4.19.251-rc1 review

This is the start of the stable review cycle for the 4.19.251 release. There are 33 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 Thu, 07 Jul 2022 11:55:56 +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/v4.x/stable-review/patch-4.19.251-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-4.19.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 4.19.251-rc1

Daniele Palmas <[email protected]>

    net: usb: qmi_wwan: add Telit 0x1070 composition

Carlo Lobrano <[email protected]>

    net: usb: qmi_wwan: add Telit 0x1060 composition

Oleksandr Tyshchenko <[email protected]>
xen/arm: Fix race in RB-tree based P2M accounting

Roger Pau Monne <[email protected]>
xen/blkfront: force data bouncing when backend is untrusted

Roger Pau Monne <[email protected]>
xen/netfront: force data bouncing when backend is untrusted

Roger Pau Monne <[email protected]>
xen/netfront: fix leaking data in shared pages

Roger Pau Monne <[email protected]>
xen/blkfront: fix leaking data in shared pages

katrinzhou <[email protected]>
ipv6/sit: fix ipip6_tunnel_get_prl return value

kernel test robot <[email protected]>
sit: use min

Doug Berger <[email protected]>
net: dsa: bcm_sf2: force pause link settings

Yang Yingliang <[email protected]>
hwmon: (ibmaem) don't call platform_device_del() if platform_device_add() fails

Demi Marie Obenour <[email protected]>
xen/gntdev: Avoid blocking in unmap_grant_pages()

Jakub Kicinski <[email protected]>
net: tun: avoid disabling NAPI twice

Michael Walle <[email protected]>
NFC: nxp-nci: Don't issue a zero length i2c_master_read()

Krzysztof Kozlowski <[email protected]>
nfc: nfcmrvl: Fix irq_of_parse_and_map() return value

Yevhen Orlov <[email protected]>
net: bonding: fix use-after-free after 802.3ad slave unbind

Eric Dumazet <[email protected]>
net: bonding: fix possible NULL deref in rlb code

Pablo Neira Ayuso <[email protected]>
netfilter: nft_dynset: restore set element counter when failing to update

Jason Wang <[email protected]>
caif_virtio: fix race between virtio_device_ready() and ndo_open()

YueHaibing <[email protected]>
net: ipv6: unexport __init-annotated seg6_hmac_net_init()

Oliver Neukum <[email protected]>
usbnet: fix memory allocation in helpers

Kamal Heib <[email protected]>
RDMA/qedr: Fix reporting QP timeout attribute

Jakub Kicinski <[email protected]>
net: tun: stop NAPI when detaching queues

Jakub Kicinski <[email protected]>
net: tun: unlink NAPI from device on destruction

Dimitris Michailidis <[email protected]>
selftests/net: pass ipv6_args to udpgso_bench's IPv6 TCP test

Jason Wang <[email protected]>
virtio-net: fix race between ndo_open() and virtio_device_ready()

Jose Alonso <[email protected]>

    net: usb: ax88179_178a: Fix packet receiving

Duoming Zhou <[email protected]>
net: rose: fix UAF bugs caused by timer handler

Chuck Lever <[email protected]>
SUNRPC: Fix READ_PLUS crasher

Jason A. Donenfeld <[email protected]>
s390/archrandom: simplify back to earlier design and initialize earlier

Mikulas Patocka <[email protected]>
dm raid: fix KASAN warning in raid5_add_disks

Heinz Mauelshagen <[email protected]>
dm raid: fix accesses beyond end of raid member array

Chris Ye <[email protected]>
nvdimm: Fix badblocks clear off-by-one error

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

Diffstat:

 Makefile                                    |   4 +-
 arch/arm/xen/p2m.c                          |   6 +-
 arch/s390/crypto/arch_random.c              | 111 +--------------------
 arch/s390/include/asm/archrandom.h          |  21 ++--
 arch/s390/kernel/setup.c                    |   5 +
 drivers/block/xen-blkfront.c                |  56 +++++++----
 drivers/hwmon/ibmaem.c                      |  12 ++-
 drivers/infiniband/hw/qedr/qedr.h           |   1 +
 drivers/infiniband/hw/qedr/verbs.c          |   4 +-
 drivers/md/dm-raid.c                        |  34 ++++---
 drivers/md/raid5.c                          |   1 +
 drivers/net/bonding/bond_3ad.c              |   3 +-
 drivers/net/bonding/bond_alb.c              |   2 +-
 drivers/net/caif/caif_virtio.c              |  10 +-
 drivers/net/dsa/bcm_sf2.c                   |   5 +
 drivers/net/tun.c                           |  15 ++-
 drivers/net/usb/ax88179_178a.c              | 101 ++++++++++++++-----
 drivers/net/usb/qmi_wwan.c                  |   2 +
 drivers/net/usb/usbnet.c                    |   4 +-
 drivers/net/virtio_net.c                    |   8 +-
 drivers/net/xen-netfront.c                  |  52 +++++++++-
 drivers/nfc/nfcmrvl/i2c.c                   |   6 +-
 drivers/nfc/nfcmrvl/spi.c                   |   6 +-
 drivers/nfc/nxp-nci/i2c.c                   |   3 +
 drivers/nvdimm/bus.c                        |   4 +-
 drivers/xen/gntdev-common.h                 |   8 ++
 drivers/xen/gntdev.c                        | 146 +++++++++++++++++++---------
 net/ipv6/seg6_hmac.c                        |   1 -
 net/ipv6/sit.c                              |  10 +-
 net/netfilter/nft_set_hash.c                |   2 +
 net/rose/rose_timer.c                       |  34 ++++---
 net/sunrpc/xdr.c                            |   2 +-
 tools/testing/selftests/net/udpgso_bench.sh |   2 +-
 33 files changed, 406 insertions(+), 275 deletions(-)

From: Chris Ye <[email protected]>

commit ef9102004a87cb3f8b26e000a095a261fc0467d3 upstream.

nvdimm_clear_badblocks_region() validates badblock clearing requests against the span of the region, however it compares the inclusive badblock request range to the exclusive region range. Fix up the off-by-one error.

---
 drivers/nvdimm/bus.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -196,8 +196,8 @@ static int nvdimm_clear_badblocks_region
 	ndr_end = nd_region->ndr_start + nd_region->ndr_size - 1;
 
 	/* make sure we are in the region */
-	if (ctx->phys < nd_region->ndr_start
-			|| (ctx->phys + ctx->cleared) > ndr_end)
+	if (ctx->phys < nd_region->ndr_start ||
+	    (ctx->phys + ctx->cleared - 1) > ndr_end)
 		return 0;
 
 	sector = (ctx->phys - nd_region->ndr_start) / 512;

From: Heinz Mauelshagen <[email protected]>

commit 332bd0778775d0cf105c4b9e03e460b590749916 upstream.

On dm-raid table load (using raid_ctr), dm-raid allocates an array rs->devs[rs->raid_disks] for the raid device members. rs->raid_disks is defined by the number of raid metadata and image tupples passed into the target's constructor.

In the case of RAID layout changes being requested, that number can be different from the current number of members for existing raid sets as defined in their superblocks. Example RAID layout changes include:
- raid1 legs being added/removed
- raid4/5/6/10 number of stripes changed (stripe reshaping)
- takeover to higher raid level (e.g. raid5 -> raid6)

When accessing array members, rs->raid_disks must be used in control loops instead of the potentially larger value in rs->md.raid_disks. Otherwise it will cause memory access beyond the end of the rs->devs array.

Fix this by changing code that is prone to out-of-bounds access. Also fix validate_raid_redundancy() to validate all devices that are added. Also, use braces to help clean up raid_iterate_devices().

The out-of-bounds memory accesses was discovered using KASAN.

This commit was verified to pass all LVM2 RAID tests (with KASAN enabled).

---
 drivers/md/dm-raid.c |   34 ++++++++++++++++++----------------
 1 file changed, 18 insertions(+), 16 deletions(-)

@@ -998,12 +998,13 @@ static int validate_region_size(struct r
 static int validate_raid_redundancy(struct raid_set *rs)
 {
 	unsigned int i, rebuild_cnt = 0;
-	unsigned int rebuilds_per_group = 0, copies;
+	unsigned int rebuilds_per_group = 0, copies, raid_disks;
 	unsigned int group_size, last_group_start;
 
-	for (i = 0; i < rs->md.raid_disks; i++)
-		if (!test_bit(In_sync, &rs->dev[i].rdev.flags) ||
-		    !rs->dev[i].rdev.sb_page)
+	for (i = 0; i < rs->raid_disks; i++)
+		if (!test_bit(FirstUse, &rs->dev[i].rdev.flags) &&
+		    ((!test_bit(In_sync, &rs->dev[i].rdev.flags) ||
+		      !rs->dev[i].rdev.sb_page)))
 			rebuild_cnt++;
 
 	switch (rs->md.level) {

@@ -1043,8 +1044,9 @@ static int validate_raid_redundancy(stru
 		 *	    A	 A    B	   B	C
 		 *	    C	 D    D	   E	E
 		 */
+		raid_disks = min(rs->raid_disks, rs->md.raid_disks);
 		if (__is_raid10_near(rs->md.new_layout)) {
-			for (i = 0; i < rs->md.raid_disks; i++) {
+			for (i = 0; i < raid_disks; i++) {
 				if (!(i % copies))
 					rebuilds_per_group = 0;
 				if ((!rs->dev[i].rdev.sb_page ||

@@ -1067,10 +1069,10 @@ static int validate_raid_redundancy(stru
 		 * results in the need to treat the last (potentially larger)
 		 * set differently.
 		 */
-		group_size = (rs->md.raid_disks / copies);
-		last_group_start = (rs->md.raid_disks / group_size) - 1;
+		group_size = (raid_disks / copies);
+		last_group_start = (raid_disks / group_size) - 1;
 		last_group_start *= group_size;
-		for (i = 0; i < rs->md.raid_disks; i++) {
+		for (i = 0; i < raid_disks; i++) {
 			if (!(i % copies) && !(i > last_group_start))
 				rebuilds_per_group = 0;
 			if ((!rs->dev[i].rdev.sb_page ||

@@ -1585,7 +1587,7 @@ static sector_t __rdev_sectors(struct ra
 {
 	int i;
 
-	for (i = 0; i < rs->md.raid_disks; i++) {
+	for (i = 0; i < rs->raid_disks; i++) {
 		struct md_rdev *rdev = &rs->dev[i].rdev;
 
 		if (!test_bit(Journal, &rdev->flags) &&

@@ -3751,13 +3753,13 @@ static int raid_iterate_devices(struct d
 	unsigned int i;
 	int r = 0;
 
-	for (i = 0; !r && i < rs->md.raid_disks; i++)
-		if (rs->dev[i].data_dev)
-			r = fn(ti,
-				 rs->dev[i].data_dev,
-				 0, /* No offset on data devs */
-				 rs->md.dev_sectors,
-				 data);
+	for (i = 0; !r && i < rs->raid_disks; i++) {
+		if (rs->dev[i].data_dev) {
+			r = fn(ti, rs->dev[i].data_dev,
+			       0, /* No offset on data devs */
+			       rs->md.dev_sectors, data);
+		}
+	}
 
 	return r;
 }

From: Mikulas Patocka <[email protected]>

commit 617b365872a247480e9dcd50a32c8d1806b21861 upstream.

There's a KASAN warning in raid5_add_disk when running the LVM testsuite. The warning happens in the test lvconvert-raid-reshape-linear_to_raid6-single-type.sh. We fix the warning by verifying that rdev->saved_raid_disk is within limits.

---
 drivers/md/raid5.c |    1 +
 1 file changed, 1 insertion(+)

@@ -7716,6 +7716,7 @@ static int raid5_add_disk(struct mddev *
 	 */
 	if (rdev->saved_raid_disk >= 0 &&
 	    rdev->saved_raid_disk >= first &&
+	    rdev->saved_raid_disk <= last &&
 	    conf->disks[rdev->saved_raid_disk].rdev == NULL)
 		first = rdev->saved_raid_disk;
 

From: Jason A. Donenfeld <[email protected]>

commit e4f74400308cb8abde5fdc9cad609c2aba32110c upstream.

s390x appears to present two RNG interfaces:
- a "TRNG" that gathers entropy using some hardware function; and
- a "DRBG" that takes in a seed and expands it.

Previously, the TRNG was wired up to arch_get_random_{long,int}(), but it was observed that this was being called really frequently, resulting in high overhead. So it was changed to be wired up to arch_get_random_ seed_{long,int}(), which was a reasonable decision. Later on, the DRBG was then wired up to arch_get_random_{long,int}(), with a complicated buffer filling thread, to control overhead and rate.

Fortunately, none of the performance issues matter much now. The RNG always attempts to use arch_get_random_seed_{long,int}() first, which means a complicated implementation of arch_get_random_{long,int}() isn't really valuable or useful to have around. And it's only used when reseeding, which means it won't hit the high throughput complications that were faced before.

So this commit returns to an earlier design of just calling the TRNG in arch_get_random_seed_{long,int}(), and returning false in arch_get_ random_{long,int}().

Part of what makes the simplification possible is that the RNG now seeds itself using the TRNG at bootup. But this only works if the TRNG is detected early in boot, before random_init() is called. So this commit also causes that check to happen in setup_arch().

---
 arch/s390/crypto/arch_random.c     |  111 -------------------------------------
 arch/s390/include/asm/archrandom.h |   21 ++++---
 arch/s390/kernel/setup.c           |    5 +
 3 files changed, 18 insertions(+), 119 deletions(-)

@@ -2,126 +2,17 @@
 /*
  * s390 arch random implementation.
  *
- * Copyright IBM Corp. 2017, 2018
+ * Copyright IBM Corp. 2017, 2020
  * Author(s): Harald Freudenberger
- *
- * The s390_arch_random_generate() function may be called from random.c
- * in interrupt context. So this implementation does the best to be very
- * fast. There is a buffer of random data which is asynchronously checked
- * and filled by a workqueue thread.
- * If there are enough bytes in the buffer the s390_arch_random_generate()
- * just delivers these bytes. Otherwise false is returned until the
- * worker thread refills the buffer.
- * The worker fills the rng buffer by pulling fresh entropy from the
- * high quality (but slow) true hardware random generator. This entropy
- * is then spread over the buffer with an pseudo random generator PRNG.
- * As the arch_get_random_seed_long() fetches 8 bytes and the calling
- * function add_interrupt_randomness() counts this as 1 bit entropy the
- * distribution needs to make sure there is in fact 1 bit entropy contained
- * in 8 bytes of the buffer. The current values pull 32 byte entropy
- * and scatter this into a 2048 byte buffer. So 8 byte in the buffer
- * will contain 1 bit of entropy.
- * The worker thread is rescheduled based on the charge level of the
- * buffer but at least with 500 ms delay to avoid too much CPU consumption.
- * So the max. amount of rng data delivered via arch_get_random_seed is
- * limited to 4k bytes per second.
  */
 
 #include <linux/kernel.h>
 #include <linux/atomic.h>
 #include <linux/random.h>
-#include <linux/slab.h>
 #include <linux/static_key.h>
-#include <linux/workqueue.h>
 #include <asm/cpacf.h>
 
 DEFINE_STATIC_KEY_FALSE(s390_arch_random_available);
 
 atomic64_t s390_arch_random_counter = ATOMIC64_INIT(0);
 EXPORT_SYMBOL(s390_arch_random_counter);
-
-#define ARCH_REFILL_TICKS (HZ/2)
-#define ARCH_PRNG_SEED_SIZE 32
-#define ARCH_RNG_BUF_SIZE 2048
-
-static DEFINE_SPINLOCK(arch_rng_lock);
-static u8 *arch_rng_buf;
-static unsigned int arch_rng_buf_idx;
-
-static void arch_rng_refill_buffer(struct work_struct *);
-static DECLARE_DELAYED_WORK(arch_rng_work, arch_rng_refill_buffer);
-
-bool s390_arch_random_generate(u8 *buf, unsigned int nbytes)
-{
-	/* max hunk is ARCH_RNG_BUF_SIZE */
-	if (nbytes > ARCH_RNG_BUF_SIZE)
-		return false;
-
-	/* lock rng buffer */
-	if (!spin_trylock(&arch_rng_lock))
-		return false;
-
-	/* try to resolve the requested amount of bytes from the buffer */
-	arch_rng_buf_idx -= nbytes;
-	if (arch_rng_buf_idx < ARCH_RNG_BUF_SIZE) {
-		memcpy(buf, arch_rng_buf + arch_rng_buf_idx, nbytes);
-		atomic64_add(nbytes, &s390_arch_random_counter);
-		spin_unlock(&arch_rng_lock);
-		return true;
-	}
-
-	/* not enough bytes in rng buffer, refill is done asynchronously */
-	spin_unlock(&arch_rng_lock);
-
-	return false;
-}
-EXPORT_SYMBOL(s390_arch_random_generate);
-
-static void arch_rng_refill_buffer(struct work_struct *unused)
-{
-	unsigned int delay = ARCH_REFILL_TICKS;
-
-	spin_lock(&arch_rng_lock);
-	if (arch_rng_buf_idx > ARCH_RNG_BUF_SIZE) {
-		/* buffer is exhausted and needs refill */
-		u8 seed[ARCH_PRNG_SEED_SIZE];
-		u8 prng_wa[240];
-		/* fetch ARCH_PRNG_SEED_SIZE bytes of entropy */
-		cpacf_trng(NULL, 0, seed, sizeof(seed));
-		/* blow this entropy up to ARCH_RNG_BUF_SIZE with PRNG */
-		memset(prng_wa, 0, sizeof(prng_wa));
-		cpacf_prno(CPACF_PRNO_SHA512_DRNG_SEED,
-			   &prng_wa, NULL, 0, seed, sizeof(seed));
-		cpacf_prno(CPACF_PRNO_SHA512_DRNG_GEN,
-			   &prng_wa, arch_rng_buf, ARCH_RNG_BUF_SIZE, NULL, 0);
-		arch_rng_buf_idx = ARCH_RNG_BUF_SIZE;
-	}
-	delay += (ARCH_REFILL_TICKS * arch_rng_buf_idx) / ARCH_RNG_BUF_SIZE;
-	spin_unlock(&arch_rng_lock);
-
-	/* kick next check */
-	queue_delayed_work(system_long_wq, &arch_rng_work, delay);
-}
-
-static int __init s390_arch_random_init(void)
-{
-	/* all the needed PRNO subfunctions available ? */
-	if (cpacf_query_func(CPACF_PRNO, CPACF_PRNO_TRNG) &&
-	    cpacf_query_func(CPACF_PRNO, CPACF_PRNO_SHA512_DRNG_GEN)) {
-
-		/* alloc arch random working buffer */
-		arch_rng_buf = kmalloc(ARCH_RNG_BUF_SIZE, GFP_KERNEL);
-		if (!arch_rng_buf)
-			return -ENOMEM;
-
-		/* kick worker queue job to fill the random buffer */
-		queue_delayed_work(system_long_wq,
-				   &arch_rng_work, ARCH_REFILL_TICKS);
-
-		/* enable arch random to the outside world */
-		static_branch_enable(&s390_arch_random_available);
-	}
-
-	return 0;
-}
-arch_initcall(s390_arch_random_init);

@@ -2,7 +2,7 @@
 /*
  * Kernel interface for the s390 arch_random_* functions
  *
- * Copyright IBM Corp. 2017
+ * Copyright IBM Corp. 2017, 2020
  *
  * Author: Harald Freudenberger <[email protected]>
  *

@@ -15,34 +15,37 @@
 
 #include <linux/static_key.h>
 #include <linux/atomic.h>
+#include <asm/cpacf.h>
 
 DECLARE_STATIC_KEY_FALSE(s390_arch_random_available);
 extern atomic64_t s390_arch_random_counter;
 
-bool s390_arch_random_generate(u8 *buf, unsigned int nbytes);
-
-static inline bool arch_get_random_long(unsigned long *v)
+static inline bool __must_check arch_get_random_long(unsigned long *v)
 {
 	return false;
 }
 
-static inline bool arch_get_random_int(unsigned int *v)
+static inline bool __must_check arch_get_random_int(unsigned int *v)
 {
 	return false;
 }
 
-static inline bool arch_get_random_seed_long(unsigned long *v)
+static inline bool __must_check arch_get_random_seed_long(unsigned long *v)
 {
 	if (static_branch_likely(&s390_arch_random_available)) {
-		return s390_arch_random_generate((u8 *)v, sizeof(*v));
+		cpacf_trng(NULL, 0, (u8 *)v, sizeof(*v));
+		atomic64_add(sizeof(*v), &s390_arch_random_counter);
+		return true;
 	}
 	return false;
 }
 
-static inline bool arch_get_random_seed_int(unsigned int *v)
+static inline bool __must_check arch_get_random_seed_int(unsigned int *v)
 {
 	if (static_branch_likely(&s390_arch_random_available)) {
-		return s390_arch_random_generate((u8 *)v, sizeof(*v));
+		cpacf_trng(NULL, 0, (u8 *)v, sizeof(*v));
+		atomic64_add(sizeof(*v), &s390_arch_random_counter);
+		return true;
 	}
 	return false;
 }

@@ -851,6 +851,11 @@ static void __init setup_randomness(void
 	if (stsi(vmms, 3, 2, 2) == 0 && vmms->count)
 		add_device_randomness(&vmms->vm, sizeof(vmms->vm[0]) * vmms->count);
 	memblock_free((unsigned long) vmms, PAGE_SIZE);
+
+#ifdef CONFIG_ARCH_RANDOM
+	if (cpacf_query_func(CPACF_PRNO, CPACF_PRNO_TRNG))
+		static_branch_enable(&s390_arch_random_available);
+#endif
 }
 
 /*

From: Chuck Lever <[email protected]>

commit a23dd544debcda4ee4a549ec7de59e85c3c8345c upstream.

Looks like there are still cases when "space_left - frag1bytes" can legitimately exceed PAGE_SIZE. Ensure that xdr->end always remains within the current encode buffer.

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

@@ -544,7 +544,7 @@ static __be32 *xdr_get_next_encode_buffe
 	 */
 	xdr->p = (void *)p + frag2bytes;
 	space_left = xdr->buf->buflen - xdr->buf->len;
-	if (space_left - nbytes >= PAGE_SIZE)
+	if (space_left - frag1bytes >= PAGE_SIZE)
 		xdr->end = (void *)p + PAGE_SIZE;
 	else
 		xdr->end = (void *)p + space_left - frag1bytes;

From: Duoming Zhou <[email protected]>

commit 9cc02ede696272c5271a401e4f27c262359bc2f6 upstream.

There are UAF bugs in rose_heartbeat_expiry(), rose_timer_expiry() and rose_idletimer_expiry(). The root cause is that del_timer() could not stop the timer handler that is running and the refcount of sock is not managed properly.

One of the UAF bugs is shown below:

    (thread 1)          |        (thread 2)
                        |  rose_bind
                        |  rose_connect
                        |    rose_start_heartbeat
rose_release            |    (wait a time)
  case ROSE_STATE_0     |
  rose_destroy_socket   |  rose_heartbeat_expiry
    rose_stop_heartbeat |
    sock_put(sk)        |    ...
  sock_put(sk) // FREE  |
                        |    bh_lock_sock(sk) // USE

The sock is deallocated by sock_put() in rose_release() and then used by bh_lock_sock() in rose_heartbeat_expiry().

Although rose_destroy_socket() calls rose_stop_heartbeat(), it could not stop the timer that is running.

The KASAN report triggered by POC is shown below:

BUG: KASAN: use-after-free in _raw_spin_lock+0x5a/0x110
Write of size 4 at addr ffff88800ae59098 by task swapper/3/0
...
Call Trace:
 <IRQ>
 dump_stack_lvl+0xbf/0xee
 print_address_description+0x7b/0x440
 print_report+0x101/0x230
 ? irq_work_single+0xbb/0x140
 ? _raw_spin_lock+0x5a/0x110
 kasan_report+0xed/0x120
 ? _raw_spin_lock+0x5a/0x110
 kasan_check_range+0x2bd/0x2e0
 _raw_spin_lock+0x5a/0x110
 rose_heartbeat_expiry+0x39/0x370
 ? rose_start_heartbeat+0xb0/0xb0
 call_timer_fn+0x2d/0x1c0
 ? rose_start_heartbeat+0xb0/0xb0
 expire_timers+0x1f3/0x320
 __run_timers+0x3ff/0x4d0
 run_timer_softirq+0x41/0x80
 __do_softirq+0x233/0x544
 irq_exit_rcu+0x41/0xa0
 sysvec_apic_timer_interrupt+0x8c/0xb0
 </IRQ>
 <TASK>
 asm_sysvec_apic_timer_interrupt+0x1b/0x20
RIP: 0010:default_idle+0xb/0x10
RSP: 0018:ffffc9000012fea0 EFLAGS: 00000202
RAX: 000000000000bcae RBX: ffff888006660f00 RCX: 000000000000bcae
RDX: 0000000000000001 RSI: ffffffff843a11c0 RDI: ffffffff843a1180
RBP: dffffc0000000000 R08: dffffc0000000000 R09: ffffed100da36d46
R10: dfffe9100da36d47 R11: ffffffff83cf0950 R12: 0000000000000000
R13: 1ffff11000ccc1e0 R14: ffffffff8542af28 R15: dffffc0000000000
...
Allocated by task 146:
 __kasan_kmalloc+0xc4/0xf0
 sk_prot_alloc+0xdd/0x1a0
 sk_alloc+0x2d/0x4e0
 rose_create+0x7b/0x330
 __sock_create+0x2dd/0x640
 __sys_socket+0xc7/0x270
 __x64_sys_socket+0x71/0x80
 do_syscall_64+0x43/0x90
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Freed by task 152:
 kasan_set_track+0x4c/0x70
 kasan_set_free_info+0x1f/0x40
 ____kasan_slab_free+0x124/0x190
 kfree+0xd3/0x270
 __sk_destruct+0x314/0x460
 rose_release+0x2fa/0x3b0
 sock_close+0xcb/0x230
 __fput+0x2d9/0x650
 task_work_run+0xd6/0x160
 exit_to_user_mode_loop+0xc7/0xd0
 exit_to_user_mode_prepare+0x4e/0x80
 syscall_exit_to_user_mode+0x20/0x40
 do_syscall_64+0x4f/0x90
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

This patch adds refcount of sock when we use functions such as rose_start_heartbeat() and so on to start timer, and decreases the refcount of sock when timer is finished or deleted by functions such as rose_stop_heartbeat() and so on. As a result, the UAF bugs could be mitigated.

---
 net/rose/rose_timer.c |   34 +++++++++++++++++++---------------
 1 file changed, 19 insertions(+), 15 deletions(-)

@@ -34,89 +34,89 @@ static void rose_idletimer_expiry(struct
 
 void rose_start_heartbeat(struct sock *sk)
 {
-	del_timer(&sk->sk_timer);
+	sk_stop_timer(sk, &sk->sk_timer);
 
 	sk->sk_timer.function = rose_heartbeat_expiry;
 	sk->sk_timer.expires  = jiffies + 5 * HZ;
 
-	add_timer(&sk->sk_timer);
+	sk_reset_timer(sk, &sk->sk_timer, sk->sk_timer.expires);
 }
 
 void rose_start_t1timer(struct sock *sk)
 {
 	struct rose_sock *rose = rose_sk(sk);
 
-	del_timer(&rose->timer);
+	sk_stop_timer(sk, &rose->timer);
 
 	rose->timer.function = rose_timer_expiry;
 	rose->timer.expires  = jiffies + rose->t1;
 
-	add_timer(&rose->timer);
+	sk_reset_timer(sk, &rose->timer, rose->timer.expires);
 }
 
 void rose_start_t2timer(struct sock *sk)
 {
 	struct rose_sock *rose = rose_sk(sk);
 
-	del_timer(&rose->timer);
+	sk_stop_timer(sk, &rose->timer);
 
 	rose->timer.function = rose_timer_expiry;
 	rose->timer.expires  = jiffies + rose->t2;
 
-	add_timer(&rose->timer);
+	sk_reset_timer(sk, &rose->timer, rose->timer.expires);
 }
 
 void rose_start_t3timer(struct sock *sk)
 {
 	struct rose_sock *rose = rose_sk(sk);
 
-	del_timer(&rose->timer);
+	sk_stop_timer(sk, &rose->timer);
 
 	rose->timer.function = rose_timer_expiry;
 	rose->timer.expires  = jiffies + rose->t3;
 
-	add_timer(&rose->timer);
+	sk_reset_timer(sk, &rose->timer, rose->timer.expires);
 }
 
 void rose_start_hbtimer(struct sock *sk)
 {
 	struct rose_sock *rose = rose_sk(sk);
 
-	del_timer(&rose->timer);
+	sk_stop_timer(sk, &rose->timer);
 
 	rose->timer.function = rose_timer_expiry;
 	rose->timer.expires  = jiffies + rose->hb;
 
-	add_timer(&rose->timer);
+	sk_reset_timer(sk, &rose->timer, rose->timer.expires);
 }
 
 void rose_start_idletimer(struct sock *sk)
 {
 	struct rose_sock *rose = rose_sk(sk);
 
-	del_timer(&rose->idletimer);
+	sk_stop_timer(sk, &rose->idletimer);
 
 	if (rose->idle > 0) {
 		rose->idletimer.function = rose_idletimer_expiry;
 		rose->idletimer.expires  = jiffies + rose->idle;
 
-		add_timer(&rose->idletimer);
+		sk_reset_timer(sk, &rose->idletimer, rose->idletimer.expires);
 	}
 }
 
 void rose_stop_heartbeat(struct sock *sk)
 {
-	del_timer(&sk->sk_timer);
+	sk_stop_timer(sk, &sk->sk_timer);
 }
 
 void rose_stop_timer(struct sock *sk)
 {
-	del_timer(&rose_sk(sk)->timer);
+	sk_stop_timer(sk, &rose_sk(sk)->timer);
 }
 
 void rose_stop_idletimer(struct sock *sk)
 {
-	del_timer(&rose_sk(sk)->idletimer);
+	sk_stop_timer(sk, &rose_sk(sk)->idletimer);
 }
 
 static void rose_heartbeat_expiry(struct timer_list *t)

@@ -133,6 +133,7 @@ static void rose_heartbeat_expiry(struct
 		    (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) {
 			bh_unlock_sock(sk);
 			rose_destroy_socket(sk);
+			sock_put(sk);
 			return;
 		}
 		break;

@@ -155,6 +156,7 @@ static void rose_heartbeat_expiry(struct
 
 	rose_start_heartbeat(sk);
 	bh_unlock_sock(sk);
+	sock_put(sk);
 }
 
 static void rose_timer_expiry(struct timer_list *t)

@@ -184,6 +186,7 @@ static void rose_timer_expiry(struct tim
 		break;
 	}
 	bh_unlock_sock(sk);
+	sock_put(sk);
 }
 
 static void rose_idletimer_expiry(struct timer_list *t)

@@ -208,4 +211,5 @@ static void rose_idletimer_expiry(struct
 		sock_set_flag(sk, SOCK_DEAD);
 	}
 	bh_unlock_sock(sk);
+	sock_put(sk);
 }

From: Jose Alonso <[email protected]>

commit f8ebb3ac881b17712e1d5967c97ab1806b16d3d6 upstream.

This patch corrects packet receiving in ax88179_rx_fixup.

- problem observed:
ifconfig shows allways a lot of 'RX Errors' while packets are received normally.

This occurs because ax88179_rx_fixup does not recognise properly the usb urb received.
The packets are normally processed and at the end, the code exits with 'return 0', generating RX Errors.
(pkt_cnt==-2 and ptk_hdr over field rx_hdr trying to identify
another packet there)

This is a usb urb received by "tcpdump -i usbmon2 -X" on a little-endian CPU:
0x0000: eeee f8e3 3b19 87a0 94de 80e3 daac 0800

           ^         packet 1 start (pkt_len = 0x05ec)
           ^^^^      IP alignment pseudo header
                ^    ethernet packet start
           last byte ethernet packet   v
           padding (8-bytes aligned)     vvvv vvvv
  0x05e0:  c92d d444 1420 8a69 83dd 272f e82b 9811
  0x05f0:  eeee f8e3 3b19 87a0 94de 80e3 daac 0800
  ...      ^ packet 2
  0x0be0:  eeee f8e3 3b19 87a0 94de 80e3 daac 0800
  ...
  0x1130:  9d41 9171 8a38 0ec5 eeee f8e3 3b19 87a0
  ...
  0x1720:  8cfc 15ff 5e4c e85c eeee f8e3 3b19 87a0
  ...
  0x1d10:  ecfa 2a3a 19ab c78c eeee f8e3 3b19 87a0
  ...
  0x2070:  eeee f8e3 3b19 87a0 94de 80e3 daac 0800
  ...      ^ packet 7
  0x2120:  7c88 4ca5 5c57 7dcc 0d34 7577 f778 7e0a
  0x2130:  f032 e093 7489 0740 3008 ec05 0000 0080
                               ====1==== ====2====
           hdr_off             ^
           pkt_len = 0x05ec         ^^^^
           AX_RXHDR_*=0x00830  ^^^^   ^
           pkt_len = 0                        ^^^^
           AX_RXHDR_DROP_ERR=0x80000000  ^^^^   ^
  0x2140:  3008 ec05 0000 0080 3008 5805 0000 0080
  0x2150:  3008 ec05 0000 0080 3008 ec05 0000 0080
  0x2160:  3008 5803 0000 0080 3008 c800 0000 0080
           ===11==== ===12==== ===13==== ===14====
  0x2170:  0000 0000 0e00 3821
                     ^^^^ ^^^^ rx_hdr
                     ^^^^      pkt_cnt=14
                          ^^^^ hdr_off=0x2138
           ^^^^ ^^^^           padding

The dump shows that pkt_cnt is the number of entrys in the per-packet metadata. It is "2 * packet count". Each packet have two entrys. The first have a valid value (pkt_len and AX_RXHDR_*) and the second have a dummy-header 0x80000000 (pkt_len=0 with AX_RXHDR_DROP_ERR). Why exists dummy-header for each packet?!?
My guess is that this was done probably to align the entry for each packet to 64-bits and maintain compatibility with old firmware.
There is also a padding (0x00000000) before the rx_hdr to align the end of rx_hdr to 64-bit.
Note that packets have a alignment of 64-bits (8-bytes).

This patch assumes that the dummy-header and the last padding are optional. So it preserves semantics and recognises the same valid packets as the current code.

This patch was made using only the dumpfile information and tested with only one device:
0b95:1790 ASIX Electronics Corp. AX88179 Gigabit Ethernet

---
 drivers/net/usb/ax88179_178a.c |  101 ++++++++++++++++++++++++++++++-----------
 1 file changed, 76 insertions(+), 25 deletions(-)

@@ -1377,6 +1377,42 @@ static int ax88179_rx_fixup(struct usbne
 	 * are bundled into this buffer and where we can find an array of
 	 * per-packet metadata (which contains elements encoded into u16).
 	 */
+
+	/* SKB contents for current firmware:
+	 *   <packet 1> <padding>
+	 *   ...
+	 *   <packet N> <padding>
+	 *   <per-packet metadata entry 1> <dummy header>
+	 *   ...
+	 *   <per-packet metadata entry N> <dummy header>
+	 *   <padding2> <rx_hdr>
+	 *
+	 * where:
+	 *   <packet N> contains pkt_len bytes:
+	 *		2 bytes of IP alignment pseudo header
+	 *		packet received
+	 *   <per-packet metadata entry N> contains 4 bytes:
+	 *		pkt_len and fields AX_RXHDR_*
+	 *   <padding>	0-7 bytes to terminate at
+	 *		8 bytes boundary (64-bit).
+	 *   <padding2> 4 bytes to make rx_hdr terminate at
+	 *		8 bytes boundary (64-bit)
+	 *   <dummy-header> contains 4 bytes:
+	 *		pkt_len=0 and AX_RXHDR_DROP_ERR
+	 *   <rx-hdr>	contains 4 bytes:
+	 *		pkt_cnt and hdr_off (offset of
+	 *		  <per-packet metadata entry 1>)
+	 *
+	 * pkt_cnt is number of entrys in the per-packet metadata.
+	 * In current firmware there is 2 entrys per packet.
+	 * The first points to the packet and the
+	 *  second is a dummy header.
+	 * This was done probably to align fields in 64-bit and
+	 *  maintain compatibility with old firmware.
+	 * This code assumes that <dummy header> and <padding2> are
+	 *  optional.
+	 */
+
 	if (skb->len < 4)
 		return 0;
 	skb_trim(skb, skb->len - 4);

@@ -1391,51 +1427,66 @@ static int ax88179_rx_fixup(struct usbne
 	/* Make sure that the bounds of the metadata array are inside the SKB
 	 * (and in front of the counter at the end).
 	 */
-	if (pkt_cnt * 2 + hdr_off > skb->len)
+	if (pkt_cnt * 4 + hdr_off > skb->len)
 		return 0;
 	pkt_hdr = (u32 *)(skb->data + hdr_off);
 
 	/* Packets must not overlap the metadata array */
 	skb_trim(skb, hdr_off);
 
-	for (; ; pkt_cnt--, pkt_hdr++) {
+	for (; pkt_cnt > 0; pkt_cnt--, pkt_hdr++) {
+		u16 pkt_len_plus_padd;
 		u16 pkt_len;
 
 		le32_to_cpus(pkt_hdr);
 		pkt_len = (*pkt_hdr >> 16) & 0x1fff;
+		pkt_len_plus_padd = (pkt_len + 7) & 0xfff8;
 
-		if (pkt_len > skb->len)
+		/* Skip dummy header used for alignment
+		 */
+		if (pkt_len == 0)
+			continue;
+
+		if (pkt_len_plus_padd > skb->len)
 			return 0;
 
 		/* Check CRC or runt packet */
-		if (((*pkt_hdr & (AX_RXHDR_CRC_ERR | AX_RXHDR_DROP_ERR)) == 0) &&
-		    pkt_len >= 2 + ETH_HLEN) {
-			bool last = (pkt_cnt == 0);
-
-			if (last) {
-				ax_skb = skb;
-			} else {
-				ax_skb = skb_clone(skb, GFP_ATOMIC);
-				if (!ax_skb)
-					return 0;
-			}
-			ax_skb->len = pkt_len;
-			/* Skip IP alignment pseudo header */
-			skb_pull(ax_skb, 2);
-			skb_set_tail_pointer(ax_skb, ax_skb->len);
-			ax_skb->truesize = pkt_len + sizeof(struct sk_buff);
-			ax88179_rx_checksum(ax_skb, pkt_hdr);
+		if ((*pkt_hdr & (AX_RXHDR_CRC_ERR | AX_RXHDR_DROP_ERR)) ||
+		    pkt_len < 2 + ETH_HLEN) {
+			dev->net->stats.rx_errors++;
+			skb_pull(skb, pkt_len_plus_padd);
+			continue;
+		}
 
-			if (last)
-				return 1;
+		/* last packet */
+		if (pkt_len_plus_padd == skb->len) {
+			skb_trim(skb, pkt_len);
 
-			usbnet_skb_return(dev, ax_skb);
+			/* Skip IP alignment pseudo header */
+			skb_pull(skb, 2);
+
+			skb->truesize = SKB_TRUESIZE(pkt_len_plus_padd);
+			ax88179_rx_checksum(skb, pkt_hdr);
+			return 1;
 		}
 
-		/* Trim this packet away from the SKB */
-		if (!skb_pull(skb, (pkt_len + 7) & 0xFFF8))
+		ax_skb = skb_clone(skb, GFP_ATOMIC);
+		if (!ax_skb)
 			return 0;
+		skb_trim(ax_skb, pkt_len);
+
+		/* Skip IP alignment pseudo header */
+		skb_pull(ax_skb, 2);
+
+		skb->truesize = pkt_len_plus_padd +
+				SKB_DATA_ALIGN(sizeof(struct sk_buff));
+		ax88179_rx_checksum(ax_skb, pkt_hdr);
+		usbnet_skb_return(dev, ax_skb);
+
+		skb_pull(skb, pkt_len_plus_padd);
 	}
+
+	return 0;
 }
 
 static struct sk_buff *

From: Jason Wang <[email protected]>

commit 50c0ada627f56c92f5953a8bf9158b045ad026a1 upstream.

We currently call virtio_device_ready() after netdev registration. Since ndo_open() can be called immediately after register_netdev, this means there exists a race between ndo_open() and virtio_device_ready(): the driver may start to use the device before DRIVER_OK which violates the spec.

Fix this by switching to use register_netdevice() and protect the virtio_device_ready() with rtnl_lock() to make sure ndo_open() can only be called after virtio_device_ready().

Message-Id: <[email protected]>

---
 drivers/net/virtio_net.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

@@ -3105,14 +3105,20 @@ static int virtnet_probe(struct virtio_d
 		}
 	}
 
-	err = register_netdev(dev);
+	/* serialize netdev register + virtio_device_ready() with ndo_open() */
+	rtnl_lock();
+
+	err = register_netdevice(dev);
 	if (err) {
 		pr_debug("virtio_net: registering device failed\n");
+		rtnl_unlock();
 		goto free_failover;
 	}
 
 	virtio_device_ready(vdev);
 
+	rtnl_unlock();
+
 	err = virtnet_cpu_notif_add(vi);
 	if (err) {
 		pr_debug("virtio_net: registering cpu notifier failed\n");

From: Dimitris Michailidis <[email protected]>

commit b968080808f7f28b89aa495b7402ba48eb17ee93 upstream.

udpgso_bench.sh has been running its IPv6 TCP test with IPv4 arguments since its initial conmit. Looks like a typo.

---
 tools/testing/selftests/net/udpgso_bench.sh |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -57,7 +57,7 @@ run_all() {
 	run_udp "${ipv4_args}"
 
 	echo "ipv6"
-	run_tcp "${ipv4_args}"
+	run_tcp "${ipv6_args}"
 	run_udp "${ipv6_args}"
 }
 

From: Jakub Kicinski <[email protected]>

commit 3b9bc84d311104906d2b4995a9a02d7b7ddab2db upstream.

Syzbot found a race between tun file and device destruction. NAPIs live in struct tun_file which can get destroyed before the netdev so we have to del them explicitly. The current code is missing deleting the NAPI if the queue was detached first.

---
 drivers/net/tun.c |    1 +
 1 file changed, 1 insertion(+)

@@ -797,6 +797,7 @@ static void tun_detach_all(struct net_de
 		sock_put(&tfile->sk);
 	}
 	list_for_each_entry_safe(tfile, tmp, &tun->disabled, next) {
+		tun_napi_del(tfile);
 		tun_enable_queue(tfile);
 		tun_queue_purge(tfile);
 		xdp_rxq_info_unreg(&tfile->xdp_rxq);

From: Jakub Kicinski <[email protected]>

commit a8fc8cb5692aebb9c6f7afd4265366d25dcd1d01 upstream.

While looking at a syzbot report I noticed the NAPI only gets disabled before it's deleted. I think that user can detach the queue before destroying the device and the NAPI will never be stopped.

---
 drivers/net/tun.c |   11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

@@ -333,6 +333,12 @@ static void tun_napi_init(struct tun_str
 	}
 }
 
+static void tun_napi_enable(struct tun_file *tfile)
+{
+	if (tfile->napi_enabled)
+		napi_enable(&tfile->napi);
+}
+
 static void tun_napi_disable(struct tun_file *tfile)
 {
 	if (tfile->napi_enabled)

@@ -723,8 +729,10 @@ static void __tun_detach(struct tun_file
 		if (clean) {
 			RCU_INIT_POINTER(tfile->tun, NULL);
 			sock_put(&tfile->sk);
-		} else
+		} else {
 			tun_disable_queue(tun, tfile);
+			tun_napi_disable(tfile);
+		}
 
 		synchronize_net();
 		tun_flow_delete_by_queue(tun, tun->numqueues + 1);

@@ -878,6 +886,7 @@ static int tun_attach(struct tun_struct
 
 	if (tfile->detached) {
 		tun_enable_queue(tfile);
+		tun_napi_enable(tfile);
 	} else {
 		sock_hold(&tfile->sk);
 		tun_napi_init(tun, tfile, napi, napi_frags);

From: Kamal Heib <[email protected]>

commit 118f767413ada4eef7825fbd4af7c0866f883441 upstream.

Make sure to save the passed QP timeout attribute when the QP gets modified, so when calling query QP the right value is reported and not the converted value that is required by the firmware. This issue was found while running the pyverbs tests.

---
 drivers/infiniband/hw/qedr/qedr.h  |    1 +
 drivers/infiniband/hw/qedr/verbs.c |    4 +++-
 2 files changed, 4 insertions(+), 1 deletion(-)

@@ -407,6 +407,7 @@ struct qedr_qp {
 	u32 sq_psn;
 	u32 qkey;
 	u32 dest_qp_num;
+	u8 timeout;
 
 	/* Relevant to qps created from kernel space only (ULPs) */
 	u8 prev_wqe_size;

@@ -2376,6 +2376,8 @@ int qedr_modify_qp(struct ib_qp *ibqp, s
 					1 << max_t(int, attr->timeout - 8, 0);
 		else
 			qp_params.ack_timeout = 0;
+
+		qp->timeout = attr->timeout;
 	}
 
 	if (attr_mask & IB_QP_RETRY_CNT) {

@@ -2535,7 +2537,7 @@ int qedr_query_qp(struct ib_qp *ibqp,
 	rdma_ah_set_dgid_raw(&qp_attr->ah_attr, &params.dgid.bytes[0]);
 	rdma_ah_set_port_num(&qp_attr->ah_attr, 1);
 	rdma_ah_set_sl(&qp_attr->ah_attr, 0);
-	qp_attr->timeout = params.timeout;
+	qp_attr->timeout = qp->timeout;
 	qp_attr->rnr_retry = params.rnr_retry;
 	qp_attr->retry_cnt = params.retry_cnt;
 	qp_attr->min_rnr_timer = params.min_rnr_nak_timer;

From: Oliver Neukum <[email protected]>

commit e65af5403e462ccd7dff6a045a886c64da598c2e upstream.

usbnet provides some helper functions that are also used in the context of reset() operations. During a reset the other drivers on a device are unable to operate. As that can be block drivers, a driver for another interface cannot use paging in its memory allocations without risking a deadlock. Use GFP_NOIO in the helpers.

---
 drivers/net/usb/usbnet.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -1998,7 +1998,7 @@ static int __usbnet_read_cmd(struct usbn
 		   cmd, reqtype, value, index, size);
 
 	if (size) {
-		buf = kmalloc(size, GFP_KERNEL);
+		buf = kmalloc(size, GFP_NOIO);
 		if (!buf)
 			goto out;
 	}

@@ -2030,7 +2030,7 @@ static int __usbnet_write_cmd(struct usb
 		   cmd, reqtype, value, index, size);
 
 	if (data) {
-		buf = kmemdup(data, size, GFP_KERNEL);
+		buf = kmemdup(data, size, GFP_NOIO);
 		if (!buf)
 			goto out;
 	} else {

From: YueHaibing <[email protected]>

commit 53ad46169fe2996fe1b623ba6c9c4fa33847876f upstream.

As of commit 5801f064e351 ("net: ipv6: unexport __init-annotated seg6_hmac_init()"), EXPORT_SYMBOL and __init is a bad combination because the .init.text section is freed up after the initialization. Hence, modules cannot use symbols annotated __init. The access to a freed symbol may end up with kernel panic.

This remove the EXPORT_SYMBOL to fix modpost warning:

WARNING: modpost: vmlinux.o(___ksymtab+seg6_hmac_net_init+0x0): Section mismatch in reference from the variable __ksymtab_seg6_hmac_net_init to the function .init.text:seg6_hmac_net_init() The symbol seg6_hmac_net_init is exported and annotated __init
Fix this by removing the __init annotation of seg6_hmac_net_init or drop the export.

---
 net/ipv6/seg6_hmac.c |    1 -
 1 file changed, 1 deletion(-)

@@ -415,7 +415,6 @@ int __net_init seg6_hmac_net_init(struct
 
 	return 0;
 }
-EXPORT_SYMBOL(seg6_hmac_net_init);
 
 void seg6_hmac_exit(void)
 {

From: Jason Wang <[email protected]>

commit 11a37eb66812ce6a06b79223ad530eb0e1d7294d upstream.

We currently depend on probe() calling virtio_device_ready() - which happens after netdev
registration. Since ndo_open() can be called immediately after register_netdev, this means there exists a race between ndo_open() and virtio_device_ready(): the driver may start to use the device (e.g. TX) before DRIVER_OK which violates the spec.

Fix this by switching to use register_netdevice() and protect the virtio_device_ready() with rtnl_lock() to make sure ndo_open() can only be called after virtio_device_ready().

Message-Id: <[email protected]>

---
 drivers/net/caif/caif_virtio.c |   10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

@@ -727,13 +727,21 @@ static int cfv_probe(struct virtio_devic
 	/* Carrier is off until netdevice is opened */
 	netif_carrier_off(netdev);
 
+	/* serialize netdev register + virtio_device_ready() with ndo_open() */
+	rtnl_lock();
+
 	/* register Netdev */
-	err = register_netdev(netdev);
+	err = register_netdevice(netdev);
 	if (err) {
+		rtnl_unlock();
 		dev_err(&vdev->dev, "Unable to register netdev (%d)\n", err);
 		goto err;
 	}
 
+	virtio_device_ready(vdev);
+
+	rtnl_unlock();
+
 	debugfs_init(cfv);
 
 	return 0;

From: Pablo Neira Ayuso <[email protected]>

commit 05907f10e235680cc7fb196810e4ad3215d5e648 upstream.

This patch fixes a race condition.

nft_rhash_update() might fail for two reasons:

- Element already exists in the hashtable.
- Another packet won race to insert an entry in the hashtable.

In both cases, new() has already bumped the counter via atomic_add_unless(), therefore, decrement the set element counter.

---
 net/netfilter/nft_set_hash.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -145,6 +145,7 @@ static bool nft_rhash_update(struct nft_
 	/* Another cpu may race to insert the element with the same key */
 	if (prev) {
 		nft_set_elem_destroy(set, he, true);
+		atomic_dec(&set->nelems);
 		he = prev;
 	}
 

@@ -154,6 +155,7 @@ out:
 
 err2:
 	nft_set_elem_destroy(set, he, true);
+	atomic_dec(&set->nelems);
 err1:
 	return false;
 }

From: Eric Dumazet <[email protected]>

commit ab84db251c04d38b8dc7ee86e13d4050bedb1c88 upstream.

syzbot has two reports involving the same root cause.

bond_alb_initialize() must not set bond->alb_info.rlb_enabled if a memory allocation error is detected.

Report 1:

general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017]
CPU: 0 PID: 12276 Comm: kworker/u4:10 Not tainted 5.19.0-rc3-syzkaller-00132-g3b89b511ea0c #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: netns cleanup_net
RIP: 0010:rlb_clear_slave+0x10e/0x690 drivers/net/bonding/bond_alb.c:393
Code: 8e fc 83 fb ff 0f 84 74 02 00 00 e8 cc 2a 8e fc 48 8b 44 24 08 89 dd 48 c1 e5 06 4c 8d 34 28 49 8d 7e 14 48 89 f8 48 c1 e8 03 <42> 0f b6 14 20 48 89 f8 83 e0 07 83 c0 03 38 d0 7c 08 84 d2 0f 85
RSP: 0018:ffffc90018a8f678 EFLAGS: 00010203
RAX: 0000000000000002 RBX: 0000000000000000 RCX: 0000000000000000
RDX: ffff88803375bb00 RSI: ffffffff84ec4ac4 RDI: 0000000000000014
RBP: 0000000000000000 R08: 0000000000000005 R09: 00000000ffffffff
R10: 0000000000000000 R11: 0000000000000000 R12: dffffc0000000000
R13: ffff8880ac889000 R14: 0000000000000000 R15: ffff88815a668c80
FS: 0000000000000000(0000) GS:ffff8880b9a00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00005597077e10b0 CR3: 0000000026668000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
bond_alb_deinit_slave+0x43c/0x6b0 drivers/net/bonding/bond_alb.c:1663
__bond_release_one.cold+0x383/0xd53 drivers/net/bonding/bond_main.c:2370
bond_slave_netdev_event drivers/net/bonding/bond_main.c:3778 [inline]
bond_netdev_event+0x993/0xad0 drivers/net/bonding/bond_main.c:3889
notifier_call_chain+0xb5/0x200 kernel/notifier.c:87
call_netdevice_notifiers_info+0xb5/0x130 net/core/dev.c:1945
call_netdevice_notifiers_extack net/core/dev.c:1983 [inline]
call_netdevice_notifiers net/core/dev.c:1997 [inline]
unregister_netdevice_many+0x948/0x18b0 net/core/dev.c:10839
default_device_exit_batch+0x449/0x590 net/core/dev.c:11333
ops_exit_list+0x125/0x170 net/core/net_namespace.c:167
cleanup_net+0x4ea/0xb00 net/core/net_namespace.c:594
process_one_work+0x996/0x1610 kernel/workqueue.c:2289
worker_thread+0x665/0x1080 kernel/workqueue.c:2436
kthread+0x2e9/0x3a0 kernel/kthread.c:376
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:302
</TASK>

Report 2:

general protection fault, probably for non-canonical address 0xdffffc0000000006: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000030-0x0000000000000037]
CPU: 1 PID: 5206 Comm: syz-executor.1 Not tainted 5.18.0-syzkaller-12108-g58f9d52ff689 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:rlb_req_update_slave_clients+0x109/0x2f0 drivers/net/bonding/bond_alb.c:502
Code: 5d 18 8f fc 41 80 3e 00 0f 85 a5 01 00 00 89 d8 48 c1 e0 06 49 03 84 24 68 01 00 00 48 8d 78 30 49 89 c7 48 89 fa 48 c1 ea 03 <80> 3c 2a 00 0f 85 98 01 00 00 4d 39 6f 30 75 83 e8 22 18 8f fc 49
RSP: 0018:ffffc9000300ee80 EFLAGS: 00010206
RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffc90016c11000
RDX: 0000000000000006 RSI: ffffffff84eb6bf3 RDI: 0000000000000030
RBP: dffffc0000000000 R08: 0000000000000005 R09: 00000000ffffffff
R10: 0000000000000000 R11: 0000000000000000 R12: ffff888027c80c80
R13: ffff88807d7ff800 R14: ffffed1004f901bd R15: 0000000000000000
FS:  00007f6f46c58700(0000) GS:ffff8880b9b00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000020010000 CR3: 00000000516cc000 CR4: 00000000003506e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 alb_fasten_mac_swap+0x886/0xa80 drivers/net/bonding/bond_alb.c:1070
 bond_alb_handle_active_change+0x624/0x1050 drivers/net/bonding/bond_alb.c:1765
 bond_change_active_slave+0xfa1/0x29b0 drivers/net/bonding/bond_main.c:1173
 bond_select_active_slave+0x23f/0xa50 drivers/net/bonding/bond_main.c:1253
 bond_enslave+0x3b34/0x53b0 drivers/net/bonding/bond_main.c:2159
 do_set_master+0x1c8/0x220 net/core/rtnetlink.c:2577
 rtnl_newlink_create net/core/rtnetlink.c:3380 [inline]
 __rtnl_newlink+0x13ac/0x17e0 net/core/rtnetlink.c:3580
 rtnl_newlink+0x64/0xa0 net/core/rtnetlink.c:3593
 rtnetlink_rcv_msg+0x43a/0xc90 net/core/rtnetlink.c:6089
 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2501
 netlink_unicast_kernel net/netlink/af_netlink.c:1319 [inline]
 netlink_unicast+0x543/0x7f0 net/netlink/af_netlink.c:1345
 netlink_sendmsg+0x917/0xe10 net/netlink/af_netlink.c:1921
 sock_sendmsg_nosec net/socket.c:714 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:734
 ____sys_sendmsg+0x6eb/0x810 net/socket.c:2492
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2546
 __sys_sendmsg net/socket.c:2575 [inline]
 __do_sys_sendmsg net/socket.c:2584 [inline]
 __se_sys_sendmsg net/socket.c:2582 [inline]
 __x64_sys_sendmsg+0x132/0x220 net/socket.c:2582
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x46/0xb0
RIP: 0033:0x7f6f45a89109
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f6f46c58168 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
RAX: ffffffffffffffda RBX: 00007f6f45b9c030 RCX: 00007f6f45a89109
RDX: 0000000000000000 RSI: 0000000020000080 RDI: 0000000000000006
RBP: 00007f6f45ae308d R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007ffed99029af R14: 00007f6f46c58300 R15: 0000000000022000
 </TASK>

---
 drivers/net/bonding/bond_alb.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1292,12 +1292,12 @@ int bond_alb_initialize(struct bonding *
 		return res;
 
 	if (rlb_enabled) {
-		bond->alb_info.rlb_enabled = 1;
 		res = rlb_initialize(bond);
 		if (res) {
 			tlb_deinitialize(bond);
 			return res;
 		}
+		bond->alb_info.rlb_enabled = 1;
 	} else {
 		bond->alb_info.rlb_enabled = 0;
 	}

From: Yevhen Orlov <[email protected]>

commit 050133e1aa2cb49bb17be847d48a4431598ef562 upstream.

commit 0622cab0341c ("bonding: fix 802.3ad aggregator reselection"), resolve case, when there is several aggregation groups in the same bond. bond_3ad_unbind_slave will invalidate (clear) aggregator when __agg_active_ports return zero. So, ad_clear_agg can be executed even, when num_of_ports!=0. Than bond_3ad_unbind_slave can be executed again for, previously cleared aggregator. NOTE: at this time bond_3ad_unbind_slave will not update slave ports list, because lag_ports==NULL. So, here we got slave ports, pointing to freed aggregator memory.

Fix with checking actual number of ports in group (as was before commit 0622cab0341c ("bonding: fix 802.3ad aggregator reselection") ), before ad_clear_agg().

The KASAN logs are as follows:

[  767.617392] ==================================================================
[  767.630776] BUG: KASAN: use-after-free in bond_3ad_state_machine_handler+0x13dc/0x1470
[  767.638764] Read of size 2 at addr ffff00011ba9d430 by task kworker/u8:7/767
[  767.647361] CPU: 3 PID: 767 Comm: kworker/u8:7 Tainted: G           O 5.15.11 #15
[  767.655329] Hardware name: DNI AmazonGo1 A7040 board (DT)
[  767.660760] Workqueue: lacp_1 bond_3ad_state_machine_handler
[  767.666468] Call trace:
[  767.668930]  dump_backtrace+0x0/0x2d0
[  767.672625]  show_stack+0x24/0x30
[  767.675965]  dump_stack_lvl+0x68/0x84
[  767.679659]  print_address_description.constprop.0+0x74/0x2b8
[  767.685451]  kasan_report+0x1f0/0x260
[  767.689148]  __asan_load2+0x94/0xd0
[  767.692667]  bond_3ad_state_machine_handler+0x13dc/0x1470

---
 drivers/net/bonding/bond_3ad.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -2199,7 +2199,8 @@ void bond_3ad_unbind_slave(struct slave
 				temp_aggregator->num_of_ports--;
 				if (__agg_active_ports(temp_aggregator) == 0) {
 					select_new_active_agg = temp_aggregator->is_active;
-					ad_clear_agg(temp_aggregator);
+					if (temp_aggregator->num_of_ports == 0)
+						ad_clear_agg(temp_aggregator);
 					if (select_new_active_agg) {
 						netdev_info(bond->dev, "Removing an active aggregator\n");
 						/* select new active aggregator */

From: Krzysztof Kozlowski <[email protected]>

commit 5a478a653b4cca148d5c89832f007ec0809d7e6d upstream.

The irq_of_parse_and_map() returns 0 on failure, not a negative ERRNO.

---
 drivers/nfc/nfcmrvl/i2c.c |    6 +++---
 drivers/nfc/nfcmrvl/spi.c |    6 +++---
 2 files changed, 6 insertions(+), 6 deletions(-)

@@ -186,9 +186,9 @@ static int nfcmrvl_i2c_parse_dt(struct d
 		pdata->irq_polarity = IRQF_TRIGGER_RISING;
 
 	ret = irq_of_parse_and_map(node, 0);
-	if (ret < 0) {
-		pr_err("Unable to get irq, error: %d\n", ret);
-		return ret;
+	if (!ret) {
+		pr_err("Unable to get irq\n");
+		return -EINVAL;
 	}
 	pdata->irq = ret;
 

@@ -129,9 +129,9 @@ static int nfcmrvl_spi_parse_dt(struct d
 	}
 
 	ret = irq_of_parse_and_map(node, 0);
-	if (ret < 0) {
-		pr_err("Unable to get irq, error: %d\n", ret);
-		return ret;
+	if (!ret) {
+		pr_err("Unable to get irq\n");
+		return -EINVAL;
 	}
 	pdata->irq = ret;
 

From: Michael Walle <[email protected]>

commit eddd95b9423946aaacb55cac6a9b2cea8ab944fc upstream.

There are packets which doesn't have a payload. In that case, the second i2c_master_read() will have a zero length. But because the NFC controller doesn't have any data left, it will NACK the I2C read and -ENXIO will be returned. In case there is no payload, just skip the second i2c master read.

---
 drivers/nfc/nxp-nci/i2c.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -178,6 +178,9 @@ static int nxp_nci_i2c_nci_read(struct n
 
 	skb_put_data(*skb, (void *)&header, NCI_CTRL_HDR_SIZE);
 
+	if (!header.plen)
+		return 0;
+
 	r = i2c_master_recv(client, skb_put(*skb, header.plen), header.plen);
 	if (r != header.plen) {
 		nfc_err(&client->dev,

From: Jakub Kicinski <[email protected]>

commit ff1fa2081d173b01cebe2fbf0a2d0f1cee9ce4b5 upstream.

Eric reports that syzbot made short work out of my speculative fix. Indeed when queue gets detached its tfile->tun remains, so we would try to stop NAPI twice with a detach(), close() sequence.

Alternative fix would be to move tun_napi_disable() to tun_detach_all() and let the NAPI run after the queue has been detached.

---
 drivers/net/tun.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -710,7 +710,8 @@ static void __tun_detach(struct tun_file
 	tun = rtnl_dereference(tfile->tun);
 
 	if (tun && clean) {
-		tun_napi_disable(tfile);
+		if (!tfile->detached)
+			tun_napi_disable(tfile);
 		tun_napi_del(tfile);
 	}
 

From: Demi Marie Obenour <[email protected]>

commit dbe97cff7dd9f0f75c524afdd55ad46be3d15295 upstream.

unmap_grant_pages() currently waits for the pages to no longer be used. In https://github.com/QubesOS/qubes-issues/issues/7481, this lead to a deadlock against i915: i915 was waiting for gntdev's MMU notifier to finish, while gntdev was waiting for i915 to free its pages. I also believe this is responsible for various deadlocks I have experienced in the past.

Avoid these problems by making unmap_grant_pages async. This requires making it return void, as any errors will not be available when the function returns. Fortunately, the only use of the return value is a WARN_ON(), which can be replaced by a WARN_ON when the error is detected. Additionally, a failed call will not prevent further calls from being made, but this is harmless.

Because unmap_grant_pages is now async, the grant handle will be sent to INVALID_GRANT_HANDLE too late to prevent multiple unmaps of the same handle. Instead, a separate bool array is allocated for this purpose. This wastes memory, but stuffing this information in padding bytes is too fragile. Furthermore, it is necessary to grab a reference to the map before making the asynchronous call, and release the reference when the call returns.

It is also necessary to guard against reentrancy in gntdev_map_put(), and to handle the case where userspace tries to map a mapping whose contents have not all been freed yet.

---
 drivers/xen/gntdev-common.h |    8 ++
 drivers/xen/gntdev.c        |  146 ++++++++++++++++++++++++++++++--------------
 2 files changed, 109 insertions(+), 45 deletions(-)

@@ -15,6 +15,8 @@
 #include <linux/mman.h>
 #include <linux/mmu_notifier.h>
 #include <linux/types.h>
+#include <xen/interface/event_channel.h>
+#include <xen/grant_table.h>
 
 struct gntdev_dmabuf_priv;
 

@@ -61,6 +63,7 @@ struct gntdev_grant_map {
 	struct gnttab_unmap_grant_ref *unmap_ops;
 	struct gnttab_map_grant_ref   *kmap_ops;
 	struct gnttab_unmap_grant_ref *kunmap_ops;
+	bool *being_removed;
 	struct page **pages;
 	unsigned long pages_vm_start;
 

@@ -78,6 +81,11 @@ struct gntdev_grant_map {
 	/* Needed to avoid allocation in gnttab_dma_free_pages(). */
 	xen_pfn_t *frames;
 #endif
+
+	/* Number of live grants */
+	atomic_t live_grants;
+	/* Needed to avoid allocation in __unmap_grant_pages */
+	struct gntab_unmap_queue_data unmap_data;
 };
 
 struct gntdev_grant_map *gntdev_alloc_map(struct gntdev_priv *priv, int count,

From: Yang Yingliang <[email protected]>

[ Upstream commit d0e51022a025ca5350fafb8e413a6fe5d4baf833 ]

If platform_device_add() fails, it no need to call platform_device_del(), split platform_device_unregister() into platform_device_del/put(), so platform_device_put() can be called separately.

---
 drivers/hwmon/ibmaem.c |   12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

@@ -563,7 +563,7 @@ static int aem_init_aem1_inst(struct aem
 
 	res = platform_device_add(data->pdev);
 	if (res)
-		goto ipmi_err;
+		goto dev_add_err;
 
 	platform_set_drvdata(data->pdev, data);
 

@@ -611,7 +611,9 @@ hwmon_reg_err:
 	ipmi_destroy_user(data->ipmi.user);
 ipmi_err:
 	platform_set_drvdata(data->pdev, NULL);
-	platform_device_unregister(data->pdev);
+	platform_device_del(data->pdev);
+dev_add_err:
+	platform_device_put(data->pdev);
 dev_err:
 	ida_simple_remove(&aem_ida, data->id);
 id_err:

@@ -703,7 +705,7 @@ static int aem_init_aem2_inst(struct aem
 
 	res = platform_device_add(data->pdev);
 	if (res)
-		goto ipmi_err;
+		goto dev_add_err;
 
 	platform_set_drvdata(data->pdev, data);
 

@@ -751,7 +753,9 @@ hwmon_reg_err:
 	ipmi_destroy_user(data->ipmi.user);
 ipmi_err:
 	platform_set_drvdata(data->pdev, NULL);
-	platform_device_unregister(data->pdev);
+	platform_device_del(data->pdev);
+dev_add_err:
+	platform_device_put(data->pdev);
 dev_err:
 	ida_simple_remove(&aem_ida, data->id);
 id_err:

From: Doug Berger <[email protected]>

commit 7c97bc0128b2eecc703106112679a69d446d1a12 upstream.

The pause settings reported by the PHY should also be applied to the GMII port status override otherwise the switch will not generate pause frames towards the link partner despite the advertisement saying otherwise.

---
 drivers/net/dsa/bcm_sf2.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -599,6 +599,11 @@ force_link:
 		reg |= LINK_STS;
 	if (state->duplex == DUPLEX_FULL)
 		reg |= DUPLX_MODE;
+	if (state->pause & MLO_PAUSE_TXRX_MASK) {
+		if (state->pause & MLO_PAUSE_TX)
+			reg |= TXFLOW_CNTL;
+		reg |= RXFLOW_CNTL;
+	}
 
 	core_writel(priv, reg, offset);
 }

From: kernel test robot <[email protected]>

commit 284fda1eff8a8b27d2cafd7dc8fb423d13720f21 upstream.

Opportunity for min()

Generated by: scripts/coccinelle/misc/minmax.cocci

---
 net/ipv6/sit.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -316,7 +316,7 @@ static int ipip6_tunnel_get_prl(struct i
 
 	rcu_read_lock();
 
-	ca = t->prl_count < cmax ? t->prl_count : cmax;
+	ca = min(t->prl_count, cmax);
 
 	if (!kp) {
 		/* We don't try hard to allocate much memory for

From: katrinzhou <[email protected]>

commit adabdd8f6acabc0c3fdbba2e7f5a2edd9c5ef22d upstream.

When kcalloc fails, ipip6_tunnel_get_prl() should return -ENOMEM. Move the position of label "out" to return correctly.

Addresses-Coverity: ("Unused value")

---
 net/ipv6/sit.c |    8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)

@@ -314,8 +314,6 @@ static int ipip6_tunnel_get_prl(struct i
 		kcalloc(cmax, sizeof(*kp), GFP_KERNEL | __GFP_NOWARN) :
 		NULL;
 
-	rcu_read_lock();
-
 	ca = min(t->prl_count, cmax);
 
 	if (!kp) {

@@ -331,7 +329,7 @@ static int ipip6_tunnel_get_prl(struct i
 		}
 	}
 
-	c = 0;
+	rcu_read_lock();
 	for_each_prl_rcu(t->prl) {
 		if (c >= cmax)
 			break;

@@ -343,7 +341,7 @@ static int ipip6_tunnel_get_prl(struct i
 		if (kprl.addr != htonl(INADDR_ANY))
 			break;
 	}
-out:
+
 	rcu_read_unlock();
 
 	len = sizeof(*kp) * c;

@@ -352,7 +350,7 @@ out:
 		ret = -EFAULT;
 
 	kfree(kp);
-
+out:
 	return ret;
 }
 

From: Roger Pau Monne <[email protected]>

commit 2f446ffe9d737e9a844b97887919c4fda18246e7 upstream.

When allocating pages to be used for shared communication with the backend always zero them, this avoids leaking unintended data present on the pages.

This is CVE-2022-26365, part of XSA-403.

---
 drivers/block/xen-blkfront.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

@@ -301,7 +301,7 @@ static int fill_grant_buffer(struct blkf
 			goto out_of_memory;
 
 		if (info->feature_persistent) {
-			granted_page = alloc_page(GFP_NOIO);
+			granted_page = alloc_page(GFP_NOIO | __GFP_ZERO);
 			if (!granted_page) {
 				kfree(gnt_list_entry);
 				goto out_of_memory;

@@ -1744,7 +1744,7 @@ static int setup_blkring(struct xenbus_d
 	for (i = 0; i < info->nr_ring_pages; i++)
 		rinfo->ring_ref[i] = GRANT_INVALID_REF;
 
-	sring = alloc_pages_exact(ring_size, GFP_NOIO);
+	sring = alloc_pages_exact(ring_size, GFP_NOIO | __GFP_ZERO);
 	if (!sring) {
 		xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
 		return -ENOMEM;

@@ -2283,7 +2283,8 @@ static int blkfront_setup_indirect(struc
 
 		BUG_ON(!list_empty(&rinfo->indirect_pages));
 		for (i = 0; i < num; i++) {
-			struct page *indirect_page = alloc_page(GFP_KERNEL);
+			struct page *indirect_page = alloc_page(GFP_KERNEL |
+			                                        __GFP_ZERO);
 			if (!indirect_page)
 				goto out_of_memory;
 			list_add(&indirect_page->lru, &rinfo->indirect_pages);

From: Roger Pau Monne <[email protected]>

commit 307c8de2b02344805ebead3440d8feed28f2f010 upstream.

When allocating pages to be used for shared communication with the backend always zero them, this avoids leaking unintended data present on the pages.

This is CVE-2022-33740, part of XSA-403.

---
 drivers/net/xen-netfront.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -261,7 +261,7 @@ static struct sk_buff *xennet_alloc_one_
 	if (unlikely(!skb))
 		return NULL;
 
-	page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
+	page = alloc_page(GFP_ATOMIC | __GFP_NOWARN | __GFP_ZERO);
 	if (!page) {
 		kfree_skb(skb);
 		return NULL;

From: Roger Pau Monne <[email protected]>

commit 4491001c2e0fa69efbb748c96ec96b100a5cdb7e upstream.

Bounce all data on the skbs to be transmitted into zeroed pages if the backend is untrusted. This avoids leaking data present in the pages shared with the backend but not part of the skb fragments. This requires introducing a new helper in order to allocate skbs with a size multiple of XEN_PAGE_SIZE so we don't leak contiguous data on the granted pages.

Reporting whether the backend is to be trusted can be done using a module parameter, or from the xenstore frontend path as set by the toolstack when adding the device.

This is CVE-2022-33741, part of XSA-403.

---
 drivers/net/xen-netfront.c |   50 +++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 48 insertions(+), 2 deletions(-)

@@ -63,6 +63,10 @@ module_param_named(max_queues, xennet_ma
 MODULE_PARM_DESC(max_queues,
 		 "Maximum number of queues per virtual interface");
 
+static bool __read_mostly xennet_trusted = true;
+module_param_named(trusted, xennet_trusted, bool, 0644);
+MODULE_PARM_DESC(trusted, "Is the backend trusted");
+
 #define XENNET_TIMEOUT  (5 * HZ)
 
 static const struct ethtool_ops xennet_ethtool_ops;

@@ -163,6 +167,9 @@ struct netfront_info {
 	/* Is device behaving sane? */
 	bool broken;
 
+	/* Should skbs be bounced into a zeroed buffer? */
+	bool bounce;
+
 	atomic_t rx_gso_checksum_fixup;
 };
 

@@ -593,6 +600,34 @@ static void xennet_mark_tx_pending(struc
 		queue->tx_link[i] = TX_PENDING;
 }
 
+struct sk_buff *bounce_skb(const struct sk_buff *skb)
+{
+	unsigned int headerlen = skb_headroom(skb);
+	/* Align size to allocate full pages and avoid contiguous data leaks */
+	unsigned int size = ALIGN(skb_end_offset(skb) + skb->data_len,
+				  XEN_PAGE_SIZE);
+	struct sk_buff *n = alloc_skb(size, GFP_ATOMIC | __GFP_ZERO);
+
+	if (!n)
+		return NULL;
+
+	if (!IS_ALIGNED((uintptr_t)n->head, XEN_PAGE_SIZE)) {
+		WARN_ONCE(1, "misaligned skb allocated\n");
+		kfree_skb(n);
+		return NULL;
+	}
+
+	/* Set the data pointer */
+	skb_reserve(n, headerlen);
+	/* Set the tail pointer and length */
+	skb_put(n, skb->len);
+
+	BUG_ON(skb_copy_bits(skb, -headerlen, n->head, headerlen + skb->len));
+
+	skb_copy_header(n, skb);
+	return n;
+}
+
 #define MAX_XEN_SKB_FRAGS (65536 / XEN_PAGE_SIZE + 1)
 
 static netdev_tx_t xennet_start_xmit(struct sk_buff *skb, struct net_device *dev)

@@ -645,9 +680,13 @@ static netdev_tx_t xennet_start_xmit(str
 
 	/* The first req should be at least ETH_HLEN size or the packet will be
 	 * dropped by netback.
+	 *
+	 * If the backend is not trusted bounce all data to zeroed pages to
+	 * avoid exposing contiguous data on the granted page not belonging to
+	 * the skb.
 	 */
-	if (unlikely(PAGE_SIZE - offset < ETH_HLEN)) {
-		nskb = skb_copy(skb, GFP_ATOMIC);
+	if (np->bounce || unlikely(PAGE_SIZE - offset < ETH_HLEN)) {
+		nskb = bounce_skb(skb);
 		if (!nskb)
 			goto drop;
 		dev_consume_skb_any(skb);

@@ -1953,6 +1992,10 @@ static int talk_to_netback(struct xenbus
 
 	info->netdev->irq = 0;
 
+	/* Check if backend is trusted. */
+	info->bounce = !xennet_trusted ||
+		       !xenbus_read_unsigned(dev->nodename, "trusted", 1);
+
 	/* Check if backend supports multiple queues */
 	max_queues = xenbus_read_unsigned(info->xbdev->otherend,
 					  "multi-queue-max-queues", 1);

@@ -2106,6 +2149,9 @@ static int xennet_connect(struct net_dev
 	err = talk_to_netback(np->xbdev, np);
 	if (err)
 		return err;
+	if (np->bounce)
+		dev_info(&np->xbdev->dev,
+			 "bouncing transmitted data to zeroed pages\n");
 
 	/* talk_to_netback() sets the correct number of queues */
 	num_queues = dev->real_num_tx_queues;

From: Roger Pau Monne <[email protected]>

commit 2400617da7eebf9167d71a46122828bc479d64c9 upstream.

Split the current bounce buffering logic used with persistent grants into it's own option, and allow enabling it independently of persistent grants. This allows to reuse the same code paths to perform the bounce buffering required to avoid leaking contiguous data in shared pages not part of the request fragments.

Reporting whether the backend is to be trusted can be done using a module parameter, or from the xenstore frontend path as set by the toolstack when adding the device.

This is CVE-2022-33742, part of XSA-403.

---
 drivers/block/xen-blkfront.c |   49 +++++++++++++++++++++++++++++--------------
 1 file changed, 34 insertions(+), 15 deletions(-)

@@ -151,6 +151,10 @@ static unsigned int xen_blkif_max_ring_o
 module_param_named(max_ring_page_order, xen_blkif_max_ring_order, int, 0444);
 MODULE_PARM_DESC(max_ring_page_order, "Maximum order of pages to be used for the shared ring");
 
+static bool __read_mostly xen_blkif_trusted = true;
+module_param_named(trusted, xen_blkif_trusted, bool, 0644);
+MODULE_PARM_DESC(trusted, "Is the backend trusted");
+
 #define BLK_RING_SIZE(info)	\
 	__CONST_RING_SIZE(blkif, XEN_PAGE_SIZE * (info)->nr_ring_pages)
 

@@ -211,6 +215,7 @@ struct blkfront_info
 	unsigned int feature_discard:1;
 	unsigned int feature_secdiscard:1;
 	unsigned int feature_persistent:1;
+	unsigned int bounce:1;
 	unsigned int discard_granularity;
 	unsigned int discard_alignment;
 	/* Number of 4KB segments handled */

@@ -300,7 +305,7 @@ static int fill_grant_buffer(struct blkf
 		if (!gnt_list_entry)
 			goto out_of_memory;
 
-		if (info->feature_persistent) {
+		if (info->bounce) {
 			granted_page = alloc_page(GFP_NOIO | __GFP_ZERO);
 			if (!granted_page) {
 				kfree(gnt_list_entry);

@@ -320,7 +325,7 @@ out_of_memory:
 	list_for_each_entry_safe(gnt_list_entry, n,
 	                         &rinfo->grants, node) {
 		list_del(&gnt_list_entry->node);
-		if (info->feature_persistent)
+		if (info->bounce)
 			__free_page(gnt_list_entry->page);
 		kfree(gnt_list_entry);
 		i--;

@@ -366,7 +371,7 @@ static struct grant *get_grant(grant_ref
 	/* Assign a gref to this page */
 	gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head);
 	BUG_ON(gnt_list_entry->gref == -ENOSPC);
-	if (info->feature_persistent)
+	if (info->bounce)
 		grant_foreign_access(gnt_list_entry, info);
 	else {
 		/* Grant access to the GFN passed by the caller */

@@ -390,7 +395,7 @@ static struct grant *get_indirect_grant(
 	/* Assign a gref to this page */
 	gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head);
 	BUG_ON(gnt_list_entry->gref == -ENOSPC);
-	if (!info->feature_persistent) {
+	if (!info->bounce) {
 		struct page *indirect_page;
 
 		/* Fetch a pre-allocated page to use for indirect grefs */

@@ -705,7 +710,7 @@ static int blkif_queue_rw_req(struct req
 		.grant_idx = 0,
 		.segments = NULL,
 		.rinfo = rinfo,
-		.need_copy = rq_data_dir(req) && info->feature_persistent,
+		.need_copy = rq_data_dir(req) && info->bounce,
 	};
 
 	/*

@@ -1026,11 +1031,12 @@ static void xlvbd_flush(struct blkfront_
 {
 	blk_queue_write_cache(info->rq, info->feature_flush ? true : false,
 			      info->feature_fua ? true : false);
-	pr_info("blkfront: %s: %s %s %s %s %s\n",
+	pr_info("blkfront: %s: %s %s %s %s %s %s %s\n",
 		info->gd->disk_name, flush_info(info),
 		"persistent grants:", info->feature_persistent ?
 		"enabled;" : "disabled;", "indirect descriptors:",
-		info->max_indirect_segments ? "enabled;" : "disabled;");
+		info->max_indirect_segments ? "enabled;" : "disabled;",
+		"bounce buffer:", info->bounce ? "enabled" : "disabled;");
 }
 
 static int xen_translate_vdev(int vdevice, int *minor, unsigned int *offset)

@@ -1265,7 +1271,7 @@ static void blkif_free_ring(struct blkfr
 	if (!list_empty(&rinfo->indirect_pages)) {
 		struct page *indirect_page, *n;
 
-		BUG_ON(info->feature_persistent);
+		BUG_ON(info->bounce);
 		list_for_each_entry_safe(indirect_page, n, &rinfo->indirect_pages, lru) {
 			list_del(&indirect_page->lru);
 			__free_page(indirect_page);

@@ -1282,7 +1288,7 @@ static void blkif_free_ring(struct blkfr
 							  0, 0UL);
 				rinfo->persistent_gnts_c--;
 			}
-			if (info->feature_persistent)
+			if (info->bounce)
 				__free_page(persistent_gnt->page);
 			kfree(persistent_gnt);
 		}

@@ -1303,7 +1309,7 @@ static void blkif_free_ring(struct blkfr
 		for (j = 0; j < segs; j++) {
 			persistent_gnt = rinfo->shadow[i].grants_used[j];
 			gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL);
-			if (info->feature_persistent)
+			if (info->bounce)
 				__free_page(persistent_gnt->page);
 			kfree(persistent_gnt);
 		}

@@ -1493,7 +1499,7 @@ static int blkif_completion(unsigned lon
 	data.s = s;
 	num_sg = s->num_sg;
 
-	if (bret->operation == BLKIF_OP_READ && info->feature_persistent) {
+	if (bret->operation == BLKIF_OP_READ && info->bounce) {
 		for_each_sg(s->sg, sg, num_sg, i) {
 			BUG_ON(sg->offset + sg->length > PAGE_SIZE);
 

@@ -1552,7 +1558,7 @@ static int blkif_completion(unsigned lon
 				 * Add the used indirect page back to the list of
 				 * available pages for indirect grefs.
 				 */
-				if (!info->feature_persistent) {
+				if (!info->bounce) {
 					indirect_page = s->indirect_grants[i]->page;
 					list_add(&indirect_page->lru, &rinfo->indirect_pages);
 				}

@@ -1847,6 +1853,10 @@ static int talk_to_blkback(struct xenbus
 	if (!info)
 		return -ENODEV;
 
+	/* Check if backend is trusted. */
+	info->bounce = !xen_blkif_trusted ||
+		       !xenbus_read_unsigned(dev->nodename, "trusted", 1);
+
 	max_page_order = xenbus_read_unsigned(info->xbdev->otherend,
 					      "max-ring-page-order", 0);
 	ring_page_order = min(xen_blkif_max_ring_order, max_page_order);

@@ -2273,10 +2283,10 @@ static int blkfront_setup_indirect(struc
 	if (err)
 		goto out_of_memory;
 
-	if (!info->feature_persistent && info->max_indirect_segments) {
+	if (!info->bounce && info->max_indirect_segments) {
 		/*
-		 * We are using indirect descriptors but not persistent
-		 * grants, we need to allocate a set of pages that can be
+		 * We are using indirect descriptors but don't have a bounce
+		 * buffer, we need to allocate a set of pages that can be
 		 * used for mapping indirect grefs
 		 */
 		int num = INDIRECT_GREFS(grants) * BLK_RING_SIZE(info);

@@ -2376,6 +2386,8 @@ static void blkfront_gather_backend_feat
 	info->feature_persistent =
 		!!xenbus_read_unsigned(info->xbdev->otherend,
 				       "feature-persistent", 0);
+	if (info->feature_persistent)
+		info->bounce = true;
 
 	indirect_segments = xenbus_read_unsigned(info->xbdev->otherend,
 					"feature-max-indirect-segments", 0);

@@ -2751,6 +2763,13 @@ static void blkfront_delay_work(struct w
 	struct blkfront_info *info;
 	bool need_schedule_work = false;
 
+	/*
+	 * Note that when using bounce buffers but not persistent grants
+	 * there's no need to run blkfront_delay_work because grants are
+	 * revoked in blkif_completion or else an error is reported and the
+	 * connection is closed.
+	 */
+
 	mutex_lock(&blkfront_mutex);
 
 	list_for_each_entry(info, &info_list, info_list) {

From: Oleksandr Tyshchenko <[email protected]>

commit b75cd218274e01d026dc5240e86fdeb44bbed0c8 upstream.

During the PV driver life cycle the mappings are added to the RB-tree by set_foreign_p2m_mapping(), which is called from gnttab_map_refs() and are removed by clear_foreign_p2m_mapping() which is called from gnttab_unmap_refs(). As both functions end up calling __set_phys_to_machine_multi() which updates the RB-tree, this function can be called concurrently.

There is already a "p2m_lock" to protect against concurrent accesses, but the problem is that the first read of "phys_to_mach.rb_node" in __set_phys_to_machine_multi() is not covered by it, so this might lead to the incorrect mappings update (removing in our case) in RB-tree.

In my environment the related issue happens rarely and only when PV net backend is running, the xen_add_phys_to_mach_entry() claims that it cannot add new pfn <-> mfn mapping to the tree since it is already exists which results in a failure when mapping foreign pages.

But there might be other bad consequences related to the non-protected root reads such use-after-free, etc.

While at it, also fix the similar usage in __pfn_to_mfn(), so initialize "struct rb_node *n" with the "p2m_lock" held in both functions to avoid possible bad consequences.

This is CVE-2022-33744 / XSA-406.

---
 arch/arm/xen/p2m.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

@@ -61,11 +61,12 @@ out:
 
 unsigned long __pfn_to_mfn(unsigned long pfn)
 {
-	struct rb_node *n = phys_to_mach.rb_node;
+	struct rb_node *n;
 	struct xen_p2m_entry *entry;
 	unsigned long irqflags;
 
 	read_lock_irqsave(&p2m_lock, irqflags);
+	n = phys_to_mach.rb_node;
 	while (n) {
 		entry = rb_entry(n, struct xen_p2m_entry, rbnode_phys);
 		if (entry->pfn <= pfn &&

@@ -151,10 +152,11 @@ bool __set_phys_to_machine_multi(unsigne
 	int rc;
 	unsigned long irqflags;
 	struct xen_p2m_entry *p2m_entry;
-	struct rb_node *n = phys_to_mach.rb_node;
+	struct rb_node *n;
 
 	if (mfn == INVALID_P2M_ENTRY) {
 		write_lock_irqsave(&p2m_lock, irqflags);
+		n = phys_to_mach.rb_node;
 		while (n) {
 			p2m_entry = rb_entry(n, struct xen_p2m_entry, rbnode_phys);
 			if (p2m_entry->pfn <= pfn &&

From: Carlo Lobrano <[email protected]>

commit 8d17a33b076d24aa4861f336a125c888fb918605 upstream.

This patch adds support for Telit LN920 0x1060 composition

0x1060: tty, adb, rmnet, tty, tty, tty, tty

---
 drivers/net/usb/qmi_wwan.c |    1 +
 1 file changed, 1 insertion(+)

@@ -1318,6 +1318,7 @@ static const struct usb_device_id produc
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1031, 3)}, /* Telit LE910C1-EUX */
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1040, 2)},	/* Telit LE922A */
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1050, 2)},	/* Telit FN980 */
+	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1060, 2)},	/* Telit LN920 */
 	{QMI_FIXED_INTF(0x1bc7, 0x1100, 3)},	/* Telit ME910 */
 	{QMI_FIXED_INTF(0x1bc7, 0x1101, 3)},	/* Telit ME910 dual modem */
 	{QMI_FIXED_INTF(0x1bc7, 0x1200, 5)},	/* Telit LE920 */

From: Daniele Palmas <[email protected]>

commit 94f2a444f28a649926c410eb9a38afb13a83ebe0 upstream.

Add the following Telit FN990 composition:

0x1070: tty, adb, rmnet, tty, tty, tty, tty

---
 drivers/net/usb/qmi_wwan.c |    1 +
 1 file changed, 1 insertion(+)

@@ -1319,6 +1319,7 @@ static const struct usb_device_id produc
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1040, 2)},	/* Telit LE922A */
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1050, 2)},	/* Telit FN980 */
 	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1060, 2)},	/* Telit LN920 */
+	{QMI_QUIRK_SET_DTR(0x1bc7, 0x1070, 2)},	/* Telit FN990 */
 	{QMI_FIXED_INTF(0x1bc7, 0x1100, 3)},	/* Telit ME910 */
 	{QMI_FIXED_INTF(0x1bc7, 0x1101, 3)},	/* Telit ME910 dual modem */
 	{QMI_FIXED_INTF(0x1bc7, 0x1200, 5)},	/* Telit LE920 */

All tests passing for Tegra ...

Test results for stable-v4.19:

    10 builds:	10 pass, 0 fail
    22 boots:	22 pass, 0 fail
    40 tests:	40 pass, 0 fail

Linux version: 4.19.251-rc1-gb9f174a70c6f
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,

                tegra194-p2972-0000, tegra20-ventana,
                tegra210-p2371-2180, tegra30-cardhu-a04

Jon

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

## Build
* kernel: 4.19.251-rc1
* git: https://gitlab.com/Linaro/lkft/mirrors/stable/linux-stable-rc
* git branch: linux-4.19.y
* git commit: b9f174a70c6f609b6132c3a65883b16d552984b8
* git describe: v4.19.250-34-gb9f174a70c6f
* test details:
https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-4.19.y/build/v4.19.250-34-gb9f174a70c6f

## Test Regressions (compared to v4.19.250)
No test regressions found.

## Metric Regressions (compared to v4.19.250)
No metric regressions found.

## Test Fixes (compared to v4.19.250)
No test fixes found.

## Metric Fixes (compared to v4.19.250)
No metric fixes found.

## Test result summary
total: 110465, pass: 97549, fail: 274, skip: 11512, xfail: 1130

## Build Summary
* arc: 10 total, 10 passed, 0 failed
* arm: 291 total, 286 passed, 5 failed
* arm64: 58 total, 57 passed, 1 failed
* i386: 26 total, 23 passed, 3 failed
* mips: 38 total, 38 passed, 0 failed
* parisc: 12 total, 12 passed, 0 failed
* powerpc: 55 total, 54 passed, 1 failed
* s390: 12 total, 12 passed, 0 failed
* sh: 24 total, 24 passed, 0 failed
* sparc: 12 total, 12 passed, 0 failed
* x86_64: 52 total, 51 passed, 1 failed

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

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

Hi Greg,

Build test (gcc version 11.3.1 20220627):
mips: 63 configs -> no failure
arm: 115 configs -> no failure
arm64: 2 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]

[1]. https://openqa.qa.codethink.co.uk/tests/1455

--
Regards
Sudip

Build results:

	total: 157 pass: 157 fail: 0
Qemu test results:
	total: 425 pass: 425 fail: 0

Guenter

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

thanks,
-- Shuah

Tested on arm64 and x86 for 4.19.251-rc1,

Kernel repo:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
Branch: linux-4.19.y
Version: 4.19.251-rc1
Commit: b9f174a70c6f609b6132c3a65883b16d552984b8
Compiler: gcc version 7.3.0 (GCC)

arm64:
--------------------------------------------------------------------
Testcase Result Summary:
total: 8946
passed: 8946
failed: 0
timeout: 0
--------------------------------------------------------------------

x86:
--------------------------------------------------------------------
Testcase Result Summary:
total: 8946
passed: 8946
failed: 0
timeout: 0
--------------------------------------------------------------------