Messages in this thread | | | Subject | Re: [PATCHv3 4/4] arm64: dump: Add checking for writable and exectuable pages | From | Laura Abbott <> | Date | Thu, 20 Oct 2016 06:01:29 -0700 |
| |
On 10/20/2016 03:32 AM, Ard Biesheuvel wrote: > On 18 October 2016 at 23:01, Laura Abbott <labbott@redhat.com> wrote: >> >> Page mappings with full RWX permissions are a security risk. x86 >> has an option to walk the page tables and dump any bad pages. >> (See e1a58320a38d ("x86/mm: Warn on W^X mappings")). Add a similar >> implementation for arm64. >> >> Reviewed-by: Kees Cook <keescook@chromium.org> >> Reviewed-by: Mark Rutland <mark.rutland@arm.com> >> Tested-by: Mark Rutland <mark.rutland@arm.com> >> Signed-off-by: Laura Abbott <labbott@redhat.com> >> --- >> v3: Rebased for header guard fixup, whitespace fixes >> --- >> arch/arm64/Kconfig.debug | 29 +++++++++++++++++++++++ >> arch/arm64/include/asm/ptdump.h | 8 +++++++ >> arch/arm64/mm/dump.c | 52 +++++++++++++++++++++++++++++++++++++++++ >> arch/arm64/mm/mmu.c | 2 ++ >> 4 files changed, 91 insertions(+) >> >> diff --git a/arch/arm64/Kconfig.debug b/arch/arm64/Kconfig.debug >> index 21a5b74..d1ebd46 100644 >> --- a/arch/arm64/Kconfig.debug >> +++ b/arch/arm64/Kconfig.debug >> @@ -42,6 +42,35 @@ config ARM64_RANDOMIZE_TEXT_OFFSET >> of TEXT_OFFSET and platforms must not require a specific >> value. >> >> +config DEBUG_WX >> + bool "Warn on W+X mappings at boot" >> + select ARM64_PTDUMP_CORE >> + ---help--- >> + Generate a warning if any W+X mappings are found at boot. >> + >> + This is useful for discovering cases where the kernel is leaving >> + W+X mappings after applying NX, as such mappings are a security risk. >> + This check also includes UXN, which should be set on all kernel >> + mappings. >> + >> + Look for a message in dmesg output like this: >> + >> + arm64/mm: Checked W+X mappings: passed, no W+X pages found. >> + >> + or like this, if the check failed: >> + >> + arm64/mm: Checked W+X mappings: FAILED, <N> W+X pages found. >> + >> + Note that even if the check fails, your kernel is possibly >> + still fine, as W+X mappings are not a security hole in >> + themselves, what they do is that they make the exploitation >> + of other unfixed kernel bugs easier. >> + >> + There is no runtime or memory usage effect of this option >> + once the kernel has booted up - it's a one time check. >> + >> + If in doubt, say "Y". >> + >> config DEBUG_SET_MODULE_RONX >> bool "Set loadable kernel module data as NX and text as RO" >> depends on MODULES >> diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h >> index f72ee69..6afd847 100644 >> --- a/arch/arm64/include/asm/ptdump.h >> +++ b/arch/arm64/include/asm/ptdump.h >> @@ -42,5 +42,13 @@ static inline int ptdump_debugfs_register(struct ptdump_info *info, >> return 0; >> } >> #endif >> +void ptdump_check_wx(void); >> #endif /* CONFIG_ARM64_PTDUMP_CORE */ >> + >> +#ifdef CONFIG_DEBUG_WX >> +#define debug_checkwx() ptdump_check_wx() >> +#else >> +#define debug_checkwx() do { } while (0) >> +#endif >> + >> #endif /* __ASM_PTDUMP_H */ >> diff --git a/arch/arm64/mm/dump.c b/arch/arm64/mm/dump.c >> index bb36649..4913af5 100644 >> --- a/arch/arm64/mm/dump.c >> +++ b/arch/arm64/mm/dump.c >> @@ -74,6 +74,8 @@ struct pg_state { >> unsigned long start_address; >> unsigned level; >> u64 current_prot; >> + bool check_wx; >> + unsigned long wx_pages; >> }; >> >> struct prot_bits { >> @@ -202,6 +204,35 @@ static void dump_prot(struct pg_state *st, const struct prot_bits *bits, >> } >> } >> >> +static void note_prot_uxn(struct pg_state *st, unsigned long addr) >> +{ >> + if (!st->check_wx) >> + return; >> + >> + if ((st->current_prot & PTE_UXN) == PTE_UXN) >> + return; >> + >> + WARN_ONCE(1, "arm64/mm: Found non-UXN mapping at address %p/%pS\n", >> + (void *)st->start_address, (void *)st->start_address); >> + >> + st->wx_pages += (addr - st->start_address) / PAGE_SIZE; >> +} >> + >> +static void note_prot_wx(struct pg_state *st, unsigned long addr) >> +{ >> + if (!st->check_wx) >> + return; >> + if ((st->current_prot & PTE_RDONLY) == PTE_RDONLY) >> + return; >> + if ((st->current_prot & PTE_PXN) == PTE_PXN) >> + return; >> + >> + WARN_ONCE(1, "arm64/mm: Found insecure W+X mapping at address %p/%pS\n", >> + (void *)st->start_address, (void *)st->start_address); >> + >> + st->wx_pages += (addr - st->start_address) / PAGE_SIZE; >> +} >> + > > Why are these separate functions, and why is wx_pages increased twice, > potentially? > > Given how rare non-UXN kernel mappings should be, could we not just add > > if ((st->current_prot & PTE_UXN) == 0) > WARN(xxx) > > (without the _ONCE) to note_prot_wx(), and drop note_prot_uxn() entirely? > >
UXN is a separate bit from PTE_PXN/PTE_RDONLY and both pairs need to be checked. The current return == 0 logic means that one set or the other may not get checked. Rather than complicate the logic, it seemed better to have separate functions. I see your point about the wx_pages double counting so I can fix that.
>> static void note_page(struct pg_state *st, unsigned long addr, unsigned level, >> u64 val) >> { >> @@ -219,6 +250,8 @@ static void note_page(struct pg_state *st, unsigned long addr, unsigned level, >> unsigned long delta; >> >> if (st->current_prot) { >> + note_prot_uxn(st, addr); >> + note_prot_wx(st, addr); >> pt_dump_seq_printf(st->seq, "0x%016lx-0x%016lx ", >> st->start_address, addr); >> >> @@ -344,6 +377,25 @@ static struct ptdump_info kernel_ptdump_info = { >> .base_addr = VA_START, >> }; >> >> +void ptdump_check_wx(void) >> +{ >> + struct pg_state st = { >> + .seq = NULL, >> + .marker = (struct addr_marker[]) { >> + { -1, NULL}, >> + }, >> + .check_wx = true, >> + }; >> + >> + walk_pgd(&st, &init_mm, 0); >> + note_page(&st, 0, 0, 0); >> + if (st.wx_pages) >> + pr_info("Checked W+X mappings: FAILED, %lu W+X pages found\n", >> + st.wx_pages); > > Could we upgrade this to pr_warn? >
Sure
>> + else >> + pr_info("Checked W+X mappings: passed, no W+X pages found\n"); >> +} >> + >> static int ptdump_init(void) >> { >> ptdump_initialize(); >> diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c >> index 05615a3..2cbe2fe 100644 >> --- a/arch/arm64/mm/mmu.c >> +++ b/arch/arm64/mm/mmu.c >> @@ -42,6 +42,7 @@ >> #include <asm/tlb.h> >> #include <asm/memblock.h> >> #include <asm/mmu_context.h> >> +#include <asm/ptdump.h> >> >> u64 idmap_t0sz = TCR_T0SZ(VA_BITS); >> >> @@ -396,6 +397,7 @@ void mark_rodata_ro(void) >> section_size = (unsigned long)__init_begin - (unsigned long)__start_rodata; >> create_mapping_late(__pa(__start_rodata), (unsigned long)__start_rodata, >> section_size, PAGE_KERNEL_RO); >> + debug_checkwx(); >> } >> >> static void __init map_kernel_segment(pgd_t *pgd, void *va_start, void *va_end, >> -- >> 2.7.4 >>
| |