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, 120 insertions, 0 deletions

diff --git a/drivers/pci/controller/pcie-rcar.c b/drivers/pci/controller/pcie-rcar.c
new file mode 100644
index 000000000..7583699ef
--- /dev/null
+++ b/drivers/pci/controller/pcie-rcar.c
@@ -0,0 +1,120 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * PCIe driver for Renesas R-Car SoCs
+ *  Copyright (C) 2014-2020 Renesas Electronics Europe Ltd
+ *
+ * Author: Phil Edworthy <phil.edworthy@renesas.com>
+ */
+
+#include <linux/delay.h>
+#include <linux/pci.h>
+
+#include "pcie-rcar.h"
+
+void rcar_pci_write_reg(struct rcar_pcie *pcie, u32 val, unsigned int reg)
+{
+	writel(val, pcie->base + reg);
+}
+
+u32 rcar_pci_read_reg(struct rcar_pcie *pcie, unsigned int reg)
+{
+	return readl(pcie->base + reg);
+}
+
+void rcar_rmw32(struct rcar_pcie *pcie, int where, u32 mask, u32 data)
+{
+	unsigned int shift = BITS_PER_BYTE * (where & 3);
+	u32 val = rcar_pci_read_reg(pcie, where & ~3);
+
+	val &= ~(mask << shift);
+	val |= data << shift;
+	rcar_pci_write_reg(pcie, val, where & ~3);
+}
+
+int rcar_pcie_wait_for_phyrdy(struct rcar_pcie *pcie)
+{
+	unsigned int timeout = 10;
+
+	while (timeout--) {
+		if (rcar_pci_read_reg(pcie, PCIEPHYSR) & PHYRDY)
+			return 0;
+
+		msleep(5);
+	}
+
+	return -ETIMEDOUT;
+}
+
+int rcar_pcie_wait_for_dl(struct rcar_pcie *pcie)
+{
+	unsigned int timeout = 10000;
+
+	while (timeout--) {
+		if ((rcar_pci_read_reg(pcie, PCIETSTR) & DATA_LINK_ACTIVE))
+			return 0;
+
+		udelay(5);
+		cpu_relax();
+	}
+
+	return -ETIMEDOUT;
+}
+
+void rcar_pcie_set_outbound(struct rcar_pcie *pcie, int win,
+			    struct resource_entry *window)
+{
+	/* Setup PCIe address space mappings for each resource */
+	struct resource *res = window->res;
+	resource_size_t res_start;
+	resource_size_t size;
+	u32 mask;
+
+	rcar_pci_write_reg(pcie, 0x00000000, PCIEPTCTLR(win));
+
+	/*
+	 * The PAMR mask is calculated in units of 128Bytes, which
+	 * keeps things pretty simple.
+	 */
+	size = resource_size(res);
+	if (size > 128)
+		mask = (roundup_pow_of_two(size) / SZ_128) - 1;
+	else
+		mask = 0x0;
+	rcar_pci_write_reg(pcie, mask << 7, PCIEPAMR(win));
+
+	if (res->flags & IORESOURCE_IO)
+		res_start = pci_pio_to_address(res->start) - window->offset;
+	else
+		res_start = res->start - window->offset;
+
+	rcar_pci_write_reg(pcie, upper_32_bits(res_start), PCIEPAUR(win));
+	rcar_pci_write_reg(pcie, lower_32_bits(res_start) & ~0x7F,
+			   PCIEPALR(win));
+
+	/* First resource is for IO */
+	mask = PAR_ENABLE;
+	if (res->flags & IORESOURCE_IO)
+		mask |= IO_SPACE;
+
+	rcar_pci_write_reg(pcie, mask, PCIEPTCTLR(win));
+}
+
+void rcar_pcie_set_inbound(struct rcar_pcie *pcie, u64 cpu_addr,
+			   u64 pci_addr, u64 flags, int idx, bool host)
+{
+	/*
+	 * Set up 64-bit inbound regions as the range parser doesn't
+	 * distinguish between 32 and 64-bit types.
+	 */
+	if (host)
+		rcar_pci_write_reg(pcie, lower_32_bits(pci_addr),
+				   PCIEPRAR(idx));
+	rcar_pci_write_reg(pcie, lower_32_bits(cpu_addr), PCIELAR(idx));
+	rcar_pci_write_reg(pcie, flags, PCIELAMR(idx));
+
+	if (host)
+		rcar_pci_write_reg(pcie, upper_32_bits(pci_addr),
+				   PCIEPRAR(idx + 1));
+	rcar_pci_write_reg(pcie, upper_32_bits(cpu_addr), PCIELAR(idx + 1));
+	rcar_pci_write_reg(pcie, 0, PCIELAMR(idx + 1));
+}