5.9.3-rc1 review

@@ -1450,14 +1450,13 @@ static int cpufreq_online(unsigned int c
 	 */
 	if ((cpufreq_driver->flags & CPUFREQ_NEED_INITIAL_FREQ_CHECK)
 	    && has_target()) {
+		unsigned int old_freq = policy->cur;
+
 		/* Are we running at unknown frequency ? */
-		ret = cpufreq_frequency_table_get_index(policy, policy->cur);
+		ret = cpufreq_frequency_table_get_index(policy, old_freq);
 		if (ret == -EINVAL) {
-			/* Warn user and fix it */
-			pr_warn("%s: CPU%d: Running at unlisted freq: %u KHz\n",
-				__func__, policy->cpu, policy->cur);
-			ret = __cpufreq_driver_target(policy, policy->cur - 1,
-				CPUFREQ_RELATION_L);
+			ret = __cpufreq_driver_target(policy, old_freq - 1,
+						      CPUFREQ_RELATION_L);
 
 			/*
 			 * Reaching here after boot in a few seconds may not

@@ -1465,8 +1464,8 @@ static int cpufreq_online(unsigned int c
 			 * frequency for longer duration. Hence, a BUG_ON().
 			 */
 			BUG_ON(ret);
-			pr_warn("%s: CPU%d: Unlisted initial frequency changed to: %u KHz\n",
-				__func__, policy->cpu, policy->cur);
+			pr_info("%s: CPU%d: Running at unlisted initial frequency: %u KHz, changing to: %u KHz\n",
+				__func__, policy->cpu, old_freq, policy->cur);
 		}
 	}
 

@@ -896,6 +896,12 @@ static inline struct nft_expr *nft_expr_
 	return (struct nft_expr *)&rule->data[rule->dlen];
 }
 
+static inline bool nft_expr_more(const struct nft_rule *rule,
+				 const struct nft_expr *expr)
+{
+	return expr != nft_expr_last(rule) && expr->ops;
+}
+
 static inline struct nft_userdata *nft_userdata(const struct nft_rule *rule)
 {
 	return (void *)&rule->data[rule->dlen];

@@ -302,7 +302,7 @@ static void nft_rule_expr_activate(const
 	struct nft_expr *expr;
 
 	expr = nft_expr_first(rule);
-	while (expr != nft_expr_last(rule) && expr->ops) {
+	while (nft_expr_more(rule, expr)) {
 		if (expr->ops->activate)
 			expr->ops->activate(ctx, expr);
 

@@ -317,7 +317,7 @@ static void nft_rule_expr_deactivate(con
 	struct nft_expr *expr;
 
 	expr = nft_expr_first(rule);
-	while (expr != nft_expr_last(rule) && expr->ops) {
+	while (nft_expr_more(rule, expr)) {
 		if (expr->ops->deactivate)
 			expr->ops->deactivate(ctx, expr, phase);
 

@@ -3036,7 +3036,7 @@ static void nf_tables_rule_destroy(const
 	 * is called on error from nf_tables_newrule().
 	 */
 	expr = nft_expr_first(rule);
-	while (expr != nft_expr_last(rule) && expr->ops) {
+	while (nft_expr_more(rule, expr)) {
 		next = nft_expr_next(expr);
 		nf_tables_expr_destroy(ctx, expr);
 		expr = next;

@@ -37,7 +37,7 @@ struct nft_flow_rule *nft_flow_rule_crea
 	struct nft_expr *expr;
 
 	expr = nft_expr_first(rule);
-	while (expr->ops && expr != nft_expr_last(rule)) {
+	while (nft_expr_more(rule, expr)) {
 		if (expr->ops->offload_flags & NFT_OFFLOAD_F_ACTION)
 			num_actions++;
 

@@ -61,7 +61,7 @@ struct nft_flow_rule *nft_flow_rule_crea
 	ctx->net = net;
 	ctx->dep.type = NFT_OFFLOAD_DEP_UNSPEC;
 
-	while (expr->ops && expr != nft_expr_last(rule)) {
+	while (nft_expr_more(rule, expr)) {
 		if (!expr->ops->offload) {
 			err = -EOPNOTSUPP;
 			goto err_out;

@@ -1226,13 +1226,26 @@ static void io_kill_timeout(struct io_ki
 	}
 }
 
-static void io_kill_timeouts(struct io_ring_ctx *ctx)
+static bool io_task_match(struct io_kiocb *req, struct task_struct *tsk)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+
+	if (!tsk || req->task == tsk)
+		return true;
+	if ((ctx->flags & IORING_SETUP_SQPOLL) && req->task == ctx->sqo_thread)
+		return true;
+	return false;
+}
+
+static void io_kill_timeouts(struct io_ring_ctx *ctx, struct task_struct *tsk)
 {
 	struct io_kiocb *req, *tmp;
 
 	spin_lock_irq(&ctx->completion_lock);
-	list_for_each_entry_safe(req, tmp, &ctx->timeout_list, timeout.list)
-		io_kill_timeout(req);
+	list_for_each_entry_safe(req, tmp, &ctx->timeout_list, timeout.list) {
+		if (io_task_match(req, tsk))
+			io_kill_timeout(req);
+	}
 	spin_unlock_irq(&ctx->completion_lock);
 }
 

@@ -5017,7 +5030,7 @@ static bool io_poll_remove_one(struct io
 	return do_complete;
 }
 
-static void io_poll_remove_all(struct io_ring_ctx *ctx)
+static void io_poll_remove_all(struct io_ring_ctx *ctx, struct task_struct *tsk)
 {
 	struct hlist_node *tmp;
 	struct io_kiocb *req;

@@ -5028,8 +5041,10 @@ static void io_poll_remove_all(struct io
 		struct hlist_head *list;
 
 		list = &ctx->cancel_hash[i];
-		hlist_for_each_entry_safe(req, tmp, list, hash_node)
-			posted += io_poll_remove_one(req);
+		hlist_for_each_entry_safe(req, tmp, list, hash_node) {
+			if (io_task_match(req, tsk))
+				posted += io_poll_remove_one(req);
+		}
 	}
 	spin_unlock_irq(&ctx->completion_lock);
 

@@ -7989,8 +8004,8 @@ static void io_ring_ctx_wait_and_kill(st
 	percpu_ref_kill(&ctx->refs);
 	mutex_unlock(&ctx->uring_lock);
 
-	io_kill_timeouts(ctx);
-	io_poll_remove_all(ctx);
+	io_kill_timeouts(ctx, NULL);
+	io_poll_remove_all(ctx, NULL);
 
 	if (ctx->io_wq)
 		io_wq_cancel_all(ctx->io_wq);

@@ -8221,7 +8236,7 @@ static bool io_cancel_task_cb(struct io_
 	struct io_kiocb *req = container_of(work, struct io_kiocb, work);
 	struct task_struct *task = data;
 
-	return req->task == task;
+	return io_task_match(req, task);
 }
 
 static int io_uring_flush(struct file *file, void *data)

@@ -5648,6 +5648,20 @@ static int io_req_defer(struct io_kiocb
 	return -EIOCBQUEUED;
 }
 
+static void io_req_drop_files(struct io_kiocb *req)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ctx->inflight_lock, flags);
+	list_del(&req->inflight_entry);
+	if (waitqueue_active(&ctx->inflight_wait))
+		wake_up(&ctx->inflight_wait);
+	spin_unlock_irqrestore(&ctx->inflight_lock, flags);
+	req->flags &= ~REQ_F_INFLIGHT;
+	req->work.files = NULL;
+}
+
 static void __io_clean_op(struct io_kiocb *req)
 {
 	struct io_async_ctx *io = req->io;

@@ -5697,17 +5711,8 @@ static void __io_clean_op(struct io_kioc
 		req->flags &= ~REQ_F_NEED_CLEANUP;
 	}
 
-	if (req->flags & REQ_F_INFLIGHT) {
-		struct io_ring_ctx *ctx = req->ctx;
-		unsigned long flags;
-
-		spin_lock_irqsave(&ctx->inflight_lock, flags);
-		list_del(&req->inflight_entry);
-		if (waitqueue_active(&ctx->inflight_wait))
-			wake_up(&ctx->inflight_wait);
-		spin_unlock_irqrestore(&ctx->inflight_lock, flags);
-		req->flags &= ~REQ_F_INFLIGHT;
-	}
+	if (req->flags & REQ_F_INFLIGHT)
+		io_req_drop_files(req);
 }
 
 static int io_issue_sqe(struct io_kiocb *req, const struct io_uring_sqe *sqe,

@@ -265,7 +265,16 @@ struct io_ring_ctx {
 	/* IO offload */
 	struct io_wq		*io_wq;
 	struct task_struct	*sqo_thread;	/* if using sq thread polling */
-	struct mm_struct	*sqo_mm;
+
+	/*
+	 * For SQPOLL usage - we hold a reference to the parent task, so we
+	 * have access to the ->files
+	 */
+	struct task_struct	*sqo_task;
+
+	/* Only used for accounting purposes */
+	struct mm_struct	*mm_account;
+
 	wait_queue_head_t	sqo_wait;
 
 	/*

@@ -969,9 +978,10 @@ static int __io_sq_thread_acquire_mm(str
 {
 	if (!current->mm) {
 		if (unlikely(!(ctx->flags & IORING_SETUP_SQPOLL) ||
-			     !mmget_not_zero(ctx->sqo_mm)))
+			     !ctx->sqo_task->mm ||
+			     !mmget_not_zero(ctx->sqo_task->mm)))
 			return -EFAULT;
-		kthread_use_mm(ctx->sqo_mm);
+		kthread_use_mm(ctx->sqo_task->mm);
 	}
 
 	return 0;

@@ -7591,11 +7601,11 @@ static void io_unaccount_mem(struct io_r
 	if (ctx->limit_mem)
 		__io_unaccount_mem(ctx->user, nr_pages);
 
-	if (ctx->sqo_mm) {
+	if (ctx->mm_account) {
 		if (acct == ACCT_LOCKED)
-			ctx->sqo_mm->locked_vm -= nr_pages;
+			ctx->mm_account->locked_vm -= nr_pages;
 		else if (acct == ACCT_PINNED)
-			atomic64_sub(nr_pages, &ctx->sqo_mm->pinned_vm);
+			atomic64_sub(nr_pages, &ctx->mm_account->pinned_vm);
 	}
 }
 

@@ -7610,11 +7620,11 @@ static int io_account_mem(struct io_ring
 			return ret;
 	}
 
-	if (ctx->sqo_mm) {
+	if (ctx->mm_account) {
 		if (acct == ACCT_LOCKED)
-			ctx->sqo_mm->locked_vm += nr_pages;
+			ctx->mm_account->locked_vm += nr_pages;
 		else if (acct == ACCT_PINNED)
-			atomic64_add(nr_pages, &ctx->sqo_mm->pinned_vm);
+			atomic64_add(nr_pages, &ctx->mm_account->pinned_vm);
 	}
 
 	return 0;

@@ -7918,9 +7928,12 @@ static void io_ring_ctx_free(struct io_r
 {
 	io_finish_async(ctx);
 	io_sqe_buffer_unregister(ctx);
-	if (ctx->sqo_mm) {
-		mmdrop(ctx->sqo_mm);
-		ctx->sqo_mm = NULL;
+
+	if (ctx->sqo_task) {
+		put_task_struct(ctx->sqo_task);
+		ctx->sqo_task = NULL;
+		mmdrop(ctx->mm_account);
+		ctx->mm_account = NULL;
 	}
 
 	io_sqe_files_unregister(ctx);

@@ -8665,8 +8678,16 @@ static int io_uring_create(unsigned entr
 	ctx->user = user;
 	ctx->creds = get_current_cred();
 
+	ctx->sqo_task = get_task_struct(current);
+
+	/*
+	 * This is just grabbed for accounting purposes. When a process exits,
+	 * the mm is exited and dropped before the files, hence we need to hang
+	 * on to this mm purely for the purposes of being able to unaccount
+	 * memory (locked/pinned vm). It's not used for anything else.
+	 */
 	mmgrab(current->mm);
-	ctx->sqo_mm = current->mm;
+	ctx->mm_account = current->mm;
 
 	/*
 	 * Account memory _before_ installing the file descriptor. Once

@@ -553,7 +553,6 @@ enum {
 	REQ_F_BUFFER_SELECTED_BIT,
 	REQ_F_NO_FILE_TABLE_BIT,
 	REQ_F_WORK_INITIALIZED_BIT,
-	REQ_F_TASK_PINNED_BIT,
 
 	/* not a real bit, just to check we're not overflowing the space */
 	__REQ_F_LAST_BIT,

@@ -599,8 +598,6 @@ enum {
 	REQ_F_NO_FILE_TABLE	= BIT(REQ_F_NO_FILE_TABLE_BIT),
 	/* io_wq_work is initialized */
 	REQ_F_WORK_INITIALIZED	= BIT(REQ_F_WORK_INITIALIZED_BIT),
-	/* req->task is refcounted */
-	REQ_F_TASK_PINNED	= BIT(REQ_F_TASK_PINNED_BIT),
 };
 
 struct async_poll {

@@ -942,14 +939,6 @@ struct sock *io_uring_get_socket(struct
 }
 EXPORT_SYMBOL(io_uring_get_socket);
 
-static void io_get_req_task(struct io_kiocb *req)
-{
-	if (req->flags & REQ_F_TASK_PINNED)
-		return;
-	get_task_struct(req->task);
-	req->flags |= REQ_F_TASK_PINNED;
-}
-
 static inline void io_clean_op(struct io_kiocb *req)
 {
 	if (req->flags & (REQ_F_NEED_CLEANUP | REQ_F_BUFFER_SELECTED |

@@ -957,13 +946,6 @@ static inline void io_clean_op(struct io
 		__io_clean_op(req);
 }
 
-/* not idempotent -- it doesn't clear REQ_F_TASK_PINNED */
-static void __io_put_req_task(struct io_kiocb *req)
-{
-	if (req->flags & REQ_F_TASK_PINNED)
-		put_task_struct(req->task);
-}
-
 static void io_sq_thread_drop_mm(void)
 {
 	struct mm_struct *mm = current->mm;

@@ -1589,7 +1571,8 @@ static void __io_free_req_finish(struct
 {
 	struct io_ring_ctx *ctx = req->ctx;
 
-	__io_put_req_task(req);
+	put_task_struct(req->task);
+
 	if (likely(!io_is_fallback_req(req)))
 		kmem_cache_free(req_cachep, req);
 	else

@@ -1916,16 +1899,13 @@ static void io_req_free_batch(struct req
 	if (req->flags & REQ_F_LINK_HEAD)
 		io_queue_next(req);
 
-	if (req->flags & REQ_F_TASK_PINNED) {
-		if (req->task != rb->task) {
-			if (rb->task)
-				put_task_struct_many(rb->task, rb->task_refs);
-			rb->task = req->task;
-			rb->task_refs = 0;
-		}
-		rb->task_refs++;
-		req->flags &= ~REQ_F_TASK_PINNED;
+	if (req->task != rb->task) {
+		if (rb->task)
+			put_task_struct_many(rb->task, rb->task_refs);
+		rb->task = req->task;
+		rb->task_refs = 0;
 	}
+	rb->task_refs++;
 
 	WARN_ON_ONCE(io_dismantle_req(req));
 	rb->reqs[rb->to_free++] = req;

@@ -2550,9 +2530,6 @@ static int io_prep_rw(struct io_kiocb *r
 	if (kiocb->ki_flags & IOCB_NOWAIT)
 		req->flags |= REQ_F_NOWAIT;
 
-	if (kiocb->ki_flags & IOCB_DIRECT)
-		io_get_req_task(req);
-
 	if (force_nonblock)
 		kiocb->ki_flags |= IOCB_NOWAIT;
 

@@ -2564,7 +2541,6 @@ static int io_prep_rw(struct io_kiocb *r
 		kiocb->ki_flags |= IOCB_HIPRI;
 		kiocb->ki_complete = io_complete_rw_iopoll;
 		req->iopoll_completed = 0;
-		io_get_req_task(req);
 	} else {
 		if (kiocb->ki_flags & IOCB_HIPRI)
 			return -EINVAL;

@@ -3132,8 +3108,6 @@ static bool io_rw_should_retry(struct io
 	kiocb->ki_flags |= IOCB_WAITQ;
 	kiocb->ki_flags &= ~IOCB_NOWAIT;
 	kiocb->ki_waitq = wait;
-
-	io_get_req_task(req);
 	return true;
 }
 

@@ -4965,7 +4939,6 @@ static bool io_arm_poll_handler(struct i
 	apoll->double_poll = NULL;
 
 	req->flags |= REQ_F_POLLED;
-	io_get_req_task(req);
 	req->apoll = apoll;
 	INIT_HLIST_NODE(&req->hash_node);
 

@@ -5148,8 +5121,6 @@ static int io_poll_add_prep(struct io_ki
 #endif
 	poll->events = demangle_poll(events) | EPOLLERR | EPOLLHUP |
 		       (events & EPOLLEXCLUSIVE);
-
-	io_get_req_task(req);
 	return 0;
 }
 

@@ -6336,7 +6307,6 @@ static int io_submit_sqe(struct io_kiocb
 			return ret;
 		}
 		trace_io_uring_link(ctx, req, head);
-		io_get_req_task(req);
 		list_add_tail(&req->link_list, &head->link_list);
 
 		/* last request of a link, enqueue the link */

@@ -6461,6 +6431,7 @@ static int io_init_req(struct io_ring_ct
 	/* one is dropped after submission, the other at completion */
 	refcount_set(&req->refs, 2);
 	req->task = current;
+	get_task_struct(req->task);
 	req->result = 0;
 
 	if (unlikely(req->opcode >= IORING_OP_LAST))

@@ -1229,16 +1229,23 @@ static bool io_task_match(struct io_kioc
 	return false;
 }
 
-static void io_kill_timeouts(struct io_ring_ctx *ctx, struct task_struct *tsk)
+/*
+ * Returns true if we found and killed one or more timeouts
+ */
+static bool io_kill_timeouts(struct io_ring_ctx *ctx, struct task_struct *tsk)
 {
 	struct io_kiocb *req, *tmp;
+	int canceled = 0;
 
 	spin_lock_irq(&ctx->completion_lock);
 	list_for_each_entry_safe(req, tmp, &ctx->timeout_list, timeout.list) {
-		if (io_task_match(req, tsk))
+		if (io_task_match(req, tsk)) {
 			io_kill_timeout(req);
+			canceled++;
+		}
 	}
 	spin_unlock_irq(&ctx->completion_lock);
+	return canceled != 0;
 }
 
 static void __io_queue_deferred(struct io_ring_ctx *ctx)

@@ -5013,7 +5020,10 @@ static bool io_poll_remove_one(struct io
 	return do_complete;
 }
 
-static void io_poll_remove_all(struct io_ring_ctx *ctx, struct task_struct *tsk)
+/*
+ * Returns true if we found and killed one or more poll requests
+ */
+static bool io_poll_remove_all(struct io_ring_ctx *ctx, struct task_struct *tsk)
 {
 	struct hlist_node *tmp;
 	struct io_kiocb *req;

@@ -5033,6 +5043,8 @@ static void io_poll_remove_all(struct io
 
 	if (posted)
 		io_cqring_ev_posted(ctx);
+
+	return posted != 0;
 }
 
 static int io_poll_cancel(struct io_ring_ctx *ctx, __u64 sqe_addr)

@@ -8178,11 +8190,14 @@ static void io_cancel_defer_files(struct
 	}
 }
 
-static void io_uring_cancel_files(struct io_ring_ctx *ctx,
+/*
+ * Returns true if we found and killed one or more files pinning requests
+ */
+static bool io_uring_cancel_files(struct io_ring_ctx *ctx,
 				  struct files_struct *files)
 {
 	if (list_empty_careful(&ctx->inflight_list))
-		return;
+		return false;
 
 	io_cancel_defer_files(ctx, files);
 	/* cancel all at once, should be faster than doing it one by one*/

@@ -8218,6 +8233,8 @@ static void io_uring_cancel_files(struct
 		schedule();
 		finish_wait(&ctx->inflight_wait, &wait);
 	}
+
+	return true;
 }
 
 static bool io_cancel_task_cb(struct io_wq_work *work, void *data)

@@ -1344,12 +1344,24 @@ static void io_cqring_mark_overflow(stru
 	}
 }
 
+static inline bool io_match_files(struct io_kiocb *req,
+				       struct files_struct *files)
+{
+	if (!files)
+		return true;
+	if (req->flags & REQ_F_WORK_INITIALIZED)
+		return req->work.files == files;
+	return false;
+}
+
 /* Returns true if there are no backlogged entries after the flush */
-static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force)
+static bool io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
+				     struct task_struct *tsk,
+				     struct files_struct *files)
 {
 	struct io_rings *rings = ctx->rings;
+	struct io_kiocb *req, *tmp;
 	struct io_uring_cqe *cqe;
-	struct io_kiocb *req;
 	unsigned long flags;
 	LIST_HEAD(list);
 

@@ -1368,13 +1380,16 @@ static bool io_cqring_overflow_flush(str
 		ctx->cq_overflow_flushed = 1;
 
 	cqe = NULL;
-	while (!list_empty(&ctx->cq_overflow_list)) {
+	list_for_each_entry_safe(req, tmp, &ctx->cq_overflow_list, compl.list) {
+		if (tsk && req->task != tsk)
+			continue;
+		if (!io_match_files(req, files))
+			continue;
+
 		cqe = io_get_cqring(ctx);
 		if (!cqe && !force)
 			break;
 
-		req = list_first_entry(&ctx->cq_overflow_list, struct io_kiocb,
-						compl.list);
 		list_move(&req->compl.list, &list);
 		if (cqe) {
 			WRITE_ONCE(cqe->user_data, req->user_data);

@@ -1988,7 +2003,7 @@ static unsigned io_cqring_events(struct
 		if (noflush && !list_empty(&ctx->cq_overflow_list))
 			return -1U;
 
-		io_cqring_overflow_flush(ctx, false);
+		io_cqring_overflow_flush(ctx, false, NULL, NULL);
 	}
 
 	/* See comment at the top of this file */

@@ -6489,7 +6504,7 @@ static int io_submit_sqes(struct io_ring
 	/* if we have a backlog and couldn't flush it all, return BUSY */
 	if (test_bit(0, &ctx->sq_check_overflow)) {
 		if (!list_empty(&ctx->cq_overflow_list) &&
-		    !io_cqring_overflow_flush(ctx, false))
+		    !io_cqring_overflow_flush(ctx, false, NULL, NULL))
 			return -EBUSY;
 	}
 

@@ -7993,7 +8008,7 @@ static void io_ring_exit_work(struct wor
 	 */
 	do {
 		if (ctx->rings)
-			io_cqring_overflow_flush(ctx, true);
+			io_cqring_overflow_flush(ctx, true, NULL, NULL);
 		io_iopoll_try_reap_events(ctx);
 	} while (!wait_for_completion_timeout(&ctx->ref_comp, HZ/20));
 	io_ring_ctx_free(ctx);

@@ -8013,7 +8028,7 @@ static void io_ring_ctx_wait_and_kill(st
 
 	/* if we failed setting up the ctx, we might not have any rings */
 	if (ctx->rings)
-		io_cqring_overflow_flush(ctx, true);
+		io_cqring_overflow_flush(ctx, true, NULL, NULL);
 	io_iopoll_try_reap_events(ctx);
 	idr_for_each(&ctx->personality_idr, io_remove_personalities, ctx);
 

@@ -8069,12 +8084,6 @@ static bool io_match_link(struct io_kioc
 	return false;
 }
 
-static inline bool io_match_files(struct io_kiocb *req,
-				       struct files_struct *files)
-{
-	return (req->flags & REQ_F_WORK_INITIALIZED) && req->work.files == files;
-}
-
 static bool io_match_link_files(struct io_kiocb *req,
 				struct files_struct *files)
 {

@@ -8365,7 +8374,7 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned
 	ret = 0;
 	if (ctx->flags & IORING_SETUP_SQPOLL) {
 		if (!list_empty_careful(&ctx->cq_overflow_list))
-			io_cqring_overflow_flush(ctx, false);
+			io_cqring_overflow_flush(ctx, false, NULL, NULL);
 		if (flags & IORING_ENTER_SQ_WAKEUP)
 			wake_up(&ctx->sqo_wait);
 		submitted = to_submit;

@@ -62,6 +62,7 @@
 #include <linux/oom.h>
 #include <linux/compat.h>
 #include <linux/vmalloc.h>
+#include <linux/io_uring.h>
 
 #include <linux/uaccess.h>
 #include <asm/mmu_context.h>

@@ -1895,6 +1896,11 @@ static int bprm_execve(struct linux_binp
 	struct files_struct *displaced;
 	int retval;
 
+	/*
+	 * Cancel any io_uring activity across execve
+	 */
+	io_uring_task_cancel();
+
 	retval = unshare_files(&displaced);
 	if (retval)
 		return retval;

@@ -21,6 +21,7 @@
 #include <linux/rcupdate.h>
 #include <linux/close_range.h>
 #include <net/sock.h>
+#include <linux/io_uring.h>
 
 unsigned int sysctl_nr_open __read_mostly = 1024*1024;
 unsigned int sysctl_nr_open_min = BITS_PER_LONG;

@@ -452,6 +453,7 @@ void exit_files(struct task_struct *tsk)
 	struct files_struct * files = tsk->files;
 
 	if (files) {
+		io_uring_files_cancel(files);
 		task_lock(tsk);
 		tsk->files = NULL;
 		task_unlock(tsk);

454 lines not on this page · raw message

53 lines not on this page · raw message

@@ -63,6 +63,7 @@ struct sighand_struct;
 struct signal_struct;
 struct task_delay_info;
 struct task_group;
+struct io_uring_task;
 
 /*
  * Task state bitmask. NOTE! These bits are also

@@ -935,6 +936,10 @@ struct task_struct {
 	/* Open file information: */
 	struct files_struct		*files;
 
+#ifdef CONFIG_IO_URING
+	struct io_uring_task		*io_uring;
+#endif
+
 	/* Namespaces: */
 	struct nsproxy			*nsproxy;
 

@@ -114,6 +114,9 @@ struct task_struct init_task
 	.thread		= INIT_THREAD,
 	.fs		= &init_fs,
 	.files		= &init_files,
+#ifdef CONFIG_IO_URING
+	.io_uring	= NULL,
+#endif
 	.signal		= &init_signals,
 	.sighand	= &init_sighand,
 	.nsproxy	= &init_nsproxy,

@@ -95,6 +95,7 @@
 #include <linux/stackleak.h>
 #include <linux/kasan.h>
 #include <linux/scs.h>
+#include <linux/io_uring.h>
 
 #include <asm/pgalloc.h>
 #include <linux/uaccess.h>

@@ -728,6 +729,7 @@ void __put_task_struct(struct task_struc
 	WARN_ON(refcount_read(&tsk->usage));
 	WARN_ON(tsk == current);
 
+	io_uring_free(tsk);
 	cgroup_free(tsk);
 	task_numa_free(tsk, true);
 	security_task_free(tsk);

@@ -2002,6 +2004,10 @@ static __latent_entropy struct task_stru
 	p->vtime.state = VTIME_INACTIVE;
 #endif
 
+#ifdef CONFIG_IO_URING
+	p->io_uring = NULL;
+#endif
+
 #if defined(SPLIT_RSS_COUNTING)
 	memset(&p->rss_stat, 0, sizeof(p->rss_stat));
 #endif

@@ -60,6 +60,7 @@ struct io_worker {
 	const struct cred *cur_creds;
 	const struct cred *saved_creds;
 	struct files_struct *restore_files;
+	struct nsproxy *restore_nsproxy;
 	struct fs_struct *restore_fs;
 };
 

@@ -153,6 +154,7 @@ static bool __io_worker_unuse(struct io_
 
 		task_lock(current);
 		current->files = worker->restore_files;
+		current->nsproxy = worker->restore_nsproxy;
 		task_unlock(current);
 	}
 

@@ -318,6 +320,7 @@ static void io_worker_start(struct io_wq
 
 	worker->flags |= (IO_WORKER_F_UP | IO_WORKER_F_RUNNING);
 	worker->restore_files = current->files;
+	worker->restore_nsproxy = current->nsproxy;
 	worker->restore_fs = current->fs;
 	io_wqe_inc_running(wqe, worker);
 }

@@ -454,6 +457,7 @@ static void io_impersonate_work(struct i
 	if (work->files && current->files != work->files) {
 		task_lock(current);
 		current->files = work->files;
+		current->nsproxy = work->nsproxy;
 		task_unlock(current);
 	}
 	if (work->fs && current->fs != work->fs)

@@ -88,6 +88,7 @@ struct io_wq_work {
 	struct files_struct *files;
 	struct mm_struct *mm;
 	const struct cred *creds;
+	struct nsproxy *nsproxy;
 	struct fs_struct *fs;
 	unsigned long fsize;
 	unsigned flags;

@@ -5678,6 +5678,7 @@ static void io_req_drop_files(struct io_
 	spin_unlock_irqrestore(&ctx->inflight_lock, flags);
 	req->flags &= ~REQ_F_INFLIGHT;
 	put_files_struct(req->work.files);
+	put_nsproxy(req->work.nsproxy);
 	req->work.files = NULL;
 }
 

@@ -6086,6 +6087,8 @@ static int io_grab_files(struct io_kiocb
 		return 0;
 
 	req->work.files = get_files_struct(current);
+	get_nsproxy(current->nsproxy);
+	req->work.nsproxy = current->nsproxy;
 	req->flags |= REQ_F_INFLIGHT;
 
 	spin_lock_irq(&ctx->inflight_lock);

@@ -88,7 +88,7 @@ enum {
  */
 struct io_wqe {
 	struct {
-		spinlock_t lock;
+		raw_spinlock_t lock;
 		struct io_wq_work_list work_list;
 		unsigned long hash_map;
 		unsigned flags;

@@ -149,7 +149,7 @@ static bool __io_worker_unuse(struct io_
 
 	if (current->files != worker->restore_files) {
 		__acquire(&wqe->lock);
-		spin_unlock_irq(&wqe->lock);
+		raw_spin_unlock_irq(&wqe->lock);
 		dropped_lock = true;
 
 		task_lock(current);

@@ -168,7 +168,7 @@ static bool __io_worker_unuse(struct io_
 	if (worker->mm) {
 		if (!dropped_lock) {
 			__acquire(&wqe->lock);
-			spin_unlock_irq(&wqe->lock);
+			raw_spin_unlock_irq(&wqe->lock);
 			dropped_lock = true;
 		}
 		__set_current_state(TASK_RUNNING);

@@ -222,17 +222,17 @@ static void io_worker_exit(struct io_wor
 	worker->flags = 0;
 	preempt_enable();
 
-	spin_lock_irq(&wqe->lock);
+	raw_spin_lock_irq(&wqe->lock);
 	hlist_nulls_del_rcu(&worker->nulls_node);
 	list_del_rcu(&worker->all_list);
 	if (__io_worker_unuse(wqe, worker)) {
 		__release(&wqe->lock);
-		spin_lock_irq(&wqe->lock);
+		raw_spin_lock_irq(&wqe->lock);
 	}
 	acct->nr_workers--;
 	nr_workers = wqe->acct[IO_WQ_ACCT_BOUND].nr_workers +
 			wqe->acct[IO_WQ_ACCT_UNBOUND].nr_workers;
-	spin_unlock_irq(&wqe->lock);
+	raw_spin_unlock_irq(&wqe->lock);
 
 	/* all workers gone, wq exit can proceed */
 	if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs))

@@ -508,7 +508,7 @@ get_next:
 		else if (!wq_list_empty(&wqe->work_list))
 			wqe->flags |= IO_WQE_FLAG_STALLED;
 
-		spin_unlock_irq(&wqe->lock);
+		raw_spin_unlock_irq(&wqe->lock);
 		if (!work)
 			break;
 		io_assign_current_work(worker, work);

@@ -542,17 +542,17 @@ get_next:
 				io_wqe_enqueue(wqe, linked);
 
 			if (hash != -1U && !next_hashed) {
-				spin_lock_irq(&wqe->lock);
+				raw_spin_lock_irq(&wqe->lock);
 				wqe->hash_map &= ~BIT_ULL(hash);
 				wqe->flags &= ~IO_WQE_FLAG_STALLED;
 				/* skip unnecessary unlock-lock wqe->lock */
 				if (!work)
 					goto get_next;
-				spin_unlock_irq(&wqe->lock);
+				raw_spin_unlock_irq(&wqe->lock);
 			}
 		} while (work);
 
-		spin_lock_irq(&wqe->lock);
+		raw_spin_lock_irq(&wqe->lock);
 	} while (1);
 }
 

@@ -567,7 +567,7 @@ static int io_wqe_worker(void *data)
 	while (!test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
 		set_current_state(TASK_INTERRUPTIBLE);
 loop:
-		spin_lock_irq(&wqe->lock);
+		raw_spin_lock_irq(&wqe->lock);
 		if (io_wqe_run_queue(wqe)) {
 			__set_current_state(TASK_RUNNING);
 			io_worker_handle_work(worker);

@@ -578,7 +578,7 @@ loop:
 			__release(&wqe->lock);
 			goto loop;
 		}
-		spin_unlock_irq(&wqe->lock);
+		raw_spin_unlock_irq(&wqe->lock);
 		if (signal_pending(current))
 			flush_signals(current);
 		if (schedule_timeout(WORKER_IDLE_TIMEOUT))

@@ -590,11 +590,11 @@ loop:
 	}
 
 	if (test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
-		spin_lock_irq(&wqe->lock);
+		raw_spin_lock_irq(&wqe->lock);
 		if (!wq_list_empty(&wqe->work_list))
 			io_worker_handle_work(worker);
 		else
-			spin_unlock_irq(&wqe->lock);
+			raw_spin_unlock_irq(&wqe->lock);
 	}
 
 	io_worker_exit(worker);

@@ -634,9 +634,9 @@ void io_wq_worker_sleeping(struct task_s
 
 	worker->flags &= ~IO_WORKER_F_RUNNING;
 
-	spin_lock_irq(&wqe->lock);
+	raw_spin_lock_irq(&wqe->lock);
 	io_wqe_dec_running(wqe, worker);
-	spin_unlock_irq(&wqe->lock);
+	raw_spin_unlock_irq(&wqe->lock);
 }
 
 static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)

@@ -660,7 +660,7 @@ static bool create_io_worker(struct io_w
 		return false;
 	}
 
-	spin_lock_irq(&wqe->lock);
+	raw_spin_lock_irq(&wqe->lock);
 	hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
 	list_add_tail_rcu(&worker->all_list, &wqe->all_list);
 	worker->flags |= IO_WORKER_F_FREE;

@@ -669,7 +669,7 @@ static bool create_io_worker(struct io_w
 	if (!acct->nr_workers && (worker->flags & IO_WORKER_F_BOUND))
 		worker->flags |= IO_WORKER_F_FIXED;
 	acct->nr_workers++;
-	spin_unlock_irq(&wqe->lock);
+	raw_spin_unlock_irq(&wqe->lock);
 
 	if (index == IO_WQ_ACCT_UNBOUND)
 		atomic_inc(&wq->user->processes);

@@ -724,12 +724,12 @@ static int io_wq_manager(void *data)
 			if (!node_online(node))
 				continue;
 
-			spin_lock_irq(&wqe->lock);
+			raw_spin_lock_irq(&wqe->lock);
 			if (io_wqe_need_worker(wqe, IO_WQ_ACCT_BOUND))
 				fork_worker[IO_WQ_ACCT_BOUND] = true;
 			if (io_wqe_need_worker(wqe, IO_WQ_ACCT_UNBOUND))
 				fork_worker[IO_WQ_ACCT_UNBOUND] = true;
-			spin_unlock_irq(&wqe->lock);
+			raw_spin_unlock_irq(&wqe->lock);
 			if (fork_worker[IO_WQ_ACCT_BOUND])
 				create_io_worker(wq, wqe, IO_WQ_ACCT_BOUND);
 			if (fork_worker[IO_WQ_ACCT_UNBOUND])

@@ -825,10 +825,10 @@ static void io_wqe_enqueue(struct io_wqe
 	}
 
 	work_flags = work->flags;
-	spin_lock_irqsave(&wqe->lock, flags);
+	raw_spin_lock_irqsave(&wqe->lock, flags);
 	io_wqe_insert_work(wqe, work);
 	wqe->flags &= ~IO_WQE_FLAG_STALLED;
-	spin_unlock_irqrestore(&wqe->lock, flags);
+	raw_spin_unlock_irqrestore(&wqe->lock, flags);
 
 	if ((work_flags & IO_WQ_WORK_CONCURRENT) ||
 	    !atomic_read(&acct->nr_running))

@@ -955,13 +955,13 @@ static void io_wqe_cancel_pending_work(s
 	unsigned long flags;
 
 retry:
-	spin_lock_irqsave(&wqe->lock, flags);
+	raw_spin_lock_irqsave(&wqe->lock, flags);
 	wq_list_for_each(node, prev, &wqe->work_list) {
 		work = container_of(node, struct io_wq_work, list);
 		if (!match->fn(work, match->data))
 			continue;
 		io_wqe_remove_pending(wqe, work, prev);
-		spin_unlock_irqrestore(&wqe->lock, flags);
+		raw_spin_unlock_irqrestore(&wqe->lock, flags);
 		io_run_cancel(work, wqe);
 		match->nr_pending++;
 		if (!match->cancel_all)

@@ -970,7 +970,7 @@ retry:
 		/* not safe to continue after unlock */
 		goto retry;
 	}
-	spin_unlock_irqrestore(&wqe->lock, flags);
+	raw_spin_unlock_irqrestore(&wqe->lock, flags);
 }
 
 static void io_wqe_cancel_running_work(struct io_wqe *wqe,

@@ -1078,7 +1078,7 @@ struct io_wq *io_wq_create(unsigned boun
 		}
 		atomic_set(&wqe->acct[IO_WQ_ACCT_UNBOUND].nr_running, 0);
 		wqe->wq = wq;
-		spin_lock_init(&wqe->lock);
+		raw_spin_lock_init(&wqe->lock);
 		INIT_WQ_LIST(&wqe->work_list);
 		INIT_HLIST_NULLS_HEAD(&wqe->free_list, 0);
 		INIT_LIST_HEAD(&wqe->all_list);

@@ -202,7 +202,6 @@ static void io_worker_exit(struct io_wor
 {
 	struct io_wqe *wqe = worker->wqe;
 	struct io_wqe_acct *acct = io_wqe_get_acct(wqe, worker);
-	unsigned nr_workers;
 
 	/*
 	 * If we're not at zero, someone else is holding a brief reference

@@ -230,15 +229,11 @@ static void io_worker_exit(struct io_wor
 		raw_spin_lock_irq(&wqe->lock);
 	}
 	acct->nr_workers--;
-	nr_workers = wqe->acct[IO_WQ_ACCT_BOUND].nr_workers +
-			wqe->acct[IO_WQ_ACCT_UNBOUND].nr_workers;
 	raw_spin_unlock_irq(&wqe->lock);
 
-	/* all workers gone, wq exit can proceed */
-	if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs))
-		complete(&wqe->wq->done);
-
 	kfree_rcu(worker, rcu);
+	if (refcount_dec_and_test(&wqe->wq->refs))
+		complete(&wqe->wq->done);
 }
 
 static inline bool io_wqe_run_queue(struct io_wqe *wqe)

@@ -641,7 +636,7 @@ void io_wq_worker_sleeping(struct task_s
 
 static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
 {
-	struct io_wqe_acct *acct =&wqe->acct[index];
+	struct io_wqe_acct *acct = &wqe->acct[index];
 	struct io_worker *worker;
 
 	worker = kzalloc_node(sizeof(*worker), GFP_KERNEL, wqe->node);

@@ -674,6 +669,7 @@ static bool create_io_worker(struct io_w
 	if (index == IO_WQ_ACCT_UNBOUND)
 		atomic_inc(&wq->user->processes);
 
+	refcount_inc(&wq->refs);
 	wake_up_process(worker->task);
 	return true;
 }

@@ -689,28 +685,63 @@ static inline bool io_wqe_need_worker(st
 	return acct->nr_workers < acct->max_workers;
 }
 
+static bool io_wqe_worker_send_sig(struct io_worker *worker, void *data)
+{
+	send_sig(SIGINT, worker->task, 1);
+	return false;
+}
+
+/*
+ * Iterate the passed in list and call the specific function for each
+ * worker that isn't exiting
+ */
+static bool io_wq_for_each_worker(struct io_wqe *wqe,
+				  bool (*func)(struct io_worker *, void *),
+				  void *data)
+{
+	struct io_worker *worker;
+	bool ret = false;
+
+	list_for_each_entry_rcu(worker, &wqe->all_list, all_list) {
+		if (io_worker_get(worker)) {
+			/* no task if node is/was offline */
+			if (worker->task)
+				ret = func(worker, data);
+			io_worker_release(worker);
+			if (ret)
+				break;
+		}
+	}
+
+	return ret;
+}
+
+static bool io_wq_worker_wake(struct io_worker *worker, void *data)
+{
+	wake_up_process(worker->task);
+	return false;
+}
+
 /*
  * Manager thread. Tasked with creating new workers, if we need them.
  */
 static int io_wq_manager(void *data)
 {
 	struct io_wq *wq = data;
-	int workers_to_create = num_possible_nodes();
 	int node;
 
 	/* create fixed workers */
-	refcount_set(&wq->refs, workers_to_create);
+	refcount_set(&wq->refs, 1);
 	for_each_node(node) {
 		if (!node_online(node))
 			continue;
-		if (!create_io_worker(wq, wq->wqes[node], IO_WQ_ACCT_BOUND))
-			goto err;
-		workers_to_create--;
+		if (create_io_worker(wq, wq->wqes[node], IO_WQ_ACCT_BOUND))
+			continue;
+		set_bit(IO_WQ_BIT_ERROR, &wq->state);
+		set_bit(IO_WQ_BIT_EXIT, &wq->state);
+		goto out;
 	}
 
-	while (workers_to_create--)
-		refcount_dec(&wq->refs);
-
 	complete(&wq->done);
 
 	while (!kthread_should_stop()) {

@@ -742,12 +773,18 @@ static int io_wq_manager(void *data)
 	if (current->task_works)
 		task_work_run();
 
-	return 0;
-err:
-	set_bit(IO_WQ_BIT_ERROR, &wq->state);
-	set_bit(IO_WQ_BIT_EXIT, &wq->state);
-	if (refcount_sub_and_test(workers_to_create, &wq->refs))
+out:
+	if (refcount_dec_and_test(&wq->refs)) {
 		complete(&wq->done);
+		return 0;
+	}
+	/* if ERROR is set and we get here, we have workers to wake */
+	if (test_bit(IO_WQ_BIT_ERROR, &wq->state)) {
+		rcu_read_lock();
+		for_each_node(node)
+			io_wq_for_each_worker(wq->wqes[node], io_wq_worker_wake, NULL);
+		rcu_read_unlock();
+	}
 	return 0;
 }
 

@@ -854,37 +891,6 @@ void io_wq_hash_work(struct io_wq_work *
 	work->flags |= (IO_WQ_WORK_HASHED | (bit << IO_WQ_HASH_SHIFT));
 }
 
-static bool io_wqe_worker_send_sig(struct io_worker *worker, void *data)
-{
-	send_sig(SIGINT, worker->task, 1);
-	return false;
-}
-
-/*
- * Iterate the passed in list and call the specific function for each
- * worker that isn't exiting
- */
-static bool io_wq_for_each_worker(struct io_wqe *wqe,
-				  bool (*func)(struct io_worker *, void *),
-				  void *data)
-{
-	struct io_worker *worker;
-	bool ret = false;
-
-	list_for_each_entry_rcu(worker, &wqe->all_list, all_list) {
-		if (io_worker_get(worker)) {
-			/* no task if node is/was offline */
-			if (worker->task)
-				ret = func(worker, data);
-			io_worker_release(worker);
-			if (ret)
-				break;
-		}
-	}
-
-	return ret;
-}
-
 void io_wq_cancel_all(struct io_wq *wq)
 {
 	int node;

@@ -1117,12 +1123,6 @@ bool io_wq_get(struct io_wq *wq, struct
 	return refcount_inc_not_zero(&wq->use_refs);
 }
 
-static bool io_wq_worker_wake(struct io_worker *worker, void *data)
-{
-	wake_up_process(worker->task);
-	return false;
-}
-
 static void __io_wq_destroy(struct io_wq *wq)
 {
 	int node;

@@ -7536,7 +7536,6 @@ void __io_uring_free(struct task_struct
 	struct io_uring_task *tctx = tsk->io_uring;
 
 	WARN_ON_ONCE(!xa_empty(&tctx->xa));
-	xa_destroy(&tctx->xa);
 	kfree(tctx);
 	tsk->io_uring = NULL;
 }

@@ -8415,28 +8415,19 @@ static void io_uring_attempt_task_drop(s
 void __io_uring_files_cancel(struct files_struct *files)
 {
 	struct io_uring_task *tctx = current->io_uring;
-	XA_STATE(xas, &tctx->xa, 0);
+	struct file *file;
+	unsigned long index;
 
 	/* make sure overflow events are dropped */
 	tctx->in_idle = true;
 
-	do {
-		struct io_ring_ctx *ctx;
-		struct file *file;
-
-		xas_lock(&xas);
-		file = xas_next_entry(&xas, ULONG_MAX);
-		xas_unlock(&xas);
-
-		if (!file)
-			break;
-
-		ctx = file->private_data;
+	xa_for_each(&tctx->xa, index, file) {
+		struct io_ring_ctx *ctx = file->private_data;
 
 		io_uring_cancel_task_requests(ctx, files);
 		if (files)
 			io_uring_del_task_file(file);
-	} while (1);
+	}
 }
 
 static inline bool io_uring_task_idle(struct io_uring_task *tctx)

@@ -8336,27 +8336,24 @@ static void io_uring_cancel_task_request
  */
 static int io_uring_add_task_file(struct file *file)
 {
-	if (unlikely(!current->io_uring)) {
+	struct io_uring_task *tctx = current->io_uring;
+
+	if (unlikely(!tctx)) {
 		int ret;
 
 		ret = io_uring_alloc_task_context(current);
 		if (unlikely(ret))
 			return ret;
+		tctx = current->io_uring;
 	}
-	if (current->io_uring->last != file) {
-		XA_STATE(xas, &current->io_uring->xa, (unsigned long) file);
-		void *old;
+	if (tctx->last != file) {
+		void *old = xa_load(&tctx->xa, (unsigned long)file);
 
-		rcu_read_lock();
-		old = xas_load(&xas);
-		if (old != file) {
+		if (!old) {
 			get_file(file);
-			xas_lock(&xas);
-			xas_store(&xas, file);
-			xas_unlock(&xas);
+			xa_store(&tctx->xa, (unsigned long)file, file, GFP_KERNEL);
 		}
-		rcu_read_unlock();
-		current->io_uring->last = file;
+		tctx->last = file;
 	}
 
 	return 0;

@@ -8365,27 +8365,17 @@ static int io_uring_add_task_file(struct
 static void io_uring_del_task_file(struct file *file)
 {
 	struct io_uring_task *tctx = current->io_uring;
-	XA_STATE(xas, &tctx->xa, (unsigned long) file);
 
 	if (tctx->last == file)
 		tctx->last = NULL;
-
-	xas_lock(&xas);
-	file = xas_store(&xas, NULL);
-	xas_unlock(&xas);
-
+	file = xa_erase(&tctx->xa, (unsigned long)file);
 	if (file)
 		fput(file);
 }
 
 static void __io_uring_attempt_task_drop(struct file *file)
 {
-	XA_STATE(xas, &current->io_uring->xa, (unsigned long) file);
-	struct file *old;
-
-	rcu_read_lock();
-	old = xas_load(&xas);
-	rcu_read_unlock();
+	struct file *old = xa_load(&current->io_uring->xa, (unsigned long)file);
 
 	if (old == file)
 		io_uring_del_task_file(file);

@@ -62,10 +62,12 @@ efi_status_t handle_kernel_image(unsigne
 			status = efi_get_random_bytes(sizeof(phys_seed),
 						      (u8 *)&phys_seed);
 			if (status == EFI_NOT_FOUND) {
-				efi_info("EFI_RNG_PROTOCOL unavailable, no randomness supplied\n");
+				efi_info("EFI_RNG_PROTOCOL unavailable, KASLR will be disabled\n");
+				efi_nokaslr = true;
 			} else if (status != EFI_SUCCESS) {
-				efi_err("efi_get_random_bytes() failed\n");
-				return status;
+				efi_err("efi_get_random_bytes() failed (0x%lx), KASLR will be disabled\n",
+					status);
+				efi_nokaslr = true;
 			}
 		} else {
 			efi_info("KASLR disabled on kernel command line\n");

@@ -136,7 +136,7 @@ static efi_status_t update_fdt(void *ori
 	if (status)
 		goto fdt_set_fail;
 
-	if (IS_ENABLED(CONFIG_RANDOMIZE_BASE)) {
+	if (IS_ENABLED(CONFIG_RANDOMIZE_BASE) && !efi_nokaslr) {
 		efi_status_t efi_status;
 
 		efi_status = efi_get_random_bytes(sizeof(fdt_val64),

@@ -145,8 +145,6 @@ static efi_status_t update_fdt(void *ori
 			status = fdt_setprop_var(fdt, node, "kaslr-seed", fdt_val64);
 			if (status)
 				goto fdt_set_fail;
-		} else if (efi_status != EFI_NOT_FOUND) {
-			return efi_status;
 		}
 	}
 

@@ -17,7 +17,7 @@ int firmware_fallback_platform(struct fw
 	if (!(opt_flags & FW_OPT_FALLBACK_PLATFORM))
 		return -ENOENT;
 
-	rc = security_kernel_load_data(LOADING_FIRMWARE_EFI_EMBEDDED);
+	rc = security_kernel_load_data(LOADING_FIRMWARE);
 	if (rc)
 		return rc;
 

@@ -2862,7 +2862,6 @@ extern int do_pipe_flags(int *, int);
 	id(UNKNOWN, unknown)		\
 	id(FIRMWARE, firmware)		\
 	id(FIRMWARE_PREALLOC_BUFFER, firmware)	\
-	id(FIRMWARE_EFI_EMBEDDED, firmware)	\
 	id(MODULE, kernel-module)		\
 	id(KEXEC_IMAGE, kexec-image)		\
 	id(KEXEC_INITRAMFS, kexec-initramfs)	\

@@ -599,6 +599,12 @@ check_branch_predictor(const struct arm6
 	return (need_wa > 0);
 }
 
+static void
+cpu_enable_branch_predictor_hardening(const struct arm64_cpu_capabilities *cap)
+{
+	cap->matches(cap, SCOPE_LOCAL_CPU);
+}
+
 static const __maybe_unused struct midr_range tx2_family_cpus[] = {
 	MIDR_ALL_VERSIONS(MIDR_BRCM_VULCAN),
 	MIDR_ALL_VERSIONS(MIDR_CAVIUM_THUNDERX2),

@@ -890,9 +896,11 @@ const struct arm64_cpu_capabilities arm6
 	},
 #endif
 	{
+		.desc = "Branch predictor hardening",
 		.capability = ARM64_HARDEN_BRANCH_PREDICTOR,
 		.type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM,
 		.matches = check_branch_predictor,
+		.cpu_enable = cpu_enable_branch_predictor_hardening,
 	},
 #ifdef CONFIG_RANDOMIZE_BASE
 	{

@@ -457,6 +457,12 @@ out_printmsg:
 	return required;
 }
 
+static void cpu_enable_ssbd_mitigation(const struct arm64_cpu_capabilities *cap)
+{
+	if (ssbd_state != ARM64_SSBD_FORCE_DISABLE)
+		cap->matches(cap, SCOPE_LOCAL_CPU);
+}
+
 /* known invulnerable cores */
 static const struct midr_range arm64_ssb_cpus[] = {
 	MIDR_ALL_VERSIONS(MIDR_CORTEX_A35),

@@ -914,6 +920,7 @@ const struct arm64_cpu_capabilities arm6
 		.capability = ARM64_SSBD,
 		.type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM,
 		.matches = has_ssbd_mitigation,
+		.cpu_enable = cpu_enable_ssbd_mitigation,
 		.midr_range_list = arm64_ssb_cpus,
 	},
 #ifdef CONFIG_ARM64_ERRATUM_1418040

@@ -33,6 +33,7 @@
 #include <asm/hw_irq.h>
 #include <asm/io_apic.h>
 #include <asm/intel-mid.h>
+#include <asm/acpi.h>
 
 #define PCIE_CAP_OFFSET	0x100
 

@@ -53,9 +53,7 @@ void *__memmove(void *to, const void *fr
 #ifndef CONFIG_KASAN
 #define __HAVE_ARCH_MEMSET32
 #define __HAVE_ARCH_MEMSET64
-#define __HAVE_ARCH_MEMCPY_MCSAFE
 
-extern int memcpy_mcsafe(void *dst, const void *src, __kernel_size_t sz);
 extern void *__memset16(uint16_t *, uint16_t v, __kernel_size_t);
 extern void *__memset32(uint32_t *, uint32_t v, __kernel_size_t);
 extern void *__memset64(uint64_t *, uint64_t v, __kernel_size_t);

@@ -435,6 +435,32 @@ do {								\
 extern unsigned long __copy_tofrom_user(void __user *to,
 		const void __user *from, unsigned long size);
 
+#ifdef CONFIG_ARCH_HAS_COPY_MC
+unsigned long __must_check
+copy_mc_generic(void *to, const void *from, unsigned long size);
+
+static inline unsigned long __must_check
+copy_mc_to_kernel(void *to, const void *from, unsigned long size)
+{
+	return copy_mc_generic(to, from, size);
+}
+#define copy_mc_to_kernel copy_mc_to_kernel
+
+static inline unsigned long __must_check
+copy_mc_to_user(void __user *to, const void *from, unsigned long n)
+{
+	if (likely(check_copy_size(from, n, true))) {
+		if (access_ok(to, n)) {
+			allow_write_to_user(to, n);
+			n = copy_mc_generic((void *)to, from, n);
+			prevent_write_to_user(to, n);
+		}
+	}
+
+	return n;
+}
+#endif
+
 #ifdef __powerpc64__
 static inline unsigned long
 raw_copy_in_user(void __user *to, const void __user *from, unsigned long n)

@@ -523,20 +549,6 @@ raw_copy_to_user(void __user *to, const
 	return ret;
 }
 
-static __always_inline unsigned long __must_check
-copy_to_user_mcsafe(void __user *to, const void *from, unsigned long n)
-{
-	if (likely(check_copy_size(from, n, true))) {
-		if (access_ok(to, n)) {
-			allow_write_to_user(to, n);
-			n = memcpy_mcsafe((void *)to, from, n);
-			prevent_write_to_user(to, n);
-		}
-	}
-
-	return n;
-}
-
 unsigned long __arch_clear_user(void __user *addr, unsigned long size);
 
 static inline unsigned long clear_user(void __user *addr, unsigned long size)

75 lines not on this page · raw message

@@ -174,6 +174,15 @@ extern void mce_unregister_decode_chain(
 
 extern int mce_p5_enabled;
 
+#ifdef CONFIG_ARCH_HAS_COPY_MC
+extern void enable_copy_mc_fragile(void);
+unsigned long __must_check copy_mc_fragile(void *dst, const void *src, unsigned cnt);
+#else
+static inline void enable_copy_mc_fragile(void)
+{
+}
+#endif
+
 #ifdef CONFIG_X86_MCE
 int mcheck_init(void);
 void mcheck_cpu_init(struct cpuinfo_x86 *c);

75 lines not on this page · raw message

@@ -82,38 +82,6 @@ int strcmp(const char *cs, const char *c
 
 #endif
 
-#define __HAVE_ARCH_MEMCPY_MCSAFE 1
-__must_check unsigned long __memcpy_mcsafe(void *dst, const void *src,
-		size_t cnt);
-DECLARE_STATIC_KEY_FALSE(mcsafe_key);
-
-/**
- * memcpy_mcsafe - copy memory with indication if a machine check happened
- *
- * @dst:	destination address
- * @src:	source address
- * @cnt:	number of bytes to copy
- *
- * Low level memory copy function that catches machine checks
- * We only call into the "safe" function on systems that can
- * actually do machine check recovery. Everyone else can just
- * use memcpy().
- *
- * Return 0 for success, or number of bytes not copied if there was an
- * exception.
- */
-static __always_inline __must_check unsigned long
-memcpy_mcsafe(void *dst, const void *src, size_t cnt)
-{
-#ifdef CONFIG_X86_MCE
-	if (static_branch_unlikely(&mcsafe_key))
-		return __memcpy_mcsafe(dst, src, cnt);
-	else
-#endif
-		memcpy(dst, src, cnt);
-	return 0;
-}
-
 #ifdef CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE
 #define __HAVE_ARCH_MEMCPY_FLUSHCACHE 1
 void __memcpy_flushcache(void *dst, const void *src, size_t cnt);

@@ -455,6 +455,15 @@ extern __must_check long strnlen_user(co
 unsigned long __must_check clear_user(void __user *mem, unsigned long len);
 unsigned long __must_check __clear_user(void __user *mem, unsigned long len);
 
+#ifdef CONFIG_ARCH_HAS_COPY_MC
+unsigned long __must_check
+copy_mc_to_kernel(void *to, const void *from, unsigned len);
+#define copy_mc_to_kernel copy_mc_to_kernel
+
+unsigned long __must_check
+copy_mc_to_user(void *to, const void *from, unsigned len);
+#endif
+
 /*
  * movsl can be slow when source and dest are not both 8-byte aligned
  */

@@ -47,22 +47,6 @@ copy_user_generic(void *to, const void *
 }
 
 static __always_inline __must_check unsigned long
-copy_to_user_mcsafe(void *to, const void *from, unsigned len)
-{
-	unsigned long ret;
-
-	__uaccess_begin();
-	/*
-	 * Note, __memcpy_mcsafe() is explicitly used since it can
-	 * handle exceptions / faults.  memcpy_mcsafe() may fall back to
-	 * memcpy() which lacks this handling.
-	 */
-	ret = __memcpy_mcsafe(to, from, len);
-	__uaccess_end();
-	return ret;
-}
-
-static __always_inline __must_check unsigned long
 raw_copy_from_user(void *dst, const void __user *src, unsigned long size)
 {
 	return copy_user_generic(dst, (__force void *)src, size);

@@ -102,8 +86,4 @@ __copy_from_user_flushcache(void *dst, c
 	kasan_check_write(dst, size);
 	return __copy_user_flushcache(dst, src, size);
 }
-
-unsigned long
-mcsafe_handle_tail(char *to, char *from, unsigned len);
-
 #endif /* _ASM_X86_UACCESS_64_H */

@@ -40,7 +40,6 @@
 #include <linux/debugfs.h>
 #include <linux/irq_work.h>
 #include <linux/export.h>
-#include <linux/jump_label.h>
 #include <linux/set_memory.h>
 #include <linux/sync_core.h>
 #include <linux/task_work.h>

@@ -2127,7 +2126,7 @@ void mce_disable_bank(int bank)
 	and older.
  * mce=nobootlog Don't log MCEs from before booting.
  * mce=bios_cmci_threshold Don't program the CMCI threshold
- * mce=recovery force enable memcpy_mcsafe()
+ * mce=recovery force enable copy_mc_fragile()
  */
 static int __init mcheck_enable(char *str)
 {

@@ -2735,13 +2734,10 @@ static void __init mcheck_debugfs_init(v
 static void __init mcheck_debugfs_init(void) { }
 #endif
 
-DEFINE_STATIC_KEY_FALSE(mcsafe_key);
-EXPORT_SYMBOL_GPL(mcsafe_key);
-
 static int __init mcheck_late_init(void)
 {
 	if (mca_cfg.recovery)
-		static_branch_inc(&mcsafe_key);
+		enable_copy_mc_fragile();
 
 	mcheck_debugfs_init();
 

@@ -8,6 +8,7 @@
 
 #include <asm/hpet.h>
 #include <asm/setup.h>
+#include <asm/mce.h>
 
 #if defined(CONFIG_X86_IO_APIC) && defined(CONFIG_SMP) && defined(CONFIG_PCI)
 

@@ -624,10 +625,6 @@ static void amd_disable_seq_and_redirect
 DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_NB_F3,
 			amd_disable_seq_and_redirect_scrub);
 
-#if defined(CONFIG_X86_64) && defined(CONFIG_X86_MCE)
-#include <linux/jump_label.h>
-#include <asm/string_64.h>
-
 /* Ivy Bridge, Haswell, Broadwell */
 static void quirk_intel_brickland_xeon_ras_cap(struct pci_dev *pdev)
 {

@@ -636,7 +633,7 @@ static void quirk_intel_brickland_xeon_r
 	pci_read_config_dword(pdev, 0x84, &capid0);
 
 	if (capid0 & 0x10)
-		static_branch_inc(&mcsafe_key);
+		enable_copy_mc_fragile();
 }
 
 /* Skylake */

@@ -653,7 +650,7 @@ static void quirk_intel_purley_xeon_ras_
 	 * enabled, so memory machine check recovery is also enabled.
 	 */
 	if ((capid0 & 0xc0) == 0xc0 || (capid5 & 0x1e0))
-		static_branch_inc(&mcsafe_key);
+		enable_copy_mc_fragile();
 
 }
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x0ec3, quirk_intel_brickland_xeon_ras_cap);

@@ -661,7 +658,6 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_IN
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x6fc0, quirk_intel_brickland_xeon_ras_cap);
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x2083, quirk_intel_purley_xeon_ras_cap);
 #endif
-#endif
 
 bool x86_apple_machine;
 EXPORT_SYMBOL(x86_apple_machine);

82 lines not on this page · raw message

@@ -56,27 +56,6 @@ unsigned long clear_user(void __user *to
 }
 EXPORT_SYMBOL(clear_user);
 
-/*
- * Similar to copy_user_handle_tail, probe for the write fault point,
- * but reuse __memcpy_mcsafe in case a new read error is encountered.
- * clac() is handled in _copy_to_iter_mcsafe().
- */
-__visible notrace unsigned long
-mcsafe_handle_tail(char *to, char *from, unsigned len)
-{
-	for (; len; --len, to++, from++) {
-		/*
-		 * Call the assembly routine back directly since
-		 * memcpy_mcsafe() may silently fallback to memcpy.
-		 */
-		unsigned long rem = __memcpy_mcsafe(to, from, 1);
-
-		if (rem)
-			break;
-	}
-	return len;
-}
-
 #ifdef CONFIG_ARCH_HAS_UACCESS_FLUSHCACHE
 /**
  * clean_cache_range - write back a cache range with CLWB

@@ -49,7 +49,7 @@ do {								\
 #define pmem_assign(dest, src)	((dest) = (src))
 #endif
 
-#if defined(__HAVE_ARCH_MEMCPY_MCSAFE) && defined(DM_WRITECACHE_HAS_PMEM)
+#if IS_ENABLED(CONFIG_ARCH_HAS_COPY_MC) && defined(DM_WRITECACHE_HAS_PMEM)
 #define DM_WRITECACHE_HANDLE_HARDWARE_ERRORS
 #endif
 

@@ -992,7 +992,8 @@ static void writecache_resume(struct dm_
 	}
 	wc->freelist_size = 0;
 
-	r = memcpy_mcsafe(&sb_seq_count, &sb(wc)->seq_count, sizeof(uint64_t));
+	r = copy_mc_to_kernel(&sb_seq_count, &sb(wc)->seq_count,
+			      sizeof(uint64_t));
 	if (r) {
 		writecache_error(wc, r, "hardware memory error when reading superblock: %d", r);
 		sb_seq_count = cpu_to_le64(0);

@@ -1008,7 +1009,8 @@ static void writecache_resume(struct dm_
 			e->seq_count = -1;
 			continue;
 		}
-		r = memcpy_mcsafe(&wme, memory_entry(wc, e), sizeof(struct wc_memory_entry));
+		r = copy_mc_to_kernel(&wme, memory_entry(wc, e),
+				      sizeof(struct wc_memory_entry));
 		if (r) {
 			writecache_error(wc, r, "hardware memory error when reading metadata entry %lu: %d",
 					 (unsigned long)b, r);

@@ -1206,7 +1208,7 @@ static void bio_copy_block(struct dm_wri
 
 		if (rw == READ) {
 			int r;
-			r = memcpy_mcsafe(buf, data, size);
+			r = copy_mc_to_kernel(buf, data, size);
 			flush_dcache_page(bio_page(bio));
 			if (unlikely(r)) {
 				writecache_error(wc, r, "hardware memory error when reading data: %d", r);

@@ -2349,7 +2351,7 @@ invalid_optional:
 		}
 	}
 
-	r = memcpy_mcsafe(&s, sb(wc), sizeof(struct wc_memory_superblock));
+	r = copy_mc_to_kernel(&s, sb(wc), sizeof(struct wc_memory_superblock));
 	if (r) {
 		ti->error = "Hardware memory error when reading superblock";
 		goto bad;

@@ -2360,7 +2362,8 @@ invalid_optional:
 			ti->error = "Unable to initialize device";
 			goto bad;
 		}
-		r = memcpy_mcsafe(&s, sb(wc), sizeof(struct wc_memory_superblock));
+		r = copy_mc_to_kernel(&s, sb(wc),
+				      sizeof(struct wc_memory_superblock));
 		if (r) {
 			ti->error = "Hardware memory error when reading superblock";
 			goto bad;

@@ -268,7 +268,7 @@ static int nsio_rw_bytes(struct nd_names
 	if (rw == READ) {
 		if (unlikely(is_bad_pmem(&nsio->bb, sector, sz_align)))
 			return -EIO;
-		if (memcpy_mcsafe(buf, nsio->addr + offset, size) != 0)
+		if (copy_mc_to_kernel(buf, nsio->addr + offset, size) != 0)
 			return -EIO;
 		return 0;
 	}

@@ -125,7 +125,7 @@ static blk_status_t read_pmem(struct pag
 	while (len) {
 		mem = kmap_atomic(page);
 		chunk = min_t(unsigned int, len, PAGE_SIZE - off);
-		rem = memcpy_mcsafe(mem + off, pmem_addr, chunk);
+		rem = copy_mc_to_kernel(mem + off, pmem_addr, chunk);
 		kunmap_atomic(mem);
 		if (rem)
 			return BLK_STS_IOERR;

@@ -304,7 +304,7 @@ static long pmem_dax_direct_access(struc
 
 /*
  * Use the 'no check' versions of copy_from_iter_flushcache() and
- * copy_to_iter_mcsafe() to bypass HARDENED_USERCOPY overhead. Bounds
+ * copy_mc_to_iter() to bypass HARDENED_USERCOPY overhead. Bounds
  * checking, both file offset and device offset, is handled by
  * dax_iomap_actor()
  */

@@ -317,7 +317,7 @@ static size_t pmem_copy_from_iter(struct
 static size_t pmem_copy_to_iter(struct dax_device *dax_dev, pgoff_t pgoff,
 		void *addr, size_t bytes, struct iov_iter *i)
 {
-	return _copy_to_iter_mcsafe(addr, bytes, i);
+	return _copy_mc_to_iter(addr, bytes, i);
 }
 
 static const struct dax_operations pmem_dax_ops = {

@@ -161,20 +161,13 @@ extern int bcmp(const void *,const void
 #ifndef __HAVE_ARCH_MEMCHR
 extern void * memchr(const void *,int,__kernel_size_t);
 #endif
-#ifndef __HAVE_ARCH_MEMCPY_MCSAFE
-static inline __must_check unsigned long memcpy_mcsafe(void *dst,
-		const void *src, size_t cnt)
-{
-	memcpy(dst, src, cnt);
-	return 0;
-}
-#endif
 #ifndef __HAVE_ARCH_MEMCPY_FLUSHCACHE
 static inline void memcpy_flushcache(void *dst, const void *src, size_t cnt)
 {
 	memcpy(dst, src, cnt);
 }
 #endif
+
 void *memchr_inv(const void *s, int c, size_t n);
 char *strreplace(char *s, char old, char new);
 

@@ -179,6 +179,19 @@ copy_in_user(void __user *to, const void
 }
 #endif
 
+#ifndef copy_mc_to_kernel
+/*
+ * Without arch opt-in this generic copy_mc_to_kernel() will not handle
+ * #MC (or arch equivalent) during source read.
+ */
+static inline unsigned long __must_check
+copy_mc_to_kernel(void *dst, const void *src, size_t cnt)
+{
+	memcpy(dst, src, cnt);
+	return 0;
+}
+#endif
+
 static __always_inline void pagefault_disabled_inc(void)
 {
 	current->pagefault_disabled++;

@@ -185,10 +185,10 @@ size_t _copy_from_iter_flushcache(void *
 #define _copy_from_iter_flushcache _copy_from_iter_nocache
 #endif
 
-#ifdef CONFIG_ARCH_HAS_UACCESS_MCSAFE
-size_t _copy_to_iter_mcsafe(const void *addr, size_t bytes, struct iov_iter *i);
+#ifdef CONFIG_ARCH_HAS_COPY_MC
+size_t _copy_mc_to_iter(const void *addr, size_t bytes, struct iov_iter *i);
 #else
-#define _copy_to_iter_mcsafe _copy_to_iter
+#define _copy_mc_to_iter _copy_to_iter
 #endif
 
 static __always_inline __must_check

@@ -201,12 +201,12 @@ size_t copy_from_iter_flushcache(void *a
 }
 
 static __always_inline __must_check
-size_t copy_to_iter_mcsafe(void *addr, size_t bytes, struct iov_iter *i)
+size_t copy_mc_to_iter(void *addr, size_t bytes, struct iov_iter *i)
 {
 	if (unlikely(!check_copy_size(addr, bytes, true)))
 		return 0;
 	else
-		return _copy_to_iter_mcsafe(addr, bytes, i);
+		return _copy_mc_to_iter(addr, bytes, i);
 }
 
 size_t iov_iter_zero(size_t bytes, struct iov_iter *);

117 lines not on this page · raw message

@@ -1,13 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _MCSAFE_TEST_H_
-#define _MCSAFE_TEST_H_
-
-.macro MCSAFE_TEST_CTL
-.endm
-
-.macro MCSAFE_TEST_SRC reg count target
-.endm
-
-.macro MCSAFE_TEST_DST reg count target
-.endm
-#endif /* _MCSAFE_TEST_H_ */

@@ -548,8 +548,8 @@ static const char *uaccess_safe_builtin[
 	"__ubsan_handle_shift_out_of_bounds",
 	/* misc */
 	"csum_partial_copy_generic",
-	"__memcpy_mcsafe",
-	"mcsafe_handle_tail",
+	"copy_mc_fragile",
+	"copy_mc_fragile_handle_tail",
 	"ftrace_likely_update", /* CONFIG_TRACE_BRANCH_PROFILING */
 	NULL
 };

@@ -1,24 +0,0 @@
-/*
- * From code in arch/x86/lib/usercopy_64.c, copied to keep tools/ copy
- * of the kernel's arch/x86/lib/memcpy_64.s used in 'perf bench mem memcpy'
- * happy.
- */
-#include <linux/types.h>
-
-unsigned long __memcpy_mcsafe(void *dst, const void *src, size_t cnt);
-unsigned long mcsafe_handle_tail(char *to, char *from, unsigned len);
-
-unsigned long mcsafe_handle_tail(char *to, char *from, unsigned len)
-{
-	for (; len; --len, to++, from++) {
-		/*
-		 * Call the assembly routine back directly since
-		 * memcpy_mcsafe() may silently fallback to memcpy.
-		 */
-		unsigned long rem = __memcpy_mcsafe(to, from, 1);
-
-		if (rem)
-			break;
-	}
-	return len;
-}

111 lines not on this page · raw message

@@ -45,6 +45,8 @@ void enable_copy_mc_fragile(void)
 #define copy_mc_fragile_enabled (0)
 #endif
 
+unsigned long copy_mc_enhanced_fast_string(void *dst, const void *src, unsigned len);
+
 /**
  * copy_mc_to_kernel - memory copy that handles source exceptions
  *

@@ -52,9 +54,11 @@ void enable_copy_mc_fragile(void)
  * @src:	source address
  * @len:	number of bytes to copy
  *
- * Call into the 'fragile' version on systems that have trouble
- * actually do machine check recovery. Everyone else can just
- * use memcpy().
+ * Call into the 'fragile' version on systems that benefit from avoiding
+ * corner case poison consumption scenarios, For example, accessing
+ * poison across 2 cachelines with a single instruction. Almost all
+ * other uses case can use copy_mc_enhanced_fast_string() for a fast
+ * recoverable copy, or fallback to plain memcpy.
  *
  * Return 0 for success, or number of bytes not copied if there was an
  * exception.

@@ -63,6 +67,8 @@ unsigned long __must_check copy_mc_to_ke
 {
 	if (copy_mc_fragile_enabled)
 		return copy_mc_fragile(dst, src, len);
+	if (static_cpu_has(X86_FEATURE_ERMS))
+		return copy_mc_enhanced_fast_string(dst, src, len);
 	memcpy(dst, src, len);
 	return 0;
 }

@@ -72,11 +78,19 @@ unsigned long __must_check copy_mc_to_us
 {
 	unsigned long ret;
 
-	if (!copy_mc_fragile_enabled)
-		return copy_user_generic(dst, src, len);
+	if (copy_mc_fragile_enabled) {
+		__uaccess_begin();
+		ret = copy_mc_fragile(dst, src, len);
+		__uaccess_end();
+		return ret;
+	}
+
+	if (static_cpu_has(X86_FEATURE_ERMS)) {
+		__uaccess_begin();
+		ret = copy_mc_enhanced_fast_string(dst, src, len);
+		__uaccess_end();
+		return ret;
+	}
 
-	__uaccess_begin();
-	ret = copy_mc_fragile(dst, src, len);
-	__uaccess_end();
-	return ret;
+	return copy_user_generic(dst, src, len);
 }

@@ -550,6 +550,7 @@ static const char *uaccess_safe_builtin[
 	"csum_partial_copy_generic",
 	"copy_mc_fragile",
 	"copy_mc_fragile_handle_tail",
+	"copy_mc_enhanced_fast_string",
 	"ftrace_likely_update", /* CONFIG_TRACE_BRANCH_PROFILING */
 	NULL
 };

@@ -141,6 +141,9 @@ static int efivarfs_callback(efi_char16_
 
 	name[len + EFI_VARIABLE_GUID_LEN+1] = '\0';
 
+	/* replace invalid slashes like kobject_set_name_vargs does for /sys/firmware/efi/vars. */
+	strreplace(name, '/', '!');
+
 	inode = efivarfs_get_inode(sb, d_inode(root), S_IFREG | 0644, 0,
 				   is_removable);
 	if (!inode)

@@ -9566,7 +9566,10 @@ int bnxt_open_nic(struct bnxt *bp, bool
 {
 	int rc = 0;
 
-	rc = __bnxt_open_nic(bp, irq_re_init, link_re_init);
+	if (test_bit(BNXT_STATE_ABORT_ERR, &bp->state))
+		rc = -EIO;
+	if (!rc)
+		rc = __bnxt_open_nic(bp, irq_re_init, link_re_init);
 	if (rc) {
 		netdev_err(bp->dev, "nic open fail (rc: %x)\n", rc);
 		dev_close(bp->dev);

@@ -11790,15 +11790,16 @@ static void bnxt_remove_one(struct pci_d
 	if (BNXT_PF(bp))
 		bnxt_sriov_disable(bp);
 
+	if (BNXT_PF(bp))
+		devlink_port_type_clear(&bp->dl_port);
+	pci_disable_pcie_error_reporting(pdev);
+	unregister_netdev(dev);
 	clear_bit(BNXT_STATE_IN_FW_RESET, &bp->state);
+	/* Flush any pending tasks */
 	bnxt_cancel_sp_work(bp);
 	bp->sp_event = 0;
 
 	bnxt_dl_fw_reporters_destroy(bp, true);
-	if (BNXT_PF(bp))
-		devlink_port_type_clear(&bp->dl_port);
-	pci_disable_pcie_error_reporting(pdev);
-	unregister_netdev(dev);
 	bnxt_dl_unregister(bp);
 	bnxt_shutdown_tc(bp);
 

@@ -1158,16 +1158,6 @@ static void bnxt_queue_sp_work(struct bn
 		schedule_work(&bp->sp_task);
 }
 
-static void bnxt_cancel_sp_work(struct bnxt *bp)
-{
-	if (BNXT_PF(bp)) {
-		flush_workqueue(bnxt_pf_wq);
-	} else {
-		cancel_work_sync(&bp->sp_task);
-		cancel_delayed_work_sync(&bp->fw_reset_task);
-	}
-}
-
 static void bnxt_sched_reset(struct bnxt *bp, struct bnxt_rx_ring_info *rxr)
 {
 	if (!rxr->bnapi->in_reset) {

@@ -11796,7 +11786,8 @@ static void bnxt_remove_one(struct pci_d
 	unregister_netdev(dev);
 	clear_bit(BNXT_STATE_IN_FW_RESET, &bp->state);
 	/* Flush any pending tasks */
-	bnxt_cancel_sp_work(bp);
+	cancel_work_sync(&bp->sp_task);
+	cancel_delayed_work_sync(&bp->fw_reset_task);
 	bp->sp_event = 0;
 
 	bnxt_dl_fw_reporters_destroy(bp, true);

@@ -12530,6 +12530,9 @@ static pci_ers_result_t bnxt_io_error_de
 		return PCI_ERS_RESULT_DISCONNECT;
 	}
 
+	if (state == pci_channel_io_frozen)
+		set_bit(BNXT_STATE_PCI_CHANNEL_IO_FROZEN, &bp->state);
+
 	if (netif_running(netdev))
 		bnxt_close(netdev);
 

@@ -12556,7 +12559,7 @@ static pci_ers_result_t bnxt_io_slot_res
 {
 	struct net_device *netdev = pci_get_drvdata(pdev);
 	struct bnxt *bp = netdev_priv(netdev);
-	int err = 0;
+	int err = 0, off;
 	pci_ers_result_t result = PCI_ERS_RESULT_DISCONNECT;
 
 	netdev_info(bp->dev, "PCI Slot Reset\n");

@@ -12568,6 +12571,20 @@ static pci_ers_result_t bnxt_io_slot_res
 			"Cannot re-enable PCI device after reset.\n");
 	} else {
 		pci_set_master(pdev);
+		/* Upon fatal error, our device internal logic that latches to
+		 * BAR value is getting reset and will restore only upon
+		 * rewritting the BARs.
+		 *
+		 * As pci_restore_state() does not re-write the BARs if the
+		 * value is same as saved value earlier, driver needs to
+		 * write the BARs to 0 to force restore, in case of fatal error.
+		 */
+		if (test_and_clear_bit(BNXT_STATE_PCI_CHANNEL_IO_FROZEN,
+				       &bp->state)) {
+			for (off = PCI_BASE_ADDRESS_0;
+			     off <= PCI_BASE_ADDRESS_5; off += 4)
+				pci_write_config_dword(bp->pdev, off, 0);
+		}
 		pci_restore_state(pdev);
 		pci_save_state(pdev);
 

@@ -1736,6 +1736,7 @@ struct bnxt {
 #define BNXT_STATE_ABORT_ERR	5
 #define BNXT_STATE_FW_FATAL_COND	6
 #define BNXT_STATE_DRV_REGISTERED	7
+#define BNXT_STATE_PCI_CHANNEL_IO_FROZEN	8
 
 #define BNXT_NO_FW_ACCESS(bp)					\
 	(test_bit(BNXT_STATE_FW_FATAL_COND, &(bp)->state) ||	\

@@ -4296,7 +4296,8 @@ static int bnxt_hwrm_do_send_msg(struct
 	u32 bar_offset = BNXT_GRCPF_REG_CHIMP_COMM;
 	u16 dst = BNXT_HWRM_CHNL_CHIMP;
 
-	if (BNXT_NO_FW_ACCESS(bp))
+	if (BNXT_NO_FW_ACCESS(bp) &&
+	    le16_to_cpu(req->req_type) != HWRM_FUNC_RESET)
 		return -EBUSY;
 
 	if (msg_len > BNXT_HWRM_MAX_REQ_LEN) {

@@ -1514,7 +1514,6 @@ static void add_to_reap_list(struct sock
 	struct chtls_sock *csk = sk->sk_user_data;
 
 	local_bh_disable();
-	bh_lock_sock(sk);
 	release_tcp_port(sk); /* release the port immediately */
 
 	spin_lock(&reap_list_lock);

@@ -1523,7 +1522,6 @@ static void add_to_reap_list(struct sock
 	if (!csk->passive_reap_next)
 		schedule_work(&reap_task);
 	spin_unlock(&reap_list_lock);
-	bh_unlock_sock(sk);
 	local_bh_enable();
 }
 

@@ -772,14 +772,13 @@ static int chtls_pass_open_rpl(struct ch
 	if (rpl->status != CPL_ERR_NONE) {
 		pr_info("Unexpected PASS_OPEN_RPL status %u for STID %u\n",
 			rpl->status, stid);
-		return CPL_RET_BUF_DONE;
+	} else {
+		cxgb4_free_stid(cdev->tids, stid, listen_ctx->lsk->sk_family);
+		sock_put(listen_ctx->lsk);
+		kfree(listen_ctx);
+		module_put(THIS_MODULE);
 	}
-	cxgb4_free_stid(cdev->tids, stid, listen_ctx->lsk->sk_family);
-	sock_put(listen_ctx->lsk);
-	kfree(listen_ctx);
-	module_put(THIS_MODULE);
-
-	return 0;
+	return CPL_RET_BUF_DONE;
 }
 
 static int chtls_close_listsrv_rpl(struct chtls_dev *cdev, struct sk_buff *skb)

@@ -796,15 +795,13 @@ static int chtls_close_listsrv_rpl(struc
 	if (rpl->status != CPL_ERR_NONE) {
 		pr_info("Unexpected CLOSE_LISTSRV_RPL status %u for STID %u\n",
 			rpl->status, stid);
-		return CPL_RET_BUF_DONE;
+	} else {
+		cxgb4_free_stid(cdev->tids, stid, listen_ctx->lsk->sk_family);
+		sock_put(listen_ctx->lsk);
+		kfree(listen_ctx);
+		module_put(THIS_MODULE);
 	}
-
-	cxgb4_free_stid(cdev->tids, stid, listen_ctx->lsk->sk_family);
-	sock_put(listen_ctx->lsk);
-	kfree(listen_ctx);
-	module_put(THIS_MODULE);
-
-	return 0;
+	return CPL_RET_BUF_DONE;
 }
 
 static void chtls_purge_wr_queue(struct sock *sk)

@@ -1585,6 +1585,7 @@ skip_copy:
 			tp->urg_data = 0;
 
 		if ((avail + offset) >= skb->len) {
+			struct sk_buff *next_skb;
 			if (ULP_SKB_CB(skb)->flags & ULPCB_FLAG_TLS_HDR) {
 				tp->copied_seq += skb->len;
 				hws->rcvpld = skb->hdr_len;

@@ -1595,8 +1596,10 @@ skip_copy:
 			chtls_free_skb(sk, skb);
 			buffers_freed++;
 			hws->copied_seq = 0;
-			if (copied >= target &&
-			    !skb_peek(&sk->sk_receive_queue))
+			next_skb = skb_peek(&sk->sk_receive_queue);
+			if (copied >= target && !next_skb)
+				break;
+			if (ULP_SKB_CB(next_skb)->flags & ULPCB_FLAG_TLS_HDR)
 				break;
 		}
 	} while (len > 0);

@@ -145,13 +145,13 @@ static int configure_filter_smac(struct
 	int err;
 
 	/* do a set-tcb for smac-sel and CWR bit.. */
-	err = set_tcb_tflag(adap, f, f->tid, TF_CCTRL_CWR_S, 1, 1);
-	if (err)
-		goto smac_err;
-
 	err = set_tcb_field(adap, f, f->tid, TCB_SMAC_SEL_W,
 			    TCB_SMAC_SEL_V(TCB_SMAC_SEL_M),
 			    TCB_SMAC_SEL_V(f->smt->idx), 1);
+	if (err)
+		goto smac_err;
+
+	err = set_tcb_tflag(adap, f, f->tid, TF_CCTRL_CWR_S, 1, 1);
 	if (!err)
 		return 0;
 

@@ -865,6 +865,7 @@ int set_filter_wr(struct adapter *adapte
 		      FW_FILTER_WR_DIRSTEERHASH_V(f->fs.dirsteerhash) |
 		      FW_FILTER_WR_LPBK_V(f->fs.action == FILTER_SWITCH) |
 		      FW_FILTER_WR_DMAC_V(f->fs.newdmac) |
+		      FW_FILTER_WR_SMAC_V(f->fs.newsmac) |
 		      FW_FILTER_WR_INSVLAN_V(f->fs.newvlan == VLAN_INSERT ||
 					     f->fs.newvlan == VLAN_REWRITE) |
 		      FW_FILTER_WR_RMVLAN_V(f->fs.newvlan == VLAN_REMOVE ||

@@ -882,7 +883,7 @@ int set_filter_wr(struct adapter *adapte
 		 FW_FILTER_WR_OVLAN_VLD_V(f->fs.val.ovlan_vld) |
 		 FW_FILTER_WR_IVLAN_VLDM_V(f->fs.mask.ivlan_vld) |
 		 FW_FILTER_WR_OVLAN_VLDM_V(f->fs.mask.ovlan_vld));
-	fwr->smac_sel = 0;
+	fwr->smac_sel = f->smt->idx;
 	fwr->rx_chan_rx_rpl_iq =
 		htons(FW_FILTER_WR_RX_CHAN_V(0) |
 		      FW_FILTER_WR_RX_RPL_IQ_V(adapter->sge.fw_evtq.abs_id));