| Message ID | 20191212151656.26151-4-dja@axtens.net (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | KASAN for powerpc64 radix | expand |
On Fri, Dec 13, 2019 at 2:19 AM Daniel Axtens <dja@axtens.net> wrote: > > KASAN support on Book3S is a bit tricky to get right: > > - It would be good to support inline instrumentation so as to be able to > catch stack issues that cannot be caught with outline mode. > > - Inline instrumentation requires a fixed offset. > > - Book3S runs code in real mode after booting. Most notably a lot of KVM > runs in real mode, and it would be good to be able to instrument it. > > - Because code runs in real mode after boot, the offset has to point to > valid memory both in and out of real mode. > > [For those not immersed in ppc64, in real mode, the top nibble or 2 bits > (depending on radix/hash mmu) of the address is ignored. The linear > mapping is placed at 0xc000000000000000. This means that a pointer to > part of the linear mapping will work both in real mode, where it will be > interpreted as a physical address of the form 0x000..., and out of real > mode, where it will go via the linear mapping.] > How does hash or radix mmu mode effect how many bits are ignored in real mode? > One approach is just to give up on inline instrumentation. This way all > checks can be delayed until after everything set is up correctly, and the > address-to-shadow calculations can be overridden. However, the features and > speed boost provided by inline instrumentation are worth trying to do > better. > > If _at compile time_ it is known how much contiguous physical memory a > system has, the top 1/8th of the first block of physical memory can be set > aside for the shadow. This is a big hammer and comes with 3 big > consequences: > > - there's no nice way to handle physically discontiguous memory, so only > the first physical memory block can be used. > > - kernels will simply fail to boot on machines with less memory than > specified when compiling. > > - kernels running on machines with more memory than specified when > compiling will simply ignore the extra memory. > > Implement and document KASAN this way. The current implementation is Radix > only. > > Despite the limitations, it can still find bugs, > e.g. http://patchwork.ozlabs.org/patch/1103775/ > > At the moment, this physical memory limit must be set _even for outline > mode_. This may be changed in a later series - a different implementation > could be added for outline mode that dynamically allocates shadow at a > fixed offset. For example, see https://patchwork.ozlabs.org/patch/795211/ > > Suggested-by: Michael Ellerman <mpe@ellerman.id.au> > Cc: Balbir Singh <bsingharora@gmail.com> # ppc64 out-of-line radix version > Cc: Christophe Leroy <christophe.leroy@c-s.fr> # ppc32 version > Signed-off-by: Daniel Axtens <dja@axtens.net> > > --- > Changes since v2: > > - Address feedback from Christophe around cleanups and docs. > - Address feedback from Balbir: at this point I don't have a good solution > for the issues you identify around the limitations of the inline implementation > but I think that it's worth trying to get the stack instrumentation support. > I'm happy to have an alternative and more flexible outline mode - I had > envisoned this would be called 'lightweight' mode as it imposes fewer restrictions. > I've linked to your implementation. I think it's best to add it in a follow-up series. > - Made the default PHYS_MEM_SIZE_FOR_KASAN value 1024MB. I think most people have > guests with at least that much memory in the Radix 64s case so it's a much > saner default - it means that if you just turn on KASAN without reading the > docs you're much more likely to have a bootable kernel, which you will never > have if the value is set to zero! I'm happy to bikeshed the value if we want. > > Changes since v1: > - Landed kasan vmalloc support upstream > - Lots of feedback from Christophe. > > Changes since the rfc: > > - Boots real and virtual hardware, kvm works. > > - disabled reporting when we're checking the stack for exception > frames. The behaviour isn't wrong, just incompatible with KASAN. > > - Documentation! > > - Dropped old module stuff in favour of KASAN_VMALLOC. > > The bugs with ftrace and kuap were due to kernel bloat pushing > prom_init calls to be done via the plt. Because we did not have > a relocatable kernel, and they are done very early, this caused > everything to explode. Compile with CONFIG_RELOCATABLE! > --- > Documentation/dev-tools/kasan.rst | 8 +- > Documentation/powerpc/kasan.txt | 112 +++++++++++++++++- > arch/powerpc/Kconfig | 3 + > arch/powerpc/Kconfig.debug | 21 ++++ > arch/powerpc/Makefile | 11 ++ > arch/powerpc/include/asm/book3s/64/hash.h | 4 + > arch/powerpc/include/asm/book3s/64/pgtable.h | 7 ++ > arch/powerpc/include/asm/book3s/64/radix.h | 5 + > arch/powerpc/include/asm/kasan.h | 21 +++- > arch/powerpc/kernel/process.c | 8 ++ > arch/powerpc/kernel/prom.c | 64 +++++++++- > arch/powerpc/mm/kasan/Makefile | 3 +- > .../mm/kasan/{kasan_init_32.c => init_32.c} | 0 > arch/powerpc/mm/kasan/init_book3s_64.c | 72 +++++++++++ > 14 files changed, 330 insertions(+), 9 deletions(-) > rename arch/powerpc/mm/kasan/{kasan_init_32.c => init_32.c} (100%) > create mode 100644 arch/powerpc/mm/kasan/init_book3s_64.c > > diff --git a/Documentation/dev-tools/kasan.rst b/Documentation/dev-tools/kasan.rst > index 4af2b5d2c9b4..d99dc580bc11 100644 > --- a/Documentation/dev-tools/kasan.rst > +++ b/Documentation/dev-tools/kasan.rst > @@ -22,8 +22,9 @@ global variables yet. > Tag-based KASAN is only supported in Clang and requires version 7.0.0 or later. > > Currently generic KASAN is supported for the x86_64, arm64, xtensa and s390 > -architectures. It is also supported on 32-bit powerpc kernels. Tag-based KASAN > -is supported only on arm64. > +architectures. It is also supported on powerpc, for 32-bit kernels, and for > +64-bit kernels running under the Radix MMU. Tag-based KASAN is supported only > +on arm64. > > Usage > ----- > @@ -256,7 +257,8 @@ CONFIG_KASAN_VMALLOC > ~~~~~~~~~~~~~~~~~~~~ > > With ``CONFIG_KASAN_VMALLOC``, KASAN can cover vmalloc space at the > -cost of greater memory usage. Currently this is only supported on x86. > +cost of greater memory usage. Currently this is optional on x86, and > +required on 64-bit powerpc. > > This works by hooking into vmalloc and vmap, and dynamically > allocating real shadow memory to back the mappings. > diff --git a/Documentation/powerpc/kasan.txt b/Documentation/powerpc/kasan.txt > index a85ce2ff8244..f134a91600ad 100644 > --- a/Documentation/powerpc/kasan.txt > +++ b/Documentation/powerpc/kasan.txt > @@ -1,4 +1,4 @@ > -KASAN is supported on powerpc on 32-bit only. > +KASAN is supported on powerpc on 32-bit and Radix 64-bit only. > > 32 bit support > ============== > @@ -10,3 +10,113 @@ fixmap area and occupies one eighth of the total kernel virtual memory space. > > Instrumentation of the vmalloc area is not currently supported, but modules > are. > + > +64 bit support > +============== > + > +Currently, only the radix MMU is supported. There have been versions for Book3E > +processors floating around on the mailing list, but nothing has been merged. > + > +KASAN support on Book3S is a bit tricky to get right: > + > + - It would be good to support inline instrumentation so as to be able to catch > + stack issues that cannot be caught with outline mode. > + > + - Inline instrumentation requires a fixed offset. > + > + - Book3S runs code in real mode after booting. Most notably a lot of KVM runs > + in real mode, and it would be good to be able to instrument it. > + > + - Because code runs in real mode after boot, the offset has to point to > + valid memory both in and out of real mode. > + > +One approach is just to give up on inline instrumentation. This way all checks > +can be delayed until after everything set is up correctly, and the > +address-to-shadow calculations can be overridden. However, the features and > +speed boost provided by inline instrumentation are worth trying to do better. > + > +If _at compile time_ it is known how much contiguous physical memory a system > +has, the top 1/8th of the first block of physical memory can be set aside for > +the shadow. This is a big hammer and comes with 3 big consequences: > + > + - there's no nice way to handle physically discontiguous memory, so only the > + first physical memory block can be used. > + > + - kernels will simply fail to boot on machines with less memory than specified > + when compiling. > + > + - kernels running on machines with more memory than specified when compiling > + will simply ignore the extra memory. > + > +At the moment, this physical memory limit must be set _even for outline mode_. > +This may be changed in a future version - a different implementation could be > +added for outline mode that dynamically allocates shadow at a fixed offset. > +For example, see https://patchwork.ozlabs.org/patch/795211/ > + > +This value is configured in CONFIG_PHYS_MEM_SIZE_FOR_KASAN. > + > +Tips > +---- > + > + - Compile with CONFIG_RELOCATABLE. > + > + In development, boot hangs were observed when building with ftrace and KUAP > + on. These ended up being due to kernel bloat pushing prom_init calls to be > + done via the PLT. Because the kernel was not relocatable, and the calls are > + done very early, this caused execution to jump off into somewhere > + invalid. Enabling relocation fixes this. > + > +NUMA/discontiguous physical memory > +---------------------------------- > + > +Currently the code cannot really deal with discontiguous physical memory. Only > +physical memory that is contiguous from physical address zero can be used. The > +size of that memory, not total memory, must be specified when configuring the > +kernel. > + > +Discontiguous memory can occur on machines with memory spread across multiple > +nodes. For example, on a Talos II with 64GB of RAM: > + > + - 32GB runs from 0x0 to 0x0000_0008_0000_0000, > + - then there's a gap, > + - then the final 32GB runs from 0x0000_2000_0000_0000 to 0x0000_2008_0000_0000 > + > +This can create _significant_ issues: > + > + - If the machine is treated as having 64GB of _contiguous_ RAM, the > + instrumentation would assume that it ran from 0x0 to > + 0x0000_0010_0000_0000. The last 1/8th - 0x0000_000e_0000_0000 to > + 0x0000_0010_0000_0000 would be reserved as the shadow region. But when the > + kernel tried to access any of that, it would be trying to access pages that > + are not physically present. > + > + - If the shadow region size is based on the top address, then the shadow > + region would be 0x2008_0000_0000 / 8 = 0x0401_0000_0000 bytes = 4100 GB of > + memory, clearly more than the 64GB of RAM physically present. > + > +Therefore, the code currently is restricted to dealing with memory in the node > +starting at 0x0. For this system, that's 32GB. If a contiguous physical memory > +size greater than the size of the first contiguous region of memory is > +specified, the system will be unable to boot or even print an error message. > + > +The layout of a system's memory can be observed in the messages that the Radix > +MMU prints on boot. The Talos II discussed earlier has: > + > +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) > +radix-mmu: Mapped 0x0000000040000000-0x0000000800000000 with 1.00 GiB pages > +radix-mmu: Mapped 0x0000200000000000-0x0000200800000000 with 1.00 GiB pages > + > +As discussed, this system would be configured for 32768 MB. > + > +Another system prints: > + > +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) > +radix-mmu: Mapped 0x0000000040000000-0x0000002000000000 with 1.00 GiB pages > +radix-mmu: Mapped 0x0000200000000000-0x0000202000000000 with 1.00 GiB pages > + > +This machine has more memory: 0x0000_0040_0000_0000 total, but only > +0x0000_0020_0000_0000 is physically contiguous from zero, so it would be > +configured for 131072 MB of physically contiguous memory. > + > +This restriction currently also affects outline mode, but this could be > +changed in future if an alternative outline implementation is added. > diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig > index 6987b0832e5f..2561446e85a8 100644 > --- a/arch/powerpc/Kconfig > +++ b/arch/powerpc/Kconfig > @@ -173,6 +173,9 @@ config PPC > select HAVE_ARCH_HUGE_VMAP if PPC_BOOK3S_64 && PPC_RADIX_MMU > select HAVE_ARCH_JUMP_LABEL > select HAVE_ARCH_KASAN if PPC32 > + select HAVE_ARCH_KASAN if PPC_BOOK3S_64 && PPC_RADIX_MMU > + select HAVE_ARCH_KASAN_VMALLOC if PPC_BOOK3S_64 && PPC_RADIX_MMU > + select KASAN_VMALLOC if KASAN && PPC_BOOK3S_64 > select HAVE_ARCH_KGDB > select HAVE_ARCH_MMAP_RND_BITS > select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT > diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug > index 4e1d39847462..5c454f8fa24b 100644 > --- a/arch/powerpc/Kconfig.debug > +++ b/arch/powerpc/Kconfig.debug > @@ -394,6 +394,27 @@ config PPC_FAST_ENDIAN_SWITCH > help > If you're unsure what this is, say N. > > +config PHYS_MEM_SIZE_FOR_KASAN > + int "Contiguous physical memory size for KASAN (MB)" if KASAN && PPC_BOOK3S_64 > + default 1024 > + help > + > + To get inline instrumentation support for KASAN on 64-bit Book3S > + machines, you need to know how much contiguous physical memory your > + system has. A shadow offset will be calculated based on this figure, > + which will be compiled in to the kernel. KASAN will use this offset > + to access its shadow region, which is used to verify memory accesses. > + > + If you attempt to boot on a system with less memory than you specify > + here, your system will fail to boot very early in the process. If you > + boot on a system with more memory than you specify, the extra memory > + will wasted - it will be reserved and not used. > + > + For systems with discontiguous blocks of physical memory, specify the > + size of the block starting at 0x0. You can determine this by looking > + at the memory layout info printed to dmesg by the radix MMU code > + early in boot. See Documentation/powerpc/kasan.txt. > + > config KASAN_SHADOW_OFFSET > hex > depends on KASAN > diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile > index f35730548e42..eff693527462 100644 > --- a/arch/powerpc/Makefile > +++ b/arch/powerpc/Makefile > @@ -230,6 +230,17 @@ ifdef CONFIG_476FPE_ERR46 > -T $(srctree)/arch/powerpc/platforms/44x/ppc476_modules.lds > endif > > +ifdef CONFIG_PPC_BOOK3S_64 > +# The KASAN shadow offset is such that linear map (0xc000...) is shadowed by > +# the last 8th of linearly mapped physical memory. This way, if the code uses > +# 0xc addresses throughout, accesses work both in in real mode (where the top > +# 2 bits are ignored) and outside of real mode. > +# > +# 0xc000000000000000 >> 3 = 0xa800000000000000 = 12105675798371893248 > +KASAN_SHADOW_OFFSET = $(shell echo 7 \* 1024 \* 1024 \* $(CONFIG_PHYS_MEM_SIZE_FOR_KASAN) / 8 + 12105675798371893248 | bc) > +KBUILD_CFLAGS += -DKASAN_SHADOW_OFFSET=$(KASAN_SHADOW_OFFSET)UL > +endif > + > # No AltiVec or VSX instructions when building kernel > KBUILD_CFLAGS += $(call cc-option,-mno-altivec) > KBUILD_CFLAGS += $(call cc-option,-mno-vsx) > diff --git a/arch/powerpc/include/asm/book3s/64/hash.h b/arch/powerpc/include/asm/book3s/64/hash.h > index 2781ebf6add4..fce329b8452e 100644 > --- a/arch/powerpc/include/asm/book3s/64/hash.h > +++ b/arch/powerpc/include/asm/book3s/64/hash.h > @@ -18,6 +18,10 @@ > #include <asm/book3s/64/hash-4k.h> > #endif > > +#define H_PTRS_PER_PTE (1 << H_PTE_INDEX_SIZE) > +#define H_PTRS_PER_PMD (1 << H_PMD_INDEX_SIZE) > +#define H_PTRS_PER_PUD (1 << H_PUD_INDEX_SIZE) > + > /* Bits to set in a PMD/PUD/PGD entry valid bit*/ > #define HASH_PMD_VAL_BITS (0x8000000000000000UL) > #define HASH_PUD_VAL_BITS (0x8000000000000000UL) > diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h > index b01624e5c467..209817235a44 100644 > --- a/arch/powerpc/include/asm/book3s/64/pgtable.h > +++ b/arch/powerpc/include/asm/book3s/64/pgtable.h > @@ -231,6 +231,13 @@ extern unsigned long __pmd_frag_size_shift; > #define PTRS_PER_PUD (1 << PUD_INDEX_SIZE) > #define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) > > +#define MAX_PTRS_PER_PTE ((H_PTRS_PER_PTE > R_PTRS_PER_PTE) ? \ > + H_PTRS_PER_PTE : R_PTRS_PER_PTE) > +#define MAX_PTRS_PER_PMD ((H_PTRS_PER_PMD > R_PTRS_PER_PMD) ? \ > + H_PTRS_PER_PMD : R_PTRS_PER_PMD) > +#define MAX_PTRS_PER_PUD ((H_PTRS_PER_PUD > R_PTRS_PER_PUD) ? \ > + H_PTRS_PER_PUD : R_PTRS_PER_PUD) > + > /* PMD_SHIFT determines what a second-level page table entry can map */ > #define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) > #define PMD_SIZE (1UL << PMD_SHIFT) > diff --git a/arch/powerpc/include/asm/book3s/64/radix.h b/arch/powerpc/include/asm/book3s/64/radix.h > index d97db3ad9aae..4f826259de71 100644 > --- a/arch/powerpc/include/asm/book3s/64/radix.h > +++ b/arch/powerpc/include/asm/book3s/64/radix.h > @@ -35,6 +35,11 @@ > #define RADIX_PMD_SHIFT (PAGE_SHIFT + RADIX_PTE_INDEX_SIZE) > #define RADIX_PUD_SHIFT (RADIX_PMD_SHIFT + RADIX_PMD_INDEX_SIZE) > #define RADIX_PGD_SHIFT (RADIX_PUD_SHIFT + RADIX_PUD_INDEX_SIZE) > + > +#define R_PTRS_PER_PTE (1 << RADIX_PTE_INDEX_SIZE) > +#define R_PTRS_PER_PMD (1 << RADIX_PMD_INDEX_SIZE) > +#define R_PTRS_PER_PUD (1 << RADIX_PUD_INDEX_SIZE) > + > /* > * Size of EA range mapped by our pagetables. > */ > diff --git a/arch/powerpc/include/asm/kasan.h b/arch/powerpc/include/asm/kasan.h > index 296e51c2f066..f18268cbdc33 100644 > --- a/arch/powerpc/include/asm/kasan.h > +++ b/arch/powerpc/include/asm/kasan.h > @@ -2,6 +2,9 @@ > #ifndef __ASM_KASAN_H > #define __ASM_KASAN_H > > +#include <asm/page.h> > +#include <asm/pgtable.h> > + > #ifdef CONFIG_KASAN > #define _GLOBAL_KASAN(fn) _GLOBAL(__##fn) > #define _GLOBAL_TOC_KASAN(fn) _GLOBAL_TOC(__##fn) > @@ -14,13 +17,19 @@ > > #ifndef __ASSEMBLY__ > > -#include <asm/page.h> > +#ifdef CONFIG_KASAN > +void kasan_init(void); > +#else > +static inline void kasan_init(void) { } > +#endif > > #define KASAN_SHADOW_SCALE_SHIFT 3 > > #define KASAN_SHADOW_START (KASAN_SHADOW_OFFSET + \ > (PAGE_OFFSET >> KASAN_SHADOW_SCALE_SHIFT)) > > +#ifdef CONFIG_PPC32 > + > #define KASAN_SHADOW_OFFSET ASM_CONST(CONFIG_KASAN_SHADOW_OFFSET) > > #define KASAN_SHADOW_END 0UL > @@ -30,11 +39,17 @@ > #ifdef CONFIG_KASAN > void kasan_early_init(void); > void kasan_mmu_init(void); > -void kasan_init(void); > #else > -static inline void kasan_init(void) { } > static inline void kasan_mmu_init(void) { } > #endif > +#endif > + > +#ifdef CONFIG_PPC_BOOK3S_64 > + > +#define KASAN_SHADOW_SIZE ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN * \ > + 1024 * 1024 * 1 / 8) > + > +#endif /* CONFIG_PPC_BOOK3S_64 */ > > #endif /* __ASSEMBLY */ > #endif > diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c > index 4df94b6e2f32..c60ff299f39b 100644 > --- a/arch/powerpc/kernel/process.c > +++ b/arch/powerpc/kernel/process.c > @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) > /* > * See if this is an exception frame. > * We look for the "regshere" marker in the current frame. > + * > + * KASAN may complain about this. If it is an exception frame, > + * we won't have unpoisoned the stack in asm when we set the > + * exception marker. If it's not an exception frame, who knows > + * how things are laid out - the shadow could be in any state > + * at all. Just disable KASAN reporting for now. > */ > + kasan_disable_current(); > if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) > && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { > struct pt_regs *regs = (struct pt_regs *) > @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) > regs->trap, (void *)regs->nip, (void *)lr); > firstframe = 1; > } > + kasan_enable_current(); > > sp = newsp; > } while (count++ < kstack_depth_to_print); > diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c > index 6620f37abe73..d994c7c39c8d 100644 > --- a/arch/powerpc/kernel/prom.c > +++ b/arch/powerpc/kernel/prom.c > @@ -72,6 +72,7 @@ unsigned long tce_alloc_start, tce_alloc_end; > u64 ppc64_rma_size; > #endif > static phys_addr_t first_memblock_size; > +static phys_addr_t top_phys_addr; > static int __initdata boot_cpu_count; > > static int __init early_parse_mem(char *p) > @@ -449,6 +450,26 @@ static bool validate_mem_limit(u64 base, u64 *size) > { > u64 max_mem = 1UL << (MAX_PHYSMEM_BITS); > > + /* > + * To handle the NUMA/discontiguous memory case, don't allow a block > + * to be added if it falls completely beyond the configured physical > + * memory. Print an informational message. > + * > + * Frustratingly we also see this with qemu - it seems to split the > + * specified memory into a number of smaller blocks. If this happens > + * under qemu, it probably represents misconfiguration. So we want > + * the message to be noticeable, but not shouty. > + * > + * See Documentation/powerpc/kasan.txt > + */ > + if (IS_ENABLED(CONFIG_KASAN) && > + (base >= ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20))) { > + pr_warn("KASAN: not adding memory block at %llx (size %llx)\n" > + "This could be due to discontiguous memory or kernel misconfiguration.", > + base, *size); > + return false; > + } > + > if (base >= max_mem) > return false; > if ((base + *size) > max_mem) > @@ -572,8 +593,11 @@ void __init early_init_dt_add_memory_arch(u64 base, u64 size) > > /* Add the chunk to the MEMBLOCK list */ > if (add_mem_to_memblock) { > - if (validate_mem_limit(base, &size)) > + if (validate_mem_limit(base, &size)) { > memblock_add(base, size); > + if (base + size > top_phys_addr) > + top_phys_addr = base + size; > + } > } > } > > @@ -613,6 +637,8 @@ static void __init early_reserve_mem_dt(void) > static void __init early_reserve_mem(void) > { > __be64 *reserve_map; > + phys_addr_t kasan_shadow_start; > + phys_addr_t kasan_memory_size; > > reserve_map = (__be64 *)(((unsigned long)initial_boot_params) + > fdt_off_mem_rsvmap(initial_boot_params)); > @@ -651,6 +677,42 @@ static void __init early_reserve_mem(void) > return; > } > #endif > + > + if (IS_ENABLED(CONFIG_KASAN) && IS_ENABLED(CONFIG_PPC_BOOK3S_64)) { > + kasan_memory_size = > + ((phys_addr_t)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20); > + > + if (top_phys_addr < kasan_memory_size) { > + /* > + * We are doomed. We shouldn't even be able to get this > + * far, but we do in qemu. If we continue and turn > + * relocations on, we'll take fatal page faults for > + * memory that's not physically present. Instead, > + * panic() here: it will be saved to __log_buf even if > + * it doesn't get printed to the console. > + */ > + panic("Tried to book a KASAN kernel configured for %u MB with only %llu MB! Aborting.", > + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, > + (u64)(top_phys_addr >> 20)); > + } else if (top_phys_addr > kasan_memory_size) { > + /* print a biiiig warning in hopes people notice */ > + pr_err("===========================================\n" > + "Physical memory exceeds compiled-in maximum!\n" > + "This kernel was compiled for KASAN with %u MB physical memory.\n" > + "The physical memory detected is at least %llu MB.\n" > + "Memory above the compiled limit will not be used!\n" > + "===========================================\n", > + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, > + (u64)(top_phys_addr >> 20)); > + } > + > + kasan_shadow_start = _ALIGN_DOWN(kasan_memory_size * 7 / 8, > + PAGE_SIZE); > + DBG("reserving %llx -> %llx for KASAN", > + kasan_shadow_start, top_phys_addr); > + memblock_reserve(kasan_shadow_start, > + top_phys_addr - kasan_shadow_start); > + } > } > > #ifdef CONFIG_PPC_TRANSACTIONAL_MEM > diff --git a/arch/powerpc/mm/kasan/Makefile b/arch/powerpc/mm/kasan/Makefile > index 6577897673dd..f02b15c78e4d 100644 > --- a/arch/powerpc/mm/kasan/Makefile > +++ b/arch/powerpc/mm/kasan/Makefile > @@ -2,4 +2,5 @@ > > KASAN_SANITIZE := n > > -obj-$(CONFIG_PPC32) += kasan_init_32.o > +obj-$(CONFIG_PPC32) += init_32.o > +obj-$(CONFIG_PPC_BOOK3S_64) += init_book3s_64.o > diff --git a/arch/powerpc/mm/kasan/kasan_init_32.c b/arch/powerpc/mm/kasan/init_32.c > similarity index 100% > rename from arch/powerpc/mm/kasan/kasan_init_32.c > rename to arch/powerpc/mm/kasan/init_32.c > diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/kasan/init_book3s_64.c > new file mode 100644 > index 000000000000..f961e96be136 > --- /dev/null > +++ b/arch/powerpc/mm/kasan/init_book3s_64.c > @@ -0,0 +1,72 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * KASAN for 64-bit Book3S powerpc > + * > + * Copyright (C) 2019 IBM Corporation > + * Author: Daniel Axtens <dja@axtens.net> > + */ > + > +#define DISABLE_BRANCH_PROFILING > + > +#include <linux/kasan.h> > +#include <linux/printk.h> > +#include <linux/sched/task.h> > +#include <asm/pgalloc.h> > + > +void __init kasan_init(void) > +{ > + int i; > + void *k_start = kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START); > + void *k_end = kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END); > + > + pte_t pte = __pte(__pa(kasan_early_shadow_page) | > + pgprot_val(PAGE_KERNEL) | _PAGE_PTE); > + > + if (!early_radix_enabled()) > + panic("KASAN requires radix!"); > + > + for (i = 0; i < PTRS_PER_PTE; i++) > + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, > + &kasan_early_shadow_pte[i], pte, 0); > + > + for (i = 0; i < PTRS_PER_PMD; i++) > + pmd_populate_kernel(&init_mm, &kasan_early_shadow_pmd[i], > + kasan_early_shadow_pte); > + > + for (i = 0; i < PTRS_PER_PUD; i++) > + pud_populate(&init_mm, &kasan_early_shadow_pud[i], > + kasan_early_shadow_pmd); > + > + memset(kasan_mem_to_shadow((void *)PAGE_OFFSET), KASAN_SHADOW_INIT, > + KASAN_SHADOW_SIZE); > + > + kasan_populate_early_shadow( > + kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START), > + kasan_mem_to_shadow((void *)RADIX_VMALLOC_START)); > + > + /* leave a hole here for vmalloc */ > + > + kasan_populate_early_shadow( > + kasan_mem_to_shadow((void *)RADIX_VMALLOC_END), > + kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END)); > + > + flush_tlb_kernel_range((unsigned long)k_start, (unsigned long)k_end); > + > + /* mark early shadow region as RO and wipe */ > + pte = __pte(__pa(kasan_early_shadow_page) | > + pgprot_val(PAGE_KERNEL_RO) | _PAGE_PTE); > + for (i = 0; i < PTRS_PER_PTE; i++) > + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, > + &kasan_early_shadow_pte[i], pte, 0); > + > + /* > + * clear_page relies on some cache info that hasn't been set up yet. > + * It ends up looping ~forever and blows up other data. > + * Use memset instead. > + */ > + memset(kasan_early_shadow_page, 0, PAGE_SIZE); > + > + /* Enable error messages */ > + init_task.kasan_depth = 0; > + pr_info("KASAN init done (64-bit Book3S heavyweight mode)\n"); > +} > -- > 2.20.1 >
Le 12/12/2019 à 16:16, Daniel Axtens a écrit : > KASAN support on Book3S is a bit tricky to get right: > > - It would be good to support inline instrumentation so as to be able to > catch stack issues that cannot be caught with outline mode. > > - Inline instrumentation requires a fixed offset. > > - Book3S runs code in real mode after booting. Most notably a lot of KVM > runs in real mode, and it would be good to be able to instrument it. > > - Because code runs in real mode after boot, the offset has to point to > valid memory both in and out of real mode. > > [For those not immersed in ppc64, in real mode, the top nibble or 2 bits > (depending on radix/hash mmu) of the address is ignored. The linear > mapping is placed at 0xc000000000000000. This means that a pointer to > part of the linear mapping will work both in real mode, where it will be > interpreted as a physical address of the form 0x000..., and out of real > mode, where it will go via the linear mapping.] > > One approach is just to give up on inline instrumentation. This way all > checks can be delayed until after everything set is up correctly, and the > address-to-shadow calculations can be overridden. However, the features and > speed boost provided by inline instrumentation are worth trying to do > better. > > If _at compile time_ it is known how much contiguous physical memory a > system has, the top 1/8th of the first block of physical memory can be set > aside for the shadow. This is a big hammer and comes with 3 big > consequences: > > - there's no nice way to handle physically discontiguous memory, so only > the first physical memory block can be used. > > - kernels will simply fail to boot on machines with less memory than > specified when compiling. > > - kernels running on machines with more memory than specified when > compiling will simply ignore the extra memory. > > Implement and document KASAN this way. The current implementation is Radix > only. > > Despite the limitations, it can still find bugs, > e.g. http://patchwork.ozlabs.org/patch/1103775/ > > At the moment, this physical memory limit must be set _even for outline > mode_. This may be changed in a later series - a different implementation > could be added for outline mode that dynamically allocates shadow at a > fixed offset. For example, see https://patchwork.ozlabs.org/patch/795211/ > > Suggested-by: Michael Ellerman <mpe@ellerman.id.au> > Cc: Balbir Singh <bsingharora@gmail.com> # ppc64 out-of-line radix version > Cc: Christophe Leroy <christophe.leroy@c-s.fr> # ppc32 version > Signed-off-by: Daniel Axtens <dja@axtens.net> > > --- > Changes since v2: > > - Address feedback from Christophe around cleanups and docs. > - Address feedback from Balbir: at this point I don't have a good solution > for the issues you identify around the limitations of the inline implementation > but I think that it's worth trying to get the stack instrumentation support. > I'm happy to have an alternative and more flexible outline mode - I had > envisoned this would be called 'lightweight' mode as it imposes fewer restrictions. > I've linked to your implementation. I think it's best to add it in a follow-up series. > - Made the default PHYS_MEM_SIZE_FOR_KASAN value 1024MB. I think most people have > guests with at least that much memory in the Radix 64s case so it's a much > saner default - it means that if you just turn on KASAN without reading the > docs you're much more likely to have a bootable kernel, which you will never > have if the value is set to zero! I'm happy to bikeshed the value if we want. > > Changes since v1: > - Landed kasan vmalloc support upstream > - Lots of feedback from Christophe. > > Changes since the rfc: > > - Boots real and virtual hardware, kvm works. > > - disabled reporting when we're checking the stack for exception > frames. The behaviour isn't wrong, just incompatible with KASAN. > > - Documentation! > > - Dropped old module stuff in favour of KASAN_VMALLOC. > > The bugs with ftrace and kuap were due to kernel bloat pushing > prom_init calls to be done via the plt. Because we did not have > a relocatable kernel, and they are done very early, this caused > everything to explode. Compile with CONFIG_RELOCATABLE! > --- > Documentation/dev-tools/kasan.rst | 8 +- > Documentation/powerpc/kasan.txt | 112 +++++++++++++++++- > arch/powerpc/Kconfig | 3 + > arch/powerpc/Kconfig.debug | 21 ++++ > arch/powerpc/Makefile | 11 ++ > arch/powerpc/include/asm/book3s/64/hash.h | 4 + > arch/powerpc/include/asm/book3s/64/pgtable.h | 7 ++ > arch/powerpc/include/asm/book3s/64/radix.h | 5 + > arch/powerpc/include/asm/kasan.h | 21 +++- > arch/powerpc/kernel/process.c | 8 ++ > arch/powerpc/kernel/prom.c | 64 +++++++++- > arch/powerpc/mm/kasan/Makefile | 3 +- > .../mm/kasan/{kasan_init_32.c => init_32.c} | 0 > arch/powerpc/mm/kasan/init_book3s_64.c | 72 +++++++++++ > 14 files changed, 330 insertions(+), 9 deletions(-) > rename arch/powerpc/mm/kasan/{kasan_init_32.c => init_32.c} (100%) > create mode 100644 arch/powerpc/mm/kasan/init_book3s_64.c > > diff --git a/Documentation/dev-tools/kasan.rst b/Documentation/dev-tools/kasan.rst > index 4af2b5d2c9b4..d99dc580bc11 100644 > --- a/Documentation/dev-tools/kasan.rst > +++ b/Documentation/dev-tools/kasan.rst > @@ -22,8 +22,9 @@ global variables yet. > Tag-based KASAN is only supported in Clang and requires version 7.0.0 or later. > > Currently generic KASAN is supported for the x86_64, arm64, xtensa and s390 > -architectures. It is also supported on 32-bit powerpc kernels. Tag-based KASAN > -is supported only on arm64. > +architectures. It is also supported on powerpc, for 32-bit kernels, and for > +64-bit kernels running under the Radix MMU. Tag-based KASAN is supported only > +on arm64. > > Usage > ----- > @@ -256,7 +257,8 @@ CONFIG_KASAN_VMALLOC > ~~~~~~~~~~~~~~~~~~~~ > > With ``CONFIG_KASAN_VMALLOC``, KASAN can cover vmalloc space at the > -cost of greater memory usage. Currently this is only supported on x86. > +cost of greater memory usage. Currently this is optional on x86, and > +required on 64-bit powerpc. > > This works by hooking into vmalloc and vmap, and dynamically > allocating real shadow memory to back the mappings. > diff --git a/Documentation/powerpc/kasan.txt b/Documentation/powerpc/kasan.txt > index a85ce2ff8244..f134a91600ad 100644 > --- a/Documentation/powerpc/kasan.txt > +++ b/Documentation/powerpc/kasan.txt > @@ -1,4 +1,4 @@ > -KASAN is supported on powerpc on 32-bit only. > +KASAN is supported on powerpc on 32-bit and Radix 64-bit only. > > 32 bit support > ============== > @@ -10,3 +10,113 @@ fixmap area and occupies one eighth of the total kernel virtual memory space. > > Instrumentation of the vmalloc area is not currently supported, but modules > are. > + > +64 bit support > +============== > + > +Currently, only the radix MMU is supported. There have been versions for Book3E > +processors floating around on the mailing list, but nothing has been merged. > + > +KASAN support on Book3S is a bit tricky to get right: > + > + - It would be good to support inline instrumentation so as to be able to catch > + stack issues that cannot be caught with outline mode. > + > + - Inline instrumentation requires a fixed offset. > + > + - Book3S runs code in real mode after booting. Most notably a lot of KVM runs > + in real mode, and it would be good to be able to instrument it. > + > + - Because code runs in real mode after boot, the offset has to point to > + valid memory both in and out of real mode. > + > +One approach is just to give up on inline instrumentation. This way all checks > +can be delayed until after everything set is up correctly, and the > +address-to-shadow calculations can be overridden. However, the features and > +speed boost provided by inline instrumentation are worth trying to do better. > + > +If _at compile time_ it is known how much contiguous physical memory a system > +has, the top 1/8th of the first block of physical memory can be set aside for > +the shadow. This is a big hammer and comes with 3 big consequences: > + > + - there's no nice way to handle physically discontiguous memory, so only the > + first physical memory block can be used. > + > + - kernels will simply fail to boot on machines with less memory than specified > + when compiling. > + > + - kernels running on machines with more memory than specified when compiling > + will simply ignore the extra memory. > + > +At the moment, this physical memory limit must be set _even for outline mode_. > +This may be changed in a future version - a different implementation could be > +added for outline mode that dynamically allocates shadow at a fixed offset. > +For example, see https://patchwork.ozlabs.org/patch/795211/ > + > +This value is configured in CONFIG_PHYS_MEM_SIZE_FOR_KASAN. > + > +Tips > +---- > + > + - Compile with CONFIG_RELOCATABLE. > + > + In development, boot hangs were observed when building with ftrace and KUAP > + on. These ended up being due to kernel bloat pushing prom_init calls to be > + done via the PLT. Because the kernel was not relocatable, and the calls are > + done very early, this caused execution to jump off into somewhere > + invalid. Enabling relocation fixes this. > + > +NUMA/discontiguous physical memory > +---------------------------------- > + > +Currently the code cannot really deal with discontiguous physical memory. Only > +physical memory that is contiguous from physical address zero can be used. The > +size of that memory, not total memory, must be specified when configuring the > +kernel. > + > +Discontiguous memory can occur on machines with memory spread across multiple > +nodes. For example, on a Talos II with 64GB of RAM: > + > + - 32GB runs from 0x0 to 0x0000_0008_0000_0000, > + - then there's a gap, > + - then the final 32GB runs from 0x0000_2000_0000_0000 to 0x0000_2008_0000_0000 > + > +This can create _significant_ issues: > + > + - If the machine is treated as having 64GB of _contiguous_ RAM, the > + instrumentation would assume that it ran from 0x0 to > + 0x0000_0010_0000_0000. The last 1/8th - 0x0000_000e_0000_0000 to > + 0x0000_0010_0000_0000 would be reserved as the shadow region. But when the > + kernel tried to access any of that, it would be trying to access pages that > + are not physically present. > + > + - If the shadow region size is based on the top address, then the shadow > + region would be 0x2008_0000_0000 / 8 = 0x0401_0000_0000 bytes = 4100 GB of > + memory, clearly more than the 64GB of RAM physically present. > + > +Therefore, the code currently is restricted to dealing with memory in the node > +starting at 0x0. For this system, that's 32GB. If a contiguous physical memory > +size greater than the size of the first contiguous region of memory is > +specified, the system will be unable to boot or even print an error message. > + > +The layout of a system's memory can be observed in the messages that the Radix > +MMU prints on boot. The Talos II discussed earlier has: > + > +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) > +radix-mmu: Mapped 0x0000000040000000-0x0000000800000000 with 1.00 GiB pages > +radix-mmu: Mapped 0x0000200000000000-0x0000200800000000 with 1.00 GiB pages > + > +As discussed, this system would be configured for 32768 MB. > + > +Another system prints: > + > +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) > +radix-mmu: Mapped 0x0000000040000000-0x0000002000000000 with 1.00 GiB pages > +radix-mmu: Mapped 0x0000200000000000-0x0000202000000000 with 1.00 GiB pages > + > +This machine has more memory: 0x0000_0040_0000_0000 total, but only > +0x0000_0020_0000_0000 is physically contiguous from zero, so it would be > +configured for 131072 MB of physically contiguous memory. > + > +This restriction currently also affects outline mode, but this could be > +changed in future if an alternative outline implementation is added. > diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig > index 6987b0832e5f..2561446e85a8 100644 > --- a/arch/powerpc/Kconfig > +++ b/arch/powerpc/Kconfig > @@ -173,6 +173,9 @@ config PPC > select HAVE_ARCH_HUGE_VMAP if PPC_BOOK3S_64 && PPC_RADIX_MMU > select HAVE_ARCH_JUMP_LABEL > select HAVE_ARCH_KASAN if PPC32 > + select HAVE_ARCH_KASAN if PPC_BOOK3S_64 && PPC_RADIX_MMU > + select HAVE_ARCH_KASAN_VMALLOC if PPC_BOOK3S_64 && PPC_RADIX_MMU > + select KASAN_VMALLOC if KASAN && PPC_BOOK3S_64 This one should go somewhere else, most likely in the 'config PPC_BOOK3S_64' section in Kconfig.cputype Here it is more or less dedicated to capabilities. If you don't want to move it, at least you have to keep alphabetical order in this section. > select HAVE_ARCH_KGDB > select HAVE_ARCH_MMAP_RND_BITS > select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT > diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug > index 4e1d39847462..5c454f8fa24b 100644 > --- a/arch/powerpc/Kconfig.debug > +++ b/arch/powerpc/Kconfig.debug > @@ -394,6 +394,27 @@ config PPC_FAST_ENDIAN_SWITCH > help > If you're unsure what this is, say N. > > +config PHYS_MEM_SIZE_FOR_KASAN > + int "Contiguous physical memory size for KASAN (MB)" if KASAN && PPC_BOOK3S_64 > + default 1024 > + help > + > + To get inline instrumentation support for KASAN on 64-bit Book3S > + machines, you need to know how much contiguous physical memory your > + system has. A shadow offset will be calculated based on this figure, > + which will be compiled in to the kernel. KASAN will use this offset > + to access its shadow region, which is used to verify memory accesses. > + > + If you attempt to boot on a system with less memory than you specify > + here, your system will fail to boot very early in the process. If you > + boot on a system with more memory than you specify, the extra memory > + will wasted - it will be reserved and not used. > + > + For systems with discontiguous blocks of physical memory, specify the > + size of the block starting at 0x0. You can determine this by looking > + at the memory layout info printed to dmesg by the radix MMU code > + early in boot. See Documentation/powerpc/kasan.txt. > + > config KASAN_SHADOW_OFFSET > hex > depends on KASAN > diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile > index f35730548e42..eff693527462 100644 > --- a/arch/powerpc/Makefile > +++ b/arch/powerpc/Makefile > @@ -230,6 +230,17 @@ ifdef CONFIG_476FPE_ERR46 > -T $(srctree)/arch/powerpc/platforms/44x/ppc476_modules.lds > endif > > +ifdef CONFIG_PPC_BOOK3S_64 > +# The KASAN shadow offset is such that linear map (0xc000...) is shadowed by > +# the last 8th of linearly mapped physical memory. This way, if the code uses > +# 0xc addresses throughout, accesses work both in in real mode (where the top > +# 2 bits are ignored) and outside of real mode. > +# > +# 0xc000000000000000 >> 3 = 0xa800000000000000 = 12105675798371893248 > +KASAN_SHADOW_OFFSET = $(shell echo 7 \* 1024 \* 1024 \* $(CONFIG_PHYS_MEM_SIZE_FOR_KASAN) / 8 + 12105675798371893248 | bc) > +KBUILD_CFLAGS += -DKASAN_SHADOW_OFFSET=$(KASAN_SHADOW_OFFSET)UL > +endif > + > # No AltiVec or VSX instructions when building kernel > KBUILD_CFLAGS += $(call cc-option,-mno-altivec) > KBUILD_CFLAGS += $(call cc-option,-mno-vsx) > diff --git a/arch/powerpc/include/asm/book3s/64/hash.h b/arch/powerpc/include/asm/book3s/64/hash.h > index 2781ebf6add4..fce329b8452e 100644 > --- a/arch/powerpc/include/asm/book3s/64/hash.h > +++ b/arch/powerpc/include/asm/book3s/64/hash.h > @@ -18,6 +18,10 @@ > #include <asm/book3s/64/hash-4k.h> > #endif > > +#define H_PTRS_PER_PTE (1 << H_PTE_INDEX_SIZE) > +#define H_PTRS_PER_PMD (1 << H_PMD_INDEX_SIZE) > +#define H_PTRS_PER_PUD (1 << H_PUD_INDEX_SIZE) > + > /* Bits to set in a PMD/PUD/PGD entry valid bit*/ > #define HASH_PMD_VAL_BITS (0x8000000000000000UL) > #define HASH_PUD_VAL_BITS (0x8000000000000000UL) > diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h > index b01624e5c467..209817235a44 100644 > --- a/arch/powerpc/include/asm/book3s/64/pgtable.h > +++ b/arch/powerpc/include/asm/book3s/64/pgtable.h > @@ -231,6 +231,13 @@ extern unsigned long __pmd_frag_size_shift; > #define PTRS_PER_PUD (1 << PUD_INDEX_SIZE) > #define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) > > +#define MAX_PTRS_PER_PTE ((H_PTRS_PER_PTE > R_PTRS_PER_PTE) ? \ > + H_PTRS_PER_PTE : R_PTRS_PER_PTE) > +#define MAX_PTRS_PER_PMD ((H_PTRS_PER_PMD > R_PTRS_PER_PMD) ? \ > + H_PTRS_PER_PMD : R_PTRS_PER_PMD) > +#define MAX_PTRS_PER_PUD ((H_PTRS_PER_PUD > R_PTRS_PER_PUD) ? \ > + H_PTRS_PER_PUD : R_PTRS_PER_PUD) > + > /* PMD_SHIFT determines what a second-level page table entry can map */ > #define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) > #define PMD_SIZE (1UL << PMD_SHIFT) > diff --git a/arch/powerpc/include/asm/book3s/64/radix.h b/arch/powerpc/include/asm/book3s/64/radix.h > index d97db3ad9aae..4f826259de71 100644 > --- a/arch/powerpc/include/asm/book3s/64/radix.h > +++ b/arch/powerpc/include/asm/book3s/64/radix.h > @@ -35,6 +35,11 @@ > #define RADIX_PMD_SHIFT (PAGE_SHIFT + RADIX_PTE_INDEX_SIZE) > #define RADIX_PUD_SHIFT (RADIX_PMD_SHIFT + RADIX_PMD_INDEX_SIZE) > #define RADIX_PGD_SHIFT (RADIX_PUD_SHIFT + RADIX_PUD_INDEX_SIZE) > + > +#define R_PTRS_PER_PTE (1 << RADIX_PTE_INDEX_SIZE) > +#define R_PTRS_PER_PMD (1 << RADIX_PMD_INDEX_SIZE) > +#define R_PTRS_PER_PUD (1 << RADIX_PUD_INDEX_SIZE) > + > /* > * Size of EA range mapped by our pagetables. > */ > diff --git a/arch/powerpc/include/asm/kasan.h b/arch/powerpc/include/asm/kasan.h > index 296e51c2f066..f18268cbdc33 100644 > --- a/arch/powerpc/include/asm/kasan.h > +++ b/arch/powerpc/include/asm/kasan.h > @@ -2,6 +2,9 @@ > #ifndef __ASM_KASAN_H > #define __ASM_KASAN_H > > +#include <asm/page.h> > +#include <asm/pgtable.h> > + > #ifdef CONFIG_KASAN > #define _GLOBAL_KASAN(fn) _GLOBAL(__##fn) > #define _GLOBAL_TOC_KASAN(fn) _GLOBAL_TOC(__##fn) > @@ -14,13 +17,19 @@ > > #ifndef __ASSEMBLY__ > > -#include <asm/page.h> > +#ifdef CONFIG_KASAN > +void kasan_init(void); > +#else > +static inline void kasan_init(void) { } > +#endif > > #define KASAN_SHADOW_SCALE_SHIFT 3 > > #define KASAN_SHADOW_START (KASAN_SHADOW_OFFSET + \ > (PAGE_OFFSET >> KASAN_SHADOW_SCALE_SHIFT)) > > +#ifdef CONFIG_PPC32 > + > #define KASAN_SHADOW_OFFSET ASM_CONST(CONFIG_KASAN_SHADOW_OFFSET) > > #define KASAN_SHADOW_END 0UL > @@ -30,11 +39,17 @@ > #ifdef CONFIG_KASAN > void kasan_early_init(void); > void kasan_mmu_init(void); > -void kasan_init(void); > #else > -static inline void kasan_init(void) { } > static inline void kasan_mmu_init(void) { } > #endif > +#endif > + > +#ifdef CONFIG_PPC_BOOK3S_64 > + > +#define KASAN_SHADOW_SIZE ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN * \ > + 1024 * 1024 * 1 / 8) > + > +#endif /* CONFIG_PPC_BOOK3S_64 */ > > #endif /* __ASSEMBLY */ > #endif > diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c > index 4df94b6e2f32..c60ff299f39b 100644 > --- a/arch/powerpc/kernel/process.c > +++ b/arch/powerpc/kernel/process.c > @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) > /* > * See if this is an exception frame. > * We look for the "regshere" marker in the current frame. > + * > + * KASAN may complain about this. If it is an exception frame, > + * we won't have unpoisoned the stack in asm when we set the > + * exception marker. If it's not an exception frame, who knows > + * how things are laid out - the shadow could be in any state > + * at all. Just disable KASAN reporting for now. > */ > + kasan_disable_current(); > if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) > && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { > struct pt_regs *regs = (struct pt_regs *) > @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) > regs->trap, (void *)regs->nip, (void *)lr); > firstframe = 1; > } > + kasan_enable_current(); If this is really a concern for all targets including PPC32, should it be a separate patch with a Fixes: tag to be applied back in stable as well ? > > sp = newsp; > } while (count++ < kstack_depth_to_print); > diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c > index 6620f37abe73..d994c7c39c8d 100644 > --- a/arch/powerpc/kernel/prom.c > +++ b/arch/powerpc/kernel/prom.c > @@ -72,6 +72,7 @@ unsigned long tce_alloc_start, tce_alloc_end; > u64 ppc64_rma_size; > #endif > static phys_addr_t first_memblock_size; > +static phys_addr_t top_phys_addr; > static int __initdata boot_cpu_count; > > static int __init early_parse_mem(char *p) > @@ -449,6 +450,26 @@ static bool validate_mem_limit(u64 base, u64 *size) > { > u64 max_mem = 1UL << (MAX_PHYSMEM_BITS); > > + /* > + * To handle the NUMA/discontiguous memory case, don't allow a block > + * to be added if it falls completely beyond the configured physical > + * memory. Print an informational message. > + * > + * Frustratingly we also see this with qemu - it seems to split the > + * specified memory into a number of smaller blocks. If this happens > + * under qemu, it probably represents misconfiguration. So we want > + * the message to be noticeable, but not shouty. > + * > + * See Documentation/powerpc/kasan.txt > + */ > + if (IS_ENABLED(CONFIG_KASAN) && > + (base >= ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20))) { > + pr_warn("KASAN: not adding memory block at %llx (size %llx)\n" > + "This could be due to discontiguous memory or kernel misconfiguration.", > + base, *size); > + return false; > + } > + > if (base >= max_mem) > return false; > if ((base + *size) > max_mem) > @@ -572,8 +593,11 @@ void __init early_init_dt_add_memory_arch(u64 base, u64 size) > > /* Add the chunk to the MEMBLOCK list */ > if (add_mem_to_memblock) { > - if (validate_mem_limit(base, &size)) > + if (validate_mem_limit(base, &size)) { > memblock_add(base, size); > + if (base + size > top_phys_addr) > + top_phys_addr = base + size; > + } Can we use max() here ? Something like top_phys_addr = max(base + size, top_phys_addr); > } > } > > @@ -613,6 +637,8 @@ static void __init early_reserve_mem_dt(void) > static void __init early_reserve_mem(void) > { > __be64 *reserve_map; > + phys_addr_t kasan_shadow_start; > + phys_addr_t kasan_memory_size; > > reserve_map = (__be64 *)(((unsigned long)initial_boot_params) + > fdt_off_mem_rsvmap(initial_boot_params)); > @@ -651,6 +677,42 @@ static void __init early_reserve_mem(void) > return; > } > #endif > + > + if (IS_ENABLED(CONFIG_KASAN) && IS_ENABLED(CONFIG_PPC_BOOK3S_64)) { > + kasan_memory_size = > + ((phys_addr_t)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20); > + > + if (top_phys_addr < kasan_memory_size) { > + /* > + * We are doomed. We shouldn't even be able to get this > + * far, but we do in qemu. If we continue and turn > + * relocations on, we'll take fatal page faults for > + * memory that's not physically present. Instead, > + * panic() here: it will be saved to __log_buf even if > + * it doesn't get printed to the console. > + */ > + panic("Tried to book a KASAN kernel configured for %u MB with only %llu MB! Aborting.", book ==> boot ? > + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, > + (u64)(top_phys_addr >> 20)); > + } else if (top_phys_addr > kasan_memory_size) { > + /* print a biiiig warning in hopes people notice */ > + pr_err("===========================================\n" > + "Physical memory exceeds compiled-in maximum!\n" > + "This kernel was compiled for KASAN with %u MB physical memory.\n" > + "The physical memory detected is at least %llu MB.\n" > + "Memory above the compiled limit will not be used!\n" > + "===========================================\n", > + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, > + (u64)(top_phys_addr >> 20)); > + } > + > + kasan_shadow_start = _ALIGN_DOWN(kasan_memory_size * 7 / 8, > + PAGE_SIZE); Can't this fit on a single line ? powerpc allows 90 chars. > + DBG("reserving %llx -> %llx for KASAN", > + kasan_shadow_start, top_phys_addr); > + memblock_reserve(kasan_shadow_start, > + top_phys_addr - kasan_shadow_start); Same ? > + } > } > > #ifdef CONFIG_PPC_TRANSACTIONAL_MEM > diff --git a/arch/powerpc/mm/kasan/Makefile b/arch/powerpc/mm/kasan/Makefile > index 6577897673dd..f02b15c78e4d 100644 > --- a/arch/powerpc/mm/kasan/Makefile > +++ b/arch/powerpc/mm/kasan/Makefile > @@ -2,4 +2,5 @@ > > KASAN_SANITIZE := n > > -obj-$(CONFIG_PPC32) += kasan_init_32.o > +obj-$(CONFIG_PPC32) += init_32.o Shouldn't we do ppc32 name change in another patch ? > +obj-$(CONFIG_PPC_BOOK3S_64) += init_book3s_64.o > diff --git a/arch/powerpc/mm/kasan/kasan_init_32.c b/arch/powerpc/mm/kasan/init_32.c > similarity index 100% > rename from arch/powerpc/mm/kasan/kasan_init_32.c > rename to arch/powerpc/mm/kasan/init_32.c > diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/kasan/init_book3s_64.c > new file mode 100644 > index 000000000000..f961e96be136 > --- /dev/null > +++ b/arch/powerpc/mm/kasan/init_book3s_64.c > @@ -0,0 +1,72 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * KASAN for 64-bit Book3S powerpc > + * > + * Copyright (C) 2019 IBM Corporation > + * Author: Daniel Axtens <dja@axtens.net> > + */ > + > +#define DISABLE_BRANCH_PROFILING > + > +#include <linux/kasan.h> > +#include <linux/printk.h> > +#include <linux/sched/task.h> > +#include <asm/pgalloc.h> > + > +void __init kasan_init(void) > +{ > + int i; > + void *k_start = kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START); > + void *k_end = kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END); > + > + pte_t pte = __pte(__pa(kasan_early_shadow_page) | > + pgprot_val(PAGE_KERNEL) | _PAGE_PTE); Can't we do something with existing helpers ? Something like: pte = pte_mkpte(pfn_pte(virt_to_pfn(kasan_early_shadow_page), PAGE_KERNEL)); > + > + if (!early_radix_enabled()) > + panic("KASAN requires radix!"); > + > + for (i = 0; i < PTRS_PER_PTE; i++) > + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, > + &kasan_early_shadow_pte[i], pte, 0); > + > + for (i = 0; i < PTRS_PER_PMD; i++) > + pmd_populate_kernel(&init_mm, &kasan_early_shadow_pmd[i], > + kasan_early_shadow_pte); > + > + for (i = 0; i < PTRS_PER_PUD; i++) > + pud_populate(&init_mm, &kasan_early_shadow_pud[i], > + kasan_early_shadow_pmd); > + > + memset(kasan_mem_to_shadow((void *)PAGE_OFFSET), KASAN_SHADOW_INIT, > + KASAN_SHADOW_SIZE); > + > + kasan_populate_early_shadow( > + kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START), > + kasan_mem_to_shadow((void *)RADIX_VMALLOC_START)); > + > + /* leave a hole here for vmalloc */ > + > + kasan_populate_early_shadow( > + kasan_mem_to_shadow((void *)RADIX_VMALLOC_END), > + kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END)); > + > + flush_tlb_kernel_range((unsigned long)k_start, (unsigned long)k_end); > + > + /* mark early shadow region as RO and wipe */ > + pte = __pte(__pa(kasan_early_shadow_page) | > + pgprot_val(PAGE_KERNEL_RO) | _PAGE_PTE); Same comment as above, use helpers ? > + for (i = 0; i < PTRS_PER_PTE; i++) > + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, > + &kasan_early_shadow_pte[i], pte, 0); > + > + /* > + * clear_page relies on some cache info that hasn't been set up yet. > + * It ends up looping ~forever and blows up other data. > + * Use memset instead. > + */ > + memset(kasan_early_shadow_page, 0, PAGE_SIZE); > + > + /* Enable error messages */ > + init_task.kasan_depth = 0; > + pr_info("KASAN init done (64-bit Book3S heavyweight mode)\n"); > +} > Christophe
Hi Christophe, I'm working through your feedback, thank you. Regarding this one: >> --- a/arch/powerpc/kernel/process.c >> +++ b/arch/powerpc/kernel/process.c >> @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >> /* >> * See if this is an exception frame. >> * We look for the "regshere" marker in the current frame. >> + * >> + * KASAN may complain about this. If it is an exception frame, >> + * we won't have unpoisoned the stack in asm when we set the >> + * exception marker. If it's not an exception frame, who knows >> + * how things are laid out - the shadow could be in any state >> + * at all. Just disable KASAN reporting for now. >> */ >> + kasan_disable_current(); >> if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) >> && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { >> struct pt_regs *regs = (struct pt_regs *) >> @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >> regs->trap, (void *)regs->nip, (void *)lr); >> firstframe = 1; >> } >> + kasan_enable_current(); > > If this is really a concern for all targets including PPC32, should it > be a separate patch with a Fixes: tag to be applied back in stable as well ? I've managed to repro this by commening out the kasan_disable/enable lines, and just booting in qemu without a disk attached: sudo qemu-system-ppc64 -accel kvm -m 2G -M pseries -cpu power9 -kernel ./vmlinux -nographic -chardev stdio,id=charserial0,mux=on -device spapr-vty,chardev=charserial0,reg=0x30000000 -mon chardev=charserial0,mode=readline -nodefaults -smp 2 ... [ 0.210740] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) [ 0.210789] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 [ 0.210844] Call Trace: [ 0.210866] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) [ 0.210915] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c [ 0.210958] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac [ 0.211005] ================================================================== [ 0.211054] BUG: KASAN: stack-out-of-bounds in show_stack+0x438/0x580 [ 0.211095] Read of size 8 at addr c00000006a483b00 by task swapper/0/1 [ 0.211134] [ 0.211152] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 [ 0.211207] Call Trace: [ 0.211225] [c00000006a483680] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) [ 0.211274] [c00000006a4836d0] [c0000000008f877c] print_address_description.isra.10+0x7c/0x470 [ 0.211330] [c00000006a483760] [c0000000008f8e7c] __kasan_report+0x1bc/0x244 [ 0.211380] [c00000006a483830] [c0000000008f6eb8] kasan_report+0x18/0x30 [ 0.211422] [c00000006a483850] [c0000000008fa5d4] __asan_report_load8_noabort+0x24/0x40 [ 0.211471] [c00000006a483870] [c00000000003d448] show_stack+0x438/0x580 [ 0.211512] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 [ 0.211553] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c [ 0.211595] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac [ 0.211644] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 [ 0.211694] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 [ 0.211745] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c [ 0.211787] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 [ 0.211834] [ 0.211851] Allocated by task 0: [ 0.211878] save_stack+0x2c/0xe0 [ 0.211904] __kasan_kmalloc.isra.16+0x11c/0x150 [ 0.211937] kmem_cache_alloc_node+0x114/0x3b0 [ 0.211971] copy_process+0x5b8/0x6410 [ 0.211996] _do_fork+0x130/0xbf0 [ 0.212022] kernel_thread+0xdc/0x130 [ 0.212047] rest_init+0x44/0x184 [ 0.212072] start_kernel+0x77c/0x7dc [ 0.212098] start_here_common+0x1c/0x20 [ 0.212122] [ 0.212139] Freed by task 0: [ 0.212163] (stack is not available) [ 0.212187] [ 0.212205] The buggy address belongs to the object at c00000006a480000 [ 0.212205] which belongs to the cache thread_stack of size 16384 [ 0.212285] The buggy address is located 15104 bytes inside of [ 0.212285] 16384-byte region [c00000006a480000, c00000006a484000) [ 0.212356] The buggy address belongs to the page: [ 0.212391] page:c00c0000001a9200 refcount:1 mapcount:0 mapping:c00000006a019e00 index:0x0 compound_mapcount: 0 [ 0.212455] raw: 007ffff000010200 5deadbeef0000100 5deadbeef0000122 c00000006a019e00 [ 0.212504] raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000 [ 0.212551] page dumped because: kasan: bad access detected [ 0.212583] [ 0.212600] addr c00000006a483b00 is located in stack of task swapper/0/1 at offset 0 in frame: [ 0.212656] mount_block_root+0x0/0x7ac [ 0.212681] [ 0.212698] this frame has 1 object: [ 0.212722] [32, 64) 'b' [ 0.212723] [ 0.212755] Memory state around the buggy address: [ 0.212788] c00000006a483a00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 0.212836] c00000006a483a80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 0.212884] >c00000006a483b00: f1 f1 f1 f1 00 00 00 00 f3 f3 f3 f3 00 00 00 00 [ 0.212931] ^ [ 0.212957] c00000006a483b80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 0.213005] c00000006a483c00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 0.213052] ================================================================== [ 0.213100] Disabling lock debugging due to kernel taint [ 0.213134] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 [ 0.213182] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 [ 0.213231] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c [ 0.213272] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 Is that something that reproduces on ppc32? I don't see it running the test_kasan tests, so I guess that matches up with your experience. Regards, Daniel > >> >> sp = newsp; >> } while (count++ < kstack_depth_to_print); >> diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c >> index 6620f37abe73..d994c7c39c8d 100644 >> --- a/arch/powerpc/kernel/prom.c >> +++ b/arch/powerpc/kernel/prom.c >> @@ -72,6 +72,7 @@ unsigned long tce_alloc_start, tce_alloc_end; >> u64 ppc64_rma_size; >> #endif >> static phys_addr_t first_memblock_size; >> +static phys_addr_t top_phys_addr; >> static int __initdata boot_cpu_count; >> >> static int __init early_parse_mem(char *p) >> @@ -449,6 +450,26 @@ static bool validate_mem_limit(u64 base, u64 *size) >> { >> u64 max_mem = 1UL << (MAX_PHYSMEM_BITS); >> >> + /* >> + * To handle the NUMA/discontiguous memory case, don't allow a block >> + * to be added if it falls completely beyond the configured physical >> + * memory. Print an informational message. >> + * >> + * Frustratingly we also see this with qemu - it seems to split the >> + * specified memory into a number of smaller blocks. If this happens >> + * under qemu, it probably represents misconfiguration. So we want >> + * the message to be noticeable, but not shouty. >> + * >> + * See Documentation/powerpc/kasan.txt >> + */ >> + if (IS_ENABLED(CONFIG_KASAN) && >> + (base >= ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20))) { >> + pr_warn("KASAN: not adding memory block at %llx (size %llx)\n" >> + "This could be due to discontiguous memory or kernel misconfiguration.", >> + base, *size); >> + return false; >> + } >> + >> if (base >= max_mem) >> return false; >> if ((base + *size) > max_mem) >> @@ -572,8 +593,11 @@ void __init early_init_dt_add_memory_arch(u64 base, u64 size) >> >> /* Add the chunk to the MEMBLOCK list */ >> if (add_mem_to_memblock) { >> - if (validate_mem_limit(base, &size)) >> + if (validate_mem_limit(base, &size)) { >> memblock_add(base, size); >> + if (base + size > top_phys_addr) >> + top_phys_addr = base + size; >> + } > > Can we use max() here ? Something like > > top_phys_addr = max(base + size, top_phys_addr); > >> } >> } >> >> @@ -613,6 +637,8 @@ static void __init early_reserve_mem_dt(void) >> static void __init early_reserve_mem(void) >> { >> __be64 *reserve_map; >> + phys_addr_t kasan_shadow_start; >> + phys_addr_t kasan_memory_size; >> >> reserve_map = (__be64 *)(((unsigned long)initial_boot_params) + >> fdt_off_mem_rsvmap(initial_boot_params)); >> @@ -651,6 +677,42 @@ static void __init early_reserve_mem(void) >> return; >> } >> #endif >> + >> + if (IS_ENABLED(CONFIG_KASAN) && IS_ENABLED(CONFIG_PPC_BOOK3S_64)) { >> + kasan_memory_size = >> + ((phys_addr_t)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20); >> + >> + if (top_phys_addr < kasan_memory_size) { >> + /* >> + * We are doomed. We shouldn't even be able to get this >> + * far, but we do in qemu. If we continue and turn >> + * relocations on, we'll take fatal page faults for >> + * memory that's not physically present. Instead, >> + * panic() here: it will be saved to __log_buf even if >> + * it doesn't get printed to the console. >> + */ >> + panic("Tried to book a KASAN kernel configured for %u MB with only %llu MB! Aborting.", > > book ==> boot ? > >> + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, >> + (u64)(top_phys_addr >> 20)); >> + } else if (top_phys_addr > kasan_memory_size) { >> + /* print a biiiig warning in hopes people notice */ >> + pr_err("===========================================\n" >> + "Physical memory exceeds compiled-in maximum!\n" >> + "This kernel was compiled for KASAN with %u MB physical memory.\n" >> + "The physical memory detected is at least %llu MB.\n" >> + "Memory above the compiled limit will not be used!\n" >> + "===========================================\n", >> + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, >> + (u64)(top_phys_addr >> 20)); >> + } >> + >> + kasan_shadow_start = _ALIGN_DOWN(kasan_memory_size * 7 / 8, >> + PAGE_SIZE); > > Can't this fit on a single line ? powerpc allows 90 chars. > >> + DBG("reserving %llx -> %llx for KASAN", >> + kasan_shadow_start, top_phys_addr); >> + memblock_reserve(kasan_shadow_start, >> + top_phys_addr - kasan_shadow_start); > > Same ? > >> + } >> } >> >> #ifdef CONFIG_PPC_TRANSACTIONAL_MEM >> diff --git a/arch/powerpc/mm/kasan/Makefile b/arch/powerpc/mm/kasan/Makefile >> index 6577897673dd..f02b15c78e4d 100644 >> --- a/arch/powerpc/mm/kasan/Makefile >> +++ b/arch/powerpc/mm/kasan/Makefile >> @@ -2,4 +2,5 @@ >> >> KASAN_SANITIZE := n >> >> -obj-$(CONFIG_PPC32) += kasan_init_32.o >> +obj-$(CONFIG_PPC32) += init_32.o > > Shouldn't we do ppc32 name change in another patch ? > >> +obj-$(CONFIG_PPC_BOOK3S_64) += init_book3s_64.o >> diff --git a/arch/powerpc/mm/kasan/kasan_init_32.c b/arch/powerpc/mm/kasan/init_32.c >> similarity index 100% >> rename from arch/powerpc/mm/kasan/kasan_init_32.c >> rename to arch/powerpc/mm/kasan/init_32.c >> diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/kasan/init_book3s_64.c >> new file mode 100644 >> index 000000000000..f961e96be136 >> --- /dev/null >> +++ b/arch/powerpc/mm/kasan/init_book3s_64.c >> @@ -0,0 +1,72 @@ >> +// SPDX-License-Identifier: GPL-2.0 >> +/* >> + * KASAN for 64-bit Book3S powerpc >> + * >> + * Copyright (C) 2019 IBM Corporation >> + * Author: Daniel Axtens <dja@axtens.net> >> + */ >> + >> +#define DISABLE_BRANCH_PROFILING >> + >> +#include <linux/kasan.h> >> +#include <linux/printk.h> >> +#include <linux/sched/task.h> >> +#include <asm/pgalloc.h> >> + >> +void __init kasan_init(void) >> +{ >> + int i; >> + void *k_start = kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START); >> + void *k_end = kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END); >> + >> + pte_t pte = __pte(__pa(kasan_early_shadow_page) | >> + pgprot_val(PAGE_KERNEL) | _PAGE_PTE); > > Can't we do something with existing helpers ? Something like: > > pte = pte_mkpte(pfn_pte(virt_to_pfn(kasan_early_shadow_page), PAGE_KERNEL)); > >> + >> + if (!early_radix_enabled()) >> + panic("KASAN requires radix!"); >> + >> + for (i = 0; i < PTRS_PER_PTE; i++) >> + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, >> + &kasan_early_shadow_pte[i], pte, 0); >> + >> + for (i = 0; i < PTRS_PER_PMD; i++) >> + pmd_populate_kernel(&init_mm, &kasan_early_shadow_pmd[i], >> + kasan_early_shadow_pte); >> + >> + for (i = 0; i < PTRS_PER_PUD; i++) >> + pud_populate(&init_mm, &kasan_early_shadow_pud[i], >> + kasan_early_shadow_pmd); >> + >> + memset(kasan_mem_to_shadow((void *)PAGE_OFFSET), KASAN_SHADOW_INIT, >> + KASAN_SHADOW_SIZE); >> + >> + kasan_populate_early_shadow( >> + kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START), >> + kasan_mem_to_shadow((void *)RADIX_VMALLOC_START)); >> + >> + /* leave a hole here for vmalloc */ >> + >> + kasan_populate_early_shadow( >> + kasan_mem_to_shadow((void *)RADIX_VMALLOC_END), >> + kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END)); >> + >> + flush_tlb_kernel_range((unsigned long)k_start, (unsigned long)k_end); >> + >> + /* mark early shadow region as RO and wipe */ >> + pte = __pte(__pa(kasan_early_shadow_page) | >> + pgprot_val(PAGE_KERNEL_RO) | _PAGE_PTE); > > Same comment as above, use helpers ? > >> + for (i = 0; i < PTRS_PER_PTE; i++) >> + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, >> + &kasan_early_shadow_pte[i], pte, 0); >> + >> + /* >> + * clear_page relies on some cache info that hasn't been set up yet. >> + * It ends up looping ~forever and blows up other data. >> + * Use memset instead. >> + */ >> + memset(kasan_early_shadow_page, 0, PAGE_SIZE); >> + >> + /* Enable error messages */ >> + init_task.kasan_depth = 0; >> + pr_info("KASAN init done (64-bit Book3S heavyweight mode)\n"); >> +} >> > > Christophe
Daniel Axtens <dja@axtens.net> writes: > Hi Christophe, > > I'm working through your feedback, thank you. Regarding this one: > >>> --- a/arch/powerpc/kernel/process.c >>> +++ b/arch/powerpc/kernel/process.c >>> @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >>> /* >>> * See if this is an exception frame. >>> * We look for the "regshere" marker in the current frame. >>> + * >>> + * KASAN may complain about this. If it is an exception frame, >>> + * we won't have unpoisoned the stack in asm when we set the >>> + * exception marker. If it's not an exception frame, who knows >>> + * how things are laid out - the shadow could be in any state >>> + * at all. Just disable KASAN reporting for now. >>> */ >>> + kasan_disable_current(); >>> if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) >>> && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { >>> struct pt_regs *regs = (struct pt_regs *) >>> @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >>> regs->trap, (void *)regs->nip, (void *)lr); >>> firstframe = 1; >>> } >>> + kasan_enable_current(); >> >> If this is really a concern for all targets including PPC32, should it >> be a separate patch with a Fixes: tag to be applied back in stable as well ? > > I've managed to repro this by commening out the kasan_disable/enable > lines, and just booting in qemu without a disk attached: > > sudo qemu-system-ppc64 -accel kvm -m 2G -M pseries -cpu power9 -kernel ./vmlinux -nographic -chardev stdio,id=charserial0,mux=on -device spapr-vty,chardev=charserial0,reg=0x30000000 -mon chardev=charserial0,mode=readline -nodefaults -smp 2 > > ... > > [ 0.210740] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) > [ 0.210789] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 > [ 0.210844] Call Trace: > [ 0.210866] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) > [ 0.210915] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c > [ 0.210958] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac > [ 0.211005] ================================================================== > [ 0.211054] BUG: KASAN: stack-out-of-bounds in show_stack+0x438/0x580 > [ 0.211095] Read of size 8 at addr c00000006a483b00 by task swapper/0/1 > [ 0.211134] > [ 0.211152] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 > [ 0.211207] Call Trace: > [ 0.211225] [c00000006a483680] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) > [ 0.211274] [c00000006a4836d0] [c0000000008f877c] print_address_description.isra.10+0x7c/0x470 > [ 0.211330] [c00000006a483760] [c0000000008f8e7c] __kasan_report+0x1bc/0x244 > [ 0.211380] [c00000006a483830] [c0000000008f6eb8] kasan_report+0x18/0x30 > [ 0.211422] [c00000006a483850] [c0000000008fa5d4] __asan_report_load8_noabort+0x24/0x40 > [ 0.211471] [c00000006a483870] [c00000000003d448] show_stack+0x438/0x580 > [ 0.211512] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 > [ 0.211553] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c > [ 0.211595] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac > [ 0.211644] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 > [ 0.211694] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 > [ 0.211745] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c > [ 0.211787] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 > [ 0.211834] > [ 0.211851] Allocated by task 0: > [ 0.211878] save_stack+0x2c/0xe0 > [ 0.211904] __kasan_kmalloc.isra.16+0x11c/0x150 > [ 0.211937] kmem_cache_alloc_node+0x114/0x3b0 > [ 0.211971] copy_process+0x5b8/0x6410 > [ 0.211996] _do_fork+0x130/0xbf0 > [ 0.212022] kernel_thread+0xdc/0x130 > [ 0.212047] rest_init+0x44/0x184 > [ 0.212072] start_kernel+0x77c/0x7dc > [ 0.212098] start_here_common+0x1c/0x20 > [ 0.212122] > [ 0.212139] Freed by task 0: > [ 0.212163] (stack is not available) > [ 0.212187] > [ 0.212205] The buggy address belongs to the object at c00000006a480000 > [ 0.212205] which belongs to the cache thread_stack of size 16384 > [ 0.212285] The buggy address is located 15104 bytes inside of > [ 0.212285] 16384-byte region [c00000006a480000, c00000006a484000) > [ 0.212356] The buggy address belongs to the page: > [ 0.212391] page:c00c0000001a9200 refcount:1 mapcount:0 mapping:c00000006a019e00 index:0x0 compound_mapcount: 0 > [ 0.212455] raw: 007ffff000010200 5deadbeef0000100 5deadbeef0000122 c00000006a019e00 > [ 0.212504] raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000 > [ 0.212551] page dumped because: kasan: bad access detected > [ 0.212583] > [ 0.212600] addr c00000006a483b00 is located in stack of task swapper/0/1 at offset 0 in frame: > [ 0.212656] mount_block_root+0x0/0x7ac > [ 0.212681] > [ 0.212698] this frame has 1 object: > [ 0.212722] [32, 64) 'b' > [ 0.212723] > [ 0.212755] Memory state around the buggy address: > [ 0.212788] c00000006a483a00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 > [ 0.212836] c00000006a483a80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 > [ 0.212884] >c00000006a483b00: f1 f1 f1 f1 00 00 00 00 f3 f3 f3 f3 00 00 00 00 > [ 0.212931] ^ > [ 0.212957] c00000006a483b80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 > [ 0.213005] c00000006a483c00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 > [ 0.213052] ================================================================== > [ 0.213100] Disabling lock debugging due to kernel taint > [ 0.213134] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 > [ 0.213182] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 > [ 0.213231] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c > [ 0.213272] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 > > Is that something that reproduces on ppc32? > > I don't see it running the test_kasan tests, so I guess that matches up > with your experience. I've debugged this a bit further. If I put a dump_stack() in kernel_init() right before I call kernel_init_freeable(), I don't see the splat. But if I put a dump_stack() immediately inside kernel_init_freeable() I do see the splat. I wonder if some early init code isn't setting up the stack quite right? I don't see this in walking stacks that contain an interrupt frame, so I think the correct thing is to tear out this code and debug the weird stack frame stuff around kernel_init_freeable in parallel. Thanks for your attention to detail. Regards, Daniel > > Regards, > Daniel > > > >> >>> >>> sp = newsp; >>> } while (count++ < kstack_depth_to_print); >>> diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c >>> index 6620f37abe73..d994c7c39c8d 100644 >>> --- a/arch/powerpc/kernel/prom.c >>> +++ b/arch/powerpc/kernel/prom.c >>> @@ -72,6 +72,7 @@ unsigned long tce_alloc_start, tce_alloc_end; >>> u64 ppc64_rma_size; >>> #endif >>> static phys_addr_t first_memblock_size; >>> +static phys_addr_t top_phys_addr; >>> static int __initdata boot_cpu_count; >>> >>> static int __init early_parse_mem(char *p) >>> @@ -449,6 +450,26 @@ static bool validate_mem_limit(u64 base, u64 *size) >>> { >>> u64 max_mem = 1UL << (MAX_PHYSMEM_BITS); >>> >>> + /* >>> + * To handle the NUMA/discontiguous memory case, don't allow a block >>> + * to be added if it falls completely beyond the configured physical >>> + * memory. Print an informational message. >>> + * >>> + * Frustratingly we also see this with qemu - it seems to split the >>> + * specified memory into a number of smaller blocks. If this happens >>> + * under qemu, it probably represents misconfiguration. So we want >>> + * the message to be noticeable, but not shouty. >>> + * >>> + * See Documentation/powerpc/kasan.txt >>> + */ >>> + if (IS_ENABLED(CONFIG_KASAN) && >>> + (base >= ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20))) { >>> + pr_warn("KASAN: not adding memory block at %llx (size %llx)\n" >>> + "This could be due to discontiguous memory or kernel misconfiguration.", >>> + base, *size); >>> + return false; >>> + } >>> + >>> if (base >= max_mem) >>> return false; >>> if ((base + *size) > max_mem) >>> @@ -572,8 +593,11 @@ void __init early_init_dt_add_memory_arch(u64 base, u64 size) >>> >>> /* Add the chunk to the MEMBLOCK list */ >>> if (add_mem_to_memblock) { >>> - if (validate_mem_limit(base, &size)) >>> + if (validate_mem_limit(base, &size)) { >>> memblock_add(base, size); >>> + if (base + size > top_phys_addr) >>> + top_phys_addr = base + size; >>> + } >> >> Can we use max() here ? Something like >> >> top_phys_addr = max(base + size, top_phys_addr); >> >>> } >>> } >>> >>> @@ -613,6 +637,8 @@ static void __init early_reserve_mem_dt(void) >>> static void __init early_reserve_mem(void) >>> { >>> __be64 *reserve_map; >>> + phys_addr_t kasan_shadow_start; >>> + phys_addr_t kasan_memory_size; >>> >>> reserve_map = (__be64 *)(((unsigned long)initial_boot_params) + >>> fdt_off_mem_rsvmap(initial_boot_params)); >>> @@ -651,6 +677,42 @@ static void __init early_reserve_mem(void) >>> return; >>> } >>> #endif >>> + >>> + if (IS_ENABLED(CONFIG_KASAN) && IS_ENABLED(CONFIG_PPC_BOOK3S_64)) { >>> + kasan_memory_size = >>> + ((phys_addr_t)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20); >>> + >>> + if (top_phys_addr < kasan_memory_size) { >>> + /* >>> + * We are doomed. We shouldn't even be able to get this >>> + * far, but we do in qemu. If we continue and turn >>> + * relocations on, we'll take fatal page faults for >>> + * memory that's not physically present. Instead, >>> + * panic() here: it will be saved to __log_buf even if >>> + * it doesn't get printed to the console. >>> + */ >>> + panic("Tried to book a KASAN kernel configured for %u MB with only %llu MB! Aborting.", >> >> book ==> boot ? >> >>> + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, >>> + (u64)(top_phys_addr >> 20)); >>> + } else if (top_phys_addr > kasan_memory_size) { >>> + /* print a biiiig warning in hopes people notice */ >>> + pr_err("===========================================\n" >>> + "Physical memory exceeds compiled-in maximum!\n" >>> + "This kernel was compiled for KASAN with %u MB physical memory.\n" >>> + "The physical memory detected is at least %llu MB.\n" >>> + "Memory above the compiled limit will not be used!\n" >>> + "===========================================\n", >>> + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, >>> + (u64)(top_phys_addr >> 20)); >>> + } >>> + >>> + kasan_shadow_start = _ALIGN_DOWN(kasan_memory_size * 7 / 8, >>> + PAGE_SIZE); >> >> Can't this fit on a single line ? powerpc allows 90 chars. >> >>> + DBG("reserving %llx -> %llx for KASAN", >>> + kasan_shadow_start, top_phys_addr); >>> + memblock_reserve(kasan_shadow_start, >>> + top_phys_addr - kasan_shadow_start); >> >> Same ? >> >>> + } >>> } >>> >>> #ifdef CONFIG_PPC_TRANSACTIONAL_MEM >>> diff --git a/arch/powerpc/mm/kasan/Makefile b/arch/powerpc/mm/kasan/Makefile >>> index 6577897673dd..f02b15c78e4d 100644 >>> --- a/arch/powerpc/mm/kasan/Makefile >>> +++ b/arch/powerpc/mm/kasan/Makefile >>> @@ -2,4 +2,5 @@ >>> >>> KASAN_SANITIZE := n >>> >>> -obj-$(CONFIG_PPC32) += kasan_init_32.o >>> +obj-$(CONFIG_PPC32) += init_32.o >> >> Shouldn't we do ppc32 name change in another patch ? >> >>> +obj-$(CONFIG_PPC_BOOK3S_64) += init_book3s_64.o >>> diff --git a/arch/powerpc/mm/kasan/kasan_init_32.c b/arch/powerpc/mm/kasan/init_32.c >>> similarity index 100% >>> rename from arch/powerpc/mm/kasan/kasan_init_32.c >>> rename to arch/powerpc/mm/kasan/init_32.c >>> diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/kasan/init_book3s_64.c >>> new file mode 100644 >>> index 000000000000..f961e96be136 >>> --- /dev/null >>> +++ b/arch/powerpc/mm/kasan/init_book3s_64.c >>> @@ -0,0 +1,72 @@ >>> +// SPDX-License-Identifier: GPL-2.0 >>> +/* >>> + * KASAN for 64-bit Book3S powerpc >>> + * >>> + * Copyright (C) 2019 IBM Corporation >>> + * Author: Daniel Axtens <dja@axtens.net> >>> + */ >>> + >>> +#define DISABLE_BRANCH_PROFILING >>> + >>> +#include <linux/kasan.h> >>> +#include <linux/printk.h> >>> +#include <linux/sched/task.h> >>> +#include <asm/pgalloc.h> >>> + >>> +void __init kasan_init(void) >>> +{ >>> + int i; >>> + void *k_start = kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START); >>> + void *k_end = kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END); >>> + >>> + pte_t pte = __pte(__pa(kasan_early_shadow_page) | >>> + pgprot_val(PAGE_KERNEL) | _PAGE_PTE); >> >> Can't we do something with existing helpers ? Something like: >> >> pte = pte_mkpte(pfn_pte(virt_to_pfn(kasan_early_shadow_page), PAGE_KERNEL)); >> >>> + >>> + if (!early_radix_enabled()) >>> + panic("KASAN requires radix!"); >>> + >>> + for (i = 0; i < PTRS_PER_PTE; i++) >>> + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, >>> + &kasan_early_shadow_pte[i], pte, 0); >>> + >>> + for (i = 0; i < PTRS_PER_PMD; i++) >>> + pmd_populate_kernel(&init_mm, &kasan_early_shadow_pmd[i], >>> + kasan_early_shadow_pte); >>> + >>> + for (i = 0; i < PTRS_PER_PUD; i++) >>> + pud_populate(&init_mm, &kasan_early_shadow_pud[i], >>> + kasan_early_shadow_pmd); >>> + >>> + memset(kasan_mem_to_shadow((void *)PAGE_OFFSET), KASAN_SHADOW_INIT, >>> + KASAN_SHADOW_SIZE); >>> + >>> + kasan_populate_early_shadow( >>> + kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START), >>> + kasan_mem_to_shadow((void *)RADIX_VMALLOC_START)); >>> + >>> + /* leave a hole here for vmalloc */ >>> + >>> + kasan_populate_early_shadow( >>> + kasan_mem_to_shadow((void *)RADIX_VMALLOC_END), >>> + kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END)); >>> + >>> + flush_tlb_kernel_range((unsigned long)k_start, (unsigned long)k_end); >>> + >>> + /* mark early shadow region as RO and wipe */ >>> + pte = __pte(__pa(kasan_early_shadow_page) | >>> + pgprot_val(PAGE_KERNEL_RO) | _PAGE_PTE); >> >> Same comment as above, use helpers ? >> >>> + for (i = 0; i < PTRS_PER_PTE; i++) >>> + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, >>> + &kasan_early_shadow_pte[i], pte, 0); >>> + >>> + /* >>> + * clear_page relies on some cache info that hasn't been set up yet. >>> + * It ends up looping ~forever and blows up other data. >>> + * Use memset instead. >>> + */ >>> + memset(kasan_early_shadow_page, 0, PAGE_SIZE); >>> + >>> + /* Enable error messages */ >>> + init_task.kasan_depth = 0; >>> + pr_info("KASAN init done (64-bit Book3S heavyweight mode)\n"); >>> +} >>> >> >> Christophe
>> [For those not immersed in ppc64, in real mode, the top nibble or 2 bits >> (depending on radix/hash mmu) of the address is ignored. The linear >> mapping is placed at 0xc000000000000000. This means that a pointer to >> part of the linear mapping will work both in real mode, where it will be >> interpreted as a physical address of the form 0x000..., and out of real >> mode, where it will go via the linear mapping.] >> > > How does hash or radix mmu mode effect how many bits are ignored in real mode? Bah, you're picking on details that I picked up from random conversations in the office rather than from reading the spec! :P The ISA suggests that real addresses space is limited to at most 64 bits. ISAv3, Book III s5.7: | * Host real address space size is 2^m bytes, m <= 60; | see Note 1. | * Guest real address space size is 2 m bytes, m <= 60; | see Notes 1 and 2. ... | Notes: | 1. The value of m is implementation-dependent (sub- | ject to the maximum given above). When used to | address storage or to represent a guest real | address, the high-order 60-m bits of the “60-bit” | real address must be zeros. | 2. The hypervisor may assign a guest real address | space size for each partition that uses Radix Tree | translation. Accesses to guest real storage out- | side this range but still mappable by the second | level Radix Tree will cause an HISI or HDSI. | Accesses to storage outside the mappable range | will have boundedly undefined results. However, it doesn't follow from that passage that the top 4 bits are always ignored when translations are off ('real mode'): see for example the discussion of the HRMOR in s 5.7.3 and s 5.7.3.1. I think I got the 'top 2 bits on radix' thing from the discussion of 'quadrants' in arch/powerpc/include/asm/book3s/64/radix.h, which in turn is discussed in s 5.7.5.1. Table 20 in particular is really helpful for understanding it. But it's not especially relevant to what I'm actually doing here. I think to fully understand all of what's going on I would need to spend some serious time with the entirety of s5.7, because there a lot of quirks about how storage works! But I think for our purposes it suffices to say: The kernel installs a linear mapping at effective address c000... onward. This is a one-to-one mapping with physical memory from 0000... onward. Because of how memory accesses work on powerpc 64-bit Book3S, a kernel pointer in the linear map accesses the same memory both with translations on (accessing as an 'effective address'), and with translations off (accessing as a 'real address'). This works in both guests and the hypervisor. For more details, see s5.7 of Book III of version 3 of the ISA, in particular the Storage Control Overview, s5.7.3, and s5.7.5 - noting that this KASAN implementation currently only supports Radix. Thanks for your attention to detail! Regards, Daniel
On 12/18/2019 04:32 AM, Daniel Axtens wrote: > Daniel Axtens <dja@axtens.net> writes: > >> Hi Christophe, >> >> I'm working through your feedback, thank you. Regarding this one: >> >>>> --- a/arch/powerpc/kernel/process.c >>>> +++ b/arch/powerpc/kernel/process.c >>>> @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >>>> /* >>>> * See if this is an exception frame. >>>> * We look for the "regshere" marker in the current frame. >>>> + * >>>> + * KASAN may complain about this. If it is an exception frame, >>>> + * we won't have unpoisoned the stack in asm when we set the >>>> + * exception marker. If it's not an exception frame, who knows >>>> + * how things are laid out - the shadow could be in any state >>>> + * at all. Just disable KASAN reporting for now. >>>> */ >>>> + kasan_disable_current(); >>>> if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) >>>> && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { >>>> struct pt_regs *regs = (struct pt_regs *) >>>> @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) >>>> regs->trap, (void *)regs->nip, (void *)lr); >>>> firstframe = 1; >>>> } >>>> + kasan_enable_current(); >>> >>> If this is really a concern for all targets including PPC32, should it >>> be a separate patch with a Fixes: tag to be applied back in stable as well ? >> >> I've managed to repro this by commening out the kasan_disable/enable >> lines, and just booting in qemu without a disk attached: >> >> sudo qemu-system-ppc64 -accel kvm -m 2G -M pseries -cpu power9 -kernel ./vmlinux -nographic -chardev stdio,id=charserial0,mux=on -device spapr-vty,chardev=charserial0,reg=0x30000000 -mon chardev=charserial0,mode=readline -nodefaults -smp 2 >> >> ... >> >> [ 0.210740] Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) >> [ 0.210789] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 >> [ 0.210844] Call Trace: >> [ 0.210866] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) >> [ 0.210915] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c >> [ 0.210958] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac >> [ 0.211005] ================================================================== >> [ 0.211054] BUG: KASAN: stack-out-of-bounds in show_stack+0x438/0x580 >> [ 0.211095] Read of size 8 at addr c00000006a483b00 by task swapper/0/1 >> [ 0.211134] >> [ 0.211152] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.5.0-rc1-next-20191213-16824-g469a24fbdb34 #12 >> [ 0.211207] Call Trace: >> [ 0.211225] [c00000006a483680] [c000000001f74f48] dump_stack+0xfc/0x154 (unreliable) >> [ 0.211274] [c00000006a4836d0] [c0000000008f877c] print_address_description.isra.10+0x7c/0x470 >> [ 0.211330] [c00000006a483760] [c0000000008f8e7c] __kasan_report+0x1bc/0x244 >> [ 0.211380] [c00000006a483830] [c0000000008f6eb8] kasan_report+0x18/0x30 >> [ 0.211422] [c00000006a483850] [c0000000008fa5d4] __asan_report_load8_noabort+0x24/0x40 >> [ 0.211471] [c00000006a483870] [c00000000003d448] show_stack+0x438/0x580 >> [ 0.211512] [c00000006a4839b0] [c000000001f74f48] dump_stack+0xfc/0x154 >> [ 0.211553] [c00000006a483a00] [c00000000025411c] panic+0x258/0x59c >> [ 0.211595] [c00000006a483aa0] [c0000000024870b0] mount_block_root+0x648/0x7ac >> [ 0.211644] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 >> [ 0.211694] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 >> [ 0.211745] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c >> [ 0.211787] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 >> [ 0.211834] >> [ 0.211851] Allocated by task 0: >> [ 0.211878] save_stack+0x2c/0xe0 >> [ 0.211904] __kasan_kmalloc.isra.16+0x11c/0x150 >> [ 0.211937] kmem_cache_alloc_node+0x114/0x3b0 >> [ 0.211971] copy_process+0x5b8/0x6410 >> [ 0.211996] _do_fork+0x130/0xbf0 >> [ 0.212022] kernel_thread+0xdc/0x130 >> [ 0.212047] rest_init+0x44/0x184 >> [ 0.212072] start_kernel+0x77c/0x7dc >> [ 0.212098] start_here_common+0x1c/0x20 >> [ 0.212122] >> [ 0.212139] Freed by task 0: >> [ 0.212163] (stack is not available) >> [ 0.212187] >> [ 0.212205] The buggy address belongs to the object at c00000006a480000 >> [ 0.212205] which belongs to the cache thread_stack of size 16384 >> [ 0.212285] The buggy address is located 15104 bytes inside of >> [ 0.212285] 16384-byte region [c00000006a480000, c00000006a484000) >> [ 0.212356] The buggy address belongs to the page: >> [ 0.212391] page:c00c0000001a9200 refcount:1 mapcount:0 mapping:c00000006a019e00 index:0x0 compound_mapcount: 0 >> [ 0.212455] raw: 007ffff000010200 5deadbeef0000100 5deadbeef0000122 c00000006a019e00 >> [ 0.212504] raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000 >> [ 0.212551] page dumped because: kasan: bad access detected >> [ 0.212583] >> [ 0.212600] addr c00000006a483b00 is located in stack of task swapper/0/1 at offset 0 in frame: >> [ 0.212656] mount_block_root+0x0/0x7ac >> [ 0.212681] >> [ 0.212698] this frame has 1 object: >> [ 0.212722] [32, 64) 'b' >> [ 0.212723] >> [ 0.212755] Memory state around the buggy address: >> [ 0.212788] c00000006a483a00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 >> [ 0.212836] c00000006a483a80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 >> [ 0.212884] >c00000006a483b00: f1 f1 f1 f1 00 00 00 00 f3 f3 f3 f3 00 00 00 00 >> [ 0.212931] ^ >> [ 0.212957] c00000006a483b80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 >> [ 0.213005] c00000006a483c00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 >> [ 0.213052] ================================================================== >> [ 0.213100] Disabling lock debugging due to kernel taint >> [ 0.213134] [c00000006a483be0] [c000000002487784] prepare_namespace+0x1ec/0x240 >> [ 0.213182] [c00000006a483c60] [c00000000248669c] kernel_init_freeable+0x7f4/0x870 >> [ 0.213231] [c00000006a483da0] [c000000000011f30] kernel_init+0x3c/0x15c >> [ 0.213272] [c00000006a483e20] [c00000000000bebc] ret_from_kernel_thread+0x5c/0x80 >> >> Is that something that reproduces on ppc32? >> >> I don't see it running the test_kasan tests, so I guess that matches up >> with your experience. > > I've debugged this a bit further. If I put a dump_stack() in > kernel_init() right before I call kernel_init_freeable(), I don't see > the splat. But if I put a dump_stack() immediately inside > kernel_init_freeable() I do see the splat. I wonder if some early init > code isn't setting up the stack quite right? > > I don't see this in walking stacks that contain an interrupt frame, so I > think the correct thing is to tear out this code and debug the weird > stack frame stuff around kernel_init_freeable in parallel. > > Thanks for your attention to detail. > I added a dump_stack() at the start of kernel_init_freeable() and I get nothing more than what follows: [ 0.000000] Activating Kernel Userspace Execution Prevention [ 0.000000] Activating Kernel Userspace Access Protection [ 0.000000] Linux version 5.5.0-rc2-s3k-dev-00932-gf5a548a2b0bc-dirty (root@po16098vm.idsi0.si.c-s.fr) (gcc version 5.5.0 (GCC)) #2596 PREEMPT Wed Dec 18 09:05:02 UTC 2019 [ 0.000000] KASAN init done [ 0.000000] Using CMPC885 machine description [ 0.000000] ----------------------------------------------------- [ 0.000000] phys_mem_size = 0x8000000 [ 0.000000] dcache_bsize = 0x10 [ 0.000000] icache_bsize = 0x10 [ 0.000000] cpu_features = 0x0000000000000100 [ 0.000000] possible = 0x0000000000000120 [ 0.000000] always = 0x0000000000000000 [ 0.000000] cpu_user_features = 0x84000000 0x00000000 [ 0.000000] mmu_features = 0x00000002 [ 0.000000] ----------------------------------------------------- [ 0.000000] SMC microcode patch installed [ 0.000000] Top of RAM: 0x8000000, Total RAM: 0x8000000 [ 0.000000] Memory hole size: 0MB [ 0.000000] Zone ranges: [ 0.000000] Normal [mem 0x0000000000000000-0x0000000007ffffff] [ 0.000000] Movable zone start for each node [ 0.000000] Early memory node ranges [ 0.000000] node 0: [mem 0x0000000000000000-0x0000000007ffffff] [ 0.000000] Initmem setup node 0 [mem 0x0000000000000000-0x0000000007ffffff] [ 0.000000] On node 0 totalpages: 8192 [ 0.000000] Normal zone: 16 pages used for memmap [ 0.000000] Normal zone: 0 pages reserved [ 0.000000] Normal zone: 8192 pages, LIFO batch:0 [ 0.000000] MMU: Allocated 76 bytes of context maps for 16 contexts [ 0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768 [ 0.000000] pcpu-alloc: [0] 0 [ 0.000000] Built 1 zonelists, mobility grouping on. Total pages: 8176 [ 0.000000] Kernel command line: console=ttyCPM0,115200N8 ip=192.168.0.3:192.168.0.1::255.0.0.0:vgoip:eth0:off [ 0.000000] Dentry cache hash table entries: 16384 (order: 2, 65536 bytes, linear) [ 0.000000] Inode-cache hash table entries: 8192 (order: 1, 32768 bytes, linear) [ 0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off [ 0.000000] Memory: 95520K/131072K available (8640K kernel code, 1840K rwdata, 2720K rodata, 656K init, 4679K bss, 35552K reserved, 0K cma-reserved) [ 0.000000] Kernel virtual memory layout: [ 0.000000] * 0xf8000000..0x00000000 : kasan shadow mem [ 0.000000] * 0xf7afc000..0xf7ffc000 : fixmap [ 0.000000] * 0xc9000000..0xf7afc000 : vmalloc & ioremap [ 0.000000] SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=1 [ 0.000000] rcu: Preemptible hierarchical RCU implementation. [ 0.000000] Tasks RCU enabled. [ 0.000000] rcu: RCU calculated value of scheduler-enlistment delay is 10 jiffies. [ 0.000000] NR_IRQS: 512, nr_irqs: 512, preallocated irqs: 16 [ 0.000000] Decrementer Frequency = 0x7de290 [ 0.000000] time_init: decrementer frequency = 8.250000 MHz [ 0.000000] time_init: processor frequency = 132.000000 MHz [ 0.000153] clocksource: timebase: mask: 0xffffffffffffffff max_cycles: 0x1e717863c, max_idle_ns: 440795202213 ns [ 0.000336] clocksource: timebase mult[79364d93] shift[24] registered [ 0.000569] clockevent: decrementer mult[21cac08] shift[32] cpu[0] [ 0.246805] printk: console [ttyCPM0] enabled [ 0.251462] pid_max: default: 32768 minimum: 301 [ 0.259445] Mount-cache hash table entries: 4096 (order: 0, 16384 bytes, linear) [ 0.267030] Mountpoint-cache hash table entries: 4096 (order: 0, 16384 bytes, linear) [ 0.295520] CPU: 0 PID: 1 Comm: swapper Not tainted 5.5.0-rc2-s3k-dev-00932-gf5a548a2b0bc-dirty #2596 [ 0.304710] Call Trace: [ 0.307115] [c5121ed8] [c0b19290] kernel_init_freeable+0x20/0x240 (unreliable) [ 0.314283] [c5121f18] [c0003ddc] kernel_init+0x18/0x10c [ 0.319728] [c5121f38] [c00121cc] ret_from_kernel_thread+0x14/0x1c [ 0.337686] rcu: Hierarchical SRCU implementation. [ 0.350228] devtmpfs: initialized [ 0.533520] device: 'platform': device_add [ 0.535188] bus: 'platform': registered ... Christophe
diff --git a/Documentation/dev-tools/kasan.rst b/Documentation/dev-tools/kasan.rst index 4af2b5d2c9b4..d99dc580bc11 100644 --- a/Documentation/dev-tools/kasan.rst +++ b/Documentation/dev-tools/kasan.rst @@ -22,8 +22,9 @@ global variables yet. Tag-based KASAN is only supported in Clang and requires version 7.0.0 or later. Currently generic KASAN is supported for the x86_64, arm64, xtensa and s390 -architectures. It is also supported on 32-bit powerpc kernels. Tag-based KASAN -is supported only on arm64. +architectures. It is also supported on powerpc, for 32-bit kernels, and for +64-bit kernels running under the Radix MMU. Tag-based KASAN is supported only +on arm64. Usage ----- @@ -256,7 +257,8 @@ CONFIG_KASAN_VMALLOC ~~~~~~~~~~~~~~~~~~~~ With ``CONFIG_KASAN_VMALLOC``, KASAN can cover vmalloc space at the -cost of greater memory usage. Currently this is only supported on x86. +cost of greater memory usage. Currently this is optional on x86, and +required on 64-bit powerpc. This works by hooking into vmalloc and vmap, and dynamically allocating real shadow memory to back the mappings. diff --git a/Documentation/powerpc/kasan.txt b/Documentation/powerpc/kasan.txt index a85ce2ff8244..f134a91600ad 100644 --- a/Documentation/powerpc/kasan.txt +++ b/Documentation/powerpc/kasan.txt @@ -1,4 +1,4 @@ -KASAN is supported on powerpc on 32-bit only. +KASAN is supported on powerpc on 32-bit and Radix 64-bit only. 32 bit support ============== @@ -10,3 +10,113 @@ fixmap area and occupies one eighth of the total kernel virtual memory space. Instrumentation of the vmalloc area is not currently supported, but modules are. + +64 bit support +============== + +Currently, only the radix MMU is supported. There have been versions for Book3E +processors floating around on the mailing list, but nothing has been merged. + +KASAN support on Book3S is a bit tricky to get right: + + - It would be good to support inline instrumentation so as to be able to catch + stack issues that cannot be caught with outline mode. + + - Inline instrumentation requires a fixed offset. + + - Book3S runs code in real mode after booting. Most notably a lot of KVM runs + in real mode, and it would be good to be able to instrument it. + + - Because code runs in real mode after boot, the offset has to point to + valid memory both in and out of real mode. + +One approach is just to give up on inline instrumentation. This way all checks +can be delayed until after everything set is up correctly, and the +address-to-shadow calculations can be overridden. However, the features and +speed boost provided by inline instrumentation are worth trying to do better. + +If _at compile time_ it is known how much contiguous physical memory a system +has, the top 1/8th of the first block of physical memory can be set aside for +the shadow. This is a big hammer and comes with 3 big consequences: + + - there's no nice way to handle physically discontiguous memory, so only the + first physical memory block can be used. + + - kernels will simply fail to boot on machines with less memory than specified + when compiling. + + - kernels running on machines with more memory than specified when compiling + will simply ignore the extra memory. + +At the moment, this physical memory limit must be set _even for outline mode_. +This may be changed in a future version - a different implementation could be +added for outline mode that dynamically allocates shadow at a fixed offset. +For example, see https://patchwork.ozlabs.org/patch/795211/ + +This value is configured in CONFIG_PHYS_MEM_SIZE_FOR_KASAN. + +Tips +---- + + - Compile with CONFIG_RELOCATABLE. + + In development, boot hangs were observed when building with ftrace and KUAP + on. These ended up being due to kernel bloat pushing prom_init calls to be + done via the PLT. Because the kernel was not relocatable, and the calls are + done very early, this caused execution to jump off into somewhere + invalid. Enabling relocation fixes this. + +NUMA/discontiguous physical memory +---------------------------------- + +Currently the code cannot really deal with discontiguous physical memory. Only +physical memory that is contiguous from physical address zero can be used. The +size of that memory, not total memory, must be specified when configuring the +kernel. + +Discontiguous memory can occur on machines with memory spread across multiple +nodes. For example, on a Talos II with 64GB of RAM: + + - 32GB runs from 0x0 to 0x0000_0008_0000_0000, + - then there's a gap, + - then the final 32GB runs from 0x0000_2000_0000_0000 to 0x0000_2008_0000_0000 + +This can create _significant_ issues: + + - If the machine is treated as having 64GB of _contiguous_ RAM, the + instrumentation would assume that it ran from 0x0 to + 0x0000_0010_0000_0000. The last 1/8th - 0x0000_000e_0000_0000 to + 0x0000_0010_0000_0000 would be reserved as the shadow region. But when the + kernel tried to access any of that, it would be trying to access pages that + are not physically present. + + - If the shadow region size is based on the top address, then the shadow + region would be 0x2008_0000_0000 / 8 = 0x0401_0000_0000 bytes = 4100 GB of + memory, clearly more than the 64GB of RAM physically present. + +Therefore, the code currently is restricted to dealing with memory in the node +starting at 0x0. For this system, that's 32GB. If a contiguous physical memory +size greater than the size of the first contiguous region of memory is +specified, the system will be unable to boot or even print an error message. + +The layout of a system's memory can be observed in the messages that the Radix +MMU prints on boot. The Talos II discussed earlier has: + +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) +radix-mmu: Mapped 0x0000000040000000-0x0000000800000000 with 1.00 GiB pages +radix-mmu: Mapped 0x0000200000000000-0x0000200800000000 with 1.00 GiB pages + +As discussed, this system would be configured for 32768 MB. + +Another system prints: + +radix-mmu: Mapped 0x0000000000000000-0x0000000040000000 with 1.00 GiB pages (exec) +radix-mmu: Mapped 0x0000000040000000-0x0000002000000000 with 1.00 GiB pages +radix-mmu: Mapped 0x0000200000000000-0x0000202000000000 with 1.00 GiB pages + +This machine has more memory: 0x0000_0040_0000_0000 total, but only +0x0000_0020_0000_0000 is physically contiguous from zero, so it would be +configured for 131072 MB of physically contiguous memory. + +This restriction currently also affects outline mode, but this could be +changed in future if an alternative outline implementation is added. diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index 6987b0832e5f..2561446e85a8 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -173,6 +173,9 @@ config PPC select HAVE_ARCH_HUGE_VMAP if PPC_BOOK3S_64 && PPC_RADIX_MMU select HAVE_ARCH_JUMP_LABEL select HAVE_ARCH_KASAN if PPC32 + select HAVE_ARCH_KASAN if PPC_BOOK3S_64 && PPC_RADIX_MMU + select HAVE_ARCH_KASAN_VMALLOC if PPC_BOOK3S_64 && PPC_RADIX_MMU + select KASAN_VMALLOC if KASAN && PPC_BOOK3S_64 select HAVE_ARCH_KGDB select HAVE_ARCH_MMAP_RND_BITS select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT diff --git a/arch/powerpc/Kconfig.debug b/arch/powerpc/Kconfig.debug index 4e1d39847462..5c454f8fa24b 100644 --- a/arch/powerpc/Kconfig.debug +++ b/arch/powerpc/Kconfig.debug @@ -394,6 +394,27 @@ config PPC_FAST_ENDIAN_SWITCH help If you're unsure what this is, say N. +config PHYS_MEM_SIZE_FOR_KASAN + int "Contiguous physical memory size for KASAN (MB)" if KASAN && PPC_BOOK3S_64 + default 1024 + help + + To get inline instrumentation support for KASAN on 64-bit Book3S + machines, you need to know how much contiguous physical memory your + system has. A shadow offset will be calculated based on this figure, + which will be compiled in to the kernel. KASAN will use this offset + to access its shadow region, which is used to verify memory accesses. + + If you attempt to boot on a system with less memory than you specify + here, your system will fail to boot very early in the process. If you + boot on a system with more memory than you specify, the extra memory + will wasted - it will be reserved and not used. + + For systems with discontiguous blocks of physical memory, specify the + size of the block starting at 0x0. You can determine this by looking + at the memory layout info printed to dmesg by the radix MMU code + early in boot. See Documentation/powerpc/kasan.txt. + config KASAN_SHADOW_OFFSET hex depends on KASAN diff --git a/arch/powerpc/Makefile b/arch/powerpc/Makefile index f35730548e42..eff693527462 100644 --- a/arch/powerpc/Makefile +++ b/arch/powerpc/Makefile @@ -230,6 +230,17 @@ ifdef CONFIG_476FPE_ERR46 -T $(srctree)/arch/powerpc/platforms/44x/ppc476_modules.lds endif +ifdef CONFIG_PPC_BOOK3S_64 +# The KASAN shadow offset is such that linear map (0xc000...) is shadowed by +# the last 8th of linearly mapped physical memory. This way, if the code uses +# 0xc addresses throughout, accesses work both in in real mode (where the top +# 2 bits are ignored) and outside of real mode. +# +# 0xc000000000000000 >> 3 = 0xa800000000000000 = 12105675798371893248 +KASAN_SHADOW_OFFSET = $(shell echo 7 \* 1024 \* 1024 \* $(CONFIG_PHYS_MEM_SIZE_FOR_KASAN) / 8 + 12105675798371893248 | bc) +KBUILD_CFLAGS += -DKASAN_SHADOW_OFFSET=$(KASAN_SHADOW_OFFSET)UL +endif + # No AltiVec or VSX instructions when building kernel KBUILD_CFLAGS += $(call cc-option,-mno-altivec) KBUILD_CFLAGS += $(call cc-option,-mno-vsx) diff --git a/arch/powerpc/include/asm/book3s/64/hash.h b/arch/powerpc/include/asm/book3s/64/hash.h index 2781ebf6add4..fce329b8452e 100644 --- a/arch/powerpc/include/asm/book3s/64/hash.h +++ b/arch/powerpc/include/asm/book3s/64/hash.h @@ -18,6 +18,10 @@ #include <asm/book3s/64/hash-4k.h> #endif +#define H_PTRS_PER_PTE (1 << H_PTE_INDEX_SIZE) +#define H_PTRS_PER_PMD (1 << H_PMD_INDEX_SIZE) +#define H_PTRS_PER_PUD (1 << H_PUD_INDEX_SIZE) + /* Bits to set in a PMD/PUD/PGD entry valid bit*/ #define HASH_PMD_VAL_BITS (0x8000000000000000UL) #define HASH_PUD_VAL_BITS (0x8000000000000000UL) diff --git a/arch/powerpc/include/asm/book3s/64/pgtable.h b/arch/powerpc/include/asm/book3s/64/pgtable.h index b01624e5c467..209817235a44 100644 --- a/arch/powerpc/include/asm/book3s/64/pgtable.h +++ b/arch/powerpc/include/asm/book3s/64/pgtable.h @@ -231,6 +231,13 @@ extern unsigned long __pmd_frag_size_shift; #define PTRS_PER_PUD (1 << PUD_INDEX_SIZE) #define PTRS_PER_PGD (1 << PGD_INDEX_SIZE) +#define MAX_PTRS_PER_PTE ((H_PTRS_PER_PTE > R_PTRS_PER_PTE) ? \ + H_PTRS_PER_PTE : R_PTRS_PER_PTE) +#define MAX_PTRS_PER_PMD ((H_PTRS_PER_PMD > R_PTRS_PER_PMD) ? \ + H_PTRS_PER_PMD : R_PTRS_PER_PMD) +#define MAX_PTRS_PER_PUD ((H_PTRS_PER_PUD > R_PTRS_PER_PUD) ? \ + H_PTRS_PER_PUD : R_PTRS_PER_PUD) + /* PMD_SHIFT determines what a second-level page table entry can map */ #define PMD_SHIFT (PAGE_SHIFT + PTE_INDEX_SIZE) #define PMD_SIZE (1UL << PMD_SHIFT) diff --git a/arch/powerpc/include/asm/book3s/64/radix.h b/arch/powerpc/include/asm/book3s/64/radix.h index d97db3ad9aae..4f826259de71 100644 --- a/arch/powerpc/include/asm/book3s/64/radix.h +++ b/arch/powerpc/include/asm/book3s/64/radix.h @@ -35,6 +35,11 @@ #define RADIX_PMD_SHIFT (PAGE_SHIFT + RADIX_PTE_INDEX_SIZE) #define RADIX_PUD_SHIFT (RADIX_PMD_SHIFT + RADIX_PMD_INDEX_SIZE) #define RADIX_PGD_SHIFT (RADIX_PUD_SHIFT + RADIX_PUD_INDEX_SIZE) + +#define R_PTRS_PER_PTE (1 << RADIX_PTE_INDEX_SIZE) +#define R_PTRS_PER_PMD (1 << RADIX_PMD_INDEX_SIZE) +#define R_PTRS_PER_PUD (1 << RADIX_PUD_INDEX_SIZE) + /* * Size of EA range mapped by our pagetables. */ diff --git a/arch/powerpc/include/asm/kasan.h b/arch/powerpc/include/asm/kasan.h index 296e51c2f066..f18268cbdc33 100644 --- a/arch/powerpc/include/asm/kasan.h +++ b/arch/powerpc/include/asm/kasan.h @@ -2,6 +2,9 @@ #ifndef __ASM_KASAN_H #define __ASM_KASAN_H +#include <asm/page.h> +#include <asm/pgtable.h> + #ifdef CONFIG_KASAN #define _GLOBAL_KASAN(fn) _GLOBAL(__##fn) #define _GLOBAL_TOC_KASAN(fn) _GLOBAL_TOC(__##fn) @@ -14,13 +17,19 @@ #ifndef __ASSEMBLY__ -#include <asm/page.h> +#ifdef CONFIG_KASAN +void kasan_init(void); +#else +static inline void kasan_init(void) { } +#endif #define KASAN_SHADOW_SCALE_SHIFT 3 #define KASAN_SHADOW_START (KASAN_SHADOW_OFFSET + \ (PAGE_OFFSET >> KASAN_SHADOW_SCALE_SHIFT)) +#ifdef CONFIG_PPC32 + #define KASAN_SHADOW_OFFSET ASM_CONST(CONFIG_KASAN_SHADOW_OFFSET) #define KASAN_SHADOW_END 0UL @@ -30,11 +39,17 @@ #ifdef CONFIG_KASAN void kasan_early_init(void); void kasan_mmu_init(void); -void kasan_init(void); #else -static inline void kasan_init(void) { } static inline void kasan_mmu_init(void) { } #endif +#endif + +#ifdef CONFIG_PPC_BOOK3S_64 + +#define KASAN_SHADOW_SIZE ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN * \ + 1024 * 1024 * 1 / 8) + +#endif /* CONFIG_PPC_BOOK3S_64 */ #endif /* __ASSEMBLY */ #endif diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c index 4df94b6e2f32..c60ff299f39b 100644 --- a/arch/powerpc/kernel/process.c +++ b/arch/powerpc/kernel/process.c @@ -2081,7 +2081,14 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) /* * See if this is an exception frame. * We look for the "regshere" marker in the current frame. + * + * KASAN may complain about this. If it is an exception frame, + * we won't have unpoisoned the stack in asm when we set the + * exception marker. If it's not an exception frame, who knows + * how things are laid out - the shadow could be in any state + * at all. Just disable KASAN reporting for now. */ + kasan_disable_current(); if (validate_sp(sp, tsk, STACK_INT_FRAME_SIZE) && stack[STACK_FRAME_MARKER] == STACK_FRAME_REGS_MARKER) { struct pt_regs *regs = (struct pt_regs *) @@ -2091,6 +2098,7 @@ void show_stack(struct task_struct *tsk, unsigned long *stack) regs->trap, (void *)regs->nip, (void *)lr); firstframe = 1; } + kasan_enable_current(); sp = newsp; } while (count++ < kstack_depth_to_print); diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c index 6620f37abe73..d994c7c39c8d 100644 --- a/arch/powerpc/kernel/prom.c +++ b/arch/powerpc/kernel/prom.c @@ -72,6 +72,7 @@ unsigned long tce_alloc_start, tce_alloc_end; u64 ppc64_rma_size; #endif static phys_addr_t first_memblock_size; +static phys_addr_t top_phys_addr; static int __initdata boot_cpu_count; static int __init early_parse_mem(char *p) @@ -449,6 +450,26 @@ static bool validate_mem_limit(u64 base, u64 *size) { u64 max_mem = 1UL << (MAX_PHYSMEM_BITS); + /* + * To handle the NUMA/discontiguous memory case, don't allow a block + * to be added if it falls completely beyond the configured physical + * memory. Print an informational message. + * + * Frustratingly we also see this with qemu - it seems to split the + * specified memory into a number of smaller blocks. If this happens + * under qemu, it probably represents misconfiguration. So we want + * the message to be noticeable, but not shouty. + * + * See Documentation/powerpc/kasan.txt + */ + if (IS_ENABLED(CONFIG_KASAN) && + (base >= ((u64)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20))) { + pr_warn("KASAN: not adding memory block at %llx (size %llx)\n" + "This could be due to discontiguous memory or kernel misconfiguration.", + base, *size); + return false; + } + if (base >= max_mem) return false; if ((base + *size) > max_mem) @@ -572,8 +593,11 @@ void __init early_init_dt_add_memory_arch(u64 base, u64 size) /* Add the chunk to the MEMBLOCK list */ if (add_mem_to_memblock) { - if (validate_mem_limit(base, &size)) + if (validate_mem_limit(base, &size)) { memblock_add(base, size); + if (base + size > top_phys_addr) + top_phys_addr = base + size; + } } } @@ -613,6 +637,8 @@ static void __init early_reserve_mem_dt(void) static void __init early_reserve_mem(void) { __be64 *reserve_map; + phys_addr_t kasan_shadow_start; + phys_addr_t kasan_memory_size; reserve_map = (__be64 *)(((unsigned long)initial_boot_params) + fdt_off_mem_rsvmap(initial_boot_params)); @@ -651,6 +677,42 @@ static void __init early_reserve_mem(void) return; } #endif + + if (IS_ENABLED(CONFIG_KASAN) && IS_ENABLED(CONFIG_PPC_BOOK3S_64)) { + kasan_memory_size = + ((phys_addr_t)CONFIG_PHYS_MEM_SIZE_FOR_KASAN << 20); + + if (top_phys_addr < kasan_memory_size) { + /* + * We are doomed. We shouldn't even be able to get this + * far, but we do in qemu. If we continue and turn + * relocations on, we'll take fatal page faults for + * memory that's not physically present. Instead, + * panic() here: it will be saved to __log_buf even if + * it doesn't get printed to the console. + */ + panic("Tried to book a KASAN kernel configured for %u MB with only %llu MB! Aborting.", + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, + (u64)(top_phys_addr >> 20)); + } else if (top_phys_addr > kasan_memory_size) { + /* print a biiiig warning in hopes people notice */ + pr_err("===========================================\n" + "Physical memory exceeds compiled-in maximum!\n" + "This kernel was compiled for KASAN with %u MB physical memory.\n" + "The physical memory detected is at least %llu MB.\n" + "Memory above the compiled limit will not be used!\n" + "===========================================\n", + CONFIG_PHYS_MEM_SIZE_FOR_KASAN, + (u64)(top_phys_addr >> 20)); + } + + kasan_shadow_start = _ALIGN_DOWN(kasan_memory_size * 7 / 8, + PAGE_SIZE); + DBG("reserving %llx -> %llx for KASAN", + kasan_shadow_start, top_phys_addr); + memblock_reserve(kasan_shadow_start, + top_phys_addr - kasan_shadow_start); + } } #ifdef CONFIG_PPC_TRANSACTIONAL_MEM diff --git a/arch/powerpc/mm/kasan/Makefile b/arch/powerpc/mm/kasan/Makefile index 6577897673dd..f02b15c78e4d 100644 --- a/arch/powerpc/mm/kasan/Makefile +++ b/arch/powerpc/mm/kasan/Makefile @@ -2,4 +2,5 @@ KASAN_SANITIZE := n -obj-$(CONFIG_PPC32) += kasan_init_32.o +obj-$(CONFIG_PPC32) += init_32.o +obj-$(CONFIG_PPC_BOOK3S_64) += init_book3s_64.o diff --git a/arch/powerpc/mm/kasan/kasan_init_32.c b/arch/powerpc/mm/kasan/init_32.c similarity index 100% rename from arch/powerpc/mm/kasan/kasan_init_32.c rename to arch/powerpc/mm/kasan/init_32.c diff --git a/arch/powerpc/mm/kasan/init_book3s_64.c b/arch/powerpc/mm/kasan/init_book3s_64.c new file mode 100644 index 000000000000..f961e96be136 --- /dev/null +++ b/arch/powerpc/mm/kasan/init_book3s_64.c @@ -0,0 +1,72 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * KASAN for 64-bit Book3S powerpc + * + * Copyright (C) 2019 IBM Corporation + * Author: Daniel Axtens <dja@axtens.net> + */ + +#define DISABLE_BRANCH_PROFILING + +#include <linux/kasan.h> +#include <linux/printk.h> +#include <linux/sched/task.h> +#include <asm/pgalloc.h> + +void __init kasan_init(void) +{ + int i; + void *k_start = kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START); + void *k_end = kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END); + + pte_t pte = __pte(__pa(kasan_early_shadow_page) | + pgprot_val(PAGE_KERNEL) | _PAGE_PTE); + + if (!early_radix_enabled()) + panic("KASAN requires radix!"); + + for (i = 0; i < PTRS_PER_PTE; i++) + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, + &kasan_early_shadow_pte[i], pte, 0); + + for (i = 0; i < PTRS_PER_PMD; i++) + pmd_populate_kernel(&init_mm, &kasan_early_shadow_pmd[i], + kasan_early_shadow_pte); + + for (i = 0; i < PTRS_PER_PUD; i++) + pud_populate(&init_mm, &kasan_early_shadow_pud[i], + kasan_early_shadow_pmd); + + memset(kasan_mem_to_shadow((void *)PAGE_OFFSET), KASAN_SHADOW_INIT, + KASAN_SHADOW_SIZE); + + kasan_populate_early_shadow( + kasan_mem_to_shadow((void *)RADIX_KERN_VIRT_START), + kasan_mem_to_shadow((void *)RADIX_VMALLOC_START)); + + /* leave a hole here for vmalloc */ + + kasan_populate_early_shadow( + kasan_mem_to_shadow((void *)RADIX_VMALLOC_END), + kasan_mem_to_shadow((void *)RADIX_VMEMMAP_END)); + + flush_tlb_kernel_range((unsigned long)k_start, (unsigned long)k_end); + + /* mark early shadow region as RO and wipe */ + pte = __pte(__pa(kasan_early_shadow_page) | + pgprot_val(PAGE_KERNEL_RO) | _PAGE_PTE); + for (i = 0; i < PTRS_PER_PTE; i++) + __set_pte_at(&init_mm, (unsigned long)kasan_early_shadow_page, + &kasan_early_shadow_pte[i], pte, 0); + + /* + * clear_page relies on some cache info that hasn't been set up yet. + * It ends up looping ~forever and blows up other data. + * Use memset instead. + */ + memset(kasan_early_shadow_page, 0, PAGE_SIZE); + + /* Enable error messages */ + init_task.kasan_depth = 0; + pr_info("KASAN init done (64-bit Book3S heavyweight mode)\n"); +}
KASAN support on Book3S is a bit tricky to get right: - It would be good to support inline instrumentation so as to be able to catch stack issues that cannot be caught with outline mode. - Inline instrumentation requires a fixed offset. - Book3S runs code in real mode after booting. Most notably a lot of KVM runs in real mode, and it would be good to be able to instrument it. - Because code runs in real mode after boot, the offset has to point to valid memory both in and out of real mode. [For those not immersed in ppc64, in real mode, the top nibble or 2 bits (depending on radix/hash mmu) of the address is ignored. The linear mapping is placed at 0xc000000000000000. This means that a pointer to part of the linear mapping will work both in real mode, where it will be interpreted as a physical address of the form 0x000..., and out of real mode, where it will go via the linear mapping.] One approach is just to give up on inline instrumentation. This way all checks can be delayed until after everything set is up correctly, and the address-to-shadow calculations can be overridden. However, the features and speed boost provided by inline instrumentation are worth trying to do better. If _at compile time_ it is known how much contiguous physical memory a system has, the top 1/8th of the first block of physical memory can be set aside for the shadow. This is a big hammer and comes with 3 big consequences: - there's no nice way to handle physically discontiguous memory, so only the first physical memory block can be used. - kernels will simply fail to boot on machines with less memory than specified when compiling. - kernels running on machines with more memory than specified when compiling will simply ignore the extra memory. Implement and document KASAN this way. The current implementation is Radix only. Despite the limitations, it can still find bugs, e.g. http://patchwork.ozlabs.org/patch/1103775/ At the moment, this physical memory limit must be set _even for outline mode_. This may be changed in a later series - a different implementation could be added for outline mode that dynamically allocates shadow at a fixed offset. For example, see https://patchwork.ozlabs.org/patch/795211/ Suggested-by: Michael Ellerman <mpe@ellerman.id.au> Cc: Balbir Singh <bsingharora@gmail.com> # ppc64 out-of-line radix version Cc: Christophe Leroy <christophe.leroy@c-s.fr> # ppc32 version Signed-off-by: Daniel Axtens <dja@axtens.net> --- Changes since v2: - Address feedback from Christophe around cleanups and docs. - Address feedback from Balbir: at this point I don't have a good solution for the issues you identify around the limitations of the inline implementation but I think that it's worth trying to get the stack instrumentation support. I'm happy to have an alternative and more flexible outline mode - I had envisoned this would be called 'lightweight' mode as it imposes fewer restrictions. I've linked to your implementation. I think it's best to add it in a follow-up series. - Made the default PHYS_MEM_SIZE_FOR_KASAN value 1024MB. I think most people have guests with at least that much memory in the Radix 64s case so it's a much saner default - it means that if you just turn on KASAN without reading the docs you're much more likely to have a bootable kernel, which you will never have if the value is set to zero! I'm happy to bikeshed the value if we want. Changes since v1: - Landed kasan vmalloc support upstream - Lots of feedback from Christophe. Changes since the rfc: - Boots real and virtual hardware, kvm works. - disabled reporting when we're checking the stack for exception frames. The behaviour isn't wrong, just incompatible with KASAN. - Documentation! - Dropped old module stuff in favour of KASAN_VMALLOC. The bugs with ftrace and kuap were due to kernel bloat pushing prom_init calls to be done via the plt. Because we did not have a relocatable kernel, and they are done very early, this caused everything to explode. Compile with CONFIG_RELOCATABLE! --- Documentation/dev-tools/kasan.rst | 8 +- Documentation/powerpc/kasan.txt | 112 +++++++++++++++++- arch/powerpc/Kconfig | 3 + arch/powerpc/Kconfig.debug | 21 ++++ arch/powerpc/Makefile | 11 ++ arch/powerpc/include/asm/book3s/64/hash.h | 4 + arch/powerpc/include/asm/book3s/64/pgtable.h | 7 ++ arch/powerpc/include/asm/book3s/64/radix.h | 5 + arch/powerpc/include/asm/kasan.h | 21 +++- arch/powerpc/kernel/process.c | 8 ++ arch/powerpc/kernel/prom.c | 64 +++++++++- arch/powerpc/mm/kasan/Makefile | 3 +- .../mm/kasan/{kasan_init_32.c => init_32.c} | 0 arch/powerpc/mm/kasan/init_book3s_64.c | 72 +++++++++++ 14 files changed, 330 insertions(+), 9 deletions(-) rename arch/powerpc/mm/kasan/{kasan_init_32.c => init_32.c} (100%) create mode 100644 arch/powerpc/mm/kasan/init_book3s_64.c