powerpc: Further Strict RWX support
Subsystems (3)
v12 cover powerpc: Further Strict RWX support
Adding more Strict RWX support on powerpc, in particular Strict Module RWX. Thanks for all of the feedback everyone.
It is now rebased on linux-next + powerpc/64s/radix: Enable huge vmalloc mappings
(https://lore.kernel.org/linuxppc-dev/[email protected]/)
For reference the previous revision is available here:
https://lore.kernel.org/linuxppc-dev/[email protected]/
The changes in v12 for each patch:
Christophe Leroy (2):
powerpc/mm: implement set_memory_attr()
powerpc/32: use set_memory_attr()
Jordan Niethe (3):
powerpc/lib/code-patching: Set up Strict RWX patching earlier powerpc/bpf: Remove bpf_jit_free()
powerpc/bpf: Write protect JIT code
Russell Currey (3):
powerpc/mm: Implement set_memory() routines
v12: - change_page_attr() back to taking an action value
- disallow operating on huge pages
powerpc/kprobes: Mark newly allocated probes as ROX
v12: - Switch from __vmalloc_node_range() to module_alloc()
powerpc: Set ARCH_HAS_STRICT_MODULE_RWX
Some patches were dropped from this revision:
powerpc: Always define MODULES_{VADDR,END}
- Can use what Nick did with "Enable huge vmalloc mappings"
arch/powerpc/Kconfig | 2 + arch/powerpc/include/asm/set_memory.h | 34 +++++++ arch/powerpc/kernel/kprobes.c | 17 ++++ arch/powerpc/kernel/module.c | 4 +- arch/powerpc/lib/code-patching.c | 12 +-- arch/powerpc/mm/Makefile | 2 +- arch/powerpc/mm/pageattr.c | 133 ++++++++++++++++++++++++++ arch/powerpc/mm/pgtable_32.c | 60 ++---------- arch/powerpc/net/bpf_jit_comp.c | 13 +-- 9 files changed, 205 insertions(+), 72 deletions(-) create mode 100644 arch/powerpc/include/asm/set_memory.h create mode 100644 arch/powerpc/mm/pageattr.c
Signature · 1 line
2.25.1
PATCH v12 1/8 powerpc/mm: Implement set_memory() routines
From: Russell Currey <[email protected]>
The set_memory_{ro/rw/nx/x}() functions are required for STRICT_MODULE_RWX, and are generally useful primitives to have. This implementation is designed to be generic across powerpc's many MMUs. It's possible that this could be optimised to be faster for specific MMUs.
This implementation does not handle cases where the caller is attempting to change the mapping of the page it is executing from, or if another CPU is concurrently using the page being altered. These cases likely shouldn't happen, but a more complex implementation with MMU-specific code could safely handle them.
On hash, the linear mapping is not kept in the linux pagetable, so this will not change the protection if used on that range. Currently these functions are not used on the linear map so just WARN for now.
apply_to_existing_page_range() does not work on huge pages so for now disallow changing the protection of huge pages.
[jpn: - Allow set memory functions to be used without Strict RWX
- Hash: Disallow certain regions
- Have change_page_attr() take function pointers to manipulate ptes
- Radix: Add ptesync after set_pte_at()]
---
v10: WARN if trying to change the hash linear map
v11: - Update copywrite dates
- Allow set memory functions to be used without Strict RWX
- Hash: Disallow certain regions and add comment explaining why
- Have change_page_attr() take function pointers to manipulate ptes
- Clarify change_page_attr()'s comment
- Radix: Add ptesync after set_pte_at()
v12: - change_page_attr() back to taking an action value
- disallow operating on huge pages
--- arch/powerpc/Kconfig | 1 + arch/powerpc/include/asm/set_memory.h | 32 +++++++++ arch/powerpc/mm/Makefile | 2 +- arch/powerpc/mm/pageattr.c | 100 ++++++++++++++++++++++++++ 4 files changed, 134 insertions(+), 1 deletion(-) create mode 100644 arch/powerpc/include/asm/set_memory.h create mode 100644 arch/powerpc/mm/pageattr.c
arch/powerpc/Kconfig +1 -0
@@ -138,6 +138,7 @@ config PPC select ARCH_HAS_MEMBARRIER_CALLBACKS select ARCH_HAS_MEMBARRIER_SYNC_CORE select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE && PPC_BOOK3S_64 + select ARCH_HAS_SET_MEMORY select ARCH_HAS_STRICT_KERNEL_RWX if ((PPC_BOOK3S_64 || PPC32) && !HIBERNATION) select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST select ARCH_HAS_UACCESS_FLUSHCACHE
arch/powerpc/include/asm/set_memory.h new file +32 -0
@@ -0,0 +1,32 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_POWERPC_SET_MEMORY_H +#define _ASM_POWERPC_SET_MEMORY_H + +#define SET_MEMORY_RO 0 +#define SET_MEMORY_RW 1 +#define SET_MEMORY_NX 2 +#define SET_MEMORY_X 3 + +int change_memory_attr (unsigned long addr,int numpages,long action); + +static inline int set_memory_ro (unsigned long addr,int numpages) +{ +return change_memory_attr (addr, numpages,SET_MEMORY_RO ); +} + +static inline int set_memory_rw (unsigned long addr,int numpages) +{ +return change_memory_attr (addr, numpages,SET_MEMORY_RW ); +} + +static inline int set_memory_nx (unsigned long addr,int numpages) +{ +return change_memory_attr (addr, numpages,SET_MEMORY_NX ); +} + +static inline int set_memory_x (unsigned long addr,int numpages) +{ +return change_memory_attr (addr, numpages,SET_MEMORY_X ); +} + +#endif
arch/powerpc/mm/Makefile +1 -1
@@ -5,7 +5,7 @@ ccflags-$(CONFIG_PPC64) := $(NO_MINIMAL_TOC) -obj-y := fault.o mem.o pgtable.o mmap.o maccess.o \ +obj-y := fault.o mem.o pgtable.o mmap.o maccess.o pageattr.o \ init_$(BITS).o pgtable_$(BITS).o \ pgtable-frag.o ioremap.o ioremap_$(BITS).o \ init-common.o mmu_context.o drmem.o \
arch/powerpc/mm/pageattr.c new file +100 -0
100 lines not on this page · raw message
Signature · 1 line
2.25.1
PATCH v12 2/8 powerpc/lib/code-patching: Set up Strict RWX patching earlier
In reply to Jordan Niethesetup_text_poke_area() is a late init call so it runs before mark_rodata_ro() and after the init calls. This lets all the init code patching simply write to their locations. In the future, kprobes is going to allocate its instruction pages RO which means they will need setup_text__poke_area() to have been already called for their code patching. However, init_kprobes() (which allocates and patches some instruction pages) is an early init call so it happens before setup_text__poke_area().
start_kernel() calls poking_init() before any of the init calls. On powerpc, poking_init() is currently a nop. setup_text_poke_area() relies on kernel virtual memory, cpu hotplug and per_cpu_areas being setup. setup_per_cpu_areas(), boot_cpu_hotplug_init() and mm_init() are called before poking_init().
Turn setup_text_poke_area() into poking_init().
---
v9: New to series
--- arch/powerpc/lib/code-patching.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-)
arch/powerpc/lib/code-patching.c +4 -8
@@ -70,14 +70,11 @@ static int text_area_cpu_down(unsigned int cpu) }/* -* Run as a late init call. This allows all the boot time patching to be done -* simply by patching the code, and then we're called here prior to -* mark_rodata_ro(), which happens after all init calls are run. 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 someone tries -* to use patch_instruction(). +* 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 +* someone tries to use patch_instruction(). */ -static int __init setup_text_poke_area(void) +int __init poking_init(void) { BUG_ON(!cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "powerpc/text_poke:online", text_area_cpu_up,
@@ -85,7 +82,6 @@ static int __init setup_text_poke_area(void)return 0 ; } -late_initcall (setup_text_poke_area);/* * This can be called for kernel text or a module.
Signature · 1 line
2.25.1
Re: [PATCH v12 2/8] powerpc/lib/code-patching: Set up Strict RWX patching earlier
Le 06/05/2021 à 04:34, Jordan Niethe a écrit :
Quoted 15 lines
setup_text_poke_area()is a late init call so it runs beforemark_rodata_ro()and after the init calls. This lets all the init code patching simply write to their locations. In the future, kprobes is going to allocate its instruction pages RO which means they will needsetup_text__poke_area()to have been already called for their code patching. However,init_kprobes()(which allocates and patches some instruction pages) is an early init call so it happens beforesetup_text__poke_area().
start_kernel()callspoking_init()before any of the init calls. On powerpc,poking_init()is currently a nop.setup_text_poke_area()relies on kernel virtual memory, cpu hotplug andper_cpu_areasbeing setup.setup_per_cpu_areas(),boot_cpu_hotplug_init()andmm_init()are called beforepoking_init().
Quoted 4 lines
Turn
setup_text_poke_area()intopoking_init().Reviewed-by Russell Currey <[email protected]>Signed-off-by Jordan Niethe <[email protected]>
Quoted 38 lines
---
v9: New to series--- arch/powerpc/lib/code-patching.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-)arch/powerpc/lib/code-patching.c +4
-8@@ -70,14 +70,11 @@ static int text_area_cpu_down(unsigned int cpu) }/* -* Run as a late init call. This allows all the boot time patching to be done -* simply by patching the code, and then we're called here prior to -* mark_rodata_ro(), which happens after all init calls are run. 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 someone tries -* to use patch_instruction(). +* 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 +* someone tries to use patch_instruction(). */ -static int __init setup_text_poke_area(void) +int __init poking_init(void) { BUG_ON(!cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "powerpc/text_poke:online", text_area_cpu_up,@@ -85,7 +82,6 @@ static int __init setup_text_poke_area(void)return 0 ; } -late_initcall (setup_text_poke_area);/* * This can be called for kernel text or a module.
PATCH v12 3/8 powerpc/kprobes: Mark newly allocated probes as ROX
In reply to Jordan NietheFrom: Russell Currey <[email protected]>
Add the arch specific insn page allocator for powerpc. This allocates ROX pages if STRICT_KERNEL_RWX is enabled. These pages are only written to with patch_instruction() which is able to write RO pages.
---
v9: - vmalloc_exec() no longer exists
- Set the page to RW before freeing it
v10: - use __vmalloc_node_range()
v11: - Neaten up
v12: - Switch from __vmalloc_node_range() to module_alloc()
--- arch/powerpc/kernel/kprobes.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
arch/powerpc/kernel/kprobes.c +17 -0
@@ -19,11 +19,13 @@#include <linux/extable.h> #include <linux/kdebug.h> #include <linux/slab.h> +#include <linux/moduleloader.h> #include <asm/code-patching.h> #include <asm/cacheflush.h> #include <asm/sstep.h> #include <asm/sections.h> #include <asm/inst.h> +#include <asm/set_memory.h> #include <linux/uaccess.h> DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
@@ -103,6 +105,21 @@ kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset)return addr; } +void *alloc_insn_page (void ) +{ +void *page; + + page =module_alloc (PAGE_SIZE ); +if (!page) +return NULL ; + +if (IS_ENABLED (CONFIG_STRICT_MODULE_RWX )) { +set_memory_ro ((unsigned long )page,1 ); +set_memory_x ((unsigned long )page,1 ); + } +return page; +} +int arch_prepare_kprobe (struct kprobe *p) {int ret =0 ;
Signature · 1 line
2.25.1
Re: [PATCH v12 3/8] powerpc/kprobes: Mark newly allocated probes as ROX
Le 06/05/2021 à 04:34, Jordan Niethe a écrit :
Quoted 42 lines
From: Russell Currey <[email protected]>
Add the arch specific insn page allocator for powerpc. This allocates ROX pages if
STRICT_KERNEL_RWXis enabled. These pages are only written to withpatch_instruction()which is able to write RO pages.Reviewed-by Daniel Axtens <[email protected]>Signed-off-by Russell Currey <[email protected]>Signed-off-by Christophe Leroy <[email protected]>[jpn: Reword commit message, switch to __vmalloc_node_range()]Signed-off-by Jordan Niethe <[email protected]>---
v9: -vmalloc_exec()no longer exists
- Set the page to RW before freeing it
v10: - use__vmalloc_node_range()
v11: - Neaten up
v12: - Switch from__vmalloc_node_range()tomodule_alloc()--- arch/powerpc/kernel/kprobes.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)arch/powerpc/kernel/kprobes.c +2
-0@@ -19,11 +19,13 @@#include <linux/extable.h> #include <linux/kdebug.h> #include <linux/slab.h> +#include <linux/moduleloader.h> #include <asm/code-patching.h> #include <asm/cacheflush.h> #include <asm/sstep.h> #include <asm/sections.h> #include <asm/inst.h> +#include <asm/set_memory.h> #include <linux/uaccess.h> DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;@@ -103,6 +105,21 @@ kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset)return addr; }
arch/powerpc/kernel/kprobes.c kprobe_opcode_t *kprobe_lookup_name(const char *name, unsigned int offset) · line 116
Quoted 10 lines
arch/powerpc/kernel/kprobes.c +9
-0+void *alloc_insn_page (void ) +{ +void *page; + + page =module_alloc (PAGE_SIZE ); +if (!page) +return NULL ; + +if (IS_ENABLED (CONFIG_STRICT_MODULE_RWX )) {
I'm not sure this test is OK, because strict kernel/module rwx can be disabled at boottime. There is a global variable 'rodata_enabled' to reflect that.
We have a helper in powerpc asm/mmu.h called strict_kernel_rwx_enabled() to check it.
Quoted 10 lines
arch/powerpc/kernel/kprobes.c +6
-0+set_memory_ro ((unsigned long )page,1 ); +set_memory_x ((unsigned long )page,1 ); + } +return page; +} +int arch_prepare_kprobe (struct kprobe *p) {int ret =0 ;
PATCH v12 4/8 powerpc/bpf: Remove bpf_jit_free()
In reply to Jordan NietheCommit 74451e66d516 ("bpf: make jited programs visible in traces") added a default bpf_jit_free() implementation. Powerpc did not use the default bpf_jit_free() as powerpc did not set the images read-only. The default bpf_jit_free() called bpf_jit_binary_unlock_ro() is why it could not be used for powerpc.
Commit d53d2f78cead ("bpf: Use vmalloc special flag") moved keeping track of read-only memory to vmalloc. This included removing bpf_jit_binary_unlock_ro(). Therefore there is no reason powerpc needs its own bpf_jit_free(). Remove it.
---
v11: New to series
--- arch/powerpc/net/bpf_jit_comp.c | 12 ------------ 1 file changed, 12 deletions(-)
arch/powerpc/net/bpf_jit_comp.c +0 -12
@@ -257,15 +257,3 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *fp)return fp; } - -/* Overriding bpf_jit_free() as we don't set images read-only. */ -void bpf_jit_free (struct bpf_prog *fp) -{ -unsigned long addr = (unsigned long )fp->bpf_func &PAGE_MASK ; -struct bpf_binary_header *bpf_hdr = (void *)addr; - -if (fp->jited) -bpf_jit_binary_free (bpf_hdr); - -bpf_prog_unlock_free (fp); -}
Signature · 1 line
2.25.1
Re: [PATCH v12 4/8] powerpc/bpf: Remove bpf_jit_free()
Le 06/05/2021 à 04:34, Jordan Niethe a écrit :
Quoted 6 lines
Commit
74451e66d516("bpf: make jited programs visible in traces") added a defaultbpf_jit_free()implementation. Powerpc did not use the defaultbpf_jit_free()as powerpc did not set the images read-only. The defaultbpf_jit_free()calledbpf_jit_binary_unlock_ro()is why it could not be used for powerpc.
Quoted 6 lines
Commit
d53d2f78cead("bpf: Use vmalloc special flag") moved keeping track of read-only memory to vmalloc. This included removingbpf_jit_binary_unlock_ro(). Therefore there is no reason powerpc needs its ownbpf_jit_free(). Remove it.Signed-off-by Jordan Niethe <[email protected]>
Quoted 27 lines
---
v11: New to series--- arch/powerpc/net/bpf_jit_comp.c | 12 ------------ 1 file changed, 12 deletions(-)arch/powerpc/net/bpf_jit_comp.c +0
-12@@ -257,15 +257,3 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *fp)return fp; } - -/* Overriding bpf_jit_free() as we don't set images read-only. */ -void bpf_jit_free (struct bpf_prog *fp) -{ -unsigned long addr = (unsigned long )fp->bpf_func &PAGE_MASK ; -struct bpf_binary_header *bpf_hdr = (void *)addr; - -if (fp->jited) -bpf_jit_binary_free (bpf_hdr); - -bpf_prog_unlock_free (fp); -}
PATCH v12 5/8 powerpc/bpf: Write protect JIT code
In reply to Jordan NietheAdd the necessary call to bpf_jit_binary_lock_ro() to remove write and add exec permissions to the JIT image after it has finished being written.
Without CONFIG_STRICT_MODULE_RWX the image will be writable and executable until the call to bpf_jit_binary_lock_ro().
---
v10: New to series
v11: Remove CONFIG_STRICT_MODULE_RWX conditional
--- arch/powerpc/net/bpf_jit_comp.c | 1 + 1 file changed, 1 insertion(+)
arch/powerpc/net/bpf_jit_comp.c +1 -0
@@ -237,6 +237,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *fp) fp->jited_len = alloclen;bpf_flush_icache (bpf_hdr, (u8 *)bpf_hdr + (bpf_hdr->pages *PAGE_SIZE )); +bpf_jit_binary_lock_ro (bpf_hdr);if (!fp->is_func || extra_pass) {bpf_prog_fill_jited_linfo (fp, addrs); out_addrs:
Signature · 1 line
2.25.1
Re: [PATCH v12 5/8] powerpc/bpf: Write protect JIT code
Le 06/05/2021 à 04:34, Jordan Niethe a écrit :
Quoted 8 lines
Add the necessary call to
bpf_jit_binary_lock_ro()to remove write and add exec permissions to the JIT image after it has finished being written.Without
CONFIG_STRICT_MODULE_RWXthe image will be writable and executable until the call tobpf_jit_binary_lock_ro().Signed-off-by Jordan Niethe <[email protected]>
Quoted 20 lines
---
v10: New to series
v11: RemoveCONFIG_STRICT_MODULE_RWXconditional--- arch/powerpc/net/bpf_jit_comp.c | 1 + 1 file changed, 1 insertion(+)arch/powerpc/net/bpf_jit_comp.c +1
-0@@ -237,6 +237,7 @@ struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *fp) fp->jited_len = alloclen;bpf_flush_icache (bpf_hdr, (u8 *)bpf_hdr + (bpf_hdr->pages *PAGE_SIZE )); +bpf_jit_binary_lock_ro (bpf_hdr);if (!fp->is_func || extra_pass) {bpf_prog_fill_jited_linfo (fp, addrs); out_addrs:
PATCH v12 6/8 powerpc: Set ARCH_HAS_STRICT_MODULE_RWX
In reply to Jordan NietheFrom: Russell Currey <[email protected]>
To enable strict module RWX on powerpc, set:
CONFIG_STRICT_MODULE_RWX=y
You should also have CONFIG_STRICT_KERNEL_RWX=y set to have any real security benefit.
ARCH_HAS_STRICT_MODULE_RWX is set to require ARCH_HAS_STRICT_KERNEL_RWX. This is due to a quirk in arch/Kconfig and arch/powerpc/Kconfig that makes STRICT_MODULE_RWX *on by default* in configurations where STRICT_KERNEL_RWX is *unavailable*.
Since this doesn't make much sense, and module RWX without kernel RWX doesn't make much sense, having the same dependencies as kernel RWX works around this problem.
With STRICT_MODULE_RWX, now make module_alloc() allocate pages with KERNEL_PAGE protection rather than KERNEL_PAGE_EXEC.
Book32s/32 processors with a hash mmu (i.e. 604 core) can not set memory protection on a page by page basis so do not enable.
[jpn: - predicate on !PPC_BOOK3S_604
- make module_alloc() use PAGE_KERNEL protection]
---
v10: - Predicate on !PPC_BOOK3S_604
- Make module_alloc() use PAGE_KERNEL protection
v11: - Neaten up
--- arch/powerpc/Kconfig | 1 + arch/powerpc/kernel/module.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-)
arch/powerpc/Kconfig +1 -0
@@ -140,6 +140,7 @@ config PPC select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE && PPC_BOOK3S_64 select ARCH_HAS_SET_MEMORY select ARCH_HAS_STRICT_KERNEL_RWX if ((PPC_BOOK3S_64 || PPC32) && !HIBERNATION) + select ARCH_HAS_STRICT_MODULE_RWX if ARCH_HAS_STRICT_KERNEL_RWX && !PPC_BOOK3S_604 select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST select ARCH_HAS_UACCESS_FLUSHCACHE select ARCH_HAS_COPY_MC if PPC64
arch/powerpc/kernel/module.c +3 -1
@@ -92,12 +92,14 @@ int module_finalize(const Elf_Ehdr *hdr, static __always_inline void *__module_alloc (unsigned long size,unsigned long start,unsigned long end) { +pgprot_t prot =IS_ENABLED (CONFIG_STRICT_MODULE_RWX ) ?PAGE_KERNEL : +PAGE_KERNEL_EXEC ;/* * Don't do huge page allocations for modules yet until more testing * is done. STRICT_MODULE_RWX may require extra work to support this * too. */ -return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL ,PAGE_KERNEL_EXEC , +return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL , prot,VM_FLUSH_RESET_PERMS |VM_NO_HUGE_VMAP ,NUMA_NO_NODE ,__builtin_return_address (0 )); }
Signature · 1 line
2.25.1
Re: [PATCH v12 6/8] powerpc: Set ARCH_HAS_STRICT_MODULE_RWX
Le 06/05/2021 à 04:34, Jordan Niethe a écrit :
Quoted 54 lines
From: Russell Currey <[email protected]>
To enable strict module RWX on powerpc, set:
CONFIG_STRICT_MODULE_RWX=yYou should also have
CONFIG_STRICT_KERNEL_RWX=yset to have any real security benefit.
ARCH_HAS_STRICT_MODULE_RWXis set to requireARCH_HAS_STRICT_KERNEL_RWX. This is due to a quirk inarch/Kconfigandarch/powerpc/Kconfigthat makesSTRICT_MODULE_RWX*on by default* in configurations whereSTRICT_KERNEL_RWXis *unavailable*.Since this doesn't make much sense, and module RWX without kernel RWX doesn't make much sense, having the same dependencies as kernel RWX works around this problem.
With
STRICT_MODULE_RWX, now makemodule_alloc()allocate pages withKERNEL_PAGEprotection rather thanKERNEL_PAGE_EXEC.Book32s/32 processors with a hash mmu (i.e. 604 core) can not set memory protection on a page by page basis so do not enable.
Signed-off-by Russell Currey <[email protected]>[jpn: - predicate on !PPC_BOOK3S_604
- makemodule_alloc()usePAGE_KERNELprotection]Signed-off-by Jordan Niethe <[email protected]>---
v10: - Predicate on !PPC_BOOK3S_604
- Makemodule_alloc()usePAGE_KERNELprotection
v11: - Neaten up--- arch/powerpc/Kconfig | 1 + arch/powerpc/kernel/module.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-)arch/powerpc/Kconfig +1
-0@@ -140,6 +140,7 @@ config PPC select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE && PPC_BOOK3S_64 select ARCH_HAS_SET_MEMORY select ARCH_HAS_STRICT_KERNEL_RWX if ((PPC_BOOK3S_64 || PPC32) && !HIBERNATION) + select ARCH_HAS_STRICT_MODULE_RWX if ARCH_HAS_STRICT_KERNEL_RWX && !PPC_BOOK3S_604 select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST select ARCH_HAS_UACCESS_FLUSHCACHE select ARCH_HAS_COPY_MC if PPC64arch/powerpc/kernel/module.c +0
-0
arch/powerpc/kernel/module.c int module_finalize(const Elf_Ehdr *hdr, · line 96
Quoted 5 lines
arch/powerpc/kernel/module.c +2
-0@@ -92,12 +92,14 @@ int module_finalize(const Elf_Ehdr *hdr, static __always_inline void *__module_alloc (unsigned long size,unsigned long start,unsigned long end) { +pgprot_t prot =IS_ENABLED (CONFIG_STRICT_MODULE_RWX ) ?PAGE_KERNEL : +PAGE_KERNEL_EXEC ;
I'm not sure this test is OK, because strict kernel/module rwx can be disabled at boottime. There is a global variable 'rodata_enabled' to reflect that.
We have a helper in powerpc asm/mmu.h called strict_kernel_rwx_enabled() to check it.
Quoted 11 lines
arch/powerpc/kernel/module.c +1
-1/* * Don't do huge page allocations for modules yet until more testing * is done. STRICT_MODULE_RWX may require extra work to support this * too. */ -return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL ,PAGE_KERNEL_EXEC , +return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL , prot,VM_FLUSH_RESET_PERMS |VM_NO_HUGE_VMAP ,NUMA_NO_NODE ,__builtin_return_address (0 )); }
Re: [PATCH v12 6/8] powerpc: Set ARCH_HAS_STRICT_MODULE_RWX
Quoted 67 lines · Christophe Leroy
Quoted 54 lines · Jordan Niethe
From: Russell Currey <[email protected]>
To enable strict module RWX on powerpc, set:
CONFIG_STRICT_MODULE_RWX=yYou should also have
CONFIG_STRICT_KERNEL_RWX=yset to have any real security benefit.
ARCH_HAS_STRICT_MODULE_RWXis set to requireARCH_HAS_STRICT_KERNEL_RWX. This is due to a quirk inarch/Kconfigandarch/powerpc/Kconfigthat makesSTRICT_MODULE_RWX*on by default* in configurations whereSTRICT_KERNEL_RWXis *unavailable*.Since this doesn't make much sense, and module RWX without kernel RWX doesn't make much sense, having the same dependencies as kernel RWX works around this problem.
With
STRICT_MODULE_RWX, now makemodule_alloc()allocate pages withKERNEL_PAGEprotection rather thanKERNEL_PAGE_EXEC.Book32s/32 processors with a hash mmu (i.e. 604 core) can not set memory protection on a page by page basis so do not enable.
Signed-off-by Russell Currey <[email protected]>[jpn: - predicate on !PPC_BOOK3S_604
- makemodule_alloc()usePAGE_KERNELprotection]Signed-off-by Jordan Niethe <[email protected]>---
v10: - Predicate on !PPC_BOOK3S_604
- Makemodule_alloc()usePAGE_KERNELprotection
v11: - Neaten up--- arch/powerpc/Kconfig | 1 + arch/powerpc/kernel/module.c | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-)arch/powerpc/Kconfig +1
-0@@ -140,6 +140,7 @@ config PPC select ARCH_HAS_SCALED_CPUTIME if VIRT_CPU_ACCOUNTING_NATIVE && PPC_BOOK3S_64 select ARCH_HAS_SET_MEMORY select ARCH_HAS_STRICT_KERNEL_RWX if ((PPC_BOOK3S_64 || PPC32) && !HIBERNATION) + select ARCH_HAS_STRICT_MODULE_RWX if ARCH_HAS_STRICT_KERNEL_RWX && !PPC_BOOK3S_604 select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST select ARCH_HAS_UACCESS_FLUSHCACHE select ARCH_HAS_COPY_MC if PPC64arch/powerpc/kernel/module.c +0
-0arch/powerpc/kernel/module.c int module_finalize(const Elf_Ehdr *hdr, · line 96
Quoted 5 lines
arch/powerpc/kernel/module.c +2
-0@@ -92,12 +92,14 @@ int module_finalize(const Elf_Ehdr *hdr, static __always_inline void *__module_alloc (unsigned long size,unsigned long start,unsigned long end) { +pgprot_t prot =IS_ENABLED (CONFIG_STRICT_MODULE_RWX ) ?PAGE_KERNEL : +PAGE_KERNEL_EXEC ;I'm not sure this test is OK, because strict kernel/module rwx can be disabled at boottime. There is a global variable '
rodata_enabled' to reflect that.
Quoted 1 line
We have a helper in powerpc
asm/mmu.hcalledstrict_kernel_rwx_enabled()to check it.
Thanks, I will change to that.
Quoted 13 lines
Quoted 13 lines
arch/powerpc/kernel/module.c +1
-1/* * Don't do huge page allocations for modules yet until more testing * is done. STRICT_MODULE_RWX may require extra work to support this * too. */ -return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL ,PAGE_KERNEL_EXEC , +return __vmalloc_node_range (size,1 , start, end,GFP_KERNEL , prot,VM_FLUSH_RESET_PERMS |VM_NO_HUGE_VMAP ,NUMA_NO_NODE ,__builtin_return_address (0 )); }
PATCH v12 7/8 powerpc/mm: implement set_memory_attr()
In reply to Jordan NietheFrom: Christophe Leroy <[email protected]>
In addition to the set_memory_xx() functions which allows to change the memory attributes of not (yet) used memory regions, implement a set_memory_attr() function to:
- set the final memory protection after init on currently used kernel regions.
- enable/disable kernel memory regions in the scope of DEBUG_PAGEALLOC.
Unlike the set_memory_xx() which can act in three step as the regions are unused, this function must modify 'on the fly' as the kernel is executing from them. At the moment only PPC32 will use it and changing page attributes on the fly is not an issue.
--- arch/powerpc/include/asm/set_memory.h | 2 ++ arch/powerpc/mm/pageattr.c | 33 +++++++++++++++++++++++++++ 2 files changed, 35 insertions(+)
arch/powerpc/include/asm/set_memory.h +2 -0
@@ -29,4 +29,6 @@ static inline int set_memory_x(unsigned long addr, int numpages) return change_memory_attr(addr, numpages, SET_MEMORY_X); } +int set_memory_attr (unsigned long addr,int numpages,pgprot_t prot); +#endif
arch/powerpc/mm/pageattr.c +33 -0
@@ -98,3 +98,36 @@ int change_memory_attr(unsigned long addr, int numpages, long action)return apply_to_existing_page_range (&init_mm, start, size, change_page_attr, (void *)action); } + +/* +* Set the attributes of a page: +* +* This function is used by PPC32 at the end of init to set final kernel memory +* protection. It includes changing the maping of the page it is executing from +* and data pages it is using. +*/ +static int set_page_attr (pte_t *ptep,unsigned long addr,void *data) +{ +pgprot_t prot =__pgprot ((unsigned long )data); + +spin_lock (&init_mm.page_table_lock); + +set_pte_at (&init_mm, addr, ptep,pte_modify (*ptep, prot)); +flush_tlb_kernel_range (addr, addr +PAGE_SIZE ); + +spin_unlock (&init_mm.page_table_lock); + +return 0 ; +} + +int set_memory_attr (unsigned long addr,int numpages,pgprot_t prot) +{ +unsigned long start =ALIGN_DOWN (addr,PAGE_SIZE ); +unsigned long sz = numpages *PAGE_SIZE ; + +if (numpages <=0 ) +return 0 ; + +return apply_to_existing_page_range (&init_mm, start, sz, set_page_attr, + (void *)pgprot_val (prot)); +}
Signature · 1 line
2.25.1
PATCH v12 8/8 powerpc/32: use set_memory_attr()
In reply to Jordan NietheFrom: Christophe Leroy <[email protected]>
Use set_memory_attr() instead of the PPC32 specific change_page_attr()
change_page_attr() was checking that the address was not mapped by blocks and was handling highmem, but that's unneeded because the affected pages can't be in highmem and block mapping verification is already done by the callers.
--- arch/powerpc/mm/pgtable_32.c | 60 ++++++------------------------------ 1 file changed, 10 insertions(+), 50 deletions(-)
arch/powerpc/mm/pgtable_32.c +10 -50
@@ -23,6 +23,7 @@#include <linux/highmem.h> #include <linux/memblock.h> #include <linux/slab.h> +#include <linux/set_memory.h> #include <asm/pgalloc.h> #include <asm/fixmap.h>
@@ -132,64 +133,20 @@ void __init mapin_ram(void) } } -static int __change_page_attr_noflush (struct page *page,pgprot_t prot) -{ -pte_t *kpte; -unsigned long address; - -BUG_ON (PageHighMem (page)); - address = (unsigned long )page_address (page); - -if (v_block_mapped (address)) -return 0 ; - kpte =virt_to_kpte (address); -if (!kpte) -return -EINVAL ; -__set_pte_at (&init_mm, address, kpte,mk_pte (page, prot),0 ); - -return 0 ; -} - -/* -* Change the page attributes of an page in the linear mapping. -* -* THIS DOES NOTHING WITH BAT MAPPINGS, DEBUG USE ONLY -*/ -static int change_page_attr (struct page *page,int numpages,pgprot_t prot) -{ -int i, err =0 ; -unsigned long flags; -struct page *start = page; - -local_irq_save (flags); -for (i =0 ; i < numpages; i++, page++) { - err =__change_page_attr_noflush (page, prot); -if (err) -break ; - } -wmb (); -local_irq_restore (flags); -flush_tlb_kernel_range ((unsigned long )page_address (start), - (unsigned long )page_address (page)); -return err; -} -void mark_initmem_nx (void ) { -struct page *page =virt_to_page (_sinittext);unsigned long numpages =PFN_UP ((unsigned long )_einittext) -PFN_DOWN ((unsigned long )_sinittext);if (v_block_mapped ((unsigned long )_sinittext))mmu_mark_initmem_nx ();else -change_page_attr (page, numpages,PAGE_KERNEL ); +set_memory_attr ((unsigned long )_sinittext, numpages,PAGE_KERNEL ); }#ifdef CONFIG_STRICT_KERNEL_RWX void mark_rodata_ro (void ) { -struct page *page;unsigned long numpages;if (v_block_mapped ((unsigned long )_stext +1 )) {
@@ -198,20 +155,18 @@ void mark_rodata_ro(void)return ; } - page =virt_to_page (_stext); numpages =PFN_UP ((unsigned long )_etext) -PFN_DOWN ((unsigned long )_stext); -change_page_attr (page, numpages,PAGE_KERNEL_ROX ); +set_memory_attr ((unsigned long )_stext, numpages,PAGE_KERNEL_ROX );/* * mark .rodata as read only. Use __init_begin rather than __end_rodata * to cover NOTES and EXCEPTION_TABLE. */ - page =virt_to_page (__start_rodata); numpages =PFN_UP ((unsigned long )__init_begin) -PFN_DOWN ((unsigned long )__start_rodata); -change_page_attr (page, numpages,PAGE_KERNEL_RO ); +set_memory_attr ((unsigned long )__start_rodata, numpages,PAGE_KERNEL_RO );// mark_initmem_nx() should have already run by now ptdump_check_wx ();
@@ -221,9 +176,14 @@ void mark_rodata_ro(void)#ifdef CONFIG_DEBUG_PAGEALLOC void __kernel_map_pages (struct page *page,int numpages,int enable) { +unsigned long addr = (unsigned long )page_address (page); +if (PageHighMem (page))return ; -change_page_attr (page, numpages, enable ?PAGE_KERNEL :__pgprot (0 )); +if (enable) +set_memory_attr (addr, numpages,PAGE_KERNEL ); +else +set_memory_attr (addr, numpages,__pgprot (0 )); }#endif /* CONFIG_DEBUG_PAGEALLOC */
Signature · 1 line
2.25.1