From: Rick Edgecombe <rick.p.edgecombe@intel.com>
To: daniel@iogearbox.net, ast@fb.com
Cc: netdev@vger.kernel.org, ard.biesheuvel@linaro.org,
dave.hansen@intel.com, kristen@linux.intel.com,
Rick Edgecombe <rick.p.edgecombe@intel.com>
Subject: [RFC PATCH 1/4] bpf, x64: Implement BPF call retpoline
Date: Tue, 5 Feb 2019 14:51:00 -0800 [thread overview]
Message-ID: <20190205225103.28296-2-rick.p.edgecombe@intel.com> (raw)
In-Reply-To: <20190205225103.28296-1-rick.p.edgecombe@intel.com>
Add x86 call retpoline sequence from the "Intel Retpoline: A Branch Target
Injection Mitigation White Paper" for BPF JIT compiler. Unlike the paper
it uses RBX instead of RAX since RAX is part of the BPF calling
convetions.
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Alexei Starovoitov <ast@fb.com>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@intel.com>
---
arch/x86/include/asm/nospec-branch.h | 45 ++++++++++++++++++++++++++--
1 file changed, 43 insertions(+), 2 deletions(-)
diff --git a/arch/x86/include/asm/nospec-branch.h b/arch/x86/include/asm/nospec-branch.h
index dad12b767ba0..70b3f6534134 100644
--- a/arch/x86/include/asm/nospec-branch.h
+++ b/arch/x86/include/asm/nospec-branch.h
@@ -320,6 +320,8 @@ DECLARE_STATIC_KEY_FALSE(switch_mm_always_ibpb);
#endif /* __ASSEMBLY__ */
+#ifdef CONFIG_RETPOLINE
+# ifdef CONFIG_X86_64
/*
* Below is used in the eBPF JIT compiler and emits the byte sequence
* for the following assembly:
@@ -341,8 +343,6 @@ DECLARE_STATIC_KEY_FALSE(switch_mm_always_ibpb);
* jmp *%rax for x86_64
* jmp *%edx for x86_32
*/
-#ifdef CONFIG_RETPOLINE
-# ifdef CONFIG_X86_64
# define RETPOLINE_RAX_BPF_JIT_SIZE 17
# define RETPOLINE_RAX_BPF_JIT() \
do { \
@@ -355,6 +355,44 @@ do { \
EMIT4(0x48, 0x89, 0x04, 0x24); /* mov %rax,(%rsp) */ \
EMIT1(0xC3); /* retq */ \
} while (0)
+
+/* Modified from "Intel Retpoline: A Branch Target Injection Mitigation White
+ * Paper" to use RBX register in order to not intefere with BPF calling
+ * conventions:
+ *
+ * jmp label2
+ *
+ * label0:
+ * call label1
+ *
+ * capture_ret_spec:
+ * pause
+ * lfence
+ * jmp capture_ret_spec
+ *
+ * label1:
+ * mov %rbx,(%rsp)
+ * ret
+ *
+ * label2:
+ * call label0
+ */
+# define RETPOLINE_RBX_BPF_JIT_CALL_SIZE 24
+# define RETPOLINE_RBX_BPF_JIT_CALL() \
+do { \
+ EMIT2(0xEB, 0x11); /* jump label2 */ \
+ /* label2: */ \
+ EMIT1_off32(0xE8, 7); /* call label1 */ \
+ /* capture_ret_spec: */ \
+ EMIT2(0xF3, 0x90); /* pause */ \
+ EMIT3(0x0F, 0xAE, 0xE8); /* lfence */ \
+ EMIT2(0xEB, 0xF9); /* jmp capture_ret_spec */ \
+ /* label1: */ \
+ EMIT4(0x48, 0x89, 0x1c, 0x24); /* mov %rbx,(%rsp) */ \
+ EMIT1(0xC3); /* ret */ \
+ /* label2: */ \
+ EMIT1_off32(0xE8, -22); /* call label0 */ \
+} while (0)
# else /* !CONFIG_X86_64 */
# define RETPOLINE_EDX_BPF_JIT() \
do { \
@@ -373,6 +411,9 @@ do { \
# define RETPOLINE_RAX_BPF_JIT_SIZE 2
# define RETPOLINE_RAX_BPF_JIT() \
EMIT2(0xFF, 0xE0); /* jmp *%rax */
+# define RETPOLINE_RBX_BPF_JIT_CALL_SIZE 2
+# define RETPOLINE_RBX_BPF_JIT_CALL() \
+ EMIT2(0xff, 0xD3) /* call *%rbx */
# else /* !CONFIG_X86_64 */
# define RETPOLINE_EDX_BPF_JIT() \
EMIT2(0xFF, 0xE2) /* jmp *%edx */
--
2.17.1
next prev parent reply other threads:[~2019-02-05 22:54 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-02-05 22:50 [RFC PATCH 0/4] Initial support for allocating BPF JITs in vmalloc for x86 Rick Edgecombe
2019-02-05 22:51 ` Rick Edgecombe [this message]
2019-02-05 22:51 ` [RFC PATCH 2/4] bpf, x64: Increase distance for bpf calls Rick Edgecombe
2019-02-05 22:51 ` [RFC PATCH 3/4] bpf: Charge bpf jit limit in bpf_jit_alloc_exec Rick Edgecombe
2019-02-05 22:51 ` [RFC PATCH 4/4] bpf, x64: Enable unprivlidged jit in vmalloc Rick Edgecombe
2019-02-06 0:35 ` [RFC PATCH 0/4] Initial support for allocating BPF JITs in vmalloc for x86 Alexei Starovoitov
2019-02-06 1:11 ` Edgecombe, Rick P
2019-02-06 1:40 ` Alexei Starovoitov
2019-02-06 15:38 ` Daniel Borkmann
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190205225103.28296-2-rick.p.edgecombe@intel.com \
--to=rick.p.edgecombe@intel.com \
--cc=ard.biesheuvel@linaro.org \
--cc=ast@fb.com \
--cc=daniel@iogearbox.net \
--cc=dave.hansen@intel.com \
--cc=kristen@linux.intel.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox