x86: Fix alternatives and kprobes to remap write-protected kernel text

Reenable kprobes and alternative patching when the kernel text is write
protected by DEBUG_RODATA

Add a general utility function to change write protected text.  The new
function remaps the code using vmap to write it and takes care of CPU
synchronization.  It also does CLFLUSH to make icache recovery faster.

There are some limitations on when the function can be used, see the
comment.

This is a newer version that also changes the paravirt_ops code.
text_poke also supports multi byte patching now.

Contains bug fixes from Zach Amsden and suggestions from Mathieu
Desnoyers.

Cc: Jan Beulich <jbeulich@novell.com>
Cc: Jeremy Fitzhardinge <jeremy@goop.org>
Cc: Mathieu Desnoyers <compudj@krystal.dyndns.org>
Cc: Zach Amsden <zach@vmware.com>
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

4 files changed, 53 insertions, 27 deletions

diff --git a/arch/i386/kernel/alternative.c b/arch/i386/kernel/alternative.c
index 0695be538de5..206ea2ca63cc 100644
--- a/arch/i386/kernel/alternative.c
+++ b/arch/i386/kernel/alternative.c
@@ -2,8 +2,12 @@
 #include <linux/sched.h>
 #include <linux/spinlock.h>
 #include <linux/list.h>
+#include <linux/kprobes.h>
+#include <linux/mm.h>
+#include <linux/vmalloc.h>
 #include <asm/alternative.h>
 #include <asm/sections.h>
+#include <asm/pgtable.h>
 
 #ifdef CONFIG_HOTPLUG_CPU
 static int smp_alt_once;
@@ -150,7 +154,7 @@ static void nop_out(void *insns, unsigned int len)
                 unsigned int noplen = len;
                 if (noplen > ASM_NOP_MAX)
                         noplen = ASM_NOP_MAX;
-                memcpy(insns, noptable[noplen], noplen);
+                text_poke(insns, noptable[noplen], noplen);
                 insns += noplen;
                 len -= noplen;
         }
@@ -202,7 +206,7 @@ static void alternatives_smp_lock(u8 **start, u8 **end, u8 *text, u8 *text_end)
                         continue;
                 if (*ptr > text_end)
                         continue;
-                **ptr = 0xf0; /* lock prefix */
+                text_poke(*ptr, ((unsigned char []){0xf0}), 1); /* add lock prefix */
         };
 }
 
@@ -360,10 +364,6 @@ void apply_paravirt(struct paravirt_patch_site *start,
                 /* Pad the rest with nops */
                 nop_out(p->instr + used, p->len - used);
         }
-
-        /* Sync to be conservative, in case we patched following
-         * instructions */
-        sync_core();
 }
 extern struct paravirt_patch_site __start_parainstructions[],
         __stop_parainstructions[];
@@ -406,3 +406,31 @@ void __init alternative_instructions(void)
         apply_paravirt(__parainstructions, __parainstructions_end);
         local_irq_restore(flags);
 }
+
+/*
+ * Warning:
+ * When you use this code to patch more than one byte of an instruction
+ * you need to make sure that other CPUs cannot execute this code in parallel.
+ * Also no thread must be currently preempted in the middle of these instructions.
+ * And on the local CPU you need to be protected again NMI or MCE handlers
+ * seeing an inconsistent instruction while you patch.
+ */
+void __kprobes text_poke(void *oaddr, unsigned char *opcode, int len)
+{
+        u8 *addr = oaddr;
+        if (!pte_write(*lookup_address((unsigned long)addr))) {
+                struct page *p[2] = { virt_to_page(addr), virt_to_page(addr+PAGE_SIZE) };
+                addr = vmap(p, 2, VM_MAP, PAGE_KERNEL);
+                if (!addr)
+                        return;
+                addr += ((unsigned long)oaddr) % PAGE_SIZE;
+        }
+        memcpy(addr, opcode, len);
+        sync_core();
+        /* Not strictly needed, but can speed CPU recovery up. Ignore cross cacheline
+           case. */
+        if (cpu_has_clflush)
+                asm("clflush (%0) " :: "r" (oaddr) : "memory");
+        if (addr != oaddr)
+                vunmap(addr);
+}
diff --git a/arch/i386/kernel/kprobes.c b/arch/i386/kernel/kprobes.c
index dde828a333c3..448a50b1324c 100644
--- a/arch/i386/kernel/kprobes.c
+++ b/arch/i386/kernel/kprobes.c
@@ -35,6 +35,7 @@
 #include <asm/cacheflush.h>
 #include <asm/desc.h>
 #include <asm/uaccess.h>
+#include <asm/alternative.h>
 
 void jprobe_return_end(void);
 
@@ -169,16 +170,12 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 
 void __kprobes arch_arm_kprobe(struct kprobe *p)
 {
-        *p->addr = BREAKPOINT_INSTRUCTION;
-        flush_icache_range((unsigned long) p->addr,
-                           (unsigned long) p->addr + sizeof(kprobe_opcode_t));
+        text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1);
 }
 
 void __kprobes arch_disarm_kprobe(struct kprobe *p)
 {
-        *p->addr = p->opcode;
-        flush_icache_range((unsigned long) p->addr,
-                           (unsigned long) p->addr + sizeof(kprobe_opcode_t));
+        text_poke(p->addr, &p->opcode, 1);
 }
 
 void __kprobes arch_remove_kprobe(struct kprobe *p)
diff --git a/arch/i386/kernel/paravirt.c b/arch/i386/kernel/paravirt.c
index 53f07a8275e3..79c167fcaee9 100644
--- a/arch/i386/kernel/paravirt.c
+++ b/arch/i386/kernel/paravirt.c
@@ -124,20 +124,28 @@ unsigned paravirt_patch_ignore(unsigned len)
         return len;
 }
 
+struct branch {
+        unsigned char opcode;
+        u32 delta;
+} __attribute__((packed));
+
 unsigned paravirt_patch_call(void *target, u16 tgt_clobbers,
                              void *site, u16 site_clobbers,
                              unsigned len)
 {
         unsigned char *call = site;
         unsigned long delta = (unsigned long)target - (unsigned long)(call+5);
+        struct branch b;
 
         if (tgt_clobbers & ~site_clobbers)
                 return len;     /* target would clobber too much for this site */
         if (len < 5)
                 return len;     /* call too long for patch site */
 
-        *call++ = 0xe8;         /* call */
-        *(unsigned long *)call = delta;
+        b.opcode = 0xe8; /* call */
+        b.delta = delta;
+        BUILD_BUG_ON(sizeof(b) != 5);
+        text_poke(call, (unsigned char *)&b, 5);
 
         return 5;
 }
@@ -150,8 +158,9 @@ unsigned paravirt_patch_jmp(void *target, void *site, unsigned len)
         if (len < 5)
                 return len;     /* call too long for patch site */
 
-        *jmp++ = 0xe9;          /* jmp */
-        *(unsigned long *)jmp = delta;
+        b.opcode = 0xe9;        /* jmp */
+        b.delta = delta;
+        text_poke(call, (unsigned char *)&b, 5);
 
         return 5;
 }
diff --git a/arch/i386/mm/init.c b/arch/i386/mm/init.c
index e1a9a805c445..c3b9905af2d5 100644
--- a/arch/i386/mm/init.c
+++ b/arch/i386/mm/init.c
@@ -800,17 +800,9 @@ void mark_rodata_ro(void)
         unsigned long start = PFN_ALIGN(_text);
         unsigned long size = PFN_ALIGN(_etext) - start;
 
-#ifndef CONFIG_KPROBES
-#ifdef CONFIG_HOTPLUG_CPU
-        /* It must still be possible to apply SMP alternatives. */
-        if (num_possible_cpus() <= 1)
-#endif
-        {
-                change_page_attr(virt_to_page(start),
-                                 size >> PAGE_SHIFT, PAGE_KERNEL_RX);
-                printk("Write protecting the kernel text: %luk\n", size >> 10);
-        }
-#endif
+        change_page_attr(virt_to_page(start),
+                         size >> PAGE_SHIFT, PAGE_KERNEL_RX);
+        printk("Write protecting the kernel text: %luk\n", size >> 10);
         start += size;
         size = (unsigned long)__end_rodata - start;
         change_page_attr(virt_to_page(start),