Skip to contentSkip to conversation

x86/dumpstack: don't dump kernel memory based on usermode RIP

PATCH Superseded Jann Horn 2 participants 2 messages Aug 28, 2018

Subsystems (3)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Jann Horn

author
#0 lore raw

PATCH 1/1 x86/dumpstack: don't dump kernel memory based on usermode RIP

show_opcodes() is used both for dumping kernel instructions and for dumping user instructions. If userspace causes #PF by jumping to a kernel address, show_opcodes() can be reached with regs->ip controlled by the user, pointing to kernel code. Make sure that userspace can't trick us into dumping kernel memory into dmesg.

Fixes 7cccf0725cf7 ("x86/dumpstack: Add a show_ip() function")
---
 arch/x86/include/asm/stacktrace.h |  2 +-
 arch/x86/kernel/dumpstack.c       | 13 ++++++++++---
 arch/x86/mm/fault.c               |  2 +-
 3 files changed, 12 insertions(+), 5 deletions(-)
arch/x86/include/asm/stacktrace.h +1 -1
index b6dc698f992a..f335aad404a4 100644
@@ -111,6 +111,6 @@ static inline unsigned long caller_frame_pointer(void)
 	return (unsigned long)frame;
 }
 
-void show_opcodes(u8 *rip, const char *loglvl);
+void show_opcodes(struct pt_regs *regs, const char *loglvl);
 void show_ip(struct pt_regs *regs, const char *loglvl);
 #endif /* _ASM_X86_STACKTRACE_H */
arch/x86/kernel/dumpstack.c +10 -3
index 9c8652974f8e..5d2700c98b29 100644
@@ -89,14 +89,21 @@ static void printk_stack_address(unsigned long address, int reliable,
  * Thus, the 2/3rds prologue and 64 byte OPCODE_BUFSIZE is just a random
  * guesstimate in attempt to achieve all of the above.
  */
-void show_opcodes(u8 *rip, const char *loglvl)
+void show_opcodes(struct pt_regs *regs, const char *loglvl)
 {
 #define PROLOGUE_SIZE 42
 #define EPILOGUE_SIZE 21
 #define OPCODE_BUFSIZE (PROLOGUE_SIZE + 1 + EPILOGUE_SIZE)
 	u8 opcodes[OPCODE_BUFSIZE];
+	u8 *prologue = (u8 *)regs->ip - PROLOGUE_SIZE;
+	/*
+	 * Make sure userspace isn't trying to trick us into dumping kernel
+	 * memory by pointing the userspace instruction pointer at it.
+	 */
+	bool bad_ip = user_mode(regs) &&
+		      __range_not_ok(prologue, OPCODE_BUFSIZE, TASK_SIZE_MAX);
 
-	if (probe_kernel_read(opcodes, rip - PROLOGUE_SIZE, OPCODE_BUFSIZE)) {
+	if (bad_ip || probe_kernel_read(opcodes, prologue, OPCODE_BUFSIZE)) {
 		printk("%sCode: Bad RIP value.\n", loglvl);
 	} else {
 		printk("%sCode: %" __stringify(PROLOGUE_SIZE) "ph <%02x> %"
@@ -112,7 +119,7 @@ void show_ip(struct pt_regs *regs, const char *loglvl)
 #else
 	printk("%sRIP: %04x:%pS\n", loglvl, (int)regs->cs, (void *)regs->ip);
 #endif
-	show_opcodes((u8 *)regs->ip, loglvl);
+	show_opcodes(regs, loglvl);
 }
 
 void show_iret_regs(struct pt_regs *regs)
arch/x86/mm/fault.c +1 -1
index b9123c497e0a..47bebfe6efa7 100644
@@ -837,7 +837,7 @@ show_signal_msg(struct pt_regs *regs, unsigned long error_code,
 
 	printk(KERN_CONT "\n");
 
-	show_opcodes((u8 *)regs->ip, loglvl);
+	show_opcodes(regs, loglvl);
 }
 
 static void
Signature · 1 line
2.19.0.rc0.228.g281dcd1b4d0-goog
Reviewed-by on v1 1/1
#1 lore raw
Quoted 9 lines · Jann Horn

show_opcodes() is used both for dumping kernel instructions and for dumping user instructions. If userspace causes #PF by jumping to a kernel address, show_opcodes() can be reached with regs->ip controlled by the user, pointing to kernel code. Make sure that userspace can't trick us into dumping kernel memory into dmesg.

Fixes 7cccf0725cf7 ("x86/dumpstack: Add a show_ip() function")

Thanks for finding this!

-Kees

Quoted 71 lines
---
 arch/x86/include/asm/stacktrace.h |  2 +-
 arch/x86/kernel/dumpstack.c       | 13 ++++++++++---
 arch/x86/mm/fault.c               |  2 +-
 3 files changed, 12 insertions(+), 5 deletions(-)
arch/x86/include/asm/stacktrace.h +1 -1
index b6dc698f992a..f335aad404a4 100644
@@ -111,6 +111,6 @@ static inline unsigned long caller_frame_pointer(void)
        return (unsigned long)frame;
 }
 
-void show_opcodes(u8 *rip, const char *loglvl);
+void show_opcodes(struct pt_regs *regs, const char *loglvl);
 void show_ip(struct pt_regs *regs, const char *loglvl);
 #endif /* _ASM_X86_STACKTRACE_H */
arch/x86/kernel/dumpstack.c +10 -3
index 9c8652974f8e..5d2700c98b29 100644
@@ -89,14 +89,21 @@ static void printk_stack_address(unsigned long address, int reliable,
  * Thus, the 2/3rds prologue and 64 byte OPCODE_BUFSIZE is just a random
  * guesstimate in attempt to achieve all of the above.
  */
-void show_opcodes(u8 *rip, const char *loglvl)
+void show_opcodes(struct pt_regs *regs, const char *loglvl)
 {
 #define PROLOGUE_SIZE 42
 #define EPILOGUE_SIZE 21
 #define OPCODE_BUFSIZE (PROLOGUE_SIZE + 1 + EPILOGUE_SIZE)
        u8 opcodes[OPCODE_BUFSIZE];
+       u8 *prologue = (u8 *)regs->ip - PROLOGUE_SIZE;
+       /*
+        * Make sure userspace isn't trying to trick us into dumping kernel
+        * memory by pointing the userspace instruction pointer at it.
+        */
+       bool bad_ip = user_mode(regs) &&
+                     __range_not_ok(prologue, OPCODE_BUFSIZE, TASK_SIZE_MAX);
 
-       if (probe_kernel_read(opcodes, rip - PROLOGUE_SIZE, OPCODE_BUFSIZE)) {
+       if (bad_ip || probe_kernel_read(opcodes, prologue, OPCODE_BUFSIZE)) {
                printk("%sCode: Bad RIP value.\n", loglvl);
        } else {
                printk("%sCode: %" __stringify(PROLOGUE_SIZE) "ph <%02x> %"
@@ -112,7 +119,7 @@ void show_ip(struct pt_regs *regs, const char *loglvl)
 #else
        printk("%sRIP: %04x:%pS\n", loglvl, (int)regs->cs, (void *)regs->ip);
 #endif
-       show_opcodes((u8 *)regs->ip, loglvl);
+       show_opcodes(regs, loglvl);
 }
 
 void show_iret_regs(struct pt_regs *regs)
arch/x86/mm/fault.c +1 -2
index b9123c497e0a..47bebfe6efa7 100644
@@ -837,7 +837,7 @@ show_signal_msg(struct pt_regs *regs, unsigned long error_code,
 
        printk(KERN_CONT "\n");
 
-       show_opcodes((u8 *)regs->ip, loglvl);
+       show_opcodes(regs, loglvl);
 }
 
 static void
--

2.19.0.rc0.228.g281dcd1b4d0-goog

Signature · 2 lines
Kees Cook
Pixel Security