Merge tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-tracegrafted

Pull tracing fixes from Steven Rostedt:

 - Fix error output of boot instance creation failure

   Currently if a boot instance creation fails, instead of printing out
   the name of the instance that failed, it prints "(null)". That is
   because it prints "cur_str" that had already been processed by
   strsep(). Print the saved name instead.

   While at it, print the error code of the failure.

 - Fix use-after-free for same named historgrams

   Histograms can be named so that they can be used in multiple events.
   But if the named histogram has a variable attached, the second event
   that uses the named histogram which duplicates it and needs to free
   the original after duplication leaves the old variable in place and
   still visible. If another histogram uses than variable, it will use
   the stale one which will try to reference the freed duplicate
   histogram and crash the kernel.

   Free the duplicate variables along with the duplicated histogram
   data.

 - Check return value of kthread_run() in event self test

   The events self tests uses a kthread for testing but does not check
   if it succeeded in creating a kthread. If the kthread creation were
   to fail, the code will still try to call kthread_stop() on the error
   returned.

 - Fix race between reading trace_pipe and updating subbuffer size

   If a user is reading the trace_pipe file at the same time they update
   the ring buffer sub-buffer size, can cause the trace_pipe read to
   read stale data. Add trace_access_lock() around updating the ring
   buffer sub-buffer size.

 - Fix eventfs_inode on failure path in creation of the events directory

   In the creation of the "events" directory, if after allocating the
   eventfs_inode a failure is detected, it calls cleanup_ei() which
   calls free_ei(). The free_ei() will test if eventfs_inode being freed
   has no children. It is a bug if it does. But on the failure case of
   the creation of the "events" directory, the children lists have not
   yet been initialized and the free will trigger a warning because
   list_empty() on an uninitialized list returns false.

   Move the initialization into init_ei() where it makes more sense and
   makes sure that a created eventfs_inode has its lists initialized
   upon creation.

 - Check return value of kthread_run() in ftrace direct sample code

   The sample code that shows how to use the ftrace direct calls does
   not test the return of kthread_run() to see if it succeeds. Return a
   failure if the kthread_run() doesn't succeed.

 - Clear user events state on fork in case of alloc failure

   On fork, the child gets a pointer to the parent's user events state.
   It makes a copy of it then updates the child's pointer to it. But if
   the allocation fails, the duplication function leaves the child with
   a pointer to its parent's descriptor. When the child cleans up its
   data, it will free the parent's descriptor while the parent is still
   using it.

   In the duplication function, set the child's user_event_mm to NULL
   before testing if the allocation succeeded, and when it exits it will
   not free the parent's descriptor.

 - Fix retry exhaustion in simple ring buffer reader swap

   simple_ring_buffer_swap_reader_page() starts with retry set to 8 and
   post-decrements it only after a failed link replacement. On the final
   attempt, a successful replacement leaves retry at zero, while a
   failed replacement leaves it at -1.

   But the check for success expects the retry value to be non-zero and
   exits with an error on zero. This is the opposite result. Fix it.

 - Fail nicely when the remote swap_reader_page() returns an error

   Currently, if the swap_reader_page() of a remote buffer fails, it
   triggers a WARN_ON_ONCE() and continues normally. Instead, have it
   exit with an error and a pr_warn() print instead of a full WARNING.

* tag 'trace-v7.3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  ring-buffer: Stop remote reader update when page swap fails
  tracing: Fix retry exhaustion in simple ring buffer reader swap
  tracing/user_events: Clear copied tracing state before fork duplication
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-multi-modify
  samples/ftrace: Fix kthread_stop() on ERR_PTR in ftrace-direct-modify
  eventfs: Initialize ei->children and ei->list in init_ei()
  tracing: Fix use-after-free in trace_pipe read on sub-buffer order change
  tracing: Fix crash passing ERR_PTR to kthread_stop()
  tracing: Fix use-after-free with same-name named triggers
  tracing: Fix logged instance name on creation failure

1 files changed, 372 insertions, 0 deletions

diff --git a/fs/smb/client/compress.c b/fs/smb/client/compress.c
new file mode 100644
index 000000000..8f0860970
--- /dev/null
+++ b/fs/smb/client/compress.c
@@ -0,0 +1,372 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2024, SUSE LLC
+ *
+ * Authors: Enzo Matsumiya <ematsumiya@suse.de>
+ *
+ * This file implements I/O compression support for SMB2 messages (SMB 3.1.1 only).
+ * See compress/ for implementation details of each algorithm.
+ *
+ * References:
+ * MS-SMB2 "3.1.4.4 Compressing the Message"
+ * MS-SMB2 "3.1.5.3 Decompressing the Chained Message"
+ * MS-XCA - for details of the supported algorithms
+ */
+#include <linux/slab.h>
+#include <linux/kernel.h>
+#include <linux/uio.h>
+#include <linux/sort.h>
+
+#include "cifsglob.h"
+#include "../common/smb2pdu.h"
+#include "cifsproto.h"
+#include "smb2proto.h"
+
+#include "../common/compress/lz77.h"
+#include "compress.h"
+
+/*
+ * The heuristic_*() functions below try to determine data compressibility.
+ *
+ * Derived from fs/btrfs/compression.c, changing coding style, some parameters, and removing
+ * unused parts.
+ *
+ * Read that file for better and more detailed explanation of the calculations.
+ *
+ * The algorithms are ran in a collected sample of the input (uncompressed) data.
+ * The sample is formed of 2K reads in PAGE_SIZE intervals, with a maximum size of 4M.
+ *
+ * Parsing the sample goes from "low-hanging fruits" (fastest algorithms, likely compressible)
+ * to "need more analysis" (likely uncompressible).
+ */
+
+struct bucket {
+	unsigned int count;
+};
+
+static inline size_t pow4(size_t n)
+{
+	return n * n * n * n;
+}
+
+/*
+ * has_low_entropy() - Compute Shannon entropy of the sampled data.
+ * @bkt:	Bytes counts of the sample.
+ * @slen:	Size of the sample.
+ *
+ * Return: true if the level (percentage of number of bits that would be required to
+ *	   compress the data) is below the minimum threshold.
+ *
+ * Note:
+ * There _is_ an entropy level here that's > 65 (minimum threshold) that would indicate a
+ * possibility of compression, but compressing, or even further analysing, it would waste so much
+ * resources that it's simply not worth it.
+ *
+ * Also Shannon entropy is the last computed heuristic; if we got this far and ended up
+ * with uncertainty, just stay on the safe side and call it uncompressible.
+ */
+static bool has_low_entropy(struct bucket *bkt, size_t slen)
+{
+	const size_t threshold = 65, max_entropy = 8 * ilog2(16);
+	size_t i, p, p2, len, sum = 0;
+
+	len = ilog2(pow4(slen));
+
+	for (i = 0; i < 256 && bkt[i].count > 0; i++) {
+		p = bkt[i].count;
+		p2 = ilog2(pow4(p));
+		sum += p * (len - p2);
+	}
+
+	sum /= slen;
+
+	return ((sum * 100 / max_entropy) <= threshold);
+}
+
+#define BYTE_DIST_BAD		0
+#define BYTE_DIST_GOOD		1
+#define BYTE_DIST_MAYBE		2
+/*
+ * calc_byte_distribution() - Compute byte distribution on the sampled data.
+ * @bkt:	Byte counts of the sample.
+ * @slen:	Size of the sample.
+ *
+ * Return:
+ * BYTE_DIST_BAD:	A "hard no" for compression -- a computed uniform distribution of
+ *			the bytes (e.g. random or encrypted data).
+ * BYTE_DIST_GOOD:	High probability (normal (Gaussian) distribution) of the data being
+ *			compressible.
+ * BYTE_DIST_MAYBE:	When computed byte distribution resulted in "low > n < high"
+ *			grounds.  has_low_entropy() should be used for a final decision.
+ */
+static int calc_byte_distribution(struct bucket *bkt, size_t slen)
+{
+	const size_t low = 64, high = 200, threshold = slen * 90 / 100;
+	size_t sum = 0;
+	int i;
+
+	for (i = 0; i < low; i++)
+		sum += bkt[i].count;
+
+	if (sum > threshold)
+		return BYTE_DIST_BAD;
+
+	for (; i < high && bkt[i].count > 0; i++) {
+		sum += bkt[i].count;
+		if (sum > threshold)
+			break;
+	}
+
+	if (i <= low)
+		return BYTE_DIST_GOOD;
+
+	if (i >= high)
+		return BYTE_DIST_BAD;
+
+	return BYTE_DIST_MAYBE;
+}
+
+static bool is_mostly_ascii(const struct bucket *bkt)
+{
+	size_t count = 0;
+	int i;
+
+	for (i = 0; i < 256; i++)
+		if (bkt[i].count > 0)
+			/* Too many non-ASCII (0-63) bytes. */
+			if (++count > 64)
+				return false;
+
+	return true;
+}
+
+static bool has_repeated_data(const u8 *sample, size_t len)
+{
+	size_t s = len / 2;
+
+	return (!memcmp(&sample[0], &sample[s], s));
+}
+
+static int cmp_bkt(const void *_a, const void *_b)
+{
+	const struct bucket *a = _a, *b = _b;
+
+	/* Reverse sort. */
+	if (a->count > b->count)
+		return -1;
+
+	return 1;
+}
+
+/*
+ * Collect some 2K samples with 2K gaps between.
+ */
+static int collect_sample(const struct iov_iter *source, ssize_t max, u8 *sample)
+{
+	struct iov_iter iter = *source;
+	size_t s = 0;
+
+	while (iov_iter_count(&iter) >= SZ_2K) {
+		size_t part = umin(umin(iov_iter_count(&iter), SZ_2K), max);
+		size_t n;
+
+		n = copy_from_iter(sample + s, part, &iter);
+		if (n != part)
+			return -EFAULT;
+
+		s += n;
+		max -= n;
+
+		if (iov_iter_count(&iter) < PAGE_SIZE - SZ_2K)
+			break;
+
+		iov_iter_advance(&iter, SZ_2K);
+	}
+
+	return s;
+}
+
+/*
+ * is_compressible() - Determines if a chunk of data is compressible.
+ * @data: Iterator containing uncompressed data.
+ *
+ * Return: true if @data is compressible, false otherwise.
+ *
+ * Tests shows that this function is quite reliable in predicting data compressibility,
+ * matching close to 1:1 with the behaviour of LZ77 compression success and failures.
+ */
+static bool is_compressible(const struct iov_iter *data)
+{
+	const size_t read_size = SZ_2K, bkt_size = 256, max = SZ_4M;
+	struct bucket *bkt = NULL;
+	size_t len;
+	u8 *sample;
+	bool ret = false;
+	int i;
+
+	/* Preventive double check -- already checked in should_compress(). */
+	len = iov_iter_count(data);
+	if (unlikely(len < read_size))
+		return ret;
+
+	if (len - read_size > max)
+		len = max;
+
+	sample = kvzalloc(len, GFP_KERNEL);
+	if (!sample) {
+		WARN_ON_ONCE(1);
+
+		return ret;
+	}
+
+	/* Sample 2K bytes per page of the uncompressed data. */
+	i = collect_sample(data, len, sample);
+	if (i <= 0) {
+		WARN_ON_ONCE(1);
+
+		goto out;
+	}
+
+	len = i;
+	ret = true;
+
+	if (has_repeated_data(sample, len))
+		goto out;
+
+	bkt = kzalloc_objs(*bkt, bkt_size);
+	if (!bkt) {
+		WARN_ON_ONCE(1);
+		ret = false;
+
+		goto out;
+	}
+
+	for (i = 0; i < len; i++)
+		bkt[sample[i]].count++;
+
+	if (is_mostly_ascii(bkt))
+		goto out;
+
+	/* Sort in descending order */
+	sort(bkt, bkt_size, sizeof(*bkt), cmp_bkt, NULL);
+
+	i = calc_byte_distribution(bkt, len);
+	if (i != BYTE_DIST_MAYBE) {
+		ret = !!i;
+
+		goto out;
+	}
+
+	ret = has_low_entropy(bkt, len);
+out:
+	kvfree(sample);
+	kfree(bkt);
+
+	return ret;
+}
+
+/*
+ * should_compress() - Determines if a request (write) or the response to a
+ *		       request (read) should be compressed.
+ * @tcon: tcon of the request is being sent to
+ * @rqst: request to evaluate
+ *
+ * Return: true iff:
+ * - compression was successfully negotiated with server
+ * - server has enabled compression for the share
+ * - it's a read or write request
+ * - (write only) request length is >= SMB_COMPRESS_MIN_LEN
+ * - (write only) is_compressible() returns 1
+ *
+ * Return false otherwise.
+ */
+bool should_compress(const struct cifs_tcon *tcon, const struct smb_rqst *rq)
+{
+	const struct smb2_hdr *shdr = rq->rq_iov->iov_base;
+
+	if (unlikely(!tcon || !tcon->ses || !tcon->ses->server))
+		return false;
+
+	if (!tcon->ses->server->compression.enabled)
+		return false;
+
+	if (!(tcon->share_flags & SMB2_SHAREFLAG_COMPRESS_DATA))
+		return false;
+
+	if (shdr->Command == SMB2_WRITE) {
+		const struct smb2_write_req *wreq = rq->rq_iov->iov_base;
+
+		if (le32_to_cpu(wreq->Length) < SMB_COMPRESS_MIN_LEN)
+			return false;
+
+		return is_compressible(&rq->rq_iter);
+	}
+
+	return (shdr->Command == SMB2_READ);
+}
+
+int smb_compress(struct TCP_Server_Info *server, struct smb_rqst *rq, compress_send_fn send_fn)
+{
+	struct iov_iter iter;
+	u32 slen, dlen;
+	void *src, *dst = NULL;
+	int ret;
+
+	if (!server || !rq || !rq->rq_iov || !rq->rq_iov->iov_base)
+		return -EINVAL;
+
+	if (rq->rq_iov->iov_len != sizeof(struct smb2_write_req))
+		return -EINVAL;
+
+	slen = iov_iter_count(&rq->rq_iter);
+	src = kvzalloc(slen, GFP_KERNEL);
+	if (!src) {
+		ret = -ENOMEM;
+		goto err_free;
+	}
+
+	/* Keep the original iter intact. */
+	iter = rq->rq_iter;
+
+	if (!copy_from_iter_full(src, slen, &iter)) {
+		ret = smb_EIO(smb_eio_trace_compress_copy);
+		goto err_free;
+	}
+
+	dlen = smb_lz77_compressed_alloc_size(slen);
+	dst = kvzalloc(dlen, GFP_KERNEL);
+	if (!dst) {
+		ret = -ENOMEM;
+		goto err_free;
+	}
+
+	ret = smb_lz77_compress(src, slen, dst, &dlen);
+	if (!ret) {
+		struct smb2_compression_hdr hdr = { 0 };
+		struct smb_rqst comp_rq = { .rq_nvec = 3, };
+		struct kvec iov[3];
+
+		hdr.ProtocolId = SMB2_COMPRESSION_TRANSFORM_ID;
+		hdr.OriginalCompressedSegmentSize = cpu_to_le32(slen);
+		hdr.CompressionAlgorithm = SMB3_COMPRESS_LZ77;
+		hdr.Flags = SMB2_COMPRESSION_FLAG_NONE;
+		hdr.Offset = cpu_to_le32(rq->rq_iov[0].iov_len);
+
+		iov[0].iov_base = &hdr;
+		iov[0].iov_len = sizeof(hdr);
+		iov[1] = rq->rq_iov[0];
+		iov[2].iov_base = dst;
+		iov[2].iov_len = dlen;
+
+		comp_rq.rq_iov = iov;
+
+		ret = send_fn(server, 1, &comp_rq);
+	} else if (ret == -EMSGSIZE || dlen >= slen) {
+		ret = send_fn(server, 1, rq);
+	}
+err_free:
+	kvfree(dst);
+	kvfree(src);
+
+	return ret;
+}