The patch below does not apply to the 4.4-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <stable@xxxxxxxxxxxxxxx>. thanks, greg k-h ------------------ original commit in Linus's tree ------------------ >From 47da29763ec9a153b9b685bff9db659e4e09e494 Mon Sep 17 00:00:00 2001 From: Johannes Berg <johannes.berg@xxxxxxxxx> Date: Wed, 13 Jan 2021 22:08:02 +0100 Subject: [PATCH] um: mm: check more comprehensively for stub changes If userspace tries to change the stub, we need to kill it, because otherwise it can escape the virtual machine. In a few cases the stub checks weren't good, e.g. if userspace just tries to mmap(0x100000 - 0x1000, 0x3000, ...) it could succeed to get a new private/anonymous mapping replacing the stubs. Fix this by checking everywhere, and checking for _overlap_, not just direct changes. Cc: stable@xxxxxxxxxxxxxxx Fixes: 3963333fe676 ("uml: cover stubs with a VMA") Signed-off-by: Johannes Berg <johannes.berg@xxxxxxxxx> Signed-off-by: Richard Weinberger <richard@xxxxxx> diff --git a/arch/um/kernel/tlb.c b/arch/um/kernel/tlb.c index 61776790cd67..89468da6bf88 100644 --- a/arch/um/kernel/tlb.c +++ b/arch/um/kernel/tlb.c @@ -125,6 +125,9 @@ static int add_mmap(unsigned long virt, unsigned long phys, unsigned long len, struct host_vm_op *last; int fd = -1, ret = 0; + if (virt + len > STUB_START && virt < STUB_END) + return -EINVAL; + if (hvc->userspace) fd = phys_mapping(phys, &offset); else @@ -162,7 +165,7 @@ static int add_munmap(unsigned long addr, unsigned long len, struct host_vm_op *last; int ret = 0; - if ((addr >= STUB_START) && (addr < STUB_END)) + if (addr + len > STUB_START && addr < STUB_END) return -EINVAL; if (hvc->index != 0) { @@ -192,6 +195,9 @@ static int add_mprotect(unsigned long addr, unsigned long len, struct host_vm_op *last; int ret = 0; + if (addr + len > STUB_START && addr < STUB_END) + return -EINVAL; + if (hvc->index != 0) { last = &hvc->ops[hvc->index - 1]; if ((last->type == MPROTECT) && @@ -472,6 +478,10 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long address) struct mm_id *mm_id; address &= PAGE_MASK; + + if (address >= STUB_START && address < STUB_END) + goto kill; + pgd = pgd_offset(mm, address); if (!pgd_present(*pgd)) goto kill;