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, 140 insertions, 0 deletions

diff --git a/drivers/firmware/sysfb_simplefb.c b/drivers/firmware/sysfb_simplefb.c
new file mode 100644
index 000000000..71f542e37
--- /dev/null
+++ b/drivers/firmware/sysfb_simplefb.c
@@ -0,0 +1,140 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Generic System Framebuffers
+ * Copyright (c) 2012-2013 David Herrmann <dh.herrmann@gmail.com>
+ */
+
+/*
+ * simple-framebuffer probing
+ * Try to convert "screen_info" into a "simple-framebuffer" compatible mode.
+ * If the mode is incompatible, we return "false" and let the caller create
+ * legacy nodes instead.
+ */
+
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/platform_data/simplefb.h>
+#include <linux/platform_device.h>
+#include <linux/screen_info.h>
+#include <linux/sysfb.h>
+
+static const char simplefb_resname[] = "BOOTFB";
+static const struct simplefb_format formats[] = SIMPLEFB_FORMATS;
+
+/* try parsing screen_info into a simple-framebuffer mode struct */
+__init bool sysfb_parse_mode(const struct screen_info *si,
+			     struct simplefb_platform_data *mode)
+{
+	__u8 type;
+	u32 bits_per_pixel;
+	unsigned int i;
+
+	type = si->orig_video_isVGA;
+	if (type != VIDEO_TYPE_VLFB && type != VIDEO_TYPE_EFI)
+		return false;
+
+	bits_per_pixel = __screen_info_lfb_bits_per_pixel(si);
+
+	for (i = 0; i < ARRAY_SIZE(formats); ++i) {
+		const struct simplefb_format *f = &formats[i];
+
+		if (f->transp.length)
+			continue; /* transparent formats are unsupported by VESA/EFI */
+
+		if (bits_per_pixel == f->bits_per_pixel &&
+		    si->red_size == f->red.length &&
+		    si->red_pos == f->red.offset &&
+		    si->green_size == f->green.length &&
+		    si->green_pos == f->green.offset &&
+		    si->blue_size == f->blue.length &&
+		    si->blue_pos == f->blue.offset) {
+			mode->format = f->name;
+			mode->width = si->lfb_width;
+			mode->height = si->lfb_height;
+			mode->stride = si->lfb_linelength;
+			return true;
+		}
+	}
+
+	return false;
+}
+
+__init struct platform_device *sysfb_create_simplefb(const struct screen_info *si,
+						     const struct simplefb_platform_data *mode,
+						     struct device *parent)
+{
+	struct platform_device *pd;
+	struct resource res;
+	u64 base, size;
+	u32 length;
+	int ret;
+
+	/*
+	 * If the 64BIT_BASE capability is set, ext_lfb_base will contain the
+	 * upper half of the base address. Assemble the address, then make sure
+	 * it is valid and we can actually access it.
+	 */
+	base = si->lfb_base;
+	if (si->capabilities & VIDEO_CAPABILITY_64BIT_BASE)
+		base |= (u64)si->ext_lfb_base << 32;
+	if (!base || (u64)(resource_size_t)base != base) {
+		printk(KERN_DEBUG "sysfb: inaccessible VRAM base\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	/*
+	 * Don't use lfb_size as IORESOURCE size, since it may contain the
+	 * entire VMEM, and thus require huge mappings. Use just the part we
+	 * need, that is, the part where the framebuffer is located. But verify
+	 * that it does not exceed the advertised VMEM.
+	 * Note that in case of VBE, the lfb_size is shifted by 16 bits for
+	 * historical reasons.
+	 */
+	size = si->lfb_size;
+	if (si->orig_video_isVGA == VIDEO_TYPE_VLFB)
+		size <<= 16;
+	length = mode->height * mode->stride;
+	if (length > size) {
+		printk(KERN_WARNING "sysfb: VRAM smaller than advertised\n");
+		return ERR_PTR(-EINVAL);
+	}
+	length = PAGE_ALIGN(length);
+
+	/* setup IORESOURCE_MEM as framebuffer memory */
+	memset(&res, 0, sizeof(res));
+	res.flags = IORESOURCE_MEM;
+	res.name = simplefb_resname;
+	res.start = base;
+	res.end = res.start + length - 1;
+	if (res.end <= res.start)
+		return ERR_PTR(-EINVAL);
+
+	pd = platform_device_alloc("simple-framebuffer", 0);
+	if (!pd)
+		return ERR_PTR(-ENOMEM);
+
+	pd->dev.parent = parent;
+
+	sysfb_set_efifb_fwnode(si, pd);
+
+	ret = platform_device_add_resources(pd, &res, 1);
+	if (ret)
+		goto err_put_device;
+
+	ret = platform_device_add_data(pd, mode, sizeof(*mode));
+	if (ret)
+		goto err_put_device;
+
+	ret = platform_device_add(pd);
+	if (ret)
+		goto err_put_device;
+
+	return pd;
+
+err_put_device:
+	platform_device_put(pd);
+
+	return ERR_PTR(ret);
+}