x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 132 insertions, 0 deletions

diff --git a/Documentation/crypto/sha3.rst b/Documentation/crypto/sha3.rst
new file mode 100644
index 000000000..250669c98
--- /dev/null
+++ b/Documentation/crypto/sha3.rst
@@ -0,0 +1,132 @@
+.. SPDX-License-Identifier: GPL-2.0-or-later
+
+.. _sha3:
+
+==========================
+SHA-3 Algorithm Collection
+==========================
+
+.. contents::
+
+Overview
+========
+
+The SHA-3 family of algorithms, as specified in NIST FIPS-202 [1]_, contains six
+algorithms based on the Keccak sponge function.  The differences between them
+are: the "rate" (how much of the state buffer gets updated with new data between
+invocations of the Keccak function and analogous to the "block size"), what
+domain separation suffix gets appended to the input data, and how much output
+data is extracted at the end.  The Keccak sponge function is designed such that
+arbitrary amounts of output can be obtained for certain algorithms.
+
+Four digest algorithms are provided:
+
+ - SHA3-224
+ - SHA3-256
+ - SHA3-384
+ - SHA3-512
+
+Additionally, two Extendable-Output Functions (XOFs) are provided:
+
+ - SHAKE128
+ - SHAKE256
+
+The SHA-3 library API supports all six of these algorithms.  The four digest
+algorithms are also supported by the crypto_shash and crypto_ahash APIs.
+
+This document describes the SHA-3 library API.
+
+
+Digests
+=======
+
+The following functions compute SHA-3 digests::
+
+	void sha3_224(const u8 *in, size_t in_len, u8 out[SHA3_224_DIGEST_SIZE]);
+	void sha3_256(const u8 *in, size_t in_len, u8 out[SHA3_256_DIGEST_SIZE]);
+	void sha3_384(const u8 *in, size_t in_len, u8 out[SHA3_384_DIGEST_SIZE]);
+	void sha3_512(const u8 *in, size_t in_len, u8 out[SHA3_512_DIGEST_SIZE]);
+
+For users that need to pass in data incrementally, an incremental API is also
+provided.  The incremental API uses the following struct::
+
+	struct sha3_ctx { ... };
+
+Initialization is done with one of::
+
+	void sha3_224_init(struct sha3_ctx *ctx);
+	void sha3_256_init(struct sha3_ctx *ctx);
+	void sha3_384_init(struct sha3_ctx *ctx);
+	void sha3_512_init(struct sha3_ctx *ctx);
+
+Input data is then added with any number of calls to::
+
+	void sha3_update(struct sha3_ctx *ctx, const u8 *in, size_t in_len);
+
+Finally, the digest is generated using::
+
+	void sha3_final(struct sha3_ctx *ctx, u8 *out);
+
+which also zeroizes the context.  The length of the digest is determined by the
+initialization function that was called.
+
+
+Extendable-Output Functions
+===========================
+
+The following functions compute the SHA-3 extendable-output functions (XOFs)::
+
+	void shake128(const u8 *in, size_t in_len, u8 *out, size_t out_len);
+	void shake256(const u8 *in, size_t in_len, u8 *out, size_t out_len);
+
+For users that need to provide the input data incrementally and/or receive the
+output data incrementally, an incremental API is also provided.  The incremental
+API uses the following struct::
+
+	struct shake_ctx { ... };
+
+Initialization is done with one of::
+
+	void shake128_init(struct shake_ctx *ctx);
+	void shake256_init(struct shake_ctx *ctx);
+
+Input data is then added with any number of calls to::
+
+	void shake_update(struct shake_ctx *ctx, const u8 *in, size_t in_len);
+
+Finally, the output data is extracted with any number of calls to::
+
+	void shake_squeeze(struct shake_ctx *ctx, u8 *out, size_t out_len);
+
+and telling it how much data should be extracted.  Note that performing multiple
+squeezes, with the output laid consecutively in a buffer, gets exactly the same
+output as doing a single squeeze for the combined amount over the same buffer.
+
+More input data cannot be added after squeezing has started.
+
+Once all the desired output has been extracted, zeroize the context::
+
+	void shake_zeroize_ctx(struct shake_ctx *ctx);
+
+
+Testing
+=======
+
+To test the SHA-3 code, use sha3_kunit (CONFIG_CRYPTO_LIB_SHA3_KUNIT_TEST).
+
+Since the SHA-3 algorithms are FIPS-approved, when the kernel is booted in FIPS
+mode the SHA-3 library also performs a simple self-test.  This is purely to meet
+a FIPS requirement.  Normal testing done by kernel developers and integrators
+should use the much more comprehensive KUnit test suite instead.
+
+
+References
+==========
+
+.. [1] https://nvlpubs.nist.gov/nistpubs/FIPS/NIST.FIPS.202.pdf
+
+
+API Function Reference
+======================
+
+.. kernel-doc:: include/crypto/sha3.h