kernel/drivers/misc/lkdtm_core.c, branch linux-4.16.y

Merge tag 'usercopy-v4.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux

2018-02-04T00:25:42Z

Pull hardened usercopy whitelisting from Kees Cook: "Currently, hardened usercopy performs dynamic bounds checking on slab cache objects. This is good, but still leaves a lot of kernel memory available to be copied to/from userspace in the face of bugs. To further restrict what memory is available for copying, this creates a way to whitelist specific areas of a given slab cache object for copying to/from userspace, allowing much finer granularity of access control. Slab caches that are never exposed to userspace can declare no whitelist for their objects, thereby keeping them unavailable to userspace via dynamic copy operations. (Note, an implicit form of whitelisting is the use of constant sizes in usercopy operations and get_user()/put_user(); these bypass all hardened usercopy checks since these sizes cannot change at runtime.) This new check is WARN-by-default, so any mistakes can be found over the next several releases without breaking anyone's system. The series has roughly the following sections: - remove %p and improve reporting with offset - prepare infrastructure and whitelist kmalloc - update VFS subsystem with whitelists - update SCSI subsystem with whitelists - update network subsystem with whitelists - update process memory with whitelists - update per-architecture thread_struct with whitelists - update KVM with whitelists and fix ioctl bug - mark all other allocations as not whitelisted - update lkdtm for more sensible test overage" * tag 'usercopy-v4.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux: (38 commits) lkdtm: Update usercopy tests for whitelisting usercopy: Restrict non-usercopy caches to size 0 kvm: x86: fix KVM_XEN_HVM_CONFIG ioctl kvm: whitelist struct kvm_vcpu_arch arm: Implement thread_struct whitelist for hardened usercopy arm64: Implement thread_struct whitelist for hardened usercopy x86: Implement thread_struct whitelist for hardened usercopy fork: Provide usercopy whitelisting for task_struct fork: Define usercopy region in thread_stack slab caches fork: Define usercopy region in mm_struct slab caches net: Restrict unwhitelisted proto caches to size 0 sctp: Copy struct sctp_sock.autoclose to userspace using put_user() sctp: Define usercopy region in SCTP proto slab cache caif: Define usercopy region in caif proto slab cache ip: Define usercopy region in IP proto slab cache net: Define usercopy region in struct proto slab cache scsi: Define usercopy region in scsi_sense_cache slab cache cifs: Define usercopy region in cifs_request slab cache vxfs: Define usercopy region in vxfs_inode slab cache ufs: Define usercopy region in ufs_inode_cache slab cache ...

lkdtm: fix handle_irq_event symbol for INT_HW_IRQ_EN

2018-01-25T09:41:05Z

Function handle_IRQ_event was retired in v2.6.39 and replaced with handle_irq_event but nobody changed it in lkdtm so INT_HW_IRQ_EN has been broken for a while. Fixes: 33b054b867b8 ("genirq: Remove handle_IRQ_event") Signed-off-by: Travis Brown Signed-off-by: Ivan Delalande Acked-by: Kees Cook Signed-off-by: Greg Kroah-Hartman

lkdtm: Update usercopy tests for whitelisting

2018-01-15T20:08:09Z

This updates the USERCOPY_HEAP_FLAG_* tests to USERCOPY_HEAP_WHITELIST_*, since the final form of usercopy whitelisting ended up using an offset/size window instead of the earlier proposed allocation flags. Signed-off-by: Kees Cook

lkdtm: Constify the crashtypes table

2017-10-20T13:51:39Z

Since nothing in the crashtypes table ever changes, mark it const. Adds some missing "static" markings as well. Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman

lkdtm: Convert from jprobe to kprobe

2017-10-20T13:50:55Z

The jprobe subsystem is being removed, so convert to using kprobe instead. Cc: Masami Hiramatsu Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman

lkdtm: Add -fstack-protector-strong test

2017-08-15T19:27:35Z

There wasn't an LKDTM test to distinguish between -fstack-protector and -fstack-protector-strong in use. This adds CORRUPT_STACK_STRONG to see the difference. Also adjusts the stack-clobber value to 0xff so execution won't potentially jump into userspace when the stack protector is missing. Signed-off-by: Kees Cook

lkdtm: Test VMAP_STACK allocates leading/trailing guard pages

2017-08-04T20:04:21Z

Two new tests STACK_GUARD_PAGE_LEADING and STACK_GUARD_PAGE_TRAILING attempt to read the byte before and after, respectively, of the current stack frame, which should fault. Signed-off-by: Kees Cook

lkdtm: Provide timing tests for atomic_t vs refcount_t

2017-07-26T21:38:04Z

While not a crash test, this does provide two tight atomic_t and refcount_t loops for performance comparisons: cd /sys/kernel/debug/provoke-crash perf stat -B -- cat <(echo ATOMIC_TIMING) > DIRECT perf stat -B -- cat <(echo REFCOUNT_TIMING) > DIRECT Looking a CPU cycles is the best way to example the fast-path (rather than instruction counts, since conditional jumps will be executed but will be negligible due to branch-prediction). Signed-off-by: Kees Cook

lkdtm: Provide more complete coverage for REFCOUNT tests

2017-07-26T21:38:03Z

The existing REFCOUNT_* LKDTM tests were designed only for testing a narrow portion of CONFIG_REFCOUNT_FULL. This moves the tests to their own file and expands their testing to poke each boundary condition. Since the protections (CONFIG_REFCOUNT_FULL and x86-fast) use different saturation values and reach-zero behavior, those have to be build-time set so the tests can actually validate things are happening at the right places. Notably, the x86-fast protection will fail REFCOUNT_INC_ZERO and REFCOUNT_ADD_ZERO since those conditions are not checked (only overflow is critical to protecting refcount_t). CONFIG_REFCOUNT_FULL will warn for each REFCOUNT_*_NEGATIVE test since it provides zero-pinning behaviors (which allows it to pass REFCOUNT_INC_ZERO and REFCOUNT_ADD_ZERO). Signed-off-by: Kees Cook

lkdtm: add bad USER_DS test

2017-04-08T15:56:02Z

This adds CORRUPT_USER_DS to check that the get_fs() test on syscall return (via __VERIFY_PRE_USERMODE_STATE) still sees USER_DS. Since trying to deal with values other than USER_DS and KERNEL_DS across all architectures in a safe way is not sensible, this sets KERNEL_DS, but since that could be extremely dangerous if the protection is not present, it also raises SIGKILL for current, so that no matter what, the process will die. A successful test will be visible with a BUG(), like all the other LKDTM tests. Signed-off-by: Kees Cook Signed-off-by: Greg Kroah-Hartman