)
+ symbol_end = (unsigned long)_end;
+ else
+ symbol_end = (unsigned long)_etext;
+#endif
+ }
+
+ *symbolsize = symbol_end - symbol_start;
+ *offset = addr - symbol_start;
+
+ return low;
+}
+
+static int is_ksym_addr(struct lguest *lg,
+ unsigned long addr)
+{
+ /* need to look up the segs */
+ return 1;
+}
+
+static unsigned int get_symbol_offset(struct lguest *lg,
+ struct guest_ksym_stuff *kstuff,
+ unsigned long pos)
+{
+ const u8 *name;
+ int i;
+ unsigned long idx;
+
+ idx = lgread_u32(lg, (unsigned long)&kstuff->markers[pos>>8]);
+
+ /* use the closest marker we have. We have markers every 256 positions,
+ * so that should be close enough */
+ name = &kstuff->names[ idx ];
+
+ /* sequentially scan all the symbols up to the point we're searching for.
+ * Every symbol is stored in a [<len>][<len> bytes of data]
format, so we
+ * just need to add the len to the current pointer for every symbol we
+ * wish to skip */
+ for(i = 0; i < (pos&0xFF); i++) {
+ u8 c;
+ c = lgread_u8(lg, (unsigned long)name);
+ name = name + c + 1;
+ }
+
+ return name - kstuff->names;
+}
+
+static const char *lguest_syms_lookup(struct lguest *lg,
+ unsigned long addr,
+ unsigned long *symbolsize,
+ unsigned long *offset,
+ char **modname, char *namebuf)
+{
+ struct lguest_data *data = lg->lguest_data;
+ struct guest_ksym_stuff kstuff;
+ const char *msym;
+ unsigned long *ptr;
+ int i;
+
+ kstuff.addresses = (unsigned long*)lgread_u32(lg,
+ (unsigned long)&data->kallsyms_addresses);
+ kstuff.num_syms = lgread_u32(lg, (unsigned
long)&data->kallsyms_num_syms);
+ kstuff.names = (u8*)lgread_u32(lg, (unsigned
long)&data->kallsyms_names);
+ kstuff.token_table = (u8*)lgread_u32(lg,
+ (unsigned long)&data->kallsyms_token_table);
+ kstuff.token_index = (u16*)lgread_u32(lg,
+ (unsigned long)&data->kallsyms_token_index);
+ kstuff.markers = (unsigned long*)lgread_u32(lg,
+ (unsigned long)&data->kallsyms_markers);
+
+ if (!kstuff.addresses || !kstuff.num_syms || !kstuff.names ||
+ !kstuff.token_table || !kstuff.token_index || !kstuff.markers)
+ return NULL;
+
+ /* FIXME: Validate all the kstuff here!! */
+
+ ptr = kmalloc(sizeof(unsigned long)*kstuff.num_syms, GFP_KERNEL);
+ if (!ptr)
+ return NULL;
+
+ for (i=0; i < kstuff.num_syms; i++) {
+ /* FIXME: do this better! */
+ ptr[i] = lgread_u32(lg, (unsigned long)&kstuff.addresses[i]);
+ if (i && ptr[i] < ptr[i-1]) {
+ kill_guest(lg, "bad kallsyms table\n");
+ kstuff.addresses = ptr;
+ goto out;
+ }
+ }
+ kstuff.addresses = ptr;
+
+ namebuf[KSYM_NAME_LEN] = 0;
+ namebuf[0] = 0;
+
+ if (is_ksym_addr(lg, addr)) {
+ unsigned long pos;
+
+ pos = get_symbol_pos(lg, &kstuff, addr, symbolsize, offset);
+ if (pos == (unsigned long)(-1UL))
+ goto out;
+
+ /* Grab name */
+ expand_symbol(lg, &kstuff,
+ get_symbol_offset(lg, &kstuff, pos), namebuf);
+ *modname = NULL;
+ kfree(kstuff.addresses);
+ return namebuf;
+ }
+
+ /* see if it's in a module */
+ msym = module_address_lookup(addr, symbolsize, offset, modname);
+ if (msym) {
+ kfree(kstuff.addresses);
+ return strncpy(namebuf, msym, KSYM_NAME_LEN);
+ }
+
+out:
+ kfree(kstuff.addresses);
+ return NULL;
+}
+
+void lguest_print_address(struct lguest *lg, unsigned long address)
+{
+ unsigned long offset = 0, symsize;
+ const char *symname;
+ char *modname;
+ char *delim = ":";
+ char namebuf[KSYM_NAME_LEN+1];
+
+ symname = lguest_syms_lookup(lg, address, &symsize, &offset,
+ &modname, namebuf);
+ if (!symname) {
+ printk(" [<%016lx>]\n", address);
+ return;
+ }
+ if (!modname)
+ modname = delim = "";
+ printk(" [<%016lx>] %s%s%s%s+0x%lx/0x%lx\n",
+ address, delim, modname, delim, symname, offset, symsize);
+
+}
+
+void lguest_dump_trace(struct lguest *lg, struct lguest_regs *regs)
+{
+ unsigned long stack = regs->esp;
+ unsigned long stack_end = (regs->esp & PAGE_MASK) + PAGE_SIZE;
+ unsigned long addr;
+ struct lguest_text_ptr *segs;
+
+ printk("Stack Dump:\n");
+ if (stack < lg->page_offset) {
+ printk(" <USER STACK>\n");
+ goto out;
+ }
+
+ segs = get_text_segs(lg);
+ if (!segs)
+ return;
+
+ for (; stack < stack_end; stack += sizeof(stack)) {
+ addr = lgread_u32(lg, guest_pa(lg, stack));
+ if (is_text_seg(segs, addr)) {
+ lguest_print_address(lg, addr);
+ }
+ }
+
+ put_text_segs(segs);
+
+out:
+ printk("=======\n");
+}
+
Index: linux-2.6.21-rc5-mm2/include/linux/lguest.h
==================================================================---
linux-2.6.21-rc5-mm2.orig/include/linux/lguest.h
+++ linux-2.6.21-rc5-mm2/include/linux/lguest.h
@@ -54,6 +54,13 @@ struct hcall_ring
u32 eax, edx, ebx, ecx;
};
+/* for kallsyms dumping */
+struct lguest_text_ptr {
+ unsigned long next; /* guest pa address of next pointer */
+ unsigned long start;
+ unsigned long end;
+};
+
/* All the good stuff happens here: guest registers it with LGUEST_INIT */
struct lguest_data
{
@@ -76,6 +83,16 @@ struct lguest_data
/* ID of this guest (used by network driver to set ethernet address) */
u16 guestid;
+ /* pa address of lguest_text_ptr addresses */
+ unsigned long text;
+/* If the kernel has kallsyms, we can use it to do backtraces of a guest */
+ unsigned long kallsyms_addresses;
+ unsigned long kallsyms_num_syms;
+ unsigned long kallsyms_names;
+ unsigned long kallsyms_token_table;
+ unsigned long kallsyms_token_index;
+ unsigned long kallsyms_markers;
+
/* Fields initialized by the guest at boot: */
/* Instruction range to suppress interrupts even if enabled */
unsigned long noirq_start, noirq_end;