Skip to contentSkip to conversation

Use per-CPU temporary mappings for patching on Radix MMU

PATCH In review Christopher M. Riedl 3 participants 24 messages Jul 13, 2021 – Aug 11, 2021

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

v5 cover Use per-CPU temporary mappings for patching on Radix MMU

When compiled with CONFIG_STRICT_KERNEL_RWX, the kernel must create temporary mappings when patching itself. These mappings temporarily override the strict RWX text protections to permit a write. Currently, powerpc allocates a per-CPU VM area for patching. Patching occurs as follows:

1. Map page in per-CPU VM area w/ PAGE_KERNEL protection
2. Patch text
3. Remove the temporary mapping

While the VM area is per-CPU, the mapping is actually inserted into the kernel page tables. Presumably, this could allow another CPU to access the normally write-protected text - either malicously or accidentally - via this same mapping if the address of the VM area is known. Ideally, the mapping should be kept local to the CPU doing the patching [0].

x86 introduced "temporary mm" structs which allow the creation of mappings local to a particular CPU [1]. This series intends to bring the notion of a temporary mm to powerpc's Book3s64 Radix MMU and harden it by using such a mapping for patching a kernel with strict RWX permissions.

The first four patches implement an LKDTM test "proof-of-concept" which exploits the potential vulnerability (ie. the temporary mapping during patching is exposed in the kernel page tables and accessible by other CPUs) using a simple brute-force approach. This test is implemented for both powerpc and x86_64. The test passes on powerpc Radix with this new series, fails on upstream powerpc, passes on upstream x86_64, and fails on an older (ancient) x86_64 tree without the x86_64 temporary mm patches. The remaining patches add support for and use a temporary mm for code patching on powerpc with the Radix MMU.

Tested boot, ftrace, and repeated LKDTM "hijack":
- QEMU+KVM (host: POWER9 Blackbird): Radix MMU w/ KUAP
- QEMU+KVM (host: POWER9 Blackbird): Hash MMU

Tested repeated LKDTM "hijack":
- QEMU+KVM (host: AMD desktop): x86_64 upstream
- QEMU+KVM (host: AMD desktop): x86_64 w/o percpu temp mm to verify the LKDTM "hijack" test fails

Tested boot and ftrace:
- QEMU+TCG: ppc44x (bamboo)
- QEMU+TCG: g5 (mac99)

I also tested with various extra config options enabled as suggested in section 12) in Documentation/process/submit-checklist.rst.

v5: * Only support Book3s64 Radix MMU for now. There are some issues with
the previous implementation on the Hash MMU as pointed out by Nick Piggin. Fixing these is not trivial so we only support the Radix MMU for now. I tried using realmode (no data translation) to patch with Hash to at least avoid exposing the patch mapping to other CPUs but this doesn't reliably work either since we cannot access vmalloc'd space in realmode.
* Use percpu variables for the patching_mm and patching_addr. This avoids the need for synchronization mechanisms entirely. Thanks to Peter Zijlstra for pointing out text_mutex which unfortunately didn't work out without larger code changes in powerpc. Also thanks to Daniel Axtens for comments about using percpu variables for the *percpu* temp mm things off list.

v4: * It's time to revisit this series again since @jpn and @mpe fixed
our known STRICT_*_RWX bugs on powerpc/64s.
* Rebase on linuxppc/next:
commit ee1bc694fbaec ("powerpc/kvm: Fix build error when PPC_MEM_KEYS/PPC_PSERIES=n")
* Completely rework how map_patch() works on book3s64 Hash MMU
* Split the LKDTM x86_64 and powerpc bits into separate patches
* Annotate commit messages with changes from v3 instead of
listing them here completely out-of context...

v3: * Rebase on linuxppc/next: commit 9123e3a74ec7 ("Linux 5.9-rc1")
* Move temporary mm implementation into code-patching.c where it belongs
* Implement LKDTM hijacker test on x86_64 (on IBM time oof) Do
* not use address zero for the patching address in the temporary mm (thanks @dja for pointing this out!)
* Wrap the LKDTM test w/ CONFIG_SMP as suggested by Christophe Leroy
* Comments to clarify PTE pre-allocation and patching addr selection

v2: * Rebase on linuxppc/next:
commit 105fb38124a4 ("powerpc/8xx: Modify ptep_get()")
* Always dirty pte when mapping patch
* Use ppc_inst_len instead of sizeof on instructions
* Declare LKDTM patching addr accessor in header where it belongs

v1: * Rebase on linuxppc/next (4336b9337824)
* Save and restore second hw watchpoint
* Use new ppc_inst_* functions for patching check and in LKDTM test

rfc-v2: * Many fixes and improvements mostly based on extensive feedback
and testing by Christophe Leroy (thanks!).
* Make patching_mm and patching_addr static and move '__ro_after_init' to after the variable name (more common in other parts of the kernel)
* Use 'asm/debug.h' header instead of 'asm/hw_breakpoint.h' to fix PPC64e compile
* Add comment explaining why we use BUG_ON() during the init call to setup for patching later
* Move ptep into patch_mapping to avoid walking page tables a second time when unmapping the temporary mapping
* Use KUAP under non-radix, also manually dirty the PTE for patch mapping on non-BOOK3S_64 platforms
* Properly return any error from __patch_instruction
* Do not use 'memcmp' where a simple comparison is appropriate
* Simplify expression for patch address by removing pointer maths
* Add LKDTM test

[0]: https://github.com/linuxppc/issues/issues/224
[1]: https://lore.kernel.org/kernel-hardening/[email protected]/

Christopher M. Riedl (8):
powerpc: Add LKDTM accessor for patching addr lkdtm/powerpc: Add test to hijack a patch mapping
x86_64: Add LKDTM accessor for patching addr lkdtm/x86_64: Add test to hijack a patch mapping powerpc/64s: Introduce temporary mm for Radix MMU
powerpc: Rework and improve STRICT_KERNEL_RWX patching powerpc/64s: Initialize and use a temporary mm for patching on Radix lkdtm/powerpc: Fix code patching hijack test

 arch/powerpc/include/asm/code-patching.h |   4 +
 arch/powerpc/include/asm/debug.h         |   1 +
 arch/powerpc/kernel/process.c            |   5 +
 arch/powerpc/lib/code-patching.c         | 240 ++++++++++++++++++++---
 arch/x86/include/asm/text-patching.h     |   4 +
 arch/x86/kernel/alternative.c            |   7 +
 drivers/misc/lkdtm/core.c                |   1 +
 drivers/misc/lkdtm/lkdtm.h               |   1 +
 drivers/misc/lkdtm/perms.c               | 143 ++++++++++++++
 9 files changed, 378 insertions(+), 28 deletions(-)
Signature · 1 line
2.26.1
#1 lore raw

PATCH v5 1/8 powerpc: Add LKDTM accessor for patching addr

When live patching with STRICT_KERNEL_RWX a mapping is installed at a "patching address" with temporary write permissions. Provide a LKDTM-only accessor function for this address in preparation for a LKDTM test which attempts to "hijack" this mapping by writing to it from another CPU.

---
 arch/powerpc/include/asm/code-patching.h | 4 ++++
 arch/powerpc/lib/code-patching.c         | 7 +++++++
 2 files changed, 11 insertions(+)
arch/powerpc/include/asm/code-patching.h +4 -0
index a95f63788c6b1..16fbc58a4932f 100644
@@ -184,4 +184,8 @@ static inline unsigned long ppc_kallsyms_lookup_name(const char *name)
 #define PPC_INST_STD_LR		PPC_RAW_STD(_R0, _R1, PPC_LR_STKOFF)
 #endif /* CONFIG_PPC64 */
 
+#if IS_BUILTIN(CONFIG_LKDTM) && IS_ENABLED(CONFIG_STRICT_KERNEL_RWX)
+unsigned long read_cpu_patching_addr(unsigned int cpu);
+#endif
+
 #endif /* _ASM_POWERPC_CODE_PATCHING_H */
arch/powerpc/lib/code-patching.c +7 -0
index f9a3019e37b43..54b6157d44e95 100644
@@ -47,6 +47,13 @@ int raw_patch_instruction(u32 *addr, struct ppc_inst instr)
 #ifdef CONFIG_STRICT_KERNEL_RWX
 static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
 
+#if IS_BUILTIN(CONFIG_LKDTM)
+unsigned long read_cpu_patching_addr(unsigned int cpu)
+{
+	return (unsigned long)(per_cpu(text_poke_area, cpu))->addr;
+}
+#endif
+
 static int text_area_cpu_up(unsigned int cpu)
 {
 	struct vm_struct *area;
Signature · 1 line
2.26.1
#2 lore raw

PATCH v5 2/8 lkdtm/powerpc: Add test to hijack a patch mapping

In reply to Christopher M. Riedl

When live patching with STRICT_KERNEL_RWX the CPU doing the patching must temporarily remap the page(s) containing the patch site with +W permissions. While this temporary mapping is in use, another CPU could write to the same mapping and maliciously alter kernel text. Implement a LKDTM test to attempt to exploit such an opening during code patching. The test is implemented on powerpc and requires LKDTM built into the kernel (building LKDTM as a module is insufficient).

The LKDTM "hijack" test works as follows:

1. A CPU executes an infinite loop to patch an instruction. This is the "patching" CPU.
2. Another CPU attempts to write to the address of the temporary mapping used by the "patching" CPU. This other CPU is the "hijacker" CPU. The hijack either fails with a fault/error or succeeds, in which case some kernel text is now overwritten.

The virtual address of the temporary patch mapping is provided via an LKDTM-specific accessor to the hijacker CPU. This test assumes a hypothetical situation where this address was leaked previously.

How to run the test:

mount -t debugfs none /sys/kernel/debug
(echo HIJACK_PATCH > /sys/kernel/debug/provoke-crash/DIRECT)

A passing test indicates that it is not possible to overwrite kernel text from another CPU by using the temporary mapping established by a CPU for patching.

---

v5: * Use u32* instead of struct ppc_inst* based on new series in
upstream.

v4: * Separate the powerpc and x86_64 bits into individual patches.
* Use __put_kernel_nofault() when attempting to hijack the mapping
* Use raw_smp_processor_id() to avoid triggering the BUG() when calling smp_processor_id() in preemptible code - the only thing that matters is that one of the threads is bound to a different CPU - we are not using smp_processor_id() to access any per-cpu data or similar where preemption should be disabled.
* Rework the patching_cpu() kthread stop condition to avoid:
https://lwn.net/Articles/628628/

---
 drivers/misc/lkdtm/core.c  |   1 +
 drivers/misc/lkdtm/lkdtm.h |   1 +
 drivers/misc/lkdtm/perms.c | 134 +++++++++++++++++++++++++++++++++++++
 3 files changed, 136 insertions(+)
drivers/misc/lkdtm/core.c +1 -0
index 8024b6a5cc7fc..fbcb95eda337b 100644
@@ -147,6 +147,7 @@ static const struct crashtype crashtypes[] = {
 	CRASHTYPE(WRITE_RO),
 	CRASHTYPE(WRITE_RO_AFTER_INIT),
 	CRASHTYPE(WRITE_KERN),
+	CRASHTYPE(HIJACK_PATCH),
 	CRASHTYPE(REFCOUNT_INC_OVERFLOW),
 	CRASHTYPE(REFCOUNT_ADD_OVERFLOW),
 	CRASHTYPE(REFCOUNT_INC_NOT_ZERO_OVERFLOW),
drivers/misc/lkdtm/lkdtm.h +1 -0
index 99f90d3e5e9cb..87e7e6136d962 100644
@@ -62,6 +62,7 @@ void lkdtm_EXEC_USERSPACE(void);
 void lkdtm_EXEC_NULL(void);
 void lkdtm_ACCESS_USERSPACE(void);
 void lkdtm_ACCESS_NULL(void);
+void lkdtm_HIJACK_PATCH(void);
 
 /* refcount.c */
 void lkdtm_REFCOUNT_INC_OVERFLOW(void);
drivers/misc/lkdtm/perms.c +134 -0

146 lines not on this page · raw message

Signature · 1 line
2.26.1
#3 lore raw

Re: [PATCH v5 2/8] lkdtm/powerpc: Add test to hijack a patch mapping

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 89 lines

When live patching with STRICT_KERNEL_RWX the CPU doing the patching must temporarily remap the page(s) containing the patch site with +W permissions. While this temporary mapping is in use, another CPU could write to the same mapping and maliciously alter kernel text. Implement a LKDTM test to attempt to exploit such an opening during code patching. The test is implemented on powerpc and requires LKDTM built into the kernel (building LKDTM as a module is insufficient).

The LKDTM "hijack" test works as follows:

1. A CPU executes an infinite loop to patch an instruction. This is the "patching" CPU.
2. Another CPU attempts to write to the address of the temporary mapping used by the "patching" CPU. This other CPU is the "hijacker" CPU. The hijack either fails with a fault/error or succeeds, in which case some kernel text is now overwritten.

The virtual address of the temporary patch mapping is provided via an LKDTM-specific accessor to the hijacker CPU. This test assumes a hypothetical situation where this address was leaked previously.

How to run the test:

mount -t debugfs none /sys/kernel/debug
(echo HIJACK_PATCH > /sys/kernel/debug/provoke-crash/DIRECT)

A passing test indicates that it is not possible to overwrite kernel text from another CPU by using the temporary mapping established by a CPU for patching.

---

v5: * Use u32* instead of struct ppc_inst* based on new series in
upstream.

v4: * Separate the powerpc and x86_64 bits into individual patches.
* Use __put_kernel_nofault() when attempting to hijack the mapping
* Use raw_smp_processor_id() to avoid triggering the BUG() when calling smp_processor_id() in preemptible code - the only thing that matters is that one of the threads is bound to a different CPU - we are not using smp_processor_id() to access any per-cpu data or similar where preemption should be disabled.
* Rework the patching_cpu() kthread stop condition to avoid:
https://lwn.net/Articles/628628/

---
  drivers/misc/lkdtm/core.c  |   1 +
  drivers/misc/lkdtm/lkdtm.h |   1 +
  drivers/misc/lkdtm/perms.c | 134 +++++++++++++++++++++++++++++++++++++
  3 files changed, 136 insertions(+)
drivers/misc/lkdtm/core.c +1 -0
index 8024b6a5cc7fc..fbcb95eda337b 100644
@@ -147,6 +147,7 @@ static const struct crashtype crashtypes[] = {
  	CRASHTYPE(WRITE_RO),
  	CRASHTYPE(WRITE_RO_AFTER_INIT),
  	CRASHTYPE(WRITE_KERN),
+	CRASHTYPE(HIJACK_PATCH),
  	CRASHTYPE(REFCOUNT_INC_OVERFLOW),
  	CRASHTYPE(REFCOUNT_ADD_OVERFLOW),
  	CRASHTYPE(REFCOUNT_INC_NOT_ZERO_OVERFLOW),
drivers/misc/lkdtm/lkdtm.h +1 -0
index 99f90d3e5e9cb..87e7e6136d962 100644
@@ -62,6 +62,7 @@ void lkdtm_EXEC_USERSPACE(void);
  void lkdtm_EXEC_NULL(void);
  void lkdtm_ACCESS_USERSPACE(void);
  void lkdtm_ACCESS_NULL(void);
+void lkdtm_HIJACK_PATCH(void);
 
  /* refcount.c */
  void lkdtm_REFCOUNT_INC_OVERFLOW(void);
drivers/misc/lkdtm/perms.c +1 -0
index 2dede2ef658f3..39e7456852229 100644
@@ -9,6 +9,7 @@
  #include <linux/vmalloc.h>
  #include <linux/mman.h>
  #include <linux/uaccess.h>
+#include <linux/kthread.h>
  #include <asm/cacheflush.h>
 
  /* Whether or not to fill the target memory area with do_nothing(). */

drivers/misc/lkdtm/perms.c void lkdtm_ACCESS_NULL(void) · line 227

Quoted 5 lines
drivers/misc/lkdtm/perms.c +2 -0
@@ -222,6 +223,139 @@ void lkdtm_ACCESS_NULL(void)
  	pr_err("FAIL: survived bad write\n");
  }
 
+#if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
+	defined(CONFIG_PPC))

I think this test shouldn't be limited to CONFIG_PPC and shouldn't be limited to CONFIG_STRICT_KERNEL_RWX. It should be there all the time.

Also why limiting it to IS_BUILTIN(CONFIG_LKDTM) ?

Quoted 135 lines
drivers/misc/lkdtm/perms.c +131 -0
+/*
+ * This is just a dummy location to patch-over.
+ */
+static void patching_target(void)
+{
+	return;
+}
+
+#include <asm/code-patching.h>
+const u32 *patch_site = (const u32 *)&patching_target;
+
+static inline int lkdtm_do_patch(u32 data)
+{
+	return patch_instruction((u32 *)patch_site, ppc_inst(data));
+}
+
+static inline u32 lkdtm_read_patch_site(void)
+{
+	return READ_ONCE(*patch_site);
+}
+
+/* Returns True if the write succeeds */
+static inline bool lkdtm_try_write(u32 data, u32 *addr)
+{
+	__put_kernel_nofault(addr, &data, u32, err);
+	return true;
+
+err:
+	return false;
+}
+
+static int lkdtm_patching_cpu(void *data)
+{
+	int err = 0;
+	u32 val = 0xdeadbeef;
+
+	pr_info("starting patching_cpu=%d\n", raw_smp_processor_id());
+
+	do {
+		err = lkdtm_do_patch(val);
+	} while (lkdtm_read_patch_site() == val && !err && !kthread_should_stop());
+
+	if (err)
+		pr_warn("XFAIL: patch_instruction returned error: %d\n", err);
+
+	while (!kthread_should_stop()) {
+		set_current_state(TASK_INTERRUPTIBLE);
+		schedule();
+	}
+
+	return err;
+}
+
+void lkdtm_HIJACK_PATCH(void)
+{
+	struct task_struct *patching_kthrd;
+	int patching_cpu, hijacker_cpu, attempts;
+	unsigned long addr;
+	bool hijacked;
+	const u32 bad_data = 0xbad00bad;
+	const u32 original_insn = lkdtm_read_patch_site();
+
+	if (!IS_ENABLED(CONFIG_SMP)) {
+		pr_err("XFAIL: this test requires CONFIG_SMP\n");
+		return;
+	}
+
+	if (num_online_cpus() < 2) {
+		pr_warn("XFAIL: this test requires at least two cpus\n");
+		return;
+	}
+
+	hijacker_cpu = raw_smp_processor_id();
+	patching_cpu = cpumask_any_but(cpu_online_mask, hijacker_cpu);
+
+	patching_kthrd = kthread_create_on_node(&lkdtm_patching_cpu, NULL,
+						cpu_to_node(patching_cpu),
+						"lkdtm_patching_cpu");
+	kthread_bind(patching_kthrd, patching_cpu);
+	wake_up_process(patching_kthrd);
+
+	addr = offset_in_page(patch_site) | read_cpu_patching_addr(patching_cpu);
+
+	pr_info("starting hijacker_cpu=%d\n", hijacker_cpu);
+	for (attempts = 0; attempts < 100000; ++attempts) {
+		/* Try to write to the other CPU's temp patch mapping */
+		hijacked = lkdtm_try_write(bad_data, (u32 *)addr);
+
+		if (hijacked) {
+			if (kthread_stop(patching_kthrd)) {
+				pr_info("hijack attempts: %d\n", attempts);
+				pr_err("XFAIL: error stopping patching cpu\n");
+				return;
+			}
+			break;
+		}
+	}
+	pr_info("hijack attempts: %d\n", attempts);
+
+	if (hijacked) {
+		if (lkdtm_read_patch_site() == bad_data)
+			pr_err("overwrote kernel text\n");
+		/*
+		 * There are window conditions where the hijacker cpu manages to
+		 * write to the patch site but the site gets overwritten again by
+		 * the patching cpu. We still consider that a "successful" hijack
+		 * since the hijacker cpu did not fault on the write.
+		 */
+		pr_err("FAIL: wrote to another cpu's patching area\n");
+	} else {
+		kthread_stop(patching_kthrd);
+	}
+
+	/* Restore the original data to be able to run the test again */
+	lkdtm_do_patch(original_insn);
+}
+
+#else
+
+void lkdtm_HIJACK_PATCH(void)
+{
+	if (!IS_ENABLED(CONFIG_PPC))
+		pr_err("XFAIL: this test only runs on powerpc\n");
+	if (!IS_ENABLED(CONFIG_STRICT_KERNEL_RWX))
+		pr_err("XFAIL: this test requires CONFIG_STRICT_KERNEL_RWX\n");
+	if (!IS_BUILTIN(CONFIG_LKDTM))
+		pr_err("XFAIL: this test requires CONFIG_LKDTM=y (not =m!)\n");
+}
+
+#endif
+
  void __init lkdtm_perms_init(void)
  {
  	/* Make sure we can write to __ro_after_init values during __init */
#4 lore raw

Re: [PATCH v5 2/8] lkdtm/powerpc: Add test to hijack a patch mapping

Quoted 103 lines · Christophe Leroy
Quoted 89 lines · Christopher M. Riedl

When live patching with STRICT_KERNEL_RWX the CPU doing the patching must temporarily remap the page(s) containing the patch site with +W permissions. While this temporary mapping is in use, another CPU could write to the same mapping and maliciously alter kernel text. Implement a LKDTM test to attempt to exploit such an opening during code patching. The test is implemented on powerpc and requires LKDTM built into the kernel (building LKDTM as a module is insufficient).

The LKDTM "hijack" test works as follows:

1. A CPU executes an infinite loop to patch an instruction. This is the "patching" CPU.
2. Another CPU attempts to write to the address of the temporary mapping used by the "patching" CPU. This other CPU is the "hijacker" CPU. The hijack either fails with a fault/error or succeeds, in which case some kernel text is now overwritten.

The virtual address of the temporary patch mapping is provided via an LKDTM-specific accessor to the hijacker CPU. This test assumes a hypothetical situation where this address was leaked previously.

How to run the test:

mount -t debugfs none /sys/kernel/debug
(echo HIJACK_PATCH > /sys/kernel/debug/provoke-crash/DIRECT)

A passing test indicates that it is not possible to overwrite kernel text from another CPU by using the temporary mapping established by a CPU for patching.

---

v5: * Use u32* instead of struct ppc_inst* based on new series in
upstream.

v4: * Separate the powerpc and x86_64 bits into individual patches.
* Use __put_kernel_nofault() when attempting to hijack the mapping
* Use raw_smp_processor_id() to avoid triggering the BUG() when calling smp_processor_id() in preemptible code - the only thing that matters is that one of the threads is bound to a different CPU - we are not using smp_processor_id() to access any per-cpu data or similar where preemption should be disabled.
* Rework the patching_cpu() kthread stop condition to avoid:
https://lwn.net/Articles/628628/

---
  drivers/misc/lkdtm/core.c  |   1 +
  drivers/misc/lkdtm/lkdtm.h |   1 +
  drivers/misc/lkdtm/perms.c | 134 +++++++++++++++++++++++++++++++++++++
  3 files changed, 136 insertions(+)
drivers/misc/lkdtm/core.c +1 -0
index 8024b6a5cc7fc..fbcb95eda337b 100644
@@ -147,6 +147,7 @@ static const struct crashtype crashtypes[] = {
  	CRASHTYPE(WRITE_RO),
  	CRASHTYPE(WRITE_RO_AFTER_INIT),
  	CRASHTYPE(WRITE_KERN),
+	CRASHTYPE(HIJACK_PATCH),
  	CRASHTYPE(REFCOUNT_INC_OVERFLOW),
  	CRASHTYPE(REFCOUNT_ADD_OVERFLOW),
  	CRASHTYPE(REFCOUNT_INC_NOT_ZERO_OVERFLOW),
drivers/misc/lkdtm/lkdtm.h +1 -0
index 99f90d3e5e9cb..87e7e6136d962 100644
@@ -62,6 +62,7 @@ void lkdtm_EXEC_USERSPACE(void);
  void lkdtm_EXEC_NULL(void);
  void lkdtm_ACCESS_USERSPACE(void);
  void lkdtm_ACCESS_NULL(void);
+void lkdtm_HIJACK_PATCH(void);
 
  /* refcount.c */
  void lkdtm_REFCOUNT_INC_OVERFLOW(void);
drivers/misc/lkdtm/perms.c +1 -0
index 2dede2ef658f3..39e7456852229 100644
@@ -9,6 +9,7 @@
  #include <linux/vmalloc.h>
  #include <linux/mman.h>
  #include <linux/uaccess.h>
+#include <linux/kthread.h>
  #include <asm/cacheflush.h>
 
  /* Whether or not to fill the target memory area with do_nothing(). */

drivers/misc/lkdtm/perms.c void lkdtm_ACCESS_NULL(void) · line 227

Quoted 5 lines
drivers/misc/lkdtm/perms.c +2 -0
@@ -222,6 +223,139 @@ void lkdtm_ACCESS_NULL(void)
  	pr_err("FAIL: survived bad write\n");
  }
 
+#if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
+	defined(CONFIG_PPC))

I think this test shouldn't be limited to CONFIG_PPC and shouldn't be limited to
CONFIG_STRICT_KERNEL_RWX. It should be there all the time.

Quoted 2 lines

Also why limiting it to IS_BUILTIN(CONFIG_LKDTM) ?

The test needs read_cpu_patching_addr() which definitely cannot be exposed outside of the kernel (ie. builtin).

Quoted 135 lines
Quoted 135 lines
drivers/misc/lkdtm/perms.c +131 -0
+/*
+ * This is just a dummy location to patch-over.
+ */
+static void patching_target(void)
+{
+	return;
+}
+
+#include <asm/code-patching.h>
+const u32 *patch_site = (const u32 *)&patching_target;
+
+static inline int lkdtm_do_patch(u32 data)
+{
+	return patch_instruction((u32 *)patch_site, ppc_inst(data));
+}
+
+static inline u32 lkdtm_read_patch_site(void)
+{
+	return READ_ONCE(*patch_site);
+}
+
+/* Returns True if the write succeeds */
+static inline bool lkdtm_try_write(u32 data, u32 *addr)
+{
+	__put_kernel_nofault(addr, &data, u32, err);
+	return true;
+
+err:
+	return false;
+}
+
+static int lkdtm_patching_cpu(void *data)
+{
+	int err = 0;
+	u32 val = 0xdeadbeef;
+
+	pr_info("starting patching_cpu=%d\n", raw_smp_processor_id());
+
+	do {
+		err = lkdtm_do_patch(val);
+	} while (lkdtm_read_patch_site() == val && !err && !kthread_should_stop());
+
+	if (err)
+		pr_warn("XFAIL: patch_instruction returned error: %d\n", err);
+
+	while (!kthread_should_stop()) {
+		set_current_state(TASK_INTERRUPTIBLE);
+		schedule();
+	}
+
+	return err;
+}
+
+void lkdtm_HIJACK_PATCH(void)
+{
+	struct task_struct *patching_kthrd;
+	int patching_cpu, hijacker_cpu, attempts;
+	unsigned long addr;
+	bool hijacked;
+	const u32 bad_data = 0xbad00bad;
+	const u32 original_insn = lkdtm_read_patch_site();
+
+	if (!IS_ENABLED(CONFIG_SMP)) {
+		pr_err("XFAIL: this test requires CONFIG_SMP\n");
+		return;
+	}
+
+	if (num_online_cpus() < 2) {
+		pr_warn("XFAIL: this test requires at least two cpus\n");
+		return;
+	}
+
+	hijacker_cpu = raw_smp_processor_id();
+	patching_cpu = cpumask_any_but(cpu_online_mask, hijacker_cpu);
+
+	patching_kthrd = kthread_create_on_node(&lkdtm_patching_cpu, NULL,
+						cpu_to_node(patching_cpu),
+						"lkdtm_patching_cpu");
+	kthread_bind(patching_kthrd, patching_cpu);
+	wake_up_process(patching_kthrd);
+
+	addr = offset_in_page(patch_site) | read_cpu_patching_addr(patching_cpu);
+
+	pr_info("starting hijacker_cpu=%d\n", hijacker_cpu);
+	for (attempts = 0; attempts < 100000; ++attempts) {
+		/* Try to write to the other CPU's temp patch mapping */
+		hijacked = lkdtm_try_write(bad_data, (u32 *)addr);
+
+		if (hijacked) {
+			if (kthread_stop(patching_kthrd)) {
+				pr_info("hijack attempts: %d\n", attempts);
+				pr_err("XFAIL: error stopping patching cpu\n");
+				return;
+			}
+			break;
+		}
+	}
+	pr_info("hijack attempts: %d\n", attempts);
+
+	if (hijacked) {
+		if (lkdtm_read_patch_site() == bad_data)
+			pr_err("overwrote kernel text\n");
+		/*
+		 * There are window conditions where the hijacker cpu manages to
+		 * write to the patch site but the site gets overwritten again by
+		 * the patching cpu. We still consider that a "successful" hijack
+		 * since the hijacker cpu did not fault on the write.
+		 */
+		pr_err("FAIL: wrote to another cpu's patching area\n");
+	} else {
+		kthread_stop(patching_kthrd);
+	}
+
+	/* Restore the original data to be able to run the test again */
+	lkdtm_do_patch(original_insn);
+}
+
+#else
+
+void lkdtm_HIJACK_PATCH(void)
+{
+	if (!IS_ENABLED(CONFIG_PPC))
+		pr_err("XFAIL: this test only runs on powerpc\n");
+	if (!IS_ENABLED(CONFIG_STRICT_KERNEL_RWX))
+		pr_err("XFAIL: this test requires CONFIG_STRICT_KERNEL_RWX\n");
+	if (!IS_BUILTIN(CONFIG_LKDTM))
+		pr_err("XFAIL: this test requires CONFIG_LKDTM=y (not =m!)\n");
+}
+
+#endif
+
  void __init lkdtm_perms_init(void)
  {
  	/* Make sure we can write to __ro_after_init values during __init */
#5 lore raw

Re: [PATCH v5 2/8] lkdtm/powerpc: Add test to hijack a patch mapping

Quoted 14 lines · Christopher M. Riedl
Quoted 13 lines · Christophe Leroy
Quoted 10 lines · Christopher M. Riedl

When live patching with STRICT_KERNEL_RWX the CPU doing the patching must temporarily remap the page(s) containing the patch site with +W permissions. While this temporary mapping is in use, another CPU could write to the same mapping and maliciously alter kernel text. Implement a LKDTM test to attempt to exploit such an opening during code patching. The test is implemented on powerpc and requires LKDTM built into the kernel (building LKDTM as a module is insufficient).

The LKDTM "hijack" test works as follows:

Quoted 12 lines
Quoted 4 lines
Quoted 4 lines

1. A CPU executes an infinite loop to patch an instruction. This is the "patching" CPU.
2. Another CPU attempts to write to the address of the temporary mapping used by the "patching" CPU. This other CPU is the

Quoted 8 lines

drivers/misc/lkdtm/perms.c void lkdtm_ACCESS_NULL(void) · line 227

Quoted 5 lines

"hijacker" CPU. The hijack either fails with a fault/error or succeeds, in which case some kernel text is now overwritten.
[...]

drivers/misc/lkdtm/perms.c +2 -0
+#if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
+	defined(CONFIG_PPC))

I think this test shouldn't be limited to CONFIG_PPC and shouldn't be limited to CONFIG_STRICT_KERNEL_RWX. It should be there all the time.

Agreed: if the machinery exists to provide this defense on even one arch/config/whatever combo, I'd like LKDTM to test for it. This lets use compare defenses across different combinations more easily, and means folks must answer questions like "why doesn't $combination provide $defense?"

Quoted 4 lines
Quoted 1 line

Also why limiting it to IS_BUILTIN(CONFIG_LKDTM) ?

The test needs read_cpu_patching_addr() which definitely cannot be exposed outside of the kernel (ie. builtin).

FWIW, I'm okay with this. There isn't a solution that feels entirely "right", so either a build-time requirement like this, or using an exception for modules like this:

arch/x86/kernel/cpu/common.c:#if IS_MODULE(CONFIG_LKDTM) arch/x86/kernel/cpu/common.c-EXPORT_SYMBOL_GPL(native_write_cr4); arch/x86/kernel/cpu/common.c-#endif

I think neither is great. Another idea is maybe using a name-spaced export, like:

EXPORT_SYMBOL_NS_GPL(native_write_cr4, LKDTM);

But that still means it gets exposed to malicious discovery, so probably not.

I suspect the best is to just do the BUILTIN check, since building LKDTM as a module on a _production_ kernel is rare if it exists at all. The only downside is needing to completely reboot to perform updated tests, but then, I frequently find myself breaking the kernel badly on bad tests, so I have to reboot anyway. ;)

-Kees

Signature · 1 line
Kees Cook
#6 lore raw

PATCH v5 3/8 x86_64: Add LKDTM accessor for patching addr

In reply to Christopher M. Riedl

When live patching with STRICT_KERNEL_RWX a mapping is installed at a "patching address" with temporary write permissions. Provide a LKDTM-only accessor function for this address in preparation for a LKDTM test which attempts to "hijack" this mapping by writing to it from another CPU.

---
 arch/x86/include/asm/text-patching.h | 4 ++++
 arch/x86/kernel/alternative.c        | 7 +++++++
 2 files changed, 11 insertions(+)
arch/x86/include/asm/text-patching.h +4 -0
index b7421780e4e92..f0caf9ee13bd8 100644
@@ -167,4 +167,8 @@ void int3_emulate_ret(struct pt_regs *regs)
 }
 #endif /* !CONFIG_UML_X86 */
 
+#if IS_BUILTIN(CONFIG_LKDTM)
+unsigned long read_cpu_patching_addr(unsigned int cpu);
+#endif
+
 #endif /* _ASM_X86_TEXT_PATCHING_H */
arch/x86/kernel/alternative.c +7 -0
index e9da3dc712541..28bb92b695639 100644
@@ -773,6 +773,13 @@ static inline void unuse_temporary_mm(temp_mm_state_t prev_state)
 __ro_after_init struct mm_struct *poking_mm;
 __ro_after_init unsigned long poking_addr;
 
+#if IS_BUILTIN(CONFIG_LKDTM)
+unsigned long read_cpu_patching_addr(unsigned int cpu)
+{
+	return poking_addr;
+}
+#endif
+
 static void *__text_poke(void *addr, const void *opcode, size_t len)
 {
 	bool cross_page_boundary = offset_in_page(addr) + len > PAGE_SIZE;
Signature · 1 line
2.26.1
#7 lore raw

PATCH v5 4/8 lkdtm/x86_64: Add test to hijack a patch mapping

In reply to Christopher M. Riedl

A previous commit implemented an LKDTM test on powerpc to exploit the temporary mapping established when patching code with STRICT_KERNEL_RWX enabled. Extend the test to work on x86_64 as well.

---
 drivers/misc/lkdtm/perms.c | 26 ++++++++++++++++++++++----
 1 file changed, 22 insertions(+), 4 deletions(-)
drivers/misc/lkdtm/perms.c +22 -4
index 39e7456852229..41e87e5f9cc86 100644
@@ -224,7 +224,7 @@ void lkdtm_ACCESS_NULL(void)
 }
 
 #if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
-	defined(CONFIG_PPC))
+	(defined(CONFIG_PPC) || defined(CONFIG_X86_64)))
 /*
  * This is just a dummy location to patch-over.
  */
@@ -233,12 +233,25 @@ static void patching_target(void)
 	return;
 }
 
-#include <asm/code-patching.h>
 const u32 *patch_site = (const u32 *)&patching_target;
 
+#ifdef CONFIG_PPC
+#include <asm/code-patching.h>
+#endif
+
+#ifdef CONFIG_X86_64
+#include <asm/text-patching.h>
+#endif
+
 static inline int lkdtm_do_patch(u32 data)
 {
+#ifdef CONFIG_PPC
 	return patch_instruction((u32 *)patch_site, ppc_inst(data));
+#endif
+#ifdef CONFIG_X86_64
+	text_poke((void *)patch_site, &data, sizeof(u32));
+	return 0;
+#endif
 }
 
 static inline u32 lkdtm_read_patch_site(void)
@@ -249,11 +262,16 @@ static inline u32 lkdtm_read_patch_site(void)
 /* Returns True if the write succeeds */
 static inline bool lkdtm_try_write(u32 data, u32 *addr)
 {
+#ifdef CONFIG_PPC
 	__put_kernel_nofault(addr, &data, u32, err);
 	return true;
 
 err:
 	return false;
+#endif
+#ifdef CONFIG_X86_64
+	return !__put_user(data, addr);
+#endif
 }
 
 static int lkdtm_patching_cpu(void *data)
@@ -346,8 +364,8 @@ void lkdtm_HIJACK_PATCH(void)
 
 void lkdtm_HIJACK_PATCH(void)
 {
-	if (!IS_ENABLED(CONFIG_PPC))
-		pr_err("XFAIL: this test only runs on powerpc\n");
+	if (!IS_ENABLED(CONFIG_PPC) && !IS_ENABLED(CONFIG_X86_64))
+		pr_err("XFAIL: this test only runs on powerpc and x86_64\n");
 	if (!IS_ENABLED(CONFIG_STRICT_KERNEL_RWX))
 		pr_err("XFAIL: this test requires CONFIG_STRICT_KERNEL_RWX\n");
 	if (!IS_BUILTIN(CONFIG_LKDTM))
Signature · 1 line
2.26.1
#8 lore raw

Re: [PATCH v5 4/8] lkdtm/x86_64: Add test to hijack a patch mapping

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 68 lines

A previous commit implemented an LKDTM test on powerpc to exploit the temporary mapping established when patching code with STRICT_KERNEL_RWX enabled. Extend the test to work on x86_64 as well.

---
  drivers/misc/lkdtm/perms.c | 26 ++++++++++++++++++++++----
  1 file changed, 22 insertions(+), 4 deletions(-)
drivers/misc/lkdtm/perms.c +20 -2
index 39e7456852229..41e87e5f9cc86 100644
@@ -224,7 +224,7 @@ void lkdtm_ACCESS_NULL(void)
  }
 
  #if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
-	defined(CONFIG_PPC))
+	(defined(CONFIG_PPC) || defined(CONFIG_X86_64)))
  /*
   * This is just a dummy location to patch-over.
   */
@@ -233,12 +233,25 @@ static void patching_target(void)
  	return;
  }
 
-#include <asm/code-patching.h>
  const u32 *patch_site = (const u32 *)&patching_target;
 
+#ifdef CONFIG_PPC
+#include <asm/code-patching.h>
+#endif
+
+#ifdef CONFIG_X86_64
+#include <asm/text-patching.h>
+#endif
+
  static inline int lkdtm_do_patch(u32 data)
  {
+#ifdef CONFIG_PPC
  	return patch_instruction((u32 *)patch_site, ppc_inst(data));
+#endif
+#ifdef CONFIG_X86_64
+	text_poke((void *)patch_site, &data, sizeof(u32));
+	return 0;
+#endif
  }
 
  static inline u32 lkdtm_read_patch_site(void)
@@ -249,11 +262,16 @@ static inline u32 lkdtm_read_patch_site(void)
  /* Returns True if the write succeeds */
  static inline bool lkdtm_try_write(u32 data, u32 *addr)
  {
+#ifdef CONFIG_PPC
  	__put_kernel_nofault(addr, &data, u32, err);
  	return true;
 
  err:
  	return false;
+#endif
+#ifdef CONFIG_X86_64
+	return !__put_user(data, addr);
+#endif
  }
 
  static int lkdtm_patching_cpu(void *data)

drivers/misc/lkdtm/perms.c void lkdtm_HIJACK_PATCH(void) · line 371

Quoted 10 lines
drivers/misc/lkdtm/perms.c +2 -2
@@ -346,8 +364,8 @@ void lkdtm_HIJACK_PATCH(void)
 
  void lkdtm_HIJACK_PATCH(void)
  {
-	if (!IS_ENABLED(CONFIG_PPC))
-		pr_err("XFAIL: this test only runs on powerpc\n");
+	if (!IS_ENABLED(CONFIG_PPC) && !IS_ENABLED(CONFIG_X86_64))
+		pr_err("XFAIL: this test only runs on powerpc and x86_64\n");
  	if (!IS_ENABLED(CONFIG_STRICT_KERNEL_RWX))
  		pr_err("XFAIL: this test requires CONFIG_STRICT_KERNEL_RWX\n");
  	if (!IS_BUILTIN(CONFIG_LKDTM))

Instead of spreading arch specific stuff into LKDTM, wouldn't it make sence to define common a common API ? Because the day another arch like arm64 implements it own approach, do we add specific functions again and again into LKDTM ?

Also, I find it odd to define tests only when they can succeed. For other tests like ACCESS_USERSPACE, they are there all the time, regardless of whether we have selected CONFIG_PPC_KUAP or not. I think it should be the same here, have it all there time, if CONFIG_STRICT_KERNEL_RWX is selected the test succeeds otherwise it fails, but it is always there.

Christophe

#9 lore raw

Re: [PATCH v5 4/8] lkdtm/x86_64: Add test to hijack a patch mapping

Quoted 71 lines · Christophe Leroy
Quoted 68 lines · Christopher M. Riedl

A previous commit implemented an LKDTM test on powerpc to exploit the temporary mapping established when patching code with STRICT_KERNEL_RWX enabled. Extend the test to work on x86_64 as well.

---
  drivers/misc/lkdtm/perms.c | 26 ++++++++++++++++++++++----
  1 file changed, 22 insertions(+), 4 deletions(-)
drivers/misc/lkdtm/perms.c +20 -2
index 39e7456852229..41e87e5f9cc86 100644
@@ -224,7 +224,7 @@ void lkdtm_ACCESS_NULL(void)
  }
 
  #if (IS_BUILTIN(CONFIG_LKDTM) && defined(CONFIG_STRICT_KERNEL_RWX) && \
-	defined(CONFIG_PPC))
+	(defined(CONFIG_PPC) || defined(CONFIG_X86_64)))
  /*
   * This is just a dummy location to patch-over.
   */
@@ -233,12 +233,25 @@ static void patching_target(void)
  	return;
  }
 
-#include <asm/code-patching.h>
  const u32 *patch_site = (const u32 *)&patching_target;
 
+#ifdef CONFIG_PPC
+#include <asm/code-patching.h>
+#endif
+
+#ifdef CONFIG_X86_64
+#include <asm/text-patching.h>
+#endif
+
  static inline int lkdtm_do_patch(u32 data)
  {
+#ifdef CONFIG_PPC
  	return patch_instruction((u32 *)patch_site, ppc_inst(data));
+#endif
+#ifdef CONFIG_X86_64
+	text_poke((void *)patch_site, &data, sizeof(u32));
+	return 0;
+#endif
  }
 
  static inline u32 lkdtm_read_patch_site(void)
@@ -249,11 +262,16 @@ static inline u32 lkdtm_read_patch_site(void)
  /* Returns True if the write succeeds */
  static inline bool lkdtm_try_write(u32 data, u32 *addr)
  {
+#ifdef CONFIG_PPC
  	__put_kernel_nofault(addr, &data, u32, err);
  	return true;
 
  err:
  	return false;
+#endif
+#ifdef CONFIG_X86_64
+	return !__put_user(data, addr);
+#endif
  }
 
  static int lkdtm_patching_cpu(void *data)
@@ -346,8 +364,8 @@ void lkdtm_HIJACK_PATCH(void)
 
Quoted 16 lines

drivers/misc/lkdtm/perms.c void lkdtm_HIJACK_PATCH(void) · line 371

Quoted 10 lines
drivers/misc/lkdtm/perms.c +2 -2
  void lkdtm_HIJACK_PATCH(void)
  {
-	if (!IS_ENABLED(CONFIG_PPC))
-		pr_err("XFAIL: this test only runs on powerpc\n");
+	if (!IS_ENABLED(CONFIG_PPC) && !IS_ENABLED(CONFIG_X86_64))
+		pr_err("XFAIL: this test only runs on powerpc and x86_64\n");
  	if (!IS_ENABLED(CONFIG_STRICT_KERNEL_RWX))
  		pr_err("XFAIL: this test requires CONFIG_STRICT_KERNEL_RWX\n");
  	if (!IS_BUILTIN(CONFIG_LKDTM))

Instead of spreading arch specific stuff into LKDTM, wouldn't it make sence to define common a
common API ? Because the day another arch like arm64 implements it own approach, do we add specific
functions again and again into LKDTM ?

Hmm a common patch/poke kernel API is probably out of scope for this series? I do agree though - since you suggested splitting the series maybe that's something I can add along with the LKDTM patches.

Quoted 9 lines

Also, I find it odd to define tests only when they can succeed. For other tests like
ACCESS_USERSPACE, they are there all the time, regardless of whether we have selected
CONFIG_PPC_KUAP or not. I think it should be the same here, have it all there time, if
CONFIG_STRICT_KERNEL_RWX is selected the test succeeds otherwise it fails, but it is always there.

I followed the approach in lkdtm_DOUBLE_FAULT and others in drivers/misc/lkdtm/bugs.c. I suppose it doesn't hurt to always build the test irrespective of CONFIG_STRICT_KERNEL_RWX.

Quoted 2 lines

Christophe

#10 lore raw

PATCH v5 5/8 powerpc/64s: Introduce temporary mm for Radix MMU

In reply to Christopher M. Riedl

x86 supports the notion of a temporary mm which restricts access to temporary PTEs to a single CPU. A temporary mm is useful for situations where a CPU needs to perform sensitive operations (such as patching a STRICT_KERNEL_RWX kernel) requiring temporary mappings without exposing said mappings to other CPUs. Another benefit is that other CPU TLBs do not need to be flushed when the temporary mm is torn down.

Mappings in the temporary mm can be set in the userspace portion of the address-space.

Interrupts must be disabled while the temporary mm is in use. HW breakpoints, which may have been set by userspace as watchpoints on addresses now within the temporary mm, are saved and disabled when loading the temporary mm. The HW breakpoints are restored when unloading the temporary mm. All HW breakpoints are indiscriminately disabled while the temporary mm is in use.

Based on x86 implementation:

commit cefa929c034e
("x86/mm: Introduce temporary mm structs")

---

v5: * Drop support for using a temporary mm on Book3s64 Hash MMU.

v4: * Pass the prev mm instead of NULL to switch_mm_irqs_off() when
using/unusing the temp mm as suggested by Jann Horn to keep the context.active counter in-sync on mm/nohash.
* Disable SLB preload in the temporary mm when initializing the temp_mm struct.
* Include asm/debug.h header to fix build issue with ppc44x_defconfig.

---
 arch/powerpc/include/asm/debug.h |  1 +
 arch/powerpc/kernel/process.c    |  5 +++
 arch/powerpc/lib/code-patching.c | 56 ++++++++++++++++++++++++++++++++
 3 files changed, 62 insertions(+)
arch/powerpc/include/asm/debug.h +1 -0
index 86a14736c76c3..dfd82635ea8b3 100644
@@ -46,6 +46,7 @@ static inline int debugger_fault_handler(struct pt_regs *regs) { return 0; }
 #endif
 
 void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk);
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk);
 bool ppc_breakpoint_available(void);
 #ifdef CONFIG_PPC_ADV_DEBUG_REGS
 extern void do_send_trap(struct pt_regs *regs, unsigned long address,
arch/powerpc/kernel/process.c +5 -0
index 185beb2905801..a0776200772e8 100644
@@ -865,6 +865,11 @@ static inline int set_breakpoint_8xx(struct arch_hw_breakpoint *brk)
 	return 0;
 }
 
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk)
+{
+	memcpy(brk, this_cpu_ptr(&current_brk[nr]), sizeof(*brk));
+}
+
 void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk)
 {
 	memcpy(this_cpu_ptr(&current_brk[nr]), brk, sizeof(*brk));
arch/powerpc/lib/code-patching.c +56 -0

68 lines not on this page · raw message

Signature · 1 line
2.26.1
#11 lore raw

Re: [PATCH v5 5/8] powerpc/64s: Introduce temporary mm for Radix MMU

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 10 lines

x86 supports the notion of a temporary mm which restricts access to temporary PTEs to a single CPU. A temporary mm is useful for situations where a CPU needs to perform sensitive operations (such as patching a STRICT_KERNEL_RWX kernel) requiring temporary mappings without exposing said mappings to other CPUs. Another benefit is that other CPU TLBs do not need to be flushed when the temporary mm is torn down.

Mappings in the temporary mm can be set in the userspace portion of the address-space.

Quoted 6 lines

Interrupts must be disabled while the temporary mm is in use. HW breakpoints, which may have been set by userspace as watchpoints on addresses now within the temporary mm, are saved and disabled when loading the temporary mm. The HW breakpoints are restored when unloading the temporary mm. All HW breakpoints are indiscriminately disabled while the temporary mm is in use.

Can you explain more about that breakpoint stuff ? Why is it a special case here at all ? Isn't it the same when you switch from one user task to another one ? x86 commit doesn't say anythink about breakpoints.

Quoted 100 lines

Based on x86 implementation:

commit cefa929c034e
("x86/mm: Introduce temporary mm structs")

---

v5: * Drop support for using a temporary mm on Book3s64 Hash MMU.

v4: * Pass the prev mm instead of NULL to switch_mm_irqs_off() when
using/unusing the temp mm as suggested by Jann Horn to keep the context.active counter in-sync on mm/nohash.
* Disable SLB preload in the temporary mm when initializing the temp_mm struct.
* Include asm/debug.h header to fix build issue with ppc44x_defconfig.

---
  arch/powerpc/include/asm/debug.h |  1 +
  arch/powerpc/kernel/process.c    |  5 +++
  arch/powerpc/lib/code-patching.c | 56 ++++++++++++++++++++++++++++++++
  3 files changed, 62 insertions(+)
arch/powerpc/include/asm/debug.h +1 -0
index 86a14736c76c3..dfd82635ea8b3 100644
@@ -46,6 +46,7 @@ static inline int debugger_fault_handler(struct pt_regs *regs) { return 0; }
  #endif
 
  void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk);
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk);
  bool ppc_breakpoint_available(void);
  #ifdef CONFIG_PPC_ADV_DEBUG_REGS
  extern void do_send_trap(struct pt_regs *regs, unsigned long address,
arch/powerpc/kernel/process.c +5 -0
index 185beb2905801..a0776200772e8 100644
@@ -865,6 +865,11 @@ static inline int set_breakpoint_8xx(struct arch_hw_breakpoint *brk)
  	return 0;
  }
 
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk)
+{
+	memcpy(brk, this_cpu_ptr(&current_brk[nr]), sizeof(*brk));
+}
+
  void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk)
  {
  	memcpy(this_cpu_ptr(&current_brk[nr]), brk, sizeof(*brk));
arch/powerpc/lib/code-patching.c +32 -0
index 54b6157d44e95..3122d8e4cc013 100644
@@ -17,6 +17,9 @@
  #include <asm/code-patching.h>
  #include <asm/setup.h>
  #include <asm/inst.h>
+#include <asm/mmu_context.h>
+#include <asm/debug.h>
+#include <asm/tlb.h>
 
  static int __patch_instruction(u32 *exec_addr, struct ppc_inst instr, u32 *patch_addr)
  {
@@ -45,6 +48,59 @@ int raw_patch_instruction(u32 *addr, struct ppc_inst instr)
  }
 
  #ifdef CONFIG_STRICT_KERNEL_RWX
+
+struct temp_mm {
+	struct mm_struct *temp;
+	struct mm_struct *prev;
+	struct arch_hw_breakpoint brk[HBP_NUM_MAX];
+};
+
+static inline void init_temp_mm(struct temp_mm *temp_mm, struct mm_struct *mm)
+{
+	/* We currently only support temporary mm on the Book3s64 Radix MMU */
+	WARN_ON(!radix_enabled());
+
+	temp_mm->temp = mm;
+	temp_mm->prev = NULL;
+	memset(&temp_mm->brk, 0, sizeof(temp_mm->brk));
+}
+
+static inline void use_temporary_mm(struct temp_mm *temp_mm)
+{
+	lockdep_assert_irqs_disabled();
+
+	temp_mm->prev = current->active_mm;
+	switch_mm_irqs_off(temp_mm->prev, temp_mm->temp, current);
+
+	WARN_ON(!mm_is_thread_local(temp_mm->temp));
+
+	if (ppc_breakpoint_available()) {
+		struct arch_hw_breakpoint null_brk = {0};
+		int i = 0;

arch/powerpc/lib/code-patching.c int raw_patch_instruction(u32 *addr, struct ppc_inst instr) · line 89

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+
+		for (; i < nr_wp_slots(); ++i) {
+			__get_breakpoint(i, &temp_mm->brk[i]);
+			if (temp_mm->brk[i].type != 0)
+				__set_breakpoint(i, &null_brk);
+		}
+	}
+}
+
+static inline void unuse_temporary_mm(struct temp_mm *temp_mm)

not sure about the naming.

Maybe start_using_temp_mm() and stop_using_temp_mm() would be more explicit.

Quoted 8 lines
arch/powerpc/lib/code-patching.c +7 -0
+{
+	lockdep_assert_irqs_disabled();
+
+	switch_mm_irqs_off(temp_mm->temp, temp_mm->prev, current);
+
+	if (ppc_breakpoint_available()) {
+		int i = 0;

arch/powerpc/lib/code-patching.c int raw_patch_instruction(u32 *addr, struct ppc_inst instr) · line 106

Quoted 10 lines
arch/powerpc/lib/code-patching.c +7 -0
+
+		for (; i < nr_wp_slots(); ++i)
+			if (temp_mm->brk[i].type != 0)
+				__set_breakpoint(i, &temp_mm->brk[i]);
+	}
+}
+
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
 
  #if IS_BUILTIN(CONFIG_LKDTM)

You'll probably get a bisecting hasard with those unused 'static inline' functions in a .c file because that patch alone probably fails build.

#12 lore raw

Re: [PATCH v5 5/8] powerpc/64s: Introduce temporary mm for Radix MMU

Quoted 20 lines · Christophe Leroy
Quoted 10 lines · Christopher M. Riedl

x86 supports the notion of a temporary mm which restricts access to temporary PTEs to a single CPU. A temporary mm is useful for situations where a CPU needs to perform sensitive operations (such as patching a STRICT_KERNEL_RWX kernel) requiring temporary mappings without exposing said mappings to other CPUs. Another benefit is that other CPU TLBs do not need to be flushed when the temporary mm is torn down.

Mappings in the temporary mm can be set in the userspace portion of the address-space.

Quoted 6 lines

Interrupts must be disabled while the temporary mm is in use. HW breakpoints, which may have been set by userspace as watchpoints on addresses now within the temporary mm, are saved and disabled when loading the temporary mm. The HW breakpoints are restored when unloading the temporary mm. All HW breakpoints are indiscriminately disabled while the temporary mm is in use.

Quoted 6 lines

Can you explain more about that breakpoint stuff ? Why is it a special case here at all ? Isn't it
the same when you switch from one user task to another one ? x86 commit doesn't say anythink about
breakpoints.

We do not check if the breakpoint is on a kernel address (perf can do this IIUC) and just disable all of them. I had to dig, but x86 has a comment with their implementation at arch/x86/kernel/alternative.c:743.

I can reword that part of the commit message if it's unclear.

Quoted 113 lines
Quoted 100 lines

Based on x86 implementation:

commit cefa929c034e
("x86/mm: Introduce temporary mm structs")

---

v5: * Drop support for using a temporary mm on Book3s64 Hash MMU.

v4: * Pass the prev mm instead of NULL to switch_mm_irqs_off() when
using/unusing the temp mm as suggested by Jann Horn to keep the context.active counter in-sync on mm/nohash.
* Disable SLB preload in the temporary mm when initializing the temp_mm struct.
* Include asm/debug.h header to fix build issue with ppc44x_defconfig.

---
  arch/powerpc/include/asm/debug.h |  1 +
  arch/powerpc/kernel/process.c    |  5 +++
  arch/powerpc/lib/code-patching.c | 56 ++++++++++++++++++++++++++++++++
  3 files changed, 62 insertions(+)
arch/powerpc/include/asm/debug.h +1 -0
index 86a14736c76c3..dfd82635ea8b3 100644
@@ -46,6 +46,7 @@ static inline int debugger_fault_handler(struct pt_regs *regs) { return 0; }
  #endif
 
  void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk);
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk);
  bool ppc_breakpoint_available(void);
  #ifdef CONFIG_PPC_ADV_DEBUG_REGS
  extern void do_send_trap(struct pt_regs *regs, unsigned long address,
arch/powerpc/kernel/process.c +5 -0
index 185beb2905801..a0776200772e8 100644
@@ -865,6 +865,11 @@ static inline int set_breakpoint_8xx(struct arch_hw_breakpoint *brk)
  	return 0;
  }
 
+void __get_breakpoint(int nr, struct arch_hw_breakpoint *brk)
+{
+	memcpy(brk, this_cpu_ptr(&current_brk[nr]), sizeof(*brk));
+}
+
  void __set_breakpoint(int nr, struct arch_hw_breakpoint *brk)
  {
  	memcpy(this_cpu_ptr(&current_brk[nr]), brk, sizeof(*brk));
arch/powerpc/lib/code-patching.c +32 -0
index 54b6157d44e95..3122d8e4cc013 100644
@@ -17,6 +17,9 @@
  #include <asm/code-patching.h>
  #include <asm/setup.h>
  #include <asm/inst.h>
+#include <asm/mmu_context.h>
+#include <asm/debug.h>
+#include <asm/tlb.h>
 
  static int __patch_instruction(u32 *exec_addr, struct ppc_inst instr, u32 *patch_addr)
  {
@@ -45,6 +48,59 @@ int raw_patch_instruction(u32 *addr, struct ppc_inst instr)
  }
 
  #ifdef CONFIG_STRICT_KERNEL_RWX
+
+struct temp_mm {
+	struct mm_struct *temp;
+	struct mm_struct *prev;
+	struct arch_hw_breakpoint brk[HBP_NUM_MAX];
+};
+
+static inline void init_temp_mm(struct temp_mm *temp_mm, struct mm_struct *mm)
+{
+	/* We currently only support temporary mm on the Book3s64 Radix MMU */
+	WARN_ON(!radix_enabled());
+
+	temp_mm->temp = mm;
+	temp_mm->prev = NULL;
+	memset(&temp_mm->brk, 0, sizeof(temp_mm->brk));
+}
+
+static inline void use_temporary_mm(struct temp_mm *temp_mm)
+{
+	lockdep_assert_irqs_disabled();
+
+	temp_mm->prev = current->active_mm;
+	switch_mm_irqs_off(temp_mm->prev, temp_mm->temp, current);
+
+	WARN_ON(!mm_is_thread_local(temp_mm->temp));
+
+	if (ppc_breakpoint_available()) {
+		struct arch_hw_breakpoint null_brk = {0};
+		int i = 0;

arch/powerpc/lib/code-patching.c int raw_patch_instruction(u32 *addr, struct ppc_inst instr) · line 89

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+
+		for (; i < nr_wp_slots(); ++i) {
+			__get_breakpoint(i, &temp_mm->brk[i]);
+			if (temp_mm->brk[i].type != 0)
+				__set_breakpoint(i, &null_brk);
+		}
+	}
+}
+
+static inline void unuse_temporary_mm(struct temp_mm *temp_mm)

not sure about the naming.

Quoted 3 lines

Maybe start_using_temp_mm() and stop_using_temp_mm() would be more explicit.

Hehe I think we've discussed this before - naming things is hard :) I'll take your suggestions for the next spin.

Quoted 17 lines
Quoted 17 lines
arch/powerpc/lib/code-patching.c +14 -0
+{
+	lockdep_assert_irqs_disabled();
+
+	switch_mm_irqs_off(temp_mm->temp, temp_mm->prev, current);
+
+	if (ppc_breakpoint_available()) {
+		int i = 0;
+
+		for (; i < nr_wp_slots(); ++i)
+			if (temp_mm->brk[i].type != 0)
+				__set_breakpoint(i, &temp_mm->brk[i]);
+	}
+}
+
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
 
Quoted 6 lines

arch/powerpc/lib/code-patching.c int raw_patch_instruction(u32 *addr, struct ppc_inst instr) · line 106

Quoted 2 lines
arch/powerpc/lib/code-patching.c +0 -0
  #if IS_BUILTIN(CONFIG_LKDTM)

You'll probably get a bisecting hasard with those unused 'static inline' functions in a .c file
because that patch alone probably fails build.

I just built the patch without any issue. The compiler only complains for unused 'static' (non-inline) functions right?

#13 lore raw

PATCH v5 6/8 powerpc: Rework and improve STRICT_KERNEL_RWX patching

In reply to Christopher M. Riedl

Rework code-patching with STRICT_KERNEL_RWX to prepare for the next patch which uses a temporary mm for patching under the Book3s64 Radix MMU. Make improvements by adding a WARN_ON when the patchsite doesn't match after patching and return the error from __patch_instruction() properly.

---

v5: * New to series.

---
 arch/powerpc/lib/code-patching.c | 51 +++++++++++++++++---------------
 1 file changed, 27 insertions(+), 24 deletions(-)
arch/powerpc/lib/code-patching.c +27 -24
index 3122d8e4cc013..9f2eba9b70ee4 100644
@@ -102,11 +102,12 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
 }
 
 static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
+static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
 
 #if IS_BUILTIN(CONFIG_LKDTM)
 unsigned long read_cpu_patching_addr(unsigned int cpu)
 {
-	return (unsigned long)(per_cpu(text_poke_area, cpu))->addr;
+	return per_cpu(cpu_patching_addr, cpu);
 }
 #endif
 
@@ -121,6 +122,7 @@ static int text_area_cpu_up(unsigned int cpu)
 		return -1;
 	}
 	this_cpu_write(text_poke_area, area);
+	this_cpu_write(cpu_patching_addr, (unsigned long)area->addr);
 
 	return 0;
 }
@@ -146,7 +148,7 @@ void __init poking_init(void)
 /*
  * This can be called for kernel text or a module.
  */
-static int map_patch_area(void *addr, unsigned long text_poke_addr)
+static int map_patch_area(void *addr)
 {
 	unsigned long pfn;
 	int err;
@@ -156,17 +158,20 @@ static int map_patch_area(void *addr, unsigned long text_poke_addr)
 	else
 		pfn = __pa_symbol(addr) >> PAGE_SHIFT;
 
-	err = map_kernel_page(text_poke_addr, (pfn << PAGE_SHIFT), PAGE_KERNEL);
+	err = map_kernel_page(__this_cpu_read(cpu_patching_addr),
+			      (pfn << PAGE_SHIFT), PAGE_KERNEL);
 
-	pr_devel("Mapped addr %lx with pfn %lx:%d\n", text_poke_addr, pfn, err);
+	pr_devel("Mapped addr %lx with pfn %lx:%d\n",
+		 __this_cpu_read(cpu_patching_addr), pfn, err);
 	if (err)
 		return -1;
 
 	return 0;
 }
 
-static inline int unmap_patch_area(unsigned long addr)
+static inline int unmap_patch_area(void)
 {
+	unsigned long addr = __this_cpu_read(cpu_patching_addr);
 	pte_t *ptep;
 	pmd_t *pmdp;
 	pud_t *pudp;
@@ -175,23 +180,23 @@ static inline int unmap_patch_area(unsigned long addr)
 
 	pgdp = pgd_offset_k(addr);
 	if (unlikely(!pgdp))
-		return -EINVAL;
+		goto out_err;
 
 	p4dp = p4d_offset(pgdp, addr);
 	if (unlikely(!p4dp))
-		return -EINVAL;
+		goto out_err;
 
 	pudp = pud_offset(p4dp, addr);
 	if (unlikely(!pudp))
-		return -EINVAL;
+		goto out_err;
 
 	pmdp = pmd_offset(pudp, addr);
 	if (unlikely(!pmdp))
-		return -EINVAL;
+		goto out_err;
 
 	ptep = pte_offset_kernel(pmdp, addr);
 	if (unlikely(!ptep))
-		return -EINVAL;
+		goto out_err;
 
 	pr_devel("clearing mm %p, pte %p, addr %lx\n", &init_mm, ptep, addr);
 
@@ -202,15 +207,17 @@ static inline int unmap_patch_area(unsigned long addr)
 	flush_tlb_kernel_range(addr, addr + PAGE_SIZE);
 
 	return 0;
+
+out_err:
+	pr_warn("failed to unmap %lx\n", addr);
+	return -EINVAL;
 }
 
 static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
 {
-	int err;
+	int err, rc = 0;
 	u32 *patch_addr = NULL;
 	unsigned long flags;
-	unsigned long text_poke_addr;
-	unsigned long kaddr = (unsigned long)addr;
 
 	/*
 	 * During early early boot patch_instruction is called
@@ -222,24 +229,20 @@ static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
 
 	local_irq_save(flags);
 
-	text_poke_addr = (unsigned long)__this_cpu_read(text_poke_area)->addr;
-	if (map_patch_area(addr, text_poke_addr)) {
-		err = -1;
+	err = map_patch_area(addr);
+	if (err)
 		goto out;
-	}
-
-	patch_addr = (u32 *)(text_poke_addr + (kaddr & ~PAGE_MASK));
 
-	__patch_instruction(addr, instr, patch_addr);
+	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
+	rc = __patch_instruction(addr, instr, patch_addr);
 
-	err = unmap_patch_area(text_poke_addr);
-	if (err)
-		pr_warn("failed to unmap %lx\n", text_poke_addr);
+	err = unmap_patch_area();
 
 out:
 	local_irq_restore(flags);
+	WARN_ON(!ppc_inst_equal(ppc_inst_read(addr), instr));
 
-	return err;
+	return rc ? rc : err;
 }
 #else /* !CONFIG_STRICT_KERNEL_RWX */
 
Signature · 1 line
2.26.1
#14 lore raw

Re: [PATCH v5 6/8] powerpc: Rework and improve STRICT_KERNEL_RWX patching

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 104 lines

Rework code-patching with STRICT_KERNEL_RWX to prepare for the next patch which uses a temporary mm for patching under the Book3s64 Radix MMU. Make improvements by adding a WARN_ON when the patchsite doesn't match after patching and return the error from __patch_instruction() properly.

---

v5: * New to series.

---
  arch/powerpc/lib/code-patching.c | 51 +++++++++++++++++---------------
  1 file changed, 27 insertions(+), 24 deletions(-)
arch/powerpc/lib/code-patching.c +15 -10
index 3122d8e4cc013..9f2eba9b70ee4 100644
@@ -102,11 +102,12 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
  }
 
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
+static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
 
  #if IS_BUILTIN(CONFIG_LKDTM)
  unsigned long read_cpu_patching_addr(unsigned int cpu)
  {
-	return (unsigned long)(per_cpu(text_poke_area, cpu))->addr;
+	return per_cpu(cpu_patching_addr, cpu);
  }
  #endif
 
@@ -121,6 +122,7 @@ static int text_area_cpu_up(unsigned int cpu)
  		return -1;
  	}
  	this_cpu_write(text_poke_area, area);
+	this_cpu_write(cpu_patching_addr, (unsigned long)area->addr);
 
  	return 0;
  }
@@ -146,7 +148,7 @@ void __init poking_init(void)
  /*
   * This can be called for kernel text or a module.
   */
-static int map_patch_area(void *addr, unsigned long text_poke_addr)
+static int map_patch_area(void *addr)
  {
  	unsigned long pfn;
  	int err;
@@ -156,17 +158,20 @@ static int map_patch_area(void *addr, unsigned long text_poke_addr)
  	else
  		pfn = __pa_symbol(addr) >> PAGE_SHIFT;
 
-	err = map_kernel_page(text_poke_addr, (pfn << PAGE_SHIFT), PAGE_KERNEL);
+	err = map_kernel_page(__this_cpu_read(cpu_patching_addr),
+			      (pfn << PAGE_SHIFT), PAGE_KERNEL);
 
-	pr_devel("Mapped addr %lx with pfn %lx:%d\n", text_poke_addr, pfn, err);
+	pr_devel("Mapped addr %lx with pfn %lx:%d\n",
+		 __this_cpu_read(cpu_patching_addr), pfn, err);
  	if (err)
  		return -1;
 
  	return 0;
  }
 
-static inline int unmap_patch_area(unsigned long addr)
+static inline int unmap_patch_area(void)
  {
+	unsigned long addr = __this_cpu_read(cpu_patching_addr);
  	pte_t *ptep;
  	pmd_t *pmdp;
  	pud_t *pudp;
@@ -175,23 +180,23 @@ static inline int unmap_patch_area(unsigned long addr)
 
  	pgdp = pgd_offset_k(addr);
  	if (unlikely(!pgdp))
-		return -EINVAL;
+		goto out_err;
 
  	p4dp = p4d_offset(pgdp, addr);
  	if (unlikely(!p4dp))
-		return -EINVAL;
+		goto out_err;
 
  	pudp = pud_offset(p4dp, addr);
  	if (unlikely(!pudp))
-		return -EINVAL;
+		goto out_err;
 
  	pmdp = pmd_offset(pudp, addr);
  	if (unlikely(!pmdp))
-		return -EINVAL;
+		goto out_err;
 
  	ptep = pte_offset_kernel(pmdp, addr);
  	if (unlikely(!ptep))
-		return -EINVAL;
+		goto out_err;
 
  	pr_devel("clearing mm %p, pte %p, addr %lx\n", &init_mm, ptep, addr);
 

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(unsigned long addr) · line 213

Quoted 7 lines
arch/powerpc/lib/code-patching.c +4 -0
@@ -202,15 +207,17 @@ static inline int unmap_patch_area(unsigned long addr)
  	flush_tlb_kernel_range(addr, addr + PAGE_SIZE);
 
  	return 0;
+
+out_err:
+	pr_warn("failed to unmap %lx\n", addr);
+	return -EINVAL;

Can you keep that in the caller of unmap_patch_area() instead of all those goto stuff ?

Quoted 29 lines
arch/powerpc/lib/code-patching.c +4 -10
  }
 
  static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
  {
-	int err;
+	int err, rc = 0;
  	u32 *patch_addr = NULL;
  	unsigned long flags;
-	unsigned long text_poke_addr;
-	unsigned long kaddr = (unsigned long)addr;
 
  	/*
  	 * During early early boot patch_instruction is called
@@ -222,24 +229,20 @@ static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
 
  	local_irq_save(flags);
 
-	text_poke_addr = (unsigned long)__this_cpu_read(text_poke_area)->addr;
-	if (map_patch_area(addr, text_poke_addr)) {
-		err = -1;
+	err = map_patch_area(addr);
+	if (err)
  		goto out;
-	}
-
-	patch_addr = (u32 *)(text_poke_addr + (kaddr & ~PAGE_MASK));
 
-	__patch_instruction(addr, instr, patch_addr);
+	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));

arch/powerpc/lib/code-patching.c static int do_patch_instruction(u32 *addr, struct ppc_inst instr) · line 243

Quoted 10 lines
arch/powerpc/lib/code-patching.c +3 -3
+	rc = __patch_instruction(addr, instr, patch_addr);
 
-	err = unmap_patch_area(text_poke_addr);
-	if (err)
-		pr_warn("failed to unmap %lx\n", text_poke_addr);
+	err = unmap_patch_area();
 
  out:
  	local_irq_restore(flags);
+	WARN_ON(!ppc_inst_equal(ppc_inst_read(addr), instr));

Why adding that WARN_ON(), what could make that happen that is worth a WARN_ON() ?

Patching is quite a critical fast path, I'm not sure we want to afford too many checks during patching, we want it quick at first.

Quoted 7 lines
arch/powerpc/lib/code-patching.c +1 -1
-	return err;
+	return rc ? rc : err;
  }
  #else /* !CONFIG_STRICT_KERNEL_RWX */
 
 
#15 lore raw

Re: [PATCH v5 6/8] powerpc: Rework and improve STRICT_KERNEL_RWX patching

Quoted 109 lines · Christophe Leroy
Quoted 106 lines · Christopher M. Riedl

Rework code-patching with STRICT_KERNEL_RWX to prepare for the next patch which uses a temporary mm for patching under the Book3s64 Radix MMU. Make improvements by adding a WARN_ON when the patchsite doesn't match after patching and return the error from __patch_instruction() properly.

---

v5: * New to series.

---
  arch/powerpc/lib/code-patching.c | 51 +++++++++++++++++---------------
  1 file changed, 27 insertions(+), 24 deletions(-)
arch/powerpc/lib/code-patching.c +15 -10
index 3122d8e4cc013..9f2eba9b70ee4 100644
@@ -102,11 +102,12 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
  }
 
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
+static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
 
  #if IS_BUILTIN(CONFIG_LKDTM)
  unsigned long read_cpu_patching_addr(unsigned int cpu)
  {
-	return (unsigned long)(per_cpu(text_poke_area, cpu))->addr;
+	return per_cpu(cpu_patching_addr, cpu);
  }
  #endif
 
@@ -121,6 +122,7 @@ static int text_area_cpu_up(unsigned int cpu)
  		return -1;
  	}
  	this_cpu_write(text_poke_area, area);
+	this_cpu_write(cpu_patching_addr, (unsigned long)area->addr);
 
  	return 0;
  }
@@ -146,7 +148,7 @@ void __init poking_init(void)
  /*
   * This can be called for kernel text or a module.
   */
-static int map_patch_area(void *addr, unsigned long text_poke_addr)
+static int map_patch_area(void *addr)
  {
  	unsigned long pfn;
  	int err;
@@ -156,17 +158,20 @@ static int map_patch_area(void *addr, unsigned long text_poke_addr)
  	else
  		pfn = __pa_symbol(addr) >> PAGE_SHIFT;
 
-	err = map_kernel_page(text_poke_addr, (pfn << PAGE_SHIFT), PAGE_KERNEL);
+	err = map_kernel_page(__this_cpu_read(cpu_patching_addr),
+			      (pfn << PAGE_SHIFT), PAGE_KERNEL);
 
-	pr_devel("Mapped addr %lx with pfn %lx:%d\n", text_poke_addr, pfn, err);
+	pr_devel("Mapped addr %lx with pfn %lx:%d\n",
+		 __this_cpu_read(cpu_patching_addr), pfn, err);
  	if (err)
  		return -1;
 
  	return 0;
  }
 
-static inline int unmap_patch_area(unsigned long addr)
+static inline int unmap_patch_area(void)
  {
+	unsigned long addr = __this_cpu_read(cpu_patching_addr);
  	pte_t *ptep;
  	pmd_t *pmdp;
  	pud_t *pudp;
@@ -175,23 +180,23 @@ static inline int unmap_patch_area(unsigned long addr)
 
  	pgdp = pgd_offset_k(addr);
  	if (unlikely(!pgdp))
-		return -EINVAL;
+		goto out_err;
 
  	p4dp = p4d_offset(pgdp, addr);
  	if (unlikely(!p4dp))
-		return -EINVAL;
+		goto out_err;
 
  	pudp = pud_offset(p4dp, addr);
  	if (unlikely(!pudp))
-		return -EINVAL;
+		goto out_err;
 
  	pmdp = pmd_offset(pudp, addr);
  	if (unlikely(!pmdp))
-		return -EINVAL;
+		goto out_err;
 
  	ptep = pte_offset_kernel(pmdp, addr);
  	if (unlikely(!ptep))
-		return -EINVAL;
+		goto out_err;
 
  	pr_devel("clearing mm %p, pte %p, addr %lx\n", &init_mm, ptep, addr);
 
@@ -202,15 +207,17 @@ static inline int unmap_patch_area(unsigned long addr)
  	flush_tlb_kernel_range(addr, addr + PAGE_SIZE);
 
Quoted 9 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(unsigned long addr) · line 213

Quoted 5 lines
arch/powerpc/lib/code-patching.c +4 -0
  	return 0;
+
+out_err:
+	pr_warn("failed to unmap %lx\n", addr);
+	return -EINVAL;

Can you keep that in the caller of unmap_patch_area() instead of all those goto stuff ?

Yeah I think that's fair. I'll do this in the next spin.

Quoted 36 lines
Quoted 36 lines
arch/powerpc/lib/code-patching.c +6 -13
  }
 
  static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
  {
-	int err;
+	int err, rc = 0;
  	u32 *patch_addr = NULL;
  	unsigned long flags;
-	unsigned long text_poke_addr;
-	unsigned long kaddr = (unsigned long)addr;
 
  	/*
  	 * During early early boot patch_instruction is called
@@ -222,24 +229,20 @@ static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
 
  	local_irq_save(flags);
 
-	text_poke_addr = (unsigned long)__this_cpu_read(text_poke_area)->addr;
-	if (map_patch_area(addr, text_poke_addr)) {
-		err = -1;
+	err = map_patch_area(addr);
+	if (err)
  		goto out;
-	}
-
-	patch_addr = (u32 *)(text_poke_addr + (kaddr & ~PAGE_MASK));
 
-	__patch_instruction(addr, instr, patch_addr);
+	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
+	rc = __patch_instruction(addr, instr, patch_addr);
 
-	err = unmap_patch_area(text_poke_addr);
-	if (err)
-		pr_warn("failed to unmap %lx\n", text_poke_addr);
+	err = unmap_patch_area();
 
Quoted 6 lines

arch/powerpc/lib/code-patching.c static int do_patch_instruction(u32 *addr, struct ppc_inst instr) · line 243

Quoted 3 lines
arch/powerpc/lib/code-patching.c +1 -0
  out:
  	local_irq_restore(flags);
+	WARN_ON(!ppc_inst_equal(ppc_inst_read(addr), instr));

Why adding that WARN_ON(), what could make that happen that is worth a WARN_ON() ?

Failing to patch something could cause very strange issues later, so explicitly calling out a failure when it happens is warranted IMO.

Quoted 4 lines

Patching is quite a critical fast path, I'm not sure we want to afford too many checks during
patching, we want it quick at first.

Hmm, I'd prefer to measure the impact first - if it's a huge degradation then sure we can drop the WARN_ON()... I'll add some data with the next spin.

Quoted 8 lines
Quoted 8 lines
arch/powerpc/lib/code-patching.c +1 -1
-	return err;
+	return rc ? rc : err;
  }
  #else /* !CONFIG_STRICT_KERNEL_RWX */
 
 
#16 lore raw

PATCH v5 7/8 powerpc/64s: Initialize and use a temporary mm for patching on Radix

In reply to Christopher M. Riedl

When code patching a STRICT_KERNEL_RWX kernel the page containing the address to be patched is temporarily mapped as writeable. Currently, a per-cpu vmalloc patch area is used for this purpose. While the patch area is per-cpu, the temporary page mapping is inserted into the kernel page tables for the duration of patching. The mapping is exposed to CPUs other than the patching CPU - this is undesirable from a hardening perspective. Use a temporary mm instead which keeps the mapping local to the CPU doing the patching.

Use the poking_init init hook to prepare a temporary mm and patching address. Initialize the temporary mm by copying the init mm. Choose a randomized patching address inside the temporary mm userspace address space. The patching address is randomized between PAGE_SIZE and DEFAULT_MAP_WINDOW-PAGE_SIZE.

Bits of entropy with 64K page size on BOOK3S_64:

bits of entropy = log2(DEFAULT_MAP_WINDOW_USER64 / PAGE_SIZE)

PAGE_SIZE=64K, DEFAULT_MAP_WINDOW_USER64=128TB bits of entropy = log2(128TB / 64K) bits of entropy = 31

The upper limit is DEFAULT_MAP_WINDOW due to how the Book3s64 Hash MMU operates - by default the space above DEFAULT_MAP_WINDOW is not available. Currently the Hash MMU does not use a temporary mm so technically this upper limit isn't necessary; however, a larger randomization range does not further "harden" this overall approach and future work may introduce patching with a temporary mm on Hash as well.

Randomization occurs only once during initialization at boot for each possible CPU in the system.

Introduce two new functions, map_patch() and unmap_patch(), to respectively create and remove the temporary mapping with write permissions at patching_addr. Map the page with PAGE_KERNEL to set EAA[0] for the PTE which ignores the AMR (so no need to unlock/lock KUAP) according to PowerISA v3.0b Figure 35 on Radix.

Based on x86 implementation:

commit 4fc19708b165
("x86/alternatives: Initialize temporary mm for patching")

and:

commit b3fd8e83ada0
("x86/alternatives: Use temporary mm for text poking")

---

v5: * Only support Book3s64 Radix MMU for now.
* Use a per-cpu datastructure to hold the patching_addr and patching_mm to avoid the need for a synchronization lock/mutex.

v4: * In the previous series this was two separate patches: one to init
the temporary mm in poking_init() (unused in powerpc at the time) and the other to use it for patching (which removed all the per-cpu vmalloc code). Now that we use poking_init() in the existing per-cpu vmalloc approach, that separation doesn't work as nicely anymore so I just merged the two patches into one.
* Preload the SLB entry and hash the page for the patching_addr when using Hash on book3s64 to avoid taking an SLB and Hash fault during patching. The previous implementation was a hack which changed current->mm to allow the SLB and Hash fault handlers to work with the temporary mm since both of those code-paths always assume mm == current->mm.
* Also (hmm - seeing a trend here) with the book3s64 Hash MMU we have to manage the mm->context.active_cpus counter and mm cpumask since they determine (via mm_is_thread_local()) if the TLB flush in pte_clear() is local or not - it should always be local when we're using the temporary mm. On book3s64's Radix MMU we can just call local_flush_tlb_mm().
* Use HPTE_USE_KERNEL_KEY on Hash to avoid costly lock/unlock of KUAP.

---
 arch/powerpc/lib/code-patching.c | 132 +++++++++++++++++++++++++++++--
 1 file changed, 125 insertions(+), 7 deletions(-)
arch/powerpc/lib/code-patching.c +125 -7
index 9f2eba9b70ee4..027dabd42b8dd 100644
@@ -11,6 +11,7 @@
 #include <linux/cpuhotplug.h>
 #include <linux/slab.h>
 #include <linux/uaccess.h>
+#include <linux/random.h>
 
 #include <asm/tlbflush.h>
 #include <asm/page.h>
@@ -103,6 +104,7 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
 
 static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
 static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
+static DEFINE_PER_CPU(struct mm_struct *, cpu_patching_mm);
 
 #if IS_BUILTIN(CONFIG_LKDTM)
 unsigned long read_cpu_patching_addr(unsigned int cpu)
@@ -133,6 +135,51 @@ static int text_area_cpu_down(unsigned int cpu)
 	return 0;
 }
 
+static __always_inline void __poking_init_temp_mm(void)
+{
+	int cpu;
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct mm_struct *patching_mm;
+	unsigned long patching_addr;
+
+	for_each_possible_cpu(cpu) {
+		/*
+		 * Some parts of the kernel (static keys for example) depend on
+		 * successful code patching. Code patching under
+		 * STRICT_KERNEL_RWX requires this setup - otherwise we cannot
+		 * patch at all. We use BUG_ON() here and later since an early
+		 * failure is preferred to buggy behavior and/or strange
+		 * crashes later.
+		 */
+		patching_mm = copy_init_mm();
+		BUG_ON(!patching_mm);
+
+		per_cpu(cpu_patching_mm, cpu) = patching_mm;
+
+		/*
+		 * Choose a randomized, page-aligned address from the range:
+		 * [PAGE_SIZE, DEFAULT_MAP_WINDOW - PAGE_SIZE] The lower
+		 * address bound is PAGE_SIZE to avoid the zero-page.  The
+		 * upper address bound is DEFAULT_MAP_WINDOW - PAGE_SIZE to
+		 * stay under DEFAULT_MAP_WINDOW with the Book3s64 Hash MMU.
+		 */
+		patching_addr = PAGE_SIZE + ((get_random_long() & PAGE_MASK)
+				% (DEFAULT_MAP_WINDOW - 2 * PAGE_SIZE));
+
+		per_cpu(cpu_patching_addr, cpu) = patching_addr;
+
+		/*
+		 * PTE allocation uses GFP_KERNEL which means we need to
+		 * pre-allocate the PTE here because we cannot do the
+		 * allocation during patching when IRQs are disabled.
+		 */
+		ptep = get_locked_pte(patching_mm, patching_addr, &ptl);
+		BUG_ON(!ptep);
+		pte_unmap_unlock(ptep, ptl);
+	}
+}
+
 /*
  * Although BUG_ON() is rude, in this case it should only happen if ENOMEM, and
  * we judge it as being preferable to a kernel that will crash later when
@@ -140,6 +187,11 @@ static int text_area_cpu_down(unsigned int cpu)
  */
 void __init poking_init(void)
 {
+	if (radix_enabled()) {
+		__poking_init_temp_mm();
+		return;
+	}
+
 	BUG_ON(!cpuhp_setup_state(CPUHP_AP_ONLINE_DYN,
 		"powerpc/text_poke:online", text_area_cpu_up,
 		text_area_cpu_down));
@@ -213,30 +265,96 @@ static inline int unmap_patch_area(void)
 	return -EINVAL;
 }
 
+struct patch_mapping {
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct temp_mm temp_mm;
+};
+
+/*
+ * This can be called for kernel text or a module.
+ */
+static int map_patch(const void *addr, struct patch_mapping *patch_mapping)
+{
+	struct page *page;
+	pte_t pte;
+	pgprot_t pgprot;
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	if (is_vmalloc_or_module_addr(addr))
+		page = vmalloc_to_page(addr);
+	else
+		page = virt_to_page(addr);
+
+	patch_mapping->ptep = get_locked_pte(patching_mm, patching_addr,
+					     &patch_mapping->ptl);
+	if (unlikely(!patch_mapping->ptep)) {
+		pr_warn("map patch: failed to allocate pte for patching\n");
+		return -1;
+	}
+
+	pgprot = PAGE_KERNEL;
+	pte = mk_pte(page, pgprot);
+	pte = pte_mkdirty(pte);
+	set_pte_at(patching_mm, patching_addr, patch_mapping->ptep, pte);
+
+	init_temp_mm(&patch_mapping->temp_mm, patching_mm);
+	use_temporary_mm(&patch_mapping->temp_mm);
+
+	return 0;
+}
+
+static void unmap_patch(struct patch_mapping *patch_mapping)
+{
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	pte_clear(patching_mm, patching_addr, patch_mapping->ptep);
+
+	local_flush_tlb_mm(patching_mm);
+
+	pte_unmap_unlock(patch_mapping->ptep, patch_mapping->ptl);
+
+	unuse_temporary_mm(&patch_mapping->temp_mm);
+}
+
 static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
 {
 	int err, rc = 0;
 	u32 *patch_addr = NULL;
 	unsigned long flags;
+	struct patch_mapping patch_mapping;
 
 	/*
-	 * During early early boot patch_instruction is called
-	 * when text_poke_area is not ready, but we still need
-	 * to allow patching. We just do the plain old patching
+	 * During early early boot patch_instruction is called when the
+	 * patching_mm/text_poke_area is not ready, but we still need to allow
+	 * patching. We just do the plain old patching.
 	 */
-	if (!this_cpu_read(text_poke_area))
-		return raw_patch_instruction(addr, instr);
+	if (radix_enabled()) {
+		if (!this_cpu_read(cpu_patching_mm))
+			return raw_patch_instruction(addr, instr);
+	} else {
+		if (!this_cpu_read(text_poke_area))
+			return raw_patch_instruction(addr, instr);
+	}
 
 	local_irq_save(flags);
 
-	err = map_patch_area(addr);
+	if (radix_enabled())
+		err = map_patch(addr, &patch_mapping);
+	else
+		err = map_patch_area(addr);
 	if (err)
 		goto out;
 
 	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
 	rc = __patch_instruction(addr, instr, patch_addr);
 
-	err = unmap_patch_area();
+	if (radix_enabled())
+		unmap_patch(&patch_mapping);
+	else
+		err = unmap_patch_area();
 
 out:
 	local_irq_restore(flags);
Signature · 1 line
2.26.1
#17 lore raw

Re: [PATCH v5 7/8] powerpc/64s: Initialize and use a temporary mm for patching on Radix

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 114 lines

When code patching a STRICT_KERNEL_RWX kernel the page containing the address to be patched is temporarily mapped as writeable. Currently, a per-cpu vmalloc patch area is used for this purpose. While the patch area is per-cpu, the temporary page mapping is inserted into the kernel page tables for the duration of patching. The mapping is exposed to CPUs other than the patching CPU - this is undesirable from a hardening perspective. Use a temporary mm instead which keeps the mapping local to the CPU doing the patching.

Use the poking_init init hook to prepare a temporary mm and patching address. Initialize the temporary mm by copying the init mm. Choose a randomized patching address inside the temporary mm userspace address space. The patching address is randomized between PAGE_SIZE and DEFAULT_MAP_WINDOW-PAGE_SIZE.

Bits of entropy with 64K page size on BOOK3S_64:

bits of entropy = log2(DEFAULT_MAP_WINDOW_USER64 / PAGE_SIZE)

PAGE_SIZE=64K, DEFAULT_MAP_WINDOW_USER64=128TB bits of entropy = log2(128TB / 64K)
bits of entropy = 31

The upper limit is DEFAULT_MAP_WINDOW due to how the Book3s64 Hash MMU operates - by default the space above DEFAULT_MAP_WINDOW is not available. Currently the Hash MMU does not use a temporary mm so technically this upper limit isn't necessary; however, a larger randomization range does not further "harden" this overall approach and future work may introduce patching with a temporary mm on Hash as well.

Randomization occurs only once during initialization at boot for each possible CPU in the system.

Introduce two new functions, map_patch() and unmap_patch(), to respectively create and remove the temporary mapping with write permissions at patching_addr. Map the page with PAGE_KERNEL to set EAA[0] for the PTE which ignores the AMR (so no need to unlock/lock KUAP) according to PowerISA v3.0b Figure 35 on Radix.

Based on x86 implementation:

commit 4fc19708b165
("x86/alternatives: Initialize temporary mm for patching")

and:

commit b3fd8e83ada0
("x86/alternatives: Use temporary mm for text poking")

---

v5: * Only support Book3s64 Radix MMU for now.
* Use a per-cpu datastructure to hold the patching_addr and patching_mm to avoid the need for a synchronization lock/mutex.

v4: * In the previous series this was two separate patches: one to init
the temporary mm in poking_init() (unused in powerpc at the time) and the other to use it for patching (which removed all the per-cpu vmalloc code). Now that we use poking_init() in the existing per-cpu vmalloc approach, that separation doesn't work as nicely anymore so I just merged the two patches into one.
* Preload the SLB entry and hash the page for the patching_addr when using Hash on book3s64 to avoid taking an SLB and Hash fault during patching. The previous implementation was a hack which changed current->mm to allow the SLB and Hash fault handlers to work with the temporary mm since both of those code-paths always assume mm == current->mm.
* Also (hmm - seeing a trend here) with the book3s64 Hash MMU we have to manage the mm->context.active_cpus counter and mm cpumask since they determine (via mm_is_thread_local()) if the TLB flush in pte_clear() is local or not - it should always be local when we're using the temporary mm. On book3s64's Radix MMU we can just call local_flush_tlb_mm().
* Use HPTE_USE_KERNEL_KEY on Hash to avoid costly lock/unlock of KUAP.

---
  arch/powerpc/lib/code-patching.c | 132 +++++++++++++++++++++++++++++--
  1 file changed, 125 insertions(+), 7 deletions(-)
arch/powerpc/lib/code-patching.c +11 -0
index 9f2eba9b70ee4..027dabd42b8dd 100644
@@ -11,6 +11,7 @@
  #include <linux/cpuhotplug.h>
  #include <linux/slab.h>
  #include <linux/uaccess.h>
+#include <linux/random.h>
 
  #include <asm/tlbflush.h>
  #include <asm/page.h>
@@ -103,6 +104,7 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
 
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
  static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
+static DEFINE_PER_CPU(struct mm_struct *, cpu_patching_mm);
 
  #if IS_BUILTIN(CONFIG_LKDTM)
  unsigned long read_cpu_patching_addr(unsigned int cpu)
@@ -133,6 +135,51 @@ static int text_area_cpu_down(unsigned int cpu)
  	return 0;
  }
 
+static __always_inline void __poking_init_temp_mm(void)
+{
+	int cpu;
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct mm_struct *patching_mm;
+	unsigned long patching_addr;
+
+	for_each_possible_cpu(cpu) {

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 156

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+		/*
+		 * Some parts of the kernel (static keys for example) depend on
+		 * successful code patching. Code patching under
+		 * STRICT_KERNEL_RWX requires this setup - otherwise we cannot
+		 * patch at all. We use BUG_ON() here and later since an early
+		 * failure is preferred to buggy behavior and/or strange
+		 * crashes later.
+		 */
+		patching_mm = copy_init_mm();
+		BUG_ON(!patching_mm);

Read https://www.kernel.org/doc/html/latest/process/deprecated.html#bug-and-bug-on and https://github.com/linuxppc/issues/issues/88

Avoid BUG_ON()s thanks.

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 168

Quoted 12 lines
arch/powerpc/lib/code-patching.c +12 -0
+
+		per_cpu(cpu_patching_mm, cpu) = patching_mm;
+
+		/*
+		 * Choose a randomized, page-aligned address from the range:
+		 * [PAGE_SIZE, DEFAULT_MAP_WINDOW - PAGE_SIZE] The lower
+		 * address bound is PAGE_SIZE to avoid the zero-page.  The
+		 * upper address bound is DEFAULT_MAP_WINDOW - PAGE_SIZE to
+		 * stay under DEFAULT_MAP_WINDOW with the Book3s64 Hash MMU.
+		 */
+		patching_addr = PAGE_SIZE + ((get_random_long() & PAGE_MASK)
+				% (DEFAULT_MAP_WINDOW - 2 * PAGE_SIZE));

% should be at the end of first line and the second line alignment should match open parenthesis in first line.

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 178

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+
+		per_cpu(cpu_patching_addr, cpu) = patching_addr;
+
+		/*
+		 * PTE allocation uses GFP_KERNEL which means we need to
+		 * pre-allocate the PTE here because we cannot do the
+		 * allocation during patching when IRQs are disabled.
+		 */
+		ptep = get_locked_pte(patching_mm, patching_addr, &ptl);
+		BUG_ON(!ptep);

Avoid BUG_ON() please

Quoted 37 lines
arch/powerpc/lib/code-patching.c +23 -0
+		pte_unmap_unlock(ptep, ptl);
+	}
+}
+
  /*
   * Although BUG_ON() is rude, in this case it should only happen if ENOMEM, and
   * we judge it as being preferable to a kernel that will crash later when
@@ -140,6 +187,11 @@ static int text_area_cpu_down(unsigned int cpu)
   */
  void __init poking_init(void)
  {
+	if (radix_enabled()) {
+		__poking_init_temp_mm();
+		return;
+	}
+
  	BUG_ON(!cpuhp_setup_state(CPUHP_AP_ONLINE_DYN,
  		"powerpc/text_poke:online", text_area_cpu_up,
  		text_area_cpu_down));
@@ -213,30 +265,96 @@ static inline int unmap_patch_area(void)
  	return -EINVAL;
  }
 
+struct patch_mapping {
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct temp_mm temp_mm;
+};
+
+/*
+ * This can be called for kernel text or a module.
+ */
+static int map_patch(const void *addr, struct patch_mapping *patch_mapping)
+{
+	struct page *page;
+	pte_t pte;
+	pgprot_t pgprot;

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 291

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	if (is_vmalloc_or_module_addr(addr))
+		page = vmalloc_to_page(addr);
+	else
+		page = virt_to_page(addr);
+
+	patch_mapping->ptep = get_locked_pte(patching_mm, patching_addr,
+					     &patch_mapping->ptl);

Not sure you need to split this line, checkpatch now allows 100 chars per line.

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 293

Quoted 2 lines
arch/powerpc/lib/code-patching.c +2 -0
+	if (unlikely(!patch_mapping->ptep)) {
+		pr_warn("map patch: failed to allocate pte for patching\n");

That's a lot better than all above BUG_ONs

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 299

Quoted 6 lines
arch/powerpc/lib/code-patching.c +6 -0
+		return -1;
+	}
+
+	pgprot = PAGE_KERNEL;
+	pte = mk_pte(page, pgprot);
+	pte = pte_mkdirty(pte);

I'm sure you can do

	pte = pte_mkdirty(mk_pte(page, PAGE_KERNEL));

And indeed PAGE_KERNEL already includes _PAGE_DIRTY, so all you should need is

	pte = mk_pte(page, PAGE_KERNEL);

Or even just

	set_pte_at(patching_mm, patching_addr, patch_mapping->ptep, mk_pte(page, PAGE_KERNEL));
Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+	set_pte_at(patching_mm, patching_addr, patch_mapping->ptep, pte);
+
+	init_temp_mm(&patch_mapping->temp_mm, patching_mm);
+	use_temporary_mm(&patch_mapping->temp_mm);
+
+	return 0;
+}
+
+static void unmap_patch(struct patch_mapping *patch_mapping)
+{

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 319

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	pte_clear(patching_mm, patching_addr, patch_mapping->ptep);
+
+	local_flush_tlb_mm(patching_mm);
+
+	pte_unmap_unlock(patch_mapping->ptep, patch_mapping->ptl);
+
+	unuse_temporary_mm(&patch_mapping->temp_mm);

Shouldn't you stop using it before unmapping/unlocking it ?

Quoted 22 lines
arch/powerpc/lib/code-patching.c +9 -5
+}
+
  static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
  {
  	int err, rc = 0;
  	u32 *patch_addr = NULL;
  	unsigned long flags;
+	struct patch_mapping patch_mapping;
 
  	/*
-	 * During early early boot patch_instruction is called
-	 * when text_poke_area is not ready, but we still need
-	 * to allow patching. We just do the plain old patching
+	 * During early early boot patch_instruction is called when the
+	 * patching_mm/text_poke_area is not ready, but we still need to allow
+	 * patching. We just do the plain old patching.
  	 */
-	if (!this_cpu_read(text_poke_area))
-		return raw_patch_instruction(addr, instr);
+	if (radix_enabled()) {
+		if (!this_cpu_read(cpu_patching_mm))
+			return raw_patch_instruction(addr, instr);

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 345

Quoted 10 lines
arch/powerpc/lib/code-patching.c +6 -1
+	} else {
+		if (!this_cpu_read(text_poke_area))
+			return raw_patch_instruction(addr, instr);
+	}
 
  	local_irq_save(flags);
 
-	err = map_patch_area(addr);
+	if (radix_enabled())
+		err = map_patch(addr, &patch_mapping);

Maybe call it map_patch_mm() or map_patch_mapping() ?

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 355

Quoted 11 lines
arch/powerpc/lib/code-patching.c +4 -1
+	else
+		err = map_patch_area(addr);
  	if (err)
  		goto out;
 
  	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
  	rc = __patch_instruction(addr, instr, patch_addr);
 
-	err = unmap_patch_area();
+	if (radix_enabled())
+		unmap_patch(&patch_mapping);

No err ? Would be better if it could return something, allthough always 0.

And same comment about naming.

Quoted 6 lines
arch/powerpc/lib/code-patching.c +2 -0
+	else
+		err = unmap_patch_area();
 
  out:
  	local_irq_restore(flags);
#18 lore raw

Re: [PATCH v5 7/8] powerpc/64s: Initialize and use a temporary mm for patching on Radix

Quoted 133 lines · Christophe Leroy
Quoted 114 lines · Christopher M. Riedl

When code patching a STRICT_KERNEL_RWX kernel the page containing the address to be patched is temporarily mapped as writeable. Currently, a per-cpu vmalloc patch area is used for this purpose. While the patch area is per-cpu, the temporary page mapping is inserted into the kernel page tables for the duration of patching. The mapping is exposed to CPUs other than the patching CPU - this is undesirable from a hardening perspective. Use a temporary mm instead which keeps the mapping local to the CPU doing the patching.

Use the poking_init init hook to prepare a temporary mm and patching address. Initialize the temporary mm by copying the init mm. Choose a randomized patching address inside the temporary mm userspace address space. The patching address is randomized between PAGE_SIZE and DEFAULT_MAP_WINDOW-PAGE_SIZE.

Bits of entropy with 64K page size on BOOK3S_64:

bits of entropy = log2(DEFAULT_MAP_WINDOW_USER64 / PAGE_SIZE)

PAGE_SIZE=64K, DEFAULT_MAP_WINDOW_USER64=128TB bits of entropy = log2(128TB / 64K)
bits of entropy = 31

The upper limit is DEFAULT_MAP_WINDOW due to how the Book3s64 Hash MMU operates - by default the space above DEFAULT_MAP_WINDOW is not available. Currently the Hash MMU does not use a temporary mm so technically this upper limit isn't necessary; however, a larger randomization range does not further "harden" this overall approach and future work may introduce patching with a temporary mm on Hash as well.

Randomization occurs only once during initialization at boot for each possible CPU in the system.

Introduce two new functions, map_patch() and unmap_patch(), to respectively create and remove the temporary mapping with write permissions at patching_addr. Map the page with PAGE_KERNEL to set EAA[0] for the PTE which ignores the AMR (so no need to unlock/lock KUAP) according to PowerISA v3.0b Figure 35 on Radix.

Based on x86 implementation:

commit 4fc19708b165
("x86/alternatives: Initialize temporary mm for patching")

and:

commit b3fd8e83ada0
("x86/alternatives: Use temporary mm for text poking")

---

v5: * Only support Book3s64 Radix MMU for now.
* Use a per-cpu datastructure to hold the patching_addr and patching_mm to avoid the need for a synchronization lock/mutex.

v4: * In the previous series this was two separate patches: one to init
the temporary mm in poking_init() (unused in powerpc at the time) and the other to use it for patching (which removed all the per-cpu vmalloc code). Now that we use poking_init() in the existing per-cpu vmalloc approach, that separation doesn't work as nicely anymore so I just merged the two patches into one.
* Preload the SLB entry and hash the page for the patching_addr when using Hash on book3s64 to avoid taking an SLB and Hash fault during patching. The previous implementation was a hack which changed current->mm to allow the SLB and Hash fault handlers to work with the temporary mm since both of those code-paths always assume mm == current->mm.
* Also (hmm - seeing a trend here) with the book3s64 Hash MMU we have to manage the mm->context.active_cpus counter and mm cpumask since they determine (via mm_is_thread_local()) if the TLB flush in pte_clear() is local or not - it should always be local when we're using the temporary mm. On book3s64's Radix MMU we can just call local_flush_tlb_mm().
* Use HPTE_USE_KERNEL_KEY on Hash to avoid costly lock/unlock of KUAP.

---
  arch/powerpc/lib/code-patching.c | 132 +++++++++++++++++++++++++++++--
  1 file changed, 125 insertions(+), 7 deletions(-)
arch/powerpc/lib/code-patching.c +11 -0
index 9f2eba9b70ee4..027dabd42b8dd 100644
@@ -11,6 +11,7 @@
  #include <linux/cpuhotplug.h>
  #include <linux/slab.h>
  #include <linux/uaccess.h>
+#include <linux/random.h>
 
  #include <asm/tlbflush.h>
  #include <asm/page.h>
@@ -103,6 +104,7 @@ static inline void unuse_temporary_mm(struct temp_mm *temp_mm)
 
  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
  static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
+static DEFINE_PER_CPU(struct mm_struct *, cpu_patching_mm);
 
  #if IS_BUILTIN(CONFIG_LKDTM)
  unsigned long read_cpu_patching_addr(unsigned int cpu)
@@ -133,6 +135,51 @@ static int text_area_cpu_down(unsigned int cpu)
  	return 0;
  }
 
+static __always_inline void __poking_init_temp_mm(void)
+{
+	int cpu;
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct mm_struct *patching_mm;
+	unsigned long patching_addr;
+
+	for_each_possible_cpu(cpu) {

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 156

Quoted 10 lines
arch/powerpc/lib/code-patching.c +10 -0
+		/*
+		 * Some parts of the kernel (static keys for example) depend on
+		 * successful code patching. Code patching under
+		 * STRICT_KERNEL_RWX requires this setup - otherwise we cannot
+		 * patch at all. We use BUG_ON() here and later since an early
+		 * failure is preferred to buggy behavior and/or strange
+		 * crashes later.
+		 */
+		patching_mm = copy_init_mm();
+		BUG_ON(!patching_mm);

Read
https://www.kernel.org/doc/html/latest/process/deprecated.html#bug-and-bug-on
and
https://github.com/linuxppc/issues/issues/88

Quoted 2 lines

Avoid BUG_ON()s thanks.

Fine, @mpe's reply on the GH issue says the check is probably redundant:

"In general we don't need to BUG_ON(!ptr), the MMU will catch NULL pointer dereferences for us."

Quoted 8 lines
Quoted 8 lines
arch/powerpc/lib/code-patching.c +8 -0
+
+		per_cpu(cpu_patching_mm, cpu) = patching_mm;
+
+		/*
+		 * Choose a randomized, page-aligned address from the range:
+		 * [PAGE_SIZE, DEFAULT_MAP_WINDOW - PAGE_SIZE] The lower
+		 * address bound is PAGE_SIZE to avoid the zero-page.  The
+		 * upper address bound is DEFAULT_MAP_WINDOW - PAGE_SIZE to
Quoted 8 lines

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 168

Quoted 4 lines
arch/powerpc/lib/code-patching.c +4 -0
+		 * stay under DEFAULT_MAP_WINDOW with the Book3s64 Hash MMU.
+		 */
+		patching_addr = PAGE_SIZE + ((get_random_long() & PAGE_MASK)
+				% (DEFAULT_MAP_WINDOW - 2 * PAGE_SIZE));

% should be at the end of first line and the second line alignment should match open parenthesis in
first line.

Ok - thanks!

Quoted 14 lines

arch/powerpc/lib/code-patching.c static int text_area_cpu_down(unsigned int cpu) · line 178

Quoted 11 lines
arch/powerpc/lib/code-patching.c +10 -0
+
+		per_cpu(cpu_patching_addr, cpu) = patching_addr;
+
+		/*
+		 * PTE allocation uses GFP_KERNEL which means we need to
+		 * pre-allocate the PTE here because we cannot do the
+		 * allocation during patching when IRQs are disabled.
+		 */
+		ptep = get_locked_pte(patching_mm, patching_addr, &ptl);
+		BUG_ON(!ptep);

Avoid BUG_ON() please

Yup, I'll remove these in the next spin.

Quoted 43 lines
Quoted 43 lines
arch/powerpc/lib/code-patching.c +28 -0
+		pte_unmap_unlock(ptep, ptl);
+	}
+}
+
  /*
   * Although BUG_ON() is rude, in this case it should only happen if ENOMEM, and
   * we judge it as being preferable to a kernel that will crash later when
@@ -140,6 +187,11 @@ static int text_area_cpu_down(unsigned int cpu)
   */
  void __init poking_init(void)
  {
+	if (radix_enabled()) {
+		__poking_init_temp_mm();
+		return;
+	}
+
  	BUG_ON(!cpuhp_setup_state(CPUHP_AP_ONLINE_DYN,
  		"powerpc/text_poke:online", text_area_cpu_up,
  		text_area_cpu_down));
@@ -213,30 +265,96 @@ static inline int unmap_patch_area(void)
  	return -EINVAL;
  }
 
+struct patch_mapping {
+	spinlock_t *ptl; /* for protecting pte table */
+	pte_t *ptep;
+	struct temp_mm temp_mm;
+};
+
+/*
+ * This can be called for kernel text or a module.
+ */
+static int map_patch(const void *addr, struct patch_mapping *patch_mapping)
+{
+	struct page *page;
+	pte_t pte;
+	pgprot_t pgprot;
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	if (is_vmalloc_or_module_addr(addr))
+		page = vmalloc_to_page(addr);
Quoted 9 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 291

Quoted 5 lines
arch/powerpc/lib/code-patching.c +5 -0
+	else
+		page = virt_to_page(addr);
+
+	patch_mapping->ptep = get_locked_pte(patching_mm, patching_addr,
+					     &patch_mapping->ptl);

Not sure you need to split this line, checkpatch now allows 100 chars per line.

I prefer sticking to 80 columns unless readability *really* improves by going over that limit.

Quoted 21 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 293

Quoted 3 lines
arch/powerpc/lib/code-patching.c +2 -0
+	if (unlikely(!patch_mapping->ptep)) {
+		pr_warn("map patch: failed to allocate pte for patching\n");

That's a lot better than all above BUG_ONs

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 299

Quoted 6 lines
arch/powerpc/lib/code-patching.c +6 -0
+		return -1;
+	}
+
+	pgprot = PAGE_KERNEL;
+	pte = mk_pte(page, pgprot);
+	pte = pte_mkdirty(pte);

I'm sure you can do

pte = pte_mkdirty(mk_pte(page, PAGE_KERNEL));

And indeed PAGE_KERNEL already includes _PAGE_DIRTY, so all you should need is

Quoted 7 lines

pte = mk_pte(page, PAGE_KERNEL);

Or even just

set_pte_at(patching_mm, patching_addr, patch_mapping->ptep, mk_pte(page, PAGE_KERNEL));

Ok, I'll consolidate this in the next spin. Thanks!

Quoted 15 lines
Quoted 15 lines
arch/powerpc/lib/code-patching.c +14 -0
+	set_pte_at(patching_mm, patching_addr, patch_mapping->ptep, pte);
+
+	init_temp_mm(&patch_mapping->temp_mm, patching_mm);
+	use_temporary_mm(&patch_mapping->temp_mm);
+
+	return 0;
+}
+
+static void unmap_patch(struct patch_mapping *patch_mapping)
+{
+	struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
+	unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
+
+	pte_clear(patching_mm, patching_addr, patch_mapping->ptep);
Quoted 9 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 319

Quoted 6 lines
arch/powerpc/lib/code-patching.c +6 -0
+
+	local_flush_tlb_mm(patching_mm);
+
+	pte_unmap_unlock(patch_mapping->ptep, patch_mapping->ptl);
+
+	unuse_temporary_mm(&patch_mapping->temp_mm);

Shouldn't you stop using it before unmapping/unlocking it ?

Yes I think you're right - IIRC I had to do this for the Hash MMU (which we don't support w/ this verion of the series anymore anyways). I'll revisit this for the next spin.

Quoted 28 lines
Quoted 28 lines
arch/powerpc/lib/code-patching.c +13 -5
+}
+
  static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
  {
  	int err, rc = 0;
  	u32 *patch_addr = NULL;
  	unsigned long flags;
+	struct patch_mapping patch_mapping;
 
  	/*
-	 * During early early boot patch_instruction is called
-	 * when text_poke_area is not ready, but we still need
-	 * to allow patching. We just do the plain old patching
+	 * During early early boot patch_instruction is called when the
+	 * patching_mm/text_poke_area is not ready, but we still need to allow
+	 * patching. We just do the plain old patching.
  	 */
-	if (!this_cpu_read(text_poke_area))
-		return raw_patch_instruction(addr, instr);
+	if (radix_enabled()) {
+		if (!this_cpu_read(cpu_patching_mm))
+			return raw_patch_instruction(addr, instr);
+	} else {
+		if (!this_cpu_read(text_poke_area))
+			return raw_patch_instruction(addr, instr);
+	}
 
Quoted 7 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 345

Quoted 5 lines
arch/powerpc/lib/code-patching.c +2 -1
  	local_irq_save(flags);
 
-	err = map_patch_area(addr);
+	if (radix_enabled())
+		err = map_patch(addr, &patch_mapping);

Maybe call it map_patch_mm() or map_patch_mapping() ?

Yes that does sound better.

Quoted 9 lines
Quoted 9 lines
arch/powerpc/lib/code-patching.c +2 -0
+	else
+		err = map_patch_area(addr);
  	if (err)
  		goto out;
 
  	patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
  	rc = __patch_instruction(addr, instr, patch_addr);
 
Quoted 6 lines

arch/powerpc/lib/code-patching.c static inline int unmap_patch_area(void) · line 355

Quoted 3 lines
arch/powerpc/lib/code-patching.c +2 -1
-	err = unmap_patch_area();
+	if (radix_enabled())
+		unmap_patch(&patch_mapping);

No err ? Would be better if it could return something, allthough always 0.

Ok I'll do that.

Quoted 3 lines

And same comment about naming.

Yes I'll use your suggested names.

Quoted 6 lines
Quoted 6 lines
arch/powerpc/lib/code-patching.c +2 -0
+	else
+		err = unmap_patch_area();
 
  out:
  	local_irq_restore(flags);
#19 lore raw

PATCH v5 8/8 lkdtm/powerpc: Fix code patching hijack test

In reply to Christopher M. Riedl

Code patching on powerpc with a STRICT_KERNEL_RWX uses a userspace address in a temporary mm on Radix now. Use __put_user() to avoid write failures due to KUAP when attempting a "hijack" on the patching address. __put_user() also works with the non-userspace, vmalloc-based patching address on non-Radix MMUs.

---
 drivers/misc/lkdtm/perms.c | 9 ---------
 1 file changed, 9 deletions(-)
drivers/misc/lkdtm/perms.c +0 -9
index 41e87e5f9cc86..da6a34a0a49fb 100644
@@ -262,16 +262,7 @@ static inline u32 lkdtm_read_patch_site(void)
 /* Returns True if the write succeeds */
 static inline bool lkdtm_try_write(u32 data, u32 *addr)
 {
-#ifdef CONFIG_PPC
-	__put_kernel_nofault(addr, &data, u32, err);
-	return true;
-
-err:
-	return false;
-#endif
-#ifdef CONFIG_X86_64
 	return !__put_user(data, addr);
-#endif
 }
 
 static int lkdtm_patching_cpu(void *data)
Signature · 1 line
2.26.1
#20 lore raw

Re: [PATCH v5 8/8] lkdtm/powerpc: Fix code patching hijack test

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 5 lines

Code patching on powerpc with a STRICT_KERNEL_RWX uses a userspace address in a temporary mm on Radix now. Use __put_user() to avoid write failures due to KUAP when attempting a "hijack" on the patching address. __put_user() also works with the non-userspace, vmalloc-based patching address on non-Radix MMUs.

It is not really clean to use __put_user() on non user address, allthought it works by change.

I think it would be better to do something like

	if (is_kernel_addr(addr))
		copy_to_kernel_nofault(...);
	else
		copy_to_user_nofault(...);
Quoted 28 lines
---
  drivers/misc/lkdtm/perms.c | 9 ---------
  1 file changed, 9 deletions(-)
drivers/misc/lkdtm/perms.c +0 -9
index 41e87e5f9cc86..da6a34a0a49fb 100644
@@ -262,16 +262,7 @@ static inline u32 lkdtm_read_patch_site(void)
  /* Returns True if the write succeeds */
  static inline bool lkdtm_try_write(u32 data, u32 *addr)
  {
-#ifdef CONFIG_PPC
-	__put_kernel_nofault(addr, &data, u32, err);
-	return true;
-
-err:
-	return false;
-#endif
-#ifdef CONFIG_X86_64
  	return !__put_user(data, addr);
-#endif
  }
 
  static int lkdtm_patching_cpu(void *data)
#21 lore raw

Re: [PATCH v5 8/8] lkdtm/powerpc: Fix code patching hijack test

Quoted 12 lines · Christophe Leroy
Quoted 5 lines · Christopher M. Riedl

Code patching on powerpc with a STRICT_KERNEL_RWX uses a userspace address in a temporary mm on Radix now. Use __put_user() to avoid write failures due to KUAP when attempting a "hijack" on the patching address. __put_user() also works with the non-userspace, vmalloc-based patching address on non-Radix MMUs.

It is not really clean to use __put_user() on non user address, allthought it works by change.

Quoted 7 lines

I think it would be better to do something like

if (is_kernel_addr(addr))
copy_to_kernel_nofault(...);
else
copy_to_user_nofault(...);

Yes that looks much better. I'll pick this up and try it for the next spin. Thanks!

Quoted 30 lines
Quoted 30 lines
---
  drivers/misc/lkdtm/perms.c | 9 ---------
  1 file changed, 9 deletions(-)
drivers/misc/lkdtm/perms.c +0 -9
index 41e87e5f9cc86..da6a34a0a49fb 100644
@@ -262,16 +262,7 @@ static inline u32 lkdtm_read_patch_site(void)
  /* Returns True if the write succeeds */
  static inline bool lkdtm_try_write(u32 data, u32 *addr)
  {
-#ifdef CONFIG_PPC
-	__put_kernel_nofault(addr, &data, u32, err);
-	return true;
-
-err:
-	return false;
-#endif
-#ifdef CONFIG_X86_64
  	return !__put_user(data, addr);
-#endif
  }
 
  static int lkdtm_patching_cpu(void *data)
#22 lore raw
In reply to Christopher M. Riedl

Le 13/07/2021 à 07:31, Christopher M. Riedl a écrit :

Quoted 21 lines

When compiled with CONFIG_STRICT_KERNEL_RWX, the kernel must create temporary mappings when patching itself. These mappings temporarily override the strict RWX text protections to permit a write. Currently, powerpc allocates a per-CPU VM area for patching. Patching occurs as follows:

1. Map page in per-CPU VM area w/ PAGE_KERNEL protection
2. Patch text
3. Remove the temporary mapping

While the VM area is per-CPU, the mapping is actually inserted into the kernel page tables. Presumably, this could allow another CPU to access the normally write-protected text - either malicously or accidentally - via this same mapping if the address of the VM area is known. Ideally, the mapping should be kept local to the CPU doing the patching [0].

x86 introduced "temporary mm" structs which allow the creation of mappings local to a particular CPU [1]. This series intends to bring the notion of a temporary mm to powerpc's Book3s64 Radix MMU and harden it by using such a mapping for patching a kernel with strict RWX permissions.

Quoted 9 lines

The first four patches implement an LKDTM test "proof-of-concept" which exploits the potential vulnerability (ie. the temporary mapping during patching is exposed in the kernel page tables and accessible by other CPUs) using a simple brute-force approach. This test is implemented for both powerpc and x86_64. The test passes on powerpc Radix with this new series, fails on upstream powerpc, passes on upstream x86_64, and fails on an older (ancient) x86_64 tree without the x86_64 temporary mm patches. The remaining patches add support for and use a temporary mm for code patching on powerpc with the Radix MMU.

I think four first patches (together with last one) are quite independent from the heart of the series itself which is patches 5, 6, 7. Maybe you should split that series it two series ? After all those selftests are nice to have but are not absolutely necessary, that would help getting forward I think.

Quoted 17 lines

Tested boot, ftrace, and repeated LKDTM "hijack":
- QEMU+KVM (host: POWER9 Blackbird): Radix MMU w/ KUAP
- QEMU+KVM (host: POWER9 Blackbird): Hash MMU

Tested repeated LKDTM "hijack":
- QEMU+KVM (host: AMD desktop): x86_64 upstream
- QEMU+KVM (host: AMD desktop): x86_64 w/o percpu temp mm to verify the LKDTM "hijack" test fails

Tested boot and ftrace:
- QEMU+TCG: ppc44x (bamboo)
- QEMU+TCG: g5 (mac99)

I also tested with various extra config options enabled as suggested in section 12) in Documentation/process/submit-checklist.rst.

Quoted 7 lines

v5: * Only support Book3s64 Radix MMU for now. There are some issues with
the previous implementation on the Hash MMU as pointed out by Nick Piggin. Fixing these is not trivial so we only support the Radix MMU for now. I tried using realmode (no data translation) to patch with Hash to at least avoid exposing the patch mapping to other CPUs but this doesn't reliably work either since we cannot access vmalloc'd space in realmode.

So you now accept to have two different mode depending on the platform ?
As far as I remember I commented some time ago that non SMP didn't need that feature and you were reluctant to have two different implementations. What made you change your mind ? (just curious).

Quoted 79 lines

* Use percpu variables for the patching_mm and patching_addr. This avoids the need for synchronization mechanisms entirely. Thanks to Peter Zijlstra for pointing out text_mutex which unfortunately didn't work out without larger code changes in powerpc. Also thanks to Daniel Axtens for comments about using percpu variables for the *percpu* temp mm things off list.

v4: * It's time to revisit this series again since @jpn and @mpe fixed
our known STRICT_*_RWX bugs on powerpc/64s.
* Rebase on linuxppc/next:
commit ee1bc694fbaec ("powerpc/kvm: Fix build error when PPC_MEM_KEYS/PPC_PSERIES=n")
* Completely rework how map_patch() works on book3s64 Hash MMU
* Split the LKDTM x86_64 and powerpc bits into separate patches
* Annotate commit messages with changes from v3 instead of
listing them here completely out-of context...

v3: * Rebase on linuxppc/next: commit 9123e3a74ec7 ("Linux 5.9-rc1")
* Move temporary mm implementation into code-patching.c where it belongs
* Implement LKDTM hijacker test on x86_64 (on IBM time oof) Do
* not use address zero for the patching address in the temporary mm (thanks @dja for pointing this out!)
* Wrap the LKDTM test w/ CONFIG_SMP as suggested by Christophe Leroy
* Comments to clarify PTE pre-allocation and patching addr selection

v2: * Rebase on linuxppc/next:
commit 105fb38124a4 ("powerpc/8xx: Modify ptep_get()")
* Always dirty pte when mapping patch
* Use ppc_inst_len instead of sizeof on instructions
* Declare LKDTM patching addr accessor in header where it belongs

v1: * Rebase on linuxppc/next (4336b9337824)
* Save and restore second hw watchpoint
* Use new ppc_inst_* functions for patching check and in LKDTM test

rfc-v2: * Many fixes and improvements mostly based on extensive feedback
and testing by Christophe Leroy (thanks!).
* Make patching_mm and patching_addr static and move '__ro_after_init' to after the variable name (more common in other parts of the kernel)
* Use 'asm/debug.h' header instead of 'asm/hw_breakpoint.h' to fix PPC64e compile
* Add comment explaining why we use BUG_ON() during the init call to setup for patching later
* Move ptep into patch_mapping to avoid walking page tables a second time when unmapping the temporary mapping
* Use KUAP under non-radix, also manually dirty the PTE for patch mapping on non-BOOK3S_64 platforms
* Properly return any error from __patch_instruction
* Do not use 'memcmp' where a simple comparison is appropriate
* Simplify expression for patch address by removing pointer maths
* Add LKDTM test

[0]: https://github.com/linuxppc/issues/issues/224
[1]: https://lore.kernel.org/kernel-hardening/[email protected]/

Christopher M. Riedl (8):
powerpc: Add LKDTM accessor for patching addr lkdtm/powerpc: Add test to hijack a patch mapping
x86_64: Add LKDTM accessor for patching addr lkdtm/x86_64: Add test to hijack a patch mapping powerpc/64s: Introduce temporary mm for Radix MMU
powerpc: Rework and improve STRICT_KERNEL_RWX patching powerpc/64s: Initialize and use a temporary mm for patching on Radix lkdtm/powerpc: Fix code patching hijack test

  arch/powerpc/include/asm/code-patching.h |   4 +
  arch/powerpc/include/asm/debug.h         |   1 +
  arch/powerpc/kernel/process.c            |   5 +
  arch/powerpc/lib/code-patching.c         | 240 ++++++++++++++++++++---
  arch/x86/include/asm/text-patching.h     |   4 +
  arch/x86/kernel/alternative.c            |   7 +
  drivers/misc/lkdtm/core.c                |   1 +
  drivers/misc/lkdtm/lkdtm.h               |   1 +
  drivers/misc/lkdtm/perms.c               | 143 ++++++++++++++
  9 files changed, 378 insertions(+), 28 deletions(-)
#23 lore raw
Quoted 34 lines · Christophe Leroy
Quoted 21 lines · Christopher M. Riedl

When compiled with CONFIG_STRICT_KERNEL_RWX, the kernel must create temporary mappings when patching itself. These mappings temporarily override the strict RWX text protections to permit a write. Currently, powerpc allocates a per-CPU VM area for patching. Patching occurs as follows:

1. Map page in per-CPU VM area w/ PAGE_KERNEL protection
2. Patch text
3. Remove the temporary mapping

While the VM area is per-CPU, the mapping is actually inserted into the kernel page tables. Presumably, this could allow another CPU to access the normally write-protected text - either malicously or accidentally - via this same mapping if the address of the VM area is known. Ideally, the mapping should be kept local to the CPU doing the patching [0].

x86 introduced "temporary mm" structs which allow the creation of mappings local to a particular CPU [1]. This series intends to bring the notion of a temporary mm to powerpc's Book3s64 Radix MMU and harden it by using such a mapping for patching a kernel with strict RWX permissions.

Quoted 9 lines

The first four patches implement an LKDTM test "proof-of-concept" which exploits the potential vulnerability (ie. the temporary mapping during patching is exposed in the kernel page tables and accessible by other CPUs) using a simple brute-force approach. This test is implemented for both powerpc and x86_64. The test passes on powerpc Radix with this new series, fails on upstream powerpc, passes on upstream x86_64, and fails on an older (ancient) x86_64 tree without the x86_64 temporary mm patches. The remaining patches add support for and use a temporary mm for code patching on powerpc with the Radix MMU.

Quoted 8 lines

I think four first patches (together with last one) are quite independent from the heart of the
series itself which is patches 5, 6, 7. Maybe you should split that series it two series ? After all
those selftests are nice to have but are not absolutely necessary, that would help getting forward I
think.

Hmm you're probably right. The selftest at least proves there is a potential attack which I think is necessary for any hardening related series/patch. I'll split the series into separate powerpc temp mm and LKDTM series for the next spin.

Quoted 25 lines
Quoted 17 lines

Tested boot, ftrace, and repeated LKDTM "hijack":
- QEMU+KVM (host: POWER9 Blackbird): Radix MMU w/ KUAP
- QEMU+KVM (host: POWER9 Blackbird): Hash MMU

Tested repeated LKDTM "hijack":
- QEMU+KVM (host: AMD desktop): x86_64 upstream
- QEMU+KVM (host: AMD desktop): x86_64 w/o percpu temp mm to verify the LKDTM "hijack" test fails

Tested boot and ftrace:
- QEMU+TCG: ppc44x (bamboo)
- QEMU+TCG: g5 (mac99)

I also tested with various extra config options enabled as suggested in section 12) in Documentation/process/submit-checklist.rst.

Quoted 7 lines

v5: * Only support Book3s64 Radix MMU for now. There are some issues with
the previous implementation on the Hash MMU as pointed out by Nick Piggin. Fixing these is not trivial so we only support the Radix MMU for now. I tried using realmode (no data translation) to patch with Hash to at least avoid exposing the patch mapping to other CPUs but this doesn't reliably work either since we cannot access vmalloc'd space in realmode.

Quoted 1 line

So you now accept to have two different mode depending on the platform ?

By necessity yes.

Quoted 5 lines

As far as I remember I commented some time ago that non SMP didn't need that feature and you were
reluctant to have two different implementations. What made you change your mind ? (just curious).

The book3s64 hash mmu support is a pain ;) Supporting both the temp-mm and vmalloc implementations turned out to be relatively simple - I initially thought this would be messier. For now we will support both; however, in the future I'd still like to implement the percpu temp-mm support for the Hash MMU as well. I suppose we could re-evaluate then if we want/need both implementations (I know you're in favor of keeping the vmalloc-based approach for performance reasons on non-SMP).

Quoted 80 lines
Quoted 80 lines

* Use percpu variables for the patching_mm and patching_addr. This avoids the need for synchronization mechanisms entirely. Thanks to Peter Zijlstra for pointing out text_mutex which unfortunately didn't work out without larger code changes in powerpc. Also thanks to Daniel Axtens for comments about using percpu variables for the *percpu* temp mm things off list.

v4: * It's time to revisit this series again since @jpn and @mpe fixed
our known STRICT_*_RWX bugs on powerpc/64s.
* Rebase on linuxppc/next:
commit ee1bc694fbaec ("powerpc/kvm: Fix build error when PPC_MEM_KEYS/PPC_PSERIES=n")
* Completely rework how map_patch() works on book3s64 Hash MMU
* Split the LKDTM x86_64 and powerpc bits into separate patches
* Annotate commit messages with changes from v3 instead of
listing them here completely out-of context...

v3: * Rebase on linuxppc/next: commit 9123e3a74ec7 ("Linux 5.9-rc1")
* Move temporary mm implementation into code-patching.c where it belongs
* Implement LKDTM hijacker test on x86_64 (on IBM time oof) Do
* not use address zero for the patching address in the temporary mm (thanks @dja for pointing this out!)
* Wrap the LKDTM test w/ CONFIG_SMP as suggested by Christophe Leroy
* Comments to clarify PTE pre-allocation and patching addr selection

v2: * Rebase on linuxppc/next:
commit 105fb38124a4 ("powerpc/8xx: Modify ptep_get()")
* Always dirty pte when mapping patch
* Use ppc_inst_len instead of sizeof on instructions
* Declare LKDTM patching addr accessor in header where it belongs

v1: * Rebase on linuxppc/next (4336b9337824)
* Save and restore second hw watchpoint
* Use new ppc_inst_* functions for patching check and in LKDTM test

rfc-v2: * Many fixes and improvements mostly based on extensive feedback
and testing by Christophe Leroy (thanks!).
* Make patching_mm and patching_addr static and move '__ro_after_init' to after the variable name (more common in other parts of the kernel)
* Use 'asm/debug.h' header instead of 'asm/hw_breakpoint.h' to fix PPC64e compile
* Add comment explaining why we use BUG_ON() during the init call to setup for patching later
* Move ptep into patch_mapping to avoid walking page tables a second time when unmapping the temporary mapping
* Use KUAP under non-radix, also manually dirty the PTE for patch mapping on non-BOOK3S_64 platforms
* Properly return any error from __patch_instruction
* Do not use 'memcmp' where a simple comparison is appropriate
* Simplify expression for patch address by removing pointer maths
* Add LKDTM test

[0]: https://github.com/linuxppc/issues/issues/224
[1]: https://lore.kernel.org/kernel-hardening/[email protected]/

Christopher M. Riedl (8):
powerpc: Add LKDTM accessor for patching addr lkdtm/powerpc: Add test to hijack a patch mapping
x86_64: Add LKDTM accessor for patching addr lkdtm/x86_64: Add test to hijack a patch mapping powerpc/64s: Introduce temporary mm for Radix MMU
powerpc: Rework and improve STRICT_KERNEL_RWX patching powerpc/64s: Initialize and use a temporary mm for patching on Radix lkdtm/powerpc: Fix code patching hijack test

  arch/powerpc/include/asm/code-patching.h |   4 +
  arch/powerpc/include/asm/debug.h         |   1 +
  arch/powerpc/kernel/process.c            |   5 +
  arch/powerpc/lib/code-patching.c         | 240 ++++++++++++++++++++---
  arch/x86/include/asm/text-patching.h     |   4 +
  arch/x86/kernel/alternative.c            |   7 +
  drivers/misc/lkdtm/core.c                |   1 +
  drivers/misc/lkdtm/lkdtm.h               |   1 +
  drivers/misc/lkdtm/perms.c               | 143 ++++++++++++++
  9 files changed, 378 insertions(+), 28 deletions(-)