Before this patch, we were clearing pt_regs->r8..r11 on stack.
We can as well just store actual r8..r11 registers there:
they came from userspace, we leak no information by showing them to ptrace.
This allows to get rid of one insn ("xor %eax,%eax").
Not a big deal, but still...
After call to syscall_trace_enter(), before this patch we were restoring
clobbered registers and jump to code which converts 32-bit syscall
ABI to 64-bit C ABI. This is unnecessary work, we can combine both
steps into one (similar to what audit code does already).
Also, we do not need "<ABI>_do_call" labels anymore.
text data bss dec hex filename
1391 0 0 1391 56f entry_64_compat.o.before
1327 0 0 1327 52f entry_64_compat.o.after
Signed-off-by: Denys Vlasenko <dvlas...@redhat.com>
CC: Linus Torvalds <torva...@linux-foundation.org>
CC: Steven Rostedt <rost...@goodmis.org>
CC: Ingo Molnar <mi...@kernel.org>
CC: Borislav Petkov <b...@alien8.de>
CC: "H. Peter Anvin" <h...@zytor.com>
CC: Andy Lutomirski <l...@amacapital.net>
CC: Oleg Nesterov <o...@redhat.com>
CC: Frederic Weisbecker <fweis...@gmail.com>
CC: Alexei Starovoitov <a...@plumgrid.com>
CC: Will Drewry <w...@chromium.org>
CC: Kees Cook <keesc...@chromium.org>
CC: x...@kernel.org
CC: linux-kernel@vger.kernel.org
---
arch/x86/entry/entry_64_compat.S | 121 ++++++++++++++++++++++-----------------
1 file changed, 70 insertions(+), 51 deletions(-)
diff --git a/arch/x86/entry/entry_64_compat.S b/arch/x86/entry/entry_64_compat.S
index d83e7e3..9d40cae 100644
--- a/arch/x86/entry/entry_64_compat.S
+++ b/arch/x86/entry/entry_64_compat.S
@@ -110,7 +110,6 @@ sysenter_flags_fixed:
testl $_TIF_WORK_SYSCALL_ENTRY, ASM_THREAD_INFO(TI_flags, %rsp,
SIZEOF_PTREGS)
jnz sysenter_tracesys
-sysenter_do_call:
/* 32-bit syscall -> 64-bit C ABI argument conversion */
movl %edi, %r8d /* arg5 */
movl %ebp, %r9d /* arg6 */
@@ -248,24 +247,31 @@ sysenter_tracesys:
testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),
ASM_THREAD_INFO(TI_flags, %rsp, SIZEOF_PTREGS)
jz sysenter_auditsys
#endif
- SAVE_EXTRA_REGS
- xorl %eax, %eax /* Do not leak kernel information */
- movq %rax, R11(%rsp)
- movq %rax, R10(%rsp)
- movq %rax, R9(%rsp)
- movq %rax, R8(%rsp)
+ /*
+ * Ptrace needs complete pt_regs. Fill members we didn't save earlier.
+ * r8..r15 are not really meaningful, but leaving uninitialized data
+ * there may leak kernel data to ptrace.
+ */
+ movq %r11, R11(%rsp)
+ movq %r10, R10(%rsp)
+ movq %r9, R9(%rsp)
+ movq %r8, R8(%rsp)
+ SAVE_EXTRA_REGS /* pt_regs->bp, bx, r12-15 */
movq %rsp, %rdi /* &pt_regs -> arg1 */
call syscall_trace_enter
-
- /* Reload arg registers from stack. (see sysenter_tracesys) */
- movl RCX(%rsp), %ecx
- movl RDX(%rsp), %edx
- movl RSI(%rsp), %esi
- movl RDI(%rsp), %edi
- movl %eax, %eax /* zero extension */
-
- RESTORE_EXTRA_REGS
- jmp sysenter_do_call
+ /*
+ * We are going to jump back to syscall dispatch.
+ * Prepare syscall args as required by 64-bit C ABI.
+ * We must read data from pt_regs: ptrace could have changed it.
+ */
+ movl %eax, %eax /* syscall number: zero extend it */
+ movl RBX(%rsp), %edi /* arg1 */
+ movl RCX(%rsp), %esi /* arg2 */
+ movl RDX(%rsp), %edx /* arg3 */
+ movl RSI(%rsp), %ecx /* arg4 */
+ movl RDI(%rsp), %r8d /* arg5 */
+ movl RBP(%rsp), %r9d /* arg6 */
+ jmp sysenter_dispatch
ENDPROC(entry_SYSENTER_compat)
/*
@@ -339,7 +348,6 @@ ENTRY(entry_SYSCALL_compat)
testl $_TIF_WORK_SYSCALL_ENTRY, ASM_THREAD_INFO(TI_flags, %rsp,
SIZEOF_PTREGS)
jnz cstar_tracesys
-cstar_do_call:
/* 32-bit syscall -> 64-bit C ABI argument conversion */
movl %edi, %r8d /* arg5 */
movl %ebp, %r9d /* arg6 */
@@ -402,24 +410,31 @@ cstar_tracesys:
testl $(_TIF_WORK_SYSCALL_ENTRY & ~_TIF_SYSCALL_AUDIT),
ASM_THREAD_INFO(TI_flags, %rsp, SIZEOF_PTREGS)
jz cstar_auditsys
#endif
- SAVE_EXTRA_REGS
- xorl %eax, %eax /* Do not leak kernel information */
- movq %rax, R11(%rsp)
- movq %rax, R10(%rsp)
- movq %rax, R9(%rsp)
- movq %rax, R8(%rsp)
+ /*
+ * Ptrace needs complete pt_regs. Fill members we didn't save earlier.
+ * r8..r15 are not really meaningful, but leaving uninitialized data
+ * there may leak kernel data to ptrace.
+ */
+ movq %r11, R11(%rsp)
+ movq %r10, R10(%rsp)
+ movq %r9, R9(%rsp)
+ movq %r8, R8(%rsp)
+ SAVE_EXTRA_REGS /* pt_regs->bp, bx, r12-15 */
movq %rsp, %rdi /* &pt_regs -> arg1 */
call syscall_trace_enter
-
- /* Reload arg registers from stack. (see sysenter_tracesys) */
- movl RCX(%rsp), %ecx
- movl RDX(%rsp), %edx
- movl RSI(%rsp), %esi
- movl RDI(%rsp), %edi
- movl %eax, %eax /* zero extension */
-
- RESTORE_EXTRA_REGS
- jmp cstar_do_call
+ /*
+ * We are going to jump back to syscall dispatch.
+ * Prepare syscall args as required by 64-bit C ABI.
+ * We must read data from pt_regs: ptrace could have changed it.
+ */
+ movl %eax, %eax /* syscall number: zero extend it */
+ movl RBX(%rsp), %edi /* arg1 */
+ movl RCX(%rsp), %esi /* arg2 */
+ movl RDX(%rsp), %edx /* arg3 */
+ movl RSI(%rsp), %ecx /* arg4 */
+ movl RDI(%rsp), %r8d /* arg5 */
+ movl RBP(%rsp), %r9d /* arg6 */
+ jmp cstar_dispatch
END(entry_SYSCALL_compat)
ia32_badarg:
@@ -483,15 +498,15 @@ ENTRY(entry_INT80_compat)
orl $TS_COMPAT, ASM_THREAD_INFO(TI_status, %rsp, SIZEOF_PTREGS)
testl $_TIF_WORK_SYSCALL_ENTRY, ASM_THREAD_INFO(TI_flags, %rsp,
SIZEOF_PTREGS)
- jnz ia32_tracesys
+ jnz int80_tracesys
-ia32_do_call:
/* 32-bit syscall -> 64-bit C ABI argument conversion */
movl %edi, %r8d /* arg5 */
movl %ebp, %r9d /* arg6 */
xchg %ecx, %esi /* rsi:arg2, rcx:arg4 */
movl %ebx, %edi /* arg1 */
movl %edx, %edx /* arg3 (zero extension) */
+int80_dispatch:
cmpq $(IA32_NR_syscalls-1), %rax
ja 1f
@@ -500,24 +515,28 @@ ia32_do_call:
1:
jmp int_ret_from_sys_call
-ia32_tracesys:
- SAVE_EXTRA_REGS
- movq %rsp, %rdi /* &pt_regs -> arg1 */
+int80_tracesys:
+ /*
+ * Ptrace needs complete pt_regs. Fill members we didn't save earlier.
+ * r8..r15 are not really meaningful, but leaving uninitialized data
+ * there may leak kernel data to ptrace.
+ */
+ SAVE_EXTRA_REGS /* pt_regs->bp, bx, r12-15 */
+ movq %rsp, %rdi /* &pt_regs -> arg1 */
call syscall_trace_enter
/*
- * Reload arg registers from stack in case ptrace changed them.
- * Don't reload %eax because syscall_trace_enter() returned
- * the %rax value we should see. But do truncate it to 32 bits.
- * If it's -1 to make us punt the syscall, then (u32)-1 is still
- * an appropriately invalid value.
+ * We are going to jump back to syscall dispatch.
+ * Prepare syscall args as required by 64-bit C ABI.
+ * We must read data from pt_regs: ptrace could have changed it.
*/
- movl RCX(%rsp), %ecx
- movl RDX(%rsp), %edx
- movl RSI(%rsp), %esi
- movl RDI(%rsp), %edi
- movl %eax, %eax /* zero extension */
- RESTORE_EXTRA_REGS
- jmp ia32_do_call
+ movl %eax, %eax /* syscall number: zero extend it */
+ movl RBX(%rsp), %edi /* arg1 */
+ movl RCX(%rsp), %esi /* arg2 */
+ movl RDX(%rsp), %edx /* arg3 */
+ movl RSI(%rsp), %ecx /* arg4 */
+ movl RDI(%rsp), %r8d /* arg5 */
+ movl RBP(%rsp), %r9d /* arg6 */
+ jmp int80_dispatch
END(entry_INT80_compat)
.macro PTREGSCALL label, func
--
1.8.1.4
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
