In case of "Use After Free" kernel OOPs, free path of object
is required to debug futher.
And in most of cases object address is present in one of registers.
Thus check for register address and if it belongs to slab,
print its alloc and free path.
e.g. in below issue register r6 belongs to slab, and user afree issue
occurred on one of its derefer values:
[ 124.310386] (ptrval)
[ 124.312647] 8<--- cut here ---
[ 124.313761] Unable to handle kernel paging request at virtual address 6b6b6b6f
[ 124.315972] pgd = (ptrval)
...
[ 124.328290] pc : [<c052fc0c>] lr : [<c052fc00>] psr: 60000013
[ 124.330349] sp : c8993d28 ip : 0000bff4 fp : c8ae2020
[ 124.332071] r10: 00000000 r9 : 00000001 r8 : c1804cc8
[ 124.333803] r7 : 00000000 r6 : c8ae9180 r5 : c1804a80 r4 : c8ae2008
[ 124.335936] r3 : 6b6b6b6b r2 : 315049d6 r1 : 2d867000 r0 : c1396584
..
[ 124.365233] register r6: c8ae9180 belongs to slab object
[ 124.366364] INFO: Allocated in meminfo_proc_show+0x3c/0x500 age=1 cpu=0 pid=69
[ 124.367545] meminfo_proc_show+0x3c/0x500
[ 124.368271] seq_read_iter+0x10c/0x4bc
[ 124.368994] proc_reg_read_iter+0x74/0xa8
[ 124.369712] generic_file_splice_read+0xe8/0x178
[ 124.370496] splice_direct_to_actor+0xe0/0x2b8
[ 124.371261] do_splice_direct+0xa4/0xdc
[ 124.371917] do_sendfile+0x1c4/0x3ec
[ 124.372550] sys_sendfile64+0x128/0x130
[ 124.373109] ret_fast_syscall+0x0/0x54
[ 124.373664] 0xbe9a2de4
[ 124.374081] INFO: Freed in meminfo_proc_show+0x5c/0x500 age=1 cpu=0 pid=69
[ 124.374933] meminfo_proc_show+0x5c/0x500
[ 124.375485] seq_read_iter+0x10c/0x4bc
[ 124.376020] proc_reg_read_iter+0x74/0xa8
[ 124.376643] generic_file_splice_read+0xe8/0x178
[ 124.377331] splice_direct_to_actor+0xe0/0x2b8
[ 124.378022] do_splice_direct+0xa4/0xdc
[ 124.378633] do_sendfile+0x1c4/0x3ec
[ 124.379220] sys_sendfile64+0x128/0x130
[ 124.379822] ret_fast_syscall+0x0/0x54
[ 124.380421] 0xbe9a2de4
Co-developed-by: Vaneet Narang <[email protected]>
Signed-off-by: Vaneet Narang <[email protected]>
Signed-off-by: Maninder Singh <[email protected]>
---
arch/arm/include/asm/bug.h | 1 +
arch/arm/kernel/process.c | 18 ++++++++++++++++++
arch/arm/kernel/traps.c | 1 +
include/linux/slab.h | 14 ++++++++++++++
mm/slab.h | 7 -------
mm/slub.c | 19 +++++++++++++++++++
6 files changed, 53 insertions(+), 7 deletions(-)
diff --git a/arch/arm/include/asm/bug.h b/arch/arm/include/asm/bug.h
index 673c7dd..ba8d9d7 100644
--- a/arch/arm/include/asm/bug.h
+++ b/arch/arm/include/asm/bug.h
@@ -88,5 +88,6 @@ extern asmlinkage void c_backtrace(unsigned long fp, int pmode,
struct mm_struct;
void show_pte(const char *lvl, struct mm_struct *mm, unsigned long addr);
extern void __show_regs(struct pt_regs *);
+extern void __show_regs_alloc_free(struct pt_regs *regs);
#endif
diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c
index 5199a2b..97d2a7c 100644
--- a/arch/arm/kernel/process.c
+++ b/arch/arm/kernel/process.c
@@ -27,6 +27,7 @@
#include <linux/random.h>
#include <linux/hw_breakpoint.h>
#include <linux/leds.h>
+#include <linux/slab.h>
#include <asm/processor.h>
#include <asm/thread_notify.h>
@@ -92,6 +93,23 @@ void arch_cpu_idle_exit(void)
ledtrig_cpu(CPU_LED_IDLE_END);
}
+void __show_regs_alloc_free(struct pt_regs *regs)
+{
+ int i;
+
+ /* check for r0 - r12 only */
+ for (i = 0; i < 13; i++) {
+ unsigned long addr = regs->uregs[i];
+ void *object;
+ struct kmem_cache *cache;
+
+ if (slab_page_object(addr, &object, &cache)) {
+ printk("\nregister r%d: %lx belongs to slab object\n", i, addr);
+ print_tracking(cache, object);
+ }
+ }
+}
+
void __show_regs(struct pt_regs *regs)
{
unsigned long flags;
diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
index 17d5a78..64308e3 100644
--- a/arch/arm/kernel/traps.c
+++ b/arch/arm/kernel/traps.c
@@ -287,6 +287,7 @@ static int __die(const char *str, int err, struct pt_regs *regs)
print_modules();
__show_regs(regs);
+ __show_regs_alloc_free(regs);
pr_emerg("Process %.*s (pid: %d, stack limit = 0x%p)\n",
TASK_COMM_LEN, tsk->comm, task_pid_nr(tsk), end_of_stack(tsk));
diff --git a/include/linux/slab.h b/include/linux/slab.h
index 7ae6040..a19ba55 100644
--- a/include/linux/slab.h
+++ b/include/linux/slab.h
@@ -706,4 +706,18 @@ static inline void *kzalloc_node(size_t size, gfp_t flags, int node)
#define slab_dead_cpu NULL
#endif
+#ifdef CONFIG_SLUB_DEBUG
+bool slab_page_object(unsigned long address, void **object, struct kmem_cache **cache);
+extern void print_tracking(struct kmem_cache *s, void *object);
+#else
+static inline void print_tracking(struct kmem_cache *s, void *object)
+{
+}
+
+static inline bool slab_page_object(unsigned long address, void **object, struct kmem_cache **cache)
+{
+ return false;
+}
+#endif
+
#endif /* _LINUX_SLAB_H */
diff --git a/mm/slab.h b/mm/slab.h
index 076582f..8a072bd 100644
--- a/mm/slab.h
+++ b/mm/slab.h
@@ -208,18 +208,11 @@ static inline enum node_stat_item cache_vmstat_idx(struct kmem_cache *s)
NR_SLAB_RECLAIMABLE_B : NR_SLAB_UNRECLAIMABLE_B;
}
-#ifdef CONFIG_SLUB_DEBUG
#ifdef CONFIG_SLUB_DEBUG_ON
DECLARE_STATIC_KEY_TRUE(slub_debug_enabled);
#else
DECLARE_STATIC_KEY_FALSE(slub_debug_enabled);
#endif
-extern void print_tracking(struct kmem_cache *s, void *object);
-#else
-static inline void print_tracking(struct kmem_cache *s, void *object)
-{
-}
-#endif
/*
* Returns true if any of the specified slub_debug flags is enabled for the
diff --git a/mm/slub.c b/mm/slub.c
index 0d5fac3..1863806 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -648,6 +648,25 @@ void print_tracking(struct kmem_cache *s, void *object)
print_track("Freed", get_track(s, object, TRACK_FREE), pr_time);
}
+bool slab_page_object(unsigned long address, void **object, struct kmem_cache **cache)
+{
+ void *addr = (void *)address;
+ struct page *page;
+
+ if ((addr >= (void *)PAGE_OFFSET) &&
+ (addr < high_memory)) {
+ page = virt_to_head_page(addr);
+
+ if (PageSlab(page)) {
+ *cache = page->slab_cache;
+ *object = nearest_obj(*cache, page, addr);
+ return true;
+ }
+ }
+
+ return false;
+}
+
static void print_page_info(struct page *page)
{
pr_err("INFO: Slab 0x%p objects=%u used=%u fp=0x%p flags=0x%04lx\n",
--
1.9.1
On Wed, Feb 24, 2021 at 06:07:34PM +0530, Maninder Singh wrote:
> +bool slab_page_object(unsigned long address, void **object, struct kmem_cache **cache)
> +{
> + void *addr = (void *)address;
> + struct page *page;
> +
> + if ((addr >= (void *)PAGE_OFFSET) &&
> + (addr < high_memory)) {
> + page = virt_to_head_page(addr);
This check is not sufficient. There can be holes in the page array.
You need to use virt_addr_valid() to validate "addr".
--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!