5.10.125-rc1 review

This is the start of the stable review cycle for the 5.10.125 release. There are 11 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Sat, 25 Jun 2022 16:43:11 +0000. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.10.125-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-5.10.y
and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 5.10.125-rc1

Will Deacon <[email protected]>
arm64: mm: Don't invalidate FROM_DEVICE buffers at start of DMA transfer

Lukas Wunner <[email protected]>
serial: core: Initialize rs485 RTS polarity already on probe

Willy Tarreau <[email protected]>
tcp: drop the hash_32() part from the index calculation

Willy Tarreau <[email protected]>
tcp: increase source port perturb table to 2^16

Willy Tarreau <[email protected]>
tcp: dynamically allocate the perturb table used by source ports

Willy Tarreau <[email protected]>
tcp: add small random increments to the source port

Willy Tarreau <[email protected]>
tcp: use different parts of the port_offset for index and offset

Eric Dumazet <[email protected]>
tcp: add some entropy in __inet_hash_connect()

Marian Postevca <[email protected]>
usb: gadget: u_ether: fix regression in setting fixed MAC address

Damien Le Moal <[email protected]>
zonefs: fix zonefs_iomap_begin() for reads

Christian Borntraeger <[email protected]>
s390/mm: use non-quiescing sske for KVM switch to keyed guest

-------------

Diffstat:

 Makefile                              |  4 +-
 arch/arm64/mm/cache.S                 |  2 -
 arch/s390/mm/pgtable.c                |  2 +-
 drivers/tty/serial/serial_core.c      | 34 +++++--------
 drivers/usb/gadget/function/u_ether.c | 11 ++++-
 fs/zonefs/super.c                     | 92 +++++++++++++++++++++++------------
 net/ipv4/inet_hashtables.c            | 31 +++++++++---
 7 files changed, 109 insertions(+), 67 deletions(-)

From: Christian Borntraeger <[email protected]>

commit 3ae11dbcfac906a8c3a480e98660a823130dc16a upstream.

The switch to a keyed guest does not require a classic sske as the other guest CPUs are not accessing the key before the switch is complete. By using the NQ SSKE things are faster especially with multiple guests.

---
 arch/s390/mm/pgtable.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -734,7 +734,7 @@ void ptep_zap_key(struct mm_struct *mm,
 	pgste_val(pgste) |= PGSTE_GR_BIT | PGSTE_GC_BIT;
 	ptev = pte_val(*ptep);
 	if (!(ptev & _PAGE_INVALID) && (ptev & _PAGE_WRITE))
-		page_set_storage_key(ptev & PAGE_MASK, PAGE_DEFAULT_KEY, 1);
+		page_set_storage_key(ptev & PAGE_MASK, PAGE_DEFAULT_KEY, 0);
 	pgste_set_unlock(ptep, pgste);
 	preempt_enable();
 }

From: Damien Le Moal <[email protected]>

commit c1c1204c0d0c1dccc1310b9277fb2bd8b663d8fe upstream.

If a readahead is issued to a sequential zone file with an offset exactly equal to the current file size, the iomap type is set to IOMAP_UNWRITTEN, which will prevent an IO, but the iomap length is calculated as 0. This causes a WARN_ON() in iomap_iter():

[17309.548939] WARNING: CPU: 3 PID: 2137 at fs/iomap/iter.c:34 iomap_iter+0x9cf/0xe80
[...]
[17309.650907] RIP: 0010:iomap_iter+0x9cf/0xe80
[...]
[17309.754560] Call Trace:
[17309.757078]  <TASK>
[17309.759240]  ? lock_is_held_type+0xd8/0x130
[17309.763531]  iomap_readahead+0x1a8/0x870
[17309.767550]  ? iomap_read_folio+0x4c0/0x4c0
[17309.771817]  ? lockdep_hardirqs_on_prepare+0x400/0x400
[17309.778848]  ? lock_release+0x370/0x750
[17309.784462]  ? folio_add_lru+0x217/0x3f0
[17309.790220]  ? reacquire_held_locks+0x4e0/0x4e0
[17309.796543]  read_pages+0x17d/0xb60
[17309.801854]  ? folio_add_lru+0x238/0x3f0
[17309.807573]  ? readahead_expand+0x5f0/0x5f0
[17309.813554]  ? policy_node+0xb5/0x140
[17309.819018]  page_cache_ra_unbounded+0x27d/0x450
[17309.825439]  filemap_get_pages+0x500/0x1450
[17309.831444]  ? filemap_add_folio+0x140/0x140
[17309.837519]  ? lock_is_held_type+0xd8/0x130
[17309.843509]  filemap_read+0x28c/0x9f0
[17309.848953]  ? zonefs_file_read_iter+0x1ea/0x4d0 [zonefs]
[17309.856162]  ? trace_contention_end+0xd6/0x130
[17309.862416]  ? __mutex_lock+0x221/0x1480
[17309.868151]  ? zonefs_file_read_iter+0x166/0x4d0 [zonefs]
[17309.875364]  ? filemap_get_pages+0x1450/0x1450
[17309.881647]  ? __mutex_unlock_slowpath+0x15e/0x620
[17309.888248]  ? wait_for_completion_io_timeout+0x20/0x20
[17309.895231]  ? lock_is_held_type+0xd8/0x130
[17309.901115]  ? lock_is_held_type+0xd8/0x130
[17309.906934]  zonefs_file_read_iter+0x356/0x4d0 [zonefs]
[17309.913750]  new_sync_read+0x2d8/0x520
[17309.919035]  ? __x64_sys_lseek+0x1d0/0x1d0

Furthermore, this causes iomap_readahead() to loop forever as iomap_readahead_iter() always returns 0, making no progress.

Fix this by treating reads after the file size as access to holes, setting the iomap type to IOMAP_HOLE, the iomap addr to IOMAP_NULL_ADDR and using the length argument as is for the iomap length. To simplify the code with this change, zonefs_iomap_begin() is split into the read variant, zonefs_read_iomap_begin() and zonefs_read_iomap_ops, and the write variant, zonefs_write_iomap_begin() and zonefs_write_iomap_ops.

---
 fs/zonefs/super.c |   92 ++++++++++++++++++++++++++++++++++++------------------
 1 file changed, 62 insertions(+), 30 deletions(-)

@@ -68,15 +68,49 @@ static inline void zonefs_i_size_write(s
 		zi->i_flags &= ~ZONEFS_ZONE_OPEN;
 }
 
-static int zonefs_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
-			      unsigned int flags, struct iomap *iomap,
-			      struct iomap *srcmap)
+static int zonefs_read_iomap_begin(struct inode *inode, loff_t offset,
+				   loff_t length, unsigned int flags,
+				   struct iomap *iomap, struct iomap *srcmap)
 {
 	struct zonefs_inode_info *zi = ZONEFS_I(inode);
 	struct super_block *sb = inode->i_sb;
 	loff_t isize;
 
-	/* All I/Os should always be within the file maximum size */
+	/*
+	 * All blocks are always mapped below EOF. If reading past EOF,
+	 * act as if there is a hole up to the file maximum size.
+	 */
+	mutex_lock(&zi->i_truncate_mutex);
+	iomap->bdev = inode->i_sb->s_bdev;
+	iomap->offset = ALIGN_DOWN(offset, sb->s_blocksize);
+	isize = i_size_read(inode);
+	if (iomap->offset >= isize) {
+		iomap->type = IOMAP_HOLE;
+		iomap->addr = IOMAP_NULL_ADDR;
+		iomap->length = length;
+	} else {
+		iomap->type = IOMAP_MAPPED;
+		iomap->addr = (zi->i_zsector << SECTOR_SHIFT) + iomap->offset;
+		iomap->length = isize - iomap->offset;
+	}
+	mutex_unlock(&zi->i_truncate_mutex);
+
+	return 0;
+}
+
+static const struct iomap_ops zonefs_read_iomap_ops = {
+	.iomap_begin	= zonefs_read_iomap_begin,
+};
+
+static int zonefs_write_iomap_begin(struct inode *inode, loff_t offset,
+				    loff_t length, unsigned int flags,
+				    struct iomap *iomap, struct iomap *srcmap)
+{
+	struct zonefs_inode_info *zi = ZONEFS_I(inode);
+	struct super_block *sb = inode->i_sb;
+	loff_t isize;
+
+	/* All write I/Os should always be within the file maximum size */
 	if (WARN_ON_ONCE(offset + length > zi->i_max_size))
 		return -EIO;
 

@@ -86,7 +120,7 @@ static int zonefs_iomap_begin(struct ino
 	 * operation.
 	 */
 	if (WARN_ON_ONCE(zi->i_ztype == ZONEFS_ZTYPE_SEQ &&
-			 (flags & IOMAP_WRITE) && !(flags & IOMAP_DIRECT)))
+			 !(flags & IOMAP_DIRECT)))
 		return -EIO;
 
 	/*

@@ -95,45 +129,42 @@ static int zonefs_iomap_begin(struct ino
 	 * write pointer) and unwriten beyond.
 	 */
 	mutex_lock(&zi->i_truncate_mutex);
+	iomap->bdev = inode->i_sb->s_bdev;
+	iomap->offset = ALIGN_DOWN(offset, sb->s_blocksize);
+	iomap->addr = (zi->i_zsector << SECTOR_SHIFT) + iomap->offset;
 	isize = i_size_read(inode);
-	if (offset >= isize)
+	if (iomap->offset >= isize) {
 		iomap->type = IOMAP_UNWRITTEN;
-	else
+		iomap->length = zi->i_max_size - iomap->offset;
+	} else {
 		iomap->type = IOMAP_MAPPED;
-	if (flags & IOMAP_WRITE)
-		length = zi->i_max_size - offset;
-	else
-		length = min(length, isize - offset);
+		iomap->length = isize - iomap->offset;
+	}
 	mutex_unlock(&zi->i_truncate_mutex);
 
-	iomap->offset = ALIGN_DOWN(offset, sb->s_blocksize);
-	iomap->length = ALIGN(offset + length, sb->s_blocksize) - iomap->offset;
-	iomap->bdev = inode->i_sb->s_bdev;
-	iomap->addr = (zi->i_zsector << SECTOR_SHIFT) + iomap->offset;
-
 	return 0;
 }
 
-static const struct iomap_ops zonefs_iomap_ops = {
-	.iomap_begin	= zonefs_iomap_begin,
+static const struct iomap_ops zonefs_write_iomap_ops = {
+	.iomap_begin	= zonefs_write_iomap_begin,
 };
 
 static int zonefs_readpage(struct file *unused, struct page *page)
 {
-	return iomap_readpage(page, &zonefs_iomap_ops);
+	return iomap_readpage(page, &zonefs_read_iomap_ops);
 }
 
 static void zonefs_readahead(struct readahead_control *rac)
 {
-	iomap_readahead(rac, &zonefs_iomap_ops);
+	iomap_readahead(rac, &zonefs_read_iomap_ops);
 }
 
 /*
  * Map blocks for page writeback. This is used only on conventional zone files,
  * which implies that the page range can only be within the fixed inode size.
  */
-static int zonefs_map_blocks(struct iomap_writepage_ctx *wpc,
-			     struct inode *inode, loff_t offset)
+static int zonefs_write_map_blocks(struct iomap_writepage_ctx *wpc,
+				   struct inode *inode, loff_t offset)
 {
 	struct zonefs_inode_info *zi = ZONEFS_I(inode);
 

@@ -147,12 +178,12 @@ static int zonefs_map_blocks(struct ioma
 	    offset < wpc->iomap.offset + wpc->iomap.length)
 		return 0;
 
-	return zonefs_iomap_begin(inode, offset, zi->i_max_size - offset,
-				  IOMAP_WRITE, &wpc->iomap, NULL);
+	return zonefs_write_iomap_begin(inode, offset, zi->i_max_size - offset,
+					IOMAP_WRITE, &wpc->iomap, NULL);
 }
 
 static const struct iomap_writeback_ops zonefs_writeback_ops = {
-	.map_blocks		= zonefs_map_blocks,
+	.map_blocks		= zonefs_write_map_blocks,
 };
 
 static int zonefs_writepage(struct page *page, struct writeback_control *wbc)

@@ -182,7 +213,8 @@ static int zonefs_swap_activate(struct s
 		return -EINVAL;
 	}
 
-	return iomap_swapfile_activate(sis, swap_file, span, &zonefs_iomap_ops);
+	return iomap_swapfile_activate(sis, swap_file, span,
+				       &zonefs_read_iomap_ops);
 }
 
 static const struct address_space_operations zonefs_file_aops = {

@@ -612,7 +644,7 @@ static vm_fault_t zonefs_filemap_page_mk
 
 	/* Serialize against truncates */
 	down_read(&zi->i_mmap_sem);
-	ret = iomap_page_mkwrite(vmf, &zonefs_iomap_ops);
+	ret = iomap_page_mkwrite(vmf, &zonefs_write_iomap_ops);
 	up_read(&zi->i_mmap_sem);
 
 	sb_end_pagefault(inode->i_sb);

@@ -869,7 +901,7 @@ static ssize_t zonefs_file_dio_write(str
 	if (append)
 		ret = zonefs_file_dio_append(iocb, from);
 	else
-		ret = iomap_dio_rw(iocb, from, &zonefs_iomap_ops,
+		ret = iomap_dio_rw(iocb, from, &zonefs_write_iomap_ops,
 				   &zonefs_write_dio_ops, sync);
 	if (zi->i_ztype == ZONEFS_ZTYPE_SEQ &&
 	    (ret > 0 || ret == -EIOCBQUEUED)) {

@@ -911,7 +943,7 @@ static ssize_t zonefs_file_buffered_writ
 	if (ret <= 0)
 		goto inode_unlock;
 
-	ret = iomap_file_buffered_write(iocb, from, &zonefs_iomap_ops);
+	ret = iomap_file_buffered_write(iocb, from, &zonefs_write_iomap_ops);
 	if (ret > 0)
 		iocb->ki_pos += ret;
 	else if (ret == -EIO)

@@ -1004,7 +1036,7 @@ static ssize_t zonefs_file_read_iter(str
 			goto inode_unlock;
 		}
 		file_accessed(iocb->ki_filp);
-		ret = iomap_dio_rw(iocb, to, &zonefs_iomap_ops,
+		ret = iomap_dio_rw(iocb, to, &zonefs_read_iomap_ops,
 				   &zonefs_read_dio_ops, is_sync_kiocb(iocb));
 	} else {
 		ret = generic_file_read_iter(iocb, to);

From: Marian Postevca <[email protected]>

commit b337af3a4d6147000b7ca6b3438bf5c820849b37 upstream.

In systemd systems setting a fixed MAC address through the "dev_addr" module argument fails systematically. When checking the MAC address after the interface is created it always has the same but different MAC address to the one supplied as argument.

This is partially caused by systemd which by default will set an internally generated permanent MAC address for interfaces that are marked as having a randomly generated address.

Commit 890d5b40908bfd1a ("usb: gadget: u_ether: fix race in setting MAC address in setup phase") didn't take into account the fact that the interface must be marked as having a set MAC address when it's set as module argument.

Fixed by marking the interface with NET_ADDR_SET when the "dev_addr" module argument is supplied.

---
 drivers/usb/gadget/function/u_ether.c |   11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

@@ -772,9 +772,13 @@ struct eth_dev *gether_setup_name(struct
 	dev->qmult = qmult;
 	snprintf(net->name, sizeof(net->name), "%s%%d", netname);
 
-	if (get_ether_addr(dev_addr, net->dev_addr))
+	if (get_ether_addr(dev_addr, net->dev_addr)) {
+		net->addr_assign_type = NET_ADDR_RANDOM;
 		dev_warn(&g->dev,
 			"using random %s ethernet address\n", "self");
+	} else {
+		net->addr_assign_type = NET_ADDR_SET;
+	}
 	if (get_ether_addr(host_addr, dev->host_mac))
 		dev_warn(&g->dev,
 			"using random %s ethernet address\n", "host");

@@ -831,6 +835,9 @@ struct net_device *gether_setup_name_def
 	INIT_LIST_HEAD(&dev->tx_reqs);
 	INIT_LIST_HEAD(&dev->rx_reqs);
 
+	/* by default we always have a random MAC address */
+	net->addr_assign_type = NET_ADDR_RANDOM;
+
 	skb_queue_head_init(&dev->rx_frames);
 
 	/* network device setup */

@@ -868,7 +875,6 @@ int gether_register_netdev(struct net_de
 	g = dev->gadget;
 
 	memcpy(net->dev_addr, dev->dev_mac, ETH_ALEN);
-	net->addr_assign_type = NET_ADDR_RANDOM;
 
 	status = register_netdev(net);
 	if (status < 0) {

@@ -908,6 +914,7 @@ int gether_set_dev_addr(struct net_devic
 	if (get_ether_addr(dev_addr, new_addr))
 		return -EINVAL;
 	memcpy(dev->dev_mac, new_addr, ETH_ALEN);
+	net->addr_assign_type = NET_ADDR_SET;
 	return 0;
 }
 EXPORT_SYMBOL_GPL(gether_set_dev_addr);

From: Eric Dumazet <[email protected]>

commit c579bd1b4021c42ae247108f1e6f73dd3f08600c upstream.

Even when implementing RFC 6056 3.3.4 (Algorithm 4: Double-Hash Port Selection Algorithm), a patient attacker could still be able to collect enough state from an otherwise idle host.

Idea of this patch is to inject some noise, in the cases __inet_hash_connect() found a candidate in the first attempt.

This noise should not significantly reduce the collision avoidance, and should be zero if connection table is already well used.

Note that this is not implementing RFC 6056 3.3.5
because we think Algorithm 5 could hurt typical
workloads.

---
 net/ipv4/inet_hashtables.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -833,6 +833,11 @@ next_port:
 	return -EADDRNOTAVAIL;
 
 ok:
+	/* If our first attempt found a candidate, skip next candidate
+	 * in 1/16 of cases to add some noise.
+	 */
+	if (!i && !(prandom_u32() % 16))
+		i = 2;
 	WRITE_ONCE(table_perturb[index], READ_ONCE(table_perturb[index]) + i + 2);
 
 	/* Head lock still held and bh's disabled */

From: Willy Tarreau <[email protected]>

commit 9e9b70ae923baf2b5e8a0ea4fd0c8451801ac526 upstream.

Amit Klein suggests that we use different parts of port_offset for the table's index and the port offset so that there is no direct relation between them.

---
 net/ipv4/inet_hashtables.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -777,7 +777,7 @@ int __inet_hash_connect(struct inet_time
 	net_get_random_once(table_perturb, sizeof(table_perturb));
 	index = hash_32(port_offset, INET_TABLE_PERTURB_SHIFT);
 
-	offset = READ_ONCE(table_perturb[index]) + port_offset;
+	offset = READ_ONCE(table_perturb[index]) + (port_offset >> 32);
 	offset %= remaining;
 
 	/* In first pass we try ports of @low parity.

From: Willy Tarreau <[email protected]>

commit ca7af0402550f9a0b3316d5f1c30904e42ed257d upstream.

Here we're randomly adding between 0 and 7 random increments to the selected source port in order to add some noise in the source port selection that will make the next port less predictable.

With the default port range of 32768-60999 this means a worst case reuse scenario of 14116/8=1764 connections between two consecutive uses of the same port, with an average of 14116/4.5=3137. This code was stressed at more than 800000 connections per second to a fixed target with all connections closed by the client using RSTs (worst condition) and only 2 connections failed among 13 billion, despite the hash being reseeded every 10 seconds, indicating a perfectly safe situation.

---
 net/ipv4/inet_hashtables.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

@@ -833,11 +833,12 @@ next_port:
 	return -EADDRNOTAVAIL;
 
 ok:
-	/* If our first attempt found a candidate, skip next candidate
-	 * in 1/16 of cases to add some noise.
+	/* Here we want to add a little bit of randomness to the next source
+	 * port that will be chosen. We use a max() with a random here so that
+	 * on low contention the randomness is maximal and on high contention
+	 * it may be inexistent.
 	 */
-	if (!i && !(prandom_u32() % 16))
-		i = 2;
+	i = max_t(int, i, (prandom_u32() & 7) * 2);
 	WRITE_ONCE(table_perturb[index], READ_ONCE(table_perturb[index]) + i + 2);
 
 	/* Head lock still held and bh's disabled */

From: Willy Tarreau <[email protected]>

commit e9261476184be1abd486c9434164b2acbe0ed6c2 upstream.

We'll need to further increase the size of this table and it's likely that at some point its size will not be suitable anymore for a static table. Let's allocate it on boot from inet_hashinfo2_init(), which is called from tcp_init().

---
 net/ipv4/inet_hashtables.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

@@ -731,7 +731,8 @@ EXPORT_SYMBOL_GPL(inet_unhash);
  * privacy, this only consumes 1 KB of kernel memory.
  */
 #define INET_TABLE_PERTURB_SHIFT 8
-static u32 table_perturb[1 << INET_TABLE_PERTURB_SHIFT];
+#define INET_TABLE_PERTURB_SIZE (1 << INET_TABLE_PERTURB_SHIFT)
+static u32 *table_perturb;
 
 int __inet_hash_connect(struct inet_timewait_death_row *death_row,
 		struct sock *sk, u64 port_offset,

@@ -774,7 +775,8 @@ int __inet_hash_connect(struct inet_time
 	if (likely(remaining > 1))
 		remaining &= ~1U;
 
-	net_get_random_once(table_perturb, sizeof(table_perturb));
+	net_get_random_once(table_perturb,
+			    INET_TABLE_PERTURB_SIZE * sizeof(*table_perturb));
 	index = hash_32(port_offset, INET_TABLE_PERTURB_SHIFT);
 
 	offset = READ_ONCE(table_perturb[index]) + (port_offset >> 32);

@@ -912,6 +914,12 @@ void __init inet_hashinfo2_init(struct i
 					    low_limit,
 					    high_limit);
 	init_hashinfo_lhash2(h);
+
+	/* this one is used for source ports of outgoing connections */
+	table_perturb = kmalloc_array(INET_TABLE_PERTURB_SIZE,
+				      sizeof(*table_perturb), GFP_KERNEL);
+	if (!table_perturb)
+		panic("TCP: failed to alloc table_perturb");
 }
 
 int inet_hashinfo2_init_mod(struct inet_hashinfo *h)

From: Willy Tarreau <[email protected]>

commit 4c2c8f03a5ab7cb04ec64724d7d176d00bcc91e5 upstream.

Moshe Kol, Amit Klein, and Yossi Gilad reported being able to accurately identify a client by forcing it to emit only 40 times more connections than there are entries in the table_perturb[] table. The previous two improvements consisting in resalting the secret every 10s and adding randomness to each port selection only slightly improved the situation, and the current value of 2^8 was too small as it's not very difficult to make a client emit 10k connections in less than 10 seconds.

Thus we're increasing the perturb table from 2^8 to 2^16 so that the same precision now requires 2.6M connections, which is more difficult in this time frame and harder to hide as a background activity. The impact is that the table now uses 256 kB instead of 1 kB, which could mostly affect devices making frequent outgoing connections. However such components usually target a small set of destinations (load balancers, database clients, perf assessment tools), and in practice only a few entries will be visited, like before.

A live test at 1 million connections per second showed no performance difference from the previous value.

---
 net/ipv4/inet_hashtables.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

@@ -726,11 +726,12 @@ EXPORT_SYMBOL_GPL(inet_unhash);
  * Note that we use 32bit integers (vs RFC 'short integers')
  * because 2^16 is not a multiple of num_ephemeral and this
  * property might be used by clever attacker.
- * RFC claims using TABLE_LENGTH=10 buckets gives an improvement,
- * we use 256 instead to really give more isolation and
- * privacy, this only consumes 1 KB of kernel memory.
+ * RFC claims using TABLE_LENGTH=10 buckets gives an improvement, though
+ * attacks were since demonstrated, thus we use 65536 instead to really
+ * give more isolation and privacy, at the expense of 256kB of kernel
+ * memory.
  */
-#define INET_TABLE_PERTURB_SHIFT 8
+#define INET_TABLE_PERTURB_SHIFT 16
 #define INET_TABLE_PERTURB_SIZE (1 << INET_TABLE_PERTURB_SHIFT)
 static u32 *table_perturb;
 

From: Willy Tarreau <[email protected]>

commit e8161345ddbb66e449abde10d2fdce93f867eba9 upstream.

In commit 190cc82489f4 ("tcp: change source port randomizarion at connect() time"), the table_perturb[] array was introduced and an index was taken from the port_offset via hash_32(). But it turns out that hash_32() performs a multiplication while the input here comes from the output of SipHash in secure_seq, that is well distributed enough to avoid the need for yet another hash.

---
 net/ipv4/inet_hashtables.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -778,7 +778,7 @@ int __inet_hash_connect(struct inet_time
 
 	net_get_random_once(table_perturb,
 			    INET_TABLE_PERTURB_SIZE * sizeof(*table_perturb));
-	index = hash_32(port_offset, INET_TABLE_PERTURB_SHIFT);
+	index = port_offset & (INET_TABLE_PERTURB_SIZE - 1);
 
 	offset = READ_ONCE(table_perturb[index]) + (port_offset >> 32);
 	offset %= remaining;

From: Lukas Wunner <[email protected]>

commit 2dd8a74fddd21b95dcc60a2d3c9eaec993419d69 upstream.

RTS polarity of rs485-enabled ports is currently initialized on uart open via:

tty_port_open()
tty_port_block_til_ready()

    tty_port_raise_dtr_rts()  # if (C_BAUD(tty))
      uart_dtr_rts()
        uart_port_dtr_rts()

There's at least three problems here:

First, if no baud rate is set, RTS polarity is not initialized. That's the right thing to do for rs232, but not for rs485, which requires that RTS is deasserted unconditionally.

Second, if the DeviceTree property "linux,rs485-enabled-at-boot-time" is present, RTS should be deasserted as early as possible, i.e. on probe. Otherwise it may remain asserted until first open.

Third, even though RTS is deasserted on open and close, it may subsequently be asserted by uart_throttle(), uart_unthrottle() or uart_set_termios() because those functions aren't rs485-aware. (Only uart_tiocmset() is.)

To address these issues, move RTS initialization from uart_port_dtr_rts() to uart_configure_port(). Prevent subsequent modification of RTS polarity by moving the existing rs485 check from uart_tiocmget() to uart_update_mctrl().

That way, RTS is initialized on probe and then remains unmodified unless the uart transmits data. If rs485 is enabled at runtime (instead of at boot) through a TIOCSRS485 ioctl(), RTS is initialized by the uart driver's ->rs485_config() callback and then likewise remains unmodified.

The PL011 driver initializes RTS on uart open and prevents subsequent modification in its ->set_mctrl() callback. That code is obsoleted by the present commit, so drop it.

---
 drivers/tty/serial/serial_core.c |   34 ++++++++++++----------------------
 1 file changed, 12 insertions(+), 22 deletions(-)

@@ -144,6 +144,11 @@ uart_update_mctrl(struct uart_port *port
 	unsigned long flags;
 	unsigned int old;
 
+	if (port->rs485.flags & SER_RS485_ENABLED) {
+		set &= ~TIOCM_RTS;
+		clear &= ~TIOCM_RTS;
+	}
+
 	spin_lock_irqsave(&port->lock, flags);
 	old = port->mctrl;
 	port->mctrl = (old & ~clear) | set;

@@ -157,23 +162,10 @@ uart_update_mctrl(struct uart_port *port
 
 static void uart_port_dtr_rts(struct uart_port *uport, int raise)
 {
-	int rs485_on = uport->rs485_config &&
-		(uport->rs485.flags & SER_RS485_ENABLED);
-	int RTS_after_send = !!(uport->rs485.flags & SER_RS485_RTS_AFTER_SEND);
-
-	if (raise) {
-		if (rs485_on && RTS_after_send) {
-			uart_set_mctrl(uport, TIOCM_DTR);
-			uart_clear_mctrl(uport, TIOCM_RTS);
-		} else {
-			uart_set_mctrl(uport, TIOCM_DTR | TIOCM_RTS);
-		}
-	} else {
-		unsigned int clear = TIOCM_DTR;
-
-		clear |= (!rs485_on || RTS_after_send) ? TIOCM_RTS : 0;
-		uart_clear_mctrl(uport, clear);
-	}
+	if (raise)
+		uart_set_mctrl(uport, TIOCM_DTR | TIOCM_RTS);
+	else
+		uart_clear_mctrl(uport, TIOCM_DTR | TIOCM_RTS);
 }
 
 /*

@@ -1116,11 +1108,6 @@ uart_tiocmset(struct tty_struct *tty, un
 		goto out;
 
 	if (!tty_io_error(tty)) {
-		if (uport->rs485.flags & SER_RS485_ENABLED) {
-			set &= ~TIOCM_RTS;
-			clear &= ~TIOCM_RTS;
-		}
-
 		uart_update_mctrl(uport, set, clear);
 		ret = 0;
 	}

@@ -2429,6 +2416,9 @@ uart_configure_port(struct uart_driver *
 		 */
 		spin_lock_irqsave(&port->lock, flags);
 		port->mctrl &= TIOCM_DTR;
+		if (port->rs485.flags & SER_RS485_ENABLED &&
+		    !(port->rs485.flags & SER_RS485_RTS_AFTER_SEND))
+			port->mctrl |= TIOCM_RTS;
 		port->ops->set_mctrl(port, port->mctrl);
 		spin_unlock_irqrestore(&port->lock, flags);
 

From: Will Deacon <[email protected]>

commit c50f11c6196f45c92ca48b16a5071615d4ae0572 upstream.

Invalidating the buffer memory in arch_sync_dma_for_device() for FROM_DEVICE transfers

When using the streaming DMA API to map a buffer prior to inbound non-coherent DMA (i.e. DMA_FROM_DEVICE), we invalidate any dirty CPU cachelines so that they will not be written back during the transfer and corrupt the buffer contents written by the DMA. This, however, poses two potential problems:

(1) If the DMA transfer does not write to every byte in the buffer, then the unwritten bytes will contain stale data once the transfer has completed.

(2) If the buffer has a virtual alias in userspace, then stale data may be visible via this alias during the period between performing the cache invalidation and the DMA writes landing in memory.

Address both of these issues by cleaning (aka writing-back) the dirty lines in arch_sync_dma_for_device(DMA_FROM_DEVICE) instead of discarding them using invalidation.

---
 arch/arm64/mm/cache.S |    2 --
 1 file changed, 2 deletions(-)

@@ -228,8 +228,6 @@ SYM_FUNC_END_PI(__dma_flush_area)
  *	- dir	- DMA direction
  */
 SYM_FUNC_START_PI(__dma_map_area)
-	cmp	w2, #DMA_FROM_DEVICE
-	b.eq	__dma_inv_area
 	b	__dma_clean_area
 SYM_FUNC_END_PI(__dma_map_area)
 

Hi!

CIP testing did not find any problems here:

https://gitlab.com/cip-project/cip-testing/linux-stable-rc-ci/-/tree/linux-5.10.y

Best regards,
Pavel

DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany

On ARCH_BRCMSTB using 32-bit and 64-bit ARM kernels:

Florian

Compiled and booted on my test system. No dmesg regressions.

thanks,
-- Shuah

Tested on arm64 and x86 for 5.10.125-rc1,

Kernel repo:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
Branch: linux-5.10.y
Version: 5.10.125-rc1
Commit: 7119d4fdfdd36590d4247bd9dcab4b18fb592de8
Compiler: gcc version 7.3.0 (GCC)

arm64:
--------------------------------------------------------------------
Testcase Result Summary:
total: 9094
passed: 9094
failed: 0
timeout: 0
--------------------------------------------------------------------

x86:
--------------------------------------------------------------------
Testcase Result Summary:
total: 9094
passed: 9094
failed: 0
timeout: 0
--------------------------------------------------------------------

No new regressions for Tegra ...

Test results for stable-v5.10:

    10 builds:	10 pass, 0 fail
    28 boots:	28 pass, 0 fail
    75 tests:	74 pass, 1 fail

Linux version: 5.10.125-rc1-g99120abeed34
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,

                tegra194-p2972-0000, tegra194-p3509-0000+p3668-0000,
                tegra20-ventana, tegra210-p2371-2180,
                tegra210-p3450-0000, tegra30-cardhu-a04

Test failures: tegra194-p2972-0000: boot.py

Jon

nvpublic

Hi Greg,

Build test (gcc version 11.3.1 20220621):
mips: 63 configs -> no failure
arm: 104 configs -> no failure
arm64: 3 configs -> no failure
x86_64: 4 configs -> no failure alpha allmodconfig -> no failure powerpc allmodconfig -> no failure riscv allmodconfig -> no failure s390 allmodconfig -> no failure xtensa allmodconfig -> no failure

Boot test:
x86_64: Booted on my test laptop. No regression.
x86_64: Booted on qemu. No regression. [1]
arm64: Booted on rpi4b (4GB model). No regression. [2]

[1]. https://openqa.qa.codethink.co.uk/tests/1380
[2]. https://openqa.qa.codethink.co.uk/tests/1385

--
Regards
Sudip

Build results:

	total: 163 pass: 163 fail: 0
Qemu test results:
	total: 477 pass: 477 fail: 0

Guenter

Results from Linaro’s test farm.
No regressions on arm64, arm, x86_64, and i386.

## Build
* kernel: 5.10.125-rc1
* git: https://gitlab.com/Linaro/lkft/mirrors/stable/linux-stable-rc
* git branch: linux-5.10.y
* git commit: 99120abeed34b4814d3c0b4443283075bb65646c
* git describe: v5.10.123-96-g99120abeed34
* test details:
https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.10.y/build/v5.10.123-96-g99120abeed34

## Test Regressions (compared to v5.10.123-85-g1432bd558ac0) No test regressions found.

## Metric Regressions (compared to v5.10.123-85-g1432bd558ac0) No metric regressions found.

## Test Fixes (compared to v5.10.123-85-g1432bd558ac0) No test fixes found.

## Metric Fixes (compared to v5.10.123-85-g1432bd558ac0) No metric fixes found.

## Test result summary
total: 133307, pass: 119577, fail: 254, skip: 12678, xfail: 798

## Build Summary
* arc: 10 total, 10 passed, 0 failed
* arm: 314 total, 314 passed, 0 failed
* arm64: 58 total, 58 passed, 0 failed
* i386: 52 total, 49 passed, 3 failed
* mips: 37 total, 37 passed, 0 failed
* parisc: 12 total, 12 passed, 0 failed
* powerpc: 51 total, 51 passed, 0 failed
* riscv: 27 total, 27 passed, 0 failed
* s390: 21 total, 21 passed, 0 failed
* sh: 24 total, 24 passed, 0 failed
* sparc: 12 total, 12 passed, 0 failed
* x86_64: 56 total, 55 passed, 1 failed

## Test suites summary
* fwts
* igt-gpu-tools
* kunit
* kvm-unit-tests
* libgpiod
* libhugetlbfs
* log-parser-boot
* log-parser-test
* ltp-cap_bounds
* ltp-cap_bounds-tests
* ltp-commands
* ltp-commands-tests
* ltp-containers
* ltp-containers-tests
* ltp-controllers-tests
* ltp-cpuhotplug-tests
* ltp-crypto
* ltp-crypto-tests
* ltp-cve-tests
* ltp-dio-tests
* ltp-fcntl-locktests
* ltp-fcntl-locktests-tests
* ltp-filecaps
* ltp-filecaps-tests
* ltp-fs
* ltp-fs-tests
* ltp-fs_bind
* ltp-fs_bind-tests
* ltp-fs_perms_simple
* ltp-fs_perms_simple-tests
* ltp-fsx
* ltp-fsx-tests
* ltp-hugetlb
* ltp-hugetlb-tests
* ltp-io
* ltp-io-tests
* ltp-ipc
* ltp-ipc-tests
* ltp-math-tests
* ltp-mm-tests
* ltp-nptl
* ltp-nptl-tests
* ltp-open-posix-tests
* ltp-pty
* ltp-pty-tests
* ltp-sched
* ltp-sched-tests
* ltp-securebits
* ltp-securebits-tests
* ltp-smoke
* ltp-syscalls-tests
* ltp-tracing-tests
* network-basic-tests
* packetdrill
* perf
* perf/Zstd-perf.data-compression
* rcutorture
* ssuite
* v4l2-compliance
* vdso

--
Linaro LKFT
https://lkft.linaro.org