Skip to contentSkip to conversation

arm64: kprobes: Update blacklist checking on arm64

PATCH In review Masami Hiramatsu 3 participants 8 messages Feb 12, 2019 – Feb 15, 2019 arm64

lore
Share & tools
lore
#0 lore raw

v3 cover arm64: kprobes: Update blacklist checking on arm64

Hello,

Here is the v3 series of update of the kprobe blacklist checking on arm64.

I found that some blacklist checking code were mis-placed in arch_prepare_kprobe() and arch_within_kprobe_blacklist(). Since the blacklist just filters by symbol, smaller than the symbol, like extable must be checked in arch_prepare_kprobe(). Also, all function (symbol) level check must be done by blacklist.

For arm64, it checks the extable entry address in blacklist and exception/irqentry function in arch_prepare_kprobe(). And, RODATA check is unneeded since kernel/kprobes.c already ensures the probe address is in kernel-text area.

In v3, I rebased on the latest arm64 kernel which includes James' KVM/HYP fixes for kprobes, and fix a reported bugs in [4/4].

Changes in v3:
- [4/4] Fixes to remove redundant blacklist of kprobe_text and add blacklist on exception_text.

Thank you,

---

Masami Hiramatsu (4):
arm64: kprobes: Move extable address check into arch_prepare_kprobe()
arm64: kprobes: Remove unneeded RODATA check
arm64: kprobes: Move exception_text check in blacklist
arm64: kprobes: Use arch_populate_kprobe_blacklist()

 arch/arm64/kernel/probes/kprobes.c |   52 +++++++++++++++++++-----------------
 1 file changed, 27 insertions(+), 25 deletions(-)

--
Masami Hiramatsu (Linaro) <[email protected]>

#1 lore raw

PATCH v3 1/4 arm64: kprobes: Move extable address check into arch_prepare_kprobe()

Move extable address check into arch_prepare_kprobe() from arch_within_kprobe_blacklist().
The blacklist is exposed via debugfs as a list of symbols. The extable entries are smaller, so must be filtered out by arch_prepare_kprobe().

---
Update in v2:
- Update commit message.
- Add Reviewed-by from James.

---
 arch/arm64/kernel/probes/kprobes.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
arch/arm64/kernel/probes/kprobes.c +5 -2
index f17afb99890c..9989ec9baa11 100644
@@ -102,6 +102,10 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 
 	if (in_exception_text(probe_addr))
 		return -EINVAL;
+
+	if (search_exception_tables(probe_addr))
+		return -EINVAL;
+
 	if (probe_addr >= (unsigned long) __start_rodata &&
 	    probe_addr <= (unsigned long) __end_rodata)
 		return -EINVAL;
@@ -479,8 +483,7 @@ bool arch_within_kprobe_blacklist(unsigned long addr)
 	    (addr >= (unsigned long)__idmap_text_start &&
 	    addr < (unsigned long)__idmap_text_end) ||
 	    (addr >= (unsigned long)__hyp_text_start &&
-	    addr < (unsigned long)__hyp_text_end) ||
-	    !!search_exception_tables(addr))
+	    addr < (unsigned long)__hyp_text_end))
 		return true;
 
 	if (!is_kernel_in_hyp_mode()) {
#2 lore raw

PATCH v3 2/4 arm64: kprobes: Remove unneeded RODATA check

In reply to Masami Hiramatsu

Remove unneeded RODATA check from arch_prepare_kprobe().

Since check_kprobe_address_safe() already ensured that the probe address is in kernel text, we don't need to check whether the address in RODATA or not. That must be always false.

---
 arch/arm64/kernel/probes/kprobes.c |    6 ------
 1 file changed, 6 deletions(-)
arch/arm64/kernel/probes/kprobes.c +0 -6
index 9989ec9baa11..bd06b4b13fa9 100644
@@ -91,8 +91,6 @@ static void __kprobes arch_simulate_insn(struct kprobe *p, struct pt_regs *regs)
 int __kprobes arch_prepare_kprobe(struct kprobe *p)
 {
 	unsigned long probe_addr = (unsigned long)p->addr;
-	extern char __start_rodata[];
-	extern char __end_rodata[];
 
 	if (probe_addr & 0x3)
 		return -EINVAL;
@@ -106,10 +104,6 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 	if (search_exception_tables(probe_addr))
 		return -EINVAL;
 
-	if (probe_addr >= (unsigned long) __start_rodata &&
-	    probe_addr <= (unsigned long) __end_rodata)
-		return -EINVAL;
-
 	/* decode instruction */
 	switch (arm_kprobe_decode_insn(p->addr, &p->ainsn)) {
 	case INSN_REJECTED:	/* insn not supported */
#3 lore raw

PATCH v3 3/4 arm64: kprobes: Move exception_text check in blacklist

In reply to Masami Hiramatsu

Move exception/irqentry text address check in blacklist, since those are symbol based rejection.

If we prohibit probing on the symbols in exception_text, those should be blacklisted.

---
 arch/arm64/kernel/probes/kprobes.c |    6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
arch/arm64/kernel/probes/kprobes.c +2 -4
index bd06b4b13fa9..194262fca5cd 100644
@@ -98,9 +98,6 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 	/* copy instruction */
 	p->opcode = le32_to_cpu(*p->addr);
 
-	if (in_exception_text(probe_addr))
-		return -EINVAL;
-
 	if (search_exception_tables(probe_addr))
 		return -EINVAL;
 
@@ -477,7 +474,8 @@ bool arch_within_kprobe_blacklist(unsigned long addr)
 	    (addr >= (unsigned long)__idmap_text_start &&
 	    addr < (unsigned long)__idmap_text_end) ||
 	    (addr >= (unsigned long)__hyp_text_start &&
-	    addr < (unsigned long)__hyp_text_end))
+	    addr < (unsigned long)__hyp_text_end) ||
+	    in_exception_text(addr))
 		return true;
 
 	if (!is_kernel_in_hyp_mode()) {
#4 lore raw

PATCH v3 4/4 arm64: kprobes: Use arch_populate_kprobe_blacklist()

In reply to Masami Hiramatsu

Use arch_populate_kprobe_blacklist() instead of arch_within_kprobe_blacklist() so that we can see the full blacklisted symbols under the debugfs.

---
Changes in v3
- Do not populate blacklist in __kprobe_text in arch_populate_kprobe_blacklist(), since it is already populated in populate_kprobe_blacklist().
- Add exception entry text blacklist since those are rejected by in_exception_text().

---
 arch/arm64/kernel/probes/kprobes.c |   45 +++++++++++++++++++++---------------
 1 file changed, 26 insertions(+), 19 deletions(-)
arch/arm64/kernel/probes/kprobes.c +26 -19
index 194262fca5cd..37d913f33a89 100644
@@ -465,26 +465,33 @@ kprobe_breakpoint_handler(struct pt_regs *regs, unsigned int esr)
 	return DBG_HOOK_HANDLED;
 }
 
-bool arch_within_kprobe_blacklist(unsigned long addr)
+int __init arch_populate_kprobe_blacklist(void)
 {
-	if ((addr >= (unsigned long)__kprobes_text_start &&
-	    addr < (unsigned long)__kprobes_text_end) ||
-	    (addr >= (unsigned long)__entry_text_start &&
-	    addr < (unsigned long)__entry_text_end) ||
-	    (addr >= (unsigned long)__idmap_text_start &&
-	    addr < (unsigned long)__idmap_text_end) ||
-	    (addr >= (unsigned long)__hyp_text_start &&
-	    addr < (unsigned long)__hyp_text_end) ||
-	    in_exception_text(addr))
-		return true;
-
-	if (!is_kernel_in_hyp_mode()) {
-		if ((addr >= (unsigned long)__hyp_idmap_text_start &&
-		    addr < (unsigned long)__hyp_idmap_text_end))
-			return true;
-	}
-
-	return false;
+	int ret;
+
+	ret = kprobe_add_area_blacklist((unsigned long)__entry_text_start,
+					(unsigned long)__entry_text_end);
+	if (ret)
+		return ret;
+	ret = kprobe_add_area_blacklist((unsigned long)__irqentry_text_start,
+					(unsigned long)__irqentry_text_end);
+	if (ret)
+		return ret;
+	ret = kprobe_add_area_blacklist((unsigned long)__exception_text_start,
+					(unsigned long)__exception_text_end);
+	if (ret)
+		return ret;
+	ret = kprobe_add_area_blacklist((unsigned long)__idmap_text_start,
+					(unsigned long)__idmap_text_end);
+	if (ret)
+		return ret;
+	ret = kprobe_add_area_blacklist((unsigned long)__hyp_text_start,
+					(unsigned long)__hyp_text_end);
+	if (ret || is_kernel_in_hyp_mode())
+		return ret;
+	ret = kprobe_add_area_blacklist((unsigned long)__hyp_idmap_text_start,
+					(unsigned long)__hyp_idmap_text_end);
+	return ret;
 }
 
 void __kprobes __used *trampoline_probe_handler(struct pt_regs *regs)
Acked-by on v3 4/4
#5 lore raw

Re: [PATCH v3 4/4] arm64: kprobes: Use arch_populate_kprobe_blacklist()

Hi Masami,

Quoted 21 lines · Masami Hiramatsu

Use arch_populate_kprobe_blacklist() instead of arch_within_kprobe_blacklist() so that we can see the full blacklisted symbols under the debugfs.

---
Changes in v3
- Do not populate blacklist in __kprobe_text in arch_populate_kprobe_blacklist(), since it is already populated in populate_kprobe_blacklist().
- Add exception entry text blacklist since those are rejected by in_exception_text().

---
 arch/arm64/kernel/probes/kprobes.c |   45 +++++++++++++++++++++---------------
 1 file changed, 26 insertions(+), 19 deletions(-)
arch/arm64/kernel/probes/kprobes.c +0 -0
index 194262fca5cd..37d913f33a89 100644

arch/arm64/kernel/probes/kprobes.c kprobe_breakpoint_handler(struct pt_regs *regs, unsigned int esr) · line 468

Quoted 4 lines
arch/arm64/kernel/probes/kprobes.c +0 -1
@@ -465,26 +465,33 @@ kprobe_breakpoint_handler(struct pt_regs *regs, unsigned int esr)
 	return DBG_HOOK_HANDLED;
 }
 
-bool arch_within_kprobe_blacklist(unsigned long addr)

I think it would be useful to add a comment here so that we remember why this code is the way it is:

/*
* Provide a blacklist of symbols identifying ranges which cannot be kprobed.
* This blacklist is exposed to userspace via debugfs (kprobes/blacklist). */

With that, you can have my ack for the series:

Catalin -- can you pick these up with that comment added, please?

Cheers,

Will

#6 lore raw

Re: [PATCH v3 4/4] arm64: kprobes: Use arch_populate_kprobe_blacklist()

Quoted 32 lines · On Thu, 14 Feb 2019 15:55:22 +0000 Will Deacon <[email protected]> wrote:

Hi Masami,

Quoted 21 lines · Masami Hiramatsu

Use arch_populate_kprobe_blacklist() instead of arch_within_kprobe_blacklist() so that we can see the full blacklisted symbols under the debugfs.

---
Changes in v3
- Do not populate blacklist in __kprobe_text in arch_populate_kprobe_blacklist(), since it is already populated in populate_kprobe_blacklist().
- Add exception entry text blacklist since those are rejected by in_exception_text().

---
 arch/arm64/kernel/probes/kprobes.c |   45 +++++++++++++++++++++---------------
 1 file changed, 26 insertions(+), 19 deletions(-)
arch/arm64/kernel/probes/kprobes.c +0 -0
index 194262fca5cd..37d913f33a89 100644

arch/arm64/kernel/probes/kprobes.c kprobe_breakpoint_handler(struct pt_regs *regs, unsigned int esr) · line 468

Quoted 4 lines
arch/arm64/kernel/probes/kprobes.c +0 -1
@@ -465,26 +465,33 @@ kprobe_breakpoint_handler(struct pt_regs *regs, unsigned int esr)
 	return DBG_HOOK_HANDLED;
 }
 
-bool arch_within_kprobe_blacklist(unsigned long addr)

I think it would be useful to add a comment here so that we remember why this code is the way it is:

Quoted 4 lines

/*
* Provide a blacklist of symbols identifying ranges which cannot be kprobed.
* This blacklist is exposed to userspace via debugfs (kprobes/blacklist). */

Agreed. This looks good to me too.

Quoted 4 lines

With that, you can have my ack for the series:

Thank you!

Quoted 6 lines

Catalin -- can you pick these up with that comment added, please?

Cheers,

Will

Signature · 1 line
Masami Hiramatsu <[email protected]>
maintainer
#7 lore raw
In reply to Masami Hiramatsu

Hi Masami,

Quoted 3 lines · Masami Hiramatsu

In v3, I rebased on the latest arm64 kernel which includes James' KVM/HYP fixes for kprobes, and fix a reported bugs in [4/4].

I will queue this patches at 5.1-rc1 since the arm64 for-next/core branch is currently based on -rc3 and does not include James' fix. Even if I rebased your patches on top of -rc3, we'd still get a conflict when pushing into mainline, so I'd rather send them after -rc1.

Thanks.

Signature · 1 line
Catalin