summaryrefslogtreecommitdiff
path: root/arch/um/os-Linux/skas/mem.c
diff options
context:
space:
mode:
authorBenjamin Berg <benjamin.berg@intel.com>2025-06-02 15:00:52 +0200
committerJohannes Berg <johannes.berg@intel.com>2025-06-02 16:20:10 +0200
commite92e2552858142b60238b9828d802f128e4acccd (patch)
tree57e6a51acfc7397215131df99d1a67fc58d18349 /arch/um/os-Linux/skas/mem.c
parentbeddc9fb1cb161e1bf779b180750b648ff9690c7 (diff)
um: pass FD for memory operations when needed
Instead of always sharing the FDs with the userspace process, only hand over the FDs needed for mmap when required. The idea is that userspace might be able to force the stub into executing an mmap syscall, however, it will not be able to manipulate the control flow sufficiently to have access to an FD that would allow mapping arbitrary memory. Security wise, we need to be sure that only the expected syscalls are executed after the kernel sends FDs through the socket. This is currently not the case, as userspace can trivially jump to the rt_sigreturn syscall instruction to execute any syscall that the stub is permitted to do. With this, it can trick the kernel to send the FD, which in turn allows userspace to freely map any physical memory. As such, this is currently *not* secure. However, in principle the approach should be fine with a more strict SECCOMP filter and a careful review of the stub control flow (as userspace can prepare a stack). With some care, it is likely possible to extend the security model to SMP if desired. Signed-off-by: Benjamin Berg <benjamin.berg@intel.com> Link: https://patch.msgid.link/20250602130052.545733-8-benjamin@sipsolutions.net Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Diffstat (limited to 'arch/um/os-Linux/skas/mem.c')
-rw-r--r--arch/um/os-Linux/skas/mem.c66
1 files changed, 63 insertions, 3 deletions
diff --git a/arch/um/os-Linux/skas/mem.c b/arch/um/os-Linux/skas/mem.c
index 31bf3a52047a..8b9921ac3ef8 100644
--- a/arch/um/os-Linux/skas/mem.c
+++ b/arch/um/os-Linux/skas/mem.c
@@ -43,6 +43,16 @@ void syscall_stub_dump_error(struct mm_id *mm_idp)
print_hex_dump(UM_KERN_ERR, " syscall data: ", 0,
16, 4, sc, sizeof(*sc), 0);
+
+ if (using_seccomp) {
+ printk(UM_KERN_ERR "%s: FD map num: %d", __func__,
+ mm_idp->syscall_fd_num);
+ print_hex_dump(UM_KERN_ERR,
+ " FD map: ", 0, 16,
+ sizeof(mm_idp->syscall_fd_map[0]),
+ mm_idp->syscall_fd_map,
+ sizeof(mm_idp->syscall_fd_map), 0);
+ }
}
static inline unsigned long *check_init_stack(struct mm_id * mm_idp,
@@ -118,6 +128,9 @@ static inline long do_syscall_stub(struct mm_id *mm_idp)
mm_idp->syscall_data_len = 0;
}
+ if (using_seccomp)
+ mm_idp->syscall_fd_num = 0;
+
return mm_idp->syscall_data_len;
}
@@ -180,6 +193,44 @@ static struct stub_syscall *syscall_stub_get_previous(struct mm_id *mm_idp,
return NULL;
}
+static int get_stub_fd(struct mm_id *mm_idp, int fd)
+{
+ int i;
+
+ /* Find an FD slot (or flush and use first) */
+ if (!using_seccomp)
+ return fd;
+
+ /* Already crashed, value does not matter */
+ if (mm_idp->syscall_data_len < 0)
+ return 0;
+
+ /* Find existing FD in map if we can allocate another syscall */
+ if (mm_idp->syscall_data_len <
+ ARRAY_SIZE(((struct stub_data *)NULL)->syscall_data)) {
+ for (i = 0; i < mm_idp->syscall_fd_num; i++) {
+ if (mm_idp->syscall_fd_map[i] == fd)
+ return i;
+ }
+
+ if (mm_idp->syscall_fd_num < STUB_MAX_FDS) {
+ i = mm_idp->syscall_fd_num;
+ mm_idp->syscall_fd_map[i] = fd;
+
+ mm_idp->syscall_fd_num++;
+
+ return i;
+ }
+ }
+
+ /* FD map full or no syscall space available, continue after flush */
+ do_syscall_stub(mm_idp);
+ mm_idp->syscall_fd_map[0] = fd;
+ mm_idp->syscall_fd_num = 1;
+
+ return 0;
+}
+
int map(struct mm_id *mm_idp, unsigned long virt, unsigned long len, int prot,
int phys_fd, unsigned long long offset)
{
@@ -187,12 +238,21 @@ int map(struct mm_id *mm_idp, unsigned long virt, unsigned long len, int prot,
/* Compress with previous syscall if that is possible */
sc = syscall_stub_get_previous(mm_idp, STUB_SYSCALL_MMAP, virt);
- if (sc && sc->mem.prot == prot && sc->mem.fd == phys_fd &&
+ if (sc && sc->mem.prot == prot &&
sc->mem.offset == MMAP_OFFSET(offset - sc->mem.length)) {
- sc->mem.length += len;
- return 0;
+ int prev_fd = sc->mem.fd;
+
+ if (using_seccomp)
+ prev_fd = mm_idp->syscall_fd_map[sc->mem.fd];
+
+ if (phys_fd == prev_fd) {
+ sc->mem.length += len;
+ return 0;
+ }
}
+ phys_fd = get_stub_fd(mm_idp, phys_fd);
+
sc = syscall_stub_alloc(mm_idp);
sc->syscall = STUB_SYSCALL_MMAP;
sc->mem.addr = virt;