bpf: move clearing of A/X into classic to eBPF migration prologue

Back in the days where eBPF (or back then "internal BPF" ;->) was not
exposed to user space, and only the classic BPF programs internally
translated into eBPF programs, we missed the fact that for classic BPF
A and X needed to be cleared. It was fixed back then via 83d5b7ef99c9
("net: filter: initialize A and X registers"), and thus classic BPF
specifics were added to the eBPF interpreter core to work around it.

This added some confusion for JIT developers later on that take the
eBPF interpreter code as an example for deriving their JIT. F.e. in
f75298f5c3fe ("s390/bpf: clear correct BPF accumulator register"), at
least X could leak stack memory. Furthermore, since this is only needed
for classic BPF translations and not for eBPF (verifier takes care
that read access to regs cannot be done uninitialized), more complexity
is added to JITs as they need to determine whether they deal with
migrations or native eBPF where they can just omit clearing A/X in
their prologue and thus reduce image size a bit, see f.e. cde66c2d88da
("s390/bpf: Only clear A and X for converted BPF programs"). In other
cases (x86, arm64), A and X is being cleared in the prologue also for
eBPF case, which is unnecessary.

Lets move this into the BPF migration in bpf_convert_filter() where it
actually belongs as long as the number of eBPF JITs are still few. It
can thus be done generically; allowing us to remove the quirk from
__bpf_prog_run() and to slightly reduce JIT image size in case of eBPF,
while reducing code duplication on this matter in current(/future) eBPF
JITs.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Reviewed-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Tested-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Cc: Zi Shen Lim <zlim.lnx@gmail.com>
Cc: Yang Shi <yang.shi@linaro.org>
Acked-by: Yang Shi <yang.shi@linaro.org>
Acked-by: Zi Shen Lim <zlim.lnx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 9 insertions, 5 deletions

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index 75991979f667..c080e812ce85 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -193,7 +193,7 @@ struct jit_context {
 	 32 /* space for rbx, r13, r14, r15 */ + \
 	 8 /* space for skb_copy_bits() buffer */)
 
-#define PROLOGUE_SIZE 51
+#define PROLOGUE_SIZE 48
 
 /* emit x64 prologue code for BPF program and check it's size.
  * bpf_tail_call helper will skip it while jumping into another program
@@ -229,11 +229,15 @@ static void emit_prologue(u8 **pprog)
 	/* mov qword ptr [rbp-X],r15 */
 	EMIT3_off32(0x4C, 0x89, 0xBD, -STACKSIZE + 24);
 
-	/* clear A and X registers */
-	EMIT2(0x31, 0xc0); /* xor eax, eax */
-	EMIT3(0x4D, 0x31, 0xED); /* xor r13, r13 */
+	/* Clear the tail call counter (tail_call_cnt): for eBPF tail calls
+	 * we need to reset the counter to 0. It's done in two instructions,
+	 * resetting rax register to 0 (xor on eax gets 0 extended), and
+	 * moving it to the counter location.
+	 */
 
-	/* clear tail_cnt: mov qword ptr [rbp-X], rax */
+	/* xor eax, eax */
+	EMIT2(0x31, 0xc0);
+	/* mov qword ptr [rbp-X], rax */
 	EMIT3_off32(0x48, 0x89, 0x85, -STACKSIZE + 32);
 
 	BUILD_BUG_ON(cnt != PROLOGUE_SIZE);