From d4ccd54d28d3c8598e2354acc13e28c060961dbb Mon Sep 17 00:00:00 2001 From: Jann Horn Date: Thu, 17 Nov 2022 15:43:22 -0800 Subject: exit: Put an upper limit on how often we can oops Many Linux systems are configured to not panic on oops; but allowing an attacker to oops the system **really** often can make even bugs that look completely unexploitable exploitable (like NULL dereferences and such) if each crash elevates a refcount by one or a lock is taken in read mode, and this causes a counter to eventually overflow. The most interesting counters for this are 32 bits wide (like open-coded refcounts that don't use refcount_t). (The ldsem reader count on 32-bit platforms is just 16 bits, but probably nobody cares about 32-bit platforms that much nowadays.) So let's panic the system if the kernel is constantly oopsing. The speed of oopsing 2^32 times probably depends on several factors, like how long the stack trace is and which unwinder you're using; an empirically important one is whether your console is showing a graphical environment or a text console that oopses will be printed to. In a quick single-threaded benchmark, it looks like oopsing in a vfork() child with a very short stack trace only takes ~510 microseconds per run when a graphical console is active; but switching to a text console that oopses are printed to slows it down around 87x, to ~45 milliseconds per run. (Adding more threads makes this faster, but the actual oops printing happens under &die_lock on x86, so you can maybe speed this up by a factor of around 2 and then any further improvement gets eaten up by lock contention.) It looks like it would take around 8-12 days to overflow a 32-bit counter with repeated oopsing on a multi-core X86 system running a graphical environment; both me (in an X86 VM) and Seth (with a distro kernel on normal hardware in a standard configuration) got numbers in that ballpark. 12 days aren't *that* short on a desktop system, and you'd likely need much longer on a typical server system (assuming that people don't run graphical desktop environments on their servers), and this is a *very* noisy and violent approach to exploiting the kernel; and it also seems to take orders of magnitude longer on some machines, probably because stuff like EFI pstore will slow it down a ton if that's active. Signed-off-by: Jann Horn Link: https://lore.kernel.org/r/20221107201317.324457-1-jannh@google.com Reviewed-by: Luis Chamberlain Signed-off-by: Kees Cook Link: https://lore.kernel.org/r/20221117234328.594699-2-keescook@chromium.org --- Documentation/admin-guide/sysctl/kernel.rst | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'Documentation/admin-guide') diff --git a/Documentation/admin-guide/sysctl/kernel.rst b/Documentation/admin-guide/sysctl/kernel.rst index 98d1b198b2b4..09f3fb2f8585 100644 --- a/Documentation/admin-guide/sysctl/kernel.rst +++ b/Documentation/admin-guide/sysctl/kernel.rst @@ -667,6 +667,14 @@ This is the default behavior. an oops event is detected. +oops_limit +========== + +Number of kernel oopses after which the kernel should panic when +``panic_on_oops`` is not set. Setting this to 0 or 1 has the same effect +as setting ``panic_on_oops=1``. + + osrelease, ostype & version =========================== -- cgit v1.2.3 From de92f65719cd672f4b48397540b9f9eff67eca40 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 2 Dec 2022 12:59:11 -0800 Subject: exit: Allow oops_limit to be disabled In preparation for keeping oops_limit logic in sync with warn_limit, have oops_limit == 0 disable checking the Oops counter. Cc: Jann Horn Cc: Jonathan Corbet Cc: Andrew Morton Cc: Baolin Wang Cc: "Jason A. Donenfeld" Cc: Eric Biggers Cc: Huang Ying Cc: "Eric W. Biederman" Cc: Arnd Bergmann Cc: linux-doc@vger.kernel.org Signed-off-by: Kees Cook --- Documentation/admin-guide/sysctl/kernel.rst | 5 +++-- kernel/exit.c | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-) (limited to 'Documentation/admin-guide') diff --git a/Documentation/admin-guide/sysctl/kernel.rst b/Documentation/admin-guide/sysctl/kernel.rst index 09f3fb2f8585..a31d8d81ea07 100644 --- a/Documentation/admin-guide/sysctl/kernel.rst +++ b/Documentation/admin-guide/sysctl/kernel.rst @@ -671,8 +671,9 @@ oops_limit ========== Number of kernel oopses after which the kernel should panic when -``panic_on_oops`` is not set. Setting this to 0 or 1 has the same effect -as setting ``panic_on_oops=1``. +``panic_on_oops`` is not set. Setting this to 0 disables checking +the count. Setting this to 1 has the same effect as setting +``panic_on_oops=1``. The default value is 10000. osrelease, ostype & version diff --git a/kernel/exit.c b/kernel/exit.c index dc1a32149f94..deffb8e4b1b2 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -954,7 +954,7 @@ void __noreturn make_task_dead(int signr) * To make sure this can't happen, place an upper bound on how often the * kernel may oops without panic(). */ - if (atomic_inc_return(&oops_count) >= READ_ONCE(oops_limit)) + if (atomic_inc_return(&oops_count) >= READ_ONCE(oops_limit) && oops_limit) panic("Oopsed too often (kernel.oops_limit is %d)", oops_limit); /* -- cgit v1.2.3 From 9fc9e278a5c0b708eeffaf47d6eb0c82aa74ed78 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Thu, 17 Nov 2022 15:43:25 -0800 Subject: panic: Introduce warn_limit Like oops_limit, add warn_limit for limiting the number of warnings when panic_on_warn is not set. Cc: Jonathan Corbet Cc: Andrew Morton Cc: Baolin Wang Cc: "Jason A. Donenfeld" Cc: Eric Biggers Cc: Huang Ying Cc: Petr Mladek Cc: tangmeng Cc: "Guilherme G. Piccoli" Cc: Tiezhu Yang Cc: Sebastian Andrzej Siewior Cc: linux-doc@vger.kernel.org Reviewed-by: Luis Chamberlain Signed-off-by: Kees Cook Link: https://lore.kernel.org/r/20221117234328.594699-5-keescook@chromium.org --- Documentation/admin-guide/sysctl/kernel.rst | 10 ++++++++++ kernel/panic.c | 14 ++++++++++++++ 2 files changed, 24 insertions(+) (limited to 'Documentation/admin-guide') diff --git a/Documentation/admin-guide/sysctl/kernel.rst b/Documentation/admin-guide/sysctl/kernel.rst index a31d8d81ea07..179bd303b585 100644 --- a/Documentation/admin-guide/sysctl/kernel.rst +++ b/Documentation/admin-guide/sysctl/kernel.rst @@ -1509,6 +1509,16 @@ entry will default to 2 instead of 0. 2 Unprivileged calls to ``bpf()`` are disabled = ============================================================= + +warn_limit +========== + +Number of kernel warnings after which the kernel should panic when +``panic_on_warn`` is not set. Setting this to 0 disables checking +the warning count. Setting this to 1 has the same effect as setting +``panic_on_warn=1``. The default value is 0. + + watchdog ======== diff --git a/kernel/panic.c b/kernel/panic.c index cfa354322d5f..f4403fc14f67 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -58,6 +58,7 @@ bool crash_kexec_post_notifiers; int panic_on_warn __read_mostly; unsigned long panic_on_taint; bool panic_on_taint_nousertaint = false; +static unsigned int warn_limit __read_mostly; int panic_timeout = CONFIG_PANIC_TIMEOUT; EXPORT_SYMBOL_GPL(panic_timeout); @@ -88,6 +89,13 @@ static struct ctl_table kern_panic_table[] = { .extra2 = SYSCTL_ONE, }, #endif + { + .procname = "warn_limit", + .data = &warn_limit, + .maxlen = sizeof(warn_limit), + .mode = 0644, + .proc_handler = proc_douintvec, + }, { } }; @@ -203,8 +211,14 @@ static void panic_print_sys_info(bool console_flush) void check_panic_on_warn(const char *origin) { + static atomic_t warn_count = ATOMIC_INIT(0); + if (panic_on_warn) panic("%s: panic_on_warn set ...\n", origin); + + if (atomic_inc_return(&warn_count) >= READ_ONCE(warn_limit) && warn_limit) + panic("%s: system warned too often (kernel.warn_limit is %d)", + origin, warn_limit); } /** -- cgit v1.2.3