LU-12564 libcfs: Use vfree_atomic instead of vfree

Since vfree is unsafe to use in atomic context, implement our own
libcfs_vfree_atomic heavily based on code from linux 4.10 commit
bf22e37a641327e34681b7b6959d9646e3886770

We can't use the one in the kernel because it's not exported.

unconditionally use it in *_FREE_LARGE() macros since in_atomic()
is not recommended to be used outside of core kernel code.

Not everything is present on 3.10 (rhel7) so we also add
llist primitive and a replacement for raw_cpu_ptr there.

Change-Id: I50892f231e54a284f4d8a14d910ea9ab2fbe6a16
Signed-off-by: Oleg Drokin <green@whamcloud.com>
Reviewed-on: https://review.whamcloud.com/40136
Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
Reviewed-by: Aurelien Degremont <degremoa@amazon.com>
Tested-by: jenkins <devops@whamcloud.com>
Reviewed-by: Neil Brown <neilb@suse.de>
Tested-by: Maloo <maloo@whamcloud.com>

diff --git a/libcfs/include/libcfs/libcfs.h b/libcfs/include/libcfs/libcfs.h
index a172dda..2bbc820 100644 (file)
--- a/libcfs/include/libcfs/libcfs.h
+++ b/libcfs/include/libcfs/libcfs.h
@@ -129,4 +129,7 @@ do {                                                                        \
                );                                                      \
 } while (0)
 
+/* atomic-context safe vfree */
+void libcfs_vfree_atomic(const void *addr);
+
 #endif /* _LIBCFS_LIBCFS_H_ */

diff --git a/libcfs/include/libcfs/libcfs_private.h b/libcfs/include/libcfs/libcfs_private.h
index cbc003d..9c48a13 100644 (file)
--- a/libcfs/include/libcfs/libcfs_private.h
+++ b/libcfs/include/libcfs/libcfs_private.h
@@ -214,6 +214,9 @@ do {                                                                            \
 #define LIBCFS_CPT_ALLOC(ptr, cptab, cpt, size)                                    \
        LIBCFS_CPT_ALLOC_GFP(ptr, cptab, cpt, size, GFP_NOFS)
 
+void init_libcfs_vfree_atomic(void);
+void exit_libcfs_vfree_atomic(void);
+
 #define LIBCFS_FREE(ptr, size)                                         \
 do {                                                                   \
        int s = (size);                                                 \
@@ -226,7 +229,7 @@ do {                                                                        \
        CDEBUG(D_MALLOC, "kfreed '" #ptr "': %d at %p (tot %lld).\n",   \
               s, (ptr), libcfs_kmem_read());                           \
        if (unlikely(s > LIBCFS_VMALLOC_SIZE))                          \
-               vfree(ptr);                                             \
+               libcfs_vfree_atomic(ptr);                               \
        else                                                            \
                kfree(ptr);                                             \
 } while (0)

diff --git a/libcfs/libcfs/libcfs_mem.c b/libcfs/libcfs/libcfs_mem.c
index 3e83f50..757cc19 100644 (file)
--- a/libcfs/libcfs/libcfs_mem.c
+++ b/libcfs/libcfs/libcfs_mem.c
@@ -33,6 +33,7 @@
 
 #define DEBUG_SUBSYSTEM S_LNET
 
+#include <linux/workqueue.h>
 #include <libcfs/libcfs.h>
 
 struct cfs_var_array {
@@ -170,3 +171,62 @@ cfs_array_alloc(int count, unsigned int size)
        return (void *)&arr->va_ptrs[0];
 }
 EXPORT_SYMBOL(cfs_array_alloc);
+
+/*
+ * This is opencoding of vfree_atomic from Linux kernel added in 4.10 with
+ * minimum changes needed to work on older kernels too.
+ */
+
+#ifndef raw_cpu_ptr
+#define raw_cpu_ptr(p) __this_cpu_ptr(p)
+#endif
+
+#ifndef llist_for_each_safe
+#define llist_for_each_safe(pos, n, node)                       \
+        for ((pos) = (node); (pos) && ((n) = (pos)->next, true); (pos) = (n))
+#endif
+
+struct vfree_deferred {
+       struct llist_head list;
+       struct work_struct wq;
+};
+static DEFINE_PER_CPU(struct vfree_deferred, vfree_deferred);
+
+static void free_work(struct work_struct *w)
+{
+       struct vfree_deferred *p = container_of(w, struct vfree_deferred, wq);
+       struct llist_node *t, *llnode;
+
+       llist_for_each_safe(llnode, t, llist_del_all(&p->list))
+               vfree((void *)llnode);
+}
+
+void libcfs_vfree_atomic(const void *addr)
+{
+       struct vfree_deferred *p = raw_cpu_ptr(&vfree_deferred);
+
+       if (!addr)
+               return;
+
+       if (llist_add((struct llist_node *)addr, &p->list))
+               schedule_work(&p->wq);
+}
+EXPORT_SYMBOL(libcfs_vfree_atomic);
+
+void __init init_libcfs_vfree_atomic(void)
+{
+       int i;
+
+       for_each_possible_cpu(i) {
+               struct vfree_deferred *p;
+
+               p = &per_cpu(vfree_deferred, i);
+               init_llist_head(&p->list);
+               INIT_WORK(&p->wq, free_work);
+       }
+}
+
+void __exit exit_libcfs_vfree_atomic(void)
+{
+       flush_scheduled_work();
+}

diff --git a/libcfs/libcfs/module.c b/libcfs/libcfs/module.c
index ca2a797..51a3ce4 100644 (file)
--- a/libcfs/libcfs/module.c
+++ b/libcfs/libcfs/module.c
@@ -721,6 +721,8 @@ static int __init libcfs_init(void)
 
        cfs_arch_init();
 
+       init_libcfs_vfree_atomic();
+
        rc = libcfs_debug_init(5 * 1024 * 1024);
        if (rc < 0) {
                pr_err("LustreError: libcfs_debug_init: rc = %d\n", rc);
@@ -813,6 +815,8 @@ static void __exit libcfs_exit(void)
        rc = libcfs_debug_cleanup();
        if (rc)
                pr_err("LustreError: libcfs_debug_cleanup: rc = %d\n", rc);
+
+       exit_libcfs_vfree_atomic();
 }
 
 MODULE_AUTHOR("OpenSFS, Inc. <http://www.lustre.org/>");

diff --git a/lustre/include/obd_support.h b/lustre/include/obd_support.h
index 462ead5..847c443 100644 (file)
--- a/lustre/include/obd_support.h
+++ b/lustre/include/obd_support.h
@@ -892,7 +892,7 @@ do {                                                                              \
        if (is_vmalloc_addr(ptr)) {                                           \
                OBD_FREE_PRE(ptr, size, "vfreed");                            \
                POISON(ptr, 0x5a, size);                                      \
-               vfree(ptr);                                                   \
+               libcfs_vfree_atomic(ptr);                                     \
                POISON_PTR(ptr);                                              \
        } else {                                                              \
                OBD_FREE(ptr, size);                                          \