[RFC PATCH] powerpc/64/module: REL32 relocation range check

Subsystems: linux for powerpc (32-bit and 64-bit), the rest

STALE2869d

2 messages, 2 authors, 2018-10-22 · open the first message on its own page

[RFC PATCH] powerpc/64/module: REL32 relocation range check

From: Nicholas Piggin <npiggin@gmail.com>
Date: 2018-08-29 11:57:10

The recent module relocation overflow crash demonstrated that we
have no range checking on REL32 relative relocations. This patch
implements a basic check, the same kernel that previously oopsed
and rebooted now continues with some of these errors when loading
the module:

    module_64: x_tables: REL32 527703503449812 out of range!

Question is whether other relocations (ADDR32, REL16, TOC16, etc.)
should also have overflow checks.
---
 arch/powerpc/kernel/module_64.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c
index a2636c250b7b..2a2fb656d23b 100644
--- a/arch/powerpc/kernel/module_64.c
+++ b/arch/powerpc/kernel/module_64.c
@@ -678,7 +678,14 @@ int apply_relocate_add(Elf64_Shdr *sechdrs,
 
 		case R_PPC64_REL32:
 			/* 32 bits relative (used by relative exception tables) */
-			*(u32 *)location = value - (unsigned long)location;
+			/* Convert value to relative */
+			value -= (unsigned long)location;
+			if (value + 0x80000000 > 0xffffffff) {
+				pr_err("%s: REL32 %li out of range!\n",
+				       me->name, (long int)value);
+				return -ENOEXEC;
+			}
+			*(u32 *)location = value;
 			break;
 
 		case R_PPC64_TOCSAVE:
-- 
2.18.0

Re: [RFC] powerpc/64/module: REL32 relocation range check

From: Michael Ellerman <hidden>
Date: 2018-10-22 09:54:58

On Wed, 2018-08-29 at 11:56:56 UTC, Nicholas Piggin wrote:
quoted hunk
The recent module relocation overflow crash demonstrated that we
have no range checking on REL32 relative relocations. This patch
implements a basic check, the same kernel that previously oopsed
and rebooted now continues with some of these errors when loading
the module:

    module_64: x_tables: REL32 527703503449812 out of range!

Question is whether other relocations (ADDR32, REL16, TOC16, etc.)
should also have overflow checks.
---
 arch/powerpc/kernel/module_64.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/arch/powerpc/kernel/module_64.c b/arch/powerpc/kernel/module_64.c
index a2636c250b7b..2a2fb656d23b 100644
--- a/arch/powerpc/kernel/module_64.c
+++ b/arch/powerpc/kernel/module_64.c
@@ -678,7 +678,14 @@ int apply_relocate_add(Elf64_Shdr *sechdrs,
 
 		case R_PPC64_REL32:
 			/* 32 bits relative (used by relative exception tables) */
-			*(u32 *)location = value - (unsigned long)location;
+			/* Convert value to relative */
+			value -= (unsigned long)location;
+			if (value + 0x80000000 > 0xffffffff) {
+				pr_err("%s: REL32 %li out of range!\n",
+				       me->name, (long int)value);
+				return -ENOEXEC;
+			}
+			*(u32 *)location = value;
 			break;
 
 		case R_PPC64_TOCSAVE:
Applied to powerpc next, thanks.

https://git.kernel.org/powerpc/c/b851ba02a6f3075f0f99c60c4bc30a

cheers
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help