5.15.175-rc1 review

This is the start of the stable review cycle for the 5.15.175 release. There are 51 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, 19 Dec 2024 17:05:03 +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-5.15.175-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-5.15.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 5.15.175-rc1

Dan Carpenter <[email protected]>
ALSA: usb-audio: Fix a DMA to stack memory bug

Juergen Gross <[email protected]>
x86/xen: remove hypercall page

Juergen Gross <[email protected]>
x86/xen: use new hypercall functions instead of hypercall page

Juergen Gross <[email protected]>
x86/xen: add central hypercall functions

Juergen Gross <[email protected]>
x86/xen: don't do PV iret hypercall through hypercall page

Juergen Gross <[email protected]>
x86/static-call: provide a way to do very early static-call updates

Juergen Gross <[email protected]>
objtool/x86: allow syscall instruction

Juergen Gross <[email protected]>
x86: make get_cpu_vendor() accessible from Xen code

Juergen Gross <[email protected]>
xen/netfront: fix crash when removing device

Greg Kroah-Hartman <[email protected]>
Revert "parisc: fix a possible DMA corruption"

Nikolay Kuratov <[email protected]>
tracing/kprobes: Skip symbol counting logic for module symbols in create_local_trace_kprobe()

Eduard Zingerman <[email protected]>
bpf: sync_linked_regs() must preserve subreg_def

Nathan Chancellor <[email protected]>
blk-iocost: Avoid using clamp() on inuse in __propagate_weights()

Daniil Tatianin <[email protected]>
ACPICA: events/evxfregn: don't release the ContextMutex that was never acquired

Daniel Borkmann <[email protected]>
team: Fix feature propagation of NETIF_F_GSO_ENCAP_ALL

Daniel Borkmann <[email protected]>
bonding: Fix feature propagation of NETIF_F_GSO_ENCAP_ALL

Martin Ottens <[email protected]>
net/sched: netem: account for backlog updates from child qdisc

Paul Barker <[email protected]>
Documentation: PM: Clarify pm_runtime_resume_and_get() return value

Stefan Wahren <[email protected]>
qca_spi: Make driver probing reliable

Stefan Wahren <[email protected]>
qca_spi: Fix clock speed for multiple QCA7000

Anumula Murali Mohan Reddy <[email protected]>
cxgb4: use port number to set mac addr

Ilpo Järvinen <[email protected]>
ACPI: resource: Fix memory resource type union access

Daniel Machon <[email protected]>
net: sparx5: fix the maximum frame length register

Daniel Machon <[email protected]>
net: sparx5: fix FDMA performance issue

Eric Dumazet <[email protected]>
net: lapb: increase LAPB_HEADER_LEN

Thomas Weißschuh <[email protected]>
ptp: kvm: x86: Return EOPNOTSUPP instead of ENODEV from kvm_arch_ptp_init()

Jeremi Piotrowski <[email protected]>
ptp: kvm: Use decrypted memory in confidential guest on x86

Danielle Ratson <[email protected]>
selftests: mlxsw: sharedbuffer: Remove duplicate test cases

Danielle Ratson <[email protected]>
selftests: mlxsw: sharedbuffer: Remove h1 ingress test case

Eric Dumazet <[email protected]>
tipc: fix NULL deref in cleanup_bearer()

Remi Pommarel <[email protected]>
batman-adv: Do not let TT changes list grows indefinitely

Remi Pommarel <[email protected]>
batman-adv: Remove uninitialized data in full table TT response

Remi Pommarel <[email protected]>
batman-adv: Do not send uninitialized TT changes

Suraj Sonawane <[email protected]>
acpi: nfit: vmalloc-out-of-bounds Read in acpi_nfit_ctl

Sungjong Seo <[email protected]>
exfat: fix potential deadlock on __exfat_get_dentry_set

Michal Luczaj <[email protected]>
virtio/vsock: Fix accept_queue memory leak

Michal Luczaj <[email protected]>
bpf, sockmap: Fix update element with same

Darrick J. Wong <[email protected]>
xfs: fix scrub tracepoints when inode-rooted btrees are involved

Darrick J. Wong <[email protected]>
xfs: return from xfs_symlink_verify early on V4 filesystems

Darrick J. Wong <[email protected]>
xfs: don't drop errno values when we fail to ficlone the entire range

Darrick J. Wong <[email protected]>
xfs: update btree keys correctly when _insrec splits an inode root block

Jiasheng Jiang <[email protected]>
drm/i915: Fix memory leak by correcting cache object name in error handler

Lianqin Hu <[email protected]>
usb: gadget: u_serial: Fix the issue that gs_start_io crashed due to accessing null pointer

Vitalii Mordan <[email protected]>
usb: ehci-hcd: fix call balance of clocks handling routines

Stefan Wahren <[email protected]>
usb: dwc2: Fix HCD port connection race

Stefan Wahren <[email protected]>
usb: dwc2: hcd: Fix GetPortStatus & SetPortFeature

Stefan Wahren <[email protected]>
usb: dwc2: Fix HCD resume

Joe Hattori <[email protected]>
ata: sata_highbank: fix OF node reference leak in highbank_initialize_phys()

Mark Tomlinson <[email protected]>
usb: host: max3421-hcd: Correctly abort a USB request.

Jaakko Salo <[email protected]>
ALSA: usb-audio: Add implicit feedback quirk for Yamaha THR5

MoYuanhao <[email protected]>
tcp: check space before adding MPTCP SYN options

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

Diffstat:

 Documentation/power/runtime_pm.rst                 |  4 +-
 Makefile                                           |  4 +-
 arch/parisc/Kconfig                                |  1 -
 arch/parisc/include/asm/cache.h                    | 11 +--
 arch/x86/include/asm/processor.h                   |  2 +
 arch/x86/include/asm/static_call.h                 | 15 ++++
 arch/x86/include/asm/sync_core.h                   |  6 +-
 arch/x86/include/asm/xen/hypercall.h               | 36 ++++----
 arch/x86/kernel/cpu/common.c                       | 38 +++++----
 arch/x86/kernel/static_call.c                      | 10 +++
 arch/x86/xen/enlighten.c                           | 65 ++++++++++++++-
 arch/x86/xen/enlighten_hvm.c                       | 13 ++-
 arch/x86/xen/enlighten_pv.c                        |  4 +-
 arch/x86/xen/enlighten_pvh.c                       |  7 --
 arch/x86/xen/xen-asm.S                             | 49 +++++++++--
 arch/x86/xen/xen-head.S                            | 97 ++++++++++++++++++----
 arch/x86/xen/xen-ops.h                             |  9 ++
 block/blk-iocost.c                                 |  9 +-
 drivers/acpi/acpica/evxfregn.c                     |  2 -
 drivers/acpi/nfit/core.c                           |  7 +-
 drivers/acpi/resource.c                            |  6 +-
 drivers/ata/sata_highbank.c                        |  1 +
 drivers/gpu/drm/i915/i915_scheduler.c              |  2 +-
 drivers/net/bonding/bond_main.c                    |  1 +
 drivers/net/ethernet/chelsio/cxgb4/cxgb4.h         |  2 +-
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c    |  2 +-
 drivers/net/ethernet/chelsio/cxgb4/t4_hw.c         |  5 +-
 .../net/ethernet/microchip/sparx5/sparx5_main.c    | 11 ++-
 .../net/ethernet/microchip/sparx5/sparx5_port.c    |  2 +-
 drivers/net/ethernet/qualcomm/qca_spi.c            | 26 +++---
 drivers/net/ethernet/qualcomm/qca_spi.h            |  1 -
 drivers/net/team/team.c                            |  3 +-
 drivers/net/xen-netfront.c                         |  5 +-
 drivers/ptp/ptp_kvm_arm.c                          |  4 +
 drivers/ptp/ptp_kvm_common.c                       |  1 +
 drivers/ptp/ptp_kvm_x86.c                          | 61 +++++++++++---
 drivers/usb/dwc2/hcd.c                             | 19 ++---
 drivers/usb/gadget/function/u_serial.c             |  9 +-
 drivers/usb/host/ehci-sh.c                         |  9 +-
 drivers/usb/host/max3421-hcd.c                     | 16 ++--
 fs/exfat/dir.c                                     |  2 +-
 fs/xfs/libxfs/xfs_btree.c                          | 29 +++++--
 fs/xfs/libxfs/xfs_symlink_remote.c                 |  4 +-
 fs/xfs/scrub/trace.h                               |  2 +-
 fs/xfs/xfs_file.c                                  |  8 ++
 include/linux/compiler.h                           | 34 +++++---
 include/linux/ptp_kvm.h                            |  1 +
 include/linux/static_call.h                        |  1 +
 include/net/lapb.h                                 |  2 +-
 kernel/bpf/verifier.c                              |  5 +-
 kernel/static_call_inline.c                        |  2 +-
 kernel/trace/trace_kprobe.c                        |  2 +-
 net/batman-adv/translation-table.c                 | 58 +++++++++----
 net/core/sock_map.c                                |  1 +
 net/ipv4/tcp_output.c                              |  6 +-
 net/sched/sch_netem.c                              | 22 +++--
 net/tipc/udp_media.c                               |  7 +-
 net/vmw_vsock/virtio_transport_common.c            |  8 ++
 sound/usb/quirks.c                                 | 33 +++++---
 tools/objtool/check.c                              | 11 ++-
 .../selftests/drivers/net/mlxsw/sharedbuffer.sh    | 15 ----
 61 files changed, 589 insertions(+), 239 deletions(-)

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: MoYuanhao <[email protected]>

commit 06d64ab46f19ac12f59a1d2aa8cd196b2e4edb5b upstream.

Ensure there is enough space before adding MPTCP options in tcp_syn_options().

Without this check, 'remaining' could underflow, and causes issues. If there is not enough space, MPTCP should not be used.

---
 net/ipv4/tcp_output.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

@@ -824,8 +824,10 @@ static unsigned int tcp_syn_options(stru
 		unsigned int size;
 
 		if (mptcp_syn_options(sk, skb, &size, &opts->mptcp)) {
-			opts->options |= OPTION_MPTCP;
-			remaining -= size;
+			if (remaining >= size) {
+				opts->options |= OPTION_MPTCP;
+				remaining -= size;
+			}
 		}
 	}
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Jaakko Salo <[email protected]>

commit 82fdcf9b518b205da040046fbe7747fb3fd18657 upstream.

Use implicit feedback from the capture endpoint to fix popping sounds during playback.

---
 sound/usb/quirks.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1768,6 +1768,8 @@ static const struct usb_audio_quirk_flag
 		   QUIRK_FLAG_CTL_MSG_DELAY_1M | QUIRK_FLAG_IGNORE_CTL_ERROR),
 	DEVICE_FLG(0x046d, 0x09a4, /* Logitech QuickCam E 3500 */
 		   QUIRK_FLAG_CTL_MSG_DELAY_1M | QUIRK_FLAG_IGNORE_CTL_ERROR),
+	DEVICE_FLG(0x0499, 0x1506, /* Yamaha THR5 */
+		   QUIRK_FLAG_GENERIC_IMPLICIT_FB),
 	DEVICE_FLG(0x0499, 0x1509, /* Steinberg UR22 */
 		   QUIRK_FLAG_GENERIC_IMPLICIT_FB),
 	DEVICE_FLG(0x0499, 0x3108, /* Yamaha YIT-W12TX */

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Mark Tomlinson <[email protected]>

commit 0d2ada05227881f3d0722ca2364e3f7a860a301f upstream.

If the current USB request was aborted, the spi thread would not respond to any further requests. This is because the "curr_urb" pointer would not become NULL, so no further requests would be taken off the queue. The solution here is to set the "urb_done" flag, as this will cause the correct handling of the URB. Also clear interrupts that should only be expected if an URB is in progress.

---
 drivers/usb/host/max3421-hcd.c |   16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)

@@ -785,11 +785,17 @@ max3421_check_unlink(struct usb_hcd *hcd
 				retval = 1;
 				dev_dbg(&spi->dev, "%s: URB %p unlinked=%d",
 					__func__, urb, urb->unlinked);
-				usb_hcd_unlink_urb_from_ep(hcd, urb);
-				spin_unlock_irqrestore(&max3421_hcd->lock,
-						       flags);
-				usb_hcd_giveback_urb(hcd, urb, 0);
-				spin_lock_irqsave(&max3421_hcd->lock, flags);
+				if (urb == max3421_hcd->curr_urb) {
+					max3421_hcd->urb_done = 1;
+					max3421_hcd->hien &= ~(BIT(MAX3421_HI_HXFRDN_BIT) |
+							       BIT(MAX3421_HI_RCVDAV_BIT));
+				} else {
+					usb_hcd_unlink_urb_from_ep(hcd, urb);
+					spin_unlock_irqrestore(&max3421_hcd->lock,
+							       flags);
+					usb_hcd_giveback_urb(hcd, urb, 0);
+					spin_lock_irqsave(&max3421_hcd->lock, flags);
+				}
 			}
 		}
 	}

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Joe Hattori <[email protected]>

commit 676fe1f6f74db988191dab5df3bf256908177072 upstream.

The OF node reference obtained by of_parse_phandle_with_args() is not released on early return. Add a of_node_put() call before returning.

---
 drivers/ata/sata_highbank.c |    1 +
 1 file changed, 1 insertion(+)

@@ -348,6 +348,7 @@ static int highbank_initialize_phys(stru
 			phy_nodes[phy] = phy_data.np;
 			cphy_base[phy] = of_iomap(phy_nodes[phy], 0);
 			if (cphy_base[phy] == NULL) {
+				of_node_put(phy_data.np);
 				return 0;
 			}
 			phy_count += 1;

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

commit 336f72d3cbf5cc17df2947bbbd2ba6e2509f17e8 upstream.

The Raspberry Pi can suffer on interrupt storms on HCD resume. The dwc2 driver sometimes misses to enable HCD_FLAG_HW_ACCESSIBLE before re-enabling the interrupts. This causes a situation where both handler ignore a incoming port interrupt and force the upper layers to disable the dwc2 interrupt line. This leaves the USB interface in a unusable state:

irq 66: nobody cared (try booting with the "irqpoll" option)

CPU: 0 PID: 0 Comm: swapper/0 Tainted: G W          6.10.0-rc3
Hardware name: BCM2835
Call trace:
unwind_backtrace from show_stack+0x10/0x14
show_stack from dump_stack_lvl+0x50/0x64
dump_stack_lvl from __report_bad_irq+0x38/0xc0
__report_bad_irq from note_interrupt+0x2ac/0x2f4
note_interrupt from handle_irq_event+0x88/0x8c
handle_irq_event from handle_level_irq+0xb4/0x1ac
handle_level_irq from generic_handle_domain_irq+0x24/0x34
generic_handle_domain_irq from bcm2836_chained_handle_irq+0x24/0x28
bcm2836_chained_handle_irq from generic_handle_domain_irq+0x24/0x34
generic_handle_domain_irq from generic_handle_arch_irq+0x34/0x44
generic_handle_arch_irq from __irq_svc+0x88/0xb0
Exception stack(0xc1b01f20 to 0xc1b01f68)

1f20: 0005c0d4 00000001 00000000 00000000 c1b09780 c1d6b32c c1b04e54 c1a5eae8 1f40: c1b04e90 00000000 00000000 00000000 c1d6a8a0 c1b01f70 c11d2da8 c11d4160 1f60: 60000013 ffffffff
__irq_svc from default_idle_call+0x1c/0xb0
default_idle_call from do_idle+0x21c/0x284
do_idle from cpu_startup_entry+0x28/0x2c
cpu_startup_entry from kernel_init+0x0/0x12c
handlers:
[<f539e0f4>] dwc2_handle_common_intr
[<75cd278b>] usb_hcd_irq
Disabling IRQ #66

So enable the HCD_FLAG_HW_ACCESSIBLE flag in case there is a port connection.

---
 drivers/usb/dwc2/hcd.c |    1 +
 1 file changed, 1 insertion(+)

@@ -4464,6 +4464,7 @@ static int _dwc2_hcd_resume(struct usb_h
 	 * Power Down mode.
 	 */
 	if (hprt0 & HPRT0_CONNSTS) {
+		set_bit(HCD_FLAG_HW_ACCESSIBLE, &hcd->flags);
 		hsotg->lx_state = DWC2_L0;
 		goto unlock;
 	}

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

commit a8d3e4a734599c7d0f6735f8db8a812e503395dd upstream.

On Rasperry Pis without onboard USB hub the power cycle during power connect init only disable the port but never enabled it again:

usb usb1-port1: attempt power cycle

The port relevant part in dwc2_hcd_hub_control() is skipped in case port_connect_status = 0 under the assumption the core is or will be soon in device mode. But this assumption is wrong, because after ClearPortFeature USB_PORT_FEAT_POWER the port_connect_status will also be 0 and SetPortFeature (incl. USB_PORT_FEAT_POWER) will be a no-op.

Fix the behavior of dwc2_hcd_hub_control() by replacing the port_connect_status check with dwc2_is_device_mode().

---
 drivers/usb/dwc2/hcd.c |   16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

@@ -3579,11 +3579,9 @@ static int dwc2_hcd_hub_control(struct d
 			port_status |= USB_PORT_STAT_C_OVERCURRENT << 16;
 		}
 
-		if (!hsotg->flags.b.port_connect_status) {
+		if (dwc2_is_device_mode(hsotg)) {
 			/*
-			 * The port is disconnected, which means the core is
-			 * either in device mode or it soon will be. Just
-			 * return 0's for the remainder of the port status
+			 * Just return 0's for the remainder of the port status
 			 * since the port register can't be read if the core
 			 * is in device mode.
 			 */

@@ -3653,13 +3651,11 @@ static int dwc2_hcd_hub_control(struct d
 		if (wvalue != USB_PORT_FEAT_TEST && (!windex || windex > 1))
 			goto error;
 
-		if (!hsotg->flags.b.port_connect_status) {
+		if (dwc2_is_device_mode(hsotg)) {
 			/*
-			 * The port is disconnected, which means the core is
-			 * either in device mode or it soon will be. Just
-			 * return without doing anything since the port
-			 * register can't be written if the core is in device
-			 * mode.
+			 * Just return 0's for the remainder of the port status
+			 * since the port register can't be read if the core
+			 * is in device mode.
 			 */
 			break;
 		}

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

commit 1cf1bd88f129f3bd647fead4dca270a5894274bb upstream.

On Raspberry Pis without onboard USB hub frequent device reconnects can trigger a interrupt storm after DWC2 entered host clock gating. This is caused by a race between _dwc2_hcd_suspend() and the port interrupt, which sets port_connect_status. The issue occurs if port_connect_status is still 1, but there is no connection anymore:

usb 1-1: USB disconnect, device number 25
dwc2 3f980000.usb: _dwc2_hcd_suspend: port_connect_status: 1 dwc2 3f980000.usb: Entering host clock gating. Disabling IRQ #66
irq 66: nobody cared (try booting with the "irqpoll" option)

CPU: 0 UID: 0 PID: 0 Comm: swapper/0 Not tainted 6.12.0-gc1bb81b13202-dirty #322
Hardware name: BCM2835
Call trace:
 unwind_backtrace from show_stack+0x10/0x14
 show_stack from dump_stack_lvl+0x50/0x64
 dump_stack_lvl from __report_bad_irq+0x38/0xc0
 __report_bad_irq from note_interrupt+0x2ac/0x2f4
 note_interrupt from handle_irq_event+0x88/0x8c
 handle_irq_event from handle_level_irq+0xb4/0x1ac
 handle_level_irq from generic_handle_domain_irq+0x24/0x34
 generic_handle_domain_irq from bcm2836_chained_handle_irq+0x24/0x28
 bcm2836_chained_handle_irq from generic_handle_domain_irq+0x24/0x34
 generic_handle_domain_irq from generic_handle_arch_irq+0x34/0x44
 generic_handle_arch_irq from __irq_svc+0x88/0xb0
 Exception stack(0xc1d01f20 to 0xc1d01f68)
 1f20: 0004ef3c 00000001 00000000 00000000 c1d09780 c1f6bb5c c1d04e54 c1c60ca8
 1f40: c1d04e94 00000000 00000000 c1d092a8 c1f6af20 c1d01f70 c1211b98 c1212f40
 1f60: 60000013 ffffffff
 __irq_svc from default_idle_call+0x1c/0xb0
 default_idle_call from do_idle+0x21c/0x284
 do_idle from cpu_startup_entry+0x28/0x2c
 cpu_startup_entry from kernel_init+0x0/0x12c

handlers:
[<e3a25c00>] dwc2_handle_common_intr
[<58bf98a3>] usb_hcd_irq
Disabling IRQ #66

So avoid this by reading the connection status directly.

---
 drivers/usb/dwc2/hcd.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -4378,7 +4378,7 @@ static int _dwc2_hcd_suspend(struct usb_
 	if (hsotg->bus_suspended)
 		goto skip_power_saving;
 
-	if (hsotg->flags.b.port_connect_status == 0)
+	if (!(dwc2_read_hprt0(hsotg) & HPRT0_CONNSTS))
 		goto skip_power_saving;
 
 	switch (hsotg->params.power_down) {

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Vitalii Mordan <[email protected]>

commit 97264eaaba0122a5b7e8ddd7bf4ff3ac57c2b170 upstream.

If the clocks priv->iclk and priv->fclk were not enabled in ehci_hcd_sh_probe, they should not be disabled in any path.

Conversely, if they was enabled in ehci_hcd_sh_probe, they must be disabled in all error paths to ensure proper cleanup.

Found by Linux Verification Center (linuxtesting.org) with Klever.

---
 drivers/usb/host/ehci-sh.c |    9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

@@ -120,8 +120,12 @@ static int ehci_hcd_sh_probe(struct plat
 	if (IS_ERR(priv->iclk))
 		priv->iclk = NULL;
 
-	clk_enable(priv->fclk);
-	clk_enable(priv->iclk);
+	ret = clk_enable(priv->fclk);
+	if (ret)
+		goto fail_request_resource;
+	ret = clk_enable(priv->iclk);
+	if (ret)
+		goto fail_iclk;
 
 	ret = usb_add_hcd(hcd, irq, IRQF_SHARED);
 	if (ret != 0) {

@@ -137,6 +141,7 @@ static int ehci_hcd_sh_probe(struct plat
 
 fail_add_hcd:
 	clk_disable(priv->iclk);
+fail_iclk:
 	clk_disable(priv->fclk);
 
 fail_request_resource:

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Lianqin Hu <[email protected]>

commit 4cfbca86f6a8b801f3254e0e3c8f2b1d2d64be2b upstream.

Considering that in some extreme cases,
when u_serial driver is accessed by multiple threads,
Thread A is executing the open operation and calling the gs_open, Thread B is executing the disconnect operation and calling the gserial_disconnect function,The port->port_usb pointer will be set to NULL.

E.g.

    Thread A                                 Thread B
    gs_open()                                gadget_unbind_driver()
    gs_start_io()                            composite_disconnect()
    gs_start_rx()                            gserial_disconnect()
    ...                                      ...
    spin_unlock(&port->port_lock)
    status = usb_ep_queue()                  spin_lock(&port->port_lock)
    spin_lock(&port->port_lock)              port->port_usb = NULL
    gs_free_requests(port->port_usb->in)     spin_unlock(&port->port_lock)
    Crash

This causes thread A to access a null pointer (port->port_usb is null) when calling the gs_free_requests function, causing a crash.

If port_usb is NULL, the release request will be skipped as it will be done by gserial_disconnect.

So add a null pointer check to gs_start_io before attempting to access the value of the pointer port->port_usb.

Call trace:
 gs_start_io+0x164/0x25c
 gs_open+0x108/0x13c
 tty_open+0x314/0x638
 chrdev_open+0x1b8/0x258
 do_dentry_open+0x2c4/0x700
 vfs_open+0x2c/0x3c
 path_openat+0xa64/0xc60
 do_filp_open+0xb8/0x164
 do_sys_openat2+0x84/0xf0
 __arm64_sys_openat+0x70/0x9c
 invoke_syscall+0x58/0x114
 el0_svc_common+0x80/0xe0
 do_el0_svc+0x1c/0x28
 el0_svc+0x38/0x68

---
 drivers/usb/gadget/function/u_serial.c |    9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

@@ -570,9 +570,12 @@ static int gs_start_io(struct gs_port *p
 		 * we didn't in gs_start_tx() */
 		tty_wakeup(port->port.tty);
 	} else {
-		gs_free_requests(ep, head, &port->read_allocated);
-		gs_free_requests(port->port_usb->in, &port->write_pool,
-			&port->write_allocated);
+		/* Free reqs only if we are still connected */
+		if (port->port_usb) {
+			gs_free_requests(ep, head, &port->read_allocated);
+			gs_free_requests(port->port_usb->in, &port->write_pool,
+				&port->write_allocated);
+		}
 		status = -EIO;
 	}
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Jiasheng Jiang <[email protected]>

commit 2828e5808bcd5aae7fdcd169cac1efa2701fa2dd upstream.

Replace "slab_priorities" with "slab_dependencies" in the error handler to avoid memory leak.

---
 drivers/gpu/drm/i915/i915_scheduler.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -506,6 +506,6 @@ int __init i915_scheduler_module_init(vo
 	return 0;
 
 err_priorities:
-	kmem_cache_destroy(slab_priorities);
+	kmem_cache_destroy(slab_dependencies);
 	return -ENOMEM;
 }

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Darrick J. Wong <[email protected]>

commit 6d7b4bc1c3e00b1a25b7a05141a64337b4629337 upstream.

In commit 2c813ad66a72, I partially fixed a bug wherein xfs_btree_insrec would erroneously try to update the parent's key for a block that had been split if we decided to insert the new record into the new block. The solution was to detect this situation and update the in-core key value that we pass up to the caller so that the caller will (eventually) add the new block to the parent level of the tree with the correct key.

However, I missed a subtlety about the way inode-rooted btrees work. If the full block was a maximally sized inode root block, we'll solve that fullness by moving the root block's records to a new block, resizing the root block, and updating the root to point to the new block. We don't pass a pointer to the new block to the caller because that work has already been done. The new record will /always/ land in the new block, so in this case we need to use xfs_btree_update_keys to update the keys.

This bug can theoretically manifest itself in the very rare case that we split a bmbt root block and the new record lands in the very first slot of the new block, though I've never managed to trigger it in practice. However, it is very easy to reproduce by running generic/522 with the realtime rmapbt patchset if rtinherit=1.

---
 fs/xfs/libxfs/xfs_btree.c |   29 +++++++++++++++++++++++------
 1 file changed, 23 insertions(+), 6 deletions(-)

@@ -3413,14 +3413,31 @@ xfs_btree_insrec(
 	xfs_btree_log_block(cur, bp, XFS_BB_NUMRECS);
 
 	/*
-	 * If we just inserted into a new tree block, we have to
-	 * recalculate nkey here because nkey is out of date.
+	 * Update btree keys to reflect the newly added record or keyptr.
+	 * There are three cases here to be aware of.  Normally, all we have to
+	 * do is walk towards the root, updating keys as necessary.
 	 *
-	 * Otherwise we're just updating an existing block (having shoved
-	 * some records into the new tree block), so use the regular key
-	 * update mechanism.
+	 * If the caller had us target a full block for the insertion, we dealt
+	 * with that by calling the _make_block_unfull function.  If the
+	 * "make unfull" function splits the block, it'll hand us back the key
+	 * and pointer of the new block.  We haven't yet added the new block to
+	 * the next level up, so if we decide to add the new record to the new
+	 * block (bp->b_bn != old_bn), we have to update the caller's pointer
+	 * so that the caller adds the new block with the correct key.
+	 *
+	 * However, there is a third possibility-- if the selected block is the
+	 * root block of an inode-rooted btree and cannot be expanded further,
+	 * the "make unfull" function moves the root block contents to a new
+	 * block and updates the root block to point to the new block.  In this
+	 * case, no block pointer is passed back because the block has already
+	 * been added to the btree.  In this case, we need to use the regular
+	 * key update function, just like the first case.  This is critical for
+	 * overlapping btrees, because the high key must be updated to reflect
+	 * the entire tree, not just the subtree accessible through the first
+	 * child of the root (which is now two levels down from the root).
 	 */
-	if (bp && xfs_buf_daddr(bp) != old_bn) {
+	if (!xfs_btree_ptr_is_null(cur, &nptr) &&
+	    bp && xfs_buf_daddr(bp) != old_bn) {
 		xfs_btree_get_keys(cur, block, lkey);
 	} else if (xfs_btree_needs_key_update(cur, optr)) {
 		error = xfs_btree_update_keys(cur, level);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Darrick J. Wong <[email protected]>

commit 7ce31f20a0771d71779c3b0ec9cdf474cc3c8e9a upstream.

Way back when we first implemented FICLONE for XFS, life was simple -- either the the entire remapping completed, or something happened and we had to return an errno explaining what happened. Neither of those ioctls support returning partial results, so it's all or nothing.

Then things got complicated when copy_file_range came along, because it actually can return the number of bytes copied, so commit 3f68c1f562f1e4 tried to make it so that we could return a partial result if the REMAP_FILE_CAN_SHORTEN flag is set. This is also how FIDEDUPERANGE can indicate that the kernel performed a partial deduplication.

Unfortunately, the logic is wrong if an error stops the remapping and CAN_SHORTEN is not set. Because those callers cannot return partial results, it is an error for ->remap_file_range to return a positive quantity that is less than the @len passed in. Implementations really should be returning a negative errno in this case, because that's what btrfs (which introduced FICLONE{,RANGE}) did.

Therefore, ->remap_range implementations cannot silently drop an errno that they might have when the number of bytes remapped is less than the number of bytes requested and CAN_SHORTEN is not set.

Found by running generic/562 on a 64k fsblock filesystem and wondering why it reported corrupt files.

Really-Fixes: 3f68c1f562f1e4 ("xfs: support returning partial reflink results")

---
 fs/xfs/xfs_file.c |    8 ++++++++
 1 file changed, 8 insertions(+)

@@ -1193,6 +1193,14 @@ out_unlock:
 	xfs_iunlock2_io_mmap(src, dest);
 	if (ret)
 		trace_xfs_reflink_remap_range_error(dest, ret, _RET_IP_);
+	/*
+	 * If the caller did not set CAN_SHORTEN, then it is not prepared to
+	 * handle partial results -- either the whole remap succeeds, or we
+	 * must say why it did not.  In this case, any error should be returned
+	 * to the caller.
+	 */
+	if (ret && remapped < len && !(remap_flags & REMAP_FILE_CAN_SHORTEN))
+		return ret;
 	return remapped > 0 ? remapped : ret;
 }
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Darrick J. Wong <[email protected]>

commit 7f8b718c58783f3ff0810b39e2f62f50ba2549f6 upstream.

V4 symlink blocks didn't have headers, so return early if this is a V4 filesystem.

---
 fs/xfs/libxfs/xfs_symlink_remote.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -89,8 +89,10 @@ xfs_symlink_verify(
 	struct xfs_mount	*mp = bp->b_mount;
 	struct xfs_dsymlink_hdr	*dsl = bp->b_addr;
 
+	/* no verification of non-crc buffers */
 	if (!xfs_has_crc(mp))
-		return __this_address;
+		return NULL;
+
 	if (!xfs_verify_magic(bp, dsl->sl_magic))
 		return __this_address;
 	if (!uuid_equal(&dsl->sl_uuid, &mp->m_sb.sb_meta_uuid))

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Darrick J. Wong <[email protected]>

commit ffc3ea4f3c1cc83a86b7497b0c4b0aee7de5480d upstream.

Fix a minor mistakes in the scrub tracepoints that can manifest when inode-rooted btrees are enabled. The existing code worked fine for bmap btrees, but we should tighten the code up to be less sloppy.

---
 fs/xfs/scrub/trace.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -464,7 +464,7 @@ TRACE_EVENT(xchk_ifork_btree_error,
 	TP_fast_assign(
 		xfs_fsblock_t fsbno = xchk_btree_cur_fsbno(cur, level);
 		__entry->dev = sc->mp->m_super->s_dev;
-		__entry->ino = sc->ip->i_ino;
+		__entry->ino = cur->bc_ino.ip->i_ino;
 		__entry->whichfork = cur->bc_ino.whichfork;
 		__entry->type = sc->sm->sm_type;
 		__entry->btnum = cur->bc_btnum;

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Michal Luczaj <[email protected]>

commit 75e072a390da9a22e7ae4a4e8434dfca5da499fb upstream.

Consider a sockmap entry being updated with the same socket:

	osk = stab->sks[idx];
	sock_map_add_link(psock, link, map, &stab->sks[idx]);
	stab->sks[idx] = sk;
	if (osk)
		sock_map_unref(osk, &stab->sks[idx]);

Due to sock_map_unref(), which invokes sock_map_del_link(), all the psock's links for stab->sks[idx] are torn:

	list_for_each_entry_safe(link, tmp, &psock->link, list) {
		if (link->link_raw == link_raw) {
			...
			list_del(&link->list);
			sk_psock_free_link(link);
		}
	}

And that includes the new link sock_map_add_link() added just before the unref.

This results in a sockmap holding a socket, but without the respective link. This in turn means that close(sock) won't trigger the cleanup, i.e. a closed socket will not be automatically removed from the sockmap.

Stop tearing the links when a matching link_raw is found.

---
 net/core/sock_map.c |    1 +
 1 file changed, 1 insertion(+)

@@ -158,6 +158,7 @@ static void sock_map_del_link(struct soc
 				verdict_stop = true;
 			list_del(&link->list);
 			sk_psock_free_link(link);
+			break;
 		}
 	}
 	spin_unlock_bh(&psock->link_lock);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Michal Luczaj <[email protected]>

commit d7b0ff5a866724c3ad21f2628c22a63336deec3f upstream.

As the final stages of socket destruction may be delayed, it is possible that virtio_transport_recv_listen() will be called after the accept_queue has been flushed, but before the SOCK_DONE flag has been set. As a result, sockets enqueued after the flush would remain unremoved, leading to a memory leak.

vsock_release
__vsock_release
lock
virtio_transport_release
virtio_transport_close
schedule_delayed_work(close_work)
sk_shutdown = SHUTDOWN_MASK
(!) flush accept_queue
release
virtio_transport_recv_pkt
vsock_find_bound_socket lock
if flag(SOCK_DONE) return virtio_transport_recv_listen
child = vsock_create_connected
(!) vsock_enqueue_accept(child)
release
close_work
lock
virtio_transport_do_close
set_flag(SOCK_DONE)
virtio_transport_remove_sock
vsock_remove_sock
vsock_remove_bound
release

Introduce a sk_shutdown check to disallow vsock_enqueue_accept() during socket destruction.

unreferenced object 0xffff888109e3f800 (size 2040):
comm "kworker/5:2", pid 371, jiffies 4294940105 hex dump (first 32 bytes):

    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
    28 00 0b 40 00 00 00 00 00 00 00 00 00 00 00 00  (..@............
  backtrace (crc 9e5f4e84):
    [<ffffffff81418ff1>] kmem_cache_alloc_noprof+0x2c1/0x360
    [<ffffffff81d27aa0>] sk_prot_alloc+0x30/0x120
    [<ffffffff81d2b54c>] sk_alloc+0x2c/0x4b0
    [<ffffffff81fe049a>] __vsock_create.constprop.0+0x2a/0x310
    [<ffffffff81fe6d6c>] virtio_transport_recv_pkt+0x4dc/0x9a0
    [<ffffffff81fe745d>] vsock_loopback_work+0xfd/0x140
    [<ffffffff810fc6ac>] process_one_work+0x20c/0x570
    [<ffffffff810fce3f>] worker_thread+0x1bf/0x3a0
    [<ffffffff811070dd>] kthread+0xdd/0x110
    [<ffffffff81044fdd>] ret_from_fork+0x2d/0x50
    [<ffffffff8100785a>] ret_from_fork_asm+0x1a/0x30

---
 net/vmw_vsock/virtio_transport_common.c |    8 ++++++++
 1 file changed, 8 insertions(+)

@@ -1196,6 +1196,14 @@ virtio_transport_recv_listen(struct sock
 		return -ENOMEM;
 	}
 
+	/* __vsock_release() might have already flushed accept_queue.
+	 * Subsequent enqueues would lead to a memory leak.
+	 */
+	if (sk->sk_shutdown == SHUTDOWN_MASK) {
+		virtio_transport_reset_no_sock(t, pkt);
+		return -ESHUTDOWN;
+	}
+
 	child = vsock_create_connected(sk);
 	if (!child) {
 		virtio_transport_reset_no_sock(t, pkt);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Sungjong Seo <[email protected]>

commit 89fc548767a2155231128cb98726d6d2ea1256c9 upstream.

When accessing a file with more entries than ES_MAX_ENTRY_NUM, the bh-array is allocated in __exfat_get_entry_set. The problem is that the bh-array is allocated with GFP_KERNEL. It does not make sense. In the following cases, a deadlock for sbi->s_lock between the two processes may occur.

       CPU0                CPU1
       ----                ----
  kswapd
   balance_pgdat
    lock(fs_reclaim)
                      exfat_iterate
                       lock(&sbi->s_lock)
                       exfat_readdir
                        exfat_get_uniname_from_ext_entry
                         exfat_get_dentry_set
                          __exfat_get_dentry_set
                           kmalloc_array
                            ...
                            lock(fs_reclaim)
    ...
    evict
     exfat_evict_inode
      lock(&sbi->s_lock)

To fix this, let's allocate bh-array with GFP_NOFS.

---
 fs/exfat/dir.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -878,7 +878,7 @@ struct exfat_entry_set_cache *exfat_get_
 
 	num_bh = EXFAT_B_TO_BLK_ROUND_UP(off + num_entries * DENTRY_SIZE, sb);
 	if (num_bh > ARRAY_SIZE(es->__bh)) {
-		es->bh = kmalloc_array(num_bh, sizeof(*es->bh), GFP_KERNEL);
+		es->bh = kmalloc_array(num_bh, sizeof(*es->bh), GFP_NOFS);
 		if (!es->bh) {
 			brelse(bh);
 			kfree(es);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Suraj Sonawane <[email protected]>

[ Upstream commit 265e98f72bac6c41a4492d3e30a8e5fd22fe0779 ]

Fix an issue detected by syzbot with KASAN:

BUG: KASAN: vmalloc-out-of-bounds in cmd_to_func drivers/acpi/nfit/
core.c:416 [inline]
BUG: KASAN: vmalloc-out-of-bounds in acpi_nfit_ctl+0x20e8/0x24a0

drivers/acpi/nfit/core.c:459

The issue occurs in cmd_to_func when the call_pkg->nd_reserved2 array is accessed without verifying that call_pkg points to a buffer that is appropriately sized as a struct nd_cmd_pkg. This can lead to out-of-bounds access and undefined behavior if the buffer does not have sufficient space.

To address this, a check was added in acpi_nfit_ctl() to ensure that buf is not NULL and that buf_len is less than sizeof(*call_pkg) before accessing it. This ensures safe access to the members of call_pkg, including the nd_reserved2 array.

---
 drivers/acpi/nfit/core.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

index 2575d6c51f89..9cc83df22de4 100644
@@ -454,8 +454,13 @@ int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
 	if (cmd_rc)
 		*cmd_rc = -EINVAL;
 
-	if (cmd == ND_CMD_CALL)
+	if (cmd == ND_CMD_CALL) {
+		if (!buf || buf_len < sizeof(*call_pkg))
+			return -EINVAL;
+
 		call_pkg = buf;
+	}
+
 	func = cmd_to_func(nfit_mem, cmd, call_pkg, &family);
 	if (func < 0)
 		return func;

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Remi Pommarel <[email protected]>

[ Upstream commit f2f7358c3890e7366cbcb7512b4bc8b4394b2d61 ]

The number of TT changes can be less than initially expected in batadv_tt_tvlv_container_update() (changes can be removed by batadv_tt_local_event() in ADD+DEL sequence between reading tt_diff_entries_num and actually iterating the change list under lock).

Thus tt_diff_len could be bigger than the actual changes size that need to be sent. Because batadv_send_my_tt_response sends the whole packet, uninitialized data can be interpreted as TT changes on other nodes leading to weird TT global entries on those nodes such as:

* 00:00:00:00:00:00 -1 [....] ( 0) 88:12:4e:ad:7e:ba (179) (0x45845380)
* 00:00:00:00:78:79 4092 [.W..] ( 0) 88:12:4e:ad:7e:3c (145) (0x8ebadb8b)

All of the above also applies to OGM tvlv container buffer's tvlv_len.

Remove the extra allocated space to avoid sending uninitialized TT changes in batadv_send_my_tt_response() and batadv_v_ogm_send_softif().

---
 net/batman-adv/translation-table.c | 7 +++++++
 1 file changed, 7 insertions(+)

index 660a5594a647..ec30c1d36160 100644
@@ -990,6 +990,7 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	int tt_diff_len, tt_change_len = 0;
 	int tt_diff_entries_num = 0;
 	int tt_diff_entries_count = 0;
+	size_t tt_extra_len = 0;
 	u16 tvlv_len;
 
 	tt_diff_entries_num = atomic_read(&bat_priv->tt.local_changes);

@@ -1027,6 +1028,9 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	}
 	spin_unlock_bh(&bat_priv->tt.changes_list_lock);
 
+	tt_extra_len = batadv_tt_len(tt_diff_entries_num -
+				     tt_diff_entries_count);
+
 	/* Keep the buffer for possible tt_request */
 	spin_lock_bh(&bat_priv->tt.last_changeset_lock);
 	kfree(bat_priv->tt.last_changeset);

@@ -1035,6 +1039,7 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	tt_change_len = batadv_tt_len(tt_diff_entries_count);
 	/* check whether this new OGM has no changes due to size problems */
 	if (tt_diff_entries_count > 0) {
+		tt_diff_len -= tt_extra_len;
 		/* if kmalloc() fails we will reply with the full table
 		 * instead of providing the diff
 		 */

@@ -1047,6 +1052,8 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	}
 	spin_unlock_bh(&bat_priv->tt.last_changeset_lock);
 
+	/* Remove extra packet space for OGM */
+	tvlv_len -= tt_extra_len;
 container_register:
 	batadv_tvlv_container_register(bat_priv, BATADV_TVLV_TT, 1, tt_data,
 				       tvlv_len);

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Remi Pommarel <[email protected]>

[ Upstream commit 8038806db64da15721775d6b834990cacbfcf0b2 ]

The number of entries filled by batadv_tt_tvlv_generate() can be less than initially expected in batadv_tt_prepare_tvlv_{global,local}_data() (changes can be removed by batadv_tt_local_event() in ADD+DEL sequence in the meantime as the lock held during the whole tvlv global/local data generation).

Thus tvlv_len could be bigger than the actual TT entry size that need to be sent so full table TT_RESPONSE could hold invalid TT entries such as below.

* 00:00:00:00:00:00 -1 [....] ( 0) 88:12:4e:ad:7e:ba (179) (0x45845380)
* 00:00:00:00:78:79 4092 [.W..] ( 0) 88:12:4e:ad:7e:3c (145) (0x8ebadb8b)

Remove the extra allocated space to avoid sending uninitialized entries for full table TT_RESPONSE in both batadv_send_other_tt_response() and batadv_send_my_tt_response().

---
 net/batman-adv/translation-table.c | 37 ++++++++++++++++++------------
 1 file changed, 22 insertions(+), 15 deletions(-)

index ec30c1d36160..5700a9eaa49a 100644
@@ -2754,14 +2754,16 @@ static bool batadv_tt_global_valid(const void *entry_ptr,
  *
  * Fills the tvlv buff with the tt entries from the specified hash. If valid_cb
  * is not provided then this becomes a no-op.
+ *
+ * Return: Remaining unused length in tvlv_buff.
  */
-static void batadv_tt_tvlv_generate(struct batadv_priv *bat_priv,
-				    struct batadv_hashtable *hash,
-				    void *tvlv_buff, u16 tt_len,
-				    bool (*valid_cb)(const void *,
-						     const void *,
-						     u8 *flags),
-				    void *cb_data)
+static u16 batadv_tt_tvlv_generate(struct batadv_priv *bat_priv,
+				   struct batadv_hashtable *hash,
+				   void *tvlv_buff, u16 tt_len,
+				   bool (*valid_cb)(const void *,
+						    const void *,
+						    u8 *flags),
+				   void *cb_data)
 {
 	struct batadv_tt_common_entry *tt_common_entry;
 	struct batadv_tvlv_tt_change *tt_change;

@@ -2775,7 +2777,7 @@ static void batadv_tt_tvlv_generate(struct batadv_priv *bat_priv,
 	tt_change = (struct batadv_tvlv_tt_change *)tvlv_buff;
 
 	if (!valid_cb)
-		return;
+		return tt_len;
 
 	rcu_read_lock();
 	for (i = 0; i < hash->size; i++) {

@@ -2801,6 +2803,8 @@ static void batadv_tt_tvlv_generate(struct batadv_priv *bat_priv,
 		}
 	}
 	rcu_read_unlock();
+
+	return batadv_tt_len(tt_tot - tt_num_entries);
 }
 
 /**

@@ -3076,10 +3080,11 @@ static bool batadv_send_other_tt_response(struct batadv_priv *bat_priv,
 			goto out;
 
 		/* fill the rest of the tvlv with the real TT entries */
-		batadv_tt_tvlv_generate(bat_priv, bat_priv->tt.global_hash,
-					tt_change, tt_len,
-					batadv_tt_global_valid,
-					req_dst_orig_node);
+		tvlv_len -= batadv_tt_tvlv_generate(bat_priv,
+						    bat_priv->tt.global_hash,
+						    tt_change, tt_len,
+						    batadv_tt_global_valid,
+						    req_dst_orig_node);
 	}
 
 	/* Don't send the response, if larger than fragmented packet. */

@@ -3203,9 +3208,11 @@ static bool batadv_send_my_tt_response(struct batadv_priv *bat_priv,
 			goto out;
 
 		/* fill the rest of the tvlv with the real TT entries */
-		batadv_tt_tvlv_generate(bat_priv, bat_priv->tt.local_hash,
-					tt_change, tt_len,
-					batadv_tt_local_valid, NULL);
+		tvlv_len -= batadv_tt_tvlv_generate(bat_priv,
+						    bat_priv->tt.local_hash,
+						    tt_change, tt_len,
+						    batadv_tt_local_valid,
+						    NULL);
 	}
 
 	tvlv_tt_data->flags = BATADV_TT_RESPONSE;

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Remi Pommarel <[email protected]>

[ Upstream commit fff8f17c1a6fc802ca23bbd3a276abfde8cc58e6 ]

When TT changes list is too big to fit in packet due to MTU size, an empty OGM is sent expected other node to send TT request to get the changes. The issue is that tt.last_changeset was not built thus the originator was responding with previous changes to those TT requests (see batadv_send_my_tt_response). Also the changes list was never cleaned up effectively never ending growing from this point onwards, repeatedly sending the same TT response changes over and over, and creating a new empty OGM every OGM interval expecting for the local changes to be purged.

When there is more TT changes that can fit in packet, drop all changes, send empty OGM and wait for TT request so we can respond with a full table instead.

---
 net/batman-adv/translation-table.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

index 5700a9eaa49a..6be14f9e071a 100644
@@ -990,6 +990,7 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	int tt_diff_len, tt_change_len = 0;
 	int tt_diff_entries_num = 0;
 	int tt_diff_entries_count = 0;
+	bool drop_changes = false;
 	size_t tt_extra_len = 0;
 	u16 tvlv_len;
 

@@ -997,10 +998,17 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 	tt_diff_len = batadv_tt_len(tt_diff_entries_num);
 
 	/* if we have too many changes for one packet don't send any
-	 * and wait for the tt table request which will be fragmented
+	 * and wait for the tt table request so we can reply with the full
+	 * (fragmented) table.
+	 *
+	 * The local change history should still be cleaned up so the next
+	 * TT round can start again with a clean state.
 	 */
-	if (tt_diff_len > bat_priv->soft_iface->mtu)
+	if (tt_diff_len > bat_priv->soft_iface->mtu) {
 		tt_diff_len = 0;
+		tt_diff_entries_num = 0;
+		drop_changes = true;
+	}
 
 	tvlv_len = batadv_tt_prepare_tvlv_local_data(bat_priv, &tt_data,
 						     &tt_change, &tt_diff_len);

@@ -1009,7 +1017,7 @@ static void batadv_tt_tvlv_container_update(struct batadv_priv *bat_priv)
 
 	tt_data->flags = BATADV_TT_OGM_DIFF;
 
-	if (tt_diff_len == 0)
+	if (!drop_changes && tt_diff_len == 0)
 		goto container_register;
 
 	spin_lock_bh(&bat_priv->tt.changes_list_lock);

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Eric Dumazet <[email protected]>

[ Upstream commit b04d86fff66b15c07505d226431f808c15b1703c ]

syzbot found [1] that after blamed commit, ub->ubsock->sk was NULL when attempting the atomic_dec() :

atomic_dec(&tipc_net(sock_net(ub->ubsock->sk))->wq_count);

Fix this by caching the tipc_net pointer.

[1]

Oops: general protection fault, probably for non-canonical address 0xdffffc0000000006: 0000 [#1] PREEMPT SMP KASAN PTI
KASAN: null-ptr-deref in range [0x0000000000000030-0x0000000000000037]
CPU: 0 UID: 0 PID: 5896 Comm: kworker/0:3 Not tainted 6.13.0-rc1-next-20241203-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/13/2024
Workqueue: events cleanup_bearer
 RIP: 0010:read_pnet include/net/net_namespace.h:387 [inline]
 RIP: 0010:sock_net include/net/sock.h:655 [inline]
 RIP: 0010:cleanup_bearer+0x1f7/0x280 net/tipc/udp_media.c:820
Code: 18 48 89 d8 48 c1 e8 03 42 80 3c 28 00 74 08 48 89 df e8 3c f7 99 f6 48 8b 1b 48 83 c3 30 e8 f0 e4 60 00 48 89 d8 48 c1 e8 03 <42> 80 3c 28 00 74 08 48 89 df e8 1a f7 99 f6 49 83 c7 e8 48 8b 1b
RSP: 0018:ffffc9000410fb70 EFLAGS: 00010206
RAX: 0000000000000006 RBX: 0000000000000030 RCX: ffff88802fe45a00
RDX: 0000000000000001 RSI: 0000000000000008 RDI: ffffc9000410f900
RBP: ffff88807e1f0908 R08: ffffc9000410f907 R09: 1ffff92000821f20
R10: dffffc0000000000 R11: fffff52000821f21 R12: ffff888031d19980
R13: dffffc0000000000 R14: dffffc0000000000 R15: ffff88807e1f0918
FS:  0000000000000000(0000) GS:ffff8880b8600000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000556ca050b000 CR3: 0000000031c0c000 CR4: 00000000003526f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400

---
 net/tipc/udp_media.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

index 3f5a12b85b2d..f5bd75d931c1 100644
@@ -811,6 +811,7 @@ static void cleanup_bearer(struct work_struct *work)
 {
 	struct udp_bearer *ub = container_of(work, struct udp_bearer, work);
 	struct udp_replicast *rcast, *tmp;
+	struct tipc_net *tn;
 
 	list_for_each_entry_safe(rcast, tmp, &ub->rcast.list, list) {
 		dst_cache_destroy(&rcast->dst_cache);

@@ -818,10 +819,14 @@ static void cleanup_bearer(struct work_struct *work)
 		kfree_rcu(rcast, rcu);
 	}
 
+	tn = tipc_net(sock_net(ub->ubsock->sk));
+
 	dst_cache_destroy(&ub->rcast.dst_cache);
 	udp_tunnel_sock_release(ub->ubsock);
+
+	/* Note: could use a call_rcu() to avoid another synchronize_net() */
 	synchronize_net();
-	atomic_dec(&tipc_net(sock_net(ub->ubsock->sk))->wq_count);
+	atomic_dec(&tn->wq_count);
 	kfree(ub);
 }
 

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Danielle Ratson <[email protected]>

[ Upstream commit cf3515c556907b4da290967a2a6cbbd9ee0ee723 ]

The test is sending only one packet generated with mausezahn from $h1 to $h2. However, for some reason, it is testing for non-zero maximum occupancy in both the ingress pool of $h1 and $h2. The former only passes when $h2 happens to send a packet.

Avoid intermittent failures by removing unintentional test case regarding the ingress pool of $h1.

---
 tools/testing/selftests/drivers/net/mlxsw/sharedbuffer.sh | 5 -----
 1 file changed, 5 deletions(-)

index 7d9e73a43a49..f6f5e2090891 100755
@@ -108,11 +108,6 @@ port_pool_test()
 
 	devlink sb occupancy snapshot $DEVLINK_DEV
 
-	RET=0
-	max_occ=$(sb_occ_pool_check $dl_port1 $SB_POOL_ING $exp_max_occ)
-	check_err $? "Expected iPool($SB_POOL_ING) max occupancy to be $exp_max_occ, but got $max_occ"
-	log_test "physical port's($h1) ingress pool"
-
 	RET=0
 	max_occ=$(sb_occ_pool_check $dl_port2 $SB_POOL_ING $exp_max_occ)
 	check_err $? "Expected iPool($SB_POOL_ING) max occupancy to be $exp_max_occ, but got $max_occ"

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Danielle Ratson <[email protected]>

[ Upstream commit 6c46ad4d1bb2e8ec2265296e53765190f6e32f33 ]

On both port_tc_ip_test() and port_tc_arp_test(), the max occupancy is checked on $h2 twice, when only the error message is different and does not match the check itself.

Remove the two duplicated test cases from the test.

---
 .../selftests/drivers/net/mlxsw/sharedbuffer.sh        | 10 ----------
 1 file changed, 10 deletions(-)

index f6f5e2090891..9c3c426197af 100755
@@ -131,11 +131,6 @@ port_tc_ip_test()
 
 	devlink sb occupancy snapshot $DEVLINK_DEV
 
-	RET=0
-	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
-	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
-	log_test "physical port's($h1) ingress TC - IP packet"
-
 	RET=0
 	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
 	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"

@@ -162,11 +157,6 @@ port_tc_arp_test()
 
 	devlink sb occupancy snapshot $DEVLINK_DEV
 
-	RET=0
-	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
-	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"
-	log_test "physical port's($h1) ingress TC - ARP packet"
-
 	RET=0
 	max_occ=$(sb_occ_itc_check $dl_port2 $SB_ITC $exp_max_occ)
 	check_err $? "Expected ingress TC($SB_ITC) max occupancy to be $exp_max_occ, but got $max_occ"

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Jeremi Piotrowski <[email protected]>

[ Upstream commit 6365ba64b4dbe8b59ddaeaa724b281f3787715d5 ]

KVM_HC_CLOCK_PAIRING currently fails inside SEV-SNP guests because the guest passes an address to static data to the host. In confidential computing the host can't access arbitrary guest memory so handling the hypercall runs into an "rmpfault". To make the hypercall work, the guest needs to explicitly mark the memory as decrypted. Do that in kvm_arch_ptp_init(), but retain the previous behavior for non-confidential guests to save us from having to allocate memory.

Add a new arch-specific function (kvm_arch_ptp_exit()) to free the allocation and mark the memory as encrypted again.

Stable-dep-of: 5e7aa97c7acf ("ptp: kvm: x86: Return EOPNOTSUPP instead of ENODEV from kvm_arch_ptp_init()")

---
 drivers/ptp/ptp_kvm_arm.c    |  4 +++
 drivers/ptp/ptp_kvm_common.c |  1 +
 drivers/ptp/ptp_kvm_x86.c    | 59 +++++++++++++++++++++++++++++-------
 include/linux/ptp_kvm.h      |  1 +
 4 files changed, 54 insertions(+), 11 deletions(-)

index b7d28c8dfb84..e68e6943167b 100644
@@ -22,6 +22,10 @@ int kvm_arch_ptp_init(void)
 	return 0;
 }
 
+void kvm_arch_ptp_exit(void)
+{
+}
+
 int kvm_arch_ptp_get_clock(struct timespec64 *ts)
 {
 	return kvm_arch_ptp_get_crosststamp(NULL, ts, NULL);

index fcae32f56f25..051114a59286 100644
@@ -130,6 +130,7 @@ static struct kvm_ptp_clock kvm_ptp_clock;
 static void __exit ptp_kvm_exit(void)
 {
 	ptp_clock_unregister(kvm_ptp_clock.ptp_clock);
+	kvm_arch_ptp_exit();
 }
 
 static int __init ptp_kvm_init(void)

index f960a719f0d5..c1636ce76bd2 100644
@@ -12,6 +12,7 @@ struct timespec64;
 struct clocksource;
 
 int kvm_arch_ptp_init(void);
+void kvm_arch_ptp_exit(void);
 int kvm_arch_ptp_get_clock(struct timespec64 *ts);
 int kvm_arch_ptp_get_crosststamp(u64 *cycle,
 		struct timespec64 *tspec, struct clocksource **cs);

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Thomas Weißschuh <[email protected]>

[ Upstream commit 5e7aa97c7acf171275ac02a8bb018c31b8918d13 ]

The caller, ptp_kvm_init(), emits a warning if kvm_arch_ptp_init() exits with any error which is not EOPNOTSUPP:

"fail to initialize ptp_kvm"

Replace ENODEV with EOPNOTSUPP to avoid this spurious warning, aligning with the ARM implementation.

---
 drivers/ptp/ptp_kvm_x86.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

index 902844cc1a17..5e5b2ef78547 100644
@@ -26,7 +26,7 @@ int kvm_arch_ptp_init(void)
 	long ret;
 
 	if (!kvm_para_available())
-		return -ENODEV;
+		return -EOPNOTSUPP;
 
 	if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT)) {
 		p = alloc_page(GFP_KERNEL | __GFP_ZERO);

@@ -46,14 +46,14 @@ int kvm_arch_ptp_init(void)
 
 	clock_pair_gpa = slow_virt_to_phys(clock_pair);
 	if (!pvclock_get_pvti_cpu0_va()) {
-		ret = -ENODEV;
+		ret = -EOPNOTSUPP;
 		goto err;
 	}
 
 	ret = kvm_hypercall2(KVM_HC_CLOCK_PAIRING, clock_pair_gpa,
 			     KVM_CLOCK_PAIRING_WALLCLOCK);
 	if (ret == -KVM_ENOSYS) {
-		ret = -ENODEV;
+		ret = -EOPNOTSUPP;
 		goto err;
 	}
 

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Eric Dumazet <[email protected]>

[ Upstream commit a6d75ecee2bf828ac6a1b52724aba0a977e4eaf4 ]

It is unclear if net/lapb code is supposed to be ready for 8021q.

We can at least avoid crashes like the following :

skbuff: skb_under_panic: text:ffffffff8aabe1f6 len:24 put:20 head:ffff88802824a400 data:ffff88802824a3fe tail:0x16 end:0x140 dev:nr0.2

------------[ cut here ]------------
 kernel BUG at net/core/skbuff.c:206 !
Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
CPU: 1 UID: 0 PID: 5508 Comm: dhcpcd Not tainted 6.12.0-rc7-syzkaller-00144-g66418447d27b #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/30/2024
 RIP: 0010:skb_panic net/core/skbuff.c:206 [inline]
 RIP: 0010:skb_under_panic+0x14b/0x150 net/core/skbuff.c:216
Code: 0d 8d 48 c7 c6 2e 9e 29 8e 48 8b 54 24 08 8b 0c 24 44 8b 44 24 04 4d 89 e9 50 41 54 41 57 41 56 e8 1a 6f 37 02 48 83 c4 20 90 <0f> 0b 0f 1f 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 f3
RSP: 0018:ffffc90002ddf638 EFLAGS: 00010282
RAX: 0000000000000086 RBX: dffffc0000000000 RCX: 7a24750e538ff600
RDX: 0000000000000000 RSI: 0000000000000201 RDI: 0000000000000000
RBP: ffff888034a86650 R08: ffffffff8174b13c R09: 1ffff920005bbe60
R10: dffffc0000000000 R11: fffff520005bbe61 R12: 0000000000000140
R13: ffff88802824a400 R14: ffff88802824a3fe R15: 0000000000000016
FS:  00007f2a5990d740(0000) GS:ffff8880b8700000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000110c2631fd CR3: 0000000029504000 CR4: 00000000003526f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
  skb_push+0xe5/0x100 net/core/skbuff.c:2636
  nr_header+0x36/0x320 net/netrom/nr_dev.c:69
  dev_hard_header include/linux/netdevice.h:3148 [inline]
  vlan_dev_hard_header+0x359/0x480 net/8021q/vlan_dev.c:83
  dev_hard_header include/linux/netdevice.h:3148 [inline]
  lapbeth_data_transmit+0x1f6/0x2a0 drivers/net/wan/lapbether.c:257
  lapb_data_transmit+0x91/0xb0 net/lapb/lapb_iface.c:447
  lapb_transmit_buffer+0x168/0x1f0 net/lapb/lapb_out.c:149
 lapb_establish_data_link+0x84/0xd0
 lapb_device_event+0x4e0/0x670
  notifier_call_chain+0x19f/0x3e0 kernel/notifier.c:93
 __dev_notify_flags+0x207/0x400
  dev_change_flags+0xf0/0x1a0 net/core/dev.c:8922
  devinet_ioctl+0xa4e/0x1aa0 net/ipv4/devinet.c:1188
  inet_ioctl+0x3d7/0x4f0 net/ipv4/af_inet.c:1003
  sock_do_ioctl+0x158/0x460 net/socket.c:1227
  sock_ioctl+0x626/0x8e0 net/socket.c:1346
  vfs_ioctl fs/ioctl.c:51 [inline]
  __do_sys_ioctl fs/ioctl.c:907 [inline]
  __se_sys_ioctl+0xf9/0x170 fs/ioctl.c:893
  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83

---
 include/net/lapb.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 124ee122f2c8..6c07420644e4 100644
@@ -4,7 +4,7 @@
 #include <linux/lapb.h>
 #include <linux/refcount.h>
 
-#define	LAPB_HEADER_LEN	20		/* LAPB over Ethernet + a bit more */
+#define	LAPB_HEADER_LEN MAX_HEADER		/* LAPB over Ethernet + a bit more */
 
 #define	LAPB_ACK_PENDING_CONDITION	0x01
 #define	LAPB_REJECT_CONDITION		0x02

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Daniel Machon <[email protected]>

[ Upstream commit f004f2e535e2b66ccbf5ac35f8eaadeac70ad7b7 ]

The FDMA handler is responsible for scheduling a NAPI poll, which will eventually fetch RX packets from the FDMA queue. Currently, the FDMA handler is run in a threaded context. For some reason, this kills performance. Admittedly, I did not do a thorough investigation to see exactly what causes the issue, however, I noticed that in the other driver utilizing the same FDMA engine, we run the FDMA handler in hard IRQ context.

Fix this performance issue, by running the FDMA handler in hard IRQ context, not deferring any work to a thread.

Prior to this change, the RX UDP performance was:

Interval           Transfer     Bitrate         Jitter
0.00-10.20  sec    44.6 MBytes  36.7 Mbits/sec  0.027 ms

After this change, the rx UDP performance is:

Interval           Transfer     Bitrate         Jitter
0.00-9.12   sec    1.01 GBytes  953 Mbits/sec   0.020 ms

---
 drivers/net/ethernet/microchip/sparx5/sparx5_main.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

index 174d89ee6374..ccfad6fba5b4 100644
@@ -647,12 +647,11 @@ static int sparx5_start(struct sparx5 *sparx5)
 	err = -ENXIO;
 	if (sparx5->fdma_irq >= 0) {
 		if (GCB_CHIP_ID_REV_ID_GET(sparx5->chip_id) > 0)
-			err = devm_request_threaded_irq(sparx5->dev,
-							sparx5->fdma_irq,
-							NULL,
-							sparx5_fdma_handler,
-							IRQF_ONESHOT,
-							"sparx5-fdma", sparx5);
+			err = devm_request_irq(sparx5->dev,
+					       sparx5->fdma_irq,
+					       sparx5_fdma_handler,
+					       0,
+					       "sparx5-fdma", sparx5);
 		if (!err)
 			err = sparx5_fdma_start(sparx5);
 		if (err)

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Daniel Machon <[email protected]>

[ Upstream commit ddd7ba006078a2bef5971b2dc5f8383d47f96207 ]

On port initialization, we configure the maximum frame length accepted by the receive module associated with the port. This value is currently written to the MAX_LEN field of the DEV10G_MAC_ENA_CFG register, when in fact, it should be written to the DEV10G_MAC_MAXLEN_CFG register. Fix this.

---
 drivers/net/ethernet/microchip/sparx5/sparx5_port.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 8561a7bf53e1..0d9ff3db3144 100644
@@ -1113,7 +1113,7 @@ int sparx5_port_init(struct sparx5 *sparx5,
 	spx5_inst_rmw(DEV10G_MAC_MAXLEN_CFG_MAX_LEN_SET(ETH_MAXLEN),
 		      DEV10G_MAC_MAXLEN_CFG_MAX_LEN,
 		      devinst,
-		      DEV10G_MAC_ENA_CFG(0));
+		      DEV10G_MAC_MAXLEN_CFG(0));
 
 	/* Handle Signal Detect in 10G PCS */
 	spx5_inst_wr(PCS10G_BR_PCS_SD_CFG_SD_POL_SET(sd_pol) |

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 7899ca9f3bd2b008e9a7c41f2a9f1986052d7e96 ]

In acpi_decode_space() addr->info.mem.caching is checked on main level for any resource type but addr->info.mem is part of union and thus valid only if the resource type is memory range.

Move the check inside the preceeding switch/case to only execute it when the union is of correct type.

---
 drivers/acpi/resource.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

index 3c9f93981523..0433ab8ced0d 100644
@@ -250,6 +250,9 @@ static bool acpi_decode_space(struct resource_win *win,
 	switch (addr->resource_type) {
 	case ACPI_MEMORY_RANGE:
 		acpi_dev_memresource_flags(res, len, wp);
+
+		if (addr->info.mem.caching == ACPI_PREFETCHABLE_MEMORY)
+			res->flags |= IORESOURCE_PREFETCH;
 		break;
 	case ACPI_IO_RANGE:
 		acpi_dev_ioresource_flags(res, len, iodec,

@@ -265,9 +268,6 @@ static bool acpi_decode_space(struct resource_win *win,
 	if (addr->producer_consumer == ACPI_PRODUCER)
 		res->flags |= IORESOURCE_WINDOW;
 
-	if (addr->info.mem.caching == ACPI_PREFETCHABLE_MEMORY)
-		res->flags |= IORESOURCE_PREFETCH;
-
 	return !(res->flags & IORESOURCE_DISABLED);
 }
 

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Anumula Murali Mohan Reddy <[email protected]>

[ Upstream commit 356983f569c1f5991661fc0050aa263792f50616 ]

t4_set_vf_mac_acl() uses pf to set mac addr, but t4vf_get_vf_mac_acl() uses port number to get mac addr, this leads to error when an attempt to set MAC address on VF's of PF2 and PF3.
This patch fixes the issue by using port number to set mac address.

---
 drivers/net/ethernet/chelsio/cxgb4/cxgb4.h      | 2 +-
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 2 +-
 drivers/net/ethernet/chelsio/cxgb4/t4_hw.c      | 5 +++--
 3 files changed, 5 insertions(+), 4 deletions(-)

index ecea3cdd30b3..cceb1cbe0d6a 100644
@@ -2084,7 +2084,7 @@ void t4_idma_monitor(struct adapter *adapter,
 		     struct sge_idma_monitor_state *idma,
 		     int hz, int ticks);
 int t4_set_vf_mac_acl(struct adapter *adapter, unsigned int vf,
-		      unsigned int naddr, u8 *addr);
+		      u8 start, unsigned int naddr, u8 *addr);
 void t4_tp_pio_read(struct adapter *adap, u32 *buff, u32 nregs,
 		    u32 start_index, bool sleep_ok);
 void t4_tp_tm_pio_read(struct adapter *adap, u32 *buff, u32 nregs,

index 0d9cda4ab303..21afaa81697e 100644
@@ -3247,7 +3247,7 @@ static int cxgb4_mgmt_set_vf_mac(struct net_device *dev, int vf, u8 *mac)
 
 	dev_info(pi->adapter->pdev_dev,
 		 "Setting MAC %pM on VF %d\n", mac, vf);
-	ret = t4_set_vf_mac_acl(adap, vf + 1, 1, mac);
+	ret = t4_set_vf_mac_acl(adap, vf + 1, pi->lport, 1, mac);
 	if (!ret)
 		ether_addr_copy(adap->vfinfo[vf].vf_mac_addr, mac);
 	return ret;

index c99f5920c957..94fd7d6a1109 100644
@@ -10215,11 +10215,12 @@ int t4_load_cfg(struct adapter *adap, const u8 *cfg_data, unsigned int size)
  *	t4_set_vf_mac_acl - Set MAC address for the specified VF
  *	@adapter: The adapter
  *	@vf: one of the VFs instantiated by the specified PF
+ *	@start: The start port id associated with specified VF
  *	@naddr: the number of MAC addresses
  *	@addr: the MAC address(es) to be set to the specified VF
  */
 int t4_set_vf_mac_acl(struct adapter *adapter, unsigned int vf,
-		      unsigned int naddr, u8 *addr)
+		      u8 start, unsigned int naddr, u8 *addr)
 {
 	struct fw_acl_mac_cmd cmd;
 

@@ -10234,7 +10235,7 @@ int t4_set_vf_mac_acl(struct adapter *adapter, unsigned int vf,
 	cmd.en_to_len16 = cpu_to_be32((unsigned int)FW_LEN16(cmd));
 	cmd.nmac = naddr;
 
-	switch (adapter->pf) {
+	switch (start) {
 	case 3:
 		memcpy(cmd.macaddr3, addr, sizeof(cmd.macaddr3));
 		break;

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

[ Upstream commit 4dba406fac06b009873fe7a28231b9b7e4288b09 ]

Storing the maximum clock speed in module parameter qcaspi_clkspeed has the unintended side effect that the first probed instance defines the value for all other instances. Fix this issue by storing it in max_speed_hz of the relevant SPI device.

This fix keeps the priority of the speed parameter (module parameter, device tree property, driver default). Btw this uses the opportunity to get the rid of the unused member clkspeed.

---
 drivers/net/ethernet/qualcomm/qca_spi.c | 24 ++++++++++--------------
 drivers/net/ethernet/qualcomm/qca_spi.h |  1 -
 2 files changed, 10 insertions(+), 15 deletions(-)

index 385e4c62ca03..926eec0a9532 100644
@@ -829,7 +829,6 @@ qcaspi_netdev_init(struct net_device *dev)
 
 	dev->mtu = QCAFRM_MAX_MTU;
 	dev->type = ARPHRD_ETHER;
-	qca->clkspeed = qcaspi_clkspeed;
 	qca->burst_len = qcaspi_burst_len;
 	qca->spi_thread = NULL;
 	qca->buffer_size = (dev->mtu + VLAN_ETH_HLEN + QCAFRM_HEADER_LEN +

@@ -918,17 +917,15 @@ qca_spi_probe(struct spi_device *spi)
 	legacy_mode = of_property_read_bool(spi->dev.of_node,
 					    "qca,legacy-mode");
 
-	if (qcaspi_clkspeed == 0) {
-		if (spi->max_speed_hz)
-			qcaspi_clkspeed = spi->max_speed_hz;
-		else
-			qcaspi_clkspeed = QCASPI_CLK_SPEED;
-	}
+	if (qcaspi_clkspeed)
+		spi->max_speed_hz = qcaspi_clkspeed;
+	else if (!spi->max_speed_hz)
+		spi->max_speed_hz = QCASPI_CLK_SPEED;
 
-	if ((qcaspi_clkspeed < QCASPI_CLK_SPEED_MIN) ||
-	    (qcaspi_clkspeed > QCASPI_CLK_SPEED_MAX)) {
-		dev_err(&spi->dev, "Invalid clkspeed: %d\n",
-			qcaspi_clkspeed);
+	if (spi->max_speed_hz < QCASPI_CLK_SPEED_MIN ||
+	    spi->max_speed_hz > QCASPI_CLK_SPEED_MAX) {
+		dev_err(&spi->dev, "Invalid clkspeed: %u\n",
+			spi->max_speed_hz);
 		return -EINVAL;
 	}
 

@@ -953,14 +950,13 @@ qca_spi_probe(struct spi_device *spi)
 		return -EINVAL;
 	}
 
-	dev_info(&spi->dev, "ver=%s, clkspeed=%d, burst_len=%d, pluggable=%d\n",
+	dev_info(&spi->dev, "ver=%s, clkspeed=%u, burst_len=%d, pluggable=%d\n",
 		 QCASPI_DRV_VERSION,
-		 qcaspi_clkspeed,
+		 spi->max_speed_hz,
 		 qcaspi_burst_len,
 		 qcaspi_pluggable);
 
 	spi->mode = SPI_MODE_3;
-	spi->max_speed_hz = qcaspi_clkspeed;
 	if (spi_setup(spi) < 0) {
 		dev_err(&spi->dev, "Unable to setup SPI device\n");
 		return -EFAULT;

index 58ad910068d4..b3b17bd46e12 100644
@@ -101,7 +101,6 @@ struct qcaspi {
 #endif
 
 	/* user configurable options */
-	u32 clkspeed;
 	u8 legacy_mode;
 	u16 burst_len;
 };

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Stefan Wahren <[email protected]>

[ Upstream commit becc6399ce3b724cffe9ccb7ef0bff440bb1b62b ]

The module parameter qcaspi_pluggable controls if QCA7000 signature should be checked at driver probe (current default) or not. Unfortunately this could fail in case the chip is temporary in reset, which isn't under total control by the Linux host. So disable this check per default in order to avoid unexpected probe failures.

---
 drivers/net/ethernet/qualcomm/qca_spi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 926eec0a9532..9b34ab547840 100644
@@ -67,7 +67,7 @@ MODULE_PARM_DESC(qcaspi_burst_len, "Number of data bytes per burst. Use 1-5000."
 
 #define QCASPI_PLUGGABLE_MIN 0
 #define QCASPI_PLUGGABLE_MAX 1
-static int qcaspi_pluggable = QCASPI_PLUGGABLE_MIN;
+static int qcaspi_pluggable = QCASPI_PLUGGABLE_MAX;
 module_param(qcaspi_pluggable, int, 0);
 MODULE_PARM_DESC(qcaspi_pluggable, "Pluggable SPI connection (yes/no).");
 

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Paul Barker <[email protected]>

[ Upstream commit ccb84dc8f4a02e7d30ffd388522996546b4d00e1 ]

Update the documentation to match the behaviour of the code.

pm_runtime_resume_and_get() always returns 0 on success, even if __pm_runtime_resume() returns 1.

---
 Documentation/power/runtime_pm.rst | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index d6bf84f061f4..921f56dffea7 100644
@@ -341,7 +341,9 @@ drivers/base/power/runtime.c and include/linux/pm_runtime.h:
 
   `int pm_runtime_resume_and_get(struct device *dev);`
     - run pm_runtime_resume(dev) and if successful, increment the device's
-      usage counter; return the result of pm_runtime_resume
+      usage counter; returns 0 on success (whether or not the device's
+      runtime PM status was already 'active') or the error code from
+      pm_runtime_resume() on failure.
 
   `int pm_request_idle(struct device *dev);`
     - submit a request to execute the subsystem-level idle callback for the

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Martin Ottens <[email protected]>

[ Upstream commit f8d4bc455047cf3903cd6f85f49978987dbb3027 ]

In general, 'qlen' of any classful qdisc should keep track of the number of packets that the qdisc itself and all of its children holds. In case of netem, 'qlen' only accounts for the packets in its internal tfifo. When netem is used with a child qdisc, the child qdisc can use 'qdisc_tree_reduce_backlog' to inform its parent, netem, about created or dropped SKBs. This function updates 'qlen' and the backlog statistics of netem, but netem does not account for changes made by a child qdisc. 'qlen' then indicates the wrong number of packets in the tfifo. If a child qdisc creates new SKBs during enqueue and informs its parent about this, netem's 'qlen' value is increased. When netem dequeues the newly created SKBs from the child, the 'qlen' in netem is not updated. If 'qlen' reaches the configured sch->limit, the enqueue function stops working, even though the tfifo is not full.

Reproduce the bug:
Ensure that the sender machine has GSO enabled. Configure netem as root qdisc and tbf as its child on the outgoing interface of the machine as follows:

$ tc qdisc add dev <oif> root handle 1: netem delay 100ms limit 100
$ tc qdisc add dev <oif> parent 1:0 tbf rate 50Mbit burst 1542 latency 50ms

Send bulk TCP traffic out via this interface, e.g., by running an iPerf3 client on the machine. Check the qdisc statistics:

$ tc -s qdisc show dev <oif>

Statistics after 10s of iPerf3 TCP test before the fix (note that netem's backlog > limit, netem stopped accepting packets):
qdisc netem 1: root refcnt 2 limit 1000 delay 100ms
Sent 2767766 bytes 1848 pkt (dropped 652, overlimits 0 requeues 0) backlog 4294528236b 1155p requeues 0
qdisc tbf 10: parent 1:1 rate 50Mbit burst 1537b lat 50ms
Sent 2767766 bytes 1848 pkt (dropped 327, overlimits 7601 requeues 0) backlog 0b 0p requeues 0

Statistics after the fix:
qdisc netem 1: root refcnt 2 limit 1000 delay 100ms
Sent 37766372 bytes 24974 pkt (dropped 9, overlimits 0 requeues 0) backlog 0b 0p requeues 0
qdisc tbf 10: parent 1:1 rate 50Mbit burst 1537b lat 50ms
Sent 37766372 bytes 24974 pkt (dropped 327, overlimits 96017 requeues 0) backlog 0b 0p requeues 0

tbf segments the GSO SKBs (tbf_segment) and updates the netem's 'qlen'. The interface fully stops transferring packets and "locks". In this case, the child qdisc and tfifo are empty, but 'qlen' indicates the tfifo is at its limit and no more packets are accepted.

This patch adds a counter for the entries in the tfifo. Netem's 'qlen' is only decreased when a packet is returned by its dequeue function, and not during enqueuing into the child qdisc. External updates to 'qlen' are thus accounted for and only the behavior of the backlog statistics changes. As in other qdiscs, 'qlen' then keeps track of how many packets are held in netem and all of its children. As before, sch->limit remains as the maximum number of packets in the tfifo. The same applies to netem's backlog statistics.

---
 net/sched/sch_netem.c | 22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)

index 93ed7bac9ee6..f459e34684ad 100644
@@ -77,6 +77,8 @@ struct netem_sched_data {
 	struct sk_buff	*t_head;
 	struct sk_buff	*t_tail;
 
+	u32 t_len;
+
 	/* optional qdisc for classful handling (NULL at netem init) */
 	struct Qdisc	*qdisc;
 

@@ -373,6 +375,7 @@ static void tfifo_reset(struct Qdisc *sch)
 	rtnl_kfree_skbs(q->t_head, q->t_tail);
 	q->t_head = NULL;
 	q->t_tail = NULL;
+	q->t_len = 0;
 }
 
 static void tfifo_enqueue(struct sk_buff *nskb, struct Qdisc *sch)

@@ -402,6 +405,7 @@ static void tfifo_enqueue(struct sk_buff *nskb, struct Qdisc *sch)
 		rb_link_node(&nskb->rbnode, parent, p);
 		rb_insert_color(&nskb->rbnode, &q->t_root);
 	}
+	q->t_len++;
 	sch->q.qlen++;
 }
 

@@ -508,7 +512,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch,
 			1<<(prandom_u32() % 8);
 	}
 
-	if (unlikely(sch->q.qlen >= sch->limit)) {
+	if (unlikely(q->t_len >= sch->limit)) {
 		/* re-link segs, so that qdisc_drop_all() frees them all */
 		skb->next = segs;
 		qdisc_drop_all(skb, sch, to_free);

@@ -692,8 +696,8 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch)
 tfifo_dequeue:
 	skb = __qdisc_dequeue_head(&sch->q);
 	if (skb) {
-		qdisc_qstats_backlog_dec(sch, skb);
 deliver:
+		qdisc_qstats_backlog_dec(sch, skb);
 		qdisc_bstats_update(sch, skb);
 		return skb;
 	}

@@ -709,8 +713,7 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch)
 
 		if (time_to_send <= now && q->slot.slot_next <= now) {
 			netem_erase_head(q, skb);
-			sch->q.qlen--;
-			qdisc_qstats_backlog_dec(sch, skb);
+			q->t_len--;
 			skb->next = NULL;
 			skb->prev = NULL;
 			/* skb->dev shares skb->rbnode area,

@@ -737,16 +740,21 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch)
 					if (net_xmit_drop_count(err))
 						qdisc_qstats_drop(sch);
 					qdisc_tree_reduce_backlog(sch, 1, pkt_len);
+					sch->qstats.backlog -= pkt_len;
+					sch->q.qlen--;
 				}
 				goto tfifo_dequeue;
 			}
+			sch->q.qlen--;
 			goto deliver;
 		}
 
 		if (q->qdisc) {
 			skb = q->qdisc->ops->dequeue(q->qdisc);
-			if (skb)
+			if (skb) {
+				sch->q.qlen--;
 				goto deliver;
+			}
 		}
 
 		qdisc_watchdog_schedule_ns(&q->watchdog,

@@ -756,8 +764,10 @@ static struct sk_buff *netem_dequeue(struct Qdisc *sch)
 
 	if (q->qdisc) {
 		skb = q->qdisc->ops->dequeue(q->qdisc);
-		if (skb)
+		if (skb) {
+			sch->q.qlen--;
 			goto deliver;
+		}
 	}
 	return NULL;
 }

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Daniel Borkmann <[email protected]>

[ Upstream commit 77b11c8bf3a228d1c63464534c2dcc8d9c8bf7ff ]

Drivers like mlx5 expose NIC's vlan_features such as NETIF_F_GSO_UDP_TUNNEL & NETIF_F_GSO_UDP_TUNNEL_CSUM which are later not propagated when the underlying devices are bonded and a vlan device created on top of the bond.

Right now, the more cumbersome workaround for this is to create the vlan on top of the mlx5 and then enslave the vlan devices to a bond.

To fix this, add NETIF_F_GSO_ENCAP_ALL to BOND_VLAN_FEATURES such that bond_compute_features() can probe and propagate the vlan_features from the slave devices up to the vlan device.

Given the following bond:

  # ethtool -i enp2s0f{0,1}np{0,1}
  driver: mlx5_core
  [...]

  # ethtool -k enp2s0f0np0 | grep udp
  tx-udp_tnl-segmentation: on
  tx-udp_tnl-csum-segmentation: on
  tx-udp-segmentation: on
  rx-udp_tunnel-port-offload: on
  rx-udp-gro-forwarding: off

  # ethtool -k enp2s0f1np1 | grep udp
  tx-udp_tnl-segmentation: on
  tx-udp_tnl-csum-segmentation: on
  tx-udp-segmentation: on
  rx-udp_tunnel-port-offload: on
  rx-udp-gro-forwarding: off

  # ethtool -k bond0 | grep udp
  tx-udp_tnl-segmentation: on
  tx-udp_tnl-csum-segmentation: on
  tx-udp-segmentation: on
  rx-udp_tunnel-port-offload: off [fixed]
  rx-udp-gro-forwarding: off

Before:

  # ethtool -k bond0.100 | grep udp
  tx-udp_tnl-segmentation: off [requested on]
  tx-udp_tnl-csum-segmentation: off [requested on]
  tx-udp-segmentation: on
  rx-udp_tunnel-port-offload: off [fixed]
  rx-udp-gro-forwarding: off

After:

  # ethtool -k bond0.100 | grep udp
  tx-udp_tnl-segmentation: on
  tx-udp_tnl-csum-segmentation: on
  tx-udp-segmentation: on
  rx-udp_tunnel-port-offload: off [fixed]
  rx-udp-gro-forwarding: off

Various users have run into this reporting performance issues when configuring Cilium in vxlan tunneling mode and having the combination of bond & vlan for the core devices connecting the Kubernetes cluster to the outside world.

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

index 67d153cc6a6c..75499e2967e8 100644
@@ -1384,6 +1384,7 @@ static netdev_features_t bond_fix_features(struct net_device *dev,
 
 #define BOND_VLAN_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
 				 NETIF_F_FRAGLIST | NETIF_F_GSO_SOFTWARE | \
+				 NETIF_F_GSO_ENCAP_ALL | \
 				 NETIF_F_HIGHDMA | NETIF_F_LRO)
 
 #define BOND_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Daniel Borkmann <[email protected]>

[ Upstream commit 98712844589e06d9aa305b5077169942139fd75c ]

Similar to bonding driver, add NETIF_F_GSO_ENCAP_ALL to TEAM_VLAN_FEATURES in order to support slave devices which propagate NETIF_F_GSO_UDP_TUNNEL & NETIF_F_GSO_UDP_TUNNEL_CSUM as vlan_features.

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

index 8edfc8984e2c..5e5af71a85ac 100644
@@ -979,7 +979,8 @@ static void team_port_disable(struct team *team,
 
 #define TEAM_VLAN_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \
 			    NETIF_F_FRAGLIST | NETIF_F_GSO_SOFTWARE | \
-			    NETIF_F_HIGHDMA | NETIF_F_LRO)
+			    NETIF_F_HIGHDMA | NETIF_F_LRO | \
+			    NETIF_F_GSO_ENCAP_ALL)
 
 #define TEAM_ENC_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_SG | \
 				 NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE)

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Daniil Tatianin <[email protected]>

[ Upstream commit c53d96a4481f42a1635b96d2c1acbb0a126bfd54 ]

This bug was first introduced in c27f3d011b08, where the author of the patch probably meant to do DeleteMutex instead of ReleaseMutex. The mutex leak was noticed later on and fixed in e4dfe108371, but the bogus MutexRelease line was never removed, so do it now.

---
 drivers/acpi/acpica/evxfregn.c | 2 --
 1 file changed, 2 deletions(-)

index b1ff0a8f9c14..7672d70da850 100644
@@ -201,8 +201,6 @@ acpi_remove_address_space_handler(acpi_handle device,
 
 			/* Now we can delete the handler object */
 
-			acpi_os_release_mutex(handler_obj->address_space.
-					      context_mutex);
 			acpi_ut_remove_reference(handler_obj);
 			goto unlock_and_exit;
 		}

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Nathan Chancellor <[email protected]>

[ Upstream commit 57e420c84f9ab55ba4c5e2ae9c5f6c8e1ea834d2 ]

After a recent change to clamp() and its variants [1] that increases the coverage of the check that high is greater than low because it can be done through inlining, certain build configurations (such as s390 defconfig) fail to build with clang with:

  block/blk-iocost.c:1101:11: error: call to '__compiletime_assert_557' declared with 'error' attribute: clamp() low limit 1 greater than high limit active
   1101 |                 inuse = clamp_t(u32, inuse, 1, active);
        |                         ^
  include/linux/minmax.h:218:36: note: expanded from macro 'clamp_t'
    218 | #define clamp_t(type, val, lo, hi) __careful_clamp(type, val, lo, hi)
        |                                    ^
  include/linux/minmax.h:195:2: note: expanded from macro '__careful_clamp'
    195 |         __clamp_once(type, val, lo, hi, __UNIQUE_ID(v_), __UNIQUE_ID(l_), __UNIQUE_ID(h_))
        |         ^
  include/linux/minmax.h:188:2: note: expanded from macro '__clamp_once'
    188 |         BUILD_BUG_ON_MSG(statically_true(ulo > uhi),                            \
        |         ^

__propagate_weights() is called with an active value of zero in ioc_check_iocgs(), which results in the high value being less than the low value, which is undefined because the value returned depends on the order of the comparisons.

The purpose of this expression is to ensure inuse is not more than active and at least 1. This could be written more simply with a ternary expression that uses min(inuse, active) as the condition so that the value of that condition can be used if it is not zero and one if it is. Do this conversion to resolve the error and add a comment to deter people from turning this back into clamp().

---
 block/blk-iocost.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

index 9654d1c2c20f..ba23562abc80 100644
@@ -1087,7 +1087,14 @@ static void __propagate_weights(struct ioc_gq *iocg, u32 active, u32 inuse,
 		inuse = DIV64_U64_ROUND_UP(active * iocg->child_inuse_sum,
 					   iocg->child_active_sum);
 	} else {
-		inuse = clamp_t(u32, inuse, 1, active);
+		/*
+		 * It may be tempting to turn this into a clamp expression with
+		 * a lower limit of 1 but active may be 0, which cannot be used
+		 * as an upper limit in that situation. This expression allows
+		 * active to clamp inuse unless it is 0, in which case inuse
+		 * becomes 1.
+		 */
+		inuse = min(inuse, active) ?: 1;
 	}
 
 	iocg->last_inuse = iocg->inuse;

2.39.5

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Eduard Zingerman <[email protected]>

commit e9bd9c498cb0f5843996dbe5cbce7a1836a83c70 upstream.

Range propagation must not affect subreg_def marks, otherwise the following example is rewritten by verifier incorrectly when BPF_F_TEST_RND_HI32 flag is set:

  0: call bpf_ktime_get_ns                   call bpf_ktime_get_ns
  1: r0 &= 0x7fffffff       after verifier   r0 &= 0x7fffffff
  2: w1 = w0                rewrites         w1 = w0
  3: if w0 < 10 goto +0     -------------->  r11 = 0x2f5674a6     (r)
  4: r1 >>= 32                               r11 <<= 32           (r)
  5: r0 = r1                                 r1 |= r11            (r)
  6: exit;                                   if w0 < 0xa goto pc+0
                                             r1 >>= 32
                                             r0 = r1
                                             exit

(or zero extension of w1 at (2) is missing for architectures that
require zero extension for upper register half).

The following happens w/o this patch:
- r0 is marked as not a subreg at (0);
- w1 is marked as subreg at (2);
- w1 subreg_def is overridden at (3) by copy_register_state();
- w1 is read at (5) but mark_insn_zext() does not mark (2) for zero extension, because w1 subreg_def is not set;
- because of BPF_F_TEST_RND_HI32 flag verifier inserts random value for hi32 bits of (2) (marked (r));
- this random value is read at (5).

[ shung-hsi.yu: sync_linked_regs() was called find_equal_scalars() before commit
4bf79f9be434 ("bpf: Track equal scalars history on per-instruction level"), and modification is done because there is only a single call to copy_register_state() before commit 98d7ca374ba4 ("bpf: Track delta between "linked" registers."). ]

---
 kernel/bpf/verifier.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -9223,8 +9223,11 @@ static void find_equal_scalars(struct bp
 	struct bpf_reg_state *reg;
 
 	bpf_for_each_reg_in_vstate(vstate, state, reg, ({
-		if (reg->type == SCALAR_VALUE && reg->id == known_reg->id)
+		if (reg->type == SCALAR_VALUE && reg->id == known_reg->id) {
+			s32 saved_subreg_def = reg->subreg_def;
 			copy_register_state(reg, known_reg);
+			reg->subreg_def = saved_subreg_def;
+		}
 	}));
 }
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Nikolay Kuratov <[email protected]>

commit b022f0c7e404 ("tracing/kprobes: Return EADDRNOTAVAIL when func matches several symbols") avoids checking number_of_same_symbols() for module symbol in __trace_kprobe_create(), but create_local_trace_kprobe() should avoid this
check too. Doing this check leads to ENOENT for module_name:symbol_name constructions passed over perf_event_open.

No bug in newer kernels as it was fixed more generally by
commit 9d8616034f16 ("tracing/kprobes: Add symbol counting check when module loads")

---
v1 -> v2:
* Reword commit title and message
* Send for stable instead of mainline

 kernel/trace/trace_kprobe.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1823,7 +1823,7 @@ create_local_trace_kprobe(char *func, vo
 	int ret;
 	char *event;
 
-	if (func) {
+	if (func && !strchr(func, ':')) {
 		unsigned int count;
 
 		count = number_of_same_symbols(func);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Greg Kroah-Hartman <[email protected]>

This reverts commit dadac97f066a67334268132c1e2d0fd599fbcbec which is commit 7ae04ba36b381bffe2471eff3a93edced843240f upstream.

It is reported to cause build failures.

---
 arch/parisc/Kconfig             |    1 -
 arch/parisc/include/asm/cache.h |   11 +----------
 2 files changed, 1 insertion(+), 11 deletions(-)

@@ -15,7 +15,6 @@ config PARISC
 	select ARCH_SPLIT_ARG64 if !64BIT
 	select ARCH_SUPPORTS_HUGETLBFS if PA20
 	select ARCH_SUPPORTS_MEMORY_FAILURE
-	select ARCH_HAS_CACHE_LINE_SIZE
 	select DMA_OPS
 	select RTC_CLASS
 	select RTC_DRV_GENERIC

@@ -20,16 +20,7 @@
 
 #define SMP_CACHE_BYTES L1_CACHE_BYTES
 
-#ifdef CONFIG_PA20
-#define ARCH_DMA_MINALIGN	128
-#else
-#define ARCH_DMA_MINALIGN	32
-#endif
-#define ARCH_KMALLOC_MINALIGN	16	/* ldcw requires 16-byte alignment */
-
-#define arch_slab_minalign()	((unsigned)dcache_stride)
-#define cache_line_size()	dcache_stride
-#define dma_get_cache_alignment cache_line_size
+#define ARCH_DMA_MINALIGN	L1_CACHE_BYTES
 
 #define __read_mostly __section(".data..read_mostly")
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit f9244fb55f37356f75c739c57323d9422d7aa0f8 upstream.

When removing a netfront device directly after a suspend/resume cycle it might happen that the queues have not been setup again, causing a crash during the attempt to stop the queues another time.

Fix that by checking the queues are existing before trying to stop them.

This is XSA-465 / CVE-2024-53240.

---
 drivers/net/xen-netfront.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -869,7 +869,7 @@ static netdev_tx_t xennet_start_xmit(str
 static int xennet_close(struct net_device *dev)
 {
 	struct netfront_info *np = netdev_priv(dev);
-	unsigned int num_queues = dev->real_num_tx_queues;
+	unsigned int num_queues = np->queues ? dev->real_num_tx_queues : 0;
 	unsigned int i;
 	struct netfront_queue *queue;
 	netif_tx_stop_all_queues(np->netdev);

@@ -884,6 +884,9 @@ static void xennet_destroy_queues(struct
 {
 	unsigned int i;
 
+	if (!info->queues)
+		return;
+
 	for (i = 0; i < info->netdev->real_num_tx_queues; i++) {
 		struct netfront_queue *queue = &info->queues[i];
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit efbcd61d9bebb771c836a3b8bfced8165633db7c upstream.

In order to be able to differentiate between AMD and Intel based systems for very early hypercalls without having to rely on the Xen hypercall page, make get_cpu_vendor() non-static.

Refactor early_cpu_init() for the same reason by splitting out the loop initializing cpu_devs() into an externally callable function.

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/include/asm/processor.h |    2 ++
 arch/x86/kernel/cpu/common.c     |   36 +++++++++++++++++++++---------------
 2 files changed, 23 insertions(+), 15 deletions(-)

@@ -196,6 +196,8 @@ static inline unsigned long long l1tf_pf
 	return BIT_ULL(boot_cpu_data.x86_cache_bits - 1 - PAGE_SHIFT);
 }
 
+void init_cpu_devs(void);
+void get_cpu_vendor(struct cpuinfo_x86 *c);
 extern void early_cpu_init(void);
 extern void identify_boot_cpu(void);
 extern void identify_secondary_cpu(struct cpuinfo_x86 *);

@@ -798,7 +798,7 @@ void detect_ht(struct cpuinfo_x86 *c)
 #endif
 }
 
-static void get_cpu_vendor(struct cpuinfo_x86 *c)
+void get_cpu_vendor(struct cpuinfo_x86 *c)
 {
 	char *v = c->x86_vendor_id;
 	int i;

@@ -1522,15 +1522,11 @@ static void __init early_identify_cpu(st
 	detect_nopl();
 }
 
-void __init early_cpu_init(void)
+void __init init_cpu_devs(void)
 {
 	const struct cpu_dev *const *cdev;
 	int count = 0;
 
-#ifdef CONFIG_PROCESSOR_SELECT
-	pr_info("KERNEL supported cpus:\n");
-#endif
-
 	for (cdev = __x86_cpu_dev_start; cdev < __x86_cpu_dev_end; cdev++) {
 		const struct cpu_dev *cpudev = *cdev;
 

@@ -1538,20 +1534,30 @@ void __init early_cpu_init(void)
 			break;
 		cpu_devs[count] = cpudev;
 		count++;
+	}
+}
 
+void __init early_cpu_init(void)
+{
 #ifdef CONFIG_PROCESSOR_SELECT
-		{
-			unsigned int j;
+	unsigned int i, j;
 
-			for (j = 0; j < 2; j++) {
-				if (!cpudev->c_ident[j])
-					continue;
-				pr_info("  %s %s\n", cpudev->c_vendor,
-					cpudev->c_ident[j]);
-			}
-		}
+	pr_info("KERNEL supported cpus:\n");
 #endif
+
+	init_cpu_devs();
+
+#ifdef CONFIG_PROCESSOR_SELECT
+	for (i = 0; i < X86_VENDOR_NUM && cpu_devs[i]; i++) {
+		for (j = 0; j < 2; j++) {
+			if (!cpu_devs[i]->c_ident[j])
+				continue;
+			pr_info("  %s %s\n", cpu_devs[i]->c_vendor,
+				cpu_devs[i]->c_ident[j]);
+		}
 	}
+#endif
+
 	early_identify_cpu(&boot_cpu_data);
 }
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit dda014ba59331dee4f3b773a020e109932f4bd24 upstream.

The syscall instruction is used in Xen PV mode for doing hypercalls. Allow syscall to be used in the kernel in case it is tagged with an unwind hint for objtool.

This is part of XSA-466 / CVE-2024-53241.

---
 tools/objtool/check.c |   11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

@@ -3206,10 +3206,13 @@ static int validate_branch(struct objtoo
 			break;
 
 		case INSN_CONTEXT_SWITCH:
-			if (func && (!next_insn || !next_insn->hint)) {
-				WARN_FUNC("unsupported instruction in callable function",
-					  sec, insn->offset);
-				return 1;
+			if (func) {
+				if (!next_insn || !next_insn->hint) {
+					WARN_FUNC("unsupported instruction in callable function",
+						  sec, insn->offset);
+					return 1;
+				}
+				break;
 			}
 			return 0;
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit 0ef8047b737d7480a5d4c46d956e97c190f13050 upstream.

Add static_call_update_early() for updating static-call targets in very early boot.

This will be needed for support of Xen guest type specific hypercall functions.

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/include/asm/static_call.h |   15 +++++++++++++++
 arch/x86/include/asm/sync_core.h   |    6 +++---
 arch/x86/kernel/static_call.c      |   10 ++++++++++
 include/linux/compiler.h           |   32 +++++++++++++++++++++++---------
 include/linux/static_call.h        |    1 +
 kernel/static_call_inline.c        |    2 +-
 6 files changed, 53 insertions(+), 13 deletions(-)

@@ -61,4 +61,19 @@
 
 extern bool __static_call_fixup(void *tramp, u8 op, void *dest);
 
+extern void __static_call_update_early(void *tramp, void *func);
+
+#define static_call_update_early(name, _func)				\
+({									\
+	typeof(&STATIC_CALL_TRAMP(name)) __F = (_func);			\
+	if (static_call_initialized) {					\
+		__static_call_update(&STATIC_CALL_KEY(name),		\
+				     STATIC_CALL_TRAMP_ADDR(name), __F);\
+	} else {							\
+		WRITE_ONCE(STATIC_CALL_KEY(name).func, _func);		\
+		__static_call_update_early(STATIC_CALL_TRAMP_ADDR(name),\
+					   __F);			\
+	}								\
+})
+
 #endif /* _ASM_STATIC_CALL_H */

@@ -8,7 +8,7 @@
 #include <asm/special_insns.h>
 
 #ifdef CONFIG_X86_32
-static inline void iret_to_self(void)
+static __always_inline void iret_to_self(void)
 {
 	asm volatile (
 		"pushfl\n\t"

@@ -19,7 +19,7 @@ static inline void iret_to_self(void)
 		: ASM_CALL_CONSTRAINT : : "memory");
 }
 #else
-static inline void iret_to_self(void)
+static __always_inline void iret_to_self(void)
 {
 	unsigned int tmp;
 

@@ -55,7 +55,7 @@ static inline void iret_to_self(void)
  * Like all of Linux's memory ordering operations, this is a
  * compiler barrier as well.
  */
-static inline void sync_core(void)
+static __always_inline void sync_core(void)
 {
 	/*
 	 * The SERIALIZE instruction is the most straightforward way to

@@ -2,6 +2,7 @@
 #include <linux/static_call.h>
 #include <linux/memory.h>
 #include <linux/bug.h>
+#include <asm/sync_core.h>
 #include <asm/text-patching.h>
 
 enum insn_type {

@@ -165,6 +166,15 @@ void arch_static_call_transform(void *si
 }
 EXPORT_SYMBOL_GPL(arch_static_call_transform);
 
+noinstr void __static_call_update_early(void *tramp, void *func)
+{
+	BUG_ON(system_state != SYSTEM_BOOTING);
+	BUG_ON(!early_boot_irqs_disabled);
+	BUG_ON(static_call_initialized);
+	__text_gen_insn(tramp, JMP32_INSN_OPCODE, tramp, func, JMP32_INSN_SIZE);
+	sync_core();
+}
+
 #ifdef CONFIG_RETHUNK
 /*
  * This is called by apply_returns() to fix up static call trampolines,

@@ -222,6 +222,23 @@ void ftrace_likely_update(struct ftrace_
 
 #endif /* __KERNEL__ */
 
+/**
+ * offset_to_ptr - convert a relative memory offset to an absolute pointer
+ * @off:	the address of the 32-bit offset value
+ */
+static inline void *offset_to_ptr(const int *off)
+{
+	return (void *)((unsigned long)off + *off);
+}
+
+#endif /* __ASSEMBLY__ */
+
+#ifdef CONFIG_64BIT
+#define ARCH_SEL(a,b) a
+#else
+#define ARCH_SEL(a,b) b
+#endif
+
 /*
  * Force the compiler to emit 'sym' as a symbol, so that we can reference
  * it from inline assembler. Necessary in case 'sym' could be inlined

@@ -232,16 +249,13 @@ void ftrace_likely_update(struct ftrace_
 	static void * __section(".discard.addressable") __used \
 		__UNIQUE_ID(__PASTE(__addressable_,sym)) = (void *)&sym;
 
-/**
- * offset_to_ptr - convert a relative memory offset to an absolute pointer
- * @off:	the address of the 32-bit offset value
- */
-static inline void *offset_to_ptr(const int *off)
-{
-	return (void *)((unsigned long)off + *off);
-}
+#define __ADDRESSABLE_ASM(sym)						\
+	.pushsection .discard.addressable,"aw";				\
+	.align ARCH_SEL(8,4);						\
+	ARCH_SEL(.quad, .long) __stringify(sym);			\
+	.popsection;
 
-#endif /* __ASSEMBLY__ */
+#define __ADDRESSABLE_ASM_STR(sym) __stringify(__ADDRESSABLE_ASM(sym))
 
 /* &a[0] degrades to a pointer: a different type from an array */
 #define __must_be_array(a)	BUILD_BUG_ON_ZERO(__same_type((a), &(a)[0]))

@@ -138,6 +138,7 @@
 #ifdef CONFIG_HAVE_STATIC_CALL
 #include <asm/static_call.h>
 
+extern bool static_call_initialized;
 /*
  * Either @site or @tramp can be NULL.
  */

@@ -15,7 +15,7 @@ extern struct static_call_site __start_s
 extern struct static_call_tramp_key __start_static_call_tramp_key[],
 				    __stop_static_call_tramp_key[];
 
-static bool static_call_initialized;
+bool static_call_initialized;
 
 /* mutex to protect key modules/sites */
 static DEFINE_MUTEX(static_call_mutex);

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit a2796dff62d6c6bfc5fbebdf2bee0d5ac0438906 upstream.

Instead of jumping to the Xen hypercall page for doing the iret hypercall, directly code the required sequence in xen-asm.S.

This is done in preparation of no longer using hypercall page at all, as it has shown to cause problems with speculation mitigations.

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/xen/xen-asm.S |   27 ++++++++++++++++++---------
 1 file changed, 18 insertions(+), 9 deletions(-)

@@ -170,7 +170,6 @@ SYM_CODE_START(xen_early_idt_handler_arr
 SYM_CODE_END(xen_early_idt_handler_array)
 	__FINIT
 
-hypercall_iret = hypercall_page + __HYPERVISOR_iret * 32
 /*
  * Xen64 iret frame:
  *

@@ -180,16 +179,27 @@ hypercall_iret = hypercall_page + __HYPE
  *	cs
  *	rip		<-- standard iret frame
  *
- *	flags
+ *	flags		<-- xen_iret must push from here on
  *
- *	rcx		}
- *	r11		}<-- pushed by hypercall page
- * rsp->rax		}
+ *	rcx
+ *	r11
+ * rsp->rax
  */
+.macro xen_hypercall_iret
+	pushq $0	/* Flags */
+	push %rcx
+	push %r11
+	push %rax
+	mov  $__HYPERVISOR_iret, %eax
+	syscall		/* Do the IRET. */
+#ifdef CONFIG_MITIGATION_SLS
+	int3
+#endif
+.endm
+
 SYM_CODE_START(xen_iret)
 	UNWIND_HINT_EMPTY
-	pushq $0
-	jmp hypercall_iret
+	xen_hypercall_iret
 SYM_CODE_END(xen_iret)
 
 /*

@@ -290,8 +300,7 @@ SYM_CODE_START(xen_entry_SYSENTER_compat
 	UNWIND_HINT_ENTRY
 	lea 16(%rsp), %rsp	/* strip %rcx, %r11 */
 	mov $-ENOSYS, %rax
-	pushq $0
-	jmp hypercall_iret
+	xen_hypercall_iret
 SYM_CODE_END(xen_entry_SYSENTER_compat)
 SYM_CODE_END(xen_entry_SYSCALL_compat)
 

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit b4845bb6383821a9516ce30af3a27dc873e37fd4 upstream.

Add generic hypercall functions usable for all normal (i.e. not iret) hypercalls. Depending on the guest type and the processor vendor different functions need to be used due to the to be used instruction for entering the hypervisor:

- PV guests need to use syscall
- HVM/PVH guests on Intel need to use vmcall
- HVM/PVH guests on AMD and Hygon need to use vmmcall

As PVH guests need to issue hypercalls very early during boot, there is a 4th hypercall function needed for HVM/PVH which can be used on Intel and AMD processors. It will check the vendor type and then set the Intel or AMD specific function to use via static_call().

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/include/asm/xen/hypercall.h |    3 +
 arch/x86/xen/enlighten.c             |   65 +++++++++++++++++++++++++++
 arch/x86/xen/enlighten_hvm.c         |    4 +
 arch/x86/xen/enlighten_pv.c          |    4 +
 arch/x86/xen/xen-asm.S               |   22 +++++++++
 arch/x86/xen/xen-head.S              |   82 +++++++++++++++++++++++++++++++++++
 arch/x86/xen/xen-ops.h               |    9 +++
 7 files changed, 188 insertions(+), 1 deletion(-)

@@ -88,6 +88,9 @@ struct xen_dm_op_buf;
 
 extern struct { char _entry[32]; } hypercall_page[];
 
+void xen_hypercall_func(void);
+DECLARE_STATIC_CALL(xen_hypercall, xen_hypercall_func);
+
 #define __HYPERCALL		"call hypercall_page+%c[offset]"
 #define __HYPERCALL_ENTRY(x)						\
 	[offset] "i" (__HYPERVISOR_##x * sizeof(hypercall_page[0]))

@@ -284,6 +284,10 @@ static uint32_t __init xen_platform_hvm(
 	if (xen_pv_domain())
 		return 0;
 
+	/* Set correct hypercall function. */
+	if (xen_domain)
+		xen_hypercall_setfunc();
+
 	if (xen_pvh_domain() && nopv) {
 		/* Guest booting via the Xen-PVH boot entry goes here */
 		pr_info("\"nopv\" parameter is ignored in PVH guest\n");

@@ -1207,6 +1207,9 @@ asmlinkage __visible void __init xen_sta
 
 	xen_domain_type = XEN_PV_DOMAIN;
 	xen_start_flags = xen_start_info->flags;
+	/* Interrupts are guaranteed to be off initially. */
+	early_boot_irqs_disabled = true;
+	static_call_update_early(xen_hypercall, xen_hypercall_pv);
 
 	xen_setup_features();
 

@@ -1304,7 +1307,6 @@ asmlinkage __visible void __init xen_sta
 	WARN_ON(xen_cpuhp_setup(xen_cpu_up_prepare_pv, xen_cpu_dead_pv));
 
 	local_irq_disable();
-	early_boot_irqs_disabled = true;
 
 	xen_raw_console_write("mapping kernel into physical memory\n");
 	xen_setup_kernel_pagetable((pgd_t *)xen_start_info->pt_base,

@@ -20,9 +20,31 @@
 
 #include <linux/init.h>
 #include <linux/linkage.h>
+#include <linux/objtool.h>
 #include <../entry/calling.h>
 
 /*
+ * PV hypercall interface to the hypervisor.
+ *
+ * Called via inline asm(), so better preserve %rcx and %r11.
+ *
+ * Input:
+ *	%eax: hypercall number
+ *	%rdi, %rsi, %rdx, %r10, %r8: args 1..5 for the hypercall
+ * Output: %rax
+ */
+SYM_FUNC_START(xen_hypercall_pv)
+	push %rcx
+	push %r11
+	UNWIND_HINT_SAVE
+	syscall
+	UNWIND_HINT_RESTORE
+	pop %r11
+	pop %rcx
+	RET
+SYM_FUNC_END(xen_hypercall_pv)
+
+/*
  * Enable events.  This clears the event mask and tests the pending
  * event status with one and operation.  If there are pending events,
  * then enter the hypervisor to get them handled.

@@ -164,4 +164,13 @@ void xen_hvm_post_suspend(int suspend_ca
 static inline void xen_hvm_post_suspend(int suspend_cancelled) {}
 #endif
 
+#ifdef CONFIG_XEN_PV
+void xen_hypercall_pv(void);
+#endif
+void xen_hypercall_hvm(void);
+void xen_hypercall_amd(void);
+void xen_hypercall_intel(void);
+void xen_hypercall_setfunc(void);
+void *__xen_hypercall_setfunc(void);
+
 #endif /* XEN_OPS_H */

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit b1c2cb86f4a7861480ad54bb9a58df3cbebf8e92 upstream.

Call the Xen hypervisor via the new xen_hypercall_func static-call instead of the hypercall page.

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/include/asm/xen/hypercall.h |   33 ++++++++++++++++++++-------------
 1 file changed, 20 insertions(+), 13 deletions(-)

@@ -39,9 +39,11 @@
 #include <linux/string.h>
 #include <linux/types.h>
 #include <linux/pgtable.h>
+#include <linux/instrumentation.h>
 
 #include <trace/events/xen.h>
 
+#include <asm/alternative.h>
 #include <asm/page.h>
 #include <asm/smap.h>
 #include <asm/nospec-branch.h>

@@ -91,9 +93,17 @@ extern struct { char _entry[32]; } hyper
 void xen_hypercall_func(void);
 DECLARE_STATIC_CALL(xen_hypercall, xen_hypercall_func);
 
-#define __HYPERCALL		"call hypercall_page+%c[offset]"
-#define __HYPERCALL_ENTRY(x)						\
-	[offset] "i" (__HYPERVISOR_##x * sizeof(hypercall_page[0]))
+#ifdef MODULE
+#define __ADDRESSABLE_xen_hypercall
+#else
+#define __ADDRESSABLE_xen_hypercall __ADDRESSABLE_ASM_STR(__SCK__xen_hypercall)
+#endif
+
+#define __HYPERCALL					\
+	__ADDRESSABLE_xen_hypercall			\
+	"call __SCT__xen_hypercall"
+
+#define __HYPERCALL_ENTRY(x)	"a" (x)
 
 #ifdef CONFIG_X86_32
 #define __HYPERCALL_RETREG	"eax"

@@ -151,7 +161,7 @@ DECLARE_STATIC_CALL(xen_hypercall, xen_h
 	__HYPERCALL_0ARG();						\
 	asm volatile (__HYPERCALL					\
 		      : __HYPERCALL_0PARAM				\
-		      : __HYPERCALL_ENTRY(name)				\
+		      : __HYPERCALL_ENTRY(__HYPERVISOR_ ## name)	\
 		      : __HYPERCALL_CLOBBER0);				\
 	(type)__res;							\
 })

@@ -162,7 +172,7 @@ DECLARE_STATIC_CALL(xen_hypercall, xen_h
 	__HYPERCALL_1ARG(a1);						\
 	asm volatile (__HYPERCALL					\
 		      : __HYPERCALL_1PARAM				\
-		      : __HYPERCALL_ENTRY(name)				\
+		      : __HYPERCALL_ENTRY(__HYPERVISOR_ ## name)	\
 		      : __HYPERCALL_CLOBBER1);				\
 	(type)__res;							\
 })

@@ -173,7 +183,7 @@ DECLARE_STATIC_CALL(xen_hypercall, xen_h
 	__HYPERCALL_2ARG(a1, a2);					\
 	asm volatile (__HYPERCALL					\
 		      : __HYPERCALL_2PARAM				\
-		      : __HYPERCALL_ENTRY(name)				\
+		      : __HYPERCALL_ENTRY(__HYPERVISOR_ ## name)	\
 		      : __HYPERCALL_CLOBBER2);				\
 	(type)__res;							\
 })

@@ -184,7 +194,7 @@ DECLARE_STATIC_CALL(xen_hypercall, xen_h
 	__HYPERCALL_3ARG(a1, a2, a3);					\
 	asm volatile (__HYPERCALL					\
 		      : __HYPERCALL_3PARAM				\
-		      : __HYPERCALL_ENTRY(name)				\
+		      : __HYPERCALL_ENTRY(__HYPERVISOR_ ## name)	\
 		      : __HYPERCALL_CLOBBER3);				\
 	(type)__res;							\
 })

@@ -195,7 +205,7 @@ DECLARE_STATIC_CALL(xen_hypercall, xen_h
 	__HYPERCALL_4ARG(a1, a2, a3, a4);				\
 	asm volatile (__HYPERCALL					\
 		      : __HYPERCALL_4PARAM				\
-		      : __HYPERCALL_ENTRY(name)				\
+		      : __HYPERCALL_ENTRY(__HYPERVISOR_ ## name)	\
 		      : __HYPERCALL_CLOBBER4);				\
 	(type)__res;							\
 })

@@ -209,12 +219,9 @@ xen_single_call(unsigned int call,
 	__HYPERCALL_DECLS;
 	__HYPERCALL_5ARG(a1, a2, a3, a4, a5);
 
-	if (call >= PAGE_SIZE / sizeof(hypercall_page[0]))
-		return -EINVAL;
-
-	asm volatile(CALL_NOSPEC
+	asm volatile(__HYPERCALL
 		     : __HYPERCALL_5PARAM
-		     : [thunk_target] "a" (&hypercall_page[call])
+		     : __HYPERCALL_ENTRY(call)
 		     : __HYPERCALL_CLOBBER5);
 
 	return (long)__res;

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Juergen Gross <[email protected]>

commit 7fa0da5373685e7ed249af3fa317ab1e1ba8b0a6 upstream.

The hypercall page is no longer needed. It can be removed, as from the Xen perspective it is optional.

But, from Linux's perspective, it removes naked RET instructions that escape the speculative protections that Call Depth Tracking and/or Untrain Ret are trying to achieve.

This is part of XSA-466 / CVE-2024-53241.

---
 arch/x86/include/asm/xen/hypercall.h |    2 --
 arch/x86/xen/enlighten.c             |    2 --
 arch/x86/xen/enlighten_hvm.c         |    9 +--------
 arch/x86/xen/enlighten_pvh.c         |    7 -------
 arch/x86/xen/xen-head.S              |   19 -------------------
 5 files changed, 1 insertion(+), 38 deletions(-)

@@ -88,8 +88,6 @@ struct xen_dm_op_buf;
  * there aren't more than 5 arguments...)
  */
 
-extern struct { char _entry[32]; } hypercall_page[];
-
 void xen_hypercall_func(void);
 DECLARE_STATIC_CALL(xen_hypercall, xen_hypercall_func);
 

@@ -26,8 +26,6 @@
 #include "smp.h"
 #include "pmu.h"
 
-EXPORT_SYMBOL_GPL(hypercall_page);
-
 DEFINE_STATIC_CALL(xen_hypercall, xen_hypercall_hvm);
 EXPORT_STATIC_CALL_TRAMP(xen_hypercall);
 

@@ -101,15 +101,8 @@ static void __init init_hvm_pv_info(void
 	/* PVH set up hypercall page in xen_prepare_pvh(). */
 	if (xen_pvh_domain())
 		pv_info.name = "Xen PVH";
-	else {
-		u64 pfn;
-		uint32_t msr;
-
+	else
 		pv_info.name = "Xen HVM";
-		msr = cpuid_ebx(base + 2);
-		pfn = __pa(hypercall_page);
-		wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));
-	}
 
 	xen_setup_features();
 

@@ -27,17 +27,10 @@ EXPORT_SYMBOL_GPL(xen_pvh);
 
 void __init xen_pvh_init(struct boot_params *boot_params)
 {
-	u32 msr;
-	u64 pfn;
-
 	xen_pvh = 1;
 	xen_domain_type = XEN_HVM_DOMAIN;
 	xen_start_flags = pvh_start_info.flags;
 
-	msr = cpuid_ebx(xen_cpuid_base() + 2);
-	pfn = __pa(hypercall_page);
-	wrmsr_safe(msr, (u32)pfn, (u32)(pfn >> 32));
-
 	if (xen_initial_domain())
 		x86_init.oem.arch_setup = xen_add_preferred_consoles;
 	x86_init.oem.banner = xen_banner;

@@ -146,24 +146,6 @@ SYM_FUNC_START(xen_hypercall_intel)
 SYM_FUNC_END(xen_hypercall_intel)
 	.popsection
 
-.pushsection .text
-	.balign PAGE_SIZE
-SYM_CODE_START(hypercall_page)
-	.rept (PAGE_SIZE / 32)
-		UNWIND_HINT_FUNC
-		ANNOTATE_UNRET_SAFE
-		ret
-		.skip 31, 0xcc
-	.endr
-
-#define HYPERCALL(n) \
-	.equ xen_hypercall_##n, hypercall_page + __HYPERVISOR_##n * 32; \
-	.type xen_hypercall_##n, @function; .size xen_hypercall_##n, 32
-#include <asm/xen-hypercalls.h>
-#undef HYPERCALL
-SYM_CODE_END(hypercall_page)
-.popsection
-
 	ELFNOTE(Xen, XEN_ELFNOTE_GUEST_OS,       .asciz "linux")
 	ELFNOTE(Xen, XEN_ELFNOTE_GUEST_VERSION,  .asciz "2.6")
 	ELFNOTE(Xen, XEN_ELFNOTE_XEN_VERSION,    .asciz "xen-3.0")

@@ -177,7 +159,6 @@ SYM_CODE_END(hypercall_page)
 #ifdef CONFIG_XEN_PV
 	ELFNOTE(Xen, XEN_ELFNOTE_ENTRY,          _ASM_PTR startup_xen)
 #endif
-	ELFNOTE(Xen, XEN_ELFNOTE_HYPERCALL_PAGE, _ASM_PTR hypercall_page)
 	ELFNOTE(Xen, XEN_ELFNOTE_FEATURES,
 		.ascii "!writable_page_tables|pae_pgdir_above_4gb")
 	ELFNOTE(Xen, XEN_ELFNOTE_SUPPORTED_FEATURES,

5.15-stable review patch. If anyone has any objections, please let me know.

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

From: Dan Carpenter <[email protected]>

commit f7d306b47a24367302bd4fe846854e07752ffcd9 upstream.

The usb_get_descriptor() function does DMA so we're not allowed to use a stack buffer for that. Doing DMA to the stack is not portable all architectures. Move the "new_device_descriptor" from being stored on the stack and allocate it with kmalloc() instead.

---
 sound/usb/quirks.c |   31 +++++++++++++++++++++----------
 1 file changed, 21 insertions(+), 10 deletions(-)

@@ -591,7 +591,7 @@ int snd_usb_create_quirk(struct snd_usb_
 static int snd_usb_extigy_boot_quirk(struct usb_device *dev, struct usb_interface *intf)
 {
 	struct usb_host_config *config = dev->actconfig;
-	struct usb_device_descriptor new_device_descriptor;
+	struct usb_device_descriptor *new_device_descriptor = NULL;
 	int err;
 
 	if (le16_to_cpu(get_cfg_desc(config)->wTotalLength) == EXTIGY_FIRMWARE_SIZE_OLD ||

@@ -602,15 +602,20 @@ static int snd_usb_extigy_boot_quirk(str
 				      0x10, 0x43, 0x0001, 0x000a, NULL, 0);
 		if (err < 0)
 			dev_dbg(&dev->dev, "error sending boot message: %d\n", err);
+
+		new_device_descriptor = kmalloc(sizeof(*new_device_descriptor), GFP_KERNEL);
+		if (!new_device_descriptor)
+			return -ENOMEM;
 		err = usb_get_descriptor(dev, USB_DT_DEVICE, 0,
-				&new_device_descriptor, sizeof(new_device_descriptor));
+				new_device_descriptor, sizeof(*new_device_descriptor));
 		if (err < 0)
 			dev_dbg(&dev->dev, "error usb_get_descriptor: %d\n", err);
-		if (new_device_descriptor.bNumConfigurations > dev->descriptor.bNumConfigurations)
+		if (new_device_descriptor->bNumConfigurations > dev->descriptor.bNumConfigurations)
 			dev_dbg(&dev->dev, "error too large bNumConfigurations: %d\n",
-				new_device_descriptor.bNumConfigurations);
+				new_device_descriptor->bNumConfigurations);
 		else
-			memcpy(&dev->descriptor, &new_device_descriptor, sizeof(dev->descriptor));
+			memcpy(&dev->descriptor, new_device_descriptor, sizeof(dev->descriptor));
+		kfree(new_device_descriptor);
 		err = usb_reset_configuration(dev);
 		if (err < 0)
 			dev_dbg(&dev->dev, "error usb_reset_configuration: %d\n", err);

@@ -942,7 +947,7 @@ static void mbox2_setup_48_24_magic(stru
 static int snd_usb_mbox2_boot_quirk(struct usb_device *dev)
 {
 	struct usb_host_config *config = dev->actconfig;
-	struct usb_device_descriptor new_device_descriptor;
+	struct usb_device_descriptor *new_device_descriptor = NULL;
 	int err;
 	u8 bootresponse[0x12];
 	int fwsize;

@@ -977,15 +982,21 @@ static int snd_usb_mbox2_boot_quirk(stru
 
 	dev_dbg(&dev->dev, "device initialised!\n");
 
+	new_device_descriptor = kmalloc(sizeof(*new_device_descriptor), GFP_KERNEL);
+	if (!new_device_descriptor)
+		return -ENOMEM;
+
 	err = usb_get_descriptor(dev, USB_DT_DEVICE, 0,
-		&new_device_descriptor, sizeof(new_device_descriptor));
+		new_device_descriptor, sizeof(*new_device_descriptor));
 	if (err < 0)
 		dev_dbg(&dev->dev, "error usb_get_descriptor: %d\n", err);
-	if (new_device_descriptor.bNumConfigurations > dev->descriptor.bNumConfigurations)
+	if (new_device_descriptor->bNumConfigurations > dev->descriptor.bNumConfigurations)
 		dev_dbg(&dev->dev, "error too large bNumConfigurations: %d\n",
-			new_device_descriptor.bNumConfigurations);
+			new_device_descriptor->bNumConfigurations);
 	else
-		memcpy(&dev->descriptor, &new_device_descriptor, sizeof(dev->descriptor));
+		memcpy(&dev->descriptor, new_device_descriptor, sizeof(dev->descriptor));
+
+	kfree(new_device_descriptor);
 
 	err = usb_reset_configuration(dev);
 	if (err < 0)

On ARCH_BRCMSTB using 32-bit and 64-bit ARM kernels, build tested on BMIPS_GENERIC:

Florian

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

thanks,
-- Shuah

Built and booted successfully on RISC-V RV64 (HiFive Unmatched).

All tests passing for Tegra ...

Test results for stable-v5.15:

    10 builds:	10 pass, 0 fail
    26 boots:	26 pass, 0 fail
    101 tests:	101 pass, 0 fail

Linux version: 5.15.175-rc1-g11de5dde6ebe
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

Jon

The all i386 builds failed with the gcc-12 and clang-19 toolchain builds on following branches,
- linux-6.12.y
- linux-6.6.y
- linux-6.1.y
- linux-5.15.y
- linux-5.10.y

* i386, build
- clang-19-defconfig
- gcc-13-defconfig

Build log:
-------------
i686-linux-gnu-ld: arch/x86/kernel/static_call.o: in function `__static_call_update_early':
static_call.c:(.noinstr.text+0x15): undefined reference to `static_call_initialized'

The recent commit on this file is,
x86/static-call: provide a way to do very early static-call updates commit 0ef8047b737d7480a5d4c46d956e97c190f13050 upstream.

js wrote:
Yes, the fix is at (via one hop):
- https://lore.kernel.org/all/[email protected]/

## Build
* kernel: 5.15.175-rc1
* git: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
* git commit: 11de5dde6ebe6f214c662453d6b8b6c3fd349590
* git describe: v5.15.174-52-g11de5dde6ebe
* test details:
https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.15.y/build/v5.15.174-52-g11de5dde6ebe

## Test Regressions (compared to v5.15.173-566-g4b281055ccfb)
* i386, build
- clang-19-allnoconfig
- clang-19-defconfig
- clang-19-lkftconfig
- clang-19-lkftconfig-no-kselftest-frag
- clang-19-tinyconfig
- clang-nightly-lkftconfig-kselftest
- gcc-12-allnoconfig
- gcc-12-defconfig
- gcc-12-lkftconfig
- gcc-12-lkftconfig-debug
- gcc-12-lkftconfig-kselftest
- gcc-12-lkftconfig-kunit
- gcc-12-lkftconfig-libgpiod
- gcc-12-lkftconfig-no-kselftest-frag
- gcc-12-lkftconfig-perf
- gcc-12-lkftconfig-rcutorture
- gcc-12-tinyconfig
- gcc-8-allnoconfig
- gcc-8-i386_defconfig
- gcc-8-tinyconfig

## Metric Regressions (compared to v5.15.173-566-g4b281055ccfb)

## Test Fixes (compared to v5.15.173-566-g4b281055ccfb)

## Metric Fixes (compared to v5.15.173-566-g4b281055ccfb)

## Test result summary
total: 57093, pass: 41250, fail: 3116, skip: 12683, xfail: 44

## Build Summary
* arc: 5 total, 5 passed, 0 failed
* arm: 101 total, 101 passed, 0 failed
* arm64: 28 total, 28 passed, 0 failed
* i386: 22 total, 0 passed, 22 failed
* mips: 22 total, 22 passed, 0 failed
* parisc: 3 total, 3 passed, 0 failed
* powerpc: 22 total, 22 passed, 0 failed
* riscv: 8 total, 8 passed, 0 failed
* s390: 9 total, 9 passed, 0 failed
* sh: 10 total, 10 passed, 0 failed
* sparc: 6 total, 6 passed, 0 failed
* x86_64: 24 total, 24 passed, 0 failed

## Test suites summary
* boot
* kselftest-arm64
* kselftest-breakpoints
* kselftest-capabilities
* kselftest-cgroup
* kselftest-clone3
* kselftest-core
* kselftest-cpu-hotplug
* kselftest-cpufreq
* kselftest-efivarfs
* kselftest-exec
* kselftest-filesystems
* kselftest-filesystems-binderfs
* kselftest-filesystems-epoll
* kselftest-firmware
* kselftest-fpu
* kselftest-ftrace
* kselftest-futex
* kselftest-gpio
* kselftest-intel_pstate
* kselftest-ipc
* kselftest-kcmp
* kselftest-livepatch
* kselftest-membarrier
* kselftest-memfd
* kselftest-mincore
* kselftest-mqueue
* kselftest-net
* kselftest-net-mptcp
* kselftest-openat2
* kselftest-ptrace
* kselftest-rseq
* kselftest-rtc
* kselftest-seccomp
* kselftest-sigaltstack
* kselftest-size
* kselftest-tc-testing
* kselftest-timers
* kselftest-tmpfs
* kselftest-tpm2
* kselftest-user_events
* kselftest-vDSO
* kselftest-x86
* kunit
* kvm-unit-tests
* libgpiod
* libhugetlbfs
* log-parser-boot
* log-parser-build-clang
* log-parser-build-gcc
* log-parser-test
* ltp-commands
* ltp-containers
* ltp-controllers
* ltp-crypto
* ltp-cve
* ltp-dio
* ltp-fcntl-locktests
* ltp-filecaps
* ltp-fs
* ltp-fs_bind
* ltp-fs_perms_simple
* ltp-hugetlb
* ltp-ipc
* ltp-math
* ltp-mm
* ltp-nptl
* ltp-pty
* ltp-sched
* ltp-smoke
* ltp-syscalls
* ltp-tracing
* perf
* rcutorture

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